Guliev - Winning Chess Manoeuvres (2015)

Winning Chess Manoeuvres
Sarhan Guliev
Strategic Id eas thatMasters NeverFail toFind
New InChess 2015

© 2 0 1 5 New In Chess
Published by New In Chess, Alkmaar, T he Netherlands

www.newinchess.com
All rights reserved. No part of this book may be reproduced,

stored in a retrieval system or transmitted in any form or by
any means, electronic, mechanical, photocopying, recording
or otherwise, without the prior written permission from the
publisher.
Cover design: Volken Beck

Translation: Steve Giddins
Supervisor: Peter Boel
Proofreading: Rene Olthof
Production: Anton Schermer
Have you found any errors in this book?

Please send your remarks to editors@newinchess.com . We
will collect all relevant corrections on the Errata page of our
website www.newinchess.com and implement them in a
possible next edition.
ISBN: 97 8-9 0- 5 69 1 - 5 6 8- 1
Contents
Explanation of Symbols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Foreword . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Chapter 1 The Janowski Incident or Griefout ofWit ....................18
Chapter 2 Non-Routine Exchanges .................................26
Chapter3 Connecting the Endgame with the Opening ..................3 5
Chapter4 The Battle ofthe Major Pieces .............................43
Chapter5 Dances with the Knights .................................63
Chapter6 The Shuttle Manoeuvre ..................................75
Chapter7 Positional Sacrifices (Part One) ............................81
Chapter8 Positional Sacrifices (Part Two).............................90
Chapter9 Standing on the Shoulders ofthe Classics....................109
Chapter10 Fancy some Solving? ...................................113
Chapter 11 The Pawn Prod .......................................123
Chapter12 Doubled Pawns are Cool! ................................128
Chapter13 Alekhine's Nail .......................................13 7
Chapter 14 Attack and Counterattack with a Closed Centre ...............146
Chapter15 g2-g4, and White wins .................................154
Chapter16 h2-h4, and White wins .................................159
Chapter 17 Overcoming the Blockade ...............................166
Chapter18 With the Mirror's Help .................................175
Chapter19 Fischer's Hedgehog and Morphy's Needles .................. 181
Chapter 20 March ofthe Tank Columns ..............................185
Chapter 21 The Dialectics ofWeakness............................... 196
Chapter 22 Structure and Plan .....................................204
Chapter23 Torture to any Taste ....................................212
Chapter 24 Indirect Borrowings. . ..................................221
Afterword From Ecclesiastes to Eliskases .............................228
Biography .............................................. 232

Bibliography ............................................ 232
Index of Games .......................................... 234
5
Winning Chess Manouevres
Explanation of Symbols
The chessboard
with its coordinates:
8 ± White stands slightly better

7 =F Black stands slightly better
6 ± White stands better
5 + Black stands better
4
+- White has a decisive advantage
-+ Black has a decisive advantage
3
= balanced position
good move
!! excellent move
a b c d e f g h ? bad move
?? blunder
!? interesting move
D White to move ?! dubious move
• Black to move 00 unclear
'i.t> King # mate
'iV Queen
l:t Rook
� Bishop
tLl Knight
6
Foreword
The reader could get the wrong impression about the aims of this book. He might
think that the author is trying to show that people play chess wrongly, and that
they are poorly informed about the game's classical heritage or inaccurately point
out the original sources of ideas, etc.
But that is not the case. The chess elite, with a few rare exceptions, consists of
players with a high cultural erudition in the game. It could not be otherwise. Even
if a self-taught player, by dint of sheer natural talent, achieves success in the early
phase of his career, there will come a time when he will need to 'pore over his text
books' to make further progress.
Contemporary players know a lot, read a lot, and work a lot on the game. At the
board, they are not fumbling in the dark. Their borrowings, in the great majority
of cases, are conscious, not accidental. And the fact that they cannot always state
exactly who the original source of the idea was, well, so what - nobody can know
everything.
And there is no need.
The great Indian talent Viswanathan

Viswanatha n Anand Anand does so faultlessly:
Evgeny Bareev 29 . .b:h5! gxh5 30. 'it'xh5 �eB
Linares 1 993 Possibly Black calculated as far as this
position and only considered 3 l. . l:f.f3 .
• Then 3 1 .. .fS , and the attack can be
beaten off.
But there followed:
31 .:Z.f6!
With this blockading sacrifice of the
rook, White fixes the f7-pawn in place
and the black king faces one deadly
threat after another.
Not 3 l ...�xf6 32.exf6, whilst after
3 l .. . �d2 White wins with 32 .e6! with
The threat is e5-e6. the threat of 33.l:lg6+! fxg6 34.�h8#
Black decided on mate.
28 .. h5
. So Black resigned.
He drives the bishop from the key
square e6, but further weakens his In his book of best games, Anand writes
kingside. that he was helped in finding 3l..l::f.f6 !
One feels that the position should con by knowledge of the following classic
tain a combination. How to execute it? combination.
7
Robert Fischer Akiba Rubinstei n

Pal Benko Eugene Znosko-Borovsky
New York 1 963 St Petersburg 1 909
•
0
Black to move. He should have played Black has played the opening poorly.
1 7... Ci:Je6, and the battle would all lie The position has opened up, and the
ahead. But he played: opposition of the rook on dl and queen
1 7 ... 'it'e8 1 8 . .bd4 exd4 on d8 causes Black much disquiet.
Intending the defence 1 9.e5 fS . Another basis for a combination is the
There followed, however, diagonal a2-g8; if the pawn is removed
1 9. l:l.f6! ! from dS, then after a queen check on
The same blockading sacrifice of the b3, the bishop on b7 would be hanging.
rook, fixing the f7-pawn, and on White needs to show good and accurate
1 9 ... �xf6, the move 20.e5 decides. calculation. There are so many tempt
1 9 ... ..tg8 20.e5 h6 21 . tbe2 ing options that one's eyes glaze over.
Black resigned. Rubinstein played precisely:
1 4. tillce 4!
I will ask you a question: how impor Less good is 14. �xe4 fxe4 I S .Ci:Jxe4
tant was it for Anand to know the �e8.
source game for this combination? How 14 ...fxe4 15 . .be4! dxe4 1 6. 'it'b3+
important was it to scour books and ..th8 1 7. 'it'xb7! exf3 1 8. l:l.xd7 'it'e8
databases, to establish that Fischer was Or 1 8 .. .fxg2 1 9. .l::rxd8 gxhl�+ 20.�xhl
the first to use it? .l::rfxd8 2 l .�xe7.
The answer is obvious - it was not of 1 9. l:l.xe7
the slightest importance or necessity. He
demonstrated a high level of erudition
and excellent sporting qualities. It is one
thing to know that Fischer once played
such a combination, but quite another
to recall the details in the stressful con
ditions of a sporting contest. But the
task of identifying the original source is
a job for the historian, book author and
journalist.
8
Foreword
The forcing variation has ended. White William Steinitz

has not only regained his material, but Adolf Anderssen
also has an extra piece. Vienna 1 873
Victory is quite close. However, White
needs to preserve accuracy to the very
0
end.
1 9 ... 'it'g6+ 20. �a1
It is understandable that White wants to
move his king as far away from 'enemy
eyes' as possible. But 20.\tlcl was actu
ally more accurate, and if Black replies
20 ... l:rab8 2 l ."iVd5 l::rbd8 , then 22.h5!
'ifa6 23 . .l::f.f7! , suppressing any counter
play.
20 ... :Sb8 21 . 'it'e4 The trappy 1 3 .i1t.xe7 tbxc3? 14.'ifh4!
Now 2 l .'tWd5 llbd8 22.h5 would be has to be rejected at such a top level, of
bad, because of 22 ...'ifxh5 ! 23.l:Ixh5 course (13 ... 'ifxe7). Also unconvincing is
�xd5. 1 3 .i1t.xe4 dxe4 14.ifLxe7 'iVxe7 1 5 .tbd2 ,
21 ... 'ifxe4 in reply to which, as well as 1 5 ... tbf6,
Bad is 2 1 ...fxg2 22 .'iVxg2 'ifc2 due to Black can consider the counterattack
23.�f6! . 1 5 ... tbe5!? 1 6.tbdxe4 tbd3 1 7.bl .l::t.ad8
22. :Xe4 fxg2 23. l:g1 :Xf2 24. l:f4 1 8 ."iVxa7 f5 1 9.tbg3 f4 or 1 9.tbd2 .l::f.f6 .
Things have become more complicated The number of minor pieces on the
and the extra piece does not guaran seventh rank suggests the correct idea:
tee the win. After 24 .. Jlbxb2 25 .ld.f8+ 1 3. he4 dxe4 1 4. l:.fd1 !
.l:i.xf8 26.'�xb2 l:tf2+ 27.\tlb3 \tlg8 , Black's choice is wide and difficult.
White still faces many technical diffi Going into the endgame does not offer
culties. anything good: 14 ... tbb6 1 5 Jhd8 tbxa4
24... :c2 1 6 . l::!x a8 !lxa8 1 7.tbxa4 f6 1 8 . �xf6
Weakly played. i1t.xf6 1 9.tbd2 l::rd8 20 . .l::f.c 2 , or (which
25.b3 h6 26. �e7 :ea 27. �b1 :e2 is even stronger) 1 8 .tbh4 fxg5 1 9.tbf5
28. bc5 l:d8 29. �d4 i1t.f8 20.tbxc5 l::rc 8 2 l .b4 �f7 22 .tbd4.
The counterplay on the second rank has Anderssen decides to retain the queens,
been choked off. but overlooks his opponent's 1 7th move:
There followed: 1 4 ... bg5 1 5. tlJxg5
29 ... .:r.ca 30. l:.g4 The transformation of the position
And Black resigned, since after 30 ... l:f.cc2 by means of 1 5 Jhd7 "iVc8 1 6Jhb7
there is 3 1 . l::rxg7. (1 6.tbxg5 i1t.c6) 1 6 ... i1t.xe3 1 7.fxe3
"iVxb7 1 8 .tbxe4 'ifxb2 is certainly not
Razuvaev and Murakhveri, in their in White's interests.
monograph about Rubinstein, rightly 1 5 ... 'ifxg5 1 6. :Xd7 l:.fb8 1 7. 'it'b3!
drew attention to the similarity of the Now the similarity of ideas with the
above combination with the following, game Rubinstein-Znosko-Borovsky is
earlier example: complete.
9
1 7 ... .tc6 1 8. 'ifxf7 + 'ith8 1 9.h4 of Karpov's famous manoeuvre against Kasparov
And White won. (from their first match in Moscow 1 984):
Repeating combinational schemes,

motifs and even individual moves is
a more obvious and simple case of
copying in chess. It is more difficult to
detect the connection when two players
copy each other in using the same tech
nical device. Unlike combinations, in
such cases the copying is not of a specific
move or scheme, but the actual heart or
essence, which makes up the device.
46 ... gxh4, and here, instead of the automatic
Alexander Morozevich recapture, there followed 47.ltJg2! space was
Smbat Lputian secured for manoeuvring, sufficient to achieve a
Dagomys 2007 decisive advantage.
Of course, different time, different place, different
• people, but all the same .. . »
Yes, 'all the same'... Everything looks
different, but the essence is the same. By
refraining from the obvious recapture,
White increases his scope for attack.
We would add that in the game
Morozevich-Lputian, after
25 ... hxg3 26. ttJxg3 .l:h8 27. tl::!h 5
(it is too early for 27. .l::r.x h7 .l::txh7
2 8 . .t xh 7 .tbS )
For the moment, Black has an extra 27 ... tl::!g8 28 . .1:g1
pawn, but the doubled h-pawns will White continued a large-scale attack on
sooner or later fall. His choice is there the kingside, which was crowned with
fore logical: success.
24 ... h4
By giving back material, Black hopes to As well as the purely chess element
close the h-file and reduce the pressure of copying, there is also a psycholog
on the kingside as a whole. ical factor, which can be expressed as
However, there followed: 'standing on the shoulders of the clas
25. J:h1 ! sics'. The point is that, in considering
Alexey Kuzmin's commentary is as fol a move which he understands to be
lows: risky, a player can draw unwitting sup
«Wonderful play! Manoeuvering space is more port from the great players of the past.
important than pawns. For example, in the var He can draw confidence from knowing
iation 25.gxh4 .!::f.g4 26J�hl h6 27.lt:Jg3 lt:Jg8 that a great player once played some
28.lt:Jh5 fS! 29.exf6 .te8 etc. One is reminded thing similar.
10
Foreword
The following example provides rich Then Kasparov returns to his game and
material on this theme: the concrete move 1 6 .. J�f5: «80 years on,
the rook looks to be in an even more dangerous
Anatoly Karpov position, because there are more pieces on the board.
Garry Kasparov The real issue, of course, is not the threat of
Leningrad 1 986 ... tbf6-g4. By controlling dS and enabling the
break ... d6-d5, the rook performs useful work. It
• is not easy for the opponent's passive minor pieces
to attack it. There is also a psychological aspect
to the rook's move: it is for White to free himself
from thinking about the rook and wanting to trap
it, but his resources for doing so are few. Tarrasch,
incidentally, obsessed with trapping the rook, even
tually lost his game ... »
We have an extremely subtle psycholog
ical study before us. Consider the words
of the 13th World Champion. First, he
Kasparov writes: «I looked at 16 ... �g2 admits that the positions - his own and
17.tbxg2 .l:thS, with. the threats of ... �g7-h6 Lasker's - are not entirely good. Lasker,
and ... tbf6-g4. But then, by means of 1 8.�£3 in putting his rook on e5, and then
�xf3 1 9.exf3, followed by f3-f4 and tbg2-e3, c5, was not risking a huge amount in
White obtains a favourable position, and one in chess terms. It turned out to be a diffi
Karpov's style.» cult job to trap this rook, if not down
And further: «I find it hard to explain how, right impossible. White simply lacks the
suddenly, the urge came on me to play originally, strength and resources. Kasparov, on
bravely, freshly. In playing my next move, I sur the other hand, was sending his rook
prised myself by my courage». into the heart of a lot of enemy pieces.
1 6 l:l.f5!?
...
So why did he compare his game with

«Something similar happened i n the 1 908 Tarrasch-Lasker?
world championship match between Lasker and Kasparov himself gives an honest and
Tarrasch: exhaustive answer to this question. He
wanted to post the rook on fS (especially
because normal, solid moves left White
with stable, unpleasant pressure, and
the character of the battle would have
perfectly suited the style of his power
ful opponent). At the same time, he was
frightened to do so, as any player would
have been, if they realised the danger
and the risk they were taking in doing
so, a risk that might prove too great.
This duality - wanting to play the rook
15 .. Jie5 16.tbd4 l:f.cS!? 1 7.�b3 tbb6 etc. to f5 and at the same time, fearing it
II
might be a mistake - forces the player to direct copying is all that is needed, pro
call on something extra, outside him viding it is from a trustworthy source.
self. Because the internal strength, even
of so great a player as Kasparov, could Veseli n Topalov
at some moment prove insufficient. Krishnan Sasikiran
He found this extra something in the Sofia 2007
memory of the Tarrasch-Lasker exam 1 .d4 tt::lf6 2.c4 e6 3. tt::lc3 �b4 4.e3
ple. 'Look', he says to himself, 'Lasker b6 5.�d3 �b7 6. tt::lf3 0-0 7.0-0 d5
put his rook in a similar spot, and it 8.a3 �d6 9.cxd5 exd5 1 O.b4 tt::lbd7
worked. Go for it! '
This i s something worth thinking
about.
A large separate topic is the copying of

plans and pawn structures. As we will
see in the final chapter of this book,
each is inseparable from the other. In
forming a plan, a player will in most
cases think about changes in the pawn
structure, and vice versa: a favourable
change in the pawn structure does not It would be wrong to describe this posi
occur of its own volition, but usually tion as critical - only 10 moves have
results from the successful implementa been played on each side. But there is
tion of the correct plan. no doubt that this is the moment for
In comments on games by the winner, choosing, especially choosing a plan.
one often comes across phrases such as It would be good for White to offload
'a known position', 'this position was his unemployed dark-squared bishop
known to me', etc. Even the printed and to fix the backward c7-pawn. This
word reveals the satisfaction which means that it is worth considering (in
players get from such situations. It is a some order or other) b4-b5 , a3 -a4 and
broad subject, which contains within .i.cl -a3 . Black can ignore this threat,
itself several elements. and when White takes on d6 , can take
If a strong player finds himself in a towards the centre with the c-pawn.
familiar position, then he plays the next The pawn structure would thus change.
part of the game 'with eyes wide open'. But in whose favour?
There are no secrets for him, in what Commenting on this game, Topalov
happens on the board. He knows his plan three times made an interesting admis
and what plan to fear from the opponent. sion. Before White's IIth move was the
He knows whether to exchange a certain first time:
piece or not, which pawn structure «A well-known and playable position for both sides.
changes he should go in for, etc. I saw it quite recently and remembered a game I
It is by no means necessary that such very much liked, in which Ivanchuk beat Piket.
knowledge should come from a player's This game was played in the Amber Tournament
own work or experience. Sometimes, 1999, in which I also played.»
12
Foreword
1 1 .b5 lt::le4 1 2. i.b2 :ea 1 3.a4 :es

1 4. lt::le2 as
«Played after long thought. Black fixes the
structure, and this is the exact structure which
occurred in the Ivanchuk game, that I remem
bered!»
1 5. i.a3 :ca 1 6 . .bd6 cxd6
1 7. :c1 ttJdf6 1 8.h3 .l:[e7 1 9. 'ifb3
h6 20. :Xc8 'ifxc8 21 . .1:[c1 .l:[c7
22. :Xc7 'ifxc7
29 ...'iVxc2 30.tt:Jxc2 tt:JgS 3 1 .f3 �f8

32 .Wf2 We7 33 .tDc3 tt:Je6 34-.t2Je1 Wd8
3 S . itd1 tt:Jg8 36.itb3 tt:Je7 37.t2Jd3
Wd7 38.M gS 39.hxgS hxgS 4-0.gf f6
4-l .tDcl tt:Jg7 42 .tD1 e2 fS 43 .gxfS tt:JgxfS
44-.tt:Jg3 tt:Jg7 4-S .Wg2 tt:JgfS 46.tDxfS
tt:JxfS 47.Wf2 We6 48 .tDe4! itc8
49.tDxg5+.
The rest of the game can be passed over.
White won.
23. 'ifc2!
«Again offering the exchange. This was not done I particularly want to emphasise that the
out of necessity, and I could have played 23.'iib2 . plan carried out by Ivanchuk is nothing
The point of my idea is that, after the exchange out of the ordinary and its general out
of queens, Black's position is very difficult. I can lines were known long ago. It had been
kick the knight out of ef, transfer the bishop to played by well-known GMs and by less
b3 and eventually organise piece pressure against famous players. For example, in the
the pawn on dS, at the same time advancing the game Knaak-Beliavsky (Novi Sad 1 979) ,
kingside pawns. White preferred to attack the doubled
Something similar happened in the Ivanchuk-Piket pawns with the queens still on: 1 1 .bS
game.» tDef 1 2 .'iVb3 l:i.e8 1 3 .a4 tt:Jdf6 H.ita3
aS 1 S . itxd6 cxd6 1 6 .tDe2 h6 1 7J�a2
Let us look at this Ivanchuk-Piket game, 'iVd7 1 8 .tDe1 .l::f.e 7 1 9.f3 tt:Jgs 20.tt:Jg3
which is referred to three (!) times by g6 2 1. .l::f.e2 hS 22.M tt:Je6 23.'iVb1 tDf8
Topalov. H.Wf2 !lae8 2S.itc2 'iVd8 26.t2Jd3
1 1.. �b2 (the first ten moves were as tt:Je6 27.itb3 etc.
in Topalov-Sasikiran) 1 1 ..Jle8 1 2 .'iVc2
a6 1 3 .bS aS H.af tt:Jef 1 S . ita3 .tlc8 On the other hand, in the game Leverett
1 6 . .l:tfc1 tt:Jdf6 1 7.'iVb2 tt:Jgf 1 8 .t2Jd1 h6 Bisguier (Chicago 1 98S), White con
1 9.h3 tt:Jgf6 20.l::tc2 'iVe7 2 1 . ite2 'iVe6 vincingly adopted the 'Ivanchuk plan'...
22 . .l:tac1 'iVe7 2 3 . itxd6 cxd6 H.tt:Je1 some 1 S years before Ivanchuk himself:
I:!.xc2 2S .ld.xc2 l:!c8 26.t2Jd3 l:tcf 27.'iVc1 1 1 .bS tDef 1 2 .'iVb3 l:le8 1 3 .a4 tt:Jdf6
'iVc7 28.tDe1 .l:!xc2 29.'i:Vxc2 H. ita3 aS 1 S . itxd6 cxd6 16.l:tac1 l:tc8
13
1 7. .I:f.c2 .l:!.e7 1 8 .J::i.fcl l:tec7 1 9.t2Jdl g6 M i khail Botvin n i k

20 . .a'.xc7 .!::i.xc7 2 1. .�xc7 "ifxc7 Viacheslav Ragozin
Bolshevo 1 947
22 .�c2! �xc2 23.�xc2 �c8 24.tbel Kotov's commentary in Think Like a

t2Jd2 2S.f3 tbc4 26 .�f2 �f8 27.�e2 Grandmaster. «White's plan is simple: to
�e7 28.�b3 hS 29.e4 �b7 30. �xc4 attack the weak dark squares around the
dxc4 3 l .d5 t2Jd7 32 .�e3 tbcs 33 .t2Jc3 black king, and sacrifice the bishop on g6,
t2Jb3 34.t2Jc2 �c8 3S .t2Ja3 �d7 after which the queen and rook will carry
36.tbxc4, and the rest of the moves can out the decisive attack.»
be omitted. And Botvinnik adds: «It was not so difficult
for White to find the correct plan, as he only had
So, there was no necessity for Topalov to copy Lasker's play in his game with Capablanca
to follow Ivanchuk (or anyone else, (Moscow 1935)».
of course). But he did so, and in his 27. ttJf6
thrice-repeated commentary, he If White had been worried about the
emphasises the fact. pin on the a6-fl diagonal, he would
He saw the game Ivanchuk-Piket, with have played first 27. �e4, but he evi
his own eyes, as he was at the tourna dently considered that Black could not
ment. And Topalov says to everyone: afford the move 27 ... �a6 because of
I believe him. I don't see any need to 28.l:le3! �xb5 29. �xbS followed by
invent something new, nor do I con ld.e3-h3-h7.
sider it shameful, to follow in his foot 27 .. .ti:Je7 28. l:te3 tbg8
steps. Because this idea works. An interesting moment, which is left
Iron logic. totally vague in previous annotations.
It seems that the threat to win the
The Patriarch of Soviet chess, the sixth queen with l:Ie3 -h3-h7 takes prece
World Champion Mikhail Moiseevich dence over everything else. In reality,
Botvinnik, was a great practitioner of the battle after 28 ... .I:!.xc3 29.1::f.h 3 .l::f.xd4
planned play. Many of his discoveries 30.l::f.h 7 is only beginning. Black has
have entered chess theory. But even the resource 30 ... l:!.g4!. After 3 l .�xg4
Mikhail Moiseevich was not above �xf6 or 3 l .tbxg4 �xh7 32 .t2Jf6 �h8 ,
direct copying. Why not, if it leads to White achieves nothing. He must
success?! go into the endgame: 3 l .tbd7+ �e8
Foreword
32 . .l:Ixg7 (after 32.tDf6+ 'iYxf6! 33.'iYxf6 original plan, even if the tactical com
.l:Icl + the game ends in mate: 34.�h2 plications were objectively in his favour.
lixg2+ 35 .'�h3 �hl # or 34. J.fl .l::r.xg2+ 34 ... i.dS 35 . .1:1.eg5
35 .�hl .!::l.xfl #) 32 .. Jhg5 33.l:lxg5
�xd7 34. J.b5+ �d6 35.l:lxf7 .l:ta3 .
White has an extra exchange, but Black
has his chances, as for example in the
variation 36.�c4 J.ds 37.�xd5 tDxdS
3 8 . lixg6 .l:txa2 39.g4 b5 . Instead of
36.J.c4, White can try 36.�f8 !? with the
idea of winning a piece after 36 ... lixa2
37. .l:id8+ 3 8 . .l:td7+ and 39Jhe7.
29. tiJxg8 <it>xg8 30. 'ife7 i.c6
For the second time in a row, Black
shows neither inventiveness nor tenac The preparatory work is completed and
ity. It is obvious that White plans to there is no defence against the blow
transfer the rook from b5 to the king on g6. Black's passivity in the previous
side. So why not try to hinder this? stage of the game has helped White
After 30 ... J.d5! the rook is cut off on fulfil his task.
b5 . Grabbing pawns also doesn't lead to 35 ... .l:l.xc3 36. i.xg6! .l:l.xg3 37. i.xf7+
anything good: 3 1 ."i¥xa7 lixc3 32.'iYxb6 �8
licl + 33 .�h2 "i¥h6+ 34.lih3 lihl+! Ragozin follows the path of least resist
35 .�xhl (35.�g3 'iVgS#) 35 ...'iYxh3+ ance, to the very end. After 37 ... �xf7
36.�gl 'ilxg2#. Stronger is 3 3 . �fl . but 38.'ilxg3 J.xa2 White would still face
here too, after 33 ... .l::f.dc8 Black has fully technical problems, but now it all ends
adequate counterplay. immediately:
31 . .1:1.be5 38 . .1:1.xg7 .l:l.xg7 39. i.xe6+
The rook passes the potential barrier Black resigned, and we will now look at
on d5 , but does not go to g5 at once, the source game.
because then, after 3 l ...J::[d 7, the queen
would be forced to leave the kingside. Emanuel Lasker
31 ... .1:1.d7 32.'it'h4 'it'h8 Jose Raul Capablanca
Here too, 32 ... J.d5 deserved consider Moscow 1 935
ation.
33. 'iff4 'it'g7 34 . .1:1.g3
0
White has finally gathered all of his
major pieces on the kingside. The
similarity with Lasker-Capablanca (see
below) is now total.
Incidentally, the combination could
have been played here: 34. J.xg6! fxg6
35 . .l:ixe6, and if 35 .. J:tf8 , then 36.'iVg4,
whilst if 36 ... �f7, then 36.'iVd6. But
Botvinnik did not like to change his
15
In general terms, we already know the the game. Capablanca, for his own rea
plan. White brings his major pieces to sons, resigned before the adjournment.
the kingside, masks his real target with
threats on the dark squares, and at the To complete the discussion of copying
appropriate moment, sacrifices his bish plans and structures, I will remind the
op on g6. reader of the 'mirror effect', explained
24. 'iVc1 !? :Sd8 25. l:[1 e3 .tea in Chapter 1 8 . Many structures (and
More tenacious is 25 .. .f6, although after concrete opening variations), which
26.l:ta5 a6 2 7.hxg6 hxg6 28 . .l:f.g3 White's seem completely original and inde
pressure is more than noticeable. pendent, are in reality nothing but a
26. l:[h3 ®18 mirror image of another opening or
Nothing essential is changed by structure, with colours reversed.
26 ... .l::f.h8 2 7.'iVf4.
27. 'iVh6+ l:[g7 28.hxg6 hxg6
29.hg6!
The end. The sacrifice is illusory, inci
dentally: after 29 .. .fxg6 30.'iVh8+ 'iit> f7
3 l. . l:f.f3+ Black loses at once.
Unlike Ragozin, Capablanca finds a way
to prolong the game. The leitmotiv of the
defence is the idea of sacrificing the
queen, to establish a fortress.
29 ... 'iVf6 30. l:[g5! <J;e7
The trappy 30 .. J:td5 (hoping for 3 I .�f3 A match game from 1 895 between
'iYxf3 ! 32 .gxf3 �xg5+ 33 .'iVxg5 �xg6) Chigorin and Schiffers. The first moves
is refuted by 31..l:txd5! cxd5 32 .'iVh8+ (I .e4 c5 2 .tt:::lf3) show that White was
cJile7 33 .'iYxc8 �xg6 34.'iVc7+ (this pre ready to play the Closed Sicilian. In real
liminary check is necessary, so as not to ity, though, he is playing a King's Indian
have to calculate complicated variations Attack (or King's Indian Defence with
after 34.1:f.h8 .l:txg2+ 35 .�xg2 'iVg5+) colours reversed). If you are not sure of
34 ...cJitf8 3 5 .'iVb8+ cJile7 36.1:f.h8 .l:f.xg2+ this, look at the following diagram:
37.cJitxg2 'iVg5+ 38.'iVg3 .
31 .l:[f3 'iVxf3 32.gxf3 l:[dg8
Now the resolute 33 .'iVh4! leads to the
desired outcome. In the game, there
followed:
33. ®11 :Xg6 34. :Xg6 :Xg6
And Black was indeed very close to
establishing a fortress.
The further dramatic events are well
known, although treated differently
in various books. Lasker, already very The position is almost identical, with
elderly at the time, did not undertake reversed colours (Taimanov-Fischer,
anything active and waited to adjourn Vancouver 1 9 7 1).
16
Foreword
The following pair of twins look like Alexander Alekh ine

some sort of joke. Different games, Frederick Yates
played with completely different open Hamburg 1 91 0
ings, ended up reaching an identical
position, and of a middlegame charac •
ter at that.
Fedor Duz-Khoti m i rsky

Aki ba Rubi nstein
Lodz 1 907
D
Only at this point did the games
finally diverge. Duz-Khotimirsky hap
pily played on a pawn down (but lost
anyway), while Yates ran headlong into
the prepared trap:
1 9 .hg5 20. t006 !
...
I sincerely hope you will enjoy this book!

Sarhan Guliev
Baku, April 20 1 5
17
Chapter 1
The Janowski I ncident, or

G rief out of Wit
Mark Ta imanov I would like to believe that the young

Robert Fischer Fischer was familiar with this position:
Buenos Aires 1 960
Jose Raul Capablanca
• David Janowski
New York 1 9 1 6
Despite his young age (he was still

not yet 1 7) , the American player de
fended a difficult situation with great
cold-bloodedness: A historic incident - in the diagram po
81 ... �4 82.b5 <ifi>e4 83 . .td4 .tc7 sition, Black resigned! But the position
84. <iti>c5 �d3! is in fact a draw.
Taking the king round the back is the 83 ... ®t4! 84 . .td4 �3! 85.b5 <ifi>e2!
key to the position. This and the following moves are strict
85. �c6 �c4! 86 . .tb6 .tt4 87 . .ta7 ly the only moves.
.tc7! 86. �c6 <iti>d3 87 . .tb6 .tgS
And the players agreed a draw. 88.�b7!?
Or 8 8 . �c7 �e3 89. �d6 �c4, and the
Fischer's colossal talent is not in doubt, king manages to stop the bishop reach
but the question remains: where did he ing cS.
learn this method of defence? Maybe 88... �c4 89. <ifi>a6 �b3 90. .tt2 .td8
he did not know it, and found it at the 91 . .te1 <ifi>a4! 1 -0
board? Or maybe someone had shown Once again, the king arrives in
it to him? time.
18
Chapter 1 - The Janowski Incident, or Grief out ofWit
Ruslan Ponomariov 90. i.12

Levon Aronian The direct method also works: 90 . .ic7
Lausanne 2001 .ixc7 9 1 .a7.
90 ... <;ita5 91 . �h4 <;itbS 92. �e1 !
Zugzwang. Black resigned. But he could
D
have drawn with the primitive 86 ...
g3 87.�b5 g2 88 .'1t>c6 .ig3 89.a7 .if2
90.a8'it¥ gl 'it¥.
A case of grief from wit. The only ex
cuse one can make for Ljubojevic is that
it was a blindfold game.
Rook endings are the most complicated

in chess. The great number of basic
In our day, GMs use this technique positions, the large number of rules
of by-passing with the king without which work in some positions, yet not
thinking. in others that seem very similar, the
67. <;ite8! �16 68. �b4+ <;itdS 69. <;itd7 huge number of rules, and even greater
<;itc4 70. �18 number of exceptions ...
Draw agreed. If 70 ... .i.d4, in the hope of Nobody can ignore the most important
blocking the a3 -f8 diagonal with .id4- principle of studying endgames: mira
c5 , then 7 1 .'1t>c6! . cles do not happen - one must work.
Work a great deal. Study theory, read
Anatoly Karpov good, intelligent books, study the clas
Lj ubomir Lju bojevic sics, carefully dissect the commentaries
Monaco 1 996 of the strongest masters and analysts,
and, naturally, enlarge your experience
• of playing such positions .
Arnaud Hauchard
Ruslan Ponomariov
Belfort 1 998
Chess is an area where there are as

many exceptions as there are rules. In
this example, Black confidently follows
the classical method.
86 ... <;ite4 87. <;itc5 <;itd3 88. <;itc6 <;itc4
89. �b7 <;itbS
And he lost ...
19
Black to play. If you do not know the Winning.

classical stem game, then you could To say such systematic manoeuvres
easily lose your way. But if you know it, occur often in practical play would be
then the win is not so difficult: untrue, as they are relatively rare. But
69 ... �g1 70. l:.g8+ �h1 ! 71 . l:.f8 even so, not knowing such ideas would
l:.b3+! 72. �h4 �g2 73. l:.g8+ �h2 be shameful.
74.l:.f8 l:.b4+! 75. �xh5 'iii>g 3
Also possible is 75 ... �g2 with the same Vlasti mil Jansa
ideas. Efi m Geller
76. l:.g8+ 'iii> h 3 77. l:.f8 l:.b5+! 78. 'iii>h 6 Budapest 1 970
�g3 79.b8'it'+
White does not wish to see the very
D
end of the film. The main variation is
as follows: 79 . .l:tg8+ �h4 80.:lf8 .&!b6+
8 1 .�h7 l:Ixb7+ 82 .�h6 �g3 winning.
On 79. l:!.f7, there follows 79 .. Jhb7.
79 ... :Xb8 80. :Xb8 f1 'it'
And Black won.
The method employed by Ponomariov

is known in chess composition termi
nology as a 'systematic manoeuvre'. 67.f7
Interestingly, this particular example After 67Jhc2 .l::I.b 8+ 68 .�f7 l:!.b7+ we
was first shown in an endgame study reach a theoretical draw.
(in 1 890) by a practical player, no less a 67 ... l:.g4+ 68. �h8!
player than the future World Champion Again the correct choice. After 68.�f8
Emanuel Lasker. �g2 69.�e7 .l:i:e2+ White is already
unable to return to the correct path.
68 ... l:.f4 69. l:.c6+! �h5 70. 'iii>g 7
D
l:.g4+ 71 . �h7 l:.f4 72. l:.c5+ �h4
73. �g7 l:.g4+ 74. �6
Blindly following the original example
is wrong: 74.�h6 Iif4, and White has
achieved nothing.
74 ... l:.f4+ 75. �e6!
All correct. Heading towards the enemy
king immediately with 75 .�g6 Ilg4+
76.�h6 .l::I.f4 is pointless. First, he needs
1 . 'iii>b8 l:.b2+ 2. 'iii>a8! l:.c2 3. l:.h6+ to make the black rook occupy the
'iii>a5 4. 'iii>b7 l:.b2+ 5. 'iii>a7 l:.c2 second rank.
6. l:.h5+ �a4 7. �b7 l:.b2+ 8. �a6 Incidentally, this could have been
l:.c2 9. l:.h4+ �a3 1 0. c.fi>b6 l:.b2+ achieved earlier: 7 l .�f8 l::i.g2 72 .�e7
1 1 . 'iii>a5 l:.c2 1 2. l:.h3+ �a2 l:le2+ 73 .�f6 .l::I.f2+ 74.�g7 .l:tg2+
1 3.:Xh2! :Xh2 1 4.c8'it' 75 .�h7! .l::I.f2 76.l:Ic5+, etc.
20
Chapter I- The Janowski Incident, or Grief out ofWit
75 ... .:e4+ 76. <itf5 .:e2 77. <iitg 6 adjournment of games has passed into
.:g2+ 78. <iith 6! .:t2 79 . .:c4+ <iith 3 history, never to return. Classical (that is,
80. <iitg 6 .:g2+ 81 . <iith 5 .:t2 82 . .:c3+ unhurried) time controls are becoming
<iith 2 a rarer and rarer thing. By the time a
The finish resembles that of the original position arises on the board, with only a
study. The first two moves are a pre few pieces left, the modern-day player has
amble, and then, after 69.l::tc 6+! , there neither the time nor the strength to play it
follows the working out of the main well. The player is morally and physically
theme. Now the final move is the point: exhausted, and has just seconds on the
83 . .:Xc2! clock; it is hardly surprising that instead
Black resigned. «It is good when you know the of quality play, we start to get a sham
classics!» Jansa.
- bles, and the press end up talking about a
Incidentally, we can note the culture of comedy of errors.
the Soviet GM. When faced with enter It is a rare case when we can present
ing an ending with rook against queen, an endgame without pointing out mis
he resigns at once. Nowadays, players takes, even with the strongest players.
play on in such an endgame, and not Mistakes are the rule. But there are
always without success. exceptions. The players of today operate
in completely different conditions from
In former times, players studied the those of the masters of the past. And
endgame all the time, for obvious rea rather than criticising the current tribe
sons. If it reached an ending, a game of players, it is better to concentrate on
would, as a rule, be adjourned. The play the ways in which they have enriched
ers would study reference books, and chess art.
discuss the position with friends and
colleagues. The analysis of adjourned Vesel in Topalov
positions could take a day, two, or even Alexander Beliavsky
more. And on the day of resumption, Linares 1 995
after relaxing, the player would appear.
His analysis would be tested against
D
that of his opponent, and the work of
his brigade of seconds against that of his
opponent's. Whose analysis would turn
out to be deeper and more accurate?
But the game eventually ends, whereas
the life of the endgame goes on. Those
who were analysing the position again
sit down at the board, this time to
write articles about it. Everyone took an
interest in such endgames, which were Having quickly calculated the variation
discussed just as much as openings. 73.a7 �c3 74.�b7 �b4 7S .b6 �bS ,
The culture of endgames was at its one could conclude that the diagram
highest. But the march of time goes on position is a draw. But in the game, only
and the conditions for chess change. The one more move was played:
21
73. �a7! pawns against a minor piece. Here is

And Black resigned. a classic example from the Bulgarian
GM's career.
It is appropriate to recall the classical
game Maroczy-Tarrasch (San Sebastian Veseli n Topalov
1 9 1 1), in which the Hungarian master Krishnan Sasikiran
(acknowledged as a great expert on the Sofia 2007
endgame, incidentally) was unable to
win this position: •
To win the Sofia super-tournament,

Topalov needed to win this last-round
The game ended ... game. Veselin had played energetically
44. �c6 J:.c1 + 45. �b6 J:.c4! and enterprisingly, sacrificing a piece
Also sufficient for a draw was 45 ... hl'ti in the middlegame, and at the moment
46 . .l::rx hl �xh l , but the text move is shown in the diagram, had almost
more decisive and does not give the op achieved his aim.
ponent a choice. Black to move. And he has no moves.
46. :Xh2 The bishop is immobile and the knight
The threat was 46 ... .l:f.h4, and Black even also (because of mate in one). Nor can
wins. he play 40 ...'tid7 because of 4l .'iVf8+
46 ... :Xb4+ 47. 'iPc5 J:.a4 'tif7 42 .g5+. That only leaves:
Draw. 40 ... 'iff7 41 . 'ifd8+ 'ife7
Yet only two moves were needed And, not for the first time in the game,
to win: 44. l:f.xh2! �xh2 45 .Wa6 ! ! . Topalov faced a choice ...
Maroczy probably calculated something Which piece should he keep and which
like 45 .a6 Wg3 46.Wb6 �f4 47.b5 We5 exchange? It looks as though he should
48.a7 Wd5 49.�b7 Wcs so.b6 Wbs . just eliminate queens and knights.
The non-routine king move, i n front of Three pawns, which advance like a
his passed pawn, escaped his attention. bulldozer, and the bishop is out of play.
But not Topalov's! But it turns out that this is not so simple
- the bishop has the dangerous counter
The endgame where a king and passed blow ... �b7-c8-d7 and then taking on
pawns fight against a rook is a frequent b5!
guest in tournament play. Less common There is cause for worry. Not for
is the ending with connected passed nothing were the experts, watching
22
online, predicting that White would 55.h7+) 53.axb5 a4 54.f6 �g6 55 .f7!
decide to keep as many pieces on the �xf7 56.�f5 a3 57.g6+ �e7 58.h7 a2
board as possible, especially as they 59.h8'iY al 'iY 60.'iYh4+ c;tJe8 6I .c;tJe6
could see nothing wrong with the move or 60 ... �d7 6I .g7 'iYfl + 62 .'ti'f4 'iYg2
42.'iYxb6. Nonetheless, there followed: 63 .'i¥g4.
42. 1!fg8 'ii'f7 43. 1!fd8+ 1!fe7 52.�g5!
And (to everyone's great surprise) As they say in such cases, 'accurate to
44. 1!fxe7+! ? �xe7 45. �g3 tbe6 the end'. After the tempting 52 .g5 �h7
Now everyone was again convinced 53.f6 �g6, we have a fortress.
that Topalov would play 46 .tLle2 , keep 52 ... �7
ing a piece on the board, which can Nor can he save himself by 52 ... �h7
prevent the black pawns promoting. But because of 53.�f6 �xh6 54.�e7.
he took: 53.h7 �g7 54.h8'if+ �xh8 55. <iW6
46. lbxe6!? bbS
Combined with his 44th move, this Too late!
creates a profound impression. Only a 56. �e7 �d3 57.f6 �g6 58.f7 hf7
player with great belief in himself and 59. �xf7
his calculation could play in this way. Black resigned.
46 ... �xe6 47.f4 �c8 48.f5+ <iW7
49.h5 �d7 50.h6 �g8! Yu ri Averbakh
It is too early to sacrifice: 50 ... �xb5 Semen Fu rman
5 l .axb5 a4 52 .g5 a3 53 .g6+, and the Odessa 1 960
white pawns promote first. But now it
may look as though Black has tricked
0
his opponent.
51 . �f4
The culmination.
A classic example on the same theme.

White has strengthened the position
of his pieces to the maximum, but for
the moment Black is hanging on by his
fingertips. The key to the position is the
51 ... �e8 bishop sacrifice:
Topalov's calculation turns out to be 43. �e8!
precise after all. In reply to 5 l ...�xb5 After 43 ...�b7 44 . .i.d7 we have zug
White wins with 52.g5 ! (of course, zwang, so accepting the sacrifice is
not 52.axb5 a4), and then: 52 ... �h7 essential.
(52 ... �xa4 53 .g6 �e8 54.�g5 a4 43 ... �xeS 44. �xg7
23
It looks as though the pawns will 51 . c,t;>t5 tttg a 52.g7 .it7 53. 'tte5
advance of their own accord, but this .ig6 54. 'ttd 6 .id3 55. 'ttxc6 c,i;>t7
is not the case at all. For example, after 56. '1td6
44 ... �e7 the only winning line is 45 .h4! There are too many pawns, and too
�d7 46.g5! whereas after the tempting many passed pawns, too far apart.
45S!txh6 White cannot make progress: The king and bishop cannot cope. On
45 ... �f7 46.h4 �d7 47.h5 �c8 etc. 56 ... �b5 White wins by 5 7.g8'i¥+
44 ... h5 45. c,i;>xf6 �xg8 58 .�e7, whilst after the game
Slightly lengthening his task. Simpler move 56 ... .if5 the reply 57.b5! decides.
is 45 .gxh5 �xf5 46.�xf6 �bl 47.�g7
or 46 ... �xh3 47.h6 �f8 48 .h7. And if Returning to Topalov's play, we can
45 ... �e7, then 46.�g6. note that relying on long and accurate
45 ...hxg4 46.hxg4 c,t;>ta 47.g5 'ttg 8 pieces of calculation is a character
48.g6 c,t;>ta istic of his style. The Bulgarian plays
the ending the same way he plays the
middlegame. Playing 'on general con
siderations' is foreign to him. He does
not like to manoeuvre around, dancing
from one leg to the other, but prefers to
force matters.
Sometimes such an approach leads him
astray, but often it enables him to find
the shortest way to win.
After his inaccuracy at move 45 , Vesel i n Topalov

White has significantly complicated Antonio Antu nes
his task. There is no win after 49.g7+ Candas 1 99 2
�g8 50.�g6 �d7 5 I .f6 �e6! (but
not 5 l ...�e8+ 52 .'litf5 �f7 53 .�e5),
0
whilst after 49.�e5 �g7 50.�d6 �xf5
5 1 .�xc6 �xg6! 52.b5 �e8+ he even
risks losing.
49. c,l;>g5!
One can compare this with the move
52.�g5 in Topalov-Sasikiran, don't you
think?
49 ... .id7 50.t6 .tea
Or 50 ... �g8 5 1 .�f4 �e8 , when even
getting the opposition after 52 .f7+ The diagram posltlon is like an old
�xf7 53 .gxf7+ �xf7 54.'it>f5 �e7 photo. Topalov was just 1 7 at the time,
55 .�e5 �d7 56.�f6 'it>d6 does not but his style was already fully formed.
guarantee the win. But it is easy to avoid The last move was 60 ... tLlg7-f5 . White
this line of defence: 52 .�f5 �d7+ has a choice. It is possible to retreat
53 .�e5 �e8 , and only now 54.f7+. the knight, when the endgame has a
24
drawish character, but Black can torture after both 65 .�e4 �c4 66.f5 d3 67.f6
his opponent for a hundred moves. Or d2 68.f7 dl'iY 69.f8'iV 'iYel +, 70 ...'iVfl+
one can take on fS . If this is drawn, it and 7 l ...'iVxf8 , and also 65 .�e6 d3
is drawn straightaway, but if White's 66.f5 d2 67.f6 dl 'iY 68 .f7 'ii!Vd 6+.
calculations contain the slightest error, 65 ... d3 66.f5 d2 67.f6 d1 t!f 68.f7
then it will be impossible to do any Strangely, in this position, the weaker
thing about it. That is the way with side's king is better on the g-fi.le than
pawn endings - one needs to calculate the e-fi.le. For example, with the white
them to the end, because there is no king on e7, Black would win by means
way back. of 68 ...'ifd6+ 69.<iite 8 �c6! 7 1 .f8'iV
Topalov accepted the challenge: 'iVd7#. One can find many examples
61 . tlJxf5+! exf5 62. <itd4 <itc6 on this theme in endgame studies with
62 ... �e6 63 .�c5 . pawn endings.
63. <ite5 <itc5 64. <itxf5 d4 65. '1tg6! 68 ... 'iVd8 69. <itg7 'iVg5+ 70. <ith8
This move had to be foreseen, before 'iVf6+ 71 . <itg8 t!fg6+ 72. <ith8! 'ifxf7
White exchanged knights. He loses Stalemate.
25
Chapter 2
Non-Routine Exchanges
Chess is a concrete game.
It is impossible to learn to play well, by relying solely on general principles and
considerations. Nothing can replace the basic fundamental that chess is a game of
calculation.
One can say 'a blockading knight is stronger than a bishop' , but this is untrue,
because it is too general and so inaccurate.
Or one can say that 'a blockading knight is usually stronger than a bishop' , or
'a knight occupying an outpost is usually stronger than a bishop' . That is closer to
the truth.
Or one could formulate it like this : usually, a blockading knight, or a knight oc
cupying an outpost, is definitely stronger than a bishop. Exchanging such a knight
for a bishop would be a positional mistake. But sometimes the position contains
other, more important factors. And, after all, the player's task is not to hang onto
the knight on its beautiful outpost, come what may, but simply to win the game.
For the sake of this, one can sacrifice whole pieces, let alone exchange them off.
Every rule has its exceptions, and one must adopt a sceptical attitude to them all.
Nothing should be accepted on faith, everything must be doubted - that is how
great players have always behaved. And rightly so.
If White had played 22 .a4 (so as to prevent

Robert Fischer 22. .. .ib5), Black would have replied 22 ... .ic6,
Tigran Petrosian preparing .. .ti:Jf6-d7».Polugaevsky.
Buenos Aires 1 971 Polugaevsky wrote his comments
immediately after the gl ame and all of
the experts who have commented on
0
the game since have been in agreement
with Lev Abramovich. That is, they
have said it is sad to part with the lovely
knight on c5, but it is necessary. The
arguments are all the same.
But it is not true. At least, not quite true.
In the variation 22.a4 .ic6 2 3 J le2 ti:Jd7,
White has a combination: 24.tbxa6!
l:txa6 25 . .ixa6 Iha6, and now he
22. tiJxd7+!? plays the quiet move 26 . .l:tcl ! ! with the
«Highly characteristic of Fischer. He happily parts unstoppable threat of b4-b5 (26 ... .ixa4
with his good knight, exchanging it for the bad 27.l:i.c8#). And if 23 ... g6 (stopping
bishop, and reaches a position of another type. mate) 24 . .l::tc l ti:Jd7, then 25.a5, and it
26
Chapter 2 - Non-Routine Exchanges
is hard to see what Black's next move 26. 'iiff2 h5

will be. On 25 ... '1t>g7. for example, there Now 26 ... a5 is bad because of 27. 1tbs .
is the decisive 26/bxa6 z:rxa6 27.l:r.b2 ! Botvinnik, in a bid to ease the defence,
with good winning chances for White. recommended 26 ... .!:i.b8 27.a3 aS 28 .b5
So this means that the capture on d7 a4. But it is not entirely clear what he
was not necessary. The reason for play had in mind after 28.bxa5 . If 28 ... l:r.b3 ,
ing it is somewhat different. The point then 29.'f1d2 'f1xa3 30.a6 lt:Jb6 3 1 .l:r.cc2!
is to see the move lt:Jc5xd7 in the first r;t;e7 32 .'f1a2 'f1xa2 33 .l:r.xa2 , and Black's
place. That is, overcome one's natural position remains difficult.
reluctance to exchange off the lovely 27.f4
knight, include this exchange in one's Premature is 27.r;t;e3 lt:Jes.
list of candidate moves, give it due 27... h4
attention, and finally, decide to play it, Petrosian and his second Suetin indi
after calculating concrete variations. cate, as the last line of defence, 27 ... lt:Jb6
22 ... :Xd7 23. l:l.c1 28 .'f1ee7 l:If6. How many analysts
In exchanging his knight for the 'bad' against one Fischer! But Black still has
bishop, White did not give away his absolutely no moves.
advantage, but simply transformed it, 28. 'iiff3 f5 29. �e3 d4+ 30. �d2 lLlb6
indeed, increased it considerably. The 31 .l:l.ee7 Ci:KJS 32. l:l.f7+ �e8 33. l:l.b7
threat is to bring the rook to c6. The tbxb4 34. �c4
pawn weaknesses at a6 and dS require The immediate 34.�h7 was also possible,
defending. The bishop on d3 is stronger but White has no cause to hurry. The
than the knight. threat of mate forced Petrosian to resign.
23 ... l:l.d6
In preventing the entry on the 6th rank, Almost 100 years earlier, Paul Morphy
Black has to allow the enemy rook to gave up a lovely-looking bishop for
the 7th instead. an unimpressive knight. There is little
24. l:l.c7 Ci:KJ7 25. l:l.e2 doubt that Fischer knew this game, since
he greatly admired Morphy.
J u les Arnous de Riviere

Paul Morphy
Paris 1 863
25 ... g6
Polugaevsky points out that Black is in
zugzwang. Not 25 ... lt:Jb6 26.�ee7, whilst
after 25 ... a5 there is the «unpleasant»
26.bxa5 l:r.xaS 27.'f1c8#. Therefore
Petrosian opens a bolthole for his king.
27
I have shown the diagram position to 'iVd7) 19 ... �d7 20. �d2 .l::i.ae8 2 1. .l::!.ael .
a number of grandmasters, but none White no longer has an extra pawn, but
guessed Black's next move. Even more he retains the advantage, as is not hard
than that - after seeing what was to see.
played, they all admitted that they did 1 7. 'ifd4
not even consider it. Walking into a blow.
1 4 ... hh2+! ? 1 7 ... tlJf3+!
With hindsight, one can usually work A blockading sacrifice. It is important
out the reasoning behind anything. Let to block the square f3 , preventing the
us try to reconstruct the great master's white pieces coming to the aid of their
thinking. Maybe Morphy was looking king.
at .. .f5 -f4-f3 . But after .. .f5 -f4, the likely It looks as though it is equally good to
reply is f2-f3 , and the bishop on d6 play 1 7... tbxh3 1 8.gxh3 'iVh4 1 9.Vi'd3 f3
is shut off, whilst the knight from h2 or 1 8.f3 'iVh4. But White replies 1 8 .tbd2!
emerges via g4. Therefore Black decides 'iVh4 1 9.tbf3 'iVh5 20.'it>hl , intending
on such an unstereotyped exchange. the defence tbf3-h2 and retaining
1 5. 'iti>xh2 f4 1 6. 'ifxe4 ttJgs chances of saving himself: 20 ... tbxf2+
2 l .<iit>g l tbh3+ 22 .'it>hl (stronger than
22.gxh3 �xh3 23.l::!.f2 .U.f6) 22 ... �g4
23.tt::lh 2 �e2 24. �d2 �xfl 25 . .U.xfl .
Now, however, it is all over:
1 8.gxf3 ti'h4 1 9. .1:lh1 ..bh3 20. �d2
.l:l.f6 0-1
Unexpected exchanges are also seen in

the games of present-day grandmasters.
Veseli n Topalov
Morphy's idea is not unimpeachable. Let NN
us look at the strongest defence: 17."iVd3 . Frankfurt (simultaneous) 1 997
For the present, White has a healthy
extra pawn. The immediate blow on h3
D
is a bluff. Nor is Black achieving any
thing after 1 7 .. .f3 1 8 . �xg5 'iYxg5 1 9 . .l::r.g l
'ife5+ 20.g3 �xh3 because of 2 l .'iVd4!
Vi'h5 (2 1 ...'iVf5 22.g4) 22 .'iVh4, forcing
a favourable exchange of queens. The
best chance is 1 7...'iVd6, and if 1 8 .c5
'iVe5 1 9.'iVd4, then 19 .. .'�Jf3+! 20.gxf3
�h5 . But a modicum of composure
works here too: 1 8 .f3 �xh3 (or
1 8 .. .'�Jxh3 1 9.gxh3 'iYd7 20.<;;t>g l 'iVxh3 The knight on h7 is a feeble speci
21..t'tf2 'iVg3+ 22 . .l:tg2 Vi'el+ 23 .'iVfl , men, whereas the bishop on d3 is a
and it is all over) 1 9.tbd4 (1 9.gxh3 lovely piece. But what can one do, if
28
exchanging one for the other is the As we see, Black has already parted
shortest path to victory. with one of his bishops. The position
1 7 . .hh7! �xh7 has King's Indian characteristics. It is
17 ... 'i:fxh7 18.'i:fe8+ and 1 9.'i:fxe7. not hard to guess White's plan: to pre
1 8 . .bh6! pare and carry out e4-e5, and maybe
After 18 ... gxh6 there follows 1 9. l::i.f7+. to attack on the kingside. Black's plan
The game is over and Black resigned is harder to pin down. Maybe to play
after a few more moves. ... b7-b5 ? Without a light-squared bish
op, this plan is unlikely to be realisable.
It is usual to consider that in open and The break ... e7-e6 ? Then the position
semi-open positions, the bishop is opens, and the white bishop pair are
stronger than the knight, and in closed likely to assume full strength.
positions, the other way round. But Black would like the position to retain a
what about positions which are unde closed character for as long as possible.
fined? In positions which may become But in that case, it is not clear what active
either open or closed, how does one plan he can carry out. And he does not
evaluate the minor pieces? And how, want just to mark time ... Adams shows
correspondingly, does one evaluate the something quite unexpected:
exchange of one for another? 8 ... .bc3+! ? 9.bxc3 tijgf6 1 0.13
That is a good question. It is perhaps Black has parted with the pride of his
most pressing in the opening stages position: the bishop on g7. For what?
of the game, and also in the transi Some results are already visible. White
tion between opening and middle can forget about the break b2-b4, and
game. After a few initial moves, it can the break e4-e5 is more difficult. But
be hard to know whether the position that is only half of it. It is still hard to
will assume an open or closed character, see an active plan for Black.
but decisions regarding exchanges must 1 0 ... 'ifas 1 1 . 'ii'b3 0-0-0 1 2 . .i.e2
nonetheless be made. Knowing the clas l:[dg8!? 1 3. l:[b1 Wfc7 1 4 . .i.e3 �b8
sics can help the decision-making pro 1 5.'ifc2
cess.
Vesel i n Topalov
M ichael Adams
Dortmund 1 996
1 5 ... g5!?
This, it turns out, was what Adams
had in mind. Having closed the posi
tion on the queenside and in the centre,
29
he regrouped, intending the advance An overloading combination. The

... g6-g5 -g4. bishop is split between two flanks .
We should note that Topalov failed to 25. bc4 'Wh3 26. l:lf1
divine his opponent's plan in time. In a 26 .Vi'e2 lDxe4.
later game Mchedlishvili-Gagunashvili 26 ... 'iVxe3
(Tbilisi 2002), which went 1 2 ... <.t>b8 Chess is a funny game. Adams plays
1 3 . Iibl V/i/c7 14.V/i/c2 .l:ldg8, White, a move which anybody would have
having put off castling, was the first to played in his place, giving back the
become active on the kingside: 1 S .g4!? extra piece, but retaining an extra pawn
gS 1 6 .h4!?, and retained some initiative. and the initiative. But this prolongs the
1 6.0-0 l:lg6 1 7.l:lb2 �aS! 1 8. l:lfb1 win for a long time, as White resisted
l:lb8 1 9. <ti?h1 hS 20. 'Wd2 into a rook ending and only lost after
A move which is hard to understand. So 59 moves.
far, the advance of the g- and h-pawns In the meantime, Black wins at once
has not led to the opening of lines, since with 26 ... h4! ! . The threat of transfer
after ... g5 -g4 White always has the reply ring the knight to g3 is unstoppable.
f3 -f4. Now, however, White no longer For example: 27.f5 l:i.g4 28. i.f4 lDhS
has this reply, because the e4-pawn will 29.Vi'd2 �bg8 etc.
be hanging.
Maybe Topalov, true to his style, was Anyone acquainted with the classics
seeking to draw the fire onto himself, can easily see where Adams' idea came
but in that case, both sides have similar from:
intentions.
20 ... g4 21 . 'iVc2 gxf3 22.gxf3 liJeS David Bronstein
23.f4 Tigra n Petrosian
The battle reaches its culmination. Amsterdam/Leeuwarden 1 956
23 ... 'Wc8! Black has not played the opening stages

This piece sacrifice is impossible to convincingly. In the diagram position,
accept: 24.fxe5 Vi'h3 2S . .l:lgl .l:tbg8 White has a choice between several
26.l':tbbl 'ifxe3 27.1:lxg6 .l:Ixg6 28.exf6 active plans: b2-b4 or g2-g4 (or both
Vi'f2 . The square h3 must be defended, at once) , whilst Black seems to have
but then a new blow comes: nothing in return. He is faced with total
24. j.f1 tbxc4! passivity.
30
Sensing the danger, Petrosian took the to g7, from where it takes part in the
only correct decision. He constructed a battle against the advance g2-g4.
fortress, and began by parting with his 27. l:g2 l:.eg7 28. l:bg1 tbce8 29.h3
strongest piece: h5 �-¥2
1 7 ... ..bc3! 1 8.bxc3 tbf6 Incidentally, the resources are far from
What has Black achieved? White can exhausted; the board is still full, of
forget about the plan b2-b4, for one course. One idea is to try to swap round
thing. Secondly, the advance g2-g4 is the bishop on e3 and knight on f2 ,
now more difficult to achieve. Thirdly, put the rooks on g2 and h2, and play
the bishop on g7 does not now obstruct g3-g4. But in any event, Black is excel
a black major piece on the g-file. lently prepared for this break. Bron
1 9.a4 �h8 20. tbf2 l:.g8 21 .�h1 stein evidently decided that half a point
'ife8 22. l:g1 'ifg6 23. 'ii'd 2 was better than none. Indeed, I suspect
So far, each side has fulfilled his min he reconciled himself to a draw from
imum programme. But this move is the moment Petrosian exchanged on
a micro-inaccuracy. After 23 .a5 �d7 c3.
24 . .l:!bl , or immediately 23 . .!::i.b l , Black
cannot develop his queenside without Boris Gelfand
problems. He would be forced either to Vesel in Topalov
defend b7 with a rook, or else to decide Novgorod 1 996
on the committal ... b7-b6, which gives
White a target for a4-a5 , opening a •
second front. There is a chance of tying

the black forces to defence on both
flanks at once.
23 ... ..td7 24.g3 l:ae8 25.a5 J:[e7
26. l:lab1 .tea
Not much time has elapsed since

Topalov's bitter defeat against Ad
ams (see above) , but the Bulgarian
GM demonstrated that he had already
absorbed the lesson given him in that
game.
1 6 ... tbg4! ? 1 7 . ..bg4 fxg4 1 8. tbf2
This is the point. By his delay at move ..bc3! 1 9.bxc3 h5
23, White has allowed his opponent to The tactical justification lies in the vari
deploy his forces in the most harmo ation 20.lbxe4 �e7.
nious way possible. The bishop excel 20. 'ife2 ..tf5 21 . �d2 tbc7
lently defends the b7-pawn. The rook is The knight heads for g7 and from there
not stuck on a8 or b8, but is on its way to f5 . Black has no difficulties at all.
31
«Anyone can win by utilising his friends' strengths. In the King's Indian, this idea has
The real leader is the one who wins by utilising his become a standard one, but it is rarer
enemy's strengths» Napoleon.
- in the Sicilian Defence. However, the
motifs are the same. By giving up his
Incidentally, a word about lessons bishop, Black has spoiled his opponent's
learned and not learned. One often pawn structure. If he manages to keep
hears young players saying things like the position solid, then the pawns on
'Why should I study the King's Indian c4 and e4 could prove hopelessly weak,
(Sicilian, Spanish ...)? I don't play it! '. But and the white bishops helpless behind
the same, or similar, ideas and techni their own pawns.
cal devices often crop up in completely White, in his turn, needs to strive
different openings. Pawn structures are for open play and prevent the posi
often close to identical, as are the prob tion being ossified: However, it is not
lems faced by players. Consequently, entirely clear how he should do this.
one should never restrict one's attention 1 4.15 tlJcb8!?
solely to games in one's own narrow A surprising move. An amateur, if he
opening repertoire! even saw the idea of moving the knight
Many wise and sensible things were from c6, would surely play 14 . ./t:Jas.
written on this subject by the great White would reply 1 5 .l::!.b l, taking con
trainer Vladimir Zak, in his book The trol of the square b3 . In order to get the
Paths to Self-Improvement, which I recom knight to c5 , Black must play ...b7-b6, and
mend to any young player. [Translator's then ...tt::la 5-b7. The pawn on b6 could
Note: This book has not been translated become weak and an object of attack.
into English.] Kasparov intends a piece regroup
ing, but without any weakening pawn
As an example of ideas from one open moves. He plays ...tt::l c 6-b8, then ... tt::ld 7-
ing being seen in another, let us exam c5 , ... tt::lb 8-d7-f6 and ... �c8-d7-c6. If
ine the following game. he manages to carry this out, then the
black minor pieces will occupy ideal
Jaan Ehlvest positions.
Garry Kasparov 1 5.0-0 tlJc5 1 6 . ..tc2 tt:lbd7 1 7 . .:fd1
Linares 1 991 Ivanchuk, analysing this game, pointed
out the move 1 7.'iVf2 . If 1 7... b6, then
• 1 8 .'iVg3 , whilst in reply to 1 8 ...tt::lf6 Black
must reckon with 1 8 .tt::lxe5 !? and White
achieves his desired aim of opening the
position in the centre.
Alas, Ehlvest missed this crucial oppor
tunity. After playing outwardly sensible
moves, but ones which lack a coherent
plan, he gradually drifts into a difficult
position.
1 7 ... tt:lf6 1 8. tt:ld2 ..td7 1 9 . ..tg5 ..tcs
1 2 ... .b:c3+!? 1 3.bxc3 e5 20.'fff3 0-0-0
32
The position has clarified. Black has Black has just retreated his attacked
completed his development, and queen from h4 to h6, offering the
retained all his advantages. Later, he exchange. And White now faces a
carried out the plan we are familiar choice.
with from the game Topalov-Adams: He could exchange: 3 0.'iVxh6 gxh6 ,
... l:td8-g8 , ... h7-h6 , ... g7-g5 , developed but there is a high chance that Black
his initiative and eventually won. will retain drawing chances, even
if he loses a pawn. For example:
The ability to stick to one's guns is 3 1..l:t h l �g7 3 2 . Ilh4 .l::th 8 3 3 . Iiehl
essential in any contest, including .l::ie 7 34.tLlh3 , then tLlh3 -f4-h5 -f6. It is
chess. This hardly requires much evi not very nice, but with the beautiful
dence. For example, insisting on a knight on e S , Black has something to
favourable exchange of an outwardly suffer for.
'good' bishop for an apparently 'bad' White can also avoid the exchange
one. This is clear enough. with 30.'iVc3 , but then, firstly, the king
But let us look at a more subtle case. moves away: 30 ... �f8 3 1. . l::th l 'iYf6,
Imagine two duellists. Both are equally and, secondly, it is not clear what the
good shots, so the choice of weapons queen is doing on c3. Kramnik finds a
does not matter much. But one of them third possibility:
is short-sighted, so it is obvious that the 30. 'iff4!?
main dispute will be over how many You want to exchange? Be my guest -
paces the duel will begin with. but on my terms!
The same applies to chess players. It 30 ... g5
happens sometimes that both players Naturally, 30 ... 'iVxf4 3 l .gxf4 cannot suit
are prepared to make an exchange, Black. The only pride of his position,
but they disagree on the square on the knight on eS, is forced away. But
which that exchange should take place. now, with the queens remaining on the
Whoever wins this argument can win board, Black's position is still not envi
the game. able. The queen on f4 is much stronger
than on c3 , as it takes a direct part in
Vladi m i r Kra m n i k the attack.
Sergey Tiviakov 31 .fxg6 'ifxg6 32.l:.h1 �g7 33. l:.h5
Sochi 1 990 16 34.l:.eh1 &i:Jf7 35. :Sh4
35.l::i.h 7+ 'iYxh7 36.1:1xh7+ �xh7
37.�g4! also wins.
0
35 ... l:.e5 36. ttJd3 l:.g5 37. �h5 :Xh5
38. :Xh5 l:.e7 39. l:.h7+ �8 40. 'ifh4
Black resigned.
Perhaps the still quite young Kramnik

was helped towards the correct decision
by knowing this classic example:
33
M i haii Tal 23 . . . gxh6 24 . .1:.f3 f5 25 . .1:.e1 .l:td6

M i khail Botvi n n i k 26.c3 .l:tg4
Moscow 1 960 It is clear that Black retains the initia
tive .
•
David Bronstein
Isaak Boleslavsky
Moscow 1 950
The last move in the game was 20.'ife2-

e3. Should Black exchange?
There can be no question of playing
20 .. ."ifxe3+ 2 1 .fxe3. White moves the
rook from dl to fl , plays h4-hS , and
then, if necessary, l:i.hl -gl . The pressure
on the g7 and f7-pawns would become David Bronstein often surprised his
serious. opponents with original and unstereo
Nor does Black want to play 20 ... 'ifd6. typed decisions, and this is one of
In reply, there follows 2 1 .hS �h7 22 .h6 many such examples. The position is
and White easily gets rid of his weak closed, and there is only one open file;
ness. Isn't there some way to stop this? it seems that White can only choose be
20 ... 11fh6! tween 22 .'ifxe7 CiJxe7 23.!lfbl and the
This is the correct decision. After immediate 22.I::i.fb l . But instead, there
2 I .'ifxh6 gxh6, the f-file stays closed followed:
and the pawns on f4 and h4 remain 22. 'ifc5!? .l:.b8 23 . .1:.fb1 'ifxc5
weak. Why does Black not fight for the open
It is interesting how this example is a file with 23 ... l::tab6, one may ask?
kind of reverse of the previous one - Evidently, because of 24Jhb6 l:txb6
here the queen retreats back down 2S .c4!.
the same line, rather than advancing 24.dxc5 cJtf8 25 . .1:.b5
towards the enemy queen. Thus, although Black was the first to
21 . .td3 b:d3 22.:Xd3 tLlb6 occupy the open file, it is White who
23. 11fxh6 ends up with it. And this is a direct
Conceding the opponent's wish. result of the move 22 .'ifcs.
34
Chapter 3
Connecting the Endgame with the Openi ng

In chess books, one often comes across discussions of the connection between the
opening and the middlegame. Sometimes, the connection is described as 'close'.
But what exactly does this mean?
It means that, when emerging from the opening, the player should have a very
clear and well-worked out view of his prospects, rather than just a hazy notion. As
well as calculating concrete variations (of the type 'I attack that, he defends it') , he
should understand what middlegame plan he should follow, what pawn structure
he should aim for, which pieces it is favourable to exchange and which not, etc.
When we wish to compliment a player on his positional understanding, we
often describe him as 'thinking in terms of schemes' , ' thinking in plans' . Such a
player can not only calculate variations and see combinational blows, but also look
into the far future and divine the contours of the resulting position.
Connecting the opening with the middlegame is a case of seeing into the future.
One of the great masters of planning was the sixth World Champion, Mikhail
Botvinnik. In his best games, he clearly shows the connection of the opening with
not only the middlegame, but even the endgame.
Here is a characteristic example:
Black voluntarily puts his knight on aS ,

M i khail Botvi n n i k from where (after ... a7-a6 and ... b7-bS)
Efi m Geller it can exert pressure on the c4-pawn
Moscow 1 952 and White's queenside more generally.
1 .d4 ti:Jf6 2.c4 g6 3.g3 �g7 4.�g2 Even so, the knight stands badly on aS. It
o-o s.tbc3 d6 s.tt:Jf3 tbcs 7.ds tt:Jas has no chances of returning to the main
8.tt:Jd2 cs theatre of action. Where can the knight
go to from aS? To b7? And then only
to d8 . All of these square, aS , b7 and
d8 , are ones from where it is ineffective.
The knight is offside for the long-term,
maybe for ever.
In 1 9S2, the theory of this variation
(and, indeed, the King's Indian as a
whole) was only just coming into being.
Later, it was shown that the variation is
viable, but Black must play extremely
energetically. He must attack the white
Black has played the Yugoslav Variation centre from all sides, with: ... b7-bS ,
of the King's Indian Defence, in which ... e7-e6, and if necessary, even .. .f7-
3S
fS; he must wriggle and writhe, and 2a ... eea 24 . ..txg7 tbxg7 2s . ..th2
exploit the slightest chance of activity. eea 26.t4
Only in this way can he hope that the
knight's being out of play will not play
a significant role, and that this will be
outweighed by other, more dynamic
factors.
White's play is clear in terms of general
characteristics: he should dry up the
position with exchanges and head for
the endgame. The middlegame passes
under the general aim of restricting
Black's activity. This, Botvinnik suc
ceeds in doing admirably. An important moment. The break
After 1 7 moves, the position looked like e4-eS is impossible to prevent, but it is
this: possible to exploit its downside, by pre
paring to take twice on eS, and free the
• d6-square for the knight.
A sample variation might be 26 . . .
f6 2 7.'ifc3 ltJb7 2 8 .eS fxeS 29.fxeS
dxe S ! ? 3 0 .'ifxe S , and now either
30 . . . 'iff6 3 l .'ifb8 ltJd6 3 2 .'i:Ya7 'iff2
33 .ltJdfl .tfs 34.'ifxe7 .te4 (or
3 3 .ltJefl ltJf6 34.'ifxa6 .tfs) , or else
(if Black is not in the mood to sacri
fice anything) 30 . . . 'ifd6 3 l .'i:Yb2 'ifc7,
although in this case, he must reckon
The rooks have been exchanged and with 3 2 .d6 ! ?.
Black has no entry squares on the open Of course, White is not obliged to
b-file. The knight on aS sticks out like allow all this. Instead of the immediate
a sore thumb. White has a clear plan: 28.eS he can retain pieces on the board
e4-eS. In the rest of the game, he is for a long time, and choose the most
practically playing with an extra piece. convenient moment for the break
Black's position is difficult. through. He can also include his king
But not hopeless. In the diagram posi side pawns in the attack, for example
tion, Botvinnik recommended 23 ... eS!?. with 28 .h4!?.
The move looks anti-positional: as well 26 ... tl::lf6 27.e5 eea 28. 'ifc3 f6
as one bad piece on aS, Black now has An extremely committal continuation.
a bad bishop on g7. Even so, the move By allowing his opponent to establish a
makes sense. If White does not react to huge pawn on e6, Black himself creates
it, Black has chances to barricade the the conditions for a white mating
centre and establish a fortress, whilst attack ..
after 24.dxe6 .txe6, the stranded knight In Volume 2 of his Analytical and Critical
on aS can return to the centre via c6. Works, Botvinnik wrote that « . it was dan-
. .
36
Chapter 3 - Connecting the Endgame with the Opening
gerous to exploit the chance to play 28 ... �a4 in

view of 29.4.Jg4, and White's attack grows». Con
tinuing this variation: 29 ... hS 30.4.Jh6+
�g7 3 l .e6+! f6 (3 1 ...�xh6 32 .�h8#)
32 .4.Jf7 4.Jb3 33 .fS! �as (33 ... 4.Jxd2
34.�xd2 gS 3S .�e2 with mate)
34.4.Je4 gxfS (34 ...�xc3 3S .tt.Jxc3, and
the bishop is lost) 3S .�e3 ! is enough to
convince one that Botvinnik is right.
But it is hard to understand why Black
refrained from the move 28 ... 4.Jb7. 35.f5!
After all, for the last ten moves, he has If we go back to the position of the first
been marking time, so why choose this diagram (after 8 moves) , we can see
moment to change course? Maybe he that White has realised absolutely all of
was disturbed by the concrete variation his ideas. Everything he dreamed of has
29.e6!? fxe6 (he cannot refuse to cap occurred on the board.
ture the pawn: 29 ... �a4 30.exf7+ c;tJxf7 Of course, this is not just the result of
3 1 .�h8 4.Jf6 32.4.Je4) 30.dxe6 �xe6 magic, but of the powerful and care
3 I .fS!? (after 3 1 .4.JdS 'i!fd8 , the attack fully-planned work of the sixth World
comes to nothing) 3 1 ...gxfS (hopeless Champion.
is 3 1 ...�xfS 32 .4.JxfS gxfS 3 3 . �dS+ 35 ... lt:lb3
e6 34.�xe6+ 'it>f8 3S.�h8+ �e7 Agony.
36. �xfS) 32 .4.Jds 'i!fd8 33.4.Jf4 �c8 36. 'ifb2 'ifa5 37. lt:lxb3 'ife1 38.�2
34.�dS+ e6 3S .tt.Jxe6 �xe6 36.�xe6+ 'ifd1 39. lt:lc1 gxf5 40 . .i.e2 'ifd4
�f8 37.'ifh8+ �e7 3 8 . �dS , and he 41 .'ifxd4
has to acknowledge defeat. Stronger is The time control is passed and Black
32 ... �aS! White regains the sacrificed resigned.
material: 33 .�e3 (33 .4.Jxe7+ <;t>f?) «This is the only game I managed to win against
33 ...4.Jg7 34.4.Jxe7+ <;t>fg 3S .4.JxfS �xfS my ingenious opponent»- Botvinnik.
36. �xb7, but it is still early to speak of
his winning. Alexander Baburi n
Again, I would point out that White is l mad Hakki
not obliged to force matters. Instead of Yerevan 1 996
29.e6, he has a mass of other moves:
29.g4, 29.h4 etc. In all cases, Black's
0
position remains passive, unpleasant
and difficult to defend.
29.e6 .b4 30 . .i.e4 �7
He loses a piece after 30 ... 4.Jb3 3 l .�b2!
�aS (3 1 ...4.Jc7 32.�c2) 32 .4.Jxb3 �e l
33 .'ifd2, but now he has not a single
active piece remaining.
31 . ..td3 lt:lh5 32 . .i.e2 �7 33.h4
lt:le8 34 . .i.d3 �7
37
Black has played the same vanat10n been ideal. However, this has not been
(and made the same mistakes) as in the for nothing from a spectator's point
game Botvinnik-Geller, but White has of view, as we now get an interesting,
placed his pieces somewhat differently. highly instructive endgame. Black's
His dark-squared bishop is not on the problems have not disappeared - his
long diagonal and neither of his knights knight remains out of play.
is in contact with the eS -square. 35 . ..te2 tba4 36. tt:\12 16 37 .e6 15
For the moment, White has to wait aa. tt:\d1 tt:lb6
with the advance e4-e5 , and instead we A fortress? Not quite. White certainly
get a phase of manoeuvering. isn't going to break through on the
22 . ..t11 tt:\h5 23. �h2 Yi'b7 24. Yi'c2 queenside, but on the other flank, there
Yi'c7 25 . ..td3 tt:lb7 26. ttJe2 tt:\16 remains one final break.
27. ttJg1 ttJd7 2a.tt:\13 'iVb6 29. tt:ld1 39.g4 ..ta4 40. tt:\e3 ..td4
Psychologically, an understandable Not 40 .. .fxg4 41 .hxg4.
move (White is afraid of losing his 41 .gxf5 be3 42. be3 gx15
advantage and covers the b2-square 43 . ..td3
first of all), but not an obligatory one. The activation of the knight can
If one were to copy Botvinnik 's plan, be stopped with 43 . �d2, but then
one should begin with 29. �c3 . It looks 43 ... �c2 .
as though Black can create counter 43 ... tba5 44. tt:ld2 .tea 45 . ..t12 ..tg6
play: 30 ... �xc3 30.li'xc3 �b4 3 I .li'c2 The position has undergone some
tt:lb6, but let us continue this analysis: changes. White's plan is clear: to take on
32 .e5 tt:la4 33 .tt:ldl or 32 ... �a4 33 .li'e2 fS , but in such a way that the opponent
�c3 34.tt:lg4. The counterplay comes to cannot take on c4. It seems the decision
nothing, and White develops his attack is elementary: transfer the king to c3 ,
in the classical fashion. but then there is a check on a4. Con
29 ... Yi'c7 30. r;,t>g2 tt:lb6 31 . ttJe3 tba4 sequently, longer and more complicated
32. tlJQ4 'iVb6 33.e5 'iVb2 34. Yi'xb2 manoeuvres are required.
tt:\xb2 46 . ..te1 �a 47.r;,t>t3 h6 4a. r;,t>e3 r;,t>ea
49 . ..te2 �a 50. r;,t>d3 .tea 51 . r;,t>c3
tba4+ 52. r;,t>c2 tt:lb6 53 . ..th4 ..tg6
54. �c3 tba4+ 55. r;,t>d3 tt:lb6 56 . ..td1
.tea 57 . ..tc2 ..th5 5a . ..t12 .tea
59 . ..te1 'itg7 60 . ..tb3 ..th5 61 . tt:\11
White has managed to get in the break

e4-e5, but at a high cost: the queens
have been exchanged. One can con
clude that White's manoeuvres have not
38
The key moment in the game. Or 65 ... �g8 66.ctitc2 tLlb7 67.ctitb3 , and
With his last move, White set a trap: the f5-pawn falls anyway.
6 1 . ..tLlaxc4 6 L�xc4 tLlxc4 63 .tLlg3 tLlb6 66. tillcf5 ..tgs 67. lbe3 .tea 68.f5
64.tLlxh5+, winning a piece. Strangely After a few more moves, White won.
enough, this is not the end of the
battle, which, on the contrary, flares Elmar Magerramov
up with new strength. After 64 ... ctitg6, Tony Molina
White faces a crossroads. If he hangs Dubai 2007
on to his material with 65 .tLlg3 tLlxdS
66.i.d2 ctitf6, he may face insupera
D
ble difficulties in converting. A more
interesting route is to return the piece:
65.i.h4!?. A likely development of
events then is 65 ... ctitxh5 (65 ... tLlxd5
66.tLlg3 tt:Jxf4+ 67.ctitc4 tt:Jxh3 68.i.xe7
tLlf4 69. i.xd6 tLlxe6 70.tLle2) 66. i.xe7
tLlc8 (66 ... tLlxd5 67. i.xd6) 67. i.f8 (the
alternative is 67. i.d8 ctitg6 68 .�c4 �g7
69.ctitb3 �f8 , and now 70.�a4 tLle7
is useless) 67...ctitg6 68.�c3 (68 .�c4 As Grandmaster Elmar Magerramov
tLlb6+) 68 ... h5 69.ctitb3 �f6, and again pointed out, even in the most primi
the breakthrough fails: 70.�a4 tLlb6+. tive variation 23.'iVb3 'iVxb3 24.axb3
Instead of 62 .i.xc4, the immediate aS 25 .tLlc3 �xc3 26. �xc3 a4 27.bxa4
62 .tLlg3 is more subtle, and then i.xa4 28.tLle3 , White retains some
62 ...tLlb2+ 63 .ctitc2 i.dl + 64.�xb2 chances, that's how badly both of
i.xb3 65 .�xb3 ctitg6 66.tLle2 tLlxdS Black's knights are placed.
67.�a4. White has real winning 23.lbe1 !
chances. The knight has most trouble But this is a good deal stronger. Firstly,
coping with a rook's pawn. If we White takes control of b l , preventing
manage to take on a6, the rest will not Black penetrating down the b-file, and
matter. secondly, the knight heads to f3 , from
Black, in his turn, can part with a piece where it will support the break e4-e5
in another way: 61 ...tLlxb3 62 .axb3 and the attack in general.
i.f3 63 .tLlg3 (63.tLle3 i.e4+) 63 ... �g6 23 ... 85 24.g4 f6
64. i.a5 tLlxdS!? (the passive path Black desperately wants to avoid this
64 ... tLlc8 leads to defeat: 65.i.d8 ctitf6 move, but he has no choice. The knight
66.�c3 i.e4 67.ctitb2 �g6 68 .�a3 i.c2 on h6 is completely unsuited for any
69.�a4 ctitf6 70.tLlfl !) 65 .cxd5 i.xd5. sort of counterplay, so Black re-routes
Again, there is no guarantee that it will it to f7.
be possible to realise the extra piece. 25.tbf3 tbf7
61 . ..ti:Jb7 Black is prepared for the break e4-e5.
Capitulation. But White has a wide front for opera
62. lbe3 i.g6 63. i.c2 �8 64. �c3 tions and the breakthrough can come
0.85 65. i.d3 �e8 anywhere.
39
26. 'ifc2 .th6 27. tt:Jf2 a4 28. �g3 in view of 36 .. .'ifxb7 37.t2Jxd6+ and
'ifa6 29 . .tf1 38 .t2Jxb7.
White is in no hurry. The queen has
retreated to c2, freeing the square d3 , In talking of the connection between
but it is not yet clear for which white the opening and the endgame, one
piece. Let the opponent suffer. should not forget that this is a two-way
29 ... e5 street, and the reverse direction can
And Black cannot stand the tension. have at least as much influence, if not
Now the position opens up, and there more, than the direct one.
is no need to say whom that favours. Let us assume that we have an opening
In analysis, it is interesting to test 29... variation, or a whole system, in which
gS . White has two paths to choose from. straight after the opening many pieces
He can play for the attack: 30.fxg5 tDxgS are exchanged and we go into an end
3 l . �xg5 �xgS 32 .t2Jxg5 fxgS 33.e5. Or game, with the middlegame practi
he can forget about the kingside, play cally non-existent. Let us also assume
30.f5 , burying the bishop on h6 , and that the resulting endgame is assessed
transfer the battle to the opposing flank. as unquestionably better for one side
30.dxe6 .ixe6 31 .f5! gxf5 or the other. What does this lead to? It
In such positions, winning variations leads to the variations of the opening
pop up of their own accord. Here is being adjusted and adapted, in the light
one: 3 l ...�c8 32.�xh6 t2Jxh6 33 .'i:Vb2 of the endgame. The side whom the
Wf7 34.g5 t2Jg8 35 .gxf6 t2Jxf6 36.t2JgS+ ending favours will try to fool its oppo
We7 37.t2Jxh7!. nent into exchanging pieces, whilst the
32.exf5 .td7 other will refuse all such offers and will
Or 32 ... �c8 33. �xh6 t2Jxh6 34.t2Je4 dS try to initiate complications favourable
35 .cxd5! ifxfl 36.'i:Vxa4, with a decisive to himself, seeking chances in the mid
penetration - analysis by Magerramov. dlegame, rather than heading into a
33 . .ixh6 tt::\x h 6 34. tt:Je4 �7 cheerless endgame.
Thus we have a direct link - opening
middlegame-endgame. But suppose the
endgame is re-assessed. What happens
then? We get a corresponding reassess
ment of the middlegame and even the
opening.
Siegbert Tarrasch
Richard Teichmann
San Sebastian 1 91 2
1 .e4 e6 2.d4 d5 3 . tt:Jc3 tt:Jf6 4 . .tgs
35. 'fi'b2! tt:Jga 36. 'ii'x b7! .te7 5.e5 tt:Jfd7 6 . .txe7 'ifxe7
Just as this knight ended up on the 7. 'ii'd 2 0-0 8.f4 c5 9. tt:Jf3 tt:Jc6 1 0.g3
edge in the opening, and got under a6 1 1 . .tg2 b5 1 2.0-0 cxd4 1 3. tt:Jxd4
its player's feet in the middlegame, so tt:Jxd4 1 4. 'ii'x d4 'ifcs 1 5. 'ifxc5
it dies a hopeless death. Black resigned tt:Jxc5 1 6.tt:Je2 .td7 1 7. tt:Jd4
40
The famous French endgame: the 37 ... b4 38. :h7+ <iii'd 8 39 . .td3 :ca
blockading knight on d4 against a light 40.a3 as 41 . :h8+ 1 -0
squared bishop, shut in behind a barrier
of its own pawns. For many years, even Under the influence of this and similar
whole generations, this endgame was examples, theory started to develop in
assessed as unequivocally in White's a predictable way. Many French vari
favour. He would maintain the dark ations, and even those of other open
square blockade and prepare a break ings, were developed so as to result
through on either the queenside or in a similar endgame. White would
kingside. Black is condemned to passive exchange the pawns on d4 and cS ,
defence. exchange dark-squared bishops and
The ending of the game Tarrasch queens, transfer his knight via bl -c3 -
Teichmann was the cornerstone of such b5-d4 (or bl -c3 -e2-d4) and bank on a
assessments. We do not need to spend a quiet, risk-free realisation of his advan
lot of time on it, as detailed analyses of tage. Black, as far as he could, would
the ending can be found in numerous refrain from exchanges, often at great
textbooks. cost to his position.
During the course of the game, White Gradually, the assessment of such end
. exchanged one blockader for another games became axiomatic, not requiring
(after the exchange of knights, his any proof. But in the mid-1 980s, thanks
king came to d4) , broke through on to the efforts of various representa
the kingside (with first h4-h5 and then tives of the Kharkov school of chess
g4-g5) , increased his advantage and, (especially Alexander Chernin and
not without some adventures, realised Mikhail Gurevich), this assessment of
his advantage. the endgame started to be questioned.
1 7 ... :aca 1 8 . ..W2 :c7 1 9. <iii'e3 The first game to start this process was
:ea 2o.:t2 tbb7 21 . .tt1 tba5 the following.
22.b3 h6 23 . .td3 tbc6 24. tbxc6
.txcs 25. <iii'd 4 .td7 26.g4 .tea Jan Timman
27.h4 gs 2a. :h1 <iii'g7 29.h5 :ha Alexander Chern i n
30. :th2 .td7 31 .g5 hxg5 32.fxg5 Montpellier 1 985
:Xh5 33.:Xh5 gxh5 34.:Xh5 ..Wa 1 .e4 e6 2.d4 d5 3.tbc3 tbf6 4.e5
35. :h8+ <iii'e7 36.g6 fxg6 37 . .bg6 tlJfd7 5.f4 c5 6. tbf3 tbc6 7 . .te3
41
cxd4 8.ttJxd4 �cs 9. 'ifd2 bd4 became replaced by 'quiet' and 'calm'.
1 0. hd4 ttJxd4 1 1 . 'ifxd4 'ii' b6 For example, the game Ehlvest
1 2. 'ifxb6 ttJxb6 1 3. lbb5 �e7 Gurevich (Linares 1 991), instead of
1 4.0-0-0 �d7 1 5. tbd4 14.0-0-0, went 14.\td2 .id7 1 5 .tt:ld4
tt:la4 16.b3 tt:lc5 1 7. .id3 g6 (also inter
esting is 1 7... g5 !? 1 8 .fxg5 .l:Iag8 1 9.h4
h6) 1 8 .a4 aS 1 9.rf.afl h5 20.g3 .l:Iag8
2 1 .h4, but already after 2 1 ....l::rc 8 22.�a1
.l:Ic7 23.�hb1 .l::ta 8 24.�a3 .ie8 , White
abandoned any attempt to win and
offered a draw.
This reassessment led to a process of
reassessment in the reverse direction:
endgame-middlegame-opening. Now
it is Black who offers exchanges (the
1 5 ... �4 1 6 . .td3 tbcs 1 7J:[de1 hS solidity of the endgame having been
1 8.g3 :Sg8! ? 1 9.l:hg1 gS!? demonstrated many times) and White
Once the discovery is made, it all seems who avoids them, preferring a compli
so clear and obvious. Certainly, what cated middlegame. Opening variations
could be simpler than the break ... g7-g5 previously assessed as 'cheerless' are
- after all, this idea works in many dif now described as 'solid'. Whole systems
ferent variations of the French Defence. have been re-thought and their theory
But for some reason, before the Kharkov rewritten from scratch. And this has
boys came along, nobody thought of affected not only the French Defence.
doing it in a position without queens. Thus, one of the variations of the Caro
20.f5 g4 21 .f6+ �d8 22. l:e3 h4 Kann Defence, l .e4 c6 2.d4 d5 3 .e5 c5! ?
And in the end, it was only by a miracle 4.dxc5 e 6 , which Botvinnik years ago
that White saved himself. said was no good, because White could
And there is more. It turned out that, obtain the famous 'French endgame',
if Black failed to get in ... g7-g5 , (e.g. has again become trendy. You want to
if White shuts off this counterplay by play the white side of this ending with
means of h2-h4) , the endgame is still an extra tempo? Be my guest - Black
not so bad. The term 'passive defence' does not object.
42
Chapter 4
The Battle of the Major Pieces

The avantgardists, or, as they liked to call themselves, the 'hypermoderns', have
left a large legacy in chess. Some of their ideas have not survived and now appear
unrealistic, even raising a smile, but many others have survived and become a part
of everyday chess culture. Yet at one time, these ideas caused a lot of arguments
and even shocked an unprepared chess public!
We will now take a closer at look at one typical hypermodern device, namely the
queen and bishop battery on the long diagonal. The classical positional approach
envisaged the queen working in the centre, with the bishop behind it (e.g. white
queen on d4 and bishop on b2) . But the hypermoderns put the queen in the cor
ner, and, to people's great surprise, such an arrangement brought results.
occupies this square, will come to c1

Richard Reti and c2 . The other rook, by completing
Frederick Yates the doubling on the c-file, will free f1
New York 1 924 for the knight, which will transfer via
that square to e3.
These manoeuvres take quite a bit of
0
time, it must be acknowledged. Fur
thermore, even after they have been
carried out, the pressure against dS and
eS will not be so enormous. As a result,
Black can prepare to defend the pride of
his position, his pawn centre. But the
fact is that White's plan was so novel
and unconventional for its time that
Reti's opponent was bamboozled.
By the benchmarks of the classical 1 1 . :c1
school, White has given his oppo In such positions, the hypermoderns
nent a great present - he has allowed sometimes implemented their plan in
him to create an ideal pawn centre. In an even more fanciful manner by a2-a3 ,
return, he has fianchettoed both bish l:ral -a2 and 'irYdl -a l . Strangely enough,
ops, making it clear that the centre will even this plan worked sometimes,
come under pressure in the near future. although only with pawns still on c4
The opening is not yet over and White and c6. With the c-file open, it makes
still has to complete his development. much less sense to put the rook on a2 .
So how does Reti intend to attack the 1 1 ... lbf8 1 2. :c2 i.d7
dS- and eS -pawns? To do so, he intends Instead of this, Alekhine suggested
a series of manoeuvres. The queen will 12 ... �f5 1 3 .'iVal tt::l 8 d7, but the move in
head to a l . The rook, which currently the game is not a mistake.
43
1 3. 'it'a1 ! ? ttJg6 1 4.l:.fc1 i.c6 Florin G heorg h i u

But this move is superfluous. The doubled Anatoly Karpov
rooks on the c-file do not yet pose any Luzern 1 982
threat, since they have no entry squares.
Black could have chosen between •
14 ...'iVe7 l S.tt:Jfl .l:lac8 16.lhc8 .l:r.xc8

1 7. .l:lxc8+ ..txc8 1 8 .d4 e4 19.tt:Jes 'iVc7
and the sharper 14 ...b5 l S .tt:Jfl 'iVb8!?
(Nimzowitsch's recommendation).
1 5.ttJf1 "ifd7
Again 1 5 ...'iVe7 is stronger.
1 6.ttJe3 h6
After two successive inaccuracies, Black
commits a third, this time more seri
ous. He has not divined what mischief The queen does not feel comfortable on
his opponent has in mind. d8 , opposite the .!:r.dl . The pawns could
be exchanged and the d-file opened at
any moment, and there is no time to lose.
1 4 ... l:.c7! ?
The queen heads to a8 . This manoeuvre
is not only no sort of shock, but even
looks perfectly normal - what progress
has been made in 60 years!
1 5.e4
A cunning move. White presents his
opponent with the chance to determine
the central pawn configuration.
1 7.d4! e4 1 8. ttJe5 This is not an easy choice. Thus, after
Now it becomes clear that the pieces on 1 5 ...'iVa8 1 6.exd5 exdS 1 7.tt:Jg4, new
c6 and d7 have set themselves up to be factors appear in the position - after the
attacked. The knight cannot be tolerated exchange of knights, mate is threatened
on eS, but after its exchange, the 'it'+i. on g7. Karpov takes a key decision - he
battery operates at full power. keeps the e-file closed, but opens the
1 8 ... .b:e5 1 9.dxe5 ttJh7 20.f4! two neighbouring files.
White has a serious advantage and soon 1 5 ... cxd4 1 6 . .b:d4 dxc4!?
won. 1 7. ttJdxc4 'it'a8 1 8.f3 l:.fc8
Returning to the position before 1 6 ... h6, The two sides each get on with their
one can advise that instead of this, Black own thing. The "if +i. battery is faced
should prefer 1 6 ... d4 (although then with the rock-solid pawn barrier f3 -e4
White is a little better after 1 7.ltJc4) or (exactly what Yates failed to do in the
1 6 ... tt:Jg4!?, hoping after the exchange of preceding example). On the other hand,
knights to cement the a l -h8 diagonal by White has not managed to take the initi
means of .. .f7-f6. The point eS needs to ative. One complicated position has led
be defended at all costs. to another. The chances are mutual.
44
Chapter 4 The Battle of the Major Pieces
-
Vesel i n Topa lov extremely unpleasant for the opponent.

Vlad i m i r Kra m n i k This is confirmed by the a7-pawn and
Wijk aan Zee 1 999 the g7-bishop.
1 5 ... ltJe8 1 6 . .bg7 ltJxg7 1 7.0-0 ltJe6
• • • .. . > 1 8.l:.b1 .tc6 1 9.d4 .te4 20. l:.d1 as
.t � .i i A i 21 .d5! ltJcS 22.tbcl4 .txg2 23. �xg2
i .l • - · l:.fd8 24. ltJc6
White has an indisputable advantage,
which he subsequently converted to
.. victory.
� Cjj �
� � iLA: How should the major pieces be
: .u �· arranged on an open or half-open file?
The rule is simple: the queen should
1 3 ... l:.c7 1 4 . .th6 'ifa8! ? 1 5 . .bg7 stand behind the rooks.
�xg7 1 6. 'iVd4 b5! ? Exceptions are rare. It is another matter,
A new note i n a familiar tune. The however, that one should never blindly
queen on a8 exerts pressure not only follow the standard. For example, in
along the long diagonal, but also on the positions where there are only two
a-file. major pieces on the board, rather than
1 7.cxb5 axb5 1 8.ltJxb5 l:.xc1 three, the rule of 'queen behind the
1 9.l:.xc1 'ifxa2 rooks' is less valid. But more about that
Black obtained the advantage, on later.
account of his superior structure. We start with a fully canonical example:
Again we return to the source. Alexander Alekh ine

Aaron N i m zowitsch
Richard Reti San Remo 1 930
Akiba Rubi nstein
Karlsbad 1 923
The pin on the knight at c6 is deadly.

White is close to victory and all that
1 5. 'ifa1 ! remains is to arrange the major pieces
Maybe it is the corner, but it is still harmoniously.
the best square for the queen, and 24.l:.c3! 'ifd7 25. 1:.1 c2! �8 26. "it'c1 !
45
It is done. Black could resign with a attack from the knight on b4. In any
clear conscience already, but he evi event, White should not put the queen
dently set himself the aim of not losing on c2 . That is correct. Even so, the rule
before move 30. (and the classical examples) needs to be
26... l:l.bc8 27. ..b4 b5 28. .b:b5 <oi;le8 known. The more he knows, the more
29. i.a4 <ito>d8 30.h4! easily and confidently a player plays.
A beautiful position. The board is full, If you are in a strange place, you will
yet the queen and rooks cannot move, probably find the right road sooner or
the knights cannot move and the pawn later. But if you have a map, or a reliable
moves will soon run out. After 30 ... �e8 guide? Well, obviously it is much easier.
there follows 3 l .b5. Yes, Black has made
move 30. So he resigned. William Winter
We will remain with this final posi Alexander Alekh ine
tion a moment longer, and just change Nottingham 1 936
very slightly the formation of the white
major pieces. •
White has manoeuvered his pieces

exceptionally badly and chosen very
All I have done is swap round the c2-rook strange squares for them. They cannot
and the queen. A tiny detail, but: create a single threat. Black, in his turn,
30 ... tbxb4! 31 . .bd7 can do whatever he pleases, a rare luxury.
3 l .'iVb3 �xc3 . The right plan is easy to find. We put the
31 ... tbxc2 32 . .bc8 .:Xc3 33 . .be6 queen behind the rooks, create threats
.:cs! on the e-file, and at the right moment,
It is too early for 33 ...tbxd4 34. �xc3 bring the knight at c6 into the action; at
tbe2+ 35 .\t>f2 tbxc3 because of 36 .tbd4 the moment, it is not doing anything.
g6 37.h5 . 23 ... .:0e7! 24. <oi;ld1
34. i.f7 tbxd4! After 24.tbgl White loses a pawn:
Now it is time. 24 ... .ixf4 25.klxe6 .ixd2+ 26.Vi'xd2
35 . .:Xc6 tbdxc6 36. �2 tl:lb4 'ifxe6 27.tbf3 , but maybe this was the
Black has good chances of holding the least evil.
position, which not long ago looked 24 ... 'ife8
totally hopeless. «The position of the queen behind the rooks
The reader may object that, in this last promises the maximum force in the invasion»
-
example, the queen came under a direct Chebotarev.
46
-
Oleg Chebotarev, in his main special It was necessary to play 26.'ifxd5 .l::r.xe2
ity, was a military man, a major-gen 27.'ifxa5 tt::lg3 28.tt::le S!. The position has
eral, so when it came to invasions, he become much sharper. If, for example,
spoke with full knowledge of what he 28 .. J lxel+ 29Jhel fxeS 30.fxe5 c6
was talking about. (30 .. J �e6 3 l .'iVxa7) 3 l .exd6 .l::r.xel +,
25. Wff3 then 32 .�c2! (of course, not 32.�xel
Here too, 2S.tt::lg l was worth consider 'ife2+) 32 ... 'ife4+ 33 .�b3 'ife6+ 34.d5!
ing, although after 2S ... tt::lg3 26.l::r.xe6 and it is not clear who has the advan
l:txe6 27. l:tel tt::l e 4, White's position tage. Stronger is 28 .. Jhg2 , retaining the
remains difficult. threats, but this move had to be found,
25 ... tba5! which is not entirely simple.
Now, however, it is all over:
26 ... tt:lc4! 27 . .tc1
Or 27.bxc4 'ifa4+ 28 .�cl �a3+
29.�bl .U.b6+ 30.�al 'ifc2 , mating.
27 ... tt:lce3+ 28 . .be3 tbxe3+
29. tbxe3 :Xe3 30. 'iff2 'ifbS 31 . tt:lc1
The threat was 3 l ...'iVd3+ 32.�cl �a3#.
31 ... :Xc3
White soon resigned.
In the press, thanks to these two games

26.b3 against Nimzowitsch and Winter, the
Kotov, in the second volume of his formation of queen behind rooks is
Alekhine's Chess Heritage, gives this vari now known as 'Alekhine's Gun'.
ation: 26.'ifxd5 l::txe2 27Jhe2 l:rxe2 In contemporary practice, such a for
28.'ifxa5 tt::lg3 29 . .l::r.f3 'ife4, «and Black's mation occurs often. We can identify
attack is irresistible». This variation can two typical positions:
be extended, showing more vivid
colours: 28.tt::lxf6 .l:!.xd2+! 29.�xd2 1. The battle for a rook's file, which is
'ife3+ 30.�dl 'ifd3+ 3 l .�el 'ifxfl+! currently closed, but may open at any
32.�xfl tt::l e 3+ 33 .�e2 tt::lxdS 34.tt::lxd5 moment
c6 3S.tt::lf6 �xf4, but this is not the Imagine the following structure: White
main point. pawns f4, gS , hS. Black: fS , g6, h7. The
Winter, like Nimzowitsch, does not put battle rages over the h-file. White puts
up the slightest resistance: the moves one rook on h3 or h4, then a second, and
24.�d l , 2S .'iff3 , and now the passive his queen on h l . Black cannot oppose
defence of c4, which in fact does not with a similar set-up, as his major
defend anything in reality. In our day, pieces simply do not have enough space
defensive technique is at a much higher on the h-file. He has to cede the file. At
level, and one does not win a game so the right moment, White exchanges on
easily. The opponent will find some g6, penetrates with his rook to h7 or
way to change the nature of the posi h8 , and brings his queen with decisive
tion, set some problems. effect to h6 , and the game ends.
47
The same thing can happen on the 27. lLlf1 !

opposite flank: For the moment, it is not obvious where
the knight is heading, so Spassky carries
Robert Fischer on waiting.
Boris Spassky 27 ... �e7 28. lLl1 d2 <Jtg7 29. lLlb1
Belgrade 1 99 2
o
:i i�tJ�, , , , .... .· :& �,ii�:'(Bi'
��::;,; .t. ...�;fl:� a ; ;,� .t. • a
.t. �� -� ��0)': .t. iii
.t. ii� 15, 'a , ...�ii .t.
�5, t:/1..\ �r A . ,�. >'
'. ;� ttJ � 15,

;'�
·:.•':��- '8 15, ��i
B. ;��:� ;tt m
22.l:ta3!? ttJdf6 The idea starts to reveal itself If it
Straight after the game, commenta were his move, White could imme
tors recommended 22 ... h4 B .tbfl f5 , diately exchange all the major pieces:
and then 24.exf5 gxf5 25 .tDg5 tbxg5 30.axb5 axb5 3 l..�xa8 .l::tx a8 32.ltxa8
26. �xg5 �f6 27.�h6 �g7 28. �xg7 'ifxa8 33.'ii'xa8+ �xa8 , after which he
'i!;>xg7 29.f4 e4 30.tbe3 , with only an wins a pawn with 34.tba3 . The fortress
insignificant advantage to White. Even collapses.
so, one cannot shake off the impres The long wait is over. In an attempt to
sion that White is better prepared for deflect his opponent from his inten
the opening of the centre and kingside, tion, Spassky goes in for a risky piece
and so it should not be hard to improve sacrifice:
this variation. For example, instead of 29 ... lLlxe4!?
29.f4, there is the tempting 29.tbe3, and Of course, he could simply cede the
if 29 .. .'ifg5 , then 30.axb5 axb5 3 l..�xa8 a-file: 29 .. J:tab8 30.axb5 axb5 3 1 .lta7,
�xa8 32. �xf5 ! ltxf5 33 .tDxf5+ 'ii'xf5 but one concession would most likely
34.1::!.a l , winning material: 34 ...'ii'f8 soon be followed by others.
35.'ii'g 5+ �f7 36.�a7, or 34 ... �b7 30 . .b:e4 f5
35.l:!a7 �c8 36 . .l:!.c7 �a6 37.'i:Ya2 . This move was criticised by the experts.
Spassky prefers waiting tactics: Instead, they suggested 30 ... tDf6, and
23. l:.ea1 'ifd7 24.l:[1 a2 l:[fc8 Black gets a second pawn for the piece,
25. 'it'c1 Af8 26. 'it'a1 'it'e8 without weakening the second rank.
White has set up the Alekhine Gun. Now, however, the black king turns out
But Black's defensive barriers are solid. to be in big trouble:
Thus, after 27.axb5 axb5 28 . .l:ta7 l:ha7 31 . �c2 .b:d5 32.axb5 axb5
29. .l::rx a7 l:la8 , White cannot get any 33. l:ta7! <Jtf6 34. lLlbd2 .:Xa7
thing from the a-file. In order to break 35 . .:Xa7 :aa 36.g4!
down the fortress, an additional idea is In such positions, the piece is stronger
needed. And Fischer finds it. than the pawns. Spassky soon resigned.
48
Chapter 4 - The Battle of the Major Pieces
2. The battle for and against a central not yet a threat. Black could have played
isolated pawn a neutral move, and in response to 29.e4
Suppose Black has an isolated pawn set off favourable complications with
on dS. White sets up a formation with 29 ... .tgS! After 30.l::f.d l g6 3 l .�f3 dxe4
rooks on d4 (or d3) and d2, and queen or 3 l .Vi'e2 d4! he even wins. Stronger
on d l . In doing this, he has to be care is 3 0.f4, but after 30 ... d4!? 3 l . l::f.xd4
ful that his rooks do not come under (3 I . .txd4 i.c4) 3 I ...Vi'cs 32.�dl .tb3 !
attack from the enemy minor pieces. (but not 32 . . . i.f6 33 .eS) 33 .Vi'xb3
If this can be done, the pawn on dS .l:t.xd4 34.i.xd4 .l:f.xd4 3S . .l:f.xd4 �xd4+
will be finished. Even if Black man 365�>hl i.e7 there is little doubt that
ages to set his own major pieces up in Black will hold this endgame.
an analogous fashion, White can break It is another matter to find a 'neutral
with e3 -e4 (or c3-c4) and win a pawn, move' for Black, and it may be that
thanks to the pin. Spassky played 28 ... a6 (defending the
queen in any event) because he could
Anatoly Karpov not find anything better.
Boris Spassky 29.h3
Montreal 1 979 As Karpov admitted, seeing the
weakening of b6, he spent some time
wrestling with the temptation to
0
change his plan. There was the idea of
29.�d l ! ? then i.c3 -d4 and nd3 -b3 .
Even so, after resisting temptation, the
grandmaster decided not to depart from
his main plan of laying siege to the iso
lated pawn at dS.
29 ... 1Wc6 30. �h2 1Wb5 31 .f4 f6
The threat was 32 .fS , winning the
bishop. Black could have set a trap: 3 1 ...
22Jld3! .:r.cd6 23 . .:r.td1 .:r.sd7 aS 32.fS d4! but after 32.a4! he would be
24 . .:r.1 d2 1i'b5 25. 1i'd1 the one caught in the snare: 32...'ifxa4
«Remembering Alekhine's advice, that in such 33.f5 or 32...Vi'cs 33.i.d4 �c6 34.fS .
positions, the queen should stand behind the rooks» 32. 'it'd1 'ifc6 33.g4 g5
- Karpov. As in the previous example, a long
25 ... b6 26.g3 �fa 27. �g2 �e7 period of waiting is succeeded by a
28. 1i'h5 a6 feverish search for counterplay. In Kar
It seems both players convinced them pov's opinion, Black could have con
selves that the pin along the Sth rank tinued to show patience a little longer:
means that White was threatening 33 ... i.d6 34.�hl i.c7. If White then
e3 -e4. In any event, commenting on the continues as in the game, with 3S.fS
game in the tournament book, Karpov i.f7 36.e4, then after 36 ... i.f4 the
explains Black's last move as motivated 'counterplay' would fall into Black's
by a desire to defend the queen on bS . hands of its own accord.
However, in reality, the move e3 -e4 is 34. �h1 as 35.f5 �f7
49
Black has played a typical French

Defence exchange sacrifice. The assess
ment of the position depends to a con
siderable degree on whether he will
manage to carry out the advance ... e6-
eS, freeing his light-squared bishop and
exerting maximum pressure against f2 .
Topalov played the obvious prophylactic
move ...
25. l:.e1
The point is that he threatens to take
36.e4! on c4. It is worth seeing what hap
The fruit is ripe for picking. pens if White is careless and allows
36 .. .<ifo>g7 37.exd5 'ifc7 38.l:.e2 bS Black to break into f2 , in which case
An oversight, although in any event, the he gets mated, in several ways: 2S . . .
position is already very difficult. Thus, eS 26 .dxeS? 'iYxf2+ 2 7.'l!thl �h3 !
after 38 ... �d6 , White will happily 28 .gxh3 .l:!.xf3 29. �xf3 'iYxf3 +
give up the exchange: 39.�e6! �xe6 30.�gl tt:'lf2 , or 26 .tt:'lxeS tt:'lxeS
40.fxe6, developing in return an attack 27.dxeS tt:'lxf2 2 8 . �f3 tt:'lh3 + 29.<;t>h l
on the light squares, which is beyond ifgl + 3 0 Jhgl tt:'lf2#.
doubt decisive: 40 .. J:Ie7 4I.'iff3 �eS 25 ... tat6
42 .iffs �xc3 43 . �e4! etc. After long thought, Black decides to
39 . .:.Xe7! .:.Xe7 40.d6 'ilc4 41 .b3 go over to passive defence. Something
More precise than 41..�.d4 'ife2 . Black disturbed him in the variation 2S ... eS
resigned. 26. �xc4 tt:'lxf2 . Maybe he did not like
one of the zwischenzugs 27.!lb3 !? or
One can find other examples of 27.aS!?.
'Alekhine's Gun' in any textbook. We Grandmaster Jussupow made a difficult
will look here at a comparatively rare choice. He retreats his pieces onto the
(although in its way typical) example: last three ranks, as a sort of fortress. We
will see how Topalov sets about break
Veseli n Topalov ing this down. His general plan is to put
Artu r Jussupow his major pieces on the e-file and attack
Dortmund 1 997 the weak pawn on e6. And you already
know precisely how we will arrange his
0
rooks and queen.
26. �d1 .:f.f6 27 . .:f.a2 0.17 28. 'ifa3
Once a strong player has formed his
general plan, he does not forget to pay
attention to the fine details. Here is an
example: Topalov could play 28.�e3,
but preferred to set up the positional
threat a4-aS and so deflect the enemy
knight to aS.
so
-
28 ... tba5 29. lbe5 liJxe5 30. l:.xe5 M i khail G u revich

'if dB 31 . l:.e3 'it'f8 32. '1Vc1 Alexei Shirov
White can always exchange queens, Sarajevo 2000
and the battle will be long and difficult.
It is better to keep all the major pieces •
on the board. If Black holds, one can

always return to the idea of a queen
exchange later.
32 ... tlJb3 33. 'it'b1 tba5 34.h3 b6
35. Ag4 '1Vd6 36. 'ife1 tlJc6 37.h4
tlJd8 38.g3 tlJf7 39. l:.ae2 tlJd8
This is what we spoke about at the

beginning of this chapter. One pair of
rooks have already left the board and
the 'Alekhine's Gun' formation is not so
obligatory.
On the other hand, White threatens to
activate his battery on the long diagonal
by means of e5-e6. Black needs to put
something on e6, to blockade the pawn.
26 ...'iYe6 27 . .l::f.f5 g6 is hardly the correct
White has completed his plan and has all idea, because of 28. nf6. That means:
of his pieces pressing on the e6-pawn. 26 ... l:.e6! ?
But Black turns out to be prepared for The future looks rosy. The next move
this. will be ...'iYd7-e8, then taking on e5. At
Now a new stage in the plan begins, some point, it will be the turn of the
which does not really have any direct e4- and e2-pawns.
connection to our theme. White will But White finds some clever threats, to
advance his pawn to h5, and take on distract Black from his plan.
g6, creating a new weakness. Then, 27. l:.f3 'it'e8 28.l:.d3!
after exchanging a pair of rooks, White Of course, not 28. a:fs g6 29. .l:t.f6 , in
will direct his remaining major pieces reply to which, as well as 29 ... .l::[xf6
to the newly-opened h-file. Once the 30.exf6 'ifxe4+, Black can go directly
enemy forces have been tied up, the into a queen ending with an extra pawn:
white queen can suddenly transfer to 29 .. Jhe5! 30.i.xe5 i.xe5 3 1 .'1Vf3 i.xf6
a3 . 32.'ifxf6 'iYxe4+ etc. Now, however,
In the end, Black will not be able to after 28 ... i.xe5 29. i.xe5 .l::rxe5, he loses
maintain his defences on all fronts. He at once: 30.'1Vxe5! So he must make Iuft:
resigned on move 7 5 . 28 ... h6 29. '1Vd2 ""h7
Again it is too early for 29 ... i.xe5, this
Now we will look at several examples time because of 30.i.xe5 �xeS 3 1..l::f.d 8.
of a non-standard character. 30. -tcs .:.Xes 31 . l:.d7
51

Alexander Alekhine
Buenos Aires 1 927
Looking at the last diagram, who would

have thought that White would seize
the d-file and invade the seventh rank?
Evidently, here the plan of putting the The famous first game of the world
queen behind the rook has not justified championship match. The Cuban, who
itself. Or maybe the initial position was in those days lost barely once a year
just equal. Shirov did all he could, but amidst great sensation, had made a seri
White reacted accurately and managed ous oversight in an Exchange French and
to obtain counterplay. had had to surrender a pawn. It looked
What should Black do now? The cleanest like a sensation was in the making, but
reply is 3 l ...b6, but after 32.1::txc7 bxc5 then Capa started playing at his full
33 .bxc5 Ihe4 34.il'd3 the position is strength. At the point reflected in the
close to a draw. Shirov decides on some diagram, Black has problems convert
thing different. He changes the course ing his advantage.
of the game, and gives his opponent a Admittedly, at first glance, it is not obvi
chance to fight for an advantage. But in ous what the problem is. The advance
this case, Black will also get chances. ... d5-d4-d3 is tempting. The rook sup
31 ... .J:Id5! ? 32.exd5 1i'xd7 33. 1i'd3+ ports the passed pawn from behind, and
<itg8 the queen defends the king from checks.
The moment of truth. For equality, All the pieces are ideally placed, surely?
34.e4 is fully sufficient. This is what Alekhine wrote:
34.d6 «Black could go into the variation 30 ... d4 3 1 .h5
As anticipated, Gurevich could not il'fs 32.g4 il'bs 33.il'c2+ d3 34.il'xc7
restrain his ambitions. 'i¥d7 35.il'xd7 .l:Ud7 36 . .l::f.e l d2 37. l::td l
34 ... i.d8 35.h3 1i'e6 36.e4 b6 �h6 38.�g2 (38.£4 fS!) 38...�g5 39.�g3
37. i.f2 <itf8 38.g4 <i>e8 l::td 3+ 40.£3 f6 with a winning position. White,
Then the king came to d7 and ten however, has the opportunity to reply to 30 ... d4
moves later, the d6-pawn was lost, and with 3I.il'f3!.»
with it, the game. The great Russian methodologist Peter
Romanovsky quite justifiably described
In the following classical example, get queen and rook endgames as the 'fourth
ting the queen behind the rook is even stage of the game'. In terms of material
more difficult. on the board, we have an endgame,
52
but the character of the play is defi 35. l:td8 d4 36.a4 :e1 +
nitely more that of a middlegame. After Black wins even more quickly after
30 ... d4 3 l .'iff3 d3 32 .hS, Black faces a 36 ...'ife7 37. .l::rb 8 'iVc7 3 8 . .l::ra 8 'ifc6.
complicated task. In several variations, 37. citg2 1i'c6+ 38.f3 :e3 39. 1i'd1
White gets at the enemy king with the 1i'e6 40.g4 :e2+ 41 . cith3 11Ve3
manoeuvre 'iff3 -a8 , and in others, he 42. 'ifh1 1i'f4 43.h5 l:tf2 0-1
threatens a check on the b l -h7 diago
nal. Also the pawn on f7 hangs. In beginners' books, it is written that
At the time, a great debate arose about the the best way to get a rook into play is to
assessment of the Capablanca-Alekhine place it on a central file and then open
endgame, first in magazines and then that file. Most of the time, that is what
in books. The analyses replaced one happens. However, in other, not so rare
another. Reproducing all the variations cases, the rook comes into play not via
here would fill this book, and still with the centre, but from the side. And the
no guarantee of getting to the truth. issue here is not whether this is attrac
It is more important to get across to tive or not, but its effectiveness.
the reader Alekhine's viewpoint. He
deliberately avoided the plan involving Boris Spassky
the quick push of the d-pawn, in favour Tigran Petrosian
of a plan based on consolidating the Moscow 1 966
major pieces:
30 ... :e&! ? 31 .l:.a8 :es 32.:Xa7 c5
0
33J id7 11Ve6 34. 1i'd3+ g6
Two plans are tempting. The first is a

pawn storm on the kingside with g2-g4,
If we compare this diagram with the h2-h4 etc. The second is solid centrali
previous one, we may get the impres sation with .ifl -d3 (or .ifl -c4), l:f.hl
sion that Black has dawdled. He has lost e l , etc. The problem i s that both plans
his extra pawn and has not advanced are non-concrete and do not threaten
his passed pawn. All he has done over Black, whereas he, on the other hand,
the past five moves is arrange his major has a definite plan: ... �a8-b8 , ... c6-cS ,
pieces with the queen behind the rook. and then doubling rooks on the b-file.
But it is precisely here that Alekhine, 1 3.h4! !
in his own opinion, had secured the Believe me, this move is stronger than
win. And, indeed, the game did not last it looks.
much longer: 1 3 ... l:.ab8 1 4 . .:th3! c5 1 5. 11Vg5!
53
The secret is revealed. In the position 22 ... �8 23. ..tg2 .:ca 24. .1ba7
of the last diagram, in turns out that And White easily realised his material
White's interest was not the enemy king advantage.
on g8 , but the weak black queenside
pawns. That means that the move 1 3 .h4 Vlad i m i r Kra m n i k
was not a signal to attack, but the prepa Jan Timman
ration for a transition into the endgame. Wijk aan Zee 1 999
The rook is heading to h3 , then to c3 or
even a3 , and the queens will come off.
0
The position will become technical.
In this connection, the opinion has
been expressed that, instead of 14 ...
c5, Black should have defended with
14 .. J:tb6 1 5 .�g5 l:f.fb8. By attacking b2
he forces 1 6 .b3 - then the rook's path to
a3 is closed. Mind you, it is not entirely
clear why, instead of 1 6 .b3 , White
cannot play 1 6 . i.c4!?, and if 16 .. Jhb2 ,
then 1 7. i.b3 . White has an extra pawn, Black active
1 5 ... cxd4 pieces. Likely variations are 2 1 .0-0 l:Ic2
Alas, Black cannot keep the queens 22Jld1 d4; 2 1 .<;t>d1 i.a3 and 22 ... .l::!.c4;
on the board: 1 5 ...'ifxf2 16J:tf3 'iYg1 2 1 .e3 i.a3 22.'.t>e2 �c2 , in all cases
1 7. i.b5 , winning material. with counterplay.
1 6. 'ifxf6 tlJxf6 1 7 . .lbd4 It seems that the game should end in
It is already clear that the plan of 1 3 .h4 a draw, but Vladimir Kramnik finds a
and 14.l:th3 has proved fully justi way to cause his opponent problems:
fied. White was not just trying to be 21 .f4!? ..ta3 22 . .:f1 !
original for the sake of it, but played in On 2 1 .. . .l::tc4, the same move would
accordance with the requirements of follow, whilst 2 l ...l:f.c2 fails to 22.'\t>dl .
the position. 22 ... .:c4 23 . .:t3!
The battle after the opening seemed Returning the pawn, but bringing the
unclear, but now it is one-way traffic. rook into play, in a not quite traditional
Black must passively defend his pawn manner.
weaknesses: 23 ... ..txb4 24 . ..txb4 .lbb4 25 . .:S3
1 7 ... .:b7 1 8 . .:b3!
Another brilliant positional decision
(like 1 5 .'i¥g5 !). White leaves on the
board only those pieces which he needs
to attack Black's weaknesses. The other
pieces he exchanges without hesitation.
1 8 ... .lbb3 1 9.axb3 .:as 20 . .:c4
tbe8 21 . .:S4 ttJd6 22.g3!
And that is that. The transfer ofthe bishop
to the long diagonal ends the game.
54
25 ... h5!? Black to move. His pawn on c7 is hang

Black does not want to defend passively ing, and defending it from f7 looks like
with equal material: 25 ... l::f.b 7 26.�a5 the only way.
l:id7 27.�d2 , and White will win a 1 3 ... 85
pawn again, at the first opportunity. It is hard to believe that this zwischen
26. :Xa7 l:l.b2 27.a4 :S2 zug is anything more than a one-move
(in Kramnik's opinion, 27 ... g6 was trick. Surely Black does not really intend
stronger) to bring his rook out via a6? No, he is
28.f5! much more likely to bring his queen
But now the endgame turned out to be and bishop off the back rank and then
unpleasant. White won at move 41 . develop the rook to d8 or e8.
1 4.'ifa3 l:l.f7 1 5. %:1.d1 g5! ?
In general, this 'lateral' manoeuvre of Such aggressive pawn moves are typical
a rook in the endgame is well-known - of Topalov - and this one also clears the
one only has to recall the classic games 6th rank. For whom? The rook!
Capablanca-Kupchik (Havana 1 9 1 3), 1 6 . .te3 'ife8 1 7. ltJc3 tbxc3 1 8.bxc3
Alekhine-Spielmann (New York 1 927) f4 1 9.gxf4 gxf4 20 . .td4 .tg4 21 .0-0
and many others. Kramnik's version of
it only strikes one's eye because it was a
somewhat less obvious example.
Even so, such manoeuvres make rather
more of an impression in the middle
game, or even during the phase
between opening and middlegame.
When the board is full of pieces, the lat
eral development of the rook is usually
much more difficult to achieve, if not
impossible.
But chess is so remarkable because there 21 ... :S6! ?
is always an exception to every rule. Improbably, i t has happened. The rook
turns up on a6. It was in order to play
M i khail Gu revich this that Black refrained from the
Vesel i n Topalov tempting variation 2 l ....txe2 22 . .txd5
Sarajevo 2000 �h8 23 . .txf7 �xf7 (and then, if nec
essary, ... .l::ra 8-g8).
• 22.f3
A mating attack follows after 22 . .txd5
l::rg 6! 23 . .txf7+ �xf7 24.�h1 �xe2 .
22 ... .tf5 23. %:1.f2 l:l.g6 24. 'iii't 1 hd4
25. :Xd4 l:l.fg7 26. :Xd5 .th3!
Black can celebrate having implemented
the plan begun at move 1 3 , with his
rook travelling via a8-a6-g6, and now
taking a key part in the attack. In order
55
to avoid mate, White had to give up It seems that White has not achieved
material, but this did not delay the end much. The rook stands on the open file,
for long. but it cannot create any threats, without
assistance from his other major pieces.
Alexander Khalifman Now the manoeuvre tt:lgl -h3 -f4 is
Stellan B rynell tempting, but White finds something
Leningrad 1 989 rather stronger:
1 8 . .J:lb3! ? b4
Or 1 8 ...tt:lc7 1 9.a4 b4 20.c3 , and the
0
queenside gets opened up one way or
another.
1 9.a4 tDc7 20.c3! bxc3
Alas, Black proves not to be up to the
task facing him. He could maintain the
tension with 20 ... tt:le7! ? 2 l .cxb4 axb4,
and if 22 . .ixb4, then 22 ... l:f.xa4, whilst
if 22.l:txb4, then 22 ... b5! .
21 . .J:lxc3 l:i:Je7 2 2. .J:lac1
1 5.h4! As a result of his unstereotyped play,
At first glance, the idea of this move White has seized the only open file.
seems clear. White is thinking of Black cannot extinguish the activity of
opening the h-file with h4-h5. There the enemy rooks and his defence is dif
fore, Black played ficult. All of this is the direct result of
1 5 ... h6, the manoeuvre h2-h4 and l:thl -h3-c3-
so as to meet 1 6 .h5 with 16 ... g5 . But b3-c3 !
there followed
1 6 . .J:lh3! «The third rank is the ideal route for transferring
And it turns out that the advance of the rooks from one flank to the other. In general, the
h-pawn contained another idea - the rooks, like the minor pieces, should be brought
development of the rook via h3 , as in into play as soon as possible.» These are the
Spassky-Petrosian above. words of Anatoly Karpov; the 1 2th
1 6 ... tt:\a6 World Champion evidently got to the
Weak is 16 ...'ifxc2 1 7.l:f.c3 . very heart of the problem. The crite
1 7 . .J:lc3 'ifd7 rion for the successful development of
this or that piece is speed. If the rook
can develop quickly via the central files,
then this route should usually be pre
ferred, but if the path to the central files
is closed (a 'traffic jam'), then nothing
remains but to adopt the side-street.
More rarely, it is the second rank, rather

than the third, which is used for this
lateral development.
56
M ichael Adams the implementation of Black's plan:

Vlad i m i r Kra m n i k 2 I ... .l:te7 22 . ..id3 .l:tfe8 23 . ..id2 'i:Vd6.
Linares 1 999 But there follows 24.f4! and it turns out
that the main events take place not on
• the e-file, but on the neighbouring one .
Then play turns in White's favour.
Therefore Kramnik seeks his chances in
the endgame.
22. 'ffxe8 :Xe8 23 . ..td3 ..tg4!
Continuing to seek entry squares. The
threat is ... ..ig4-e2, and if 24.f3 , then
24 ... �h5 2 S . �d2 l:tae7 26.l:tael .l::txel
27. �xel l:txel + 28 . ..ixel ..if4, and the
queenside pawns are a great source of
Black to play. How should he develop? trouble.
Actually, this is a trick question, because 24 . ..td2 ..te2
if you look at the position, you will see The alternative is 24 ... l:f.ae7 2S.f3 �hS ,
that the bishops are already developed! returning to the position given in the
Yes, they stand on the eighth rank (and previous note. Maybe instead of 26.l::!.ael ,
the bishop on c8 has not made a single a stronger move is 26.l:!.fel l::!.xel+
move), but nonetheless, they are already 27. ..ixe l ! ..ig6 28 . ..ifl . Black stands
in play: they work along the diagonals freer, but there are no direct threats.
and control squares. 25 . .l:[fe1 ! Aae7 26 . ..tf5
The major pieces are the ones which White is close to the desired draw.
remain undeveloped. The plan is born
of its own accord: double the rooks on Lajos Portisch
the e-file and bring the queen to d6. Vladimir Bagi rov
Hence the first move: Beverwijk 1 965
1 9 ... Aa7! ? 20. 'fff3
A strong side of Adams' play is his
D
patient defence in slightly worse posi
tions. Now if the direct 20 .. J:te7 2 I . ..if4
..ixf4 22 .'i:Vxf4 .!::!.fe8 , then after 23 . ..id3 ,
the rooks on the open file do not have
a single penetration square. Nor is
20 .. .'iVd6 2 l .g3 , followed by ..icl -f4,
any help.
20 ... g5!?
Preventing the opponent's plan of
exchanging dark-squared bishops. It is obvious that the bishops at bl and
21 . 'ff h 5 'ffe 8! c1 are already in play. But the rook on
We are watching some unspectacu al is not. Therefore ...
lar, but high-quality play from both 1 6.Aa2!? :ea 1 7.Aae2 g6 18 . ..t82
players. It would seem nothing can stop :c1
57
Black also tries to manoeuvre his 23.tt::lxd7 l:!xe2 24.tt::lxf6+ and 25 .�xe2 ,
rooks, but rather less successfully than winning a piece.
his opponent. One may reply that the 21 ... 'WeB 22. 'iff3
last move frees c8 for the queen, from It is as if Portisch set himself the task
where she will defend e6 and attack c3. of not sacrificing anything in this game.
That may be so, but as for the square He could also have won with 22.tt::lxf7!
d5 , where a breakthrough is likely, cJi;xf7 (nothing is changed by 22...Ihc3
Black seems to have forgotten about that 23 .�d2) 23 .dxe6+ .txe6 24.l:!xe6.
square entirely. 22 ... tiJxd5
1 9. i.h6 22..Jhc3 23 .�f4.
It is easy to see that the blows on f7 23. 'ifxf7+ �h8 24. tiJxg6+!
and e6 do not yet work. For example, Not waiting for 24 ... hxg6 2 5 . l:le4, with
1 9:�:Jxf7 �xf7 20..!he6 .txe6 2 1 . l:!xe6 an inevitable mate, Black resigned.
cJ!;g?. Therefore the bishop goes to h6,
taking away squares close to the black The manoeuvre �al -a2-e2 (or
king. l::f.a 8-a7-e7) is often seen in positions
1 9 ... i.f8 arising from the Indian defences.
Black rejected 1 9 .. .'�c8 not because
of 20.tt::lxf7 �xf7 2 l..�xe6 .txe6 Hector Rossetto
22 . .txe6+ �xe6 23.l:!xe6 �xe6, when M i khail Tal
he has enormous compensation for the Portoroz 1 958
queen (although even in this variation,
there is the move 2 l .d5 ! ? with compli •
cations which could well be in White's

favour). It is more likely that Bagirov
was put off by the move 20.c4! after
which the battery of major pieces on
the c-file will soon have to return to
their former places of deployment.
20. i.g5 i.g7
The same 'traffic jam' we spoke about

earlier prevents the a8-rook from
coming to the centre via the eighth rank.
But the seventh rank is open, and allows
the rook to reach its desired spot on the
e-file. So there is no doubt what to play:
1 6 ... :S7!? 1 7. i.e3 :Se7
The rook is developed and e4 attacked.
But there remain problems with the
21 .d5! queenside pawns.
Now the game will soon reach its finale. 1 8. tbd2
Not 2 1 ...exd5 because of 22 . .txf6 .txf6 Tactics come to Black's aid:
58
1 8 ... b3!? 1 9.axb3 will recall, the 'Berlin Wall' became the
White wrongly decides to trade blows. main battleground for the players.
It was more sensible to retreat with And what happens in this variation?
1 9.�cl bxa2 (else 20.tbc3) 20Jha2, The fact is that, after the early queen
and if 20 ... tbb3 , then 2 1 .�xc4 tbxd2 exchange, the black king loses cas
22.�xd2 . tling rights and ends up blocking the
1 9 . . . lbxb3 eighth rank, interfering with the free
With tempo, so there is no time for the movement of his rooks. As a result, the
knight to retreat to c3. advance of the rooks' pawns is often the
20. :a4 way to bring the rooks into action.
Or 20.tbxb3 cxb3 2 1 .�xb3 tbxe4. Lovers of statistics might like to work
out how many times in this match
20 ... lbxd5! 21 .exd5 ..tfs 22. 'ifxc4
ttJxd2 23. 'ifxa6 tlJe4 Kramnik played the moves ... a7 -aS and
White has won a pawn, but in a rather ... h7-h5 . And he did not once lose the
less favourable fashion. Black has con game!
solidated. His further plans involve the 1 5 ... h5! 1 6. ttJd3 cs 1 7.c4
knight travelling via e4-c5-d3 , plus Nowadays, White is usually less keen to
pressure on the e-file and on the king play this move. With the blocked pawn
side. He has full compensation for the structure on the queenside, Black's
material. light-squared bishop assumes great
importance. He can aim to remove all
As we have seen, in the golden era of the rooks from the board (which often
chess, the players often used the lateral happens), and then put his bishop on
development of the rooks. And amongst fS , when the fixed white pawns will
contemporary players, Vladimir Kramnik provide a tasty feast.
often does so: his opening repertoire Kasparov presumably avoided 1 7.tbef4
allows the use of this technical device. tbxf4 1 8.tbxf4 because of the pawn sac
rifice 1 8 ... c4! ? although according to the
Garry Kasparov authoritative analyst Igor Zaitsev it is
Vlad i m i r Kra m n i k not so easy to show compensation after
London 2000 1 9.bxc4.
1 7 ... a5! 1 8.a4
• Necessary to stop ... a5 -a4 .
1 8 ... h4!
Now after 1 9.tbef4 tbxf4 20.tbxf4 �fS
2 l. . l:i.d2 �b7 22.�el �e8 White's ini
tiative is extinguished. Therefore, he
deliberately keeps more pieces on the
board, hoping later to break through
with his e- and f-pawns.
1 9.tbc3 ..te6
Weaker is 1 9... �f5?! 20.tbd5 �b7
This was the first game of the World 2 1 .tbe3 �e4 22.f4, followed by f4-f5 .
Championship match, in which, you 20.ttJd5 '1ti>b7 21 . tiJe3 l:[hS!
59
Developing and blockading at once. can grab hold of as one flies into the
22 . ..tc3 :ea 23. l:l.d2 'iti>c8 24.14 abyss.
tl:le7! 29 ... l:l.gb8!
The pawn breakthrough is finally ruled If Black is given eight moves in a row,
out and White has no other ideas. he will promote his a7-pawn. It may not
25. tlJf2 ttJf5 sound too convincing, but even such an
It is even a shame that such high-quality idea is better than none.
work should have been performed only The rook on a8 is another story - how
so as to make a draw (from 'a position it feels right now one had better not ask.
of weakness'. But, be that as it may, a 30. ttJxf4+ 'iti>f8 31 .tlJxg6+
draw it was. White probably thought he could win
how he pleased. If not, he would hardly
The rook is a piece that feels most have considered this move, uniting
comfortable on an open file or behind Black's two isolated pawns on f7 and h7.
a passed pawn. But there are some 31 ... hxg6 32 . ..td3
situations when the player rightly places In his notes, Alekhine showed that
his own rook in a dark corner. When? White could allow his opponent's idea
You are about to see one such example. and still win: 32.l:ld2 aS 3 3 . i.d3 a4
34. l:tbl .l:f.xbl 3S.i.xbl �b8 36.i.d3 a3
Jose Raul Capablanca 37.i.e2
Alexander Alekhine
New York 1 924
«with the threat of 38.l:td3, to tie the enemy rooks

to the need to defend the a3 -pawn, after which the
realisation of the advantage would not involve any
Black is in a bad way. His f4-pawn is difficulties.»
attacked and already indefensible. There This variation needs to be extended,
is no compensation at all for this mate however: 37 ... i.a4!? intending to meet
rial loss, and no counterplay. The game the direct 38 . .i:!.d3 with 38 ... .l:f.b2 39.l:txa3
is close to being over. i.dl 40.l:!.e3 l:lxa2, and White makes
At such a moment, the player needs an no progress at all. Instead of 38 . .i:!.d3 ,
idea of some sort. Possibly artificial, it is stronger to play 3 8 .�e3 , although
possibly capable of being refuted by here too, we can 'check White's papers':
accurate play from the opponent. But 38 ... .l::te 8+ (38 ... �e7 39 . .i:!.d3) 39.�d4
an idea is essential, like some projecting nbs 40.�c3 cS or 40 . .l::f.d 3 .!::i.b2 4I..l:I.xa3
rock on the edge of a cliff, which one .l::f.xe2 42 . .l:f.xa4 l:!xg2 .
60
32 ... l:lb2+ 33.l:le2 l:lab8 34. �e4 So the threat is to retreat the queen and
l:lxe2+ 35. <iPxe2 be4 36.fxe4 rj;e7 play l::rd l -d6!. Once we realise this, we
The position has changed beyond can understand the next move.
all recognition. Now we have a rook 22 ... l:lfb8!
ending, where Black is only one pawn Now after 23 .'fWa3 there is 23 ... l:!.b3 ,
down. Seven moves ago, he could not and after 23 .'fWc7 <JJg 7 24.l:!.d6, there is
even dream of this. 24 .. J::i.c 8! 2S . .l:lxe6 l:!.xc7. The initiative
Black still has to overcome some obsta is neutralised.
cles to make a draw, but it all began with Botvinnik claimed that 22 .. Jl.fb8
the seemingly strange move 29 ... l:!.gb8 ! . was the only move, but this is hard
to believe. Even the direct 22 ...'fWxd6
M ikha i l Botvin n i k 23 . .l::txd6 aS retains chances. But there
Salo Flohr is no doubt that the rook manoeuvre in
Leningrad 1 933 the text is the best move in the position.
23.1!fxe6 fxe6 24. l:ld4
• And White offered a draw, which was
accepted. Playing for a win is danger
ous: 24 . .l:i.d6 aS 2SJhc6 a4 etc.
Now let us move to modern times.
Alexey Shirov
Viswanathan Anand
Moscow 2001
A more complicated and, in its way, •
rather beautiful example. What is the

threat and how should Black play?
To get to grips with the position, let us
start with the tempting move 22 ... aS.
The passed pawn gets moving. In reply,
23 .'fWxe6 fxe6 24.l:td6 a4 2S . .l::f.xe6 is
weak, because of 2S ... a3 26 . .l::ra l .!:i.fb8
27.l:!.d6 a2 28 . .l:lddl Ilb2 . Despite his
extra material, White risks losing. Objectively, the pos1t1on is close to
The correct reaction is 23 .'fWa3 ! . The equality. But White's game is rather
threat is 24.l::l'.d6 - with queens on, easier to play. He can play his next few
White now really does threaten to take moves 'by hand', as they say: l:!.fl -dl ,
the pawns on the sixth rank, because <JJg l -f2-f3 , e3-e4. Having strengthened
the passed pawn is securely blockaded. his position, White can turn his atten
The move 23 ... .l::f.fd8 does not save Black tion to the weaknesses on a6 and f7.
from the enemy rook's depredations Black needs to come up with some
along the sixth rank, because of the thing. He does not stand worse, but
simple 24Jhd8+ �xd8 2S .'fWxaS . he will do, if he relaxes and plays rou-
61
tinely. For example, 23 .. J ifd8 24J Hdl Black transfers his knight to a5 or e5
�d5 (counting on 25 .e4 tLlxe4 26. �xd5 and doubles rooks on the c-file. The
exd5 27. I:rxd5 k:rc8 , although even here, pawn on c3 (and potentially that on a2)
White is slightly better) 25 .tLlb6 l:txd4 ties down White's forces and distracts
26. l:txd4 .!::i.b 8 27.tLlc4. The exchange his attention.
of rooks has clearly not been in Black's 1 7. l:tfb1 !? l:[fd8
favour. Now it will be much harder to Probably here too, Black should have
defend the pawn weaknesses. preferred 1 7... tLla5 , but he had not
23 ... l:tfb8!? 24.l:tfd1 a5 25.�2 c;ith7! divined his opponent's idea.
Played with the same intention: retain 1 8. lt:Jxe6!
ing all four rooks on the board. Now it turns out that Black cannot
26. l:t1 d2 a4 27. �3 l:ta7 recapture on e6 with the queen, because
The regrouping is complete. One rook - he loses the b7-pawn.
the one on a7 - defends the pawn weak 1 8 ...fxe6 1 9. l:td1 :Xd1 + 20. :Xd1
nesses, whilst the other exerts pressure l:tf8 21 . 'ire4
on b2 . The pawn on a4 hinders b2-b3 .
Black correctly felt the turning point
and did not play routinely, but thought
about things and found an original
plan. As a result, he kept the position
within the bounds of equality.
28.e4 l:tb4 29.ltJd6 l:tb8 30. ltJc4
l:tb4 31 . ltJd6 l:tb8 32. l:tc4 112-112
Vesel in Topalov
Mainhard Moroder The manoeuvre l;1f1 -bl -dl was in no
Frankfurt 1 997 sense a loss of tempo. Thanks to it,
Black has obtained a real new weakness
0
on e6. Now Black cannot, as he could
before, surround the pawns on c3 and
a2, since he is distracted by the need to
defend his own weakness on e6.
In the end, White won.
In his great book, Aaron Nimzowitsch,

in explaining one technical device,
employed the phrase 'a film in three
In this position, White is not the one parts'. We have been watching a film in
defending. He has a healthy extra pawn, four acts. Alekhine's desperate try, Flohr's
and needs to think about how to increase brilliant defence, Anand's unstereotyped
his advantage, not how to equalise. plan and Topalov's subtle manoeuvre.
Having said that, it is true that White Classics, whose value will never dimin
has his weaknesses. After the straight ish, and contemporary examples, which
forward 1 7.tLlxe6 'ifxe6 1 8 . �fdl b6 deserve never to be forgotten.
62
Chapter 5
Dances with the Knights
analysis, with the help of the computer,

Anatoly Karpov published some exceptionally beautiful
Boris Spassky variations on the internet. It turns out
Leningrad 1 974 that 3 1 ...tLlb8 loses by force to 32 .tLlxe5!
fxeS (or 32 .. .'iYc7 33. i.f4 gS 34 . .l:t.d7!
tLlxd7 3S .'�'xe7+ �g8 36.tLlxd7)
D
A masterpiece by the positional

genius Karpov. With low-key but very
analysis diagram
clever manoeuvres, he completely
disorientates the opposing army. The 33 . .l::f.d 8 ! ! and then:
next manoeuvre especially stands out. 33 .. Jhfl 34 . .l::rg 8+ �h7 3S ."iYxg6#;
24. lLlb1 ! 33 ... .l::f.xd8 34 . .l::rf7+ �h8 3S ."iYxg6;
The knight is heading for d2 and then 33 ... i.xd8 34. l:Ixf8 �xf8 3S.i.h6+
c4 or f3 . «The cunning of this move is that 'iYg7 36.'�'d6+ �f7 37. i.xg7 i.e?
White succeeds in choosing the most appropriate 38 ."iYc7;
moment to regroup this knight to an active posi 33 ... tLla6 34 . .l::rx a8 tLlcs 3S. l:iaxf8 ! tLlxe6
tion>> Karpov.
- 36J:1tf7# !
24 ... 'Wb7 25. �h2! 32 . .:Xd8 ..bd8
One wants to award every one ofWhite's Necessary, since after 32 ... l:!.xd8 there
moves with an exclamation mark. is the decisive 33 .tLlxe5! "iVc7 (33 .. .fxe5
25 ... �g7 26.c3 lLla6 27 . .:te2 .:tf8 34. I:i.f7+) 34."iVf7+ and 3S ."iVxe7! .
28. lLld2 .td8 29. lLlf3 f6 30 . .:td2 .te7 33 . .:td1 lLlb8 34 . .tc5 .:th8 35 . .:Xd8!
31 . 'ife6 .:tad8 The final blow. After the recapture there
In his notes to this game (written in the follows 36.i.e7, so Black resigned.
pre-computer era, of course), Karpov
shows that the only way to prolong Without exaggeration, an immortal
resistance was 3 I...li:Jb8. game. But the actual manoeuvre
Our contemporary, Vasily Lebedev, tLlc3-bl -d2 had been seen before. A
an amateur but a great master of long, long time before.
63
Emanuel Lasker variation 22 .. J:tae8 23.tt::ld2 e5 24.dxe5

Harry Nelson Pillsbury �xeS 25.tt::lf3 l:Ie3 26.tt::lg5 �xg3
Paris 1 900 27.hxg3 l:le7 28.tt::lx h7 'lt>g7 (instead
of 28 ... l:Ixh7??) it is unclear how great
0
White's achievements are. The extra
pawn is doubled and weak. Black also
has his trumps: the advance of the c- and
d-pawns and lively piece counterplay.
One gets the impression that with the
move 22.tt::lb l , White was concerning
himself with matters that are not of the
first priority. The main thing is to pre
vent the freeing ... e6-e5, which could
22. tbb1 ! ? :Sea be achieved by the manoeuvre 22.l:Ig5 !?
22 ... e5 is not possible, in view of 23 .dxe5 l:Iae8 2 3 .l::!.h 5 h6 (not 23 ... e5 24.dxe5
lhe5 24 . .l::U7 .l::te 7 and now 25.�xg8+!. �e5 ?? because of mate in one) 24. .l::te 5.
Therefore Black, before playing ... e6-e5 , A sample development of events is
overprotects the square e7. 24 .. Sfo)g7 2 5 .tt::le 2 tt::lf6 26.h3, and then
23.tt:x:l2 e5 24.dxe5 :XeS 25. tbf3 tt::le 2-f4. Here, White's advantage does
.J::le3 26. tbg5 :Xg3 27.hxg3 h6 not raise any doubts.
27... .l::te 7 2 8 .tt::lx h7 l:lxh7 29. i.xh7
�xh7 is weak, because of 30.l:if7+, As for the manoeuvre tt::lc 3 -bl -d2 itself,
winning the exchange. it has taken root as an established
2a. tbf7+ �g7 29. tt:x:l6 .J:le7 standard in positional play. It is often
29 ... �b8 30.lif7+ 'lt>h8 3 I .l:th7#. employed in positions arising from the
30.tbxb7 Indian structure, when the opponent
The long manoeuvre tt::lc 3-bl -d2-f3 -g5 - has a weak pawn on d6 and the knight,
f7-d6xb7 has reached its conclusion and having developed initially to c3 , strives
White has the advantage. Admittedly, in the course of the battle to reach c4.
he later loses part of his advantage:
30 ... tbf6 31 . tbc5 ..tg4 32 . .J::lf4 .tea Alexander Khalifman
33. :S4 tbg4 34 . .b6 Thomas Luther
It looks stronger to play 34.�d2 tt::l e 3 Leningrad 1 989
3 5 .tt::lb 3! and if 35 ... tt::lxg2 , then 36.tt::ld4
0
c5 37.tt::lc 6, continuing the galloping of
the restless steed. After
34 ... ..tf5 35 . .J::lf4 tbe3 36.c3 �g6
the battle flared up with renewed
vigour, although Lasker finally won at
move 60.
Let us return to the position in the dia
gram. Unlike the Karpov-Spassky exam
ple, here the move tt::lc 3-bl raises some
questions. Even in the most primitive 1 a. tbb1 !
64
Chapter 5 - Dances with the Knights
The tactical justification for this lies in threatens White. He can get on with his
the variation 1 8 ...ifxb2 19.ifxg5+ �h8 direct concerns on the queenside.
20.e5! ifxal 2 1 .ifh6 and mate is una 1 9.tt:::lb 1 ! .:Xc1 20 . .:Xc1 .:ca
voidable. Now not 2 l .ti:Jd2 (the rook hangs),
1 8 .. J 1b8 1 9. lbd2 f6 whilst 2 1..l:f.xc8 tt:Jxc8 22 .ti:Jd2 f4 is also
Black is unable to play 19 ... b5 because bad. What should White do?
of 20.ti:Jf3 . 21 . tt:::la3 ! .:Xc1 22 . .h:c1
20.b4! Petrosian shows subtlety in carrying out
Khalifman conducts the game superbly. his plan. If now 22 ... b5 , then 23 .ti:Jc2 ,
The tempting 20.ti:Jc4 allows Black to and then 23 ... a5? 24.ti:Ja3 b4 25 .ti:Jc4.
breathe easier after 20 ...iff4. Now, The knight reaches its desired square
however, to his other problems he has after all.
another added: the need to defend the 22 ... h5 23. tt:::lc4 tt:::lc8 24. �d2!
c5 -pawn. Black wants to exchange bishops,
20 ... 'ife7 which, if he can do it, would make
Not 20 ... cxb4 2 l .ifa7, whilst after 20 ... it significantly easier to defend his
b6 there is the very strong 2 1..l:tabl ! . queenside pawns. But Petrosian is alert:
21 .bxc5 dxc5 22.a5! �d7 23. tt:::lb3 24... ith6 2 5 . itb4.
.:ca 24 . .:ac1 24 ... �16 25. tt:::la 5 b6 26. tt:::lc6
The pawn on c5 can no longer be Demonstrating enviable technique,
defended. The pawn falls, and with it Petrosian won on move 49.
the game. Black soon resigned.
The manoeuvre by which the king's
The threat of the knight transfer to c4 knight took the place of its queenside
remained just a threat in this game. In colleague, by means of gl -f3-d2-e4-c3 ,
the following example, the manoeuvre is looks fanciful and artificial, but in reality,
carried to its logical completion. it contains a deep idea. Several opening
vananons, notably in the English
Tigran Petrosian Opening, involve White fighting for the
Borislav lvkov square d5, by exchanging off his queen's
Bugojno 1 982 knight and replacing it with the other.
Let us examine a concrete example.
D
Tigran Petrosian
Florin G heorghiu
Moscow 1 967
This game is remarkable for the fact
that White plays three very important
moves. And each of them is ti:Jd5 !
1 .c4 e5 2. tt:::lc3 tt:::lc6 3. tt:::lf3 tt:::lf6 4.g3
..tb4 s. �g2 o-o 6.o-o .:ea 7.d3 h6
8. tt:::ld 5 !?
The queens have been exchanged and Rubinstein's idea, which is usually a
the typical kingside attack no longer mark of quality.
65
What should Black do? His usual plan

of advancing the f-pawn does not work.
Note how badly the �e8 and .tf8
stand. The rook should be on f8 and
the bishop somewhere on the a7-g1
diagonal. Then the two pieces would
jointly pressurise f2 . In this position,
however, this is impossible, or, at
the very least, will take an enormous
amount of time.
White's plan may look slow. The knight
8 ... .tf8 9. tlJxf6+ on c3 is a real 'devourer of tempi ',
In the game Rubinstein-Duras (Carlsbad but looking at the position, one notes
1 9 1 1), there followed 9.b3 d6 1 0 . .tb2 with surprise that this is not the case.
ltJxdS 1 1 .cxd5 tbe7 1 2 .e4 cS 1 3 .dxc6 White's forces are harmoniously placed
tbxc6 14.d4 .tg4 1 S .d5 lt:Je7 16.'itd3 , and there is no impression that he has
and Black had not equalised. None lost tempi. Meanwhile, the black pieces
theless, Petrosian's move is more pre on e8 and f8 would like to exchange
cise and fits in better with the overall places.
plan. White exchanges knights and then 1 2 ... .td7
directs his remaining knight towards c3 . Black sticks to the plan of exchanging
9 ... 1i'xf6 1 0. ttJd2!? light-squared bishops by means of
White's advantage cannot be very ...'ifd8-c8 and ... �d7-h3 . This is a well
great, but he knows what to do with known idea, but Petrosian reacts subtly
it. The first priority is the manoeuvre and as a result, Black achieves little.
lt:Jd2-e4-c3 , taking firm control of the Practice shows that another plan was
squares e4 and dS. Then, once the situ more promising: setting up a barrier
ation in the centre is under his control, with ... c7-c6 and ... d6-d5 . For exam
he will advance the b-pawn, as is typi ple, 1 2 ... .te6 1 3 .b4 a6 14.!Ib1 'ifd7
cal of English Opening positions. Once 1 S .a4 .l:tab8 16.b5 lt:Je7! ? 1 7.'ii b 3 axbS
it reaches bS , it will drive off the knight 1 8 . axb5 .th3 1 9.e4 �xg2 20.�xg2
on c6, thus increasing the scope of the lt:Jg6 2 1 .'itd1 c6 (Kasimdzhanov-I.
bishop on g2 . Sokolov, Wijk aan Zee 1 999) or even
1 0 ... d6 1 1 . ttJe4 'ifd8 1 2.ltJc3 the immediate 1 2 ... tt:Je7! ? 1 3 .b4 c6
14.b5 dS!? (Grant-L.B.Hansen, Gausdal
1 992).
We will continue the discussion of this
plan below.
1 3.b4 'ifc8
Of course, not 1 3 ... tt:Jxb4 14 . .I;lb1 , and
the rook breaks through to the seventh
rank.
1 4 . .1:b1 .th3
66
A knight appears on dS for the second

time, but how the position has changed
in the meantime. White has made great
progress and Black faces a difficult
defence.
1 9 ... ttJd4
The alternative was 19 ... lt:Je7, but one
can understand Gheorghiu - he wants
to show some activity, even if it is with
just a single piece.
20. ti:le3
1 5.e4! It is clear that in the long term, White
These days, this is a standard reaction, wants to put his bishop on b2 (or e3)
but for Gheorghiu it must have come as and take on d4. But there are nuances.
an unpleasant surprise. Just as Black is For example, after 20 ... c6 2 l . �b2 'ife6 ,
ready to exchange light-squared bishops, i t i s unfavourable to play 22. �xd4 exd4
White changes the pawn structure in 23 .lt:Jf5 , because of 23 ... d5!?. The posi
the centre. The weakening of the square tion opens up and the bishop on g7 may
d4 does not play the slightest role; more come to life again.
than that, White tempts the black knight 20 ... f5 21 . .i.b2
to this square, so as to exchange it for Now, however, things are very unfa
the bishop and leave only one pair of vourable for Black after 2 l ...f4, because
minor pieces on the board. of 22. �xd4 exd4 23 .lt:Jf5 . The pawn
1 5 ... .bg2 structure remains closed, and in such
The experts criticised Black for this a position, the knight is significantly
decision. But after I S : .. �e6, White stronger than the bishop.
could open a second front: 1 6 . �e3 �e7 21 ...fxe4 22.dxe4 1We6 23 . ..tc3 b5!?
1 7.'ifd2 and then f2-f4-f5 . Evidently, the A logical attempt to break out, which,
exchange of bishops struck Gheorghiu however, White warmly welcomes.
as the lesser evil. 24.cxb5 1Wxa2
1 6. �xg2 g6 1 7.h4 ..tg7 Or 24 ... tt:Jxb5 2S .'ifd3 lt:Jxc3 26.'ifxc3
This move is unquestionably bad. It was 'ifxa2 27.'ifxc7.
necessary to play 1 7... h5. 25. 1i'd3 1We2 26. 1Wxe2 ttJxe2
1 8.h5! g5 1 9. ttJd5 27. ttJd5
67
The knight lands on dS for the third s . . . ..tcs 6.d3 h 6 7.a3 a s a . ..tg2 d6
time, and this time with decisive effect. 9.0-0 0-0 1 0. �f6+ 'it'xf6
27 ... .:ab8 28 . .1:.fe1 tlJd4 29 . .hd4
exd4
The exchange, which was planned some
ten moves ago, is finally executed, in
the most favourable circumstances pos
sible for White. Highly characteristic of
Petrosian's manner of play!
30 . .1:.bc1 .l:.b7 31 . �c7
The curtain can be drawn down. Black
survived to the end of his time-trouble
and then resigned at move 41 .
1 1 . tlJd2! ?
Theory moves forward, propelled by The position i s very similar to what we
many factors, including shock. The have seen, but there is a small differ
game Petrosian-Gheorghiu made a great ence. The bishop is not buried on f8 ,
impression on those interested in the but takes an active part in the central
opening. It became obvious that quiet, battle.
passive play in this variation would not In addition, unlike Gheorghiu, Epishin
bring equality, and attempts were made is not prepared to surrender the centre
to seek counterplay. without a fight.
After l .c4 eS 2 .tLlc3 tbc6 3 .tbf3 tbf6 1 1 ... 'it'd8 1 2.e3 tLle7!? 1 3. tl.Je4 b7
4.g3 �b4 S . �g2 0-0 6.0-0 .!::f.e 8 7.d3 1 4. tbc3 c6!
h6 8.tLldS !?, as Petrosian played, White Black sets up a barrier to the bishop on
retains the initiative for a long time. g2 . If White continues the usual English
Therefore instead of 7... h6 Black play Opening plan with b2-b4-bS , there
ers started playing 7... �xc3 8.bxc3 e4!?. follows the reply ... d6-dS . This central
The Rubinstein-Petrosian plan could be strategy could prove more effective
played one move earlier: 7.tLldS ! ?, but than the flank diversion.
Black can also accelerate his own plan Understanding this, Topalov reacts
with 6 ... e4!?. subtly. He is first to undertake opera
tions in the centre:
The argument 'for' and 'against' contin 1 5.d4 ..te6
ues to this day. But I want to look fur With this, a complicated position has
ther at a plan I mentioned above. been reached, with chances for both
sides.
Vesel i n Topalov
Vlad i m i r Epishin And who was the first to come up with
Dos Hermanas 1 994 the plan of ... tbc6-e7 and ... c7-c6 in the
1 .c4 eS 2.tbc3 tl.Jf6 3.tl.Jf3 tbc6 4.g3 English?
..tb4 S.tlJdS!? It is hard to say, but as a possible stem
As we see, the Rubinstein idea even game, we offer the following classical
works at move S . example:
68
Erich Eliskases of b5 , whilst, in turn, maintaining his

Grigory Levenfish own blockade of d5 .
Moscow 1 936 20. tl:ld2! tl:lf6 21 . tl:lxf6+ 'ifxf6
22.tbe4 'ife7 23. tl:lc3 l:lb8 24. tl:ld5
• The knight manoeuvre f3 -d2-e4-c3
(and then to d5) is completed. Can we
call it cumbersome? Can we accuse
White of having wasted tempi? Not, not
at all. The manoeuvre has fulfilled its
task - it has broken the blockade of b5
and maintained the one on d5 .
24 ... 'ifd7 25. 'ifd3 i.b7
Black wants to exchange the blockading
1 2 ... tbe7!? 1 3. ltJQ3 c6 1 4.d4 exd4 knight on d5 for his own bishop. This
1 5 . .bd4 tl:lg6 1 6.13 d5! is a questionable decision, since now all
And Black seized the initiative. the light squares fall into White's hands.
Tempting is 25 ... .td8 , and then ... tLle6-
Let us return to the plan with the knight c7. However, even in this case, things
transferring from f3 to c3. In the open are not easy for Black. The pawn b4
ing, as they say, it looks a bit much, but advances sharply, whilst the d6-pawn is
in the middlegame, it is a different story. going nowhere.
If the structure takes on a closed char 26. l:ld1 l:lfd8 27. i.d2 i.xdS
acter, and the pawn masses cannot start 28 . .bd5 'ifbS 29. 'ifb3 'ife8
advancing quickly, the loss of tempi is 30. l:ldc1 i.b6 31 .l:.c6 tl:lc7 32. l:lbc1
unlikely to be felt. In such cases, even tLlxdS 33. 'ifxd5 'ifd7 34.b5 �cs
the most long-winded manoeuvres may An oversight in a difficult position,
turn out to be justified. typical for tournament play, or a simul,
as here. It was possible to continue the
Vesel i n Topalov game by means of 34 ...'iVb7, and if
Claus Henrici 35 . .tb4, only then 35 ... .tc5 .
Frankfurt 1 997 35. b5 l:ldc8 36. l:l1 xc5!
Black resigned.
Incidentally, this is not the only exam
ple of creative play by Veselin Topalov
in simultaneous displays which you
will find in this book. There is a general
view that top players in simuls do not
bother themselves with calculating long
variations, or the search for unusual
positional plans, but try to play more
simply. Topalov seems to be an excep
The b4-pawn is passed and the d6-pawn tion to this rule. He plays all games with
backward. The essence of the position is exceptional concentration and energy,
whether White can relieve the blockade regardless of circumstances.
69
The manoeuvre 4::lf3 -g5 -h3 (or c 6 1 8 .�h1 �b6 1 9. l:tae1 .l::tf7 20.4::le 6
... f6-g4-h6) looks ridiculous at first (Karpov-Gelfand, Dos Hermanas 1994).
sight. Why waste time transferring the 1 0. tbg5!?
knight to the edge of the board? This idea is understandable. White pre
In reality, the explanation is simple. In vents the immediate 1 0 .. .f5 (in reply to
many King's Indian-type positions, the which there is the unpleasant 1 1 .4::le 6
knight needs to move, so as to free the i1Lxe6 1 2 .dxe6 c6 1 3 .�b3) , and forces
f-pawn to advance. The key thing is ... h7-h6, hoping that later on, either
often to create a pawn mass on e4 and the h6-pawn will become weak, or
f4, or e5/f5 for Black. And the knight the weakening of the bl-h7 diagonal
just needs to get out of the way for the will have its say. These hopes are not
moment; his time will come. unfounded, I would say.
There is also another idea. In some 1 0 ... h6 1 1 .ttJh3 f5 1 2.f4
cases, the knight on f6 should move to We have reached a standard situation
fS quickly. So quickly, that the loss of for this opening variation. The two
several tempi plays no great role. Then sides' pawn masses have clashed in
the square h6 is only a temporary one. the centre. Black must decide which
As usual, we will examine several exam change of structure to go in for. There
ples, classical and half-forgotten. are many possibilities. He can take an
isolated pawn on e5, or on fS , he can
Robert H ubner take hanging pawns on e5 and fS . Or
M iguel Cuellar he can play some kind of neutral move,
Leningrad 1 973 transferring to his opponent the right to
take the decision.
The latter is in fact what Black did.
0
1 2 ... ttJf6
I am tempted to suggest 1 2 .. .fxe4
1 3 .4::lxe4 ttJfs , and if 14.fxe5 , then 14 ...
dxe5. Then the plan is as follows: one
knight goes to d4, the second to f6, the
bishop to fS . The queen can transfer via
e8 and emerge on g6 (after ... g6-g5), or
it can come to b6 (after ... c7-c6).
1 3.fxe5 dxe5 1 4.ttJf2 g5 1 5. tZJd3
A well-known variation of the King's
Indian has been played. Nowadays, it
is seen more and more rarely, as it has
been shown that by 1 0 .4::l e 1 fS 1 1 .4::ld 3
4::lf6 1 2..�.gS !? White obtains the better
game. For example, 12 ... h6 1 3 . i1Lxf6
i1Lxf6 14.f4 (a paradox: having disposed
of his dark-squared bishop, White
opens the long diagonal) 14 ... exf4
1 5 .4::lx f4 ilLeS 1 6.exf5 iiLxfS 1 7.�d2
70
Thus the knight has arrived at d3, although But that is not the main thing. The main
it took four moves instead of two. point is that there is no real need for the
The pawn on eS is hanging. 1 5 ... lt::Jg6 is manoeuvre lt::Jf3 -g5 -h3 -f2-d3 . Objec
bad because of 1 6.exf5 . But it is possible tively, it is artificial, and if there is a
to try to solve the problems tactically: chance to post the knight on d3 in two
15 .. .f4 1 6 .lt::Jxe5 fxg3 1 7.hxg3 'i!Vd6 or moves instead of four, then of course
1 5 ... c6!? 1 6 .lt::Jxe5 lt::Jxe4. that is how one should play.
However, Black continued prosaically:
1 5 ... ttJxe4 1 6. ttJxe4 fxe4 1 7. he4 Boris Spassky
Without bothering himself with such Bruno Parma
more subtle ideas as 1 7.lt::Jf2 or the Havana 1 966
zwischenzug 1 7.l:txf8+, White just
recaptures with the bishop on the
0
blockading square e4.
1 7 ... �f5
A committal and inappropriate move. A
tempting alternative was 1 7... lt::Jf5 , with
the idea of bringing the knight to d6 or
d4. The bishop on c8 can always find
itself something to do.
1 8. 1i'e2 ..be4 1 9. 1i'xe4 :Xf1 +
20 . ..t>xf1 it'f8+
Missing perhaps his last chance to We have a structure very similar to the
complicate the game: 20 ... lt::Jc 8 !? 2 l .c5 preceding example. But there is a signif
(2 1 .lt::Jxe5 'i!Ve7) 2 1 ...lt::Je 7. icant difference. The white pawn stands
21 . ..tg2 'iffS 22.tbf2 'ifxe4+ on c2 , not c4 (the opening was a Pirc),
23. ttJxe4 ttJfS 24.b3 ttJd6 25. ttJxd6 which means that White needs to play
cxd6 26. b3 l:d8 27 . .:tf1 in the centre and on the kingside, rather
Black has done much to bring about this than the queenside, where he has de
endgame, which is actually prospect ferred his pawn advance.
less and difficult for him, as one would Secondly, the black bishop has come
expect. Later, White created a passed to g4. He wants to take on f3 and then
pawn on the queenside and won easily. advance .. .f7-f5 . It is better for White to
retain his knight than his light-squared
Summing up, we can say that in this bishop. Therefore, the choice is between
structure, the idea of tt.'lf3 -g5 -h3 (and l l .lt::J e l and l l .lt::Jgs .
then lt::Jh 3 -f2-d3) is interesting, but After l l .tt.'lel �xe2 12.'i!Vxe2 fS 13.f4,
no more than that. With this knight in the game Geller-Kuzmin (Lviv 1978),
manoeuvre, White posed his oppo Black replied with the non-standard
nent definite problems, which he was 13 ... exf4 14.�xf4 �xc3 !? 1 5 .bxc3 fxe4
not able to solve. But it did involve a lot 16.'i!Vxe4 lt::Jc s and obtained a very com
of tempi and with energetic, inventive fortable endgame: 1 7.'i!Vc4 'i!Vd7 18.lt::Jd3
play, Black could have obtained coun 'i!Va4 19.'i:Yxa4 lt::Jxa4 20.c4 Ilae8 etc.
terplay. Instead of 1 3 .f4, it was necessary to look
71
at 13 .f3 f4 14 . ..tf2. There are no light Weak is 1 7. . .e4 1 8 .tLld4. Black can main
squared bishops on the board, a factor tain the tension with his pawns on e5
which significantly hampers Black in and fS for at most 3-4 moves. White can
advancing ... g6-g5-g4. On the other hand, organise his forces, keeping the pawns
according to the canons of the King's under attack, and trying to force ... e5 -e4.
Indian, the f2-square should be occupied White's advantage is certainly not great.
by a knight, not a bishop. In addition, it But his plan is clear. Each piece knows
will not be easy to arrange the plan with its job. Black, in order to maintain
the advance of the b- and c-pawns. equality, needs to play very accurately,
Perhaps the strongest reaction to 1 2 ... and show some original ideas. This
fS is 1 3 .exf5 , and if 1 3 ... lLlxf5 , then grandmaster Parma was unable to do:
14.tLle4, or 13 ... gxf5 14.f4! . But the 1 7 ... ttJe4 1 8. 'iVc2 'iVe8 1 9. ttJf2 tlJxf2
paradox is that in this structure, the 20 . .:Xf2 e4 21 . it.d4!
knight stands better on h3 than on e l ! Black would have been happy with the
1 1 .ttJg5!? be2 1 2. tlJxe2 h6 position after 2 1 .tLld4 ..txd4 22 . ..txd4
Not 12 .. .f5 1 3 .tLle6. .l:Ig8. Spassky plays more subtly: by
1 3. ttJh3 c;i;>h7 exchanging bishops, he creates the
Again, it is too early for 13 .. .f5 - the threat of tLle2xd4-e6.
h6-pawn is hanging. Now it is clear 21 ... c5 22.dxc6 bd4 23. tlJxd4
why White recaptured on e2 with the Now the position is finally clarified.
knight at move 1 2 , and not the queen. The rooks will double on the d-file,
1 4.c4 and the queen come to b4 or a3. In this
In the game Timman-Seirawan (Las way, all three major pieces will attack
Palmas 1981) White calmly played 14.b3 the d6-pawn. And the knight will come
and gradually outplayed his opponent: to the blockading square e3.
14 .. .f5 1 5 .exf5 tLlxfS (15 ...gxf5 1 6 .f4!) 23 ... tlJxc6 24. 'it'c3 'it'e7 25. ttJc2!
1 6.f3 lLlf6 1 7.tLlg3 lLlxe3 1 8 .'ifxe3 c6 .:Sd8 26 . .1:1.fd2 'it'e6 27. ttJe3
1 9.c4 cxdS 20.cxd5 'iVaS 2 l .'iVd2 'iVxd2 White's advantage has grown to deci
22Jhd2 l:!.ac8 23 .tLlf2 .l:!f7 24.tLlfe4. sive proportions. Black played a few
1 4 ... f5 1 5.f3 ttJf6 1 6.exf5 gxf5 more moves and then resigned.
After 16 ... lLlxf5 , one can follow Timman's
example and play 1 7.tLlg3 . Or 1 7.tLlc3. Attila G roszpeter
1 7.f4! Andras Adorjan
H ungary 1 983
The position is starting to clarify.
72
When this game was played, the dia A) 1 5 .'iYd1 lLlf5 1 6.lLlxf5 .ixf5
gram position was already well-known (Nenashev-Barsov, Tashkent 1 988);
to theory and was not regarded as con B) 1 5 ."iYc2 .if6 1 6 . �e3 lLlg4
taining any particularly lively ideas. 1 7. �d2 .id4 1 8 .0-0 lLlxh2! 1 9.<>t>xh2
Both sides would just complete their 'fi'h4+ 20.�g1 'ifxg3 (Semkov
development, without any conflict, Dzindzichashvili, Saint John 1 988) ;
and the result would be a quiet posi C) 1 5 . �xc5 dxc5 1 6 . 0 - 0 a 6 1 7. �f3
tion, with some advantage to White. For �d6 1 8 . �fe1 f5! (Shulman-Nikolenko,
example, 1 0 ... d6 1 1 .h3 lLlbd7 1 2 .lLlg3 Moscow 1 995) .
lLle5 1 3 . �e2 �f8 14.0-0 etc. (Keene 1 2.g4
Kraidman, Gausdal 1 983). This looks very principled, but is prob
1 0 ... ltJg4!? ably just weak; after 1 2 ... lLlxg4 1 3 . l:tg1
This move is also not new, but Adorjan's d6 , no compensation for the pawn is
interpretation of it is. Previously, the visible. Adorjan, however, is after mate,
move had been part of a plan to bring not material.
the bishop to f6 as soon as possible: 1 2 ... d6 1 3.h3
1 l . �d4 d6 1 2 .lLlg3 �f6+. But White's
space advantage remains and becomes a �< ':J. ·
permanent factor: 1 3 . �e2 lLld7 14.0-0 ,l,j i .... 11 .t. • .t.
�xd4 1 5 ."iYxd4 lLlgf6 1 6 .I:i.fe1 lLle5 .. :.J�:�;�:
·'·. ..� .
./
·, . ,.;��:��j: · ;�,
,.
1 7.�ad1 (Kir. Georgiev-P.Littlewood,
Plovdiv 1 983). ',��/2 ' il�
�: :�� ·[:�{
1 1 . ..td4 tLlh6!
But this is the real novelty. In a posi
¥j � . ·.. . . 3:.'kf' �
tion where the pieces tended to move
almost as if half-asleep, Adorjan finds l:f'� ;, c�i�:Gl� �
a non-standard and very nice idea. The 1 3 ... f5!
knight is heading for f5 ! The author of the famous book Black
Of course, the idea was subjected to is OK! demonstrates that his ideas are
widespread tests. It was shown that correct.
both 1 2 .d6 �f6 13.lLld5 "iYaS+ (Yrjola 1 4. 'ii'c2 ..tg5!
Hjartarson, Gausdal 1 984) and 1 2.lLlg3 Preventing queenside castling. Weaker
.ixa3+ 13.�e2 .ixb2 (Ree-Ligterink, is 14 . . .fxg4 1 5 .hxg4 (but not 1 5 .0-0-0
Wijk aan Zee 1 984) are insufficient. The �g5+ 16.'.t>b1 �f5) 1 5 ... �xg4 16.0-
best move is considered to be 12.'fi'd3 !? 0-0 .ig5+ 1 7.f4.
d6 (12 ...lLla6!?) 13.lLlg3, with the idea of 1 5 . ..tg2 ttJd7 1 6.gxf5 tLlf6!? 1 7.h4
taking the square f5 from the knight, hf5 1 8. 'ii'a4 ..td7 1 9. 'ii' b4 ..tf4
although even here, Black has simple, Queenside castling is impossible, and
comfortable play: 13 ...lLld7 (not 13 ... �f6+ kingside castling is suicide: 20.0-0 lLlh5
14 . .ie2 �xd4 1 5.'iYxd4 lLlf5 16.'fi'f4 tbh4 etc. Black is fully mobilised and ready
1 7.0-0 lLlg6 18.'iYb4, and the knight has for the storm.
wasted too much time, Kir. Georgiev-Qi The outcome of the game is decided,
Jingxuan, Thessaloniki 1984) 14.�e2 but we will not deprive you of the
lLlc5. Then White has tried: pleasure of seeing the end:
73
20. <M1 .te5!? 21 . .he5 :XeS a pawn on cS is not solid. If Black can
22. 'it'xd6 tt:Jfg4 23. �g1 tt:Jf5 24. 'it'b4 kick away the bishop, the cS -pawn will
tt:Jxh4 25. tt:\g3 'iff6 26. ltJce4 tt:\13+ fall. So how can we do this? If 1 S .. .''l!!V b 8,
27 . .hf3 'ifxf3 28. 'ifxb7 l:[f8 then 16.tbeS. On 1S .. .'iVe7, again there
29. 'it'xd7 tt:Jxf2! 30. l:[h2 tt:Jxe4 can follow 16.tbes , with the possi
31 . tt:Jxe4 :Xe4 32. 'ifd6 l:[g4+ 0-1 ble variation 16 ... tbxeS 1 7.i.xeS tbd7
1 8 . i.d6, and White is out of danger.
It did not take long to find a classical From the viewpoint of the battle for
analogue. ... e6-eS, the move 1 S ...tbg4 is very
good. The knight frees the square f6
Gyula Breyer for its pawn. And the knight itself is not
Siegbert Tarrasch without work - it can come via h6 to
Gothenburg 1 920 f7 or fS . What will happen after that
remains unclear for the moment.
• 1 6.h3 tt:Jh6
We have reached the critical moment
of the game. Black has shown his cards
and White must now do the same.
A tempting move is 1 7.tbes, and if
17 ... tbxeS 1 8 . i.xeS f6 1 9.i.h2 eS, then
20.i.d3 , followed by castling. Black's
centre looks very nice, of course, but it
is not mobile and his minor pieces are
1 5 ... tt:\g4!? hardly dazzling, to put it mildly. White,
«A very subtle, deeply thought-out plan. Black after completing his mobilisation, can
will organise an attack by queen, bishop and both opt for e3 -e4 (and at some point f2-f4),
knights, on the c5-pawn, and then win it. For or can concentrate wholly on the queen
this purpose, the knight on f6 needs to come to side, where he has a pawn majority.
e6. White can put up temporary resistance to this, Maybe Breyer did not like the reply
but cannot prevent the plan eventually being imple 1 7... tbb8!?. But even in this case, after
mented» - Kmoch. 1 8 .bxc6 i.xc6 1 9.tbxc6 tbxc6 20.i.bS
I will permit myself to express some 'i:Vc7 2 1 .0-0, White's position is, at a
cautious reservations about whether minimum, not worse.
the move 1 5 . ..tbg4 was really the start 1 7. tt:Jb3 f6! 1 8. 'it'a3 e5 1 9 . ..tc3 'it'c7
of such a far-reaching and deep plan 20 . ..tb2 l:[ec8 21 . 'it'a2 'it' dB
as suggested by Hans Kmoch. In the White's piece manoeuvres at this stage
old days, they liked very long-winded create a rather pitiful impression. Effec
plans, but even so, the transfer of the tively, they are running on the spot. Black,
knight to e6, on a board full of pieces, sensing his opponent's uncertainty, set
seems a bit too much. up his forces against the cS-pawn (he
Furthermore, why should Black be con transferred his bishop to e7, the queen to
cerned with such deep plans? It is clear f8 and the knight came round via h6-f7-
that a battle is going on for the advance d8-e6), took the pawn and then won the
... e6-eS . The construct with i.d4 and game in beautiful style.
74
Chapter 6
The Shuttle Manoeuvre

The shuttle manoeuvre is a frequent guest in chess problems and studies. It occurs
much less often in practical play, but always creates quite an impression when it
does.
What does it involve?
Imagine that in the opening, a piece has been developed to some square or other.
In the course of the subsequent play, it becomes clear that the piece is not active on
that square and belongs on another. Then it returns to its original square, and then
takes up a new position. In commentary, such a manoeuvre is often described as a
'shuttle manoeuvre' .
From the description, it is clear that the 'shuttler' is more often than not a piece
that moves diagonally, i.e. a bishop or queen. Let us now familiarise ourselves with
some examples.
Standard for our day, but revolution
Richard Reti ary at the time. Of course, the bishop
Vlad i m i r Vu kovic belongs on the a3 -f8 diagonal. There, it
Vienna 1 922 has no pawns in its way and no oppo
site number, and can develop its full
strength.
D
1 8 ... g6 1 9. 'ife1 ! 'lt>g7 20. 'ifh4 lbg8
21 .b3
The preparations are complete and the
break f4-f5 hangs over Black's head as a
deadly threat.
21 ... ..td7 22.g4
A good alternative was 22 .'iYg3 !?. It
seems as though the break cannot be
achieved without the aid of the g-pawn,
How should we assess this position? All but the concrete variations show the
of Black's pieces are bad, especially his opposite:
king. On the other hand, the position, as 22 ... tt:Jc6 23.f5! exfS 24.e6! fxe6 25 .tt:Jh4
is characteristic of the French Defence, gS 26.i.xc6 i.xc6 27.'�We5+ tt:Jf6
is closed, and this hides the weaknesses 28 .tt:Jxf5+! exfS 29 . .!:r.xf5 and then
of the black position. I;ibl -fl with a crush;
White's plan is clear: to break with 22 ... �h7 (so as to avoid the pin on
f4-f5 . But this is not possible at the the g-file, but now the break, surpris
moment, and 1 8 .g4 fails to 1 8 ... h5. This ingly, is in place) 23 .f5! exfS (23 ... gxf5
means that the breakthrough requires 24. i.f8 !) 24.i.f8! (creating the threat
some preparation. of 25 .tt:Jg5+! hxgS 26.'iVh3+ with mate)
1 8 . ..tc1 ! 24 ...tt:Jf6 2 5 .exf6 J::rxf8 26.'iVd6 �e8
75
27 . .l::i.fel , and the black position is fall loses: 30 ...'iVh5 3 I .tLlf2 �xf2 32 .'ifxf2
ing apart. tLle4 3 3 . �g4! 'ifxg4 34.'iff7+ Wh6
22 ... h5 23.f5! exf5 35 .'iVf8+ �g6 36Jifl etc. However,
Not 23 ... hxg4 24.f6+, and this means 30 ... .l:tf7 3 I .'iVxg5 'ifxg5 32.�xg5 tLle4
the game is approaching its end. (attacking g5 and c3) keeps the game in
24 . ..bd7 an unstable balance.
Mathematically, the most accurate (as 30.a4!?
well as the prettiest) win was 24.'ifd8 ! In Reti's footsteps. Interestingly, despite
tLlh6 (the threat was 2 5 . �f8+ �h7 the result of the game, Ponomariov was
26.tLlg5#) 25 .'iVf6+ �g8 26.tLlg5 �xb5 later sceptical about the idea of a3 -a4.
27. �f8 ! �xf8 28 .tLle6+ �e8 29.'iVd8# This is what he wrote:
or 28 ... �g8 29.'iVg7#. But Reti, as an «During the game, this move seemed natural to
experienced practical player, prefers to me: in one go, White activates his rook and bishop.
win material: But the minuses of the move are also obvious: a
24 ... hxg4 25. �f8+ �xf8 26. 'ifxh8 pawn is given away, and Black gets a passed pawn,
Black soon resigned. which could assume great power in the endgame.
30.h3!? looks good, just improving the position a
Ruslan Ponomariov tiny bit.»
Vasily l vanchuk 30 ... bxa4 31 . �c1 ! l:f7 32. b3
Linares 2002 'ifxc2?
Black wrongly forces matters. He should
wait with 32 ... tLlf5 . Note that even if
0
White gets his rook to b7, it will not be
so easy to make any further progress.
The black pieces all defend one another.
After 32 ... tLlf5 , Ponomariov extends the
variation a move and a half: 33 .'ife2
'iVg7 34. �c5, and writes that 'the
whole battle lies ahead'.
33. l:c1 'it'f5
Black's queen and knight are attacked,
A more complicated, and hard to assess and this means that the text move, like
example. If we forget the previous ex those which follow, is necessary.
ample, it is easy to think that the en
tire battle revolves around the four files
from e to h. However, there is no de
cisive blow available in this area of the
board:
30.Ir.fl flxfl+ 3 l . �xfl tLle4 or 3 l .�xfl
'iVf5+ 32.tLlf2 tLlf7, and the pawn on
g5 , having exchanged defenders, is still
alive;
30.�h3 , and if Black now goes in for
an exchange of blows, then he probably
76
Chapter 6 - The Shuttle Manoeuvre
34. j.h3! tt:le4 Mate in one is threatened, and the bishop

Or 34 ... tt:lb5 35 .tt:le5 ifxh3 36.tt:lxf7 cannot be taken: 4l...�el+ 42 .�f2 .l:hcl
�g6 37.�fl . 43 . .l::i.g7+ �h8 44.�g8+ 'it>h7 45.l:t2g7#.
35. tt:le5 ._.f2+ 36 . ...xf2 l'b:f2 Therefore, Black resigned.
37. ltJxd7
White has won a piece and realised his The next two examples are separated
advantage without any undue adven by roughly 1 0 0 years. The World
tures. Champions involved have the num
bers 1 and 0 in their numerical order
Vasily Smyslov as champions, since Boris Spassky was
Isaak Boleslavsky the tenth World Champion and William
Leningrad/Moscow 1 941 Steinitz the first.
Boris Spassky
D
Jan Hein Donner
Leiden 1 970
The end of a famous game. White has

already won the exchange, but what is
more important is that the black king is
in trouble.
37. :eg2 t2Jd7 38. l:g7+ �h8 Black's position looks perfectly tenable.
39 . .1:[7g6 �h7 40. b3! He only has one weakness, on e6, but
In the two previous examples, the Black can defend this as many times as
bishop went from the c l -h6 diagonal White can attack it.
to the a3 -f8 diagonal. This is a mirror But Spassky finds an extremely unpleasant
image. plan. He returns the bishop to its starting
Interestingly, in his 'Best Games' book, position, so as to transfer it to the a2-g8
Smyslov does not mention the dual diagonal. Once that happens, White will
40. �2g3 exd4 4l..l::f.h 3 . The surprise is be threatening the advance f4-f5.
less, when one remembers that Vasily 21 .j.f1 !
Vasilievich named his book In Search of Black has exactly two tempi, in which
Harmony. to prevent a catastrophe. With hind
Harmony in chess means the best possi sight, we can say that Donner coped
ble coordination of all the pieces on the very unfortunately with this task.
board. Here, that means coordinating The move 2 1 ...a5!? deserved considera
the out-of-play bishop on c5 . tion. The move looks like a simple trap:
40 ... exd4 41 . j.c1 after 22..�.c4 b5 23.�b3 there follows
77
23 ... a4, and the bishop is trapped. White follows a miraculous escape: 2 9. . .Wh8
would probably have replied 22.a3, 30.e8'iV+ 'iVf8 ! and Black is close to a
re-establishing the original threat, but draw.
as will become clear from the varia Unfortunately, this fantasy meets a
tions below, the inclusion of the moves cold shower. Instead of 28.h5 , White
... a7-a5 and a2-a3 would very much suit wins by 28.'iVxa7. But now imagine
Black. .. the moves ... a7-a5 and c2-c3 had been
21 ... l:td6 22. �c4 l:tad8 played - White is still better, but the
Good or bad, he had to play 22 .. .f5 , resource 'iVf3 -a3 is no longer available.
although it i s a significant concession. 24... :Xd1 + 25.:Xd1 :Xd1 + 26.'ifxd1
Spassky would probably have first The attempt to improvise a solution
strengthened his central pawn with fails, as one would expect. Now after
23 .c3 , and then transferred his rooks to 26 ... tt::lx e6, White wins with 27.'iVd6,
the e-file, combining pressure against and on 26 ...'iVe7 with 27.'iVd8 ! . There
e6 with a favourable change of pawn fore, Black resigned.
structure on the kingside, by means
of h4-h5, g3 -g4 etc. The realisation of Louis Paulsen
the resulting large positional advantage William Steinitz
would take some time, but would be Baden-Baden 1 870
highly likely. However, now everything
ends rapidly. •
I suspect that Donner, not only a grand

master but also a world-famous jour
nalist and master of paradoxical writ
ing, when playing his last two moves,
had seen a variation which promised
to relieve him of all his problems. Alas,
though, he has miscalculated.
23.f5 :Xd4 24.fxe6
Black has the so-called 'advantage of

the two bishops'. Bishops need space,
and by all the canons of positional play,
Black needs to open the position. The
first line which comes to mind is 23 ...
gxf3 24.gxf3 e4 2 5 .dxe4 i.xc3 . How
ever, continuing the variation, 26.exd5
l:lhe8 27.tt::le4 i.f5 28.dxc6, we can see
that Black has not achieved anything. He
This is perhaps what Black had seen: is unable to get an advantage from the
24 ...'iVe7 25 .'iVa3 ! Ilxdl + (but not various pins and White is out of danger.
25 . ..t2:Jg6 26.!ixd4 �xd4 27.h5) 26.nxdl Another is idea is the opposite - to
l:txdl+ 27.�xdl tt::lg 6 28 .h5 'iVxa3 do nothing concrete, but just make a
29.e7+. It looks all over, but now there move that is useful on general grounds:
78
Chapter 6 - The Shuttle Manoeuvre
23 ... �b8 (and then, perhaps, ... �b8-

a8 or ... a7-a6 and ... �b8-a7). It is not
so easy for White to find a reply which
does not spoil something.
Steinitz decides differently. He finds
a concrete plan which does not strike
one's eye at first glance.
23 ... ..tf8!?
The bishop returns to its original
square, on its way to d6. There, he
will not be under the feet of his rooks, Black has everything defended and has
which can quickly transfer to the open two bishops, a serious strength, and the
f- and g-files. The object of the attack white pawns are now very weak. White
will be the knight on g3 and the weak lost on time three moves later.
kingside pawns.
24. �a1 Let us return to the game Reti-Vukovic.
Bad is 24.'ii'xeS �xa3 . The bishop was not the only piece to
24 ... ..td6 25. lbf1 l:[df8 26. lbd2 l:[h6 make a shuttle manoeuvre in this game.
The transfer of the bishop to d6 was very
good. But his next move was already
0
sub-optimal, and his last move out of
place. The rook is loose on h6 and his
king has not left the danger zone. These
factors form the basis for counterplay.
27.c4! ..te6 28. lbb3
This indecisiveness is hard to under
stand. By means of 28.cxdS cxdS 29.d4!
White could have turned matters in
his favour, for example 29 ...'ii'xd4 (29 ...
exd4 30.tt:lb3) 30.tt:le4 'ii' b 6 3 1 .'ii'd2 «<t is obvious that White's success depends on main
l:thh8 32.�xdS etc. taining his wide pawn front. The move I I .tt:lf3
28 ... gxf3 29.gxf3 ..tc7 is tempting, but then there follows I I ...'ii' b 6
White has created the threat of c4-cS , 1 2 . �e3 'i¥b2!, and White's centre is destroyed.»
so Steinitz moves the bishop away from 1 1 . 'ifd1 !?
the fork. But it seems that he could «A very subtle move, characteristic of Reti 's fresh
have fallen into the trap: 29 ... l::tg 6 30.cS and far from routine play. By defending his centre
�xeS 3 I .tt:lxcS �xeS 32.'ii'xeS I:!.fS once and for all, White lays down a solid founda
(not immediately 32 . . Jhf3 33 .'iVxhS) tion for his later attack» Kmoch.
-
33.'ii'e2 �f4, and if 34.'ii'e S 'iVd4 However, I wish to dispute the concrete variations
3S.'ii'x hS , then 3S ...'ii'g 7! and the white offered by the Viennese master and journalist. After
queen turns from the hunter to the l l .tt:lf3 'ii'b 6, it is worth considering 1 2 .�f2!?.
hunted one. The pawn cannot be taken: 12 ... cxd4 13.cxd4
30. lbd2 l:[hf6 31 . l:[c1 �b8 32.cxd5 tt:lxd4 14. �e3, whilst entering enemy territory
cxd5 33.l:[hg1 ..td6 34. l:[g5 ..tf7 is even worse: 12 ...'ii' b2 1 3 . �d3 c4 14.�bl .
79
So it seems that the shuttle manoeuvre by the white In the game Petrosian-Barcza (Budapest
queen is not the only move. 1955), after 9 .. .'iYb6 1 o.'ifd2 tt::la6 1 Ltt::lf3
On the other hand, there is no arguing with 0-0-0 1 2 .a3 White got nice play and
Kmoch's description of Reti 's play (and that of eventually won. Theory gradually came
other great players of the period) as 'fresh and far to the conclusion that instead of 8 ... gxf6,
from routine'. Thanks to such games, the shuttle the stronger recapture was 8 ...'iYxf6,
manoeuvre has become standard. It is hard to exchanging the white queen off.
imagine many lines of the French Defence without There is an interesting psychological
it, as well as various other openings. aspect here, which deserves discussion.
Thus, in the Queen's Gambit Declined, there is the For the typical amateur, the shuttle
well-known variation manoeuvre is associated with an admis
1 .d4 dS 2.c4 e6 3.tlJc3 tLlf6 4.cxd5 sion of having made a mistake or, at the
exdS s . ..tgs very least, the loss of a tempo. How could
and now not 5 ... i.e7, but Black allows it be otherwise, since we are returning
the doubling of the pawns with: a developed piece to its original square?
S ... c6 6.e3 ..tfS 7. 'iVf3 ..tg6 8. hf6 Master Reti played 'ii!Vd 1-g4, and then
gxf6 'ii!Vg4-d1 - does this mean he lost time?
The compensation is not clear. The Not at all. With the move to g4, he
bishop pair? Free development? Maybe, attacked the weak pawn on g7, inducing
but Black still has to work to equalise. Black to defend it with his king from f8.
For his part, White has to decide how to The black player had to lose a mass of
continue his development. And here the tempi, to castle by hand. So it transpires
main attention has gone to the shuttle that White did not lose time, he gained it!
manoeuvre with the queen: The manoeuvre 'ifd1 -f3 -d1 in the QGD
takes up two tempi, but in return,
White gains a permanent advantage: the
pawns on the black kingside are hope
lessly damaged. Is this advantage worth
two tempi? The manoeuvre i.c1 -e3-
cl-a3 in the French Defence costs tempi,
but it permits us first to fix a favourable
pawn structure and then to direct the
bishop to where its long-range power is
the equivalent of at least a rook. Is this
9. 'iVd1 ! ? activity worth the tempi used?
«The Moor has done his work, and can return One could multiply the examples,
home.» The queen's job on f3 was to but they all show one thing. In chess,
force Black to double his pawns; she general considerations and specific
has nothing more to do on that square, variations form an elusive whole. The
and it is time to free the square for the essence of the game is balancing the
knight, whilst the queen herself goes to general and specific, and doing this
pastures new and seeks fresh tasks. shows the class of the player.
80
Chapter 7
Positional Sacrifices ( Part One)

The pawn sacrifice is the smallest sacrifice possible in a chess game.
A minor piece (not to mention a rook or queen) we are only prepared to sacri
fice for clear gain, but a pawn is another matter. Not for nothing does the lexicon
of chess equate the word 'sacrifice' with 'discard' . 'I discarded this pawn' , as if
brushing off a speck of dust or throwing away an old pair of trousers.
For what do we sacrifice a pawn? For whatever we want. For an open line. To
shut an enemy piece out of the game (maybe only temporarily) . For the initiative,
which in turn we may convert into something more real. For a favourable change
in the pawn structure. To reach a draw ending quickly. Etc.
Of course, one should not underestimate the importance of material in chess,
even a pawn. An incorrect pawn sacrifice can cost the game. But it is no coinci
dence that the number of pawns sacrificed is many times greater than the number
of pieces. One does not throw away pawns left, right and centre, but nor does one
stand on ceremony.
lished on e4, will cement the black posi
Vlad i m i r Kra m n i k tion. But after 22 .Vi'd6 �ac8 23 .Vi'xc7
Veselin Topalov �xc7 24.tLld6 tLle4 25 . .l:f.d3 one has to
Dortmund 1 999 admit that the initiative remains with
White.
1 8.b5 0-0 1 9.b6 'it'd8 20. lbb3 tbf6
21 .tba5 l:.b8
1 6.f5!?
«A very risky and not obligatory decision, but I
wanted to sharpen up the battle» Kramnik.
-
1 6 ... exf5 1 7 .exdS cxdS Here we can draw preliminary con

After 1 7...tt:Jxd5 Kramnik gives 1 8.tLlc4 clusions about White's pawn sacrifice.
tLlxc3 (or 1 8 ... 0-0 19.tLlxd5 cxd5 The bishop on g6 is shut out of play,
20.tLld6) 1 9.�xc3 0-0 20.d5 «with a strong which is one thing. White has obtained
initiative». This variation can be contin a numerical superiority on the queen
ued: 20 ... cxd5 2 I.Vi'xd5 tLlf6 - and it side, whilst Black's kingside majority
looks as though the knight, once estab- is immobile and can hardly become a
81
threat in the near future - point two. 24. tt::lb4 .tea

And thirdly, White has a very danger Black has managed to solve one prob
ous plan: to sacrifice the knight on b7, lem: he has brought his bishop into
take on a6 and then have his passed play. But his pawn mass on the kingside
pawns sweep all before them. is still rooted to the spot and his posi
For all this, the position remains tion remains passive. It only needs 2-3
unclear. Black does have an extra pawn, moves, before he can start moving the
after all. If Black manages to 'untan great colossus on the f- and g-files, so
gle', White will have cause to regret his it follows that White has 2-3 tempi to
choice. Everything will depend on the continue manoeuvering, in preparation
next few moves. for the sacrifice on b7.
22.a4 tbe4 25. 'iVc2!
A very important moment. It is tempt The logic of this move (which Kramnik
ing to implement the plan at once: himself described as 'key') is hard to
23 .tt:Jxb7!? �xb7 24.a5. But at this very understand without the grandmaster's
moment, Black unveils his own trumps commentary. White rejected 25 .tt:Jxb7!?
- an attack on the king. For exam .t'txb7 26. �xa6 because of the the
ple: 24 ... lLJc6 25 .'iVa4 tt:Jxc3 26 . .!:ixc3 matic 26 ... l:i.xb6! 27.cxb6 'ifxb6. More
'iVf6 27.�xa6 �e7 28.l::!.d l (28.l::!.d3 dangerous is 26.a5, but then 26 ... tt:Jc6
f4) 28 ... �h5 29.�d2 .l::te 4 (not at once 27.tt:Jxc6 �xc6 2 8 . �xa6 kle7. «It is very
29... .l::f.e l+ 30.iL.fl) 30 . .l::f.cd3 .l::f.e l+ hard to eliminate the enemy bishop from c6»,
3 1 .�f2 'iVh4+ with mate. Instead of writes Kramnik, and himself suggests
24 ... lLJc6, Kramnik analyses 24.. .f4!? a method: put a rook on bl and the
25.�xa6 .t'tb8 26.tt:Jxe4 dxe4 27.c6 tt:Jxc6 bishop on b5. But after 29.�bl there
28 . .l:r.xc6 f3 and admits that the resulting follows 29 ... lLJc3 with a fork. This is
position is not to his taste. White has why White first puts his queen on c2 .
too many pieces on the queenside, and 25 ... g6
too few defending his king. 25 ... g5 !? is more active.
The piece sacrifice should not be hur 26. :b1 :f7
ried, but requires further preparation. The preparations are complete. It is time
First of all, Kramnik transfers his knight to act.
from c3 to b4.
23. tlJa2 f6
It was worth considering 23 .. .f4!?. This
is also a typical device, incidentally -
returning the extra pawn to activate his
pieces. It is important that after 24 . .l:r.xf4
lLJf5 25.lLJxb7 �xb7 26.�xa6, Black
can solve his problems with 26 .. J lxb6!
27.cxb6 'iVxb6, with a probable draw.
On 23 ... f4 24. �xf4 lLJf5 , Kramnik
had prepared the reply 25 .'iVd3 'iYg5 27. ttJxb7!
26. �cfl �fe8 27. �dl , «and the position is In a sense, this move is the continuation
hard to judge». of the plan begun as far back as 1 6.f5!?.
82
Chapter 7 - Positional Sacrifices (Part One)
It is hard to believe, but this is so. By broken and weak, whilst White's
his pawn sacrifice at move 1 6 , White knights have a blockading square on f4
obtained a pawn majority on the queen and the open e-file is available for his
side, and then the advance b4-b5 -b6 rooks.
created a potential passed pawn. And It is hard to suggest a sensible plan for
realising this pawn majority was only Black. The doubled pawns on the f-file
possible with the aid of the sacrifice of make one think of using the outposts
the knight on b7. on e4 and g4, but how can we get the
27 ... :Xb7 28.a5 tbc6 29. tbxc6 knights there? The e7-knight is defend
..bc6 30 . ..ba6 l:b8 31 . �b5! ing fS and the other knight defending
This is what White was aiming at, when f6. It seems they will be tied up for a
he put his queen on c2 and his rook on long time.
bl . But despite this, Black found and exe
Black's blockade collapses. The passed cuted a plan to get his knights to the
pawns promote. After outpost. He transfers a knight via g8-h6-
31 ... 'WeB 32 . ..bc6 'ifxc6 33.a6 f7-g5 -e4!
the outcome of the game was decided. 1 5 ... �b8 1 6. <iti'b1 �c8 1 7.lbf4 'ifc6
Topalov resigned at move 40. 1 8.g3 lbh6 1 9.l:he1 lbf7!? 20. l:e2
l:g8 21 . l:de1
Vlasti m i l Jansa
Raymond Allen Weinstein
Helsinki 1 961
White plays too academically. Perhaps

he only noticed Black's aim too late.
If it were Black's move, he would play One more move (22 .'iVfl tbe4), and the
... tbc6-e7-f5 , plugging the holes in his position will be equalised.
position. Realising this, Jansa tries to sharpen the
1 3.15!? game, but this attempt should not have
«The positional pawn sacrifice should b e i n the been crowned with success.
arsenal of every chess player!» - Jansa. 21 ... tbg5! ? 22 . ..bg5 fxg5 23. tbe6
I would add that inserting the moves g4 24. tbxd8
1 3 . .!:Iel .l::i.e 8 is in Black's favour, as he By continuing now with 24 ...'iVe8 , Black
can then meet 14.f5 with 14 ... e5. could have emerged intact in all varia
1 3 ... exf5 14 . .i.d3 tbce7 1 5. tbe2 tions. On 2S .'iVfl or 2S .'iVh4 there fol
The changes in the structure have lows 25 ... gxf3 , nor is anything changed
favoured White. Black's pawns are by the intermediate 2S . .l:ixe7 .Jl.xe7.
83
24 ... 'iVd7 feld. This opened people's eyes, clearly

A nightmarish mistake. Perhaps Black demonstrating the plan, with the help
simply mixed up the moves, by com of which Black shows the correctness of
parison with the variation 24.'ifxh5 his pawn sacrifice.
gxf3 25.tt::lxd8 �d7.
Raymond Allen Weinstein was a very Siegbert Tarrasch
talented, developing young player, who Alexander Alekhine
was one of the chess hopes of America. Bad Pistyan 1 922
Doctors soon diagnosed a psychologi 1 .d4 tt:\f6 2.tt:\f3 e6 3.c4 c5 4.d5
cal imbalance and later he was arrested b5! ? 5.dxe6
(among other things, for attacking a Nowadays White almost exclusively
chess master, Johan Barendregt) . The plays s . .igs .
story ended with his incarceration in 5 ...fxe6 6.cxb5 d5
a mental hospital, from which he has «Black has a strong pawn centre, which will be
never emerged. supported by bishops from b7 and d6; in addition,
25. ltJe5 the f-file is opened for his rooks. These positional
White emerges with an extra exchange advantages are sufficient compensation for the sac
and Black soon resigned. rificed pawn» Kotov.
-
7.e3 ..td6 8.ttJc3 0-0 9 . ..te2 ..tb7

A special place in opening theory is 1 0.b3 tt:\bd7 1 1 . ..tb2 'iVe7 1 2.0-0
occupied by variations which involve The king will also not be able to rest
a positional pawn sacrifice. Here we peacefully on the queenside. For exam
are not talking about such lines as the ple, 1 2 .�c2 e5 1 3 .0-0-0 d4 14.exd4
King's Gambit (where sharp combina cxd4 1 5 . .ic4+ �h8 1 6 .tt::l e 2 .ie4
tive possibilities supersede positional 1 7.�d2 tt::lb 6 with a strong attack
considerations) or the Queen's Gambit, (Postojev-Harikrishna, Mainz 2006).
where the sacrificed pawn is regained 1 2 ... .:Sd8 1 3. 1t'c2
almost at once. No, here we are con
cerned with lines where the pawn sac
rifice is real, but the compensation for
the material consists of positional fac
tors.
One example is the Benko Gambit (l .d4
tt::lf6 2.c4 c5 3 .d5 b5 !?) and the related
Blumenfeld Gambit (l .d4 tt::lf6 2.c4 e6
3 .tt::lf3 c5 4.d5 b5). It is not immedi
ately obvious what Black gets for the
pawn, but it is clear he will not regain
it any time soon. Even so, theory con Both sides have completed their devel
siders the Benko to be sound, whilst opment and united their rooks. It might
the Blumenfeld is much more often appear that White has a healthy extra
declined than accepted! pawn.
The following game was an impor 1 3 ... e5! 1 4. l:[fe1 e4 1 5.tt:\d2 ltJe5
tant one for the theory of the Blumen- 1 6. tt:\d1 tt:\fg4
84
Subsequent generations of players have his kingside, but now his centre col
copied Alekhine's plan without hesita lapses.
tion. Here is a typical example: 14 . .l::tae1 21 ... d4 22 . .ic1
(instead of 14. �fe1) 14 ... e4 1 5 .lDd2 After 22 .exd4 Black wins with both
lt::le 5 1 6 .f4 exf3 1 7.lt::lxf3 lt::lfg4, and it 22 ...cxd4, and the more forcing 22 ... e3
is hard to know what to advise White 2 3 . Iig1 (or 23.lt::lxe3 lt::lx e3 24.fxe3 �g3)
(Sinadinovic-Sahovic, Nis 1 98 1). 23 ... lt::lg3+! 24.fxg3 �xg3 with mating
17 . .bg4 tbxg4 1 8. t!Jf1 threats. Not only is the h2-knight hang
The knights, pinned to the back rank, ing, so is the pawn on h3 .
defend the squares f2 and h2. 22 ... d3 23. Yi'c4+ <it>h8 24 . .ib2
But what protects g2? t!Jg3+!
1 8 ... 'ilg5! Not 2 5 .fxg3 �xg3 , attacking the knight
and rook. And this means that it is
time to draw the curtain. But Tarrasch
played
25. <it>g1
allowing the knight into e2, and
continued the hopeless resistance until
move 40.
Jeroen Pi ket
Veselin Topa lov
Madrid 1 997
The decisive strengthening of the attack.
All that remains is to bring the knight •
round via g4-h6-f5-h4 and White will

be in terrible trouble.
Speaking honestly, his position is
hopeless. Black can quietly regroup,
strengthen his position, and his oppo
nent has nothing with which to oppose
this. One rarely sees White, after fewer
than 20 moves, fall into such a passive,
hopeless position. Yet one cannot really
criticise any one of his moves. Perhaps This game began with a Benko Gambit.
he really should not have taken the In the course of the battle, the structure
pawn on move 5?! took on characteristics of the Blumen
1 9.h3 t!Jh6 20. <it>h1 t!Jf5 21 .t!Jh2 feld. Probably Piket, a cultured player,
With the intention after 2 1 ...lLlh4 of knew of the famous predecessor game.
replying 22 . .l::[g 1 , covering the vulner But he could not do anything.
able point. But the position has long 1 6 ... e5!
since come to resemble a leaky boat - By comparison with the game Tarrasch
you repair one leak and another springs Alekhine, here we can identify several
up. White somehow manages to cover nuances, which are in White's favour.
85
In particular, he has a strong knight on The real issue is more concrete fac
b5 , his rook can operate along the third tors. White has two bishops, the better
rank and his a4-pawn is both extra and pawn structure (fewer 'islands'), and his
passed. opponent has no counterplay. Almost
But even so, the game did not last long any exchange brings White closer to
and was extremely one-sided. There is the endgame, in which all of the above
nothing to say. Black, with the appropri factors will have greater significance.
ate amendments, carried out Alekhine's Almost any exchange; but not every one!
plan and White did nothing to place 20 ... �d6 21 .g3 c5!? 22. bf6 gxf6
this classical plan in doubt. 23. 'ifxf6
1 7.h3 <ith8 1 8.l:l.c1 e4 1 9. tl:\h2
d4 20.exd4 cxd4 21 . 'ifa2 l:l.ad8
22.a5 d3 23. �g4 tl:\d5 24.a6 tlJcb4
25. 'ifb3 tl:\xa6 26. tlJc7 tl:\axc7
27. 'ifxb7 l:l.b8 0-1
Now let us look at some examples of

pawn sacrifices to achieve a draw.
Viswanathan Anand is one of the great
defenders of modern chess, maybe the
greatest of all. Many young players
could learn a lot from his games, and The position has been transformed.
memorise many defensive devices, Black has sacrificed a pawn, but White
which they can use in their own games. already no longer has the bishop pair.
But who taught Anand himself? Furthermore, exchanges are no longer
in his favour, because the remaining
Alexey Shirov minor pieces on the board are opposite
Viswanatha n Anand coloured bishops, which means
Linares 2002 exchanges will increase the chances of
a draw.
• As far as the extra pawn is concerned,
formally this is the pawn on b3 . It is
unlikely to become passed. Black has
only one problem: his exposed king.
This is where White's winning chances
lie. His plan is clear: firstly, not to
exchange queens, and secondly to try
to get a rook to the g-file.
23 ... �e7 24. 'ifh6
Or 24.�c3 l::!xdl 25 Jhdl .l:f.d8 26J�el
Black's position is unpleasant, and not .l:f.d4, and Black holds the balance -
merely because of the threat of 2 I . i.e5 analysis by Sakaev.
i.d6 22 . .l::f.xd6! .l:1xd6 2 3 .c5. This tacti 24 ... 'ife5 25. l:l.de1 'iff6 26. 'ifh5 �f8
cal threat is not hard to meet. 27. l:l.e4 l:l.d4 28. l:l.xd4 'ifxd4 29.l:l.e1
86
Possibly greater chances were offered Compare the position after move 29 in
by 29. .l::f.d l , and if 29 ...'iVf6 3oJ:rd7 lld8 , the game Shirov-Anand. Isn't it remark
then 3 1..l:!.a 7, eyeing up the pawns on aS ably similar?
and f7. Had Anand seen this note? Who knows?
29 ... 'flf6 30 . .I:I.e4 -'.g7 31 . .1:1.14 1i'e7 I am sure he would have seen the game
There is still a lot of play left, but in the Botvinnik-Rabinovich, and I would
end, Black held the draw. like to think that he saw the notes also,
A pawn sacrifice, forcing a position with studied them, analysed them, found the
opposite-coloured bishops, followed by mistakes, and remembered everything.
putting the pawns on aS and cS , sup At the necessary moment, his memory
ported by the dark-squared bishop, produced the recommendation and the
stopping White creating a passed pawn Indian GM utilised it.
on the queenside - did Anand think I remember one incident. In Baku, in
this up himself at the board? the republic chess centre, a meeting
It is very possible that he did. He is a was organised between young players
player of the very highest class. But let us and the leading Azeri player of those
examine the following position. days, Elmar Magerramov, who had just
become a GM. I hung on his every word.
M i khail Botvi n n i k One piece of advice he gave, which I
l lya Rabi novich particularly remember, was 'When you
Leningrad 1 934 study the games of great players, ana
lyse not just the moves played, but also
the variations and notes. Analyse them
D
- and play them over on the board,
without fail! '
One can read something about this
in relation to Nimzowitsch also. He
advised that when studying annotated
games, one should use two chess sets,
one to play the main moves and the
other for the notes. A very sensible
proposal. If you do everything on one
In an early edition of his best games board, then, especially if the notes are
collection, Botvinnik wrote: <<...on very detailed, it is easy to get confused.
26J:td3 With two sets, that will not happen.
Black replies 26 ... i.e4 27. i.xg7 i.xg7
28.'iVxe4 .!::!.dB followed by ... a7-a5! and ensures But let us return to our subject: the
the draw, despite White's extra pawn.» pawn sacrifice to make a draw. It is rare,
In a purely chess sense, this note is not but sometimes one even sees this done
correct. White cannot play 27. i.xg7 where the sacrifice is already the second
because of 27 .. .f5 ! . So it is no surprise pawn. Why would one voluntarily go
that this note is not present in later edi into a position two pawns down?
tions of Botvinnik's notes. But that is It turns out to be possible. For exam
not the main thing. ple, certain rook endings (especially
87
where the defender's king and rook are White can easily overcome the cut
active) have drawing tendencies. And ting-off of his king: 74.c;t>el and 7S .c;t>fl ,
why suffer in a position a pawn down, but then what? If he takes the king to
when one could give up a second pawn f6, then Black checks on fl , driving
to force favourable exchanges and get to the king away from the g7-pawn, and
a known draw? then returns the rook to gl . And there
is no point in taking the king towards
Vi ktor Kortchnoi the bS -pawn, because Black defends it
Anatoly Karpov with his rook from gS . Let us try one
Moscow 1 974 tactical nuance: 74.a4!? .l:lgl+ (it seems
74 ... b4 75.a5 l:tg6 76.a6 .l::rd6+! 77.�e2
• l:Id7 78.�e3 �a3 is also sufficient for
a draw) 75 .�d2 �g2+ 76.�d3 .!:lg3+
77.�e4 bxa4 78J:i:b8+ �c2 79.g8�
l:Ixg8 80Jhg8 a3 - again a draw.
Kortchnoi plays differently, but also
fails to achieve anything:
68. �b2 l:f4 69. 'it.>c2 :ta 70. 'it.>b2
l:f2+ 71 . �c3 l:f3+ 72. �d4 l:f4+
73. 'it.>e5 l:a4 74. l:g8 .:Xa3 75.g5
�xeS 76.g6 l:g3 77. l:c8+ �b4
Black's position is unenviable. He is a 78. �6 l:f3+ 79. �e6 l:g3 80. �7
pawn down, and a second pawn hangs. �a3 81 .g7
Karpov takes a difficult decision: he Draw.
gives up the hS -pawn, but takes play
into a rook ending. To be fair, we should say that another
63 ... �c6!? 64 . .:Xh5 �d4! 65 . ..bd4 solution is possible in the diagram posi
.:Xd4 66. l:g5 l:e4 67.g4 l:a4 tion: 63 ... h4! 64.gxh4 l::!.g2 . Now what
The assessment of the position depends can White do? His bishop cannot move,
largely on that arising after 68 . .l::t.g 8 because of mate in one. After 6S.ltfl
c.t>xcS 69.g5 .l::t.g4 70.g6 l:Ig3+ 7 l .�c2 �c6 we reach a curious zugzwang (66.
c.t>c4 72 .g7 .l::tg2+ 73 .c.t>dl �b3. hS l::th 2). That leaves only 6S.ltdl+, but
then 6S ... �c7 (of course, not 6S ... �c6??
66.l:Id6+ and 67. llxf6) 66 . .id4 .ixh4
67.�b4 �c6, and Black is close to a
draw.
These last variations do not devalue

Karpov's achievement. His method of
defence deserves every attention. And
it brings to mind this classical exam
ple:
analysis diagram
88
Carl Schlechter 50. :Xg6 :Xe7 51 . .:tgc6 :Xc7

Emanuel Lasker 52. :Xc7+ 'iti>g6 53 . .:tc6+ 'iti>f7
Vienna 1 91 0 54.'iti>f3
But even the rook ending looks difficult,

Material is currently equal, but Black if not lost, at first sight. He is a pawn
has three weaknesses more. By trying down and his king is cut off on the 6th
to defend them, he almost certainly rank. The threat is c2-c4 followed by
condemns himself to defeat: 47. ..lbc4 Wf3 -f4. What is to be done?
(threatening a knight fork) , and then: 54 ... :e4! ! ss. :cs 'iti>f6 56. :Xa5 .:tc4
48 .<;!;>h3 l::rb4 49.�d5 f4 50.g4 f3 This is the basis for the classic rule of
s l .�hS+ �g7 s2 .tt:Jds+ Wg8 s3 . .l::th h7! rook endings: 'activity is more impor
or 48 .. .f4 49.tbd5 f3 50.tbb6! f2 tant than material'. The white rook has
(50 ...tbxb6 5 l .�xd6+ <;!;>hs 52.�h7#) to defend the c2-pawn and the king the
5 l .tbxa4 tbd2 52.�d7. g3 -pawn. And Black just bides his time.
47 ... .:tf7! 48. :Xd6+ 'iti>h7 49. l:[e6 57 . .:ta6+ 'iti>e5 58 . .:ta5+ 'iti>f6 59 . .:ta2
tl:lg6! 'iti>e5 60 . .:tb2 .:tc3+ 61 .�g2 'iti>f6
A well-known device: Black strives to Soon, Schlechter accepted the inevita
reach a rook ending at all costs. ble. Draw!
89
Chapter 8
Positional Sacrifices ( Part Two)

The exchange sacrifice on c3 in the Sicilian is so widespread that the reader can
easily find dozens of examples.
Why does Black sacrifice the exchange? One reason is to start a mating attack, es
pecially ifWhite has castled queenside and his king is in the region of c l -b 1 . Then
the sacrifice rips open the king's pawn cover and the attack becomes much simpler.
Another motivation is to sacrifice to go into the endgame. In the book Konturi Endsh
pilya [Translator's Note: By the well-known endgame specialist Mikhail Shereshevsky,
and published in English in two volumes, under the title 'Mastering the Endgame']
this device has an entire chapter devoted to it, so common is it. The compensation is
usually in the form of pawn weaknesses in the centre and on the queenside.
The third possibility is the most difficult to classify and the most interesting to
analyse. It is when the sacrifice transforms one unclear position into another.
But if it is unclear, why sacrifice?
To confuse the opponent. And to change the course of the battle. The benefit
seems quite intangible, but in a practical game, it can be considerable.
two bishops and two pawns for the

Robert H u bner queen. Maybe he will get a third pawn,
Viswanathan Anand if he manages to win fS without losing
Dortmund 1 996 a6, although even so, we are talking
only about compensation for the mate
• rial, not an advantage .
So why did White not play this? Evi
dently, he was shaken by the sudden
change in the position and could not
adjust in time to the new situation. In
such circumstances, it is not surprising
that mistakes follow one after another.
25 ... bxc4 26 . .b:c4 tbxe4
23 ... lb:c3!? 24.bxc3 'it'xa3 25.c4

A mistake, hard to understand at first
sight. It was not so hard to calcu
late 2S.l:le2 I:l.c8 26 .c4! bxc4 27J:!al
cxd3 (even worse is 27..."iYb2 28.l:lbl ,
and after the queen retreats, 29. l:txb7)
28.llxa3 dxe2 29."iYxe2 �xe4
(29 ... tbxe4 30.�el) 30.l:ta2. Black has
90
Chapter 8 - Positional Sacrifices (Part Two)
Now, after the fall of the central pawn,

the most White can hope for is a draw.
It was worth considering 2 7. �d5 !?
and if 27... �xd5 28 Jhd5 tLlc3 , then
29 . .l:f.Id3 l::tc8 30.�c5!. In the other
vanat10n, 27 ... tLlxd2 28. �xb7 tLlc4
29.�d5 , Black has two extra pawns,
but the white pieces have developed
considerable activity: 29 .. .'�Ya4 (29 .. J':!.c8
30. �xf7, whilst after 29 ... tLlb2 3o.nai
the queen does not have a single retreat The contours of the position are very
square, on an open board) 30 . .l:f.a l 'iYbS similar to those faced by Anand in his
3 I."iVe4 tLlb6 32.�c6 'iVb2 33.l::txa6 game 16 years later, aren't they? Similar,
with chances of a satisfactory outcome. but not identical.
27. l:[d3 'ifa4 28. l:[b3 In the counterplay developed by Anand,
Continuing along the path of least an enormous role was played by his
resistance. The last chance was 28.�b3 two bishops. This is in general a very
'iYd7 29.�M! , counting on 29 ... �xh4 common form of compensation for
30.'iYxM 'iYxfS 3 I.'iVe7 tLlf2+ 32.�gl , the exchange sacrifice on c3: the two
and, strangely, Black cannot land a deci bishops (even better is two bishops plus
sive blow. a pawn, in which case the compensa
28 ... 'ifxc4 29 . .:Xb7 'ifxc2 30 . ..ig1 tion can often be more than sufficient) .
..igS! 31 .l:[db1 ..if4 32. 1:[1 b3 dS But Portisch, as we see, does not have
It is all over. Anand has three pawns for the two bishops. So the question, for
the exchange, dominates the centre and what exactly he has given the exchange,
a mating attack is not far away. After a remains in the air. For now, White
few moves, White resigned. undoubles his pawns, consolidates his
position and settles down to realising
Evgenij Ermenkov his material advantage. Or does he?
Lajos Portisch 1 8 ... .:ca!
Skara 1 980 The consolidation will have to wait a
moment. After 19.cxb5 Black has pre
pared 19 .. J :hc2 20. �f2 .Uxf2 2 l .�xf2
"iYgs 22 ."fi'fl 'iYd2+ 23 .�gl (23 .'iYe2
'iVd4+) 23 ... axb5 , and White's pieces
are tied up on the back rank.
1 9. 'ife2 tbf6 20 . .:Sc1
The threat to take on c2 is taken care of,
now what?
20 ... h5! 21 .cxb5 axbS 22. 'ifxbS
As in the previous example, the side
which has received the exchange sacri
1 4 ... .:Xc3!? 1 5.bxc3 tbxe4 1 6. tbxe4 fice loses his head. The outcome of the
be4 1 7. be7 Wfxe7 1 8.c4 game remains unclear and White could
91
still defend, but his head was obviously 1 4.l:.xc6!? .bf3

spinning. Nothing else can explain his 14 ... bxc6 l S .tt:'lxeS.
taking an obviously poisoned pawn. 1 5 . .bf3 bxc6 1 6. 'ifc2
22 ... 1i'a7+ 23. <ith1 This is a modern game, but the com
Black has many tempting paths. One pensation is of the classical sort: two
excellent-looking line is 23 ... tt:'lg4 bishops. Admittedly, it is not quite
24.'ife2 tt:'le3 2S.�gl 'ifxa3. obvious why White refrains from taking
23 ... h4 24. 'it'b3 on aS on this move and the following.
The threat was 24 ... h3 . Perhaps he did not wish to be distracted
24 ... ltJg4 by 'trifles'.
This line was also sufficient for victory. 1 6 ... 1i'd6 1 7. l:tc1 l:tb5 1 8.e4
But the most decisive continuation was But this is wrong. Heinbuch incorrectly
23 ... l:lc3 !. By seizing the third rank, avoids the forcing variation (was he
Black combines the threat of ... hS-h4-h3 intimidated by his opponent's GM rep
with those of ... tt:'lf6-g4 and ... .!:lc3 -h3 ! . utation?) : 1 8 .'ifxc6 ifxc6 1 9.l:i.xc6 tt:'lb6
20 . .l:f.xc7 tt:'lxa4 2 l . �c6 �cS 22.�xe8
The mirror image of the sacrifice on .l::txc7 23. �xa4. He has already two
c3 is the equally common sacrifice by pawns for the exchange, on two flanks
White on c6, which usually arises in the (which means he has chances of creat
English Opening, but with the pawn ing a passed pawn), and has managed to
structure being that of a Sicilian, with retain the two bishops.
colours reversed. After missing this chance, White even
Everyone has seen or heard of classical tually struggled to make a draw.
examples on this theme. We will show
an example which perhaps has not been Vasily Pa nov
published before. This is because the Vlad i m i r Simagin
sacrifice was made by a player who is Moscow 1 943
not famous, indeed is hardly known at
all. But this makes it all the more inter •
esting to follow his efforts.
Detlef Heinbuch
Aloyzas Kvei nys
Bonn 1 995
By analogy with the previous example,

one can suggest 1 2 ... �xc3 !?. Or, so as
not to give White additional options,
12 ... �xh6 1 3 .'itxh6 I:i.xc3 !? 14.bxc3
'ifc7. Black has compensation. But
Simagin had a quite different idea.
92
Vladimir Pavlovich Simagin was a 1 7 ... llJe5 1 8.f4 ltJf3! 1 9. ltJxf3

remarkable player, who always sought The same experts who suggested not
his own way and was full of ideas. In taking the exchange at move 13 now
the position of the diagram he played a suggested returning it: 1 9.�e3!? tLlxd4
move which made everyone gasp: 20.�xd4 �xd4 2 1 .'ii'xd4. Incidentally,
1 2 ... --thB!? the final position here cannot in any
When seen for the first time, the move way be assessed as better for Black,
makes a colossal impression, of course. which shows that the sacrifice was
What is Black getting for the exchange? experimental.
Nothing, except depriving the oppo 1 9 ... .txc3
nent of his dark-squared bishop, whilst
retaining his own.
1 3 . .hf8
Under the impression created by
Simagin's wonderfully conducted
attack, many experts claimed that
White should not have taken the rook.
This is an exaggeration. The sacrifice is
no more than a practical one, the sub
ject of argument. There is no question
of Black having a decisive advantage. The key moment of this exciting game.
1 3 ... 'it'xf8 1 4. ttJd4 .tc4 1 5.g5 ltJfd7 After capturing on c3 White will have
1 6 . .th3 e6 1 7. 'iPb1 an extra rook. But as far back as Morphy,
Standard prophylaxis, and at the same we have seen that it is not the abstract
time either an oversight or careless material advantage which matters, but
ness. The most likely explanation is that the amount of material in the relevant
White did not see his opponent's 1 8th area of the board. That is the case here
move and the resulting combination. also: Black has fewer pieces, but all of
He should have started with 1 7.f4, those he has are pointing at the enemy
keeping the black knight away and king, whilst his opponent does not have
ensuring space for the white pieces to time to bring additional reserves over
manoeuvre. In reply, Black cannot play for the defence.
1 7... �a6 because of 1 8 /2Jxe6! which only For example, 20.'ifxc3 �xa2+ 2 l .�xa2
leaves 1 7... tLlcS . But then 1 8 .Wbl �a6 .l:lxc3 22 .bxc3 'ii'c 8 !? (stronger than
19.f5 tLlc4 20.�f2 and then, depending 22 ... d5 23 .exd5 tLlxdS 24 . .!::lxd5 ! exdS
on circumstances, either taking on e6 2S .tLld4), and you can be sure that the
or f5 -f6, shutting out Black's strongest queen and knight will finish the job,
piece. One can also be more concrete: before White manages to transfer some
18 .f5!? �xa2 1 9.f6 (but not 1 9.tLlxa2 more pieces to the queenside.
�xd4 20.�xd4 tLlb3+), and the coun In the game, there followed:
terplay for which Black sacrificed the 20.bxc3 d5!
exchange does not materialise. As the variations show, Black has full
However, the text move does not yet compensation for his enormous mate
spoil anything. rial deficit. Later, White failed to defend
93
in the best possible way, and then gave the enemy king, but others just found
the game away in a single move. themselves an exchange down.
The position of the last diagram illus The following game gave a new impulse
trates the difference in thinking for the study of such a sacrifice. It
between players. Master Panov evi showed that the sacrifice could be used
dently thought he was obliged to 'pick not only for the sake of a mating attack,
up the gauntlet' and take all the mate but also for positional counterplay.
rial he was offered. The modern grand
master would most likely start calculat Abram Khasin
ing the variations involving the capture Leonid Stei n
on c3, but only so as to establish that Tallinn 1 965
there was no forced win, and a lot of
0
risk. Having reached this conclusion, he
would not waste any more time delving
deeply into calculating all the lines after
20.'itxc3 or 20.bxc3 , but would turn his
attention to 20.'itxd6! .
Now what does Black do? I t i s important
that he does not have 20 ...'itg7 because
of 2 I .'itd8+!. After 20 ... ..te2 2 1 .'ilixf8+
�xf8 22 . ..tg2 , Black can regain almost
all of the material, but only to con In the diagram position, it is White who
demn himself to a cheerless endgame is attacking. He has already played f4-
a pawn down. This means that the only f5 , allowing a powerful knight into e5.
way to retain queens is 20 .. .'iVe8 , but There is no way back, and White presses
then after 2 l .bxc3 , the white queen has on consistently:
turned from a spectator on d2 into the 1 6.f6!? gxf6 1 7.gxf6 .hf6 1 8 . .th6
main defender of her king. The varia The clouds are gathering over the black
tions are not very complicated: 2 1 ...'iVa4 king. Gufeld and Lazarev, the authors of a
22 .a3 .tal+ 23 .�al 'ii'xc2 24.'iVd2 or book of Stein's best games, offer the fol
2 1 ...tDa4 22.�al tt:Jxc3 23.'ii'd 7 tt:Jxdl lowing variation: 1 8 .. J�fe8 (even weaker
(23 ...'iVf8 24.tt:Je5) 24.'itxe8+ l:!.xe8 is 1 8 ... tt:Jd7 1 9.�gl+ �h8 20 . ..tg7+ with
2 5 Jh d l .tel 26.l::i.d3 ..txd3 27.cxd3 , mate) 1 9. ..th5! Vi'e7 20.�xf6! 'iVxf6
with winning chances. 2 l . �gl+ �h8 22 . ..tg5 'iVg7 23.�h4
'iVh6 24.'iVe3! 'iVxe3 25 . ..tf6#.
The game Panov-Simagin gave rise to a This variation with the sacrifice of rook
lot of hype. In the Dragon Variation of and queen is all very glittering, but it
the Sicilian Defence, and King's Indian is not clear that White has a decisive
positions, everyone started answer blow after 22 ...'ii'f5. Probably, White
ing ..th6 with ... ..th8 , giving up the does better not to be seduced by the
exchange, but keeping the dark-squared beautiful and to play instead 20. l:i.f2 or
bishop. Simagin attracted dozens, if 20J:i:f4, bringing the second rook to the
not hundreds of followers, but not all f-file. Then Black certainly is not to be
were successful. Some managed to mate envied.
94
Leonid Stein was a great master of the An instructive moment, which again
initiative, and there is no doubt that he illustrates the difference in thinking
had seen his next move some time in between players of past generations and
advance. those of today. In their book, Gufeld
1 8 ... '1t>h8! and Lazarev award Black's 23rd move
It is interesting that the computer, usu an exclamation mark. In all probability,
ally a keen hoarder of material, rec Stein himself would have agreed with
ommends as an alternative going into this judgement. And not without some
the endgame here: 1 8 .. .'�c5!? 1 9.Vi'xc5 justification - after all, he did win the
dxc5 20. �xf8 �xf8 , and assesses the game, albeit after considerable compli
final position as better for Black. We cations.
have already pointed out that the two But in his book My Great Predecessors, Garry
bishops plus an extra pawn is often Kasparov gives the following comment:
more than sufficient compensation for «Probably even better was 23 ...tt.:lc4 24Jid3
the exchange. �e5.» So who is right?
Stein's move is stronger, though. The Both are right. Stein was not afraid of
queens are not exchanged and the c-file complications and felt like a fish in
not closed, which means that, in addi water in such positions. He won, and
tion to counterplay on the kingside, the winner is always right. But the 1 3th
Black has the typical Sicilian motifs World Champion shows that the risks
... b5-b4 and ... tt.:le5-c4. could have been avoided, and he is also
1 9 . ..bf8 :Xta 20. l:ad1 l:l.d8 21 . Ag2 correct in his way. In his variation,
Ag7 22. 'ift2 l:l.g8 instead of 24 ... �e5 , it is bad to play
The consolidation stage did not last 24 ... tt.:lxb2 25Jih3 h6 26.l::tf6! but after
long. Black is ready finally to take the 24 ... .i.e5 25 . .l::rh 3 l:lg7, it turns out that
initiative. He is stronger on both flanks. Black has retained all of his plusses (the
23.'ifh4 attack on b2 , ... b5 -b4, .. .f7-f5 , etc.) , and
After 2 3 . l:rd2 , there would probably also strengthened his king. A modern
have followed the same move as in the day grandmaster would probably play
game: 23 .. .f5 , and not 24.exf5 �xg2+ in exactly that way.
25 .Vi'xg2 because of 25 ... �h6 26.Vi'e2
�xd2 27.Vi'xd2 Vi'c6+. Francisco Briffel
23 ... f5 Alexander Kha l ifman
Moscow 1 985
95
This example is not totally contempo The resulting positions are s o com
rary, but it is from a new era, when the plicated that even very deep analysis
exchange sacrifice had become a nor does not always answer the question of
mal thing and no longer surprising to whose chances are preferable.
anyone. Here is a characteristic example:
White begins the standard attack:
1 1 .h5!? lbxh5 1 2.g4 lt:Jg3 1 3 . .1:.h2 Oleg Korneev
tbxf1 1 4 . .1:.xf1 c5 1 5. �h6 Veselin Topa lov
Cala Galdana 1 999
Against a beginner, such an attack The pawn on e4 is attacked. White is to

would have every chance of success. play and cannot be happy with 1 8 .tLld5
For example, 15 ... �xh6 1 6 .'iVxh6 tLlf6 (hoping for 18 .. Jhe4 1 9.tLlc7+) because
1 7.g5 tLlh5 1 8 . .!:ixh5! gxh5 1 9.tLld5 l:te8 of 1 8 ... �xd5 19.exd5 .l::txg4 or 1 9.'iVxd5
20.tLlf6+! exf6 2 1 .gxf6 , and mate is not 'iVxd5 20.l:i.xd5 l:i.xe4.
far away. White's chances lie in an attack on the
In 1985, Alexander Khalifman was not king. The central e- and d-files need to
yet 20 years old. I don't think he spent be opened and for this purpose, the
long on his next move. By this time, e4-pawn can be sacrificed. Let us play
he had already mastered the technical a neutral move such as 1 8 .h3, defend
device of the positional exchange sac ing g4. Let us assume Black 'bites':
rifice. 1 8 ... �xe4 1 9.tLlxe4. What should he
1 5 ... ..bd4! 1 6 . ..bf8 lbxf8 retake with? If the queen, 1 9 ... 'iVxe4,
For the exchange, Black has not one, but then with 20.tLlxe5! tbxe5 2 1 . �b6
two full pawns, plus the bishop pair. White regains all his material and
There followed enters an endgame with an extra pawn:
1 7. tt:Jd5 'ifd8 1 8.t002 �g7 1 9. 'ifg5 2 l ...'iVc6 22 . .t:l'.xe5+! �e7 23 .'iVxd6
.l:.a7 'iVxd6 24 . .l::i.xd6. Black must take with
White has run out of sensible attacking the rook: 19 ... l:i.xe4, but after 20.'iVd3
moves, and Black is quite happy to see l:i.c4 2 1 .tLld2 .l:tc6 22 .tLle4, White has
the senseless ones. achieved what he wanted. The central
files are open and his pieces have the
A special place is occupied by the sac blockading squares e4 and d5. If Black
rifice on e4. The compensation again manages to castle, White will advance
consists of the two bishops plus a pawn. his kingside pawns h3 -h4, g4-g5 etc.
96
We can conclude that with the pawn gxf6 23 .�d4 d5 24. l:txe4+. The bishop
sacrifice, White ensures himself a too: 2 l ...�e7 22.�xg7 .l:Ig8 23.f6. That
long-lasting initiative. But accepting the only leaves 2 l ...�d8 , but then 22. �xg7!
sacrifice is not obligatory; after 1 8 .h3 , �xg7 23 .�xd6 �c6 24. �xe4 'ifxd6
Black would more likely reply 1 8 ... �e7. 25 . .l:ixd6. White has a rook and two
And then what has White achieved? pawns for two minor pieces, and
Nothing. He faces the same problems as another pawn is attacked. Black's posi
one move earlier. tion is unenviable.
1 8.b3 Sadly, though, Black is not obliged to
Of course, Grandmaster Korneev saw take the knight on d4. He has at least
the exchange sacrifice. Taking into two continuations which deserve con
account all that was said above, he evi sideration. Firstly, 20 ... tt:lf6 2 l .g5 hxg5
dently wanted to force events and give 22. �xg5 !li.e7 and the game was, and
the game a concrete character. And he remains, extremely unclear. Secondly,
succeeded, but he did not manage to 20 ... !Ji.e7, allowing 2 l .f6!? tt:lxf6 (weaker
cope fully with the task of calculating is 2 l ...�xf6 22 .tLlf5 �xf5 23 .gxf5 !li.e7
the resulting variations. 24.l:f.gl) 22.tt:lf5 �xf5 23 .gxf5 .
18 ... :Xe4! 1 9. ltJxe4 ..be4
analysis diagram
Material is not the main factor in this
position. The assessment depends How should we assess this position?
on whether Black will manage to get The g-file is now open and castling is
through to the enemy king. not without danger after the capture on
Tempting is 20.�d4!? �xf3 2 L � .xe5 , h6. But Black also has his chances. In
and i f 2 l ...�xdl , then 22..�.xd6+ �d8 the sample variation 23 ... tt:le4 24.'iVd3
23 .�a5+ mating. Stronger is 2 1 ...tLlxe5! d5!? 2 5 . !Ji.c1 (25 .'iVxd5 tt:lc3+) 25 ... 0-0
22.l::ixe5+ �d7. Black's king is bad, but (he loses after 25 ... �b4 26.l:lxe4! dxe4
White does not have any more minor 27.�d7+ �f8 28 .'iYd8+, 29. l:f.xd8+,
pieces he can sacrifice, and without a and 30.Ilxh8) 26.!Ji.b2 �h4 27. .l::!e 2
sacrifice, he cannot get at the king. tLlf2 28.l:i.xf2 !li.xf2 29.�xe5 (worse is
Another line is 20.tt:ld4!? exd4 2 l . �xd4. 29.'iYxd5 'iYxd5 30Jhd5 e4 3 l .l::ie 5 e3
A lovely position has arisen. Black has 3 2 . �d4 .i:f.d8 3 3 . !li.xe3 l:tdl+, and now
an extra piece, but no moves. The not 34.�b2 l:tel , nor 34. �cl �h4!)
knight cannot move: 2 1 ...tLle5 22..�xe5 29 .. .f6 30.�d6 l::ie 8 3 1 .'iYxd5+ 'iYxd5
dxe5 23 .�d7# or 2 1 ...tLlf6 22. �xf6 3 2 . l:f.xd5 leads to an equal endgame.
97
20. tlJg1 Octavia Troianescu

Tantamount to capitulation. Probably Tigran Petrosian
White calculated the variations out Bucharest 1 953
lined in the previous note and did not
0
like any of them. But now Black has an
easy game. He completes his develop
ment, aligns his pawns in the centre (...
d6-d5) and then goes after the enemy
king himself.
20 ... ..te7 21 . Af2 0-0 22. 1i'e3
There is no sense in 22 .h4 �c8 23 . .l::l.c l
'i:Vc6 24.g5 because of the simple 24...
hxg5 25.hxg5 ltxf5 . But standing still
and awaiting the end is not the most Black has the two bishops and the bet
sensible approach, either. ter prospects. But White's position is
22 ... ti:lf6 23. Ah4 'ifc6 24 . .1:.e2 .:ca quite solid, with everything defended.
25. <itb2 d5! It is not obvious how Black can break
through. A long battle lies ahead.
25. ti:le3
This move deserves some discussion.
In the previous example, White also
provoked his opponent to sacrifice the
exchange, but there he had a specific
motive. He needed to change the quiet
course of the battle and make the game
more concrete. But what was the reason
here? White's position is worse. He has
The white rooks cannot exploit any of no concrete variations to change this sit
the files, whilst the black bishops dom uation. By provoking the exchange sac
inate the whole board. White's king is rifice, White only increases his troubles.
also weak. The game is decided. It was necessary to bide time, for exam
26 . .bf6 .bf6 27.c3 ..tgs 28. 1i'g3 ple, with 25.�h l , waiting to see what
d4! 29.cxd4 plan Black will adopt to strengthen his
Declining the dubious pleasure of position. Depending on what he does,
being mated on the edge of the board: White will then base his further plans.
29.'i:Vxe5 Vi'xc3+ 30.�a3 b4+ 3 l .�a4 And if White wanted to play actively,
ltc6+ 32 .�a5 ltd8+ 33 .Wxa6 l::ta 8+. then the only rational try was 25.f5!?.
29 ... exd4 30. l:.de1 25 ... .:Xe4! 26 . ..txe4 .be4
Not 30J lxd4 because of mate in one. What Black has done can hardly be
30 ... d3! 31 . .:Xe4 ..tc1 + 0-1 called a sacrifice. Formally, two bishops
and a pawn is fully adequate compen
By way not of a stem game, but a sation for the exchange. And all the
prototype, I offer the following classic changes in the position are in Black's
example: favour.
98
White has clear problems on the long Evidently, when considering his 29th
diagonal a8-hl . After the inevitable ... d6- move, Troianescu noticed a nice trap
d5 and ... .ig7-f8 , he will also have trou in this position: 3 1 ...'irxa4 32 .tbxe6!
ble on the adjacent diagonal a7-gl . In fxe6 33 . .l::txe4. But Petrosian, true to his
addition, he needs to watch for the breaks style, does not hurry. He methodically
...b5-b4 and ... d5-d4. And all of this with improves the position of his pieces .
a complete absence of counterplay! 31 ... :ca 32.b3 .tta
27. tlJc2 All in the same manner. After 32 .. J:k3
Perhaps he could have tried to jump out 33 .tbbs 'iVxf2+ (stronger is 33 ... 'iVb8 !?
with 27.tbc4!? but alas, after 27 ... itdS , 34.tbxc3 bxc3 3S.�de2 'iVxb3) 34.'it>xf2
the knight would have been forced back .l:i.xb3 3S.a5, White gets some hopes of
again. 28 .tbxa5 'iVb6+ and 29 ...'iVxa5 . counterplay.
27 ... d5 28. ttxl4 33. tilb5 'ifa6 34. 'ife2 'ifb6+ 35. �1
Since he has nothing real to do, he
should have very slightly improved his
king with 28.h3 ! ? and 29.<it>h2 .
28 ... b4
The position is like a picture and

deserves to be framed. All of Black's
pieces stand ideally. But how can White
be forced to resign?
29.cxb4 A paradox: in the previous course of
One of several arguable moments in the the game, Black several times refused
game. It seems as though White loses the chance to win material, and now he
after 29.axb4 axb4 30.c4 on account of sacrifices another exchange:
30 ... dxc4!? 3 1 .'iixe4 'iVb6 32J:tedl c3 35 ... :c3! 36. tilxc3 bxc3 37. :c2
33 .bxc3 bxc3 34.�d3 c2 . However, if we In reply to 37.l:ld3 , Petrosian gave the
extend this line: 3S . .l::f.c l itxd4+ 36.c;!tg2 variation 37 ... itfS!? (adding the threat
Vibl 37 . .l::f.xd4 (after 37.'iVel it all goes of a check on h3) 38.g4 itxd3 39.'iVxd3
wrong: 37 ...'iVb7+ 3 8.<it>h3 .l::f.d S 39.g4 'iVxb3 40.f5 'iVxa4 4l .fxe6 'iVf4+
itb2 40 . .l::f.xd5 'iVxdS 41. .t:f.xc2 'ifd3+) 42 .'.t>g2 'iVxg4+ 43 .'it>hl fxe6. But just
37 ... �xd4 38.'iixc2 'iVb7+ 39.'tt>g l , some now, after 44.'iVxc3 .ig7 4S .'iVe3,
doubts arise. Black has an extra pawn White has real drawing chances:
and an attack on the king, but all the 4S ... c;!tf7 46.'iVa7+; 4S ... <it>h7 46.h3
pawns are on one side, and there are not and 47.'ifxe6 (but not the immediate
many pieces left on the board. Maybe 46.'iVxe6 'iff3+ 47.'it>gl itd4+); 4S ... e5
White is suddenly making a draw? 46.'iVb3 'iVd4 47. .l::r.d l 'iVe4+ 48 .'it>gl ,
29 ... axb4 30.a4 'ifa7 31 . 'iff2 and the dS-pawn cannot be defended.
99
I suspect that, at the board, the ninth Bronstein's commentary in his famous
World Champion would have found a book on the tournament deserves to be
more convincing way of realising his reproduced in full:
advantage than the one he gave in his «Black needs to blockade the white pawns and
notes. Today the computer indicates Petrosian sacrifices the exchange, so as to free the
39 .. .'ib'c7! as stronger than 39 .. .'iVxb3. e7-square for this knight, which then comes to d5.
37... 'ifxb3 38.:ec1 ..tb4 39.g4 Admittedly, Black obtains serious compensation:
..txc2 40 . .:Xc2 'ifxa4 41 .f5 exf5 his knight on d5 and his unopposed bishop, are
42.gxf5 g5! both unusually strong.»
Black has managed to keep pawns on 25... l:te6! ?
both flanks. Now there is no problem at The difficulty of Black's position lies in
all in realising the advantage. the fact that his opponent is not obliged
43.h4 ..tc5 44.hxg5 'iff4+ 45. 'ite1 to accept the sacrifice at once. He has
'ifg3+ 46.<;i;>d1 'ifg1 + 47.'ife1 great freedom of action. Bronstein him
'ffxe1 + 48. c;i;>xe1 hxg5 49. 'ite2 ..td4 self suggested the plan of attacking the
50.l:a2 c;t>g7 51 .c;t>d3 ..te5 52.:a5 king with 26.h4, and then 27.�g3 and
c;t>f6 53 . .:Xd5 c;i;>xf5 54. c;i;>e3 f6 55. :C5 h4-h5. But Reshevsky decided first of all
c;i;>g4 56.:C4+ c;i;>g3 57.c;i;>e4 g4 0-1 to break up the black queenside pawns.
26.a4
Another motivation for an exchange Now if 26 ... b4, then after 27.d5! �xd5
sacrifice is to build a fortress. That 28.�xe6 fxe6 29.'iYxc4, there is no
is what one might call 'a draw from trace of the blockade left.
a position of weakness'. Quite often 26 ... tlJe7! 27. he6 fxe6 28. 'iff1
in such cases, a rook is given up for a An accurate move, keeping the c4-pawn
bishop, which can alone attack squares under observation from the queen.
of a certain colour. Then the blockade is After 28.'iVf2 tLldS 29. l:tg3 b4 Black
established on those squares. takes over the initiative.
Of course, everyone remembers the fol 28 ... tbd5 29. l:tf3 ..td3 30 . .:Xd3!?
lowing classic example. cxd3 31 .'ti'xd3 b4
Samuel Reshevsky
Tigran Petrosian
ZOrich 1 953
An instructive moment. Black has

played this section of the game out
standingly, but even so, he is still a long
way from equality. White has returned
the exchange and obtained in return an
1 00
extra pawn. He continued the fight for 42 .. J:th7 43 . .l:Ihl lixhl 44.�xhl , and
an advantage. then at the necessary moment, g3 -g4.
The game ended in a draw anyway. But So what do we do, resign? It is still
the reader should note that one device, early for that. Black has one good piece:
no matter how beautiful and useful it the blockading knight on e6. In addi
was, does not necessarily guarantee a tion, all of his weaknesses are on light
certain result. Two players play a game squares and can only be attacked (for
of chess. By doing a certain thing cor the moment) by the bishop on b3. This
rectly, one enhances one's chances, but means that the first priority must be to
the opponent, in turn, can also act cor eliminate that bishop.
rectly and improve his chances. And so 34 ... .1:.c4! 35. hc4 dxc4
it goes on, until the end of the game. Black only needs one move to establish
the blockade: ... ..ib7-d5 . Then he can
Alexander Areshchenko simply mark time, preparing to meet
Jan Timman the break h4-h5 with the counter ... g6-
Germany 2007 g5 . But White is up to the task:
36.d5! hd5 37 . .1:.hd1 lL\xf4
• Giving up his best blockading piece
for the useless white bishop is the very
last thing Black wants to do, but there
is no choice: 37 ... 'iiic 6 3 8 . l:f.xd5! �xd5
39.l:1dl+, 40. l:f.d6+ and 4l. . l:f.xe6.
38. <itxf4 <ite6 39.g3 .l:.c8
Black resists with all his strength. How
ever, as in the previous example, he is
still a long way from a draw. White has
White is close to victory. He has an ex the tempting plan to double rooks on the
tra pawn, the two bishops and a possi d-file, tying down the black pieces, after
ble breakthrough on the kingside. Black which he can take his king to g5 . Black
has only pawn weaknesses and depress will then have nothing with which to
ing passivity. defend the weakness on g6. That means
The pawn on b5 is hanging. 34 ... ..ic6 the only chance to resist is the desperate
is no use, as White plays 35.l:!.c3 , dou counterattack ... c4-c3 and .. Jic8-c4.
bles rooks on the c-file and wins with That is roughly what happens:
a short tactic, by taking on c6 and b5 .
The preliminary exchange 34 .. J1xci
35Jh ci , and only then 35 ... ..ic6, also
fails to save the position: 36 . ..ie2 l:!.h8
37.g3 .l:Ih7 38 . ..ig5 tbxg5 (or 38 ... l:!.h8
39. ..if6 .l:f.h7 40.h5 ! gxh5 41 . ..id3)
39.hxg5 .l:Ih3 40.<Et>f2 l::rh 2+ 41 .\t>f3
(zugzwang) 4I ....l:th8 (4I ...l:!.h5 42 .c;t>g2!
.l:txg5 43 .�h3 and 44.c;t>h4) 42.�g2
(the immediate 42 .g4 is also possible)
101
40. l:[c3 l:[c6 41 . l:[d2 l:[c8 42. l:[e3 1 9.tt::ld xfl tt::lg6 (Korotylev-S.Novikov,
c3! ? Moscow 2007), Black banked the mate
Not waiting for White to double rooks. rial in a more favourable form, thanks
43. l:[c2 l:[c4+ 44. �g5 l:[g4+ 45. �h6 to the fact that the pawn on h4 restricts
�e4 46. l:[cxc3 �xes the white rook on h3 and slows up
We can see how one difficult position White's attack.
leads to another. Even so, Black man 1 6 ... ..bf1 1 7. tbxf1 b5
aged to save half a point. Drawn on Here too, 1 7... h4 was tempting, and
move 68. then, if necessary, ... tt::l e S -g6 and ... i.g7-
f6. But Rashkovsky had a different idea
The sacrifice to take over squares of a cer in mind. In search of counterplay, he
tain colour can also be used in playing wanted to sacrifice three pawns, one
for a win. Such cases are less common, after another.
but one does come across them. 1 8. tbde3
The idea is revealed in the varia
Tigran Petrosian tion 1 8 .axb5 l::tb 8 1 9.tt::lc 3 fS !? 20.exf5
N u khim Rashkovsky tt::lg4 2 I. .I:i.xh5 .l:f.el . White has a mass
Moscow 1 976 of pawns for the exchange, but Black
has lively piece activity. Admittedly,
0
though, it is an open question whether
this play is sufficient (after, say, 22 . .l:IgS
or 22.i.d2 .l:Ia1 23 .�e4).
1 8 ... bxa4 1 9. ltJf5
The rook on f1 is under attack. After

1 6 . .!:f.e1 c4 1 7.tt::l e 3 tt::l d3 1 8 . .!:f.fl �c8
1 9.tt::lfs , in the game Podgaets-Zaichik
(USSR 1 986), a complicated position
was reached, with chances for both
sides. Instead of 16 ... c4, it was worth
considering 16 .. .f5 ! ?, when 1 7.exf5 tt::lc4 Here, we should pause for a moment.
1 8 . .l::[xe8+ �xe8 1 9 . .l::r.g3 is bad because White has sacrificed the exchange, and
of 1 9...�el + 20.tt::lf l tt::la 3 . the valuable bishop from a6, the main
Petrosian was in a fighting mood: defender of the light squares, has been
1 6. l:[h3!? removed from the board. White's knight
Later, this idea was tested a number of has a great square on fS and overall, he
times in practice. In particular, it was has excellent attacking chances. This is
established that it is not obligatory to the plus side.
take the exchange straightaway. After But an exchange is an exchange. Black's
1 6 ... h4!? 1 7.tt::le 3 i.d3 1 8 .�d1 i.xfl rooks have the excellent open b- and
1 02
e-files and his attack needs only 3-4 c l : 24 ... 'iYbl 2S .ltJg4 �b2 26.'iYh6 �c2 ,
moves to become the main factor in the and there is nothing better than per
position. In a word, Black has time to petual check: 27.'iYxh7+ �f8 28.'iVh6+
present his case and develop counterplay. �g8 . This variation can be extended:
A tempting line is 1 9 ... .l:ib8 20 . .l:r.xhS 29.�g2 'iYxcl 30.'iYh7+ �f8 3 1 .lDxeS
�b4 - from this square, the rook (with the idea of 3 1 ..JheS 32.'iYh8+!
defends the pawn on a4 and attacks lDxh8 33 . .!::f.x h8#), but after 3 1 ..Jhf2+!
e4 and b2 . But after 2 1 .lDle3, the next 32 .�xf2 'iVd2+, it is still perpetual
move is not obvious. Rashkovsky goes check.
for a different regrouping: the knight But White is winning all the same:
on g6 and bishop on eS, but this also 29.�g2! 'iYxcl 30.l:tgS! The sacrifice of
proves unfortunate. On eS, the knight a second exchange cannot be prevented
currently prevents the white knight and the rest is simple: 30 ...'iYb2 31..l:x:. g6+
coming to g4, whereas soon the white fxg6 32.'iYxg6+ �f8 33 .lDfh6, and then,
attack will develop without hindrance. for example, 33...�e7 34.'iYe6+ �d8
1 9 ... tbg6 20 . .1hh5 �e5 21 .g3 .l:b8 3S.lDf7+ 'lt;;c 7 36.'iYxe8 c4 37.lDfxeS
22. tt:J1 e3 a3 23.bxa3 1Vb6 dxeS 38.'iYc6+ �b8 39.d6 'iYb6
The black king's position deteriorates 40.'iYe8+ �b7 41 .d7, winning.
quickly and Rashkovsky tries what 24 ... 'iVb3 25. tbg4 .l:b7 26. <itg2
is perhaps his last chance: he seeks to Again the same motif. While planning
exchange queens. Even in the most a mating attack, Petrosian does not for
primitive variation 24.lDc4 'iVbl (or a moment forget about prophylaxis. It
24 ...'iYb3 2S .'tie2 'iYbl 26.'iYe3) 2S .'iYd2 may look as though this was the moment
'iYxe4 White is probably winning: for the 'brilliant combination': 26.�xh7
26.lDcxd6 .ixd6 27.'iYh6! (of course, �xh7 27.'iYh6+ �g8 28 .lDxeS, mating.
not 27.lDxd6 'i!Vel+) 27 ... 'tiel + 28.'it>g2 But Black replies 26 ...'iYc3 ! forcing the
'ife4+ 29.�h3 . But Petrosian, true to exchange of queens. And he even wins
his style, finds a way to combine aggres further material. If 27.'iYh6 'iYel +
sive play with prophylactic measures. 28 .�g2 'iYxe4+, and then taking on fS ,
24. 'it'd2 and White even loses.
Therefore Petrosian removes his king
from the back rank. Now the blow on
f7 is on the agenda.
26 ... 'it'c4 27. tt:Jxe5 'ifxe4+
Any capture of the knight is met by
28 .tiJd6.
28.f3
White also wins with 28 .lDf3 .l:ib3
29.'iVh6 'iVxf3+ 30.�h3 , but Petrosian
prefers a safer method.
The concentration of pieces on the king 28 ... 1Vxe5 29. tt:Jh6+ <itf8 30.:Xe5
side is taking on a threatening charac :XeS 31 . tbg4
ter. The counterplay is too late. It seems Rashkovsky did not want to continue the
that Black can play to pin the bishop at game without a queen. Black resigned.
1 03
Shakh riya r Mamedyarov It may appear that the white initiative

Elmir Gusei nov has come to nothing.
Baku 2000 But there is also the resource 24.tbb5 !?. If
24 ... tbf6, then 25 .1i'e2, and two pawns
on a7 and d6 are attacked. After 24 ...
0
a6 one must reckon with 2 5 .tba7!? tbf6
26.1i'e2 , and now bad, for example, is
26 ...1i'd7 27.tbc6 tbxc6 28.dxc6 1i'xc6
because of 29.tbxg7 �xg7 30.i.c3 (or
30.e5 tbd5 3 l .c4).
A very interesting game could result.
After the text move, however, White's
task is simplified.
22. 'ife2 'ifg6 23.�h3!
With the moves ... h7-h6 and ... g6-g5 , Mamedyarov was just 15 at the time
Black has weakened the light squares of this game, but how maturely and
around his king. But it is not clear how impressively he plays the game! The
to exploit these. Not 20.tbxd6 tbe5 . knight is attacked and, indirectly, also
Black needs just two moves to com the queen. Not 23 ... tbf6 because of
plete his plan: the exchange on c4 and 24. i.f5 1i'h5 25 .g4 1i'h4 26.i.e1 1i'h3
... tbd7-e5 . 27 . .!:If3 , and the queen is lost. It can be
What should White do? saved only in one way:
20. tbe3! ? 23 ... ..bc3
A n even cleaner example than the pre Now that White also has two long-range
vious one. There, as well as a complex of bishops, his advantage is beyond doubt.
light squares, another factor was Black's 24 . ..bc3 tbeS
broken kingside pawn structure (f7, h7, Also bad is 24 .. .'iYxe4 25 .1i'd2! tbe5
h5) . Here, however, White has no other 26.tbf5 tbf3+ 27.l:txf3 'ifxf3 28.tbe7+
compensation for the exchange, apart �h7 29. itf5+.
from his play on the light squares. The 25. �f5
next few moves will show whether this Another way to win was 25 .tbc4!? tbxc4
compensation is sufficient. 26. i.f5 or 25 .. .f6 26.tbxd6.
20 ... ..bf1 21 . :Xf1 'iff6 25 ... 'ifg7 26. 'ifhS f6
One can understand Black's desire to get
his queen to g6, before White's queen
appears on h5 . Even so, on g6, the black
queen will be very badly placed. With
the manoeuvre i.g2-h3 -f5 , White can
underline this fact.
Evidently, there was no alternative to
2 l ...tbe5 22 .1i'h5 tbe8 23 .tbf5 �h7 (the
immediate 23 ...tbf6 is bad because of
24.tbxh6+ �h7 25.l:txf6!). The threat is
24 ... tbf6. This move follows after 24.h4.
1 04
At a quick glance, it may appear that (�cl -b2), but how serious is this press?
Black's last line of defence is holding, It is hard to believe that it amounts to
but alas, this is an illusion. very much. Even so:
27 . ..txes dxes 28. ti::lg 4 24 ... .:Xc4!?
On 27 .. .fxe5 the same move would have This move is not an acknowledgement
followed. that Black stands badly. It shows only
28 ... l:fd8 29. Ag6! that he is happy with a draw and wants
The final trick. The pawn on f6 is to achieve this in the simplest (as far as
hanging and there is also a threat of he understands it) way.
30.tt::lx h6+ and 3 1 .tt::lfs . There is no 25.bxc4 e4 26. Ab2 Ab6!?
defence. Black resigned. An interesting moment. The knight
blockade seems tempting: 26 ... tt::lc 5. But
Returning to our earlier theme of sacri in that case, the bishop on d8 is out of
tieing the exchange to make a draw, we play. So Black first tries to exchange the
can note that an especially impressive bishop for the one on b2 , and only then
case comes when a player gives up the puts his knight on cS.
exchange without any dire necessity. 27. l:e2 ..tcs 28.h5 <iWB!
One great master of such sacrifices is Prophylaxis against a pawn break. Now
Viswanathan Anand. Even many open after 29.h6 there follows 29 ... g6, whilst
ing variations in his arsenal end with an after 29.g6, correspondingly, there is
exchange sacrifice, for a pawn, to estab 29 ... h6.
lish (for example) a permanent knight 29. Ad4
on dS and quietly wait for his opponent White effectively abandons the battle.
to offer a draw. His last chance to open the position was
And here is another example on this 29.c3 . Probably, Ponomariov decided
theme. this was too risky: 29 ... tt::lb 6 30.cxb4
axb4 etc.
Ruslan Ponomariov 29 ... hd4 30 . .:Xd4 ti::lcS
Viswanathan Anand The blockade is established, and after a
Mainz 2002 few more moves, the players agreed a
draw.
•
The positional sacrifice of the queen is

akin to the exchange sacrifice and usu
ally has the same motives. For example,
in a difficult position, the queen can be
sacrificed, to head towards a drawish
ending (such as building a fortress). In
an unclear position, the sacrifice of the
queen for a solid material equivalent can
be used to shake the opponent out of his
White's position looks to be only a little comfort zone, seize the initiative, etc.
better. He is attacking the eS -pawn, and The queen sacrifice has an even greater
is ready to increase the pressure on it psychological element than the sacrifice
1 05
of the exchange. Giving up one's strong

est piece can seem a terrible thing! One
would normally only do it in associa
tion with concrete variations. But the
opposite situation is also possible. The
opponent, when thinking about his
move, is likely to devote less time to
variations involving a queen sacrifice.
And this can cost him.
Lev Polugaevsky Black has rook and minor piece for the
Boris Gelfand queen. Formally, this is insufficient
Reggio Emilia 1 99 2 compensation. But if we look at the
position, we see that Black has very
rich counterplay, involving an attack on
0
the a2-pawn. The knight comes to b4
and the rook to d2. If Black succeeds in
taking on a2, then the passed pawn on
a3 is itself a potential queen.
A rhetorical question arises: surely Polu
gaevsky (a superb calculator, inciden
tally) would not have played 1 7.ikxf6
ikxf6 1 8 .tt:le4, if he had looked seriously
at the queen sacrifice? Of course not.
The battle has only just started in the 20. 'Wxe4
centre and the bishop on d4 is attacked. Also a very telling moment. We dis
In the game Adorjan-Horvath (Buda cussed the same point in the context
pest 1 992) there followed 1 7.tt:ldf3 'iVaS of exchange sacrifices. The player who
1 8 . ikc3 iVcS. Black has no problems. sacrifices does so consciously, in full
Polugaevsky played differently: knowledge of what is happening, but
1 7 . .bf6 .bf6 1 8.tlJe4 his opponent, on the contrary, is com
But he was seriously mistaken. What pletely shaken out of his tracks. He
was the reason for his mistake? One can often loses the thread of the game and
suggest that Lev Abramovich devoted starts to play badly.
all his time to calculating the variation Good or bad, White had to play
1 8 ... .txh4 1 9.cxd5 !? tt:lb4 20.'iVc4 tt:lxdS 20. ikxe4 and then transfer his bishop
2 1 .tt:lc5 , in which White sacrifices a to bl at all costs, to defend the a2-pawn.
piece and obtains the better chances. The following variation is possible:
The line with a queen sacrifice was 20 ... tt:lb4 2 l ."iVbl l::rd 2 22 .'iYel ikc3
treated as subsidiary and little attention 23.ikbl ikh3 24.tt:lg2 .l:Xad8 - White is
was paid to it, on the assumption that virtually stalemated, but Black still has
Black would not dare do this. to demonstrate that he has something
But Gelfand did dare: more than a repetition of the position,
1 8 ... dxe4! 1 9. :Xd8 l:fxd8 after 25 .'iVcl ikb2 26.'iVe l ikc3.
1 06
20 ... .l:.d2 21 .14 .:.Xa2 would add that the computer fully
After the loss of the a2-pawn, the game endorses the sacrifice. Indeed, is it really
is over. White launches a desperate a sacrifice at all? Surely, it is more of a
attack, which has no chance of success. transformation of the position: going
22.15 gx15 23. tlJx15 .l:.a1 24. tlJxe7+ from one with identical material to one
tlJxe7 25 . .:.Xa1 ba1 26. 'it'xb7 with a non-standard, but even so roughly
..td4+ 27.e3 he3+ 28.<itf1 .l:.a7 equal material balance. Black gets rook,
29. 'it'b8+ liJc8 knight and pawn for the queen, plus
And Black won. an elastic structure, in which it is hard
to find any weaknesses. Even such an
M iguel Najdorf enterprising player as Najdorf could not
Viacheslav Ragozin find a single point of attack or any active
Saltsjobaden 1 948 counterplay. The only thing White can
do is wait, exchange pieces and hope
• that the queen will eventually get the
chance to have her say.
20 . ..td2
On 20.i.e3 , White may not have liked
20 ... tbd5 .
20 ... tt:Je4 21 . ..te3 tt:Jd6
It seems as though Black is trying to
get his knight to c4 or, via the transit
square f5 , to d4. But his idea is deeper.
Having placed his pieces conveniently,
An example which has become an abso he wants to employ the e-pawn as a bat
lute classic. The last move was 1 7.'ifb3- tering-ram.
a3 . White intends to win a pawn, by And one more thing. Note how calmly
taking on e7 and then d6. But the e4- Ragozin plays, He never rushes things,
pawn is undefended, is it not? But that and does not try immediately to show
does not matter, as the pawn cannot be his opponent that there is sufficient
taken, since the check on f6 wins the compensation for the queen. Black has
queen. simple, convenient play and faces no
That is roughly how White's thoughts dangers at all. So why force matters?
ran. However, there followed Let the opponent suffer, trying to guess
1 7 ... tlJxe4! 1 8. tt:l16+ tlJx16 1 9.lhd8 where the final blow will fall.
.l:.1xd8 22 . .l:.c1 tt:l15 23. �14 �d5 24. �c4
This is what Grandmaster Viacheslav bc4 25 . .:.Xc4 e5 26. �g5
Ragozin said about the situation: Botvinnik made the valuable recom
«This position was assessed by many people as lost mendation to give up material and
for Black. Even among strong players, the word simplify the position: 26.l:txc6! ? bxc6
'queen' seems to exert a magical effect. However, in 27.i.xe5. Let us continue the variation:
this position, Black has the initiative, whilst over 27 ...tbh4 28. i.xg7 tLlxf3+ 29.'ifxf3
the next 25 moves, the queen stays where it is and �xg7 30.'ii'c 3+ �g8 3 I.'ii'xc6 �ac8
plays only a passive role.» 32.'ii'a4 .l::rc l+ 33 .�h2 .l::i.d 2 34.'ii'x a7
1 07
l::txb2 35.a4. While the black rooks are White could probably have defended
doubled on the a-file, dealing with the more tenaciously, but in the conditions
passed pawn, the white queen will of an over-the-board game, it is not
probably manage to give perpetual easy to hold the position. Everything
check on the squares b8, e5 and f4. is hanging, the queen is just a specta
Possibly Najdorf and other commenta tor and, most importantly of all, White
tors, hypnotised by the thought of 'the cannot create a single threat and is lim
strongest piece', incorrectly assessed ited to passive, humiliating defence.
the position and did not sense that it 30. tlJe1 l:l.ad8 31 . -tcs �e5+ 32.g3
is White, not Black, who should seek a l:l.d2! 33. l:l.c2 .bg3+ 34. <it>g2
draw here.
26 ... l:l.d1 + 27. <it>h2 h6
Now it is time to reap the harvest.

Botvinnik recommends 34 ... �xf2
28. l:l.c1 3 5 . �xf2 e3 36 .'iYc3 exf2 37 . .l:f.xd2
Finally losing control over the posi fxe l tD+! (but not 37 .. .fxe l 'iV 38 . .l::f.x d8+
tion. The transfer of the bishop to c3 is and 39.'iVxe l) 38 .�fl .l::f.xd2 39.'iVxd2
tempting, but after 28.�d2 .!:i.d8 29. �c3 tLlf3 , «and with an unusual material bal
there is the unpleasant 29 .. .lbd6. In the ance of three knights and two pawns against
long variation 29.'iVb3 .l:f.lxd2 30.tLlxd2 the queen, Black should win)). In fact, after
l:i.xd2 3 1 .'ii'xb7 tLld8 (not 3 1 ...tLlcd4 40.'iid 7, one can have doubts about
32 . .!:i.c8+ �f8 because of 33.l:rxf8+! this prognosis, but the variation is
Wxf8 34.'iVb4+ and 35 .'i!Yxd2) 32.'i!Yxa7 easy to strengthen: 36 ... .i:r.8d7, and then
:!xb2 33.l:tc8 �f6 (33 ... �d2 34.'iVa5) the e-pawn promotes almost like a
34.'iVd7 \tlg7 35 . .l:i.xd8 �xd8 36 .'iVxd8 draughts game .
.l:f.xa2 the most likely outcome is a draw, The simplest solution of all is 34 ...
though Black's play can be strength e3! 35.�xe3 tL:lxe3+ 36.'iVxe3 �xf2
ened. 37.'iVxf2 �xf2+ 3 8 .�xf2 fS or 37.'iVxd2
28 ... l:l.d7 29. �e3 e4! �xd2 38 . .l:i.xd2 �xel 39 . .l:i.d7 �aS
The first point in the plan is fulfilled. 40 . .!:lxb7 �b6 , and the endgame is
The knight is driven from the square f3 , in any case hopeless for White. But
and now Black only needs to remove or Ragozin, true to his strategy of not
exchange the bishop on e3, and then he rushing, played
will be able to invade on d2 with his 34 ... �e5
rook. and also went on to win.
1 08
Chapter 9
Standing on the Shoulders of the Classics
J u d it Polgar Emanuel Lasker

Anatoly Ka rpov Johann Bauer
Hoogeveen 2003 Amsterdam 1 889
i. t�, •
0
' ( l .t i. . ;A i
.
:.1
-l -��
��:
!J � �.
'ti' It .�.
25 . .bh7+! <iitx h7 26. 'ifh5+ 1 4. lbh5!
Without waiting for 26 ... �g8 27. i.xg7! Eliminating the knight on f6, the last
defender of the black king. In Zak's book
Lasker and Vainshtein's book Myslitel, the
following variations are offered:
14 ... h6 I S . i.xf6 i.xf6 1 6.tt:lxf6+ gxf6
1 7.�g4+ �h8 1 8 .�h4 <;t>g? 1 9.Ilf3;
14 ... d4 I S . i.xf6 i.xf6 1 6.�g4 �h8
(16 ... e5 1 7.i.e4!) 1 7.�f3 �g8 1 8 . i.xh7!
14 ...tt:le8 I S . i.xg7! tt:lxg7 1 6 .�g4 - in
each case with unavoidable mate.
The best defence was missed by both
commentators. It is obvious that this must
analysis diagram
be to move the rook from f8, so as to give
(and then one of three choices to get the king a flight square. Where should
mated: 27.. Jhe3 28 .�h8#; 27... �xg7 the rook go? If 14 ... .l:!.fe8, then I S.tt:lxg7!
28.l:!.g3+ <;t>f6 29.�g5#; 27 .. .f6 <;t>xg7 16.�g4+ �f8 1 7.�g5, and the
28. i.xf6! �xf6 29.l:tg3+ �f8 30.'iYh8+ knight cannot move because of mate in
<;t>f? 3 J .'g_g7#) , Black resigned. one. The only correct move is 14... �fd8,
when the same combination now makes
It is easy and pleasant to sacrifice a piece, no sense: I S.tt:lxg7 <;t>xg7 16.'iVg4+ �f8
when you are standing on the shoulders 1 7.�gs tt:le4 1 8.�g7+ <;t>es.
of such authorities as Lasker, Tarrasch After 14 .. Jlfd8 , Black's position remains
and Alekhine: inferior, as in the variation I S .tt:lxf6+
1 09
.ixf6 I 6 . .ixh7+ �f8 ! he loses a pawn, 1 8 ... d4!?

but at least it is not mate. Nimzowitsch was a cultured and
1 4 ... tillc: h 5 erudite player, who I am sure knew
Pay attention to the screen - the clas Lasker's combination. But here his
sic combination with the double bishop attention lapsed for a moment.
sacrifice will be seen for the first time: 1 9.exd4
1 5 . ..bh7+! �xh7 1 6. "ifxh5+ �g8 There were many ways to defend against
the threats to h2 and g2 , including these
three: I9 . .l:f.fe i , I 9.e4 and even I 9.g3
dxe3 20.�c3 . Instead, White remains
blind to the danger.
1 9 ... ..bh2+!?
«Here is a rare case where knowledge not only
brings no benefit, but even turns out to be harm
ful. If Tarrasch had not known the classical her
itage, he might, instead of I9 ... .ixh2+, have
found a less striking but more convincing way to
1 7 . ..bg7! ! �xg7 1 8. "ifg4+ win: 1 9... .ixg2! 20.�xg2 (on 20.dxc5, there
Chess is such a contrary game that is the decisive 20 ...�g5) 20 ...�g5+ 2 1 .Wf3
care is required even in the most over (if 2 1 .\t>h3, then 2 1 ...'it'h5+ 22.Wg2 'ifg4+
whelming positions. After I 8 . l:tf3 l::i.h 8 23.�hi �f4) 2 I ...kf.fe8! 22.kf.gl �f4+
I9 . .l::r.g3+ .ig5 ! White would have to 23.�g2 l:re2, and White would have had to
start from the beginning again. resign earlier than in the game.»
1 8 ... �h7 1 9. l:l.f3 e5 20.l:l.h3+ 'ifh6 There is in fact nothing to add to this
21 .lb:h6+ �xh6 22. "ifd7! extract from Neishtadt's book Siegbert
The final point - the double attack wins Tarrasch. I would only suggest you
one of the bishops, making further don't show it to children, as the sugges
resistance pointless. The fact that Black tion that knowledge can, paradoxically,
did not resign until move 3 8 testifies sometimes be a bad thing is something
only to his stubbornness. they do better to learn only in adult life.
But the spectators do not suffer here.
Aaron N i m zowitsch Tarrasch's method may take longer, but
Siegbert Tarrasch it is beautiful and long ago joined the
St Petersburg 1 91 4 classics itself. One can recall, for exam
ple, the game Spassky-Tal (Montreal
•
I979), where Mikhail Nekhemyevich
carried out the attack in the Tarrasch
manner, as well as hundreds of other
examples.
20. �xh2 1!fh4+ 21 . �g1 ..bg2!
Lasker would have been pleased.
22.f3
After 22.�xg2 �g4+ 23 .�h2 Ild5
24.�xc5 White does not get enough
I I0
Chapter 9 - Standing on the Shoulders of the Classics
for the queen, because Black has the get the chance to mate the white king
zwischenzug 24 ... l::i.h 5+! 2S .'ifxh5 on d7? This may be the only chance in
'ifxhS+, 26 ...'iYgS+ and 27...'ifxd2 . a lifetime!
29. �xf4 :Z.f8+ 30. �e5 'ifh2+
• 31 . �e6 :Z.e8+ 32. �d7 ..tb5#
Alexander Alekh ine

John Drewitt
Portsmouth 1 923
22 ... :Z.fe8!
The only move, but sufficient. The attack
runs out of steam after both 22 ...'ifhl+
23 .�f2 . and 22 .. ."ifg3 23 .tt:le4.
23. 0e4
23 .�xg2 .l::f.e2+.
23 ... 'ifh 1 + 24. c;i(f2 .bf1 25.d5 Black has sacrificed a piece for three
The best defence is counterattack pawns, the last of which he has just
and White, following this principle, captured on a3. Alekhine's reply shows
unleashes his queen + bishop battery on that he was not interested in playing for
the long diagonal. It does not help, but trifles:
other defences were no better: 2S . .!::i.xfl 20. hh7+! �xh7 21 . .l:h3+ �g8
'ifh2+; 2S .tt:lf6+ �f8 26.tt:lxe8 'ifg2+ 2 l ...�g6 n .'ifhs#.
27.\t>e3 l::i.xe8+ 28.<;t>f4 gS+ 29.\t>fs 22. hg7!?
'ifxf3+ 30.�xg5 f6+ 3 I .<;t>h4 'ifh3#. If Neishtadt had written a book about
25 ... 15 26.'ifc3 'ifg2+ 27. �e3 Alekhine, he might have had the chance
of another reference to the 'danger of
too much knowledge'. With the simple
22.'iVh5 f6 23 . .ixa3 'ifxa3 24."Viih 8+
�f7 2S .'ifxd8 'ifcl+ 26.tt:lfl .ibs
27.l:!.f3 , White wins a rook, but instead,
he prefers to go in for the Lasker com
bination. And Black, in reply... resigned!
It seems Drewitt was the opposite of
Bauer - the latter did not resign until
the very last moment, whereas the
27 ... :Xe4+! 28.fxe4 f4+ former resigns with a board full of
Cleaner is 28 ...'ifg3+ 29.�d2 'iff2+ pieces. He cannot accept the sacrifice of
30.'�d1 "Viie 2#, but perfection is not the second bishop: 22 ... �xg7 23."Viig 4+
really in place here. How often does one �f6 24."�g5# or 23 ... �f8 24 . .l:Ih8#.
111
«After the strongest move 22 .. .£6,» writes In an article in the magazine New In Chess,
Kotov in his book Shakhmatnoe Nasledie Boris Gelfand gave what in his opinion
Alekhina, «White wins with 23 . .ih6! 'i!Yh7 was the strongest continuation as being
24.'i!Yh5! .i£8 (24... .ie8 25. l::tg 3+ 'lth8 20.'ftg4 g6 2 1 .a3 . One can also consider
26 . .ig7+ and 27. .Lf6+) 25 .'i!Yg4+ 'lth8 the more active plan 20.'i!Yh5 fS (20 ... g6
26.�xf8». But what if 25 ... 'ltf7 ? White 2 1 .'i!Yh6 f6 22 . .ixg6!) 2 1 . .ic4, freeing
needs to find 26 . .ixf8 'i!Yg6 27.'i!Yxg6+ the square d3 for the rook, from where
'ltxg6 28.�h6+ 'ltf7 (28 ... 'ltf5 29 . .ie7) it is ready to transfer to g3 or h3.
29 . .ic5 'ltg7 30 . .l:lh3 with a gradual But the desire to play the Lasker combi
realisation of the advantage. nation was too great:
The cleanness of the combination in 20. bh7+!? �xh7 21 . 'iVh5+ �g8
the game Polgar-Karpov is something 22. bg7 �xg7 23. 'ifg5+ �h7
of an exception to the general rule. In 24J %d3
our day, it is practically impossible to The two bishops have happily given
carry out such a classic combination, their lives and the linear mate is threat
in such a pure form. Everybody knows ened, as in the classical examples. But
everything! And they take measures in this time Black finds a defence.
advance. The extent of defensive resist 24 ... tbe3
ance has increased many times over and The simplest, although after 24 ... �e3
the level of defence is incomparably 25.fxe3 f6 26.�h5+ 'ltg7 27J:tf3 fS , it
greater than that demonstrated by the is also hard to see any serious danger.
masters of the pas. 25.fxe3
Even if the outward contours of the

combination are similar to those of
the classic examples, the analysis often
reveals internal differences. But why
should this be a surprise? Life has
become much more complicated in the
2 1 st century and so has chess.
Boris Gelfand
Vladi m i r Kra m n i k 25 ... i.e4??
Munich 1 994 A one-move blunder, when the posi
tion is objectively close to equality. After
0
25 ...'i!Yxb2 White can give perpetual
check, or can bring in the rook from f1 ,
whereupon it is Black who starts check
ing. In addition, at the end of the variation
26 . .tf.f6 'i!Ybl+ 27.'ltf2 'i!Yxa2+ 28.'ltfl
'i!Vxg2+! 29.'i!Vxg2 �xg2+ 30.'ltxg2 'ltg7
we reach an endgame which White is
hardly in any danger of winning.
26. 'iVh4+ 1 -0
1 12
Chapter 10
Fancy some Solving?

My book, alas, is not entertaining fiction like a detective or romantic novel The
reader may be getting bored or tired, so it may be a good idea to take a break and
turn our attention to something else; to 'ventilate the brain', as they say.
I offer you 24 positions to solve. They are all, with only a few exceptions, very
simple and are divided into pairs, so that the solution to one position will suggest
the solution to the other. Later, we will return to our main theme.
Good luck!
A path for the rook

No t No 2
• •
Smothered mate
No 3 No 4
•
[]
1 13
Winning Chrss Manoeuvrrs
Multiple attacks
No S No 6
• •
The back rank

No 7 No S
Cl Cl
Don't be mean
No 9 No 10
Cl Cl
1 14
Chapter I 0 - Fancy some Solving?
An empty board and no place to go

No 1 1 No 12
This cunning discovered check

No ll No 14
0 0
All that glitters is not gold

No 15 No 16
•
0
1 15
The weak back rank

No 1 7 No 18
• •
A sprat to catch a mackerel

No 19 No 20
[J
Dying for one's comrades

No 21 No 22
•
[J
1 16
Chapter I 0 Fancy some Solving?
-
Living classics
No 23 No 24
0
0
117
Solutions
No 1
1 7 .'ifh5+! 1 8.gxh5 .l::f.h4, and White resigned (Hammer-Carlsen, Halkidiki 2003).
••
No 2
16 �h4! 1 7. �g2 �xh2+! 18 .l::f.xh2 l::r.g l# (Reiner-Steinitz, Vienna 1 860)
••• •
No 3
24 �xe3!, and now 2 5 .fxe3 tbg3+! 26.hxg3 l::[h 5# (Yudasin-Kramnik, Wijk aan
•••
Zee 1 994).
No 4
White wins prosaically after 37.tbd6 �c6+ 38 .f3 J::i.d 8 39.e7. But he found another,
more beautiful way of winning: 37.tbh6! Threatening 38.�g8+! Ihg8 39.tbgf7#.
Neither 37 ...gxh6 38.�xh7#, nor 37 ... l::r.e 7 38 .�xe7! saves Black, so he resigned
(Tal-Portisch, Biel 1976).
No 5
3 t . tbf3+! In view of 32.gxf3 �g5+ and 33 .. Jhd2 , White resigned (Gil Capape
••
Anand, Gausdal 1 986).
No 6
2 1 . i.f3! 22.gxf3 .l::f.xd2! 23.lhd2 �gS+ 24.�hl �xd2 Although he has no
••
material advantage, Black does have the initiative and went on to win (Fridstein
Smyslov, Moscow 1 945).
No 7
30.�xd7! l::txd7 3 1 ..l::f.e 8+ But definitely not 3 1 .l::tc 8+ �d8 . 3 1 . �h7 32.1d:cc8 I:f.d8
••
33 .l::f.exd8 Accuracy to the end: 3 3 . .l::f.cxd8 �cl + 34.�g2 gS , and Black can still
•
fight on, whereas after the text, it is all over (Alekhine-Colle, Paris 1 925).
No S
30.tbg6+! hxg6 It is easy to see that other captures lose. 31 .�h4+ �hS 32.�xd8+!
On 3 l ...�g8 , there would follow the same move. 32 .l::f.xd8 33. l:txd8+ 'it>h7
•••
34.�ee8. Black resigned (Mamedyarov-Timofeev, Moscow 2004).
No 9
2 1 . lhg7+! Mate is inevitable, Black resigned (Kramnik-P.Toth, Rio de Janeiro 1 991).
No 1 0
2 1 .l::f.xg7! �xg7 22.'iff6 + �f8 (or 2 2 ... �g8 23 .�xh6 f5 24.exf6) 23.i.g6 Black
•
resigned (Keres-Szabo, Budapest 1 955).
1 18
Solutions - Fancy some Solving?
No 1 1
61. f4+! Seeing 6Viiixf4 �d3 ! , White preferred to resign (Topalov-Kasparov,
•.
Linares 1 999).
No 1 2
48.'i¥d3+ �b4 49.'i¥a6! c4 The threat was mate in one. SO.'i¥b6+ �a4 S l.'i¥c6+
�aS Nor is Black saved after S l ...�b4 5 2 .c3+ �aS 53.a4 'i¥e2+ 54.�a3 . 52.a4!
Again threatening mate. 52 c3+ 53.�a3 'tte2 It looks as though Black is defend
.•.
ing, but there followed 54.'i¥b7! with the threat SS.'i¥a7+ 'iVa6 56.'i¥c5+. It is all
over (Brodsky-Toradze, USSR 1 9 8 1).
No 1 3
Can't we put the bishop on the a2-g8 diagonal, pinning the rook? Yes, if we calcu
late the variations very well!
30.�c4! On 30 ...'iVxh3 , White had prepared 3 1 .�xf7+ �h8 32.'i¥h7+! �xh7
33.�e6+ and 34.�xh3. But Black can play more strongly:
30 'i¥h4! Threatening 3 1 ...'i¥g3+, mating. 31.�xf7+ �h8 So who is trick
.•.
ing whom? 32Jhe3! White is tricking Black, after all. 32 'ifxh3 Or 32 ... �xe3
•..
33.l:txe3 l:i.xf7 34.'i¥g6 'i¥f4 3S .'i¥xh6+ �g8 3 6.'i¥xb6. 33.'i¥h7+! �xh7 34.�e6+
�g6 3S.�xh3 �xe3 36.lhe3 The realisation of the extra material did not involve
any difficulties (Zviagintsev-Topalov, Pamplona 1 995).
No 1 4
The start i s easy: 40.l:txg5! hxgS 41 .'i¥h7+ tbd7
But now what? After 42. �xd7 Black does not have to take the piece, when he
would lose at least the queen: 42 ...'ifxd7 43 .I:i.c4+ (more precise than 43 .'tixd7+
�xd7 44Jhb8) 43 ... �d8 44.'iVh8+ �e7 4S .l::f.xe4+ �f7 46.'i¥h7+ or 44 ...'i¥e8
4S .'i¥xf6+ �d7 46.'tig7+ 'tte 7 47.l:tc7+! . He can play more strongly: 42 ...'iig 8! and
the attack comes to nothing, whilst White is material down ...
42.�xd7 Even so. 42 'i¥g8! 43 . .l::rb7+!! Blow meets blow! 43 �xb7 43 .. Jhb7
.•. •..
44.'i¥xg8 �xd7 4S.'i¥g7+. 44.�c8+! �a8 4S.'i¥xg8 After a few moves, Black
acknowledged defeat (Mason-Winawer, Vienna 1 882).
No 1 5
1 7 .l::rxf2!?! The explanation for the punctuation is given below. 18.�xf2 'i¥a3!!
.•.
Definitely not 18 ... �a3 1 9.'i¥e3 ! . 19.c3 0fcourse, the queen cannot be taken: 1 9.bxa3
�xa3#. The attempt to free a flight on d2 for the king also ends badly: 1 9.'i¥c3
'i¥xa2 20.l::i.dg1 (20.�d2 �b4) 20 .. Jhb2! 2 1 .'i¥xb2 �f4+; 1 9.'i¥gs .l::txb2 20.'ttd 8+
�f7 2 1 . �h5+ g6 22.'i¥xc8 l:txa2+ 23.�d2 'ii'd 3+! 24.�e1 �b4+ 2S . .i:r.d2 �xd2+
26.�d 1 l:la1 # or 22.�d2 22 ... �b4+ 23 .�e2 �a6+ 24 . .l::f.d 3 'i¥xd3#.
19...'tixa2 The surreal 19 ... e3!? 20.�xe3 �fS fails to the no less amazing 2 1 .'ii'c 2!
'i¥xa2 22.�d3 . 20.b4 'iVai+ 21 .�c2 'ifa4+ 22.�b2 The decisive mistake, but easy
to explain. Black's attack has the force of a hurricane and White simply could not
take the pressure.
1 19
Wmning Chess Manoeums
He had to play 2Ht>ci!, and then the capture on M does not come with check:
22 ...LM 23.cxb4 lhM, which means that White has a tempo to start his coun
terplay: 24.1fgS!. The king is ready to come via d2 and e1 to a safer square, whilst
after 24...1fa2 there is 25.1fd8+ � 26.W'xc7+ ct>£8 27.W'xc8+ � 28.1i'd7+
'.t>g6 29.1i'd6+ and 30.1i'xb4. The chances of the two sides are unclear after 22 ...a5
23.W'c2 ea3+ 24.1i'b2 axM. In his book My Great PmleassoB, Kasparov recommends
going into the ending: 22 ... Afs 23.�1 1i'al+ 24.'.t>c2 e3+ 2S.'.i>b3 exd2 26.lha1
lle8 27.�6 dxell 28.llhxe1 lhe1 29.lhe1 hh2 30.i.b7 i.e4 31.hc6 '.t>f?.
underlining, however, that Black has only a minimal advantage.
After going through these variations, it makes sense to go back to the diagram
position. With the simple 1 7... i.g4 Morphy could secure a solid positional advan
tage. So, is that what he should have played? Probably, but then chess lovers would
have been deprived of 'one of the most amazing and deep combinations', as Geza
Maroczy said of this game.
22 hb4! 23.cxb4 lhb4+ 24.exb4 1fxb4+ 2S.'.i>c2 e3! White evidendy missed
•••
this resource. The bishop enters the attack, with decisive effect. 26.be3 .in+
27.11d3 ec4+ On 27.�3. the same move would follow. 28.'.i>d2 1i'a2+ 29.�d1
1!Vb1+ White resigned {Bird-Morphy, London 1858).
No 16
47.eS! fxeS 48.'1Ve4- 1Vg8 49.�6! 'Extreme measures', as in the previous game,
but this time they are effective, as well as striking. After 49 gxh6 S0.1i'xeS+ 'W'g7
•••
Sl.ltxb8+ W'g8 S2.1i'xb6 1i'dS+ S3.'.i>h2 1i'xd3 S4.1i'f6+ '.fi>g8 SS.1fb2 Black
resigned {Larsen-Botvinnik, Leiden 1970).
No 17
If his pawn stood on h3, White would probably win. Black has no normal move.
For example, after 26 ... i.g8 White wins with 27.1i'g4 l:l8e7 28.lhf6+ gxf6
29.W'xg8#, whilst after 26 ... d5 there is 27.lhf6+ gxf6 28.1i'xf6+ 1i'f7 29.'W'h8+
'.t>e7 30.W'xe5+.
But the pawn is on h2, and this small detail decides the outcome of the game:
26 1i'x6! 27.�6+ Relatively best; after accepting the sacrifice, White is mated.
•••
27 1i'xf6 28.rlf3 l:l8e6 Nothing remains of the attack. The two rooks are
•••
significandy stronger than the queen, and soon Black won (Kmoch-Reti,
Semmering 1926).
No 1 8
The end of a famous game. White had attacked inventively and sacrificed material,
but Black found a defence.
32 •£5! The threat was to capture on h6, whilst after 32 ...'.t>f8 33 . .l:r.xgS! Black
•••
could count only on a draw, at best: 33 ...1Vd3 34.1lg1 l:lbl 3S.'iVxh6+ '.t>e8 36.l:lg8+
'.i>d7 37.h3 lhgl+ 38.'.t>xgl %:t.xc3 39.l:lfs 1i'e3+ 40.1i'xe3 lhe3 41 .lhf7+ '.t>e6 etc.
33.'.t>gl The weakness of the first rank has its say. Not 33.lhfS .l::r.b l+, whilst after
33.'it'e2 the simplest is 33 ...'ifxfl+! 34.'it'xfl l:lcbS and 35 ... .l::r.b l . 33 'iVg6 34.1i'e2
•••
1 20
Solutions - Fancy some Solving?
llc6. Now, even if White regains the piece with h2-h4, Black still has enough extra
material to win, as indeed he did in the game Fischer-Smyslov, Bled 1 959.
No 1 9
31.1i'h6+! Both 3l ...�xh6 32.lDxf7+ and 33.tt:Jxd6, and 3 l ...�g8 3 2 .b5 are equally
hopeless for Black (Kramnik-Vescovi, Guarapuava 1991).
No 20
White has an extra pawn and could have won another with 26.�h8+! �xh8
27.lDxf7+ and 28.lbxe5. But in the game, there followed 26.�b2. Black, in his
turn, should have neutralised the threat with 26 ...�f5 (and if 27.e4, then 27 ... lbxe4!
28.tt:Jcxe4 f6 29.ttJh3 e5). However, he gave his opponent another chance: 26 �c6 •••
And White did not take it: 27.a3 A comedy of errors (27.�h8+!), and the most
amazing thing is that such an example of poor play should be taken from one
of the games of a world championship match (Alekhine-Euwe, Rotterdam 1 937).
This is what was written about it in the book devoted to the match, by the sixth
World Champion [Translator's Note: i.e. Botvinnik] : On move 25, Euwe committed a seri
ous mistake. By a queen sacrifice, Alekhine could have won a second pawn. However, he did not see this
possibility. On the following move, Euwe repeated his mistake, but again Alekhine missed it.
The public gasped: four mistakes in as many moves! The battle continued for another 28 moves, but the
balance was not disturbed again. Thanks to opposite-coloured bishops, Euwe managed to draw.
No 21
3l.ffi! After 3l ...�xf7 32."1Vh5+ �e7 33.�xe6+ �xe6 34Jlel+ �d6 35.'irc5+
�d7 36.1!ff5+ �d6 37.1!fe6 White gives mate, and after 32 ...g6 33 .'ifxh7+ �f8
34.1ifh8+ �e7 35.lhe6+, he wins the queen.
In the game, there followed 31 'ife8 32.lhe6 Vi'xe6 33.t2:Jd8 1i'e4 34.lbxc6.
•••
White has an extra pawn and an overwhelming position (Euwe-Alekhine, Zurich

1934).
No 22
27 tbxfl.! 28.1!fxf2 .ih.2+! It is too early for 28 ...Vi'xa4 29.i.xb8. 29.Wfl �xa4
•••
The bishop on h2 cannot be attacked (30.g3 �xh3+). the sacrificed piece will
inevitably be regained, and Black's two extra pawns are a very good argument for
resigning, as White indeed did {Ivanchuk-Topalov, Linares 1997).
No 23
The pawn weaknesses balance out and Black has the two bishops. It is not obvious
how he can be worse. His king is a little vulnerable, but the queens are off, which
means that mating attacks are impossible.
Or are they?
17.l:Hl ! The bishop has no retreats: 1 7... �g4 1 8 . i.f7+ and 1 9.i.e6+, whilst cover
ing the f7-square is no better: 1 7... i.g6 1 8 .tt:Jb5! cxb5 1 9. i.xb5#. That only leaves
17 1:1fs, but then again there follows 18.tt:Jb5! �c8 19.�f5! �xfS 20.�e6 cxbS
•••
121
21..ixc8. White has extracted the maximum out of the position: he has won a
pawn and subsequently realised it (Khalifman-Inkiov, Plovdiv 1986).
Knowing Alexander Khalifman's great erudition, one can feel sure that he realised
his combination here, knowing about the famous predecessor in the next example.
No 24
Chess historians know all about the events surrounding the following game. It was
played in 1 858 at the Paris Opera House, where Morphy had been invited by two
titled noblemen: Count Karl Brunswick and Count Isouard. They especially wanted
to play chess, whilst Morphy was more interested in listening to the music (we
even know which opera was on: Vincenzo Bellini's 'Norma').
Evidently, in order not to be distracted any more than necessary from his enjoy
ment, Morphy decided to dispose of his titled adversaries as quickly as possible,
and the latter helped him. The last move in the game was 9 ... b7-b5 . There fol
lowed:
lO.ltJxbS! cxb5 The lesser evil was to enter an endgame a pawn down: 1 0 ...�b4+
1 l .'tl!Yxb4 .ixb4+ 1 2 .c3 or 1 1 .4Jc3 . ll . .ixb5+ lt:Jbd7 Or 1 1 ...�d8 1 2.0-0-0+ r;i;;c ?
1 3 . �d3 with inevitable mate 12.0-0-0 l::td8 1 2 ... 0-0-0 13 . .ia6+ �c7 14.�b7#.
13.�d7! l::txd7 14 .l:tdl 'iYe6 15 .ixd7+ lt:Jxd7
• •
Now the celebrated moment, when the Duke and Count went into history.
16.'tl!Yb8+! lt:Jxb8 1 7. .l:td8#
1 22
Chapter 11
The Pawn Prod

The pawn prod is something exceptionally widespread, but also very badly used.
More often than not it is played when no other good move is visible, so we just
attack a piece with a pawn, for want of another move. Often it occurs when in
time-trouble - no time to think, no obvious move enters one's head, so just push a
pawn forward and what will be, will be.
A move for the sake of moving. But sometimes it is different. If there is an idea
behind the move it is usually to loosen up the enemy position. And sometimes, this
loosening can be the best way out of the position.
M i khail Botvi n n i k
Alexander Konstantinopolsky
Moscow 1 952
29.'iVg3!
A very strong move: the queen x-rays
the rook at c7. Combinational motifs
hang over the board. The position
deserves to be examined in detail.
Black has a solid, but passive posi Thus, after 29 ... tt::ld 7, there is the tempt
tion. Konstantinopolsky decided to do ing combination 30.gS "iVe7 3 l .fS ! ? and
nothing: if 3 1 ...tt::lxeS , then 32.f6+. But Black
1 9 ... .1:1.d7 20.c3 'ifc7 21 . 'iVg5 'iVdB replies 3 1 ..."iVd6! 32.fxe6 (32 .f6+ 'ltg8)
22. 'iVhS 'iff6 23.a4 b6 24 . .1:1.de1 32 ... tt::lxeS 33 .dxeS "iVd2 34.exf7 l:txf7!
.l:l.ed8 25.f4 .l:l.b7 26. 'iVf3 .l:l.c7 27.g4 3 S . J.xf7 .l::rd 3! (of course, not 3S ... �xf7
g6 28 . .1:1.e5 'iii>g 7 36.e6+) 36."iVf2 "iVxf2+ 37.'ltxf2 �xf7
While Black has been standing still, with good drawing chances.
White has advanced almost all of his Instead of 3 l .fS , more convincing is
pawns to the fourth rank. If he manages 3 1 .l:txe6! fxe6 32 .fS! and then:
to break through, then his approach 32 ... l:tcc8 33.f6+ tt::lxf6 34. t!.xe6 "iVc7
will be crowned with success, but if 3S.gxf6+ �h8 36 . .tt.e s cS 3 7."iVh4 cxd4
not, then Black will reap the rewards 3 8 . tl.e7 or 37 ... t!.e8 38 .f7;
of his cautious defence. Everything 32 ... eS 33.l:txeS! 'ii'f8 (33 ... tt::lxeS
depends on concrete variations. 34.f6+) 34.f6+ tt::lxf6 3S.gxf6+ "iVxf6
1 23
36.1::te 6 'iff7 37.'ifeS+ cJth6 38.l:f.xc6! This move also deserves to be noted.
with further material gains. It turns out that Black has loosened his
In the game, there followed opponent's queenside pawn structure
29 ... c5 30.bxc5 bxc5 and hampered the advance c3-c4, and
And here White missed the chance of a thus secured his knight a good post in
decisive breakthrough: 3 l .g5 'iYe7 32.d5 the centre. So is that job done, time to
tLld7 33 .f5! 'i:Vd6 (33 ... gxf5 34.l:f.xe6!) relax? No. Now it is time to pursue the
34.f6+ cJtg8 35 .dxe6! tLlxeS 36.e7!. next avenue of counterplay, the new
ly-opened a-file.
But let us return to the position in the 1 9. :Xa7 1i'xa7 20.c4 1i'a2! 21 .cxd5
previous diagram. Few people drew 'ifxd2 22.dxe6 fxe6
attention to Botvinnik's commentary, Black's position already deserves some
but it actually deserves the closest atten preference.
tion: «Striving for counterploy, Black should have At the moment of this game, Ruslan
continued 19... a5, which would have resulted in Ponomariov was all of eleven years old.
the opening of the a-file.>> Briefly stated, and Did he know the game Botvinnik-Kon
without an exclamation mark being stantinopolsky and the former's recom
put against the move ... a7-a5 . But it is mendation of ... a7-a5! ? I don't know, but
this move which solves all the defence's there is no doubt that Ruslan was born
problems. with an exceptional talent for positional
Now let us go forward 42 years. play.
Su rya Shekhar Ganguly Let us look at another modern example,

Ruslan Ponomariov to reinforce the idea.
Szeged 1 994
H i karu Naka m u ra
• Boris Gelfand
Moscow 201 0
The pawn structure is similar to the

previous example and Black's prob
lems are also much the same. Should he
wait, while White advances his pawns White has just retreated the knight from
and prepares a breakthrough, or should c3 to e2, further defending the pawn on
be prod the opponent, to distract him d4. He is clearly preparing the advance
from his plan? e3 -e4-e5, seizing additional space.
1 6 ... a5! 1 7.a3 axb4 1 8.axb4 :S7! 21 ... a5!
1 24
Chapter 1 1 - The Pawn Prod
As Gelfand explained after the game, the 1 9 ... b5!

idea of this move is prophylaxis. IfWhite This looks like a random pawn prod,
allows the exchange on b4, Black need but in reality, it is part of a deep stra
not only have no fear of the advance tegic idea. True, there are no knights
e4-e5, he can look forward to it, since on the board, which after the exchange
his knight will then come to dS, attack on c4 could occupy the dS outpost. But
ing the b4-pawn, which will no longer even so, the exchange on c4 is useful to
be supported by the pawn from a3. Black. In addition, he threatens ... b5-b4.
Therefore, White changed his plan: One can say that the white centre starts
22.bxa5 to crumble.
Now after the routine 22 ... .ixa3 , his The experts have been divided over
rooks quickly regroup to the open files: the best response for White. Thus, the
23.l:Ial .i.d6 24.l:Idb l , with unpleasant exchanging operation recommended
pressure. But Black was up to the task: by many, 20.cxb5 cxbS 2 l .d5 .ixc3
22 ... 'ifa7! 23.llb1 llc7 24.llb3 c5! 22.bxc3 exdS 23Jhd5 l::txdS 24 . .ixd5 ,
A timely break. Now not 25 .dxc5 is dubious. If we continue the variation:
because of 25 ... .ia4. 24 ... l::txd5 25 .'ti'xe8+ �g7 26.'ti'e3 a6
25 . .:.C3 'ifxa5 26.llc1 b6 27.h3 'i:Vc4 - the queenside weaknesses
The position has opened up and Black's could become an easy prey.
bishops have obtained some operating Yury Razuvaev, in his outstanding book
space. The defensive problems have on Rubinstein [Translator's Note: This too
been solved. And this all started with has never been translated into English]
the pawn prod ... a7 -aS ! shows that in the event of 20.cxb5 cxbS
2l .a3 'i:Vb6 «White loses the d4 pawn». That is
-
And now an interesting historical epi true, but this may nonetheless have been
sode: the Viennese master Hans Kmoch the best chance: 22.d5!? 'ifxe3 23.fxe3
lost two games within a year, to the exdS 24. .i.xg7 �xg7 25.l::td4. Now if
exact same idea. Forgetfulness about his 25 .. .f6 26.l::tadl .i.f7 or 25 ... l::td6 26.l:tadl
mistakes? Maybe. But what opponents .i.c6, then the pawn will be regained:
he was up against! 27.e4, whilst after 25 .. .f5 26.l:tcl ! the
piece activity offsets the small material
Hans Kmoch deficit. Instead of 23 ... exd5, stronger is
Akiba Rubinstei n 23 ... .ixc3 24.bxc3 exdS , but here too,
Semmering 1 926 after 25.l::td3 , and then 26.l::tadl , White
can count on a draw.
•
20.lld2
Unfortunately, White overlooks the
main threat.
20 ... b4! 21 . .bb4 hd4
Who would have thought the centre
would be demolished by the march
of the b-pawn? But that is what has
happened. Black's advantage quickly
increases.
1 25
Kmoch himself pointed out the The structure is practically identical

variation 22 .'i¥g3 �eS 23 . .l::f.xd7 .l::f.xd7 with the previous example (which, I
24.Vi'h3 .txb2 as completely hopeless would remind, was played just one year
for himself. One can recommend the earlier). The only differences are the
exchange sacrifice 22.I:r.xd4!? .l:rxd4 pawn on f4 and the absence of a light
2 3 . �c3 . But here Black does not retreat squared bishop, with which to cover
the rook - after 23 ... cS! he continues to the squares weakened by Black's pawn
dominate in the centre. jab.
22. 1i'f3 cs 23. �c3 .l:ld6! 24 . .1:lad1 22 ... b5! 23.1i'f3
es 25 . ..ic2 ..ic6 26. 11Vg3 It was possible to go over to deep
Not 26.�e4 .txe4 27.Vi'xe4 .txc3 . defence by means of 23.cS. But the end
26 ... 1i'b7 27 . ..bd4 exd4 28 . ..id3 would be the same. Kmoch preferred to
.l:le6 29.h4 .l:lde8 30.h5 aS! 31 .f4 lose quickly, but to have some fun .
11Ve7 32 . .1:lf1 .l:le3 33.1i'h2 f5! 23 ... bxc4 24. 'ifxc6 11Vxf4 25. 'ifxc4
The black position resembles a tank, e5
crushing every living thing in its way. Winning a pawn. There were also other
The white pieces are scrurrying into the solid moves: 2S ... .tes 26.g3 Vi'g4, 25 ...
corners in horror. h4 etc. In such positions, the main thing
34.hxg6 hxg6 35.1i'h6 1i'f6 36. '1fl>f2 is not to panic in the face of the many
11Ve6 37 . .1:lh1 tempting lines. One must just choose
Mate was threatened by 37 ... .l::f.xd3 one, calculate it accurately and not be
3 8 . 1:1xd3 Vi'e2+ 39.<iit>g3 Vi'xg2+ 40.�h4 distracted by other options.
Vi'g4#. In defending, White in his turn 26. 11Ve2 exd4 27 . .1:ld3
creates a threat on the open h-file. But
he is too late:
37 ... .1:lf3+!
After seeing 3 8.gxf3 Vi'e3+ with una
voidable mate, White resigned.
Hans Kmoch
Alexander Alekhine
Kecskemet 1 927
• 27 ... dxc3! 28 . .1:lxd7 .l:lxd7

In sacrificing the exchange, Alekhine
has calculated the variation 29.Vi'e8+
�h7 3 0.Vi'xd7 Vi'e4! and White
cannot go into the endgame: 3 I .Vi'd3
Vi'xd3 3 2 . !1xd3 c2 or 3 I .Vi'ds Vi'xdS
3 2 . I:f.xdS c2 3 3 . IlcS �d4+, whilst
after 3 I .Vi'xf7 c2 3 2 .Vi'xh5+ �g8 , the
checks run out and the c2-pawn again
queens.
1 26
Chapter I I - The Pawn Prod
Kmoch takes the line of least resistance 23 ... 'ita8! ? 24. <ita1 bS! 25 . .:b1 .:ba
and Black does not need to show his
class.
29.:Xd7 .i..d 4+ 30. 'ith1
Missing a mate, but after 3 0 . .l::f.xd4
'ifxd4+ 3 l .'it>fl 'iff4+ 32.'it>el 'ifxb4
everything ends very quickly.
30 ... 1Wc1 +
White resigned.
Players of later generations have taken

this device into their arsenal. They have
not just routinely copied it, by playing It seems as though Black is thinking
... b7-bS in the same structure, but have about attacking the king, but in real
also enriched it with some new points ity, this is only a bluff. His real idea is
and ideas. to play against the enemy pawn weak
nesses, and his attacking feint success
Anatoly Kremenietsky fully manages to camouflage his true
Evgeny Vasi u kov intention.
Moscow 1 981 White immediately blunders a pawn:
•
26.tlJe1
In fact, there was no cause for panic. He
could play 26 . .l::I.ddl or even 26.l:tdb2 .
Black also has his problems: for exam
ple, the rook on h8 is tied down and has
to passively defend the h6-pawn. Now,
however, it is all over:
26 ... bxc4 27.bxc4 :Xb1 + 28. 'itxb1
'iVb4+ 29 . .:b2 1Wxc4
with an easy win.
127
Chapter 12
Doubled Pawns are Cool!

When a player accepts doubled pawns, he does his pawn structure serious and
possibly irreparable damage. The doubled pawns can become weak and his overall
pawn structure spoiled.
Usually, a player is forced into such a thing. But there are also occasions when he
initiates the doubled pawns himself For what?
Garry Kasparov Sarnuei Reshevsky

Anatoly Karpov Reuben Fine
Moscow 1 985 Hastings 1 937
• •
This was game 1 3 of the World Cham 1 4... •f6 1 5. •ea :td8 1 6.e5 •hs!?
pionship match. White is threatening The essence of the idea is the same:
mate directly. after 1 7.1fxh6 gxh6 the weakness of the
1 8 ... 11ff6! 1 9. 1i'e3 11fh6! 20. 1i'xh6 kingside pawns is offset by the activity
gxh6 21 J:lfe1 �c4 22.a3 b5 of Black's pieces.
After just a couple more moves, the Reshevsky declined the present with
players agreed a draw. 1 7.Dac1
After the game, all the experts hailed but after 17...1!fxe3 1 8.f:xe3 Ilac8 the
the manoeuvre ...'iVd8-f6-h6, praising battle did not last much longer. Draw.
its strength and originality. Just think:
to voluntarily shatter one's pawn struc In reality, these two previous examples
ture, just to exchange queens. Very were not so complicated. The main dif
original! ficulty lay in considering a move which
leads to the doubling of the pawns,
But look at the next diagram. It tran and including the move in one's list of
spires that almost fifty years before the candidate moves. With careful atten
Karpov-Kasparov match, people could tion, the soundness of the manoeuvre
already play chess very well. ...li'd8-f6-h6 becomes so obvious that
1 28
Chapter 1 2 - Doubled Pawns are Cool!
the advantages outweigh the draw he opens the b-file for the white rooks. It
backs. Karpov doubled his pawns? So is clear that the queen's rook will come
what, if the game goes from a difficult to b1 , whilst the king's rook also has a
middlegame into an ending with oppo chance to use the b-file - after tt::le 4-c5
site-coloured bishops, well-known for and the exchange on cS , the rook can
their drawing tendencies? Fine did the swing over to b4.
same? Well, just try yourself to find Even so, it is hard to believe that these
a way to get at the doubled h-pawns, threats will be sufficient compensation
when all the play is taking place in the for the spoiled pawns, which makes it
centre and on the queenside, where all the more interesting to follow the
Black is, at a minimum, not worse. course of the game.
1 6 ... tlJgf6
But now some cases where a player More natural is 16 ... �e7. True, after
takes on doubled pawns, not to escape 1 7.tt::lc 5 tt::lgf6 (1 7... tt::lx c5 1 8 .dxc5, and
with a draw, but in order to play for a the rook swings over to b4, as already
win. How can this be? How can dou explained) 1 8 .tt::lxb7 we have trans
bled pawns give an advantage? The next posed into the game. There is also the
three examples will look at this. possibility of 1 7.1:1bl b6 1 8 .tt::le 5!? tt::lxeS
1 9.dxe5 (Beliavsky-Pomar Salamanca,
Ruslan Ponomariov Las Palmas 1 974). The lack of develop
Holger Grund ment of the black pieces is all the more
Rimavska Sobota 1 996 noticeable, as are the weakness on g7,
and the open d-file... Do you not get
the impression that the endgame has
turned in White's favour?
1 7.ttJd6+ <t;e7 1 8. lillc: b7 l:tab8
Possibly more accurate is 1 8 ... .l:lhc8
1 9. .l:tbl l::rc 7 20.l;Ib3 �ac8, as in the
game Larsen-Addison, Palma de
Mallorca 1 970.
1 9. tt:\a5 .l:.hc8 20. ltJe5 lillc:e5
21 .dxe5 ttJd7 22 .1:.b4!?
.
The position looks about equal. Black, On 22.tt::lc4, White may not have liked
as usual in this opening (a Caro-Kann) , 22 ... .l::i'.b5 .
has no bad pieces, nor pawn weak 22 .. .1:.xb4
.
nesses. He is just a bit behind in devel The alternative is 22 ... tt::lxe5 , and if
opment, but it is far from obvious at 23 . .l::i'.xb8 .l::txb8 24.f4, then 24 .. Jlb5 .
first glance, how to exploit this. White would probably have continued
1 5. 'ifa3!? 'ifxa3 1 6.bxa3 23 . .l:!.ab l .
Thus, White has voluntarily accepted 23.axb4 lillc:e5 24.0-0-0
doubled pawns. Again, what for? The position has finally clarified.
On the surface there are two ideas. Ponomariov has returned the extra
Firstly, he threatens a check on d6, fol material, in return for which he has
lowed by taking the b7-pawn. Secondly, undoubled his pawns.
1 29
The pawn structure is symmetrical and

cannot give either player the advantage.
White is one or two tempi behind in
development, but because the centre is
closed, this is a factor that can hardly
be exploited. That leaves the only other
factor - the open c-file. However, there
are no entry squares along it.
How can Black obtain the advantage?
1 9 ... 'ifa6!
How should we assess the endgame? This move contains not one, but two
The weaknesses on a7 and c6 require ideas. With the queen exchange, a
defence, and the king on e7 is cut off defender of c2 leaves the board, and
from the queenside. Even more impor so the black rook threatens to enter on
tantly, it cannot be brought across: the seventh rank. And secondly, the
after 24 ... .l:Id8 White wins with both b-file is also opened, and this means
25.l:txd8 'ltxd8 26.f4, and 25 . .l:tel �d5 that the second rook also acquires work
27. �xe5!. - it comes to b8 and will attack the
White's clear advantage does not mean b3-pawn.
that he doesn't have his own prob The reader may rightly respond that
lems (in particular, the h5-pawn needs the loss of the b3 -pawn is unlikely to
defending). But White's position is be fatal, since Black's extra pawn would
rather more pleasant to play. He has be the doubled a-pawns, which are
pressure and it is unlikely he would unlikely to produce a passed pawn. This
have obtained such a pleasant position is true. But it is also true that the differ
without the move 1 5 .'iYa3 ! ?. ence in activity of the pieces (especially
the rooks) after the queen exchange
Such seemingly strange moves were becomes great, and this difference in
part of the arsenal of players even in the activity is more important than the
1 9th century. doubled pawns.
20. 1!fxa6
Louis Pau lsen White goes along with fate, simply
Siegbert Tarrasch through a lack of alternatives. He loses
Nuremberg 1 888 after both 20.'�e2 l::tc2+, and 20.tbel
•
'iYxd3+ 2 I .tbxd3 �c2 22.'�g2 .!::i.d2
23.�hdl tbe3+.
20 ... bxa6 21 . 'ifilg2 Ac2 22. �c1
Ab8 23. Ab1 Ac3 24 . ..td2 Acxb3
25. :Xb3 :Xb3 26. ha5 Ab2!
Avoiding a positional trap, which was
not simple, by the way. In reply to
26 .. J:la3 , Paulsen had prepared 27. .l:lcl
�d8 28. �xd8 �xd8 29.�b l ! and the
white rook is becoming active. Now,
1 30
however, Black is prepared to meet Now, however, it is all the other way
27 . .!:i.cl with 27 ... ti:Je3+ 28. <.t'g1 tDc4!. round. It is the black rook which is the
27. .td2 i..b4 first to occupy the b-file and Black's
minor pieces which, like a swarm of
locusts, start devouring the opponent's
territory. The piece activity is more than
sufficient compensation for the doubled
pawns.
For the sake of fairness we should point
out that, unlike Paulsen, Campos' hand
was not forced with the exchange. Both
20.J::i.c l ltJcS 2 I . .¥t.d l , and 20.tDa3 were
stronger than White's choice in the
The position has clarified in Black's game.
obvious favour. The pawns on a4 and 21 . ttJd2 AbO 22 . .tc2 Ab4 23. tt::lb3
d4 are hopelessly weak. The endgame tt::lc4 24. Af2 ttJdb6 25 . .te1
is winning.
And here is a contemporary version of
this classic.
Jose Campos
Veselin Topalov
Palma de Mallorca 1 99 2
Evidently, White had only counted on

2S .. J ixa4 26.l:i.xa4 tDxa4 27.tDxaS etc.
But a surprise awaits him.
25 ... tt::lxd 5! 26.exd5? !
And he at once commits an oversight.
It is clear 26 . .¥t.xb4 ti:Jxb4 is bad, but
26.tDxaS tDxaS 27.exd5 would have
1 9 ... b6! ? allowed him to hold the game.
A s i n the previous example, I t I S not 26 ... :Xe2
about passed pawns, but piece activity. But now Black cannot be stopped.
After 1 9 ...'ifxd2 20.ti:Jxd2 Black would Topalov won in just ten more moves.
only be fighting for equality. White has
a clear plan: press with a4-a5, transfer Another class of positions is where
his rooks to the b- and c-files and create a player voluntarily accepts doubled
pressure with his minor pieces on the pawns in the centre. As a rule, these
queenside. examples are complicated and debata
20. 'Wxa5 bxa5 ble. It is usually difficult to determine
131
at first whether the pawns will be weak b5). As far as open files for the rooks
or strong. Which they are depends on are concerned, now the e-file is a draw,
circumstances. whereas the newly-opened c-file will
belong to Black.
Evgeny Alekseev
Ruslan Ponomariov The plusses of the doubled pawns are
Artek 1 999 obvious. And the minuses? One can
speak of the theoretical weakness of the
doubled pawns, but it is quite unclear
0
how they can actually be attacked. The
pawn on d6 is completely safe, and
only the d5-pawn can be got at, by
placing the knight on e3 or f4, and the
bishop on f3 . However, how is this to
be achieved in a practical game? These
manoeuvres require quite a few tempi
and the opponent is hardly likely just to
wait, while White regroups his pieces.
The game has not yet left the open In the subsequent play, the pawns on d5
ing. The e5-pawn is hanging, so White and d6 never once came under attack,
played and Black won the game.
1 1 .exd6
In reply, there is the tempting l l ...�xd6 M i khail Botvi n n i k
1 2 .0-0 tbf6 , but after 1 3 . .if4 "i¥b6 N i kolay Sorokin
14.�el White's position is the more Moscow 1 931
pleasant. He has simple piece play, with
the e-file for his major pieces and the
0
e5-outpost for his minor pieces. There
are also several prospects for attack.
More principled is 1 2 ... e5!? 1 3 .dxe5
tbxe5 14.tbxe5 .ixe5, but here too,
there are questions remaining (say,
after 1 5 . .ih6). Playing with an isolated
pawn is not to everybody's taste.
1 1 ... exd6! ?
Now, strangely, there are significantly
fewer open questions. There followed: 20. 'ife3!
1 2.0-0 ll.Jf6 1 3.b3 �d7 In Botvinnik's words:
And the position from the opening is «This far from obvious move is the strongest. After
clarified. White has no e5-outpost, the exchange of queens, which Black cannot avoid
and it is more likely that Black will (20 ..."i¥c7 2 I .tbxe5), the defects in Black's posi
threaten to put his own knight on e4 tion are more striking. The pressure on the d-file is
(either at once, or after the exchange difficult to meet, the e5-pawn becomes weak and
of light-squared bishops with ... .id7- the square f7 is also in need of defence.
1 32
In the time that I had been playing chess, this was For the moment, we have a sharp
the most subtle positional move I had managed to middlegame on the board, and in the
play.» middlegame, it is hard to overstate the
20 ... 'ffxe3 21 .fxe3 �g4 22.a5 tbc8 advantages of the doubled pawns. By
23 . .l:.c1 ! bf3 24.gxf3 moving from f7 to e6, the f-pawn has
The position has clarified in White's covered the squares dS and fS , opened
favour and after the first time control, the f-file and created the conditions for
he won. possible combinative play. One cannot
The example quoted is the most well ask more of a single move.
known in the sixth World Champion's 1 9.b4?
play. I recommend the reader himself Fischer loses his head, a rare thing
study the other examples: Botvinnik indeed! The great American GM had
Chekhover (Leningrad 1 93 8) and so lost control of the situation that he
Botvfnnik-Kan (Moscow 1939). not only fails to prevent Black's com
bination, but actually does everything
Robert Fischer to ensure that it takes place in the most
Ratm ir Kholmov favourable possible circumstances.
Havana 1 965 1 9 ... ttJd4!
The last move was 1 7.tLlf3-g5 . White's After this, one can see Black's idea in all
idea is shown by the sample varia its glory.
tion 1 7... �c5 1 8 .tLld5 ! .ixdS 1 9.exd5 Let us return to the diagram position. In
llxdS 20.tLlxh7! tLlxh7 2 I . .txh7+, and the game Matulovic-Mecking (Sousse
both 2 I ...c;itxh7 22.'ife4+ followed by 1 967), Black rushed things, playing the
23 .'ifxd5, and 2 l ...�h8 22.'ifh5 are bad. immediate 1 7... tLld4?! 1 8 .cxd4 exd4.
1 7 ... h6! 1 8.tillc:e6 fxe6 There followed the unforeseen 1 9.e5!
How should we assess the changes d3 20.tLlxe6 fxe6 2 l .exf6! dxe2 22.fxe7
which have taken place in the posi 'ifxe7 23. l::tx e2 . The three minor
tion? Theoretically, the doubled pawns pieces are certainly not weaker than
on e6 and eS are weak. In a simplified the queen, and it is White who has the
endgame, they could be attacked (espe advantage.
cially the eS-pawn), but the endgame is In analysing this variation, we can
something White, as they say, still has appreciate how seriously Fischer erred.
to survive to. 20.cxd4 exd4
1 33
He already is unable to use Matulovic's In order to break the pin on the knight
idea 2 l .e5 d3 22.exf6 because of the at f3 , Bagirov played
simple 22 ... �xf6. With the move b2-b4, 1 5.h3
Fischer opened the long diagonal, and Now it is bad to play 15 ... �xf3 1 6 .�xf3
now in this variation not only his queen 0-0 1 7. .l:idfl or even the simple 1 7.tt:ld5
is attacked, but also the rook on a I . tt:lxd5 1 8 . �xd5 . The difference in
Therefore he was forced to play: activity of the bishops is too noticeable.
21 .a3 Probably, White was expecting 1 5 ... �d7
But after or 1 5 ... �c8 , but then Black's shortage of
21 ... d3 22. i.xd3 :Xd3 development would be even greater.
Black had a firm grip on the initiative. Spassky decided instead to change the
Kholmov won the game. course of the game quite sharply:
1 5 ... �e6! ? 1 6 . ..be6 fxe6
Many years later, Ratmir Dmitrievich White could now win a pawn with
told how he had broken his pre-game 1 7.tt:lh4 �f7 1 8 .'iVg4 g5 1 9.'iYh5+
routine before this encounter and was �g8 20.'iVg6! 'iVf7 (20 ... gxh4 2 l . �xf6)
reproaching himself for this, and how 2 I .'ifxf7+ �xf7 22 .tt:lf3 . It is hard to
greatly Smyslov had come to his aid. say why he refrained from this varia
According to Kholmov, the precious tion. Perhaps he thought he was selling
novelty 1 7... h6! ? was shown to him by himself too cheaply - with the fall of
Vasily Vasilievich, together with the the e5 -pawn, the bishop on g7 is acti
subsequent knight sacrifice on d4. vated and this gives Black some hope of
counterplay.
Vladi m i r Bagirov Whatever the reason, White decided
Boris Spassky to retain the queens and bank on the
Leningrad 1 963 attack. This was a logical decision, but
one which failed to justify itself.
1 7.g4 0-0 1 8. c;ftb1
0
Or 1 8 .g5 tt:lhs 1 9.'iYe3 tt:lf4 - as often
happens, the compensation for the dou
bled pawns involves an outpost on the
newly-opened file.
1 8 ... b5 1 9.'ife3 ti::d 7 20.ti::d 2 tlJb6
The position is unrecognisable. White's
initiative has come to nothing and Black
is now developing counterplay. At the
very least he is not worse, and he went
Glancing at the diagram, it is not hard on to win the game.
to conclude that White is better. He has For Bagirov, the idea of Black's volun
completed his development, his rooks tarily accepting doubled pawns evi
are operating on the open files and dently came as a complete surprise. One
the bishop on c4 is good. Black has a can only imagine how shocked a player
poor bishop on g7 and he is 2-3 moves would have been by this idea, in a pre
behind in development. vious era.
1 34
Alexander Meek
Pau l Morphy
D
New York 1 857
A Schliemann/Jaenisch Gambit. Modern

theory happily includes the move
a . ..te3!?
«To this day, Morphy is the undisputed master with which White voluntarily invites
of open positions. His significance is so great that ... doubled pawns. The main reply is
there have been no major developments in this area 8 ... �b6. If, however, Black agrees to go
since him. Every player, from beginner to master, along with his opponent's idea
should return again and again to the games of the 8 ... he3 9.fxe3 0-0 1 0 . ..tc4+ �h8
American genius» Botvinnik.
- 1 1 . ltJg5 ..tg4 1 2. "ffe 1 !
8 ... ..te6!? 9 . ..txe6 fxe6 1 0.0-0 0-0 then in the opinion of grandmaster
In this situation, there is no question of Khalifman, White has a strong attack.
the doubled pawns representing any sort
of weakness. On the contrary, the unit Finally, a trio of examples in which the
d6-e6-e5 forms a powerful pawn group psychological aspect outweighed the
in the centre. Even so, Morphy's oppo chess aspect.
nent gets it into his head that he should
punish Black for his 'incorrect' play. Levon Aron ian
1 1 . tbg5?! "fie7 1 2. <�h1 h6 1 3. tbh3 Sarhan G u l iev
As was to be expected, White's attack has Istanbul 2003
resulted in nothing except lost tempi.
1 3 ... d5
A modern GM would probably have
gone in for the plan of 1 3 ... g5 , and then
.. YJiie 7-g7, .. Jlf8-f7, ... .l:ta8-f8 , keeping
central play in reserve.
1 4. b3 tbd8 1 5.b5
Now the game departed somewhat
from its original course, although in the
end, Morphy won anyway.
1 O.exd6 'ifxe2+ 1 1 . ..txe2
How much chess players' understand Here, the theoretical reply (so as to
ing has changed in the last 1 50 years is avoid doubled pawns) is 1 1 ... cxd6. But
shown by the following example. this is just self-delusion! In this case,
1 35
the pawns on a7, c6 and d6 are clearly passing the time control, Black con
weak. But after ducted the game to victory.
1 1 ... .hd6! One wonders whether Chernikov
the weakness of the queenside pawns is would have given up the bishop, had he
not felt at all. The first thing one noticed considered the reply 24 ... gxf6!?.
instead is the active black pieces and his
lead in development. Henrique Mecki ng
1 2. ttJd2 0-0 Vlasti m i l Hort
1 2 ... .te6 was worth considering. San Antonio 1 972
1 3.lZ:lf3 :ea
(and here ... a7-a5 -a4 looks like a worthy •
alternative) White felt obliged to agree a

draw at once. Objectively, Black's posi
tion is even preferable.
Oleg Chern i kov

Anatoly Karpov
Kuibyshev 1 970
D
Hart writes that he had foreseen from
some distance the variation
22 ... lLlh5 23 . ..txe7 :Xf3 24. 1i'xf3
24... tt::lf4, apparently with decisive
threats. But suddenly he saw the refuta
tion: 25 . .td6! tt::lxd3+ 26 .'i¥xd3 ! ! .
A lovely, unusual combination, which
it would be no sin to overlook. But why
did Black get involved in calculating
Having calculated the variation 24 . .txf6 such a complicated variation? Why not
.txf6 25.tt::le 4 .te7 26.�c7 l:f.d8 - settle for the simple 24 ...'i¥xe7 ?
26 ...'i¥e5!? - 27.'i¥c2 with the initiative, Evidently, it was a matter of psychology.
White decided to give up his long-range Black did not like the fact that after
bishop: 24 ... 'ifxe7 25 . .hb5 axb5
24 . .hf6 he would have two pairs of 'weak' dou
The reply came as a cold shower to him: bled pawns. Only after going through
24 ... gxf6!? 25. lLle4 f5! 26. tlJc3 all of the complicated variations did he
Nothing is changed by 26.tt::lc 5 .l:f.d8 return to the simple line, and having
27.'i¥c3 b6. done so, he realised that he could still
26 ... 'ifc6 27. lLlb1 live with the two doubled pawns.
Or 27.tt::l e2 'iia4 28.'i¥b2 .l:f.d8. 26.g3 lLlf6 27.'ild3 'ifc5 28.ttJd2 :C7
27 ... 'ife8 28.l:l.c4 ..tf6 29. 'ifc2 l:l.d8 Now it became obvious that the worst
30. 'iPf1 1i'b5 was behind Black.
In a few moves, the position on the 29. lLlb3 'iff2 30. 'ifd2 lLlxe4
board has changed completely. After And now he already had an extra pawn.
136
Chapter 13
Alekhine's Nail
tion of dynamic equality: 23 .tLlh6+

�h8 24.'iVh4 f5 (the square h7 is
Garry Kasparov threatened and the diagonal must be
M iguel Najdorf closed, but not with 24 ... g6? on account
Bugojno 1 982 of 2 5 . �d4) 2 5 . �d2 "i¥f6 26Jhe8
'iVxal + 27.�g2! (stronger than 2 7Jiel)
27 .. Jhe8 28 .tLlf7+ �g8 29.tLlxd6. The
0
forcing variation continues. Weak is
29 .. J:f.f8 30.tt.Jxc8 J:lxc8 3 1 . �xf5 with
a double attack on c8 and h7; after
29 ...'iVe5 30. �f4 further material losses
are also unavoidable. That only leaves
29 ...'iixa3 3 0.tLlxe8 'iVxd3 3 1 .'iVd8
'iVe4+, and White has a choice: agree
to perpetual check with 32 .�h3 'iVg4+
33 .�g2 'iVe4+ 34.�fl 'iVbi+ 35 .�g2
1 7.d5!? exd5 1 8.e5! 'ife6 1 9. l2Jd4 (only not 35 .�e2?? �a6+) 35 .. .'�e4+
'ifxe5 20. tLlf5 �f6 or continue to play for a win by means
Commendable fearlessness over the of 32 .f3 'iVe2+ 33 .'lt>h3 .
previous three moves gives way to an 21 . 'it'g4 l:tce8 22. �d2 'ii'x a1 ?
untimely burst of passivity. More inter The decisive mistake. After 22 ...'ii b2
esting is 20 ... �d6!? forcing his oppo White would have to show that he has
nent to expend a tempo on 2 1 .g3 , and more than a repetition of moves, in
then 2 1 ...�ce8 22 .'iVg4 �c8 . the variation 23 .tLlh6+ �h8 24.tLlxf7+
�g8 etc.
23.:Xa1 ba1 24. ttJxg7! .txg7
25. �h6
The check on el is not mate (26.�fl),
so Black resigned at once.
The break in the centre with d4-d5 , and

after the capture, the follow-up e4-e5!,
looks so natural in such positions that
it seems it has been around forever. But
analysis diagram that is not so - every chess idea must
White has the attack, Black two extra have its source game.
pawns. Who is better? Maybe in this case, it should be sought
It may be that we have before us a posi- in the following game.
137
Pa u l Keres After 2S .ti:JgS , Black has numerous can

Reu ben Fine didate moves: 2S ... h6, 2S ... g6, 25 ... 1ids ,
Ostend 1 937 and a number of others. Even in analy
sis, it is hard to achieve absolute clarity.
Keres considered the strongest defence
0
to be 2S ...'ii'c 3 !? getting the queen
away in advance from a knight fork on
e6 and setting up a double attack on
b3 and e 1 . It seems he was correct. A
possible development of events then is
26.1ixe6+ (White has nothing in the
variation 26.'ii'x h7+ �f8 27.ti:Jxe6+
�e7 28.'ii' h4+ 'ii'f6) 26 ...�f8 27.�fl !
(27.'ii'f4+ \te7 28.'ii'f7+ �d6)
21 .d5! exd5 27 ... �c7 28 .ti:Jxh7+ �e8 29.'ii' h S+
In reply to 2 l ...eS, White had planned �e7 30.'ii'g S+ �e8 (also interesting is
22.ti:JgS , so as then to try to play f2-f4. 30 ... �xe6 3 1 .'ii'xd8 �c8 32.'ii' h4 e3!?)
22.e5! 3 1 . 1ig4, and we have gone from one
«By driving off the only black piece which defends unclear position to another.
the king, White obtains a strong attack, which can 23. ttJgs tt::lta
hardly be defended in a practical game» Keres.
- More principled is 23 ... h6! after which
This note shows how objective Paul there is no danger in 24.ti:Jxf7 �xf7
Petrovich was in his judgement of the 2S .e6+ �g8 26.exd7 �xd7. Probably,
resulting position. White gives up a Fine did not like the possibility of
pawn and is ready to part with more 24.e6! hxgS 2S.exf7+ �xf7 26.�e7+.
material. He has a strong attack. In White's attack looks very dangerous.
analysis, there is a defence for Black, Thus, after 26 ... �g6, he mates beau
but solving the problems at the board tifully: 2 7.'it'd4 'ii'c 3 2 8 . 1ic2+! \ths
is very difficult. It is not surprising (or 28 .. .'iYxc2 29Jhg7+ �hS 30.g4+
that even such a strong practical player �h4 3 1 .'ii'e 3! �xg4 32.'ii'xgS+ �f3
as Fine (an acknowledged master of 3 3 . l:te3#) 29.'ii'd l+! (it is hard to
defence, incidentally) should be unable believe, but in the variation 29.1idl +
to cope with the problems and should �g6 30.�xg7+ �h6 3 1 . .l::te 6+ ti:Jf6!
make three mistakes in a row. White's major pieces cannot coordi
22 ... ttJd7 nate their strength) 29 ... �h6 (29 ... g4
The alternative is 22 ... ti:Je4, and then: 30.h3) 30.�1e3! 'ii'xc2 3 1..l:th 3+ �g6
23 .e6! fxe6 24Jhe4! dxe4 2S .ti:JgS . This 32.'ifhS+ �f6 3 3 .'ii'f 7#.
is the key position. Instead of 2S .ti:JgS , However, another retreat of the king,
weaker is 2 S . 1ixe6+ �h8 26.ti:Jgs h6 26 ... �g8 , appears to hold the posi
(26 ... g6 27. 1if7 ! , and the maximum tion. White can force the win of the
White has is perpetual check after queen with 27.'ii'xgS 'ii'c3 28.1ixd5+
27.ti:Jf7+ �h7 28 .ti:JgS+ \th8 (but 1ixdS 29.'ifxdS+ �h8 30.'ifhS+
not 28 ... �g6 29.1if7+ \tf6 30.ti:Je6+) �g8 3 1 .'ii'f7+ �h8 32.�7e3 , but after
29.ti:Jf7+ �h7. 32 ... ti:Jf6! 33.l:th3+ ti:Jh7 34.�xc3 bxc3
138
Chapter 1 3 - Alekhine's Nail
it is interesting to ask who is playing Lev Pol ugaevsky

for a win. Therefore, instead of the M i khail Tal
capture on dS, Keres was prepared to Moscow 1 969
play 28 .h4, making luft for his king
and at the same time preparing the
0
manoeuvre .!lel -e3 -g3 . It is still not at
all clear: 28 ...�d4 29J lle3 (or 29. l::l'.d l
�f6 30.it.xd5+ it.xdS 3 1 .�xd5+ �h8
32 . .l:ixd7 �xd7 33 .�xd7 �f8 34.f3
�xh4 with a draw) 29 ...�al+! 3 0.�h2
(30 . .l:le l 'ti'd4) 30 .. ."iYf6 3 1. . l::tg3 �xgS
32.lhg5 it.c6 33 . .l:igxg7+ �h8 , and the
most likely outcome is again a draw.
24. tlJxh7!
The most well-known and also the most
difficult example of this theme.
1 6.d5! exd5 1 7.e5! ttJc4 1 8. 'iff4
Now the threat is the standard blow
on h7, followed by tt::l g S+, etc. Conse
quently, it is no surprise that our list
of candidate moves should begin with
1 8 ... h6. Polugaevsky gives the variation
1 9.'iffs g6 20.'i¥h3 �g7 2 l .e6! fxe6
2 2 .tt::ld4! «and Black's position collapses». This
24 ... tlJxh7? is true, but stronger is 2 l ...�f6! not
This is perhaps the only move for which allowing the knight into d4. Therefore,
one can criticise Fine. By the method in the game Khuzman-Parligras (Turin
of elimination, he should have found 2006) White immediately played
24 . ..tbg6. 1 9.e6 , and after 1 9...�f6 20.'ifxf6 gxf6
In the older books, the variation is 2 l . it.f5 �c7 22.tt::ld4 obtained a barely
extended a move and a half: 25 .�h5 noticeable edge (despite the extra
tt::lf4 26."iYfs . But it is not so simple: pawn) in the endgame. Another point
after 26 ... tt::le 6 (26 ... tt::lg 6 27.e6!) 27.tt::lf6+ in favour of 1 9.e6 is the fact that after
�f8 ! (of course, not 27 ... gxf6 28.exf6 1 9 ... fxe6 20.�g4 Black cannot now
�f8 29."i!Yh7!) no decisive continuation play 20 ...'ii'f6 because of 2 1 . .:.xe6 �f7
of the attack is visible. 22 . .l:ixh6.
Now, however, it is all over: The knight could also be deprived of
2S . .:r.h3 11fc1 26. 'ifxh7+ �8 27 . .1:[e3 gS by means of 1 8 .. .f6!? But then, Black
d4 28. 11fh8+ �e7 29. 'ifxg7! must also reckon with 1 9.e6.
The rook cannot be taken because of The next defence is 18 ... �c6. Let us
mate in one, whilst after 29 ... it.d5 look at Polugaevsky's analysis: 1 9.tt::lg5
3 0.�f6+, after any king retreat, 3 l .e6! h6 20.it.h7+ �h8 2 1 .tt::lxf7+, «and now
decides. There followed: a quick defeat follows both 2 I ....l:Uf7 22.�xf7
29 ... .:r.fa 30. 'iff6+ �eo 31 .e6! 1 -0 �xh7 23 .e6 tt::ld 6 24.�d7, and 2 I ...�xh7
1 39
22.tbxd8 l:hf4 23.e6». The passage of

time has brought some corrections to
this note. In the variation 2 1 .. Jhf7
22 .iYxf7 <it>xh7 2 3 .e6 t2Jd6 24.'i¥d7,
Black defends by means of 24 ... .l::rc 8.
However, the simple 23 .'i¥xb7 is per
fectly sufficient to win. On the other
hand, after 2 l ...<it>xh7 22 .t2Jxd8 .l::rx f4
23 .e6 there is in fact no win, if one
extends the variation by just one more
move: 23 ... .l:tf8 . The attempt to repair
the variation by replacing the move 21 .h4 ! !
23 .e6 with some other move also fails: Let u s try to understand what Black
2 3 . �xd5 �c7; 23 .t2Jxc6 �xc6; 23 .t2Jxb7 is threatened with. Assume he plays
.l::f.e 4 24.l:lxe4 dxe4 25 .t2Jd8 l:Ic8 26.e6 2 1 ... tbxd 1 . It turns out that there is a mate:
t2Jd6!. 22 .h5+! �xh5 (22 ...<it>h6 23 .t2Je6+)
So, does that mean 1 8 ... .l::rc 6 is the best 23 .g4+ �g6 24.iVf5+ �h6 25 .'i¥h7+
move? Not quite. For some reason, �xg5 26.iVh5+ �f4 27.iVf5#. It is sur
Polugaevsky does not analyse the direct prising that, although it is a long time
1 9. �xh7+ �xh7 20.t2Jg5+. If the idea since Polugaevsky revealed his analysis,
of bringing the rook to c6 was at some players are still trying to add correc
moment to be able to close the 6th rank tions and improvements.
- 20 ... �g8 2 I .'iVh4 .l::th6 - then this After 2 l ...f5 , Polugaevsky suggested
does not work, as instead of 2 I .iVh4, 22.�d4! and was perfectly correct. A
White plays 2 1 .e6! I:f.xe6 (2 1 . ..fxe6 sample variation is 22 ...�e7 23 .h5+
22 .�h4) 22 . .l::f.xe6! (the knight is more (23.Ile3!? probably wins as well)
needed than the rook in this situation) 23 ... r.t>h6 (23 ... �xh5 24.�h4+ �g6
22 .. .fxe6 23 .�h4 .l:r.f6 24.iVh7+ 2 5 .'i¥h7+ r.t>xg5 26.f4+ �g4 27."iVh3#)
�f8 25 .'ifh8+ �e7 26.'i¥xg7+ �e8 24.tbf7+ �h7 25 .'iVxf5+ �g8 26.e6
(26 ... �d6 27.tbe4+!) 27.'i¥xb7 - and 'i¥f6 27.'ifxf6 gxf6 28.t2Jd6 tbc4
White should win. 29.t2Jxc8 .l::txc8 30 . .l::tg4+ <it>h7 3 1 ..l::tg 6,
1 8 ... ltJb2 winning.
Black is playing with fire. Not only does Another futile attempt to defend is
he not prevent the sacrifice on h7, but 2 1 ...'iYd7 22 .e6! fxe6 23 .'i¥g4 .l::tf6
on the contrary, he provokes his oppo 24.tbxe6+ �h6 25 . .l::te 5! g6 (25 ...'i¥f7
nent into making the combination. 26.�h5+! 'ifxh5 27.'ifxg7#, Dimov
Evidently, Tal was convinced he could Tsolov, Bulgaria cr 1 990) 26 .'iVg5+ and
beat off the attack somehow. He did not 2 7.'ifxf6 - analysis by Polugaevsky.
realise that Polugaevsky had prepared In the game Dreev-Jussupow (Mainz
the whole game at home. 2003) , Black played 2 l ...'iVe7. It is
1 9 . .b:h7+! c.ti>xh7 20. ttJQ5+ <if;!g6 clear in what way this is stronger
«The first impression is that nothing is left of the than 2 l ...'iVd7. Thus, after 22 .e6 fxe6
attack, but White has at his disposal a prepared 23 .'i:Vg4 Ilf6 24.t2Jxe6+ �h6 2 5 . l:le5 g6
move of fearful strength» - Polugaevsky. 26.'i:Vg5+ �h7 White cannot take the
1 40
rook on f6, as it is now defended. But he decides with 26.h6! .l::i.h4 (or else mate
does not need to; 26.h5! l2Jxd1 27.hxg6 in one) 27.l:!.d4!
is perfectly sufficient for victory. 26. "ifxf6 gxf6 27 . .1:.d2
Be that as it may, Dreev played more Later, Polugaevsky wrote that this
cautiously: 22 . .l::i.d 2. There followed move «slows down the tempo of the attack»,
22 ... !tc4 (on 22 ... l2Jc4, strong is 23 . .i::!.d 3 and that «27.l2Jd6 was more energetic». Lev
and then 24 . .l::tg3) , and here he could Abramovich, as always, was merciless
have won with 23 .h5+! WxhS 24.g4+ towards himself, but the variations do
with a familiar mating construction: not in fact substantiate the strength of
24 ...Wh4 2S .t2Jf3+ Wh3 26 .'ftg3# or 27.t2Jd6: 27 ... t2Jxd1 28.e7 (28.!txd1 �c7)
24 ... \t>g6 2S .'fif5+ Wh6 26.'fih7+ 28 ... .l::tc 1 ! (but not 28 ... .!::i.d4 29.exf8'fi+
\txgS 27.'fih5+ Wf4 28 .'ftf5#. And c;t>xf8 30.lte8+ Wg7 3 l .t2Jf5+ Wf7
if 23 ... \t>h6, then after 24.l2Jxf7+ �h7 32 .h6! Wg6 33 .t2Jxd4) , and then:
25 .'ftf5+ Wg8 26.e6! we reach a posi A) 29.exf8'fi+ �xf8 3 0.t2Jxb7 d4;
tion from Polugaevsky-Tal, but in a B) 29.t2Jxb7 d4;
more favourable version for White. The C) 29.h6!? �b8 30.t2Jxb7 kle8 3 l .lLJd6
rook is on d2, not under attack, and one .l::!.xe7 32.!txe7 l2Je3+ 33 .\t>h2 l2Jg4+
could hardly wish for any more. 34.Wg3 t2Jxh6 3S.!txa7 with a slightly
21 ... .1:.c4 22.h5+ ..th6 better position (Naumkin-Nevanlinna,
The same mate arises for the final time Jyvaskyla 1 993) , although one has the
after 22 ... \t>xhS 23 .g4+ Wg6 24.'fif5+ impression that Black's defence can be
Wh6 25 .'fth7+ \t>xgS 26.'fth5+ c;t>f4 strengthened.
27.'fifs# and 23 ... \t>h6 24.'fih2+ \txgS Polugaevsky's intuition did not let him
2s.'fihs+ �f4 26.'f!lfs#. down. At the board, he played the
23. tbxf7+ ..th7 24. 'it'f5+ ..tg8 25.e6! strongest move.
«Looking at this position, I could hardly believe my 27 ... .1:.c6 28. :Xb2 :ea
eyes - it had been on the board in Polugaevsky's More stubborn is 28 ... �c8 , although in
hotel room that very morning» Geller.
- this case, after 29.e7 .!::i.e 8 30.t2Jd8 l:!.c7
There are few examples of such deep 3 1 .ltbe2 (or the immediate 3 1 . ltd2)
opening preparation, in the whole of 3 l ...d4 32.l:td2 l:Xcxe7 33 . .l:lxe7 .l::rxe7
chess history. 34 . .a'.xd4 White has a healthy extra
pawn and every chance of realising it.
29. tbh6+ ..th7 30. tbf5 .l:.exe6
31 .:Xe6 :Xe6
25 ... 'iff6
Running to the endgame is the only
chance. In reply to 25 ...'fte7 White
141
This ending, which still contains some 22. 'iff4 f5

play, White turns to his advantage with
great ease.
32. l:c2! .l:tc6 33.l:e2! �c8
Or 33 ... Iic7 34. Iie6.
34. l:e7+ �h8 35. ttJh4 f5 36. tbg6+
�g8 37. :Xa7
Black resigned. An epic win! " <<,>,
:; '
Alexander Khal ifman
,'/51 8 ,•
>,
Przemyslaw Gdanski w::,· ;t;

Leningrad 1 989 23.h6! �c6
It is likely that, when he put his bishop
on d7, Black miscalculated the variation
23 ... g6 24 . .if6+ �g8 25 . .l::!.xg6+! hxg6
26.h7+ �xh7 27."iYh2+ .ih6 28 . .id3.
The threat is 29 . .l:Ihl , and then, as nec
essary, a few entries from the textbook
of two-move mating combinations:
28 ...lt:Jxd3 29."iYxh6+! �xh6 30 . .l:thl#.
24.hxg7+ hg7 25.:Xg7! �xg7
26. �f6+ �g8 27.g4! fxg4 28. 'ifg5+
�8 29. 'ifh6+ �g8 30. �e2 1 -0
1 6.h4!
This is not mate, but only preparation As for the main idea - attacking the
for a mating attack. Passive defence king, using the 'nail' on e5, keeping the
does not help - Black must prepare the enemy knight from f6 - this idea is as
break .. .f7-f6, or seek counterplay on old as the game itself.
the queenside. But Black just runs on
the spot: Alexander Alekhi n e
1 6 ... 'ifb6 1 7.h5 :tea l m re Konig
Tempting is 1 7... lt:Ja4 1 8 .lt:Jb3 .ibs. Vienna 1 922
1 8. �h4!?
Black allows 1 8 .h6, but White avoids it.
D
Does anyone know why?
1 8 ... �f8 1 9 . .1:th3 �h8 20 . .1:tg3 'ifc7
21 . .J:I.e1 �d7
Probably the result of an oversight. But
even the 'more tenacious' 2 l ...h6 22."iVf4
f5 is hardly likely to hold. White plays
23.lig6, advances g2-g4 and gives mate
- not in a couple of moves, perhaps, but
within twenty or so.
Black is totally without counterplay. 1 2.h4!?
1 42
Chapter I 3 - Alekhine's Nail
«Creating two threats at once - 13 . .l::th3 followed

by l::tg 3, and 13 .tbgs followed by �d3 -h 7-g 8 !»
- writes Kotov.
1 2 ... c5
Black liquidates one of the threats;
now after 1 3 .t2Jgs cxd4 14. �h7+ �h8
1 S . �g8 there follows 1 5 ... d3 ! Therefore
Alekhine 'settles for' the second:
1 3. l:[h3
How should the position be assessed?
«Black is not saved by 13 .. .£5 14.exf6 �xf6
lS.tbgS !» - writes Kotov. He is correct: 1 4 . .b:h6!
l S ... hxgS (15 ... cxd4 1 6 . �h7+ �h8 Since Black cannot play 14 ... gxh6
1 7. �g8! d3 1 8 Jhd3 'i¥e7 1 9 . .l::f.d 7!) 1 S .'i¥d2 , the interest of the game is
16.hxg5 �xgS 1 7. �h7+ �f7 1 8.'i¥g6+ over. And Alekhine had no peers when
or 1 7... �h8 1 8 . �e4+ and 1 9. �xb7. But it came to finishing opponents off.
this is only the start of the discussion. 1 4 ...f5 1 5.exf6 .b:f6 1 6 . ..tg5 cxd4
Instead of taking with the bishop, he 1 7. ltJe5! tbc6 1 8. 'it'e2! g6 1 9 . .b:g6
should play 14 .. Jhf6!. This is stronger �g7 20 . ..th6+ <Ji>g8 21 . tillc: c6 ..txc6
for two reasons. Firstly, the e6-pawn is 22. 'it'xe6+ <Ji>h8 23 . ..txta 'it'xf8
defended, and secondly a path is freed 24. 'iVxc6 1 -0
along the eighth rank. The standard
mating scheme now leads nowhere: Lajos Portisch
1 5 . �h7+ Wh8 1 6.t2JgS �xg2 , and on Svein Johannessen
1 7.�g8 there is simply 1 7...'i¥xg8. And Havana 1 966
if 1 6 . �g5 , then 16 .. Jhf3 ! 1 7.gxf3 cxd4
with unclear consequences.
0
In reply to 14 ... .!::!.x f6, White must try
l S .tbgS , creating the threat of 1 6 . �a6.
Black can move his bishop out of the
attack: l S ... �dS , or he can pretend that
he has fallen into the trap: 1 S ... t2Jd7!?
16.�a6 hxgS 1 7. �xb7 g4! . The conse
quences of the variation 1 8 . .l::tg3 .l::tb 8
1 9. �e4 cxd4 are again completely
unclear.
1 3 ... <Ji>h8 1 6.h4!? cs 1 7. l:[h3 ..txf3
Simply an oversight. This is what so Now any recapture on f3 would slow
often happened in Alekhine's games down the tempo of the attack. For
against significantly weaker opponents. example, 1 8 .gxf3 cxd4 1 9. �xh6 tbxeS
Like a knockout-punching boxer, the 20. l::tg3 �f6. But White (one assumes
fourth World Champion would flatten when considering his 1 6th move) had
such opponents with one or two direct in view another, more energetic contin
blows. uation:
1 43
1 8 . .b:h6! A sad oversight. He wins with 2 1 . .l:rgl+

In Hajtun's book on Portisch's best .tgs (not 2 1 ...\t>hs 22 .'i:lfc1 'iff6
games, this moment receives the 23 .dxeS), and only now 22 .dxeS. Maybe
briefest of comments: 'Alekhine-style! '. White thought the text was a simple
The author possibly did not realise just transposition, but this is not the case.
how right he was, at least in the sense 21 ... 'ifd4!
that the method of attack - h2-h4, It turns out that after 22 . .l:i.gl+ �h8 ,
z:rhl -h3 and .tcl (d2)xh6 - is definitely White has no time for 23 .'iVcl because
very similar to that employed in the of 23 ...'ifxf2+. He has to think again:
game Alekhine-Konig. But the position 22.:Xh4! 'ifxh4 23J:lg1 + ..th8
here is still more unclear. 24. 'ifc1
1 8 ... .hh4
The move 18 ... .tg4 also deserves the .i I ••
most serious attention. It looks as i
though Black loses in beautiful fash i .t.:
.· .
ion: 1 9. l:rg3 fS 20.I:rxg4 fxg4 2 1 . .th7+ A

�h8 22 . .txg7+! \t>xg7 23 .'iYg6+ \t>h8
24 . .tg8 ! But one only needs to take a
slightly closer look to find an improve
ment: 20 .. JU7! White has a rook and
bishop hanging, whilst Black has
defended all his vulnerable points. After The threat of .l:i.gl -hl looks deadly. But
2 1. . l:Ig6 ti:Jf8 or 2 l . .tc4 fxg4 22 . .txe6 it is Black's move, and for the moment,
�a6, the attack threatens to come to he does not need to stop the clocks, and
nothing. the battle continues.
1 9.gxf3 gxh6 20 . ..te2 24 ... f6
But not like this. Black had proba
bly calculated the variation 2sJ:rh1
'iYgs 26 . .l::f.x h6+ \t>g7 27. .l:i.g6+ 'itxg6
28 . .txg6 c;tJxg6, where the two rooks
are more than sufficient compensa
tion for the queen, and 27.exf6+ �xh6
28.'ifhl+ 'ifhs 29.'ifcl+ 'iigs , where
White has nothing better than perpetual
check. But the refutation is quite simple.
25. llg6!
Incidentally, in the variation 2S . .l::!.h l
By comparison with the variations 'ifgS 26.l:txh6+ \t>g7 Black also loses:
shown in the previous note, Black has 27 . .tlh7+! c;itg8 28.'ifhl 'iYxeS+ 29.c;tJfl
a piece more, but in return, the white fs 30.'ifhs 'iff6 3 l. . �h6 'ifes (3 1 ...'iYg7
rooks have an extra open file. The 32.�g6) 32J:rg6+ �f7 33 .'i¥h7+ \t>e8
attack has become simpler, e.g. 20 ... 1::f.e 8 34 . .tbS+ c;tJds 3S .'ifd7#. Since the
2 l..�gl+ �f8 22 .'ifd2 .tgs 23.f4 etc. game was played in an Olympiad,
20 ... �e5 21 .dxe5 Portisch preferred to win solidly, with
1 44
something in hand as it were, without ifcl ! both fail) 30 ...ifh2 3 I..I:i.g2! Vi'h4
alarming his teammates. After 2S . .I::tg6 32.l:tg4 with a perpetual attack on the
Black resigned immediately. queen.
The most subtle is to retreat backwards:
Let us return to the position in the last 2 S . �b1 ! .
diagram, which is highly interesting!
It can be used separately, as an excel
lent exercise in the calculation of var
iations. But in the position of Svein
Johannessen, I would have played the
move 24 ... c4! without thinking at all,
just, as it were, 'with the hand'. The
move is obviously useful. It attacks the
bishop, forcing the opponent to forget
about the threat of .l::tg 1-h1 , at least for a
moment. And what will happen after
analysis diagram
wards - we will see ...
After 24... c4, the most likely outcome To the threat of .l::!.g 1-h1 White adds
is a draw. However, White is risking as another: the queen and bishop battery
much as his opponent. Let us consider with 'ifc2 . The line which saved Black
the capture 2S. �xc4. Then 2S .. J :tfd8 in the previous variation, 2S ... c3 26J:th1
26 . .l:f.h1 .l::td 2+!. Now it turns out that the cxb2, here loses at once: 27.Vi'c2! .
Greek Gift cannot be declined - 27.�e3 But Black i s rescued by the rook sacri
ifd4# - whilst accepting it also loses: fice, with which we are already famil
27.�xd2 ifxf2+ 28.�e2 l::td 8+ (the iar: 2S ... .l:f.fd8 26 . .!::lh 1 �d2+! The next
same move would have followed after few moves can be made on 'autopilot':
28 .�d3) 29.�c2 .l::f.c 8+ and 3 0 ... 1:lxc l . 27.ifxd2 (the king cannot take on d2
The second candidate move i s to bring because of mate) 27 ...'ifxh1 28.'i!Vc2!
the bishop away with tempo: 2 S . �e4. �g7 29.Vi'h7+ �f8 30.Vi'h8+ �e7
In reply, Black cannot play 2S ... l:!.fd8 3 I .Vi'xf6+, and now both players must
26.llh1 l:td2+ because of the simple assess the risks. A draw can be agreed
27.ifxd2 'iYxh 1 28.�xa8 . But there at any moment by means of perpetual
is another idea: 2S ... c3 ! . Again, White check. Playing on is only dangerous
cannot attack the queen: 26 . .!::i.h 1 cxb2 . for White: 3 l ...�e8 (weak is 3 1 ...\t>d7
He has to spend a tempo on 26.bxc3 , 32 .'i!Vxf7+, whilst after 3 I ...�f8 White
and then 26 .. J:tac8 . White has a draw, draws with both 32.Vi'h8+, and 32.�g6)
but probably nothing more: 27. .l::tg4 32.�c2?! (overstepping the bound
(27J lh1 .l::txc3) 27 ...ifh2 (not 2 7...ifhs ary of safety) 32 ... b3 ! 33.axb3 l:r.d8
28 .Vi'f4! fS 29.exf6, and the mating (33 ...Vi'c1 34.�g6+!) 34.bxc4 (34. �g6+
attack gets a second wind) 28.l::rg2 ! I:!.xc3 Vi'dl + 3S .�e3 Vi'd2+ 36.�e4 .l::td4#)
29.'iYd2 ifh1 30 . .l::f.g 1 ! (in this situation, 34 ...Vi'a1 , and it is already Black that is
30.Vi'f4 fS 3 l .exf6 �cl ! and 3 I. . l::!g6 refusing the draw.
1 45
Chapter 14
Attack and Counterattack with a Closed

Centre
In such positions, is the early (even in the opening) advance g2-g4 not just feasible,
but even the best approach? The classical answer to this question is found in the
following game.
immediately or not) , or to h3 and then

Yu ry Averbakh f2, cementing his pawn chain with
Oscar Panno e4-f3 -g4.
Buenos Aires 1 954 Finally, the third point. The move g2-g4,
although safe for the pawn structure we
have here, does give the game a some
0
what less standard course. White must
be prepared for this. As they say, he
who sows the wind can reap the whirl
wind. One cannot push the g-pawn as
early as move 10, and then start play
ing in a standard fashion, as though one
has taken on no obligations. No! White
must continue to play without routine.
1 0 ... tbe8 1 1 .h4
1 0.g4! White played less strongly in the game
The first thing to which I wish to draw Szabo-Barcza (Budapest 1 958): 1 l .gS
attention is that the centre is closed. A fS 1 2 .gxf6?! tbxf6. The f-file is opened
central counter is the strongest response and the knight from e8 has returned to
to a flank blow, but with a blocked play. It is not obvious what White has
centre, Black is deprived of such a coun achieved.
ter. Effectively, White has a free hand. 1 1 ...f5 1 2.h5 f4 1 3.g5! .l:.f7
He can behave like a bit of a hooligan,
throwing his g-pawn forward at such
an early stage, and know that he will
come to no harm.
Secondly, White has not yet castled,
nor even moved his knight from gl .
This lends his plans a certain amount
of obscurity. He might castle kingside,
queenside, or leave his king in the
centre, depending on circumstances.
The knight, waiting until the pawn
reaches gS , may come to f3 (either 1 4 . .tg4!
1 46
Chapter 1 4 - Attack and Counterattack with a Closed Centre
Averbakh demonstrates outstand Evgeny Bareev

ing technique. The f-file is closed, the Gennadi Zaich i k
knight on e8 shut out of the game, Kharkov 1 985
the light-squared bishop cannot avoid
exchange, whilst Black's dark-squared
0
bishop is reduced to the role of a big
pawn.
1 4 ... 'ifdB 1 5 . .h:c8 'ifxc8 1 6.tbf3
i..f8 1 7. �e2!
A device which was revolutionary at the
time, but has since become common
place. White commits his king to the
centre, on a full board, not to make fun
of his opponent, but because it is the
best square. On the kingside, the king In this position the knight is not on g1 ,
would get under the feet of his own but on g2 (evidently after travelling via
pieces, and on the queenside, he could g1 -f3 -h4-g2), but the rest of the posi
come under attack after ... b7-b5. But on tion is practically as in the previous
e2, nothing threatens him. example. The main things are the
Later, the game was decided by a direct closed centre and the uncastled white
attack along the open h-file. This out king, which means that he can choose
come was logical, but not forced. his own fate.
White, having suppressed all counter 1 4.g4!? a6 1 5.a4 a5
play, is better on both flanks. He could After 1 S .. J:tb8 , Black may not have liked
attack the enemy king or play to open 1 6.a5 . But the text move, closing the
the b-file, by a4-a5 , then tt:Jc3 -a4 (or queenside as well as the centre, finally
.l:i.a1 -b1) and b2-b4. Black is condemned condemns Black to defence, since on
to complete passivity. the kingside he is clearly inferior.
1 6. i..g5 tba6 1 7. 'ifd2 tLlb4 1 8.0-0-0
And one more thing. Two years later, Note that if White had played this
there was a game Averbakh-Spassky before advancing g3 -g4, the game is
(Leningrad 1 956). The future World unlikely to have engaged our attention.
Champion, either through ignorance or 1 8 ... 'ifd7 1 9.f3 tbe8 20 . .l:[dg1 b6
for some other reason, repeated Panna's 21 .h4 f5?!
first twelve (and hardly optimal) An extremely untimely shift from
moves. Then he varied with 1 3 ...�d8 , passive defence to active play. One can
but after 14. �g4! tt:Jc7 1 S . �xc8 'iVxc8 only assume that Black had missed
1 6 .t2Jf3 he realised that he was under White's 25th move and the subsequent
the press. manoeuvre.
The rest is well-known. After some 22.gxf5 gxf5 23.h5 �h8 24. ttJh4 f4
thought, Spassky produced one of the 25. i..f 1 !
most shocking moves in chess his This move could easily be missed from
tory: 16 ... tt:Jc6!?!. But that is another afar. But one could also remember it,
story... since, after all, White played the analo-
1 47
gous manoeuvre �e2-g4 in Averbakh's already castled kingside and after the
games against Panna and Spassky. inevitable .. .f7-fS , he seems to be in as
25 ... l:tf7 26. �h3! 'ifxh3 27. :Xh3 dangerous a position as his black col
bh3 28. 1i'h2 league.
The forced queen sacrifice did not bring So why does White draw the fire upon
Black any relief and White won on himself?
move 36. 1 1 ...f5
The analogy with the variation l .d4 tbf6
In modern chess, we are continually 2.c4 g6 3 .tbc3 �g7 4.e4 d6 s .tbf3 0-0
getting rid of fears, phobias and taboos, 6.�e2 eS 7.0-0 tbc6 8 .dS tbe7 9.tbel
characteristic of classical chess. That tbe8 I O.f3 fS l l .g4!? (and then tbel -g2
which the textbooks of old said was and h2-h4, with a complete blockade of
bad, dangerous or unnecessary, is often the kingside) is hardly appropriate here,
now regarded as perfectly possible, or since in this position, after 1 2 .f3 , there
even good. is the reply 12 ...'iVh4.
This has been partly the influence of 1 2.exf5 gxf5 1 3.gxf5
the computer. Machine analysis (i.e. Now the idea starts to be revealed. By
'absolute accuracy') allows a player to the double exchange on fS , Kramnik
feel solid ground under his feet. Pre initiates a battle for the light squares,
viously, he was simply afraid to play especially e4 and e6.
this or that risky move or unusual idea. 1 3 ... ttJb6
The computer analysis can be likened A natural line is the gambit 1 3 ... tbdf6
to a powerful searchlight beam, which 14. �d3 e4!? However, as shown by
illuminates the darkness and drives out later practice, this does not bring
irrational fears. enough counterplay. For example, in
But this started even before the com the game Avrukh-Plakalovic (Plovdiv
puter era. 201 0) , White won after 1 S .tbxe4 tbxe4
1 6 . �xe4 'iVM 1 7.'iVc2 �xb2 1 8 .tbf3
Vladi m i r Kramnik 'iVhS 1 9.'ifxb2 'i¥g4+ 20.<;£(hi 'iYxe4
Rainer Knaak 2 I . llgl+ �f7 22.l::t.g3 f:f.g8 23 . .I;lxg8
Dortmund 1 99 2 �xg8 24 . .I;lgi+ �f7 2S.f:f.g3 .
1 4. ttJf3 bf5 1 5. tlJQ5 1i'e7 1 6. 'ith1
D
ttJf6 1 7. l:tg1 'ith8 1 8. 1i'd2
1 1 .g4
such was the first
«In this position? Why?!» -
reaction. After all, the white king has
1 48
1 8 ... �g6 7.g4

In the King's Indian, one cannot play This is not a sacrifice (on 7 ... tt:Jxg4 there
such 'normal' but at the same time follows 8.l:igl), but questions remain
unforced moves. Good or bad, it was nonetheless. What is White planning to
essential to try 1 8 ... e4! ?. storm? After all, the black king is not yet
1 9. l:.af1 on g8. On the other hand, if the move
1 9.tlJe6 is premature because of g2-g4 in this particular case is directed
1 9 ... ttJe4, and not 20.ttJxe4 �xe4+ towards seizing space, then Black's
21 .f3 l:lxf3 ! . remaining space is likely to suffice for
1 9 ... tt:Jh5 developing counterplay.
After the game, Knaak suggested 7 ... tt:Jf8 8.g5 ttJsd7 9.h4 tt::lg 6
19 ... �h6 20.b3 �xg5 2 1 . �xg5 ttJbd7, 1 0. �e3 h6!?
solving the problem of the knight Black actively fights back.
coming into e6, but stronger is 20.ttJxh7! 1 1 . l:[g1 hxg5 1 2.hxg5 a6 1 3. 1i'd2
and after the zwischenzug 20 ... �xe3 tt:Jdf8 1 4. l:[b1 ! ?
White replies in kind with 2 1 .tlJxf8 ! . Another unorthodox decision.
20. tt:Je6 l:[f7 21 .b3 tt:Jd7 Tempting was 14.0-0-0, but White,
Kramnik 's recommendation was having got nowhere on the kingside,
2 l ...tlJf4 22. �xf4 exf4 23.�d3 �h5 decides to break on the queenside.
24.f3 'ifh4 25 .tlJe4 �e5 , «and Black holds». 1 4 ... �d7 1 5.b4! b6 1 6.a3 1i'c8
Alas, after 25 .ttJxg7 llxg7 26.l:I.xg7 1 7. �e2 l:[h3 1 8. �d1
�xg7 27.tlJe4 this is not the case:
27 .. .'�e7 28 .�xf4, or 27 ... l:I.d8 28.a:gl+
�f8 29.�g2! �g6 30.�b2 etc.
22. �d3
With the exchange of light-squared
bishops, the position resembles more
closely the classical examples we saw
above. White's advantage does not raise
any doubts.
Vlad i m i r Kra m n i k
Al berto Mascarenhas 1 8 ... tt:Jf4!?
Rio de Janeiro 1 991 One must give Kramnik's opponent his
due. He seeks counterplay, not afraid to
sacrifice a pawn in the process.
1 9. bf4 exf4 20. 1i'xf4 tt:Jg6
The game is unclear.
In general, it must be said that the move
7.g4 in the diagram position is clearly
very experimental. It is no surprise that
it has not been repeated, with others
preferring the more cautious 7. �e2 ,
7. �d3 or 7.h3 . In all cases, White tends
1 49
to wait for Black to castle kingside and Are you interested to see the first game
only then to push the pawn to g4. where White put his king on e2 in a
similar pawn structure? It may well
Amongst contemporary grandmasters have been the following game.
who actively use this device, we will
single out Veselin Topalov. He pushes Jose Rau l Capablanca
his g-pawn especially often against Vera Mench i k
players who are significantly weaker Moscow 1 935
than himself.
0
Veselin Topa lov
Francisco J i menez Mora les
Seville 1 992
1 0.g4!? h6
A trap (the pawn on h6 cannot be taken
because of the check on h4) , but almost
any other move would have been better.
9.g4! g6 1 O .i.h6 ttJg7 1 1 . 'iVd2 a6
. 1 1 .h4 c;fo>h7 1 2. ttJg3 cS?!
1 2.a4 1Va5 Completely wrong. Black closes the
It is bad enough that Black decides to centre, without any reason, thereby
play a well-known theoretical position, depriving herself of counterplay.
a tempo down (instead of l .d4 tLlf6 2.c4 1 3 . .i.d3 'iVas
cS 3 .d5 eS 4.tLlc3 d6 S .e4 .i.e7 6.tLlf3 Once again this useless manoeuvre,
0-0 7. .i.d3 tLle8 he played 2 ... e6 3 .tLlf3 which only encourages White to gain
cS 4.d5 d6 S .tLlc3 �e7 6.e4 eS 7. .i.d3 further space with a later b2-b4.
0-0 8.h3 tLle8 , so the e-pawn reached 1 4. 1Ve2 l:[h8 1 5.a3 'iVd8 1 6.b4 b6
eS in two moves, rather than one) , but 1 7. 1Vb2 �c8 1 8.c;fo>e2!
now he sends his queen on a totally with an overwhelming advantage.
pointless errand. More in the spirit of
the chosen variation was 1 2 ... tLld7 1 3 .a5 Black learnt in King's Indian positions
lLlf6, and then ... �g8-h8 and ... tLlf6-g8 , to fight back against g2-g4 with the
driving the bishop from h6 and prepar move ... h7-h5. This seems so obvious
ing the break .. .f7-f5. that it hardly needs discussing, were it
1 3. 'ite2! ? ttJd7 1 4. 1Vc2 f6 1 5 . .i.d2 not for a 'but'.
'iVc7 1 6.a5 If White plays g2-g4 before Black has
Again, White stands better on both castled, then the reply ... h7-h5 is obvi
flanks, and his king on e2 is no hin ously tempting. But what if White
drance at all in this position. delays his g-pawn advance, until after
I SO
Black has castled kingside? Then ... h7- on g7 fully compensates for the material deficit)
h5 looks terrible. The textbooks all rec White obtained a strong attack. But it seems to me
ommend that one not advance pawns that this continuation is not dangerous for Black -
in front of one's castled king, provok he can continue 14 ... tt:lf6, retaining an excellent
ing the opponent to attack one's king, position.
which he was thinking of anyway. In my game against Gligoric at the 19 59 Candidates'
But there is an exception to every rule. tournament I played 12 .h3 and after 12 ... tt:lh7
In the King's Indian, Samisch Variation, (this is Black's idea; 13 .gxh5 gives nothing because
there is a set-up where for over half a of 13...�h4+, whilst after 13 .0-0-0 he con
century Black has confidently met g2-g4 tinues 13 ... h4 followed by ... i.g7-f6-g5. This is
with ... h7-h5, despite his king having what Gligoric played against Sherwin at Portoroz
castled. 1959, and he got the better position from the open
ing) 13 .h4. Gligoric did not find the best response
• to this novelty, and after 13 ... hxg4 14.fxg4
tt:lhf6? 1 5 . .fih3 tt:lb6 16.i.gS White obtained
the advantage. Instead of 14... tt:lhf6, Black could
have continued either 14... tt:lb6 with the threats
of ... i.c8xg4 and ... tt:lb6-c4, which would force
White to go in for very unclear play after 1 5.tt:lg3
i.xg4 16.h5 with an attack for the pawn, or
(which seems even stronger to me) 14... tt:ldf6!
1 5.g5 tt:lg4 or 15 ... tt:lh5, strengthening his posi
tion on the kingside.
1 1 h5!?
... Finally, in several games from the XXVI USSR
Who was the author of this idea? It is Championship, and also the game Szabo-Gligoric
usually considered that there were (Belgrade 1959), White played 12.gS tt:lh7
two authors. The top Yugoslav player, 13 .h4. Gurgenidze, as Black against Spassky, and
Svetozar Gligoric, introduced the move also Petrosian against Polugaevsky, started play
... h7-h5 in grandmaster practice, but on the queenside, but here the inactive position
Mikhail Tal used it in his 1 960 World of the knight on h7 had its say. Gligoric played
Championship match; furthermore, he more precisely against Szabo: 13 .. .£6!? 14.gxf6
also wrote about the ideas of the move l:txf6 15. tt:lg 1 , but then he got distracted by com
in great detail, which made it part of plicated manoeuvres, as a result of which White
the opening theory of the time. We will succeeded in placing his pieces comfortably and
cite Tal's comments in full: gradually seized the initiative. However, considera
«The move 1 l ...h5 is Gligoric's patent. Because the ble attention should be given to the exchange sacri
g4-pawn is attacked, White must do something. fice 15 .. .1:1£4!? White must accept the gift (if, for
The following continuations have been seen: 1 2 .g5, example, he continues 16.0-0-0 �h4 1 7. i.g5,
12 .h3 and 1 2 .gxh5. The last move is relatively new hoping to take the exchange in more favoura
and was played in the game Weber-Riitsch (Leipzig ble circumstances, then there follows 1 7... i.h6!
1959), in which after 1 2 . ..tt:lxh5 13 .tt:lg3 tt:lf4!? Black even wins), but then after 16 . .fixf4 exf4,
14.h4 (here, and in similar positions, it is unfa Black's positional plusses are worth the exchange.
vourable to accept the pawn sacrifice: 14 . .fixf4 Of course, there is no forced win, but the resulting
exf4 1 S.�xf4, since the activity of the bishop position has attracted many supporters for Black.»
151
And here are Svetozar's words, taken 1 3 . l:tg1 ! ? h4 14.tt:Jcl �h8 1 S.tt:'Jb3 �f6.
from his book I Play against Pieces: Sherwin, as we know, castled queen
«The move 1 l ...h5 was a completely unexpected side on move 13 and was unable to pre
idea at the end of the 1950s. When it came to vent Black's plan of activating his posi
kingside counterplay, Black thought exclusively in tion on the dark squares with ... h5-h4
terms of .. .f7-f5, which after g2-g4 only helps and ... �g7-f6-g5 . But Timman put his
White's attack. However, the move ... h7-h5 is the rook on g1 in anticipation, which three
only one which answers the logic of the situation. moves later gave him the resource 1 6 .g5 .
Black exploits the fact that the white kingside is After 1 6 ... �e7 1 7.0-0-0 f6 1 8 .gxf6 l:t.xf6
insufficiently developed, and strikes back at the 19.�e2 'iff8 20.�b1 it became clear
weaknesses created on the dark squares by the move that Black had not equalised.
g2-g4 (a major threat in the Siimisch Variation).» Black has to dispense with the plan with
It is not by accident that we have quoted ... h5 -h4, and even 1 2 ... tt:Jh7. In the game
these long extracts. I wanted the reader Ivanchuk-Kasparov (Dos Hermanas
to learn about the genesis of the move 1996), there followed:
... h7-h5 not second hand, but 'from the 1 2 ... b5 1 3.i.g5 'ifa5 1 4.lbd1 b4
horse's mouth', as it were. I would add Also very interesting was 14 ...'ifxd2+
that the variations may look to some 1 5 .�xd2 tt:Jcs 1 6.tt:'Jf2 �d7 1 7.gxh5
to be a little naive, compared with the tt:JxhS! 1 8 . �e7 I::rfc8 1 9. �xd6 tt:Jb7
annotations of modern-day grandmas 20.�e7 f6 2 1 .tt:'Jd3 �f7 22.�a3 aS
ters, but when it comes to ideas, they 23 .b4 tt:Jd6 with full compensation for
are at a very high level. the sacrificed pawn, Piket-Kasparov,
Amsterdam 1996.
The strongest players of those days 1 5. ltJQ3 tbh7! ? 1 6.gxh5 tbxg5
understood chess brilliantly. Before them 1 7. 'ifxg5 i.f6
lay virgin lands and they lay the basis of And White had to hurry to end the
contemporary chess. Of course, not all game:
of their ideas passed the test of time. For 1 8. 1i'h6 i.g7 1 9. 'ifg5 i.f6 20. 1i'h6
example, in the game Timman-Kasparov i.g7
(Amsterdam 1996), after Draw.
1 2.h3 A game Kramnik-Nijboer (Groningen
and now 12 ... tt:Jh7, White, by compar 1990) finished in a similar way, but
ison with the game Sherwin-Gligoric, even earlier: 14.tt:Jg3 (instead of 14.tt:'Jdl)
played more subtly: 14 ... tt:Jh7 l S .gxhS tt:JxgS 1 6.Vi'xg5 �f6
I 7.'iVh6 �g7 1 s .'ifgs �f6 1 9.Vi'h6 �g7,
with the same repetition of moves.
Thus, the plan of a regrouping on the

dark squares with ... tt:Jf6-h7, ... h5-
h4 and ... �g7-f6-g5 , as in the game
Sherwin-Gligoric, did not fully prove
itsel£ But the basic idea ... h7-h5 lives!
The fashion in the King's Indian Defence
is still based on it to this day.
1 52
Even so, who was the first player to play bining threats on both flanks, should
... h7-h5 in such a pawn structure? Tal achieve success .
pointed to Gligoric and the latter con 1 2 ... 'ifc8 1 3. l:.g1
firmed his own authorship. We have no More tenacious is 13 .f3 .
reason to criticise these great players 1 3 ... c6 1 4.c4 b5!?
for dishonesty. Grandmasters are only Extending the battle across the whole
human and cannot know everything. board, while White is still completely
undeveloped. From this point of view,
Szymon Wi nawer his next move is just suicide.
Rudolf Charousek 1 5.f4 exf4 1 6 . .bf4
Budapest 1 896
1 6 ... bxc4
This game, which started out as a Good enough to win, but a more forcing
Spanish, has turned into a King's Indian path was as follows: 1 6 ... cxd5 1 7.cxd5
structure (although paradoxically, the (1 7. �xd6 .l::te 8) 1 7...'i:Yc5 1 8 J:tg2 tt:JexdS!
term 'King's Indian' did not exist at 1 9.exd5 .l:lae8+ 20.l::te 2 (20.�fl 'i:Yc4+
the end of the 1 9th century). White is and 2 l . ..'i:Yxf4) 20 .. Jhe2+ 2 l .�xe2
noticeably behind in development. His .l:;ie8+ 22 .�f3 tt:JxdS with irresistible
last move was l l .g2-g4, which meets threats.
the needs of the position poorly. First 1 7.�g5 hxg4 1 8.hxg4 cxd5 1 9. l:lf1
he should have developed his queenside liJxg4
with c3-c4 and t2Jbl -c3. A hurricane has overtaken the position
Any opening of the position should and swept away all the white pawns!
favour Black, because he is better pre After 20. �xe7 tt:Jxh2 the fight did not
pared for hand-to-hand fighting than last long and Charousek soon received
his opponent. Hence the decision: congratulations on victory.
1 1 ... h5! 1 2. lLlh2 «Gligoric's patent», wrote Tal about the
White either did not sense the danger move ... h7-h5 in the King's Indian pawn
or just panicked. However, even after structure. «Charousek-Gligoric-Tal 's patent»,
the strongest line 1 2 .g5 tt:Je8 , then we would write, for the sake of histor
...'i:Yd8-c8 and ... c7-c6 , Black, by com- ical accuracy.
1 53
Chapter 15
g2-g4, and Wh ite wins

The battle for and against the advance .. .f7-f5 is possible not only in the King's
Indian, but in the Sicilian, and in other openings - in a word, anywhere where the
pawn structure develops in such a way that Black is ready to advance his pawn to
f5 , whilst White tries to stop this by all possible means. And in this battle, he often
cannot do without the move g2-g4!
Not only prophylaxis against .. .f7-f5,

Alexey Shirov but, strangely enough, also the start of
Vesel in Topalov an attack on the kingside.
Sarajevo 1 999 25 ... 'ifa4 26.c4 'ifc2
It is not clear why Black did not take
the pawn offered with 26 ...ifxa2 . The
0
threat of 27.�d2 and 28 .b4 is really
not serious, since in reply to b2-b4, the
queen can retreat to b3 with tempo.
27 . .1:ld2 'ife4 28. 'ifh3 as 29.b3 'ifg6
30. '1Ph1 1re4+ 31 . '1Pg1 1rg6
Having set up his queenside blockade,
Topalov repeats moves. Draw?
Shirov does not agree. He has delved
more deeply into the position that
White has a pawn majority on the his opponent. Black's seemingly solid
queenside, Black on the kingside. It position collapses almost at a stroke, it
seems that the plans of the two sides seems.
are determined. White will start an
advance with c2-c4, b2-b4 and finally,
a4-a5. Black, after delaying his oppo
nent with the blockade ... a6-a5 , will
shift the battle to the other flank: he
will play .. .f7-f5 , supporting his pawn
mass with his major pieces and trying
to push his e- and f-pawns at least as far
as the fourth rank, if not further.
The discussion of this is easily under
standable and rather stereotyped.
Shirov, well known for his unorthodox 32.g5!
thinking, demonstrates the latter in all The threat is 33.i.g4, and then 34. i.f5
its glory: trapping the queen. There is no sense
25.g4!? in 32 ...ife4 33.i.g4 g6 34.i.d l ! , whilst
1 54
Chapter 1 5 - g2 -g4, and White wins
after 32 .. .f5 the trouble comes from the attacking on the kingside with .. .f6-f5
other side: 33 . .ih5 . And if 32 ... .ie7, and ... e5 -e4, White intends to push his
with the idea of freeing an escape for g- and h-pawns.
the queen along the sixth rank, then 25 ... :aea
33 . .ih5 'i¥d6 34 . .ixf7+! �xf7 35 . .!::i.xf7 After the game, Georgy Lisitsin sug
�xf7 36."ifxh7 .l::i.f8 37. !1f2+ �e8 gested 25 ... a4 here.
38 .�xg7 Iixf2 39 . .ixf2, and the three 26.h4 liJc8
passed pawns, as the analysis shows, are Poorly played. The idea of the move is
stronger than the minor piece. clear: the knight will go to e7, defend
32 ... ttJe4 33 . ..td3 ttJxgS ing the weakness on g6, after the subse
Leads to loss of the queen, but the quent h4-h5. But the trouble is that the
endgame after 33 ... f5 34 . .ixe4 fxe4 knight on e7 completely disrupts the
35 .�e6+ "ifxe6 36.dxe6 is no less diffi harmony of the black pieces: it cuts the
cult for Black. communications, prevents regrouping
34 . ..bg5 e4 and obstructs almost any manoeuvre.
34 ...�xg5+ 35.Iig2 . From now on, the black pieces stum
35. l:g2! exd3 36. j.,f6 ble around blindly, tripping over this
Material losses are inevitable. There is knight.
still a long struggle ahead and Black A tempting move was 26 ...'ifg7, with
has real drawing chances, but after the idea of 27.h5 gxh5, and now not
overcoming his opponent's resistance, 28.Iie4 f5! . White could reply 27.�fl .
Shirov won on move 78. but then how does he transfer his rooks
to the kingside?
Vasily Byvshev 27.h5! ttJe7
Konstanti n Klaman
Riga 1 954 ;�� ,� a • ��:��
.. • �;!:�,;� �
��··�· . ·· �
�•· s LS '*{
D
:;t.• LS ;
>
lS
'• •ii'L fi > ::t··· 'iV �·:•ry
•
l%ts �ii,·� � ••.;•:•i!f\

1 . ;:;k�:."f!l· t;�{'>•;;;;;��c:
28.hxg6
Master Byvshev believes in the strength
of his attack, otherwise he would have
The classic analogue on the same theme. played 28.a3 !? opening a second front.
Not knowing the previous example, one 28 ... hxg6
would hardly guess where the danger The last chance was 28 ... tt:Jxg6!?. Yes,
to Black will come from. his pawn structure would have been
25.g4! spoiled, but the knight would have
Turning the game through 1 80 degrees. obtained the square f4, and, the most
It turns out that, rather than Black important thing, the 7th rank would
1 55
have been unblocked, allowing one of 1 0.g4!?

his major pieces to come to g7, defend «In those days, this move created a sensation,
ing his king. but nowadays it is almost a standard device» -
Now it is all easy. White achieves his Botvinnik.

aim with simple, natural moves. I would add that if White wishes to
29. <ifilg2 <ifilg7 30. l:h1 l:h8 31 . l:lae1 advance his kingside pawns to attack
1i'b8 the enemy king, then he needs to start
with the move g2-g4. In the game
�;.�:�:�
Mamedyarov-Ponomariov (Moscow
2010), White started with 1 0.h4, and
only after 1 0 .. J::re 8 did he play 1 l .g4.
:.; ,:. � :jl, After 1 I ...ti:Jf8 1 2 . i.xf6 i.xf6 1 3 .g5
i.e7, the pawn on h7 was defended.
·M�:.i�� (There followed 14.l::tg 1 i.e6 1 S .h5

.!:lc8 16.0-0-0 cS , with a complicated,
doubled-edged game.) Now, however,
as Botvinnik pointed out, after I O ... .!:le8
32.g5! fxg5 33. 1Vg3 :Xh1 34. :Xh1 1 1 . i.xf6 tt:'lxf6 1 2 .g5 White wins a pawn.
l:h8 35.:Xh8 <itxh8 36. 1Vxg5 <ifilh7 This is quite correct, but the variation
37.f4! needs to be extended. After 1 2 ... tt:'le4!
Black resigned, in view of 37... exf4 1 3 .tt:'lxe4 dxe4 14. i.xe4 'i:YaS+
38 .'i¥f6! 'i:Yg8 39.'iVh4#. 1 S .�e2 (not 1 S .'i:Yd2 i.b4) 1S ... i.xg5
16.i.xh7+ \t>h8 , Black obtains some
In the following example, the move compensation for the pawn. The posi
g2-g4 serves as a signal for an attack on tion has opened up, he has two bishops,
the king. The only difference is that in and the white king is far from ideally
this case, the attack is carried out with placed. If now 1 7.h4 i.h6 1 8 .tt:'lg5 , then
a stable centre, whereas in the previous after 1 8 ... i.g4+ 1 9.f3 i.hS the situation
ones, this was not quite the case. is getting out of control (Beilin-Polyak,
Moscow 1 949).
M i khail Botvi n n i k As well as 1 0 .. J:te8 , another move
Vladi m i r Alatortsev deserving consideration is 1 0 ... g6 (but
Leningrad 1 934 not I O ... h6 1 I .i.f4 and then g4-g5).
1 0 ... tbxg4
Black plays along with his opponent's
0
idea.
1 1 . bh7+ <ifilh8 1 2 . .tf4
Botvinnik himself explained the logic
of this reply. He refrained from 1 2 .h4
g6 1 3 . i.xg6 fxg6 14.'i:Yxg6 because of
14 .. .'ife8 ! . Now, however, after 1 2 ... g6
1 3 . i.xg6 fxg6 14.'i:Yxg6 Black does not
have the resource 14 ...'ife8 , as White
simply takes the knight.
1 56
Chapter 1 5 - g2 -g4, and White wins
It is a pity that another lovely variation Vlad i m i r Kra m n i k

remained behind the curtain: 1 2 ... g6 Peter Toth
1 3 . �xg6 fxg6 14.'iVxg6 .l:Ixf4!? 1 5 .exf4 Rio de Janeiro 1 991
tLlf8. It seems that Black has managed
to beat off the attack. He has two pieces
0
for a rook (and two wonderful bishops
into the bargain), and the white pawn
structure is in ruins. Who is playing for
a win here? It turns out that it is White
after all: 1 6 ."iih 5+ tLlh7 1 7. �gl ! , and
if 1 7... 4Jgf6, then 1 8 .tLle5 ! ! . The queen
cannot be taken because of mate in one,
and 1 8 ... �e6 1 9."iif7! is also bad for
Black.
1 2 ... tbdf6 1 3. �d3 llJh5 1 4.g4! tiJxg4
These are not the means with which The threat of g4-g5 leaves Black no
to put out the fire. The last chance choice.
was 1 3 ... 4Je4!? 14.t2:lxe4 dxe4 1 5 . �xe4 1 5. l:l.g1 e5 1 6.c5!
'ifaS+. By creating some tension in the An interesting nuance. Unlike
centre, Black diverts his opponent from Botvinnik, the young Kramnik carries
the mating attack. out his flank attack in the presence
1 4.h3 tbgf6 1 5. �e5 llJg8 1 6.0-0-0 of an unstable centre. Usually this
circumstance is in the defender's favour,
as he can meet the flank attack with a
counterblow in the centre, but in this
instance, it is the other way round: the
attacker is able to strengthen his assault
with a central blow.
But of course, geniuses are able to break
the rules ...
1 6 ... dxc5
Weakly played. Black is defending effec
tively without his rook on a8. It was
As Botvinnik explained, White's necessary to bring it into play as quickly
plan is to attack the g7-pawn, by the as possible: 1 6 ... �e6. Although after
manoeuvre 'iVc2-e2 and then moving 1 7.cxd6 !ixd6 1 8 . .l:!.xd6 'ifxd6 1 9.h3
the knight from f3 , which will leave the tLlf6 20.�xe5 'i:Ye7 2 l .'i:Yb2 White
enemy knight on h5 (the only defender retains enormous pressure, the battle is
of g7) attacked. still going on.
Only four more moves were needed for Now, however, it all ends quickly:
the realisation of this plan: 1 7. :Xd8+ 'ifxd8 1 8. tiJxe5 llJf6
1 6 ... llJh6 1 7. l:l.dg1 �e6 1 8. 'ife2 Or 1 8 ... t2:lxe5 1 9. �xe5 f6 20.'iVg6 'iVf8
�f5 1 9 . ..bf5 tiJxf5 20. llJh4! 2 l . �xf6.
Black resigned. 1 9 . ..tc4 �e6 20. ..txe6 fxe6
157
Nimzo-Indian Defence, White rushes to

advance his g-pawn; for example, l .c4
tbf6 2 .tbc3 e6 3 .tbf3 j,b4 4."irc2 0-0
S .g4!? or even the immediate 4.g4!?.
Completely new set-ups have arisen in
the Pirc-Modern, Philidor, and King's
Indian Defences, involving the sacri
fice g2-g4, without anyone demanding
immediate compensation for this pawn.
But Shakhriyar Mamedyarov (in his
game against Andreikin, Dagomys
21 .:Xg7+!
. 2008) amazed the chess world when, in
Black resigned, since after 2 l ...�xg7 a well-known position after l .d4 tbf6
22 .'iVg6+, he is mated. 2 .c4 d6 3 .t2Jc3 t2Jbd7 4.e4 eS S .tbf3 j,e7
6.ii.e2 0-0, he played 7.g4!? (previously
Finally, I want to correct a couple of White had always played 7.0-0 here,
points in Botvinnik's notes, cited above. without even thinking) and obtained a
Certainly, in our day, the early raid very strong initiative: 7... exd4 8 .tbxd4
with the g-pawn has become a stand tbcs 9.f3 tbe8 1 0 . j_e3 j,h4+ l l .�d2
ard device, which appears in many g6 1 2 .�c2 tbg7 1 3 .'iVd2 j,d7 14J:thgl
openings. There have been thousands tbce6 l S .tbfS!?.
of games in the Slav with the varia New times result in the borders of what
tion l .d4 dS 2.c4 c6 3 .tbf3 tbf6 4.tbc3 is acceptable being gradually expanded.
e6 S .e3 tbbd7 6.'iVc2 j,d6 7.g4!?. In Do not be ashamed if you want to play
various lines of the English Opening and g2-g4 !
1 58
Chapter 16
h2-h4, and White wins

The pawn push g2-g4 is a weapon of wide application. It can be used for combi
native purposes, as a signal for attack, or for positional ones: to seize space, or as
prophylaxis against .. .f7-f5 .
The advance h2 -h4, by contrast, is much more direct and simple. Without
camouflaging his intentions, White wants to open the h-file, break in down it with
his major pieces and give mate.
As a source example, we offer one from the play of the first World Champion.
In many books and databases, the fol

William Steinitz lowing move-order is given: l l ... ti:Jf6
Augustus Mongredien 1 2 .hxg6 tbxg6 1 3 .0-0-0 c5 14.tbg5 a6,
London 1 863 but then Black's 13th and 14th moves
lose all sense. The correct move-order is
given by Neishtadt in his Russian book
D
on the first World Champion. Yakov
Isaevich quotes a reliable historical
source: the contemporary reports on the
London 1 863 tournament, published in
the magazines of the time.
1 2.hxg6 ttJxg6 1 3.0-0-0 a6
Black's counterplay is too late, but his
position is hopeless in any event.
1 4. l2Jg5 lLlf6
Black has played the opening poorly. He After 14 ... h6 1 5 .tbxf7 .l::rxf7, both
has castled early, but for some reason 1 6.'i!Vg4, and the direct 16 . .txf7+ 'iit>xf7
put his knight on e7 instead of f6. As 1 7.tbc4 (with the threat of 1 8 .ti:Jd6+ and
for the centre, one cannot say that it is 1 8 .tbxb6) win.
closed, but it is stable.
All the preconditions for a flank attack
are present. Steinitz did not have to be
asked twice:
1 0.h4! t:i::KJ 7
The knight hurries to f6, but never gets
there. There is no sense either in 1 0 ... h5
l l .tbg5 , followed by g2-g4, but perhaps
he should have tried to ease the pres
sure with 1 0 .. .'�fc8 , and if l l .h5 , then
1 1 ....ta6. Augustus Mongn§dien, thankfully, gives
1 1 .h5 cs his opponent an opportunity to carry out
159
a combination, which will last for centu William Steinitz

ries. In our day, players play this combi Emanuel Lasker
nation likes a pianist playing a scale, but New York 1 894
they are using Steinitz's prototype.
1 5. ttJxh7! ttJxh7 1 6. :Xh7!
D
There was also a win with 1 6 .�h5 ttJf6
1 7.�xg6 �c8 1 8 .�h7! ttJxh7 1 9. ith6,
and the simple 1 8 .l::[h4 (but not 1 8 . ith6
�g4) followed by doubling rooks on
the open file. But the tempo of the
attack would be slowed, whereas now
it proceeds in a single breath.
1 6 ... �xh7 1 7. 'ff h 5+ �g8 1 8.l:th1
l:te8 1 9. 'ifxg6 'flf6 20 . .bf7+! 'ifxf7
21 .l:th8+! Black's position is almost entirely
The final blow. devoid of drawbacks, and anything
21 ... �xh8 22. 'it'xf7 one criticises sounds like nit-picking.
After all the fireworks hanging over You say Black is cramped? After the
the position, material is again equal, programmed ... d6-d5 this will be put
but Black cannot prevent White getting right. His piece coordination lacks har
his knight into the game and so he pre mony? This will be corrected in time.
ferred to resign at once. The main thing, though, is that White
lacks an active plan. He cannot impose
Steinitz was a great champion and also his will.
a great experimenter. One only has It looks as though we face a long battle,
to think of the variation of the King's with lots of quiet manoeuvering. But
Gambit named after him, in which Steinitz had a different opinion:
White's king comes to e2, with a full 1 1 .h4!? 'ffc7 1 2.tlJg5
board of pieces. If played by a beginner, This looks like a wild, unjustified
we would call it eccentric, but when a attempt to attack, which is destined
World Champion plays such a thing, we to fail. But the calculation of concrete
have to ask ourselves: why?! And ana variations shows that things are not so
lysing such lines makes us reassess the simple.
norms, and reconsider long-established In order to understand what White is
assumptions and beliefs. threatening, let us give him the move.
Steinitz himself laid out the rules and (The attack continues 1 3 .h5 !?, and after
himself broke them. Just as inventors 1 3 ... tt:Jxh5 - 14.l::[xh5! gxhS l S .�xhS
sometimes go too far, so Steinitz, as the with attacks on f7 and h7.) This is seri
founder of the new positional school, ous; it means that Black needs to stop
often went beyond the realms of the coasting and take urgent measures.
theory he had created. As a person and Let us try the attack on the knight:
as a player, he was beyond any dogmas. 1 2 ... h6. In reply, Steinitz had prepared
Even in a World Championship match, the double piece sacrifice 1 3 .g4!? hxgS
he was not afraid to take risks. 14.hxg5 ttJh7 l S .ttJfS ! ?.
1 60
Chapter 1 6 - h2 -h4, and White wins
but after 14 ... g5 , the attack is visibly in

decline.
Therefore, Steinitz prepares a second
attack with f2-f3 , g2-g4, and then,
according to circumstances, h4-h5 or
tt::l e 3 -f5 .
1 3.f3 :Sd8
Lasker tries with all his might to show
his opponent that the attack is just bluff.
Otherwise, he might have chosen 1 3 ...
h6!? and if 14.g4 hxgS 1 5 .hxg5 tt::lh 7
analysis diagram
1 6 .tt::lfs , then 1 6 ... gxf5 1 7.gxf5 tt::lx fS !
This combinational idea, in various The line d1-h5 i s closed, and the queen
forms, will hang over the position for cannot get to the h-file in one move.
a long time. After 1 5 ... gxf5 1 6.gxf5 And without that, the attack has no
tt::lx fS! 1 7.'tihs tt::lh6 1 8.gxh6 .if6 chance of success.
1 9 . .ig5! Black loses after both 1 9... .ixg5 1 4.g4!?
20.'iYg6+ �h8 2 I .'iYg7#, and 1 9 ... tt::lxg5 This move looks very simple and
20.'it'g6+ 'it>h8 2 I .'ifxf6+ <iitg 8 straightforward, but in fact, it conceals
22 .'ifg7#. The only defence is 19 ...'tid8 ! an idea which is very hard to see. White
20. l::rg 1 <iith8 , but this hangs o n a very is freeing the second rank for his queen!
slender thread. One can understand that Now if 14 ... h6, then 1 S .'ife2! hxgS
Lasker was not keen to play along with 1 6.hxg5 , and Black's position may well
his opponent, especially given that he be indefensible: 16 ... tt::lh 7 1 7.tt::lf5 ! gxfS
has at his disposal the principled reply... 1 8 .'tih2 or 1 6 ... tt::le 8 1 7.'ifh2 f6 1 8 .exd5
1 2 ... d5!? etc.
«Answer a flank attack with a central counter!» 1 4 ... dxe4 1 5.fxe4
Lasker plays like a true follower of the
Steinitz school. Meanwhile, the founder
of that school behaves as if he doubts
the truthfulness of its postulates.
All of these considerations are highly
conditional. The position has become
much sharper and concrete calculation
predominates over general dogmas.
Whoever has the greater fantasy, and
can generate more ideas, and who can
calculate variations better, will triumph.
What should White do? After 1 3 .h5 h6 The culminating moment in the game.
14.hxg6 hxgS, no decisive continuation Lasker (somewhat late) played:
is to be seen (the fact that Black has 1 5 ... h6
closed the a2-g8 diagonal is important), And after
and White is a piece down. Of course, 1 6. tlff3!
it is possible to retreat with 14.tt::lf3 , he realised that he cannot accept the
161
knight sacrifice: 1 6 ... hxgS 1 7.hxg5 tbh7

1 8.tbfS ! (this combination arises again
and again) 1 8 ...gxfS 1 9.'ifh3 .l:lfe8
20.'ifxh7+ �f8 2 l .g6! with mate in
all variations: 2 l ...fxg6 22.�h6 �xh6
23 .'ifxh6# or 2 l ...tbxg6 22.�h6 �xh6
23 .'ifxf7#. In the meantime, the threat
is 1 7.tbxf7 kf.xf7 1 8 . �xf7+ �xf7 1 9.gS .
What should be done? Lasker nonethe
less found a good defence:
1 6 ... ..te8!
analysis diagram
The battle went on, and Steinitz even
tually won. And now simply 20.�e2 . A grand
thought! White is a pawn down and
Many experts (possibly copying one the queens are off, yet even so, White
another) suggested, in the position of retains the initiative and every chance
the last diagram, 1 S ...'ifc8. of winning. His light-squared bishop is
very strong, and his other pieces sup
port it. With hindsight, the entire plan
l l .h4!? looks very unified and there is
no apparent refutation.
One would expect that, over the cen

tury and a half that separates us from
the Steinitz-Mongredien game, the
advance of the h-pawn would have
been investigated in all its subtleties and
nuances, and many new aspects found.
analysis diagram
This has indeed happened. Even being
Forcing 16 . .l::i.g l . And then 16 ... h6, and the experimenter that he was, Steinitz
the combination shown above does not would probably be surprised to see that
work, because the rook is not on the nowadays, players push the h-pawn
h-file. not just on the cusp of the opening
1 S ...'ifc8 is Neishtadt's idea. Yakov middlegame transition stage, say the
Isaevich did an enormous amount of tenth move, but even as early as move
work, analysing this position. But in five and even earlier!
this particular case, I will permit myself In our day, the opening idea h2-h4 is not
to disagree with him. After 1 S ...'ifc8 fully classical and not entirely normal,
Steinitz probably intended 16.h5! �xg4 but at the same time, not especially
(even worse is 1 6 ...tbxg4 1 7.tbxh7! �xh7 unusual either. As a concrete exam
1 8.hxg6+ �xg6 1 9.tbxg4 �xg4 20.'ifd2! ple, take the Dutch Defence, Leningrad
with mating threats. For example, 20 ... Variation. The first moves:
f6 2 l..� h6+ �xh6 22.'ifxh6#) 1 7.tbxg4 1 .d4 fS 2.tLlc3 dS 3 . ..tg5 g6 4.e3
'ifxg4 1 8.'tixg4 tbxg4 1 9.hxg6 hxg6. ..tg7
1 62
Chapter 1 6 - h2 -h4, and White wins
And suddenly: One year later, in a game Khalifman

5.h4!? Legky (played in Tashkent), the players
The game Khalifman-Lerner (Kuibyshev repeated the variation. But already by
1986) went as follows: then it was Black who enjoyed the sur
prise factor. He was well-prepared and
was able to survive the opening stage:
8 ... tLlbd7! (instead of 8 ... lLlxh5) 9.h6 (or
9.hxg6 hxg6 1 0 . .l::txh8+ �xh8 1 1 .tLle5
�f7 1 2 . �e2 'ifas 1 3 .'ifd2 tLlxeS
14. �xe5 0-0-0, and Black completes
his development, Ionov-Malaniuk,
Budapest 1 989) 9 ... �f8 1 0 .'ifd2 �f7
1 1 .tLle5 e6 1 2 .f3 �e7 1 3 .g4?! tLlxeS
14. �xe5 fxg4 1 5 .fxg4 0-0 1 6 . �e2 .
5 ... i.e6 6.tl::lf3 Here, in Khalifman's opinion, Black
6.h5 h6 is too early. could have solved all his problems with
6 ... c6 7 . ..tf4 16 ... tt:Jd7! 1 7. �g3 �gS .
Now that the bishop is no longer able to
be hit with tempo, everything is ready Alexander Khalifman played his most
for the further advance of the pawn. memorable game with this idea in
7 ... tl::lf6 8.h5! ? tlJxh5 9 . .:Xh5!? 1 984, at the USSR junior championship.
gxh5 1 O.tlJg5 ..tg8 1 1 . 'Wxh5+ �8 His opponent was Alexey Dreev. Khalif
White has sacrificed the exchange, man was Black, and his h-pawn rushed
obtaining in return simply active play. forward - as early as move 3 !
The next few moves are easy to under This i s what happened:
stand: queenside castling, and then all 1 .e4 e5 2.tlJc3 tlJc6 3.g3 h5!?
the pieces are brought to the kingside.
It will be a long time before the black
king knows any peace. One should also
mention the surprise factor: thanks to
the march of the h-pawn, the opening
phase has been effectively swallowed
up. Black finds himself having to defend
his king almost from the very first
moves. This is extremely unpleasant
from the psychological point of view.
There followed: 4. tl::lf3 h4! ? 5.tbxh4 .:Xh4! 6.gxh4
1 2. i.d3 e6 1 3.g4 h6 1 4. tLlf3 i.f7 'Wxh4 7. i.g2
1 5. 'Wh3 fxg4 1 6. 'Wxg4 tl::ld 7 Grandmaster Kryakvin, in a lengthy
1 7. i.d6+ 'itg8 1 8.0-0-0 article called 'Great masters of the
It is clear that White's attacking chances flank attack' (Chesspro 201 1) showed
have increased, and it is no longer a ques what he believes is the best defence:
tion of compensation for the material, 7. .l::tg l ! giving the following variations:
but of an actual advantage for White. 7 ... 'ifxh2 8 . .l::tg2 'ifh4 9.d3; 7 ... �c5
1 63
8.l:tg2 d6 9.d3 �h3 10.�g5 'iVh7 Suddenly the queen turns out to be in
1 l..l:f.g 1 ; 7 ... tbf6 8 .d3 'iYxh2 9. .l:tg3 d6 great trouble. The threat is 16 ... �d8
1 0 . �e3. Conclusion: Black's initiative 1 7. �g5 f6, and after any retreat of the
can be extinguished, and then the extra bishop, 1 8 ... g5 . There is no satisfactory
exchange will start to count. defence. It is interesting that, unlike the
But that is analysis. At the board, even game against Lerner, here even queen
such a subtle positional player as Dreev side castling was not needed. And the
could not solve the problems facing him. game itself was decided in 1 5 moves.
Evidently, he was shaken out of himsel£ Would this have happened, had Black
The shock factor again played its role. not pushed his h-pawn?
7 ... i.c5 8.0-0 As Khalifman's trainer, Gennady Nesis,
Hartston-Mariotti (Skopje 1972) con pointed out, 'For a long time, the game
tinued no less interestingly: 8.d4 �xd4 with Dreev was Alexander's calling
9.'ife2 �xc3+ 1 0.bxc3 d6 1 1 .0-0 g5 !?. card'. And the march of the h-pawn
Black has both combinational and tac occupied a regular place in the St Peters
tical compensation for the material. burg player's attacking arsenal.
One only needs to imagine knights on
f4 and h4, to understand that Black has In the following game, the h-pawn only
no need to force matters, but can qui manages to make one step. It does not
etly complete his development and start get any further, because it is blockaded.
transferring pieces to the kingside. But the blockade has its weaknesses.
8 ... lLlf6 9. 'iff3 d6 1 0 .• g3 'ifh7 The square g5 falls into White's hands
1 1 .d3 lLld4 1 2. i.e3 i.d7 1 3.b4 and the attack proceeds along different
One can recommend as the least evil rails. The result is a masterpiece.
1 3 . �xd4 �xd4 14.tbe2 �b6 1 5 . a4 a5
1 6 .c3. Kryakvin continues the variation: Alexander Khal ifman
16 ... 0-0-0 1 7.b4 .i:th8 1 8 .h3 (or 1 8 .bxa5 lan Rogers
tbh5 1 9.'iVf3 tLlf4! 20.h3 tbxe2+ Groningen 1 990
2 1 .'iYxe2 �xa5 22 .'iVf3 �xc3) 1 8 ... g5 .
This was how a game of the Serbian GM
Ivanisevic went. Black has a firm grip
on the initiative.
1 3 ... ..ib6 1 4.l:.ac1 lLlhS 1 5. 'ifh4 c6!
Black has mixed up the opening sys

tems, and his hybrid of the King's
Indian and Queen's Gambit cannot
bring him anything good. The pawn
on e5, which players often refer to as
a 'nail', has forced the knight from f6.
1 64
Chapter 1 6 h2 -h4, and White wins
-
The path of the h-pawn is clear and Exchanging off an unnecessary piece
there is no time to lose. and at the same time freeing the
9.h4! e2-square for the knight.
But not 9.tbxd5 tbxeS I O.tbxe7+ 'i:Vxe7 1 4 ... l:.f6 1 5. tL\f4 a6 1 6 . .h:c6 1i'xc6
1 l .dxe5 tbc6, nor 9.0-0 tbc6. 1 7.l:.c3 1i'd6 1 8. 1i'e2 tL\84
9 ... tL\b6 Khalifman considered that Black's last
Black's decision in the game Adamski chance to save himself was 1 8 ...'iY'b4,
Majer (Giessen 1 992) can only be with the idea after 1 9.il.e3 to reply
described as panic: 9 ... tbc6 I O.hS tbxd4 1 9 ...tba4 20.tbxd5 tbxc3 . He then gives
1 1 .tLlxd4 tbxeS. After 1 2 .hxg6 fxg6 1 9.�fl ! 'iVxd4 20.�e3 'iY'es 2 1. . lid1 ,
13 . ..te2 , he had no compensation for retaining the initiative.
the piece. In my opinion, the simple 1 9.a3 'iY'xd4
Rogers' play is stronger. After 1 0.h5 he 20 . .l:f.c7 is also convincing enough.
has prepared 1 0 ... ..tg4! , and if White 1 9.l:.g3 b5 20. 'itf1 ! 1i'c6
sticks to his guns, he comes up empty
handed: 1 1 .hxg6 fxg6 1 2 . .l::tx h7 ..txf3 ! ,
but not 1 2 . . .�xh7 13 .tbg5+ �g8
14.'iY'xg4. Jumping ahead, we can say
that Black continued to defend pretty
well, but sadly for him, White played
brilliantly.
1 0. tL\g5! h5
The first in a series of only moves,
which, alas, do not help Black. After
1 0 .. .f6 White wins with 1 1 .tLlxh7! �xh7
1 2 .'iY'h5+ �g8 1 3 . il.xg6, whilst after With great difficulty, Black has man
1 0 ... h6, he decides matters with 1 1 .h5! aged to set up some sort of pawn bar
hxgS 1 2 .hxg6 fxg6 (12 ... g4 13 . ..th6 rier. But it does not look at all reliable,
fxg6 14.'iY'c1) 13 . ..txg6 ..tfs (or 1 3 ... g4 and by a knight sacrifice, White breaks
14 . ..th7+ �f7 1 5 .'iVc2 �e8 1 6 .'iY'g6+ through to the enemy king.
.l::tf7 1 7. ..tg8) 14.'iY'hs ..txg6 1 5 .'iY'xg6 21 .tL\xg6 ! ! .:.Xg6 22. 1i'xh5
'iY'e8 1 6 .'iY'h7+ �f7 1 7.IIh6!. Mating motifs hang over the position
1 1 .e6 f5 everywhere. For example, 22 ... l::tf6
Again, Black has no choice. The varia 23 .'iY'h7+ �f8 24.'iY'h8+! �xh8
tions 1 1 .....txe6 1 2 .tbxe6 fxe6 1 3 . �xg6 25 .t2:lh7+ �e8 26.l:tg8+ I:tfs 27. 1:1xf8#;
and 1 1 ...f6 1 2 . �xg6 fxgS 1 3 .hxg5 22 ...'iY'e8 23 .'iY'h7+ �f8 24.h5 !th6
�xe6 14.'iY'xh5 both lead to immediate 25 .'iY'xh6 il.xh6 26.tbh7#; 22 .. J:rh6
defeat. 23 .'iY'f7+ �h8 24.tLlf3 ! l::th 7 25 .tbe5
1 2. tL\e2 1i'e8 1 3.l:.h3!? 'iY'xe6 26.'iY'e8+ 'iY'g8 27.tLlf7#.
We can see the analogous device in the Black prefers the most aesthetic way of
game Khalifman-Brynell on page 56. losing:
1 3 ... tL\c6 1 4. �b5! 22 ...f4 23. 1i'h7+ 1 -0
1 65
Chapter 17
Overcoming the Blockade

Playing against the hanging pawn pair on c6 and d5 is a great art. The recipe has
been well-known since ancient times: a blockade on the squares d4 and c5. But
what should one do if the blockade does not materialise? Then, strangely enough,
the correct strategy may consist in provoking Black into the advance ... c6-c5 .
The assessment of the move 1 6.'i:Vd4

Boris Gelfand depends on whether the advance ... c6-
Igor Khenkin c5 is favourable for Black, or not. And
Riga 1 987 how do we decide this? Only by one
criterion: will Black manage to follow
up with ... d5-d4 or, at the very least,
0
manage seriously to threaten this?
All the while the pawns are on c6 and
d5, they have little mobility, but they
are also harder to attack - the d5-pawn
is defended, and the c6-pawn is easy to
defend, for example by a bishop from d7.
But everything changes once the pawn
advances to c5. Statics give way to
dynamics. If Black manages to get in
It would be wrong to think that the ... d5-d4, he will be all right, but if not,
position is completely safe for White. then the pawn pair will come under
Thus, in the game Gavrikov-Zaichik threat. Most likely, one of them will be
(Riga 1980) , after 1 6 .tt::la4 itaS!? 1 7. .ltc3 lost.
itxc3 1 8 .tt::lxc3 �e5 1 9.tDa4 tt:Je4 1 6 ... c5
20.'iVd4 l:tgs 2 l .f4? (2 I . itfl was essen With hindsight, we can say that this is
tial) 2 I ...l:!xg3+! 22.hxg3 'i:Vg6 23 .g4 a mistake. White plays his next three
.ltxg4, Black, after sacrificing a rook, moves with tempo. Black will just not
developed a mating attack. have time to play ... d5-d4.
1 6. 1i'd4! 1 7. 1i'h4 ..if5
A move which contains many plans. The d5 -pawn could also have been
Firstly, it ties down the knight on f6, defended by 1 7... itd7 1 8 .l::f.e dl itc6. But
which cannot now come to e4 because this is still not saving Black: 1 9.t2Jxd5!
of the mate threat along the long diag tt:Jxd5 20. itf3 'iVe6 (20 ... l:f.ad8 2 l .e4)
onal. Secondly, White begins to seize 2 I .'i:Vg5 .ltes 22.itxd5 itxd5 23.itxe5
the blockade squares d4 and c5. On 'ifxe5 24.l:f.xd5 , or 2 I ...f6 22.'iVxd5 ! .
16 ... �a5 1 7. l:f.edl .ltb4 there follows 1 B . .:ed1 :ada 1 9. 'ifg5!
1 8 .tDa4 (and if 1 8 ... :r.e4, then 1 9.'iVb6), Gelfand chooses the strongest of the
inching towards the desired objective. many tempting continuations. One
1 66
Chapter 1 7 - Overcoming the Blockade
option is 1 9.tba4 tbe4 20.f3 , but after 31 . :Xh5 i.a5 32.b4

20 ...tDf6 the pawn on e3 starts to hang. Black resigned.
And in the event of the immediate A strong game. When it was played,
1 9.tbxd5 tbxdS 20. �f3 �e6 2 1 .�xc5 Gelfand was not yet 20. But even so, his
tbxe3 (more subtle is 20 ...�c6!? 2 1 .b4 best qualities: positional understand
�e6 22. l::txc5 tbxe3!) the game takes ing, accurate calculation of variations
on unnecessary complications. and erudition, were already clear to see.
1 9 ... �g6
After 1 9 ...g6, White has many ways to You may ask where this erudition comes
achieve a material advantage. The most from. The fact is that four years ear
principled of these is 20.tba4 tbe4 lier, there had been a game Gavrikov
21 .�h6 f6 22.�a3. Mochalov.
By retreating the bishop to g6, Black 1 . lLlf3 c5 2.c4 lLlf6 3.lLlc3 e6 4.e3
allows his opponent to realise his main d5 5.d4 a6 6.cxd5 exd5 7.i.e2 lLlc6
tactical idea. 8.0-0 i.d6 9.b3 0-0 1 0.i.b2 cxd4
1 1 . tbxd4 :ea 1 2. Ac1 i.c7 1 3. Ae1
1i'd6 1 4.g3 i.e6
The move order in Gelfand-Khen
kin was slightly different: 14 ... �h3
1 S .tbxc6 bxc6.
1 5.tbxc6 bxc6 1 6. i.f3 ..tf5
20. tbxd5! tbxd5 21 . �f3 'it'e7

22. :Xd5 'it'xg5 23. :Xg5 l:d2
24 . ..tc3 :Xa2 25. :Xc5 i.b6 26. l:e5
:ca 27. �b7 Ab8 28. i.d5
The forcing variations continue. White
has long since won a pawn, but from
inertia, Black keeps on playing active Now there followed:
moves. Finally, the sequence ends, but 1 7.1i'd4! c5 1 8. 'it'f4! 1i'd7 1 9.'it'a4!
almost immediately, Khenkin makes 'it'e6 20. tbxd5! tbxd5 21 . :Xc5
the decisive mistake. Aed8 22. 1i'c4
28 ... h6 29.h4 :ca And Black resigned. As we see, com
He could have prolonged his resistance pletely the same concept.
by tying the rook on eS to the defence
of the bishop: 29 ... l:r.d8 . Now, however, And what if Black cannot or does not
it all ends very quickly: play ... c6-c5 ? How do we break down
30.h5! hh5 the defences then? The classic example
30 ... �d3 3 1 . l:te7 �f8 32.l::tb 7 �cs on this theme is a game of Rubinstein's,
33 .b4 or 32...�d8 3 3 . �b4. incidentally Gelfand's favourite player.
1 67
Akiba Rubi nstein Here too, a tempting alternative was

Georg Salwe 1 9 .. .'iVb7 20.e3 �fS , taking control of
Lodz 1 908 the square c2 (but not immediately
1 9 ... ..1fs 20.e4).
20.e3! 'Wb7 21 . tt:Jc5 ttJxc5 22. l:l.xc5
D
1 8. 'Wd4!
Several moves earlier, White had
played f2-f3 and !tfl -f2 , starting a The white pieces have suddenly found
deep regrouping. This will be crowned harmony. Manoeuvres which seemed
with success when White plays e2-e3 , far-fetched, even artificial, have sud
�g2-fl and l:i.f2-c2 . But first, he must denly proved to be full of sense.
take control of the squares cS and d4, This is the usual thing, when one plays
otherwise, if the c6-d5 -pawns start over Rubinstein's games. How he man
moving, the entire manoeuvre will be aged it is a mystery, but the fact is that
pointless. White has significantly increased his
1 8 ... l:l.ee8 advantage. However, the battle is not
From here to the end of the game, over yet. Here and on the next move,
Black plays passively and Salwe does not it was essential that Black play ... a7-a5 ,
really manage to pose his opponent a preventing the press with b2-b4.
single problem. 22 ... l:l.c7 23. l:l.fc2 'Wb6 24.b4!
What should we advise instead? The Now the blockade on the queenside
pawn sacrifice 1 8 .. J:f.ae8 1 9."iVxa7 acquires a permanent look. It is already
i.fs does not achieve very much. too late to advise Black.
White gradually manages to defend 24 ... 86 25. l:l.a5
everything: 20.�fl tt::lf8 2 1 .'ti'd4 tt::le 6 White could already win a pawn with
22.'iVd l . The initiative dies off. 2S . .l::f.xd5 . But he first strengthens the
It was worth considering 1 8 .. .f6!?. If position of his pieces, realising that the
White then blindly follows his planned material is not going anywhere. Such an
moves with 19 . ..1fl 'ti'b7 20.e3 .l::f.ae8 approach would later become typical of
2 1 . .l::f.fc2 , then after 2 l ...�f5 ! 22 . .l::f.xc6 Gelfand too.
tt::le S 23 .'it'xd5+ ..1e6, he risks losing 25 ... l:l.b8
material. Of course, this variation is not A cheerless endgame awaits Black after
forced, and White can correct his origi 25 ..."iVxd4 26.exd4 .l::tb 7 (26 ... ..1c8
nal plan at any moment. 27. .l::txd5) 27.a3 (less convincing is
1 9. �f1 ! l:l.ec8 27. .l::f.xc6 l::f.xb4 28J:tcxa6 .l::f.xa6 29.l:txa6
1 68
gS) 27 ... l::rb 6. But at least this would be �f6 27 . .l:rccl 'iYeS he falls into an
an endgame with material equality. unpleasant position. He could buy his
26.a3 :&7 27. :Xc6! 'it'xc6 28. 'ifxa7 way out with the exchange 27.l':Ifcl
The realisation of the extra pawn did �xc3 28.l:lxc3 , but this is not every
not pose White any difficulties. Black body's cup of tea.
resigned at move 3 8 . 23 . .l:[e1
One can understand Smyslov's reluc
A very important game for this struc tance to abandon the game as a draw
ture, which has been deeply studied (which would be practically inevitable
from both sides - for White, to under after 23 . .l:f.fcl cS 24-.tbxcS tbxcS 2S.�xcS
stand Rubinstein's plan, and for Black, �xeS 26 . .l:rxcS 'iYxe3+ 27.'iYxe3 .l:rxe3).
to understand how to fight against it. But it is more difficult to understand
As far as the latter is concerned, it is why he did not play 23 .b4-. Perhaps he
clear that passive play is not appropri was worried by the weakening on c4-.
ate. Black must defend actively. And if For example, 23 ...tbb6 24-.tbcs 'iVd6
it is hard to prevent the blockade on and then ... tbb6-c4-, or even 23 ... tbeS ,
d4- and cS , then Black should at least be and i f 24-. �xeS 'iYxeS 2S . .l:rxc6, then
able to stop the stabilising move b2-b4-. 2S ... d4-!? and White has to part with
He should find the right moment to play material in any case: 26.exd4- 'iYbS or
... a7-aS and put one of his major pieces 26.'iYxd4- 'iYxd4- 27.exd4- �d3 .
on the b-file, establishing control over The game we are examining was played
the key square b4-. in the final Candidates' match. Smyslov
was much older than his opponent, and
Vasily Smyslov besides, Kasparov was known for his
Garry Kaspa rov phenomenal calculation of variations. It
Vilnius 1 984 is perhaps not surprising that Smyslov
adopted restriction tactics. As far as
• possible, he followed the approach of
drying the game up, avoiding direct
conflict.
Of course, Smyslov knew the Rubinstein
plan and followed it. In the diagram
position, his pawn is already at f3 , cov
ering the squares e4- and g4-, and thus
taking over part of the light-squared
bishop's job. In accordance with the
plan, the bishop belongs on f1 and the
22 ... :Sb8!? king's rook on cl . But the immediate
It was simpler to begin with 22 ... 23 . .!::tfcl , as shown above, allows the
aS, but Kasparov changes the move break 23 ... cS. The move 23 .b4- involves
order, setting a trap along the way. If agreeing to sharp and concrete play,
White grabs the pawn with 2 3 . �xa7, which Smyslov was not prepared to do.
then after 23 ... .!::i.a 8 24-.tbcs (24-.'iYd4- So he plays 2 3 . l:le l , losing a tempo, but
�f6) 24-...'iYd6! 2S .tbxd7 l:lxa7 26.tbcs defending e3, and so stopping ... c6-cS .
1 69
23 ... a5! 24 . .tf1 hS! ? A draw was soon agreed.

Having prevented b2-b4, Black does Kasparov defended extremely well in
not rest on his laurels. He actively seeks this game. But one should not think
counterplay in the centre and on the that the combination of ... .l:!.a8-b8 and
kingside. ... a7 -aS in this structure is his exclusive
2S. :ec1 know-how. Of course not. This active
defence has been known for a long time
and has even passed muster at the very
highest level, in a match for the World
Championship.
Tigran Petrosian
Boris Spassky
Moscow 1 969
An important moment. It seems that

if Black wants a draw, he can make it
at once: 2S ... cS 26.tLlxcS tbxcS , and
after all the exchanges, he takes on
e3. But the position of the pawn on aS
makes the exchange sacrifice attractive:
27.l::i.xcS! i.xcS 28 . .l::f.x cS a4 29. i.bs
etc. Stronger is 28 ... l::f.e c8 , but the pros
pect of fighting two connected passed Having removed the light-squared
pawns, supported by the two bishops, bishops from the board is a definite
obviously did not appeal to Black. achievement for Black. Even so, if White
Kasparov prefers not to force the game, manages to get one of his minor pieces
but, on the contrary, to increase the to cS and then establish the blockade
tension: with b2-b4, he can probably hope for
25 ... tbe5! an advantage. Black must hurry.
Immediately after the game, Grandmas 1 8 ... a5! 1 9 . .1:.c2
ter Evgeny Vasiukov suggested 26.';ii'f2 . Or 1 9. i.cs i.d8 20.Ilc2 'ifb7 2 1 .�gl
I think Kasparov would have been very tbd7 with rich counterplay.
happy with the position after 26 ... h4!?. 1 9 ... a4 20.tt:\d2
Smyslov now simplifies to a draw. The «<t was necessary to play 20. tbcS, although after
Rubinstein plan has not worked this 20 ... i.xcS 2 I . i.xcS White must think about
time, but half a point is better than equalising: the knight is stronger than the bishop,
nothing. the pawn on c6 is easily defended and Black has
26 . ..be5 11Vxe5 27. :Xc6 .tf6 good chances of developing an initiative on the
28 . .1:.6c5 11Vxe3+ 29. 11Vxe3 :Xe3 kingside» Boleslavsky.
-
30. :Xd5 :Xf3 31 . �e2 .l:.e3 32 . ..bh5 20 ... 11Vb7! 21 .<ofilg1 .l:.ac8 22 . .1:.fc1
..bhS 33. :Xh5 gS! tt:\d7 23. tt:lf3 cS
1 70
Black's problems are obviously behind 27 ... a4 is too early, in view of 28.b4, but
him. now the threat of ... aS -a4 forces White
to hurry.
28. ti:lf5+ .bf5 29. l:.xf5 f6 30. l:.c4
Alexander Alekh ine l:.b5 31 J ld4+ �e6 32.l:.xb5 cxb5
Jose Rau l Capa blanca After a few more moves, a draw was
Buenos Aires 1 9 27 agreed.
• Other examples well worthy of atten

tion are those in which the side with
the pawn duo c6/dS (or c3/d4) starts
a determined kingside assault straight
from the opening. As they say, the
hunter and the hunted change places.
After all, the pawn pair c6/dS (or c3/
d4) is a strong pawn centre. With quiet
play on the queenside, as we have seen,
such a centre sooner or later becomes
As you will recall, Rubinstein overcame a weakness, but what if we use it for a
his opponent in the endgame. To the determined attack on the kingside?
weakness on c6 was added another . on
a7, whilst the white rooks dominated Yu ri Razuvaev
the a-file as surely as they did the c-file. Ivan Farago
This example is exactly the same in terms Dubna 1 979
of structure. If it were his move, White
would play b3 -b4, and then transfer his
0
king's rook to c1 and c3 . Then, accord
ing to circumstances, either pressurise
the c6-pawn (if necessary, also breaking
with b4-bS), or transfer his heavy artil
lery to the a-file. Another variant is also
possible, whereby one rook stays on the
c-file and the other goes to a6 , so as to
attack both weaknesses at once.
In any event, Black will not survive. But
it is his move, and this saves him. 1 3.h4!? tba5 1 4. tt:lg5 h6 1 5. 'ifh5
21 ... a5! 22. l:.hc1 �d6 23 . .1:.1 c3 :as �d5 1 6. ti:lh7! J:.e8 1 7 . .bh6! gxh6
24. tt:lb5+ �e7 25.tbd4 �d6 1 8. 'ifxh6 f5 1 9.l:.e3 .bh4 20.l:.g3+!
White's position is still better, but it is .bg3 21 . 'ifg6+ �h8 22. tt:lf6!
not clear how he can strengthen it fur Winning.
ther. Alekhine decides to change the This game deservedly attracted a wide
structure, but this only leads to new spread press. Razuvaev's admission that
exchanges. he had done most of the work at home,
26.e4!? dxe4 27. �e3 l:.b8! and was only waiting for a chance to
171
use it in practice, only added fat to the Black was not obliged to lose so
fire. Everybody understood that it was tamely. Instead of 14 ... h6, he has the
not just a chance success. defence 14 ... .ixg5 1 S .hxg5 fS ! 1 6.gxf6
What strikes one's eye? White played (16. l::!.xe6 "iVdS) 1 6 ..."iVxf6. The conse
h2-h4 after he had castled kingside. quences of 1 S . .ixg5 "iVdS 1 6."iVg4 are
This is unusual. And the point is not also unclear. The game Poluliakhov
so much in the pawn sacrifice (few Balashov (St Petersburg 1 998) went 1 6 ...
black players are going to be tempted fS 1 7."iVg3 �ac8 1 8 J �e5 "iVd7 1 9 . .ibs
by 13 ... .ixh4 14.lt:lxh4 "iVxh4 1 S . l:te3, �c6 20. �xc6 lt:lxc6 , and the play
followed by transferring the rook to ers agreed a draw. More ambitious is
g3 or h3), but in how successfully he 1 9. .l:tae1 .l:txc3 20 . .!::!.xe6 (hoping for
camouflaged his intentions. Convention 20 ... .ie4 2 1 .�e7 "iVxd4 22. �h6), but
makes us think that 0-0 and h2-h4 are by exchanging queens with 20 ..."iVxd4
moves that do not go well together, but 2 1 .l::!.d6 f4! 22 . .l:f.xd4 fxg3 23.fxg3 .l:tc7,
Razuvaev shows that this is not always Black can still defend stoutly.
the case.
The second memorable aspect is how As well as the decentralising 1 3 ... lt:laS,
successfully the c3/d4-pawn duo fit Black has also tried a mass of other
into the overall landscape. An isolated moves:
pawn on d4 could be attacked, or the
black knight could use the square b4
as a transit, to come via dS to f6, to
defend the kingside. But the pawn on
c3 defends its colleague on d4 securely
and deprives Black of these ideas. Mean
while, the main black plan in such a
position, the blockade on the squares c4
and dS, could not be realised, because
Black never had the time - his king was
in immediate trouble!
Time brought some amendments and 1 3 ..."iVdS; 1 3 .. ..l:f.c8; 1 3 ... .if6 etc.
refinements. In particular, it was found In his turn, White over the same
that after 13 ... tbas 14.lt:lgs , period widened notably the range of
his attacking tries. Black learned to
meet the sacrifices on h7, f7 and e6,
so White tried attacking in the centre.
So, after 1 3 ... .if6, two lovely minia
tures were played: 14.lt:lgS g6 1 S ."iVg4
hS 1 6."iVg3 lt:le7 1 7. �a3 "iVd7 1 8 . .l:f.ad l ! ,
and now the choice i s 1 8 .. Jiac8
1 9.d5 ! ! (Timofeev-Eliet, Cappelle-la
Grande 2004) or 1 8 ... .l:I.fe8 1 9.d5 ! !
(Solomunovic-Zelic, Neum 2005).
1 72
The theory of this variation was dis be taken: 1 3 .hxg4 �h2+ 14.\t>xh2
cussed in detail in the book on the (nothing is changed by 14.\t>h1 "ifh4)
Caro-Kann, Panov Attack, by Karpov 14 ... "ifh4+ 1 S .\t>g1 hxg4 and mate.
and Podgaets. But we will now look at One can ignore the piece with 1 3 . �b2 ,
some examples where Black, having the but the position remains danger
c6/dS -pawn duo, tried a similar attack. ous. There is one nice, though by no
And it paid off. means obligatory variation, which
runs: 13...�c7 14."ifcl "ifd6 1 S .f4
Alexey Aleksand rov �b6 1 6.hxg4 (stronger is 1 6 . .l:If3) 1 6 ...
Jaan Ehlvest hxg4 1 7.tbd1 "ifh6 (1 7... g3 !?) 1 8 . �xg7
Bled 2002 Vi'xg7 1 9."ifxc6+ \t>e7 20."ifxa8 l:thl+!
2 1 .<it>xh1 "ifh6+ 22 .�g1 g3 or 2 1 .\t>f2
• g3+ 22 .\t>e1 �aS+, in both cases with
mating threats.
There is no doubt that Ehlvest saw these
variations, but evidently did not calcu
late them clearly enough. He preferred
not to burn his boats. He retained his
initiative, but in a quieter position.
1 2 ... 1Ve7 1 3. 1i'c2
More natural is 1 3 .Vi'd4, but White real
ised that 1 3 ...Vi'es is not a threat because
1 1 ... h5!? of 14.f4, whilst on d4, the queen could
First of all threatening 12 ... �xh2+ be hit by ... c6-cS.
(as Tal once observed, such a move 1 3 . . 0-0 1 4 . .td2 .l:[e8 1 5 . .l:[fe1
.
barely deserves an exclamation mark) g6 1 6 . .l:[ad1 ..tf5 1 7 . ..td3 bd3

1 3 .\t>xh2 tbg4+ 14.\t>g1 "ifh4 1 S . �xg4 1 8. 1i'xd3 1Ve5 1 9.f4 'ife6
hxg4 or 14.\t>g3 h4+! (more pre Black stands slightly better, but it is not
cise than 14 ..."ifd6+ or 14 ..."ifgS , after very much. Eventually, the game ended
which there would follow 1 S .f4) 1 S .\t>f3 in a draw.
(1 S .\t>h3 t2Jxf2+) 1 S ..."iff6#.
White's reply is necessary: M i ron Sher
1 2.h3 Elmar Magerramov
And now the most interesting moment. Helsinki 1 99 2
With his last move, Black set out on •
an unusual path for the game and he
would like to continue it.
The first idea which deserves consid
eration is the Razuvaev-like 1 2 ... 0-0!?,
and if 1 3 . �xhS tbxhS 14."ifxhS , then
14 ... .l::t.e 8 followed by transferring the
rook along the 6th rank to g6 or h6.
Then again, it also makes sense to
look into 1 2 ... tbg4!?. The piece cannot
1 73
The situation is slightly different. Here

it is the queen which is on d6, not the
bishop. This means that the knight
jump to g4 will involve mating threats.
But for the moment, g4 is covered,
hence:
1 2 ... h5! 1 3.e4
Just as we are taught - White meets a
wing attack with a thrust in the centre.
But the enemy central structure sud
denly turns out to be unexpectedly
solid ... 1 5 ... l005 !
1 3 .h3 does not solve the problems, There is no need to play for spectac
because of 13 ... tbg4, although then after ular effects (I S ... tbxh2 1 6.'�xh2 h4
14.g3 , it is still too early to bury White: etc.) , when there is a simple path to the
14 ... tbxe3 I S Jlei or 14 ... tbe5 I S . i.g2 objective.
h4 16.g4, and if 1 6 ... i.xg4 1 7.hxg4 h3 , 1 6. tLlf4
then 1 8.i.f3 , retaining cautious opti He is not saved by 1 6 . i.g2 d3 1 7.tbf4
mism. i.g4 1 8 .'ifel h4 1 9. i.e3 hxg3 20.hxg3
1 3 ... ttJg4 1 4.g3 tbf3+ 2 I . i.xf3 , and then Black has
By analogy with the game Razuvaev a choice: 2 I ...i.xf3 22 .tbg2 'ife6!
Farago, one can recommend 14. i.xg4 23.i.xc5 'ifh3 or 2 I ...'ifh6!.
i.xg4 I S .'ifd3 or 14 ... hxg4 I S .eS! and 16 ... �g4 17 . .bg4 hxg4 1 8.h4 d3
Black is not mating, in any event. 1 9. '1ft>g2 "it'd4 20. "it'e1 g5! 0-1
1 4 ... d4 1 5. l002 A striking crush. It would be interest
A barely noticeable, but decisive mis ing to know if anybody, when looking
take. White cuts the connections through this game, had thoughts of the
between his pieces. He should have 'weakness' of the c6/d5-pawn couplet
played I S .tba4. or the 'blockade' on d4/c5?!
1 74
Chapter 18
With the M irror's Hel p

The theme of the 'mirror' is closely caught up with general chess culture, and the
breadth of erudition of this or that chess player.
Let us assume that a player has a certain favourite variation in his opening reper
toire. He studies all its subtleties, knows all the plans, and achieves good practical
results with it. Naturally, he wants to add to his repertoire the same scheme, but
with colours reversed. He is not even put off by the fact that, if he plays the set-up
as Black, he will be a tempo down.
But it more often happens that a player, say a well-known GM, has a wide general
chess culture, which allows him to orient himself in positions he only half-knows,
or barely knows at all. And what do we mean by this 'orientation' ? We mean play
by analogy, employ means that are already known in other contexts.
But let us repeat the opening moves,

Lajos Portisch but with colours reversed: l .e4 cS
Anatoly Karpov 2 .tLlf3 tt::lc6 3 .d4 cxd4 4.tt::lxd4 g6 s .tt::l c 3
Linares 1 981 i.g7 6.tLlb3 tt::lf6 7. i.e2 d6 8 . i.e3 a6
1 .c4 e5 2.tlJc3 t2Jf6 3.g3 d5 4.cxd5 9.�d2 bS 1 0.a3 .ib7 l l .f3 0-0. This
lLJxdS 5 . ..tg2 t2Jb6 6. t2Jf3 tlJc6 7 .d3 is a variation of the Dragon Sicilian -
..te7 8.a3 ..te6 9.b4 a6 1 0 . .ib2 not a typical one, not a main line, but
'it'd7 1 1 . l2Je4 f6 1 2. 'ifc2 even so, one which leaves little doubt
as to the choice of plan. The exchange
.i i�t�; 4��; * -,:?,� :;B of dark-squared bishops with .ie3-h6,
rr��1,.t r•!• * l _ -- -· castling queenside, and the advance
· ·� ;�;;P .t I ;��;�[. of the kingside pawns with g2-g4 and
��)�� - - .• ' · '
h2-h4-h5. The order could change, but
:l;A;;.:. ;.','1�'i- r:-..
\ : i t..LJ
all of these moves will be played.
.:t..J
Karpov was of the same opinion. He
�fj� [�1��':i !:::, >��j:� t2J
...
was not disturbed by the lack of a
�·��-� !:::, tempo, resulting from being Black. On
�) ' ��·, .. ,, �
:�� ;��: ;�: the board, we have a Dragon Variation,
The game has featured a variation of which means that he must play as he
the English Opening. It is Black's move, would in the Dragon Variation.
and at first sight, he has a wide choice of 1 2 ... ..th3 1 3.0-0 bg2 1 4. �xg2
plans. What should he play? Attack on g5!? 1 5.d4
the queenside or the kingside? Advance White reacts in the classical way: he
... a6-a5 or .. .f6-f5 ? Kingside castling or meets a flank attack with a central coun
queenside? There is a mass of questions, ter. Even so, his last move involves a
to which it may seem there is no defi pawn sacrifice, and as later events show,
nite answer. this sacrifice does not prove justified.
1 75
Preferable was l S .tDcS i1.xc5 16.'i!Vxc5 Another vanatwn of the English

or 16.bxc5 . Opening, and another mirror - this
1 5 ... g4 1 6. ttJh4 exd4 1 7.tbf5 0-0-0 time of the Scheveningen Variation of
Of course, not 1 7...'iYxf5 1 8 .tDd6+. the Sicilian.
1 8. l:[fd1 1 2 ... 'ife8!? 1 3 . .td2 'ifg6
This is how White plays in the Scheve
ningen. From here, the queen eyes the
g2-square attentively, and also supports
the break ... e5-e4.
1 4 . .te1 .:Sd8 1 5. ttJb5 e4!?
The cautious 1 5 ... .!::i.d 7 looks good, but
Adams strives to seize the initiative.
1 6. ttJfd4 tbxd4 1 7. tbxd4 .tea
1 8 ... tLlc4!
It is probably this move that White
underestimated in his earlier calcula
tions. It is hard to believe that Black will
retain his material advantage in all the
possible complications, but it is defi
nitely the case: 1 9.tDxd4 tDxb2; 1 9. i1.xd4
'ifxfS 20.tDd6+ tDxd6; 1 9.'iYxc4 'ifxfS .
The game continuation is no better: A culminating moment in the game.
1 9. tbxe7+ 'ifxe7 20. tLlc5 It seems as though White can take the
20.'i!Vxc4 'ifxe4+. pawn: 1 8 . �xa5 b6 1 9.i1.el - no imme
20 ... tbxb2 21 . 'ifxb2 h5 diate threats are to be seen. But they can
The position has stabilised, with Black soon appear: 19 ... i1.b7 20.dxe4 tDf4!?. As
having an extra pawn and the attack. often happens in the Scheveningen, the
Karpov won on move 37. square g7 (usually g2) is under threat.
Possible variations are: 2 1 .exf4 �xe4;
Vladi m i r Kra m n i k 2 1 .g3 �xe4 22 .'iVxc7 tDh3+ 23.<;;t>f l
M ichael Adams �d6 24.'ifxb6 f4! 2 1 . �f3 tDxg2! (but
Las Vegas 1 999 not 2 l ...fxe4 22 .exf4 exf3 23 .'iYxg6
•
hxg6 24.tDe6) 22.exf5 (22. �xg2 �xe4)
22...'i!Vg5 23 .tDe6 tDxel+ 24.tDxg5
tDxc2 , and it is in any event not Black
who needs to worry about maintaining
the balance.
1 8.b4 axb4 1 9.axb4 c6 20 . .tf1 .tf6
21 .dxe4 fxe4
New lines have been opened for the
attack - the f-file and the c8-h3 diago
nal. Black went on to win.
1 76
Chapter 1 8 - With the Mirror's Help
Let us see how the plan of 'ftd1-e1 -g3 For a very small material investment,
and e4-e5 works in a classical Scheve Ljubojevic has developed a dangerous
ningen. initiative. He threatens to move the
knight from d4, with attacks on d6, e5
and f6. All the white pieces occupy their
Lj ubomir Lju bojevic best positions, with the exception of his
U lf Andersson king, which should be on h 1 , depriv
Wijk aan Zee 1 976 ing Black of a potential tempo along the
g1 -a7 diagonal.
Andersson defends in his own style, pas
0
sively but very cleverly. He makes four
retreating moves in a row, and each of
these rather surprisingly relieves White
of part of his initiative. After the fourth
of these moves, the attack is completely
extinguished, but then there follows
an immediate oversight, letting White
back into the game. Ljubojevic breathes
again, the attack resumes with new
1 0. 'ife1 !? 0-0 1 1 . 'ifg3 �d7 1 2.e5!? force and White wins.
A real pawn sacrifice.
«One's first thought on playing over the game is The game has been subjected to detailed
that, if everything is correct, then Black can no analyses by Timman, Dvoretsky and
longer play the Sicilian Defence. But chess has yet to other analysts. It was the centre of atten
reach that point in its development» Timman.
- tion for many years. Timman called it
1 2 ... dxe5 'the most interesting game of the last 20
Nobody has yet shown a mathematical years'. With the development of com
win, if Black declines the sacrifice and puter programs, we can now examine
plays 1 2 . ..tbe8 . this highly interesting game anew, and
Thus, in Geller-Halme (Helsinki 1 992) the range of defensive resources for
there followed: 1 3 Jhd1 (13 .tLle4 Black is significantly expanded.
deserves consideration) 13 ... tLlxd4 In general, the scope for further analy
14 . .ixd4 .ic6 (the correct reaction: sis is still immense.
now White cannot bring his knight
quickly to the kingside) 1 5 .f5 dxe5
16 . .ixe5 'iYb6+ 1 7.�h l . If Black retains
a siege mentality, then after 1 7... exf5
1 8 . !lxf5 !td8 1 9. �dfl f6 , he can suc
cessfully defend, for instance: 20 . .ic4+
�h8 2 1 . lth5 fxe5 22.l:!.xh7+ 'it'xh7,
and White has nothing better than per
petual check: 23 .'iVh3+ 'it'g6 24.'fig4+
�h6 25 .'fth3+.
1 3.fxe5 tbxe5 1 4. �f4 �d6 1 5 . .1:ad1
1 77
For the time being, Black must find Black successfully defends by means
a way to deal with the retreat of the of 16 ... tt::lf3+! (even stronger than
knight from d4. Weak is 1 5 .. .'�Jd5? 16 ... .i.xc6 1 7.llxd6 tt::lg6) 1 7.l::txf3 .i.xf4
1 6.lLlfS ! . The attempt to solve the prob 1 8 . .a.xf4 .i.xc6.
lems by tactical means with 1S ... tt::lf3+? The move 1 S ... .a.ae8!? has also not been
1 6Jhf3 eS does not work because of refuted, so, including Andersson's
1 7. �h6 tt::lh s 1 8 .lLldS ! (on 1 7... tt::lg 4, the move, we have in total four plausible
same move would follow) winning in defences: 1 S ... .a.ad8 , 1 S ... .a.fd8 , 1 S ... .a.fe8
all variations. For example, 18 ... tt::lxg3 and 1S ... .a.ae8. If you disagree, then
1 9 . .a.xg3 exd4 20.tt::lf6+ �h8 2 l . .i.xg7#. refute them!
1 5 'ifb8!?
... 1 6 .l:[d3
.
The idea of this retreat is to free c7 for Andersson's idea is revealed in the
the bishop. It has been considered that variation 1 6 .tt::lb 3 �c7! . In the game
this is the only defence, but this is not Arizmendi Martinez-Golod (Ubeda
the case. 2001) there followed 1 7.tt::lc 5 l::tc 8 (the
Even in the pre-computer age, the game deflection idea 1 7... .i.bS! ? 1 8 . .i.xb5
Neurohr-Darga (Germany Bundesliga tt::lh S! was also worth considering)
B 1 994/95) was played, in which after 1 8 .tt::l 3 e4 tt::lxe4 1 9.tt::lxe4 tt::lg 6 20 . .i.xc7
1 S ... .l::tad8 !? 1 6 .tt::lb 3 tt::ld S! 1 7.tt::lxd5 exdS Ihc7. Black gradually neutralised his
1 8 . .l::f.xd5 f6, Black had such a power opponent's initiative and won.
ful outpost on eS that he could look to
the future with confidence. Instead of
1 6 .tt::lb 3, the line 16.tt::lfs exfS 1 7. .l::f.xd6
'ifxd6 1 8 . .i.xe5 does not work, because
of 18 ...'ife6 1 9.tt::ld s tt::lg 4. Of course,
White can just play 1 6.�h 1 , but this is a
long story, and there is no question of a
sudden attack.
In connection with the German GM
Klaus Darga's idea, we should also con
sider 1 S .. J:!fd8 !?.
Furthermore, the then Moscow school Ljubojevic's move sets new problems.
boy (now also a GM) Vasily Papin The threat is 1 7.�e3 winning a piece.
found a defence, beginning with the The rook cannot be taken: 1 6 ... tt::lxd3
move 1 S ... llfe8!?. Now there is no sense 1 7. .i.xd6 'i:Va7 1 8 .'i:Vxd3 etc.
in 1 6 .tt::lfs exfS . The rook on e8 also 1 6 0.e8!?
...
turns out to be useful in the variation This retreat was also declared by the
1 6 .tt::lb 3 tt::ld S 1 7.tt::lxd5 exdS 1 8 . .a.xd5 analysts to be the only defence. On
.i.c6. Tactics also fail: 1 7. .a.xd5 (instead 1 7. .a.e3 there now follows 1 7... tt::lg 6, and
of 1 7.tt::lxd5) 1 7... exd5 1 8 .tt::lxd5 'iYxc2 the bishop on d6 is not hanging.
1 9.tt::lf6+ �h8 . Finally, in the variation Even so, Black had an alternative:
1 6 .tt::lc 6 (hoping for 16 ...'iYxc6 1 7. .a.xd6! 1 6 ... l::r.c 8 ! ? 1 7.l:re3 tt::lc4. By means of
'iYxd6 18 . .i.xe5 'iYb6+ 1 9.�h1 �f8 1 8 . .i.xd6 'ifxd6 1 9.tt::lxe6, White could
20 . .a.xf6! gxf6 2 l . .i.xf6, mating), regain the pawn, but after 1 9...'iYxg3
1 78
Chapter 18 - With the Mirror's Help
20 . .l:!.xg3 �xe6 2 l . �xc4 .l::txc4 22 . .l::!xf6 23 .'il'e3 'il'xb2! leaves the position bal
bS 23.a3 Ild8 , he is condemned to a anced.
long battle for a draw, in an inferior 1 B . .:r.c3 tlJc6!
endgame. There is also little pleasure Rather stronger than 18 .. .f6 1 9.tt:lc5 and
in 1 8 . �xc4 �xf4 1 9. .!::i:xf4. Black is not 1 8 ... �c6 1 9.tt:lxc6 (Timman gives many
obliged to go into the variation with beautiful variations, in recommending
perpetual check with 1 9 ... .l:!.xc4 20.tt:lfs 1 9.tt:lg5 !? h6 20.tt:lgxe6 fxe6 2 I .tt:lxe6
tt:lhs 2 I .tt:lh6+ �h8 22 .tt:lxf7+ '1t>g8 .l::tx f4 22 . .l:ixf4, but misses the strongest
23 .tt:lh6+, as the immediate 1 9 ... tt:lh5 ! is defence 22...�d6) 1 9 ... tt:lxc6 20. �xc7
stronger. 'il'xc7 2 I .tt:lf6+ �h8 22 .'il'xc7 tt:lxc7
1 7. tbe4 23 .tt:ld7 �fd8 (or 23 ... tt:lds 24 . .l:!.g3
tt:ld4 25.�d3 , Delanoy-Lechtynsky,
Kecskemet 1 989) 24.tt:lb6 l:tab8 25 Jhf7,
with a large advantage in the ending.
With the move 1 8 ... tt:lc6, Andersson
completes his colossal defensive work.
Nothing remains of the white attack
and he needs to concern himself with
maintaining equality.
1 9 . ..bc7
1 7 ... ..tc7! ?
Continuing the established pattern of
clever retreats.
Taking the rook looks terrifying,
although in analysis, there is no
clear refutation. Sample variations
run 1 7... tt:lxd3 1 8 . �xd6 (definitely
not 1 8 .tt:lf6+ tt:lxf6 1 9. �xd6 tt:le4,
Pinkas-Liebert, Decin 1977) 1 8 ...'il'a7!
(of course, not 1 8 ... tt:lxd6 1 9.tt:lf6+ 1 9 ... ttJxd4!
and 20.tt:lxd7) 1 9.c3 tt:lxd6 (dangerous Avoiding the last trap: 1 9... tt:lxc7 20.tt:lf6+
is 1 9... tt:lxb2 20. �xf8 �xf8 2 I .tt:lg5) '1t>h8 2 I .ifxg7+!, mating: 2 1 ...'1t>xg7
20.tt:lf6+ �h8 2 I .tt:lxd7 tt:lcs 22 .tt:lxf8. 22.f!g3+ �h6 23 . .l:r.f4 tt:lxd4 24.�h4#
White has an extra piece, but the knight or 22 .. .'iit>h8 23 J:th3 h6 24 . .!::i:x h6+ �g7
on f8 is cut off, and Black only needs 2S . .l:!.h7+ <;t>g6 26.�d3+ '1t>g5 27.h4#.
to find an accurate sequence of moves. Also bad is 19 ...'il'xc7 20.tt:lf6+ '1t>h8
After 22 ... tt:lce4 23 .'il'h4 .!::i:xf8 24. �d3 2 1 .'il'xc7 tt:lxc7 22 .tt:lxd7 tt:lxd4 2 3 . �d3
eS 2 5 . �xe4 tt:lxe4 26 . .!::i:xf7! White (probably sufficient for victory was
wins. Nor can Black be satisfied with 2 3 Jhc7 tt:lxe2+ 24.�f2 , but naturally,
22 ... 'il'b6 23 .'il'e3 .!::i:xf8 24.tt:lxe6! l:te8 White is very reluctant to allow the
25 .'il'xc5 �xeS+ 26.tt:lxc5 .!::i:xe2 27. l:td1 enemy rook onto the second rank:
tt:le8 28.Ild8. However, 22 ...'il'b6 23 .. J.tfc8 2S . .!::i:x c8+ .!::i:xc8 26.�xe2
1 79
l::f.xc2+ etc.) 23 ... �fc8 24.tLlb6 or We don't know what Ljubojevic had
23 ... tL:lcb5 24.tLlxf8 tLlxc3 25 .bxc3. planned here, but later it was found
However, the super-subtle 19...'ifa7!? that White can force a beautiful draw:
was possible, although Black has no 22..txh7+! �xh7 23 .'ifxg7+! �xg7
need of such subtlety. 24.1:1g3+ with perpetual check. White
20 . .i.d3 'ifa7 21 .tlJc5 has no choice at all: on 22 .'ifxc7 there
It was also possible to close the dan is the unpleasant 22 ...tLlb5 !.
gerous diagonal in a different way: Andersson played
2 1 .'iff2 . If Black relaxes with 2 1 ...tLlxc7, 21 ... i..b5
then he loses at once: 22.tLlf6+! (this and kept defending resourcefully. The
was the last trap) 22 ... gxf6 (22 ... �h8 game demanded enormous creativity
23 .tLlxd7) 2 3 . �xh7+! �xh7 24.'ifh4+ from both players. At move 24 Ljubo
�g7 25J:tg3#. However, any sensible jevic committed an inaccuracy, after
reply, such as 2 1 ...i.c6, 2 L. ..tb5 or 2 1 ... which Andersson even got the better
f5 , leave an extra pawn and a defensible of it. However, at move 30 the Swede
position. missed a great chance in a still compli
Ljubojevic goes for clarity. He probably cated position, and lost beautifully.
expected Andersson to take the bishop: The game created a strong impression.
2 1 ...tLlxc7. Both the attack and the defence were
rich with ideas.
In our day, the plan with an imme

diate 1 2 .e5!? is rarely seen. When it is
played, it is usually to exploit the beau
tiful drawing variation, shown in the
last diagram. Alas, such is the level of
knowledge nowadays. Once untrodden
paths turn into main roads, and end in
the gridlocked traffic of a draw.
But the idea of mirror-like schemes
analysis diagram
continues to be pushed forward.
1 80
Chapter 19
Fischer's Hedgehog and Morphy's Needles

An original plan in the Hedgehog structure is associated with Fischer's name. The
outline of the plan is well known, but the pre-history is certainly not.
But we will not jump forward.
But for the time being, the black posi
Robert Fischer tion is very solid. White needs to trans
U lf Andersson fer his knight to fS (sacrificing a piece
Siegen 1 970 in the process) to break down the
defences.
E ti' E 4 * 1 8 ... tbb6 1 9. tbc5 �c8 20. tbh4 tbd7
0
l &:' &" l 21 . ttJe4 tbf8 22.tbf5! �e6
� .t .t. Bad is 22 ... gxf5 23 .gxf5 .
23. tbc5 ttJe7 24. tbxg7 <iii>x g7 25.g5!
& �·
By exchanging the bishop on g7, White
.
opens the long diagonal a1 -h8 and

8 £::, . •. t::, ls t2J creates irresistible threats along it .
. �'iV,tD � � £3:, 8;
g . . :a: ·� In the press, this became known as the
Fischer Plan. It is worth noting in pass
1 3. <ifo>h1 ! ? 'ifd7 1 4 . .1:.g1 !? .l:.ad8 ing that this is a mirror plan, in that it
1 5. ttJe4 'iff7 1 6.g4! g6 1 7 . .1:.g3 �g7 can be used by Black as well.
1 8 . .1:.ag1
Mark Ta imanov
Artur Jussu pow
Moscow 1 982
The plan is in operation. Instead of just

hopping from one foot to the other, as
often happens in Hedgehog positions,
White gathers his forces purposefully 1 8 ... <iii>h 8!? 1 9 . .1:.c2 .l:.g8! ? 20 . .1:.cd2
on the kingside. The way he does so g5! 21 . ..td4 .l:.g6 22. tbc1
does not at first strike one as obvious, According to Mark Dvoretsky, both
and is a tribute to his ingenuity. sides exchange inaccuracies. White
181
should freeze the enemy assault with happen, when players of a lower class
22 .g4, and Black in his turn misses the find themselves faced with a situation
chance to play 2 l ...h5. that is on the edge of their understand
22 ... l:l.cg8 23. tbd3 'iff8 24. l:l.e1 g4! ing, with an unusual set-up or a plan
Etc. «Is it easy to find a plan like .. st>g8-h8, they have not seen before. There comes
... l:te8-g8, ...g7-g5 at the board? No, of course a desire to clarify the situation and take
not. But in this case, there was no need. Jussupow play into forcing channels. And this
knew the game Fischer-Andersson very well, in often ends up involving an oversight.
which the plan was first played» Dvoretsky.
- 1 9.liJxe6 fxe6 20 . ..txe6
But no, it was not the first. The combination is in itself dubious,
but after Black's reply, it leads to imme
Julio Garcia Soruco diate defeat.
Robert Fischer 20 ... lbxe4! 21 . ltJxe4 .:Xe6 0-1
Havana 1 966
It seems the source game has been
• found. Can we consider our small his
torical escapade finished? Not quite.
In 2004, I worked as trainer of the Ira
nian team. I read lectures, including
one on the Fischer Hedgehog. The lec
ture had been prepared some time ago,
and went fine, with nobody expect
ing any surprises. Then suddenly, one
of the listeners, Abulgasseem Nadjib
(an international arbiter and children's
1 4 ... '1th8! ? 1 5. tt:lce2 l:l.g8!? trainer) cheerfully showed us an even
An interesting picture: Black, as it were, earlier source. It turned out we needed
is wearing blinkers, and refuses to to go back more than 1 5 0 years.
be distracted by tempting side-roads.
Instead of 14 ... '.t>h8, the move 14 . ..tbc4 Louis Paulsen
looks more natural, while after the Pau l Morpy
knight has left c3 , 1 5 ... d5 looks excel New York 1 857
lent. But Fischer pursues his idea!
1 6.�h1 gS! 1 7.h3 l:l.g6 1 8.tlJg3 l:l.ag8 •
The set-up is complete, but the battle

is far from over. A sensible continua
tion for White is 1 9.tt.Jde2 h5 20.tt.Jc3 ,
and if 20 ... g4, then 2 1 .hxg4 hxg4 22.f4.
Probably, Black should not rush things.
He can set up a battery on the long
diagonal: after 20 ... �c6 2 1 .�d4 'i¥b7,
the whole battle is ahead.
Instead, White managed to find a way 1 5 ... �h8!? 1 6.tbd1 gS! 1 7. lLlf2 l:l.g8
of losing in two moves. These things 1 8.tbd3 g4!
1 82
Chapter 1 9 - Fischer's Hedgehog and Morphy's Needles
The real thing, eh? ... More precise is 26 . .!:i.xf7 gxh3+ 27.�hl
1 9.lbxe5 dxeS 20.hxg4 .bg4 21 . 'ii'f2 �g2 28 . .tf4! and if 28 ...'iVc6 , then
l:[g6 22. 'ii'xf7 �e6 23. 'ifxc7 .:Xg2+! 29 . .te4, whilst if 28 ... e5, then 29 . .te2 ! .
24. 'iPxg2 'ii'h3+ 25. �2 'ifh2+ 2 6... .:Xg4+ 27. �1 fS 28. 'iff2 tbeS
Faced with mate in 3 , White resigned. 29.dxe5
Morphy has played the game superbly
«< played through several hundred Morphy games up to this moment. But now he com
and was amazed at his inventiveness» Fischer.
- mits a lapse and has to begin the real
The following example is even more isation of his advantage all over again:
surprising. In this, Morphy and his 29 ... 'ifxd3+ 30. 'ife2 'ife4
opponent appear as if in a mirror - The etc. After 29 . .tf4 tLlxd3 30.�xd6 tLlxf2
'Fischer Plan' - yes, in the middle of the 3 1 . -teS+ and 32 .�xf2 , the game would
1 9th century. have ended at once.
Pa u l Morphy Paul Morphy

Adolf Anderssen Alexander M eek
Paris 1 858 New York 1 857
0 0
1 6. �h2 �h8 1 7. l:tg1 l:[g8 1 8.g4 gS In this structure, where the white pawn
Of course, Anderssen is not copying wedge d4/e5/f4 faces a black wedge
his opponent's moves just by way of a d5/e6/f5, Morphy often conducted the
parody. He wants to set up a barrier to attack with the aid of the break g2-g4.
the pawn advance, but in vain. This game is proof of that.
1 9.f4 'iff8 20. l:[g3 l:[d8 1 O.h3! tbd7 1 1 . �h2 cs 1 2.c3 c4
More tenacious is 20 ...'iVd6. 1 3. �c2 a6 1 4.tt:lf3 h6 1 5.g4 �h7
21 . tt:lf3 .hf3 22 . .:Xf3 'ifd6 23. �g2 1 6. l:[g1 l:[g8 1 7.'ife1
Black's position is difficult. He decides White has carried out the attack strictly
on a piece sacrifice, but the combina in accordance with his plan and stands
tion cannot bring success. very threateningly. But his attacking
23 ... tt:lh5 arsenal is limited. Probably, one should
Counting on 24.gxh5 gxf4+, although expect a sacrifice on g6 or f5 .
here too, White has the advantage: This makes Black's next move all the
25 .�hl fxe3 26.l:f.xf7 l:f.g3 27. -tfl tLlf8 more surprising, as he voluntarily
28 .'iYe4. removes the main defender of the weak
24.fxg5 hxgS 25.gxh5 g4 26.hxg4 squares, with his own hand.
1 83
1 7 ... ttJc6 1 8. tt::'lh 4

It is all over. On 18 ... tt::l e 7, there is
1 9.tt::lxg6! tt::lxg6 20.gxf5 exfS 2 l . �xf5 .
O n 1 8 . . .tt::lf8 , this combination i s not
so clear: 1 9.tt::lxg6 tt::lxg6 20.gxf5 exfS
2 l . �xf5 tt::le 7 22.�c2 �c8 23.tt::lg3 l::tf8 ,
and Black has slowed up the advance
f4-f5 , at least for some time. But all
questions are resolved by 1 9.gxf5 gxfS
20.tt::lxf5 exfS 2 l .�xf5+ 'iit>h8 22.�c2,
followed by the advance of the f-pawn. White has achieved a great deal. He has
The text move also fails to save Black: opened the g-file, and doubled rooks on
1 8 ... 'iff8 1 9. tbxg6! 'iftxg6 20.gxf5+ it. But he cannot succeed without get
�7 21 .fxe6+ 'iftxe6 22.f5+ ting his queen into the action.
The rest is not worth seeing. Black soon The strongest piece belongs on h4. That
resigned. means that White's plan should involve
the exchange of bishops.
Peter Leko 23. 'ife1 ! 'ilea 24 . .tf2! tba7
Wolfram von Alvensleben Either very brave or just careless. Admit
Budapest 1 990 tedly, 24 .. .'�'f7 2 5 . �h4 �g8 does not
enable him to defend either: 26.tt::lg5
'iVf8 27.tt::lx h7! l:Ixg2+ 28 . .l:f.xg2 �xh7
29. �xe7 'iVxe7 30.'iVg3 , with mate in a
few moves. But 24 ... 'iVhs 2S.tt::lgs 'iVh6
still permits him to defend stubbornly.
25 . .th4 .tbS
Here, a quicker way to the aim was
26.l::tg7 .l::f.f7 27. �f6. But even after
26. 'ifg3
White was able to conduct the game to
The structure is identical to that shown victory without difficulty:
in the previous example. White has two 26 ... :f7 27. be7 :Xe7 28. 'ifh4
possible breaks: c3 -c4 or g2-g4. But on 'iff7 29. tbg5 'Wf8 30. -thS tbd7
1 2 .c4, he must reckon with 1 2 ... cxd4 31 . .ig6 'ifg8 32. tt::'lf7+ :Xf7
1 3 .tt::lxd4 �cS 14.l:Idl gS !?, and the 33. bf7 1i'xf7 34. 'it'h6 1 -0
central pawn triangle d4-e5-f4 comes Eastern books have a rather rich style
under attack from all sides. for Europeans, but I will allow myself
Grandmaster Leko prefers to follow in to quote from one in conclusion: «The
Morphy's footsteps: great sages give their lessons to posterity, from the
1 2.h3! a6 1 3. 'ifth2 'ifd8 1 4.:g1 'ifth8 depths of the grave. They themselves may have
1 5.g4 g6 1 6. tba3 cxd4 1 7.cxd4 b5 passed on, but their creations do not decay, and
1 8. ttJc2 tt::'lb6 1 9.b3 .td7 20. :g2 b4 continue to assist teachers throughout the centuries
21 .:Sg1 aS 22.gxf5 gxf5 to come.»
1 84
Chapter 20
March of the Tank Columns

«When you play against Botvinnik, it seems as though you have a tank facing you)) - Keres once
joked, gloomily, in giving his due to the sixth World Champion's positional mas
tery.
Botvinnik studied chess wholly. He had several favourite pawn structures, which
he knew to perfection. He knew what to do in these structures in the opening,
middle game and endgame. He knew his own plans and those of his opponent. He
played the entire game with his eyes open. There were no surprises or mysteries for
him in the positions he knew. His opponents were less well prepared and they had
to seek the truth (or, rather, suffer for it) at the board, while Mikhail Moiseevich
quietly and calmly punished their mistakes.
One of his favourite pawn structures was the so-called Maroczy Bind, with pa
wns on c4 and e4 against a pawn on d6. The plan which Botvinnik worked out in
such positions looked like this:
- freeze the central pawn structure, by preventing the breaks ... b7-b5 and ... d6-d5;
- put a knight on dS;
- by positional pressure, force Black to capture this knight;
- after the exchange, take with the e4-pawn, thereby opening the e-file for his
major pieces;
- double his major pieces on the e-file, pressing on the backward e7-pawn;
- force new weaknesses in the enemy position and at the necessary moment,
quickly shift the major pieces to the kingside or queenside, where they can decide
the outcome of the game.
This plan brought Botvinnik unarguable success in many games. They are almost
all well-known: Botvinnik-Lisitsin (Leningrad 1 930), Kirillov-Botvinnik (Moscow
1 93 1) , Botvinnik-Lilienthal (Moscow 1 936) and many others.
Dozens of articles have been written on this theme, and the above examples have
appeared in many of them. There is no need to repeat them again here, and instead,
we will show a game which is less well-known.
1 85
M i khail Botvi nnik 1 3.exd5 lillc:d 4 1 4 . .bd4 b5

Khrisogon Kholodkevich 1 5 . .bg7 <iPxg7 1 6. 'ifd4+ <iPg8
Moscow 1 927
1 7.cxb5!
In the opening, Black sent his knight This is what we spoke about in the pre
on a long-winded march via b8-d7- amble to this game. White has three
e5 -c6 and consequently reached a well possibilities: take on bS , advance the
known theoretical position, two tempi pawn with 1 7.c5 or allow Black to take
down. He should by now have managed on c4 after 1 7. J:';lfel bxc4 1 8 .bxc4. Which
either to exchange on d4 and transfer possibility should he prefer?
his bishop to c6, or set up a pawn row If one is faced with this problems at the
on the sixth rank with ... a7-a6, ... e7-e6. board, one could fail to find the right
But he has not managed either. answer. It is a lottery. The greatness of
1 2. tal5 Botvinnik was that he always, or almost
Played without hesitation. From a more always, solved such problems in his
average player, one could expect the home analysis, and in the game itself,
'deeper' manoeuvre tt:Jd4-c2-e3-d5 just reproduced his prepared answer.
or completion of development with He studied structures and thought
1 2 .'ttd 2, l:tal -dl and �fl -el . in terms of schemes. His plans were
1 2 ... lillc:d 5 worked out not for 2-3 moves, but for
One is amazed by the submissiveness the whole game. In this respect, he was
of many of Botvinnik's opponents. ahead of his time.
As if under hypnosis, they follow out 1 7 ... .bb5 1 8. l:[fe1 'ifd7 1 9. l:[e3 f5
precisely the scenario which he had The following note is characteristic of
envisaged, and they only take active the sixth World Champion:
measures when it is already too late. «Black weakens his position, so as to defend the
It was worth considering the pawn sac e7-pawn from f7, although he could have done this
rifice 12 ... e6!? 1 3 .tt:Jxc6 i.xc6 I4.tt:Jxf6+ equally well from e8, and also by 19 . a5 and from
..
i.xf6 I S . i.xf6 'ifxf6 1 6.Vi'xd6 l:f.fd8 . If the square c7. Having missed the chance to play
1 7.e5, then 1 7...'iffs , and the pawn will ... a7-a5, Black gets a passive position. White's
most likely be regained, whilst after queenside pawns start to advance, which gives him
1 7.'i¥cs 'i¥c3 or 1 7.Vi'b4 'i¥b2, Black a decisive advantage.>>
gets counterplay. In any case, the game Not a single concrete variation! The
would take on a very different course. whole discussion is just about favour-
1 86
Chapter 2 0 - March of the Tank Columns
able and unfavourable changes to the 28 .'ifd4! 1txe6 29.�h8#. But Botvinnik
pawn structure. This is how he thought. was thinking differently. He is not wor
And one more thing. The game against ried by the black rook coming to cS,
Kholodkevich, if one plays it through because the exchange of this rook is
quickly on the board, looks entirely part of his plans anyway:
one-sided. Even so, Botvinnik did not 25 ... :cs 26. :c3 :Xc3
miss the opportunity to draw some les Not 26 ... .l::f.x a5 27 . .l::r.c 7, and nor is any
sons from it and later to criticise him thing changed by 26 ...'ifc8 27.�ec l .
self for an inaccuracy. Instead of 19 . .l::f.e 3 , 27. 'ifxc3 :ta 28. 'ife3 :ea 29.:c1
stronger was 1 9.a4 1ta6 20.b4, first :ca 30.:Xc8+ hc8 31 . 'ife6+!
establishing the favourable queenside 'ifxe6 32.dxe6 �g7 33.b7 ..bb7
pawn structure, and only then turning 34 . ..bb7
his attention to doubling rooks on the And after a few more moves, Black
e-file. Black, in his turn, could also have resigned.
played more strongly: 1 8 ... a5! 1 9 . .l::f.e 3 a4
20.b4 l::tc 8 with some counterplay. But Botvinnik had his own teacher. To
20.a4 i.a6 21 . :Se1 :t7 22.b4 �b7 Mikhail Moiseevich's honour, he never
23.b5 tried to claim the credit for others' suc
cesses. Thus, in this instance, he openly
admitted that he had learnt this plan
from a game of Rubinstein.
N i kolay Zubarev
Akiba Rubi nstein
Moscow 1 925
White's plan has been fully realised and

Black's position is close to being hope
less.
23 ... a6 24.b6 :ca 25.a5
This decision is also characteristic of the
pragmatist Botvinnik. When playing
positional chess, he practically never
allowed himself to be distracted by 26 ... e5! ?
tactical nuances, even when they were Not being distracted by the gain of a
objectively in his favour. pawn: 26 ... t2Jxf3+ 27. 1txf3 c4, and
I think few players would have been probably intending his next move.
able to resist the chance to finish the 27. tbxd4 exd4!
game off with a beautiful combination: Only thus. Rubinstein takes not with
25 .'iYh4 �f8 26.a5 .l::f.c s 27 . .l::f.e 6! .l::f.x aS the queen or the c-pawn, but with the
28Jhg6! hxg6 29.�h8# or 27 ... 1txd5 e-pawn.
1 87
2a.tat1 .:ea 29.14 probably have continued 38 ... .l:f.e2

The same mistake as Kholodkevich 39 . .l:f.fl fS followed by the advance of
repeated two years later. More tena the kingside pawns, which should lead
cious was 29.e4, although, as Razuvaev to a win anyway.
points out, «White is not to be envied in this 38 ... .:e2 39. 'iff1 c4! 40.b4
case either». After, for example, 29 ... dxe3 40.dxc4 d3 .
30.lt:Jxe3 'iVd4 3 l..l:f.c 3 .i:f.d8 (but not 40 ... cxd3 41 . .:tb3 .:Xf2!
3 I ....i:f.d7 3L�c6) or 30 .. .f5!? there is White resigned without waiting for
definitely nothing to envy, indeed. 42.�xf2 �dl+ and 43 ...�xb3.
29 ... 'ife6 30 . .tf3 l:ae7
And now we have the well-known dou Esteban Canal
bling of rooks on the e-file. Akiba Rubinstein
31 . tbf2 Carlsbad 1 929
31 ... .td5! 1 9.e3!?

Rubinstein is not disturbed by the A highly interesting situation. White
prospect of losing the advantage of the either knew, or maybe guessed,
bishop pair. It is much more important the threat posed by the Rubinstein
to remove the main defender of the Botvinnik plan. He sensed that Black
e2-pawn. planned to post his knight on d4, take
32. bd5 'ifxd5 33 . .:e1 'it'h5 back with the eS -pawn and build up on
34. 'it'd1 the e-file, etc.
There is nothing better: 3H�i'fl �xh2 , Decades later, theory has followed
or 34.lt:Je4 fS . the path taken intuitively by Canal. A
34 ... .:Xe2 35 . .:Xe2 'ifxe2 36. 'ifxa4 new opening scheme grew into being,
.:es known as the Hedgehog, where the
In a deliberate attempt to show his pawns are arranged along the sixth (or
whole plan to the end, Black overdoes here, the third) rank. Theory became
it a bit. Simpler was 36 ...�xb2 37.�xe8 deeper and wider, and it was shown
�xal +. that the threatening-looking Maroczy
37 . .:tb1 'iff3 38. 'ifd1 Bind could in reality be attacked and
Black's task would be complicated by even completely broken down. The pro
3 8 .�b3 (or 3 8 .�b5) , blocking the ponents of the Maroczy needed to find
breaks ... c5-c4 and ... d4-d3 . Black would new plans.
1 88
More of that later. But for now, we But now Black could play 23 ... .tb3 ! ,
can say that the theory of this pawn forcing the rook to occupy the square
structure developed on the basis of the d2, from where it would have no other
Rubinstein-Botvinnik plan. But they moves. Then Black goes 24 ... tLlfS ,
were not the first to come up with the forcing the exchange of bishop for
plan. The first explorer here was some knight, and then he can quietly double
one else. his major pieces on the e-file and pre
pare the breaks ... b6-b5 and ... c5-c4.
However, Chigorin chose ...
Jacques M ieses 23 ... ..td5 24.f5 ttJcs 25. 'iVf4 :eo
M i khail Chigorin 26. tbf3 .be4 27.dxe4 'ife7 28. tLld2
Barmen 1 905 The e-file is closed and the whole game
has taken on a chaotic appearance .
•
Another favourite Botvinnik pawn

structure was one which arises often in
the Queen's Gambit, Ragozin and, espe
cially, the Nimzo-Indian. It arises as
follows: Black brings his bishop out to
b4. After being hit by a2-a3 , the bishop
takes on c3 , and thus the b2-pawn trans
fers to c3 . Then there is an exchange of
pawns on dS, so that White gets rid of
21 ... tLld4! 22 . .bd4 his c4-pawn and Black of his e6-pawn.
More cunning is 22.fxe5!? and if We reach a position where Black has
22 ... tLlxc2 , then 23 .'i¥xc2 fxeS 24.tLlf3 , central pawns on cS and dS , and White
with some counterplay for the sacri on c3 , d4 and e3. At first glance, it does
ficed exchange. not look attractive for White. His pawns
22 ... exd4! do not look very mobile, as any advance
The moment of truth! (c3-c4 or e3-e4) leads only to further
23 . ..te4 exchanges. The dark-squared bishop
Forced, in view of the threat on cl is shut in on all sides by its own
... tLle7-f5-e3. pawns. But Botvinnik worked out a
plan, to which his opponents could find
no resistance.
In the ideal scenario, by strength or
cunning, White induces his opponent
to play ... c5 -c4. If this is not possible,
the plan can still be carried out with
tension between the pawns c5/d4.
Thus, White first plays f2-f3 , then
regroups his pieces (knight to g3 ,
queen and rooks to the e- and f-files),
and plays e3 -e4. If Black takes on e4,
1 89
White recaptures with the pawn from Why did GM Gurevich use this posi
f3 . The f-fi.le is opened and White gets a tion? Probably to wake the audience
powerful outpost on fS . If the exchange up. It was to be a long and complicated
on e4 does not take place, then the lecture, so probably he thought this
pawn advances to eS and is then joined nice combination would be a suitable
by its neighbour: f3 -f4-fS . warm-up for people.
The slow and methodical advance of But suddenly, it became apparent that
the e- and f-pawns can resemble a tank the youngsters were not familiar with
formation. Keres knew what he was this classic position. Gurevich was in
talking about, especially after he lost to shock. He could not continue the lec
Botvinnik at the XX USSR Champion ture and announced a break, before he
ship (Moscow 1 952) in precisely this went on. Later, over dinner, he voiced
structure. his astonishment to the President of
the Turkish Chess Federation, Ali Nihat
The most well-known example of this Yazici. Emotionally, he declared that all
plan is the game Botvinnik-Capablanca, of the country's trainers needed to be
AVRO 1 93 8 . This is possibly one of the brought together and warned to study
most oft-quoted games of all time and it the classics!
is hard to imagine any chess lover who
does not know the game. This would be a very appropriate time
I once worked as trainer to the Turkish to show an example from Gurevich's
national team. One summer, in the own practice. It is interesting to see
town of Urgiip, a training session was how he managed to implement the Bot
organised for promising young players, vinnik plan (jumping ahead, we can say
to which the well-known grandmaster that he did so faultlessly).
Mikhail Gurevich was invited. At the
start of the session, Mikhail put the fol M i khail Gurevich
lowing position on the demonstration C. Santos Perei ra
board: Breda 1 998
0
0
The end of the 'game for all time'

Botvinnik-Capablanca. White won by 1 2.f3!? :ea 1 3. tt::lg 3 "it'd7 1 4. "it'd3
30 . .b3! "it'xa3 31 . tLlh5+! gxh5 'it'a4 1 5 . ..tb2
32. "it'g5+ �8 33. "it'xf6+ �g8 34.e7 There is no need to hurry with the
and his king escapes the checks. advance e3 -e4. After 1 S .e4 cxd4
1 90
Chapter 20 - March of the Tank Columns
1 6.cxd4 dxe4 1 7.fxe4 .l:f.ad8 , the centre 20.'iVxa7 l::i.xe3 2 I .'ifxb6 ..a'.xe2 22.ti:Jxe2
could come under attack: 1 8 .ti:Jfs tt:Jxe4, 'ii'x a3 , with a draw.
1 8 . i.b2 ti:JcS . It is tempting to sacri 1 6 ... g6
fice the exchange with 1 8 . .!::i.x f6!? but White has not yet begun in earnest,
as practice has shown, White's activity yet Black already faces a crisis of ideas,
suffices for only a draw: 1 8 ... �xd4! which is the usual scenario in this
(stronger than 1 8 ... gxf6 1 9.ti:JfS) 1 9.'ii'e2 structure. If 16 ... cxd4 1 7.cxd4 'ifc4,
gxf6 20.'i¥g4+ �h8 (20 ... �f8!? and then 1 8 .'ii'd 2 , and again it is not easy to
White does not succeed in giving mate: suggest the next move.
2 I .ti:Jfs 'ifdl+) 2 I . i.b2 .a'.g8 (of course, 1 7 . .1:le1 cxd4 1 8.cxd4 ttJc7 1 9.e4
not 2 I ...ti:Jcs 22.ti:Jfs I:l.g8 23 .'i¥f3 with tbb5 20.e5
irresistible threats on the long diagonal)
22.'ii'f3 �g6 (P.Littlewood-Robatsch,
Borovo 1980).
1 5 ... .:aca
Routinely played. Black hopes after
the exchange on d4 to come in on c2 ,
but White can easily cover this entry
square, while the rook does not find
useful employment on the c-file.
Better is I S .. Jlad8 , trying to take play
back into the variation given in the
previous note: 1 6 .e4 dxe4 1 7.fxe4 cxd4 The Botvinnik plan in action. One tank
1 8 .cxd4 ti:JcS etc. White would proba has already reached eS, its tracks squash
bly have replied 1 6 . I:i.ael 'ii' b 3 1 7.'ii'd2 , ing all life around it. Soon, its example
and the advance e3 -e4 is again on the will be followed by another: f3 -f4-f5 .
agenda. Not wishing to die a quiet death in his
bed, Black initiates complications, but
this only accelerates his demise.
20 ... .1:lc4 21 . tbf1 tbxd4 22.tbd2
22 . l:Id I also wins.
22 ... tiJc6 23. tbxc4 dxc4 24.'ifd2!
The last subtlety. It is clear that White
would like to set up a queen and bishop
battery, but after 24.'ifc3 ti:JdS the black
pieces come alive. White finds another
way to create threats on the long diago
nal, without allowing any counterplay:
1 6 . .1:lf2! ? 24 ... tbh5 25.e6! fxe6 26. 'ifd7
Grandmaster play. Black probably Black resigned in view of 26 ... l::i.e 7
counted only on 1 6 . �ael cxd4 1 7.cxd4 27.'ii'c 8+ �f7 28.'i¥h8.
'ii'c2 (or 1 7...'ifc4 1 8 .'ii'd 2 'ii'c2)
1 8.'i¥xa6 'iixb2 1 9.�e2 (1 9.'ifxa7 �c2) Now we jump from a modern game to
1 9...'ii'c 3 (but not 1 9 ...'ii' b 3 20.e4) another classic:
191
Svetozar Gligoric Necessary to defend the centre. The

Pal Benko immediate 16.e4 fails to 1 6 ... dxe4
Budapest 1 948 1 7.fxe4 tLlxd4!.
In his notes, Gligoric returns time and
again to the problem of developing
D
the bishop on c 1 , pointing out that
he refrained from 1 6 . �d2 because of
the manoeuvre ... tLlc6-aS -b3 , and the
bishop is exchanged. This is a ques
tionable decision. In the game Szabo
Kottnauer (Groningen 1 946) , in reply
to 1 6 . �d2!? there followed exactly
this continuation: 1 6 ... tLlaS 1 7.e4 tLlb3
1 8 Jhe1 tLlxd2 , but after 1 9.ifxd2
The situation here is slightly different ife7 20.tLlfs ife6 2 1 .ifgs g6 22 .g4!
from the previous example. Black has Black had not equalised. Things are
voluntarily played ... cS-c4, taking the even worse for him after the thematic
pressure off the centre and depriving 1 8 ... bS 1 9.eS tLld7 20.f4 aS 2 1 .tLlfS
himself of counterplay on the c-file. In (Sharyadzanov-Brumen, Pula 1 999).
return, he gets the chance of the plan of 16 ... 1Wd7 1 7 . ..tb2
a queenside pawn advance: ... tLlb8-c6, The alternative is 1 7.l::ra 2 bS 1 8 . .l::re 2 aS
... b6-b5 , ... a7-aS, ... bS-b4 etc. 1 9.e4 dxe4 20.fxe4 b4, with mutual
White's plan is clear in its general fea play.
tures: f2-f3 , tLle2-g3 and e3 -e4. The 1 7 ... l:te6
intrigue lies in how White places his A critical variation for the overall assess
bishop. It can develop to d2 or b2, or it ment is 1 7... bS 18 . .l::tae1 aS 1 9.e4 dxe4
can stay where it is, keeping the second 20.fxe4 b4.
rank clear for the rook: .l::ta 1 -a2-e2 .
1 4.f3!?
Interestingly, Gligoric himself was
unsure of the accuracy of this move,
despite his convincing victory in the
game. He suggested instead 14.�b2,
then l::ra 1-e1 and only then f2-f3 . The
passivity of the bishop on b2 is only
apparent: Black, if he wants any coun
terplay, will have to advance his b-pawn
to b4, and then, after all the exchanges
analysis diagram
and the advance e3 -e4, the bishop will
play down the long diagonal. One would like to continue the attack:
We should add that in the same struc 2 l .axb4 axb4 22.cxb4 tLlxb4 23 .tLlfs
ture, Botvinnik often chose a3-a4 and tLld3 24.ifg3 , but I could not find a
�cl -a3 . clear path to an advantage. Black has
1 4 ... l:te8 1 5. ttJg3 ttJc6 1 6.1Wf2 two defences: 24 ... g6 2S.tLlh6+ �g7
1 92
Chapter 20 - March of the Tank Columns
26 . .l::ixf6 <;!;>xf6 2 7.l:!.fl+ �g7 28 . .l:lxf7+ he prepares to put the other offside too.
'ifxf7 29/bxf7 tbxb2 and 24 ... tbhS In fairness, though, we should say that
2S .tbh6+ (2S.'ifg4!?) 2S ... �h8 (but not bringing the errant steed back towards
2S ... <;!;>f8 26.'i¥d6+!, mating) 26.tbxf7+ its colleague with 24 ... tbc6 2S .tbe3 tbe7
'ifxf7 (also possible is 26 ... �g8 , forcing was also not such a great idea: 26.a4! bS
White to give perpetual) 27.'ifxd3 cxd3 2 7.axbS 'ifxbS 28.�a3 ! . The square fS
28 . .l::ixf7 tbf6 29.dS �g8 30.l:!.b7 .l:lxe4. cannot be supported by such primitive
There is the reserve line 2 l .cxb4 axb4 means. Even so, Black's pieces should
22 .dS tbes 23 . .ihes l:!.xeS 24.axb4, be heading from the queenside to the
keeping an extra pawn, but, frankly kingside, not the other way round.
speaking, this is not quite what White 25. ttJe3 tlJb6 26.g4!
was dreaming o£ The position is clarified: White has a
1 8. l:lae1 l:lae8 1 9. l:e2 g6 large advantage.
«Both here and on the next move, the plan with 26 ...1xg4 27. tl:\xg4 l:l.18 28. tl:\16+
... b6-bS, ... al-aS, ... b5-b4 was an absolute neces <it>h8
sity» Gligoric.
- The exchange sacrifice 28 ... .l::ie xf6
20. l:l.1e1 tlJaS 21 .e4 'ifb7 22.e5 t:Dd7 29.exf6 .l:txf6 does not ease his position,
23.14 in view of 30 . .l::!.e 7.
The Botvinnik plan is almost complete. 29. 'ifg2

If White manages to play f4-fS , his posi Gligoric plays in the style of Botvinnik:
tion will be completely winning. the main thing is the plan, and it
23 ...15 should not be disrupted in the search
The best chance. If now 24.exf6, then for prettiness. White is probably win
24 .. Jhe2 2S . .l:f.xe2 l:!.xe2 26."iVxe2 tbxf6 ning after 29.fS , since in the endgame
27.'ifes <;!;>f7 28.fS 'i¥e7 and nothing after 29 ... gxfS 30.'ifxfS 'iff7 3 Ltcl
remains of White's advantage. There 'i¥g6+ 32 .'ifxg6 hxg6, White can create
fore Gligoric accepts the fact that Black a mating attack: 3 3 . �h6! .l:lf7 34.l::!.e 3
has managed to set up a blockade, and tbc6 3S . .l::ih 3 etc.
he prepares to break it down with the In analysis, it is useful to study f4-fS
new break g2-g4! . on move 27 as well. The plan is a great
24. tl:\11 b5 thing, of course, but concrete variations
Astonishing carelessness. Black already are also an important part of the battle
has one inactive knight on aS , and now in chess.
1 93
29 ... tlJc6 1 0 ... h6

More tenacious is 29 ... tLlb3 , preventing Another concession. Black does not play
the bishop coming out via c l , but in his concretely and quickly falls into a diffi
own time-trouble, Black sets a trap ... cult position.
30. 'ifh3 Black's play was stronger in the game
White for some reason does not fall for Kortchnoi-O'Kelly de Galway (Bucha
it: 30.t2Jxd5 tLlxd4! . rest 1 954): l O ... tLlas 1 1 .tLlg3 tLlb3 I 2 ..l::ra 2
30 ... tai8 hS! ?. The correct strategy! It is essential
30 .. .'iYf7 3 1 .lLlxh7!. to bother White on both flanks, and not
31 . �c1 tLJca 32.<ifth1 :as 33.f5 allow him to carry out the classical plan
Finally! After in comfort. After 1 3 .g5 tLlxcl 14.'ifxc1
33 ... gxf5 34. �h6 t2Jg8 I S .h4 tLle7 1 6 .e4 i:Yc7, an unclear
Black lost on time, but his position was position arose, with mutual chances.
already completely lost. 1 1 . �g2 ltJa5 1 2.0-0 ltJb3 1 3. l:la2 0-0
1 4. ltJg3 �d7 1 5. 'ife1 .l:e8 1 6.e4!?
Ga rry Kaspa rov Kasparov is in his element. Without
Judit Polgar being deflected by prophylaxis (16.
Tilburg 1 997 h3), he immediately makes the central
breakthrough, sacrificing material.
1 6 ... dxe4 1 7.fxe4 ltJxg4
D
Black is forced to go with the flow. It
is in no way better to play 1 7... tLlxcl
1 8 .'ifxcl �xg4 I 9Jhf2 or 1 8 ... t2Jxg4
1 9.h3 Wlc7 20.e5. If the pawn were on
cS, it would be a different matter alto
gether.
1 8. �f4 'ifh4 1 9.h3 ltJf6 20.e5!
8.f3!?
For some reason, White refrains from
the usual 8 . �d3 . One cannot rule out
the possibility that he hoped to provoke
Black's next move.
8 ... c4
A concession. The bishop on f1 will
find its way into the game somehow
anyway, whilst the tension between the
pawn duo c5/d4 would very much have The critical moment in the game. It
suited Black. would appear that, by answering blow
9.tbe2 tlJc6 1 0.g4!? with blow, by means of 20 ... �xh3
Those who have studied Kasparov's play 2 l . �xh3 'ifxh3 22.l:r.h2 'ifg4 (22 ...W/e6
in this structure know that he loves to 23.�xh6! tLlg4 24. �xg7! lLlxh2
include this move in the general plan. 2S .';£;>xh2) 23.�xh6! tLlxd4! (but not
1 94
23 .. .lbd5 24. �xg7! �xg7 25 .�hl !) Black has managed to exchange queens,
24.cxd4 �xd4+ 2 5 . �e3 �xeS, Black in itself a great achievement (no joke),
can emerge intact. She has four pawns but even so, the endgame is hopeless.
for the piece, the white king is not very The worst thing is that the attack con
secure, and not so many pieces remain tinues, even without the queens. And
on the board. she no longer has any extra material
However, if we continue the variation, with which to buy White off.
it turns out that Kasparov was right. 24 ... .:.es
White's attack assumes a mating char The assessment is not changed by
acter: 26.tbf5 tbg4 27 . .l::t.h 3 �f6 (the 24 ... i.xf5 2 5 .�xf5 gxh6 26.�xh5 .
threat was 28 .�h4) 28.�b l ! (one can 25 . ..te3 ..tc6
see the 13th World Champion's love of
chessboard geometry: much weaker
is 28 .tbh6+ 'i¥xh6! 29. .l::!x h6 l::txe3)
28 ... tbxe3 29.tbe7+! �xe7 30.�h7+
�f8 3 1 .'i¥h8#.
20 ... l:.ad8 21 . 'it'f2 ti.JhS
26 . ..tf1 !
Again the chessboard geometry. The
c4-pawn hangs, and on 26 ... i.d5
White decides things with 27. i.e2 g6
28 .tbh6+ and 29.tbxf7.
26 ...f6 27 . ..bc4 ..tdS 28 . ..te2
22 . ..bh6! After a few more moves, Black resigned.
Leaving two pieces and a pawn hanging
simultaneously. There are many vari The Botvinnik plan has survived and
ations, but they are simple: 22 ...�xg3 stood the test of time. We offer several
23 .'iVxf7+ �h8 24.'iVxh5 g6 25 .�g5 games for your own independent study:
�xg5 26. i.xg5 or 23 .. .'iio>h 7 24.�xh5 Kir.Georgiev-Mitkov (Poikovsky 2001),
gxh6 2 5 .l:lf7+ �g8 26 . .l:tf6; 22 ... gxh6 Zviagintsev-Alexandrov (Poikovsky
23 .'iVxf7+ �h8 24.tbxh5; 22 ... tbxg3 2002), Ibragimov-Kacheishvili
23 .�xf7+ �h8 24.�xg7#. (Connecticut 2002) and S.Guliev-Al
22 ... .:.e7 Modiahki (Dubai 2007).
Nor is there any help from 22 ... i.e6 But, as my friend the well-known
23 .tbxh5 'i¥xh5 24. i.xg7! �xg7 journalist and publicist Zamin Hadji
25 .�f6+ �f8 26 . .l::tf4. says: «Don't go too far, as then you cannot
23.ti.Jf5 'it'xf2+ 24. l:[fxf2 return»
1 95
Chapter 21
The Dialectics of Weakness

In the Exchange Variation of the Spanish, White voluntarily gives up his bishop,
at move 4, by exchanging it for the knight. In return, he devalues the enemy
queenside pawns, specifically by doubling the c-pawns. Later, he plays d2-d4 and
generates the exchange of the d4- and e5 -pawns, after which his e-pawn becomes
a candidate passed pawn.
But Black also has his plusses. He has a solid position, and if he can open the
position, his two bishops can have their say. In addition, the e4-pawn, even if it
becomes passed, will most likely be solidly blockaded on e6.
Therefore, in practice, White often adopts a different plan, one which can
seem paradoxical. White advances not his e4-pawn, but his f-pawn. By pushing
it to f5 , he appears to give the opponent an outpost on e5 , but he also seriously
restricts the enemy minor pieces, especially the bishop on c8 . Then his plan con
sists of two elements. He can try to get his knight to e 6 , using the f5 -pawn as
support, or he can shut off the e 5 -square by e4-e5 and force further favourable
exchanges. With every exchange, the active black piece play is weakened, whilst
the static weaknesses of the latter's pawn structure remain. The ideal for White is
to reach a pawn ending, or some other ending with few pieces remaining. On an
empty board, Black, effectively a pawn down, will not be able to stop the passed
pawn.
This structure (and the plan with f4-f5) is studied in the textbooks with the
games Fischer-Unzicker and Lasker-Capablanca. Other classical examples tend to be
ignored. It is time to rectify this omission, but first, we should proceed in proper
order.
So we start with the game from the

eleventh World Champion:
Robert Fischer
Wolfgang U nzicker
Siegen 1 970
1 .e4 e5 2.tlJf3 tlJc6 3. Ab5 a6
4 . .bc6
The first step in the plan.
4 ... dxc6 5.0-0 f6 6.d4
The second point. 1 4.f5!?
6 ... exd4 7. tlJxd4 tl::e7 8.Ae3 tlJg6 «Fischer i s indebted to Lasker for this outstanding
9.tiJd2 Ad& 1 0. tiJc4 o-o 1 1 . W'd3 positional idea» Gulko.
-
tt::es 1 2. tiJxes .bes 1 3. f4 Ad6 In this highly complex position, White

The critical position. is hardly likely to be able to put his
1 96
Chapter 2 1 - The Dialectics ofWeakness
knight on e6. More likely, he intends 1 9.g4!

to attack the square e5. His bishop will An excellent move. Before transferring
come to f4, his knight to f3 (after first his queen to g3 , White defends the
putting the queen on g3), and his rook f5-pawn, creating a wedge.
to e l . Then there will follow e4-e5 . The structure in the centre and on the
The plan i s clear. I t i s up to Black t o do kingside is favourable for White. The
something about it ... black light-squared bishop (as opposed
14 ... 'fie7 to the white knight, which soon trans
But Unzicker, almost to the end of the fers to f3) cannot influence at all the
game, decides not to change the pawn coming break e4-e5. Black faces diffi
structure. cult times.
Tempting is ... c6-c5 and ... b7-b6. Since 1 9 ... 'iVd6 20. 'fig3 .J:le7
the bishop's diagonal c8-h3 has been An obvious loss of time. Both 20 ... c5
closed, he seeks a new one. But the and 20 ... g5!? would have been stronger
immediate 14 ... c5 leads to the loss of than the text.
a pawn: 1 5 .lLlb3 "i!Ve7 1 6 ."i!Vd5+ and 21 . tbf3 c5
1 7.il..xc5 .
Therefore ... c6-c5 needs to b e prepared.
Grandmaster Gulko suggests the prophy
lactic 14 .. .'lt>h8 !?. The king moves out of
the possible check, and concrete varia
tions show that Black's position is fine.
1 5. �14 b:f4
After any other move, White can
exchange on d6, undoubling the black
pawns, as happened in the game Lasker
Capablanca. Evidently Unzicker knew
the latter game and decided that such 22.e5! fxe5 23 . .J:lfe4
a transformation of the pawn structure One cannot but be amazed not only by
was unfavourable for Black. Fischer's consistency, but also Black's
1 6. lbf4 ..td7 1 7.rl.e1 sense of fatal determinism, with which
1 7."i!Vb3+ �h8 1 8 ."i!Vxb7 is bad because he permits the realisation of the white
of 1 8 ..."i!Vd6. plan.
17 ... 'ifc5 1 8.c3 .l:.ae8 Black has done nothing to make diffi
cult the advance e4-e5 or otherwise
complicate the game.
There followed the exchanges ...
23 ... ..tc6 24. lbe5 .J:lfe8 25.lbe7
:Xe7 26. tbe5
Then the remaining rooks and the
queens disappeared from the board. In
the endgame of knight vs bishop, the
f-pawn proved decisive.
Black resigned on move 42 .
1 97
Emanuel Lasker That is why it is vital to combine

Jose Rau l Capablanca this plan with prophylaxis against
St Petersburg 1 9 1 4 the manoeuvre tLlb3-d4-e6. At some
moment, Black will need to play ... c6-
c5.
0
1 3. �f4 �b7
Straight after the game, Capablanca
pointed out what, in his opinion, was
a safe path to the draw: 1 3 ... i.xf4!
14.l:f.xf4 cS ! and then lS . .l::rd l i.b7
16.l:[f2 (nothing is offered by 1 6 . .l:td7
.l:tac8 1 7.\t>f2 i.c6 or 1 7... tLlc6, then
1 8 ... tLle5 , and the rook is driven off
the 7th rank) 1 6 ... �ad8 1 7Jhd8 Irxd8
1 2.f5!? 1 8 . .1:[d2 .l:txd2 (also possible is 1 8 ... Ilc8
«White tries to restrict the activity of the enemy - White's position is in no way better)
bishops. The weakness of eS that is involved in this 1 9.tLlxd2 tLlc6 20.tLld5 tLld4 2 1 .tLlxc7
is not of great significance, since Black is not going tLlxc2 . Vainshtein suggests the improve
to be able to exploit this with his pieces in the near ment 1 7. .l:Xfd2 , with the aim of seizing
future» Zak.
- the only open file. But after 1 7 .. Jhd2
Vainshtein, in the Russian book Myslitel, 1 8 . 1:lxd2 i.c6 (covering the vulnera
discussed the move in more detail: ble square d7), and then ...tLle7-c8-d6,
«This 'anti-positional' move simultaneously breaks Black has sufficient counterplay against
three of Steinitz's rules. White is left with a back the pawn on e4. The position is such
ward pawn on e4, a hole on eS, where the black that White already is the one needing
knight will aim, and it becomes extremely difficult to be careful, otherwise he risks ending
to create a passed pawn, which is the only strategic up worse.
advantage White has in this variation, namely his According to the testimony of eye-wit
4v3 pawn majority on the kingside. nesses, the great Capa played the first
But in reality Lasker had formed a grandiose plan: part of the game very quickly, not
create an outpost on e6, enhance the scope of his doubting in the drawn outcome (the
remaining bishop and relieve Black of his bishop tournament position was such that a
pair, leaving him with the light-squared bishop, draw practically guaranteed him first
which will be nothing but a problem child. With prize) . An endgame, the two bish
out doubt, this whole plan was thought up by ops , so close to the desired goal...
Lasker in advance.» evidently, the Cuban lost his sense of
1 2 ... b6 danger and became careless, thinking
Tarrasch and Reti recommended the draw could be achieved however
1 2 ... i.d7, and then ... na8-d8 and . ..Ci:Je7- he wanted, and that there was no need
c8. Capablanca chooses a different con to look for the most exact sequence of
tinuation. Evidently, he is preparing to moves.
develop his light-squared bishop on the To be fair, one should add that White's
long diagonal. But then the white knight next move was not easy to foresee.
will get the chance to come to e6. 1 4 . .bd6! cxd6 1 5. ttJd4
1 98
We have the very pawn structure that 29. l:[g3

Unzicker was so afraid of in the previ An important moment. Capablanca puts
ous example. The c-pawns have been a question mark against his next move,
undoubled and now White cannot showing how he could have made a
create a passed pawn in the centre. draw. Zak disagreed with him, and 'in
Even so, he has a clear plan to strengthen any case', shows another way for White
his position. The knight comes into e6, to win.
covering the weakness on e4. The king Instead of the unhurried 29 . .l:lg3 , Zak
comes to f4. The rooks will double analyses the immediate break 29.gS ,
on the d-file, tying Black down to his and then: 29 ... hxgS+ 30.hxgS .l:lh8
weakness on d6. Then the a-, b-, g- and 3 1 .gxf6 �xf6 32.�xd6! llh4+ (bad is
h-pawns will advance, seizing space. 32 ... lt:Jxd6 33.eS+ �f7 34.exd6) 33 .'�g3
The main thing is that it is extremely lt:Jxd6 3H�ixh4.
difficult for Black to do anything active,
and in passivity lie the seeds of defeat.
1 5 ... :Sd8
1 S ... �c8 !?.
1 6.ttJe6 l:[d7 1 7.l:ad1 0.c8 1 8.l:[f2 b5
1 9. l:[fd2 .:de7 20.b4 <M7 21 .a3 .b8
The well-known trainer and method
ologist Alexander Konstantinopolsky
suggested the radical decision of an
exchange sacrifice: 2 1 ..J�.xe6!? 22.fxe6+
l:lxe6. Then the king occupies a role in
analysis diagram
defending the weakness on d6, and the
knight, in his turn, comes via b6 and c4,
to the square eS, cementing the position. This position arises by force. At first
Capablanca preferred to stand and wait. glance, White should have no trouble
Things should not be so bad for him as realising his advantage. He has an extra
to necessitate parting with material. pawn, and Black cannot take on fS ,
22. cJtf2 l:a7 23.g4 h6 24. l:[d3 aS because his knight hangs. The bishop
25.h4 axb4 26.axb4 l:ae7 27. cJtf3 on c8 is also very bad, closed in by his
l:[g8 28. cJtf4 g6 own pawns.
1 99
Zak gives several variations, amongst 33 ... �f8, more tenacious is 33 ... <;t>e8,
which is an unnecessarily cooperative and if 34.f6 , then 34 .. J Xf8. In his turn,
one, ending in mate: 34 ... gxf5 3S.l:lxd6 White can play more strongly: 33 . .l:txg5
l:1xe6 36 . .l::t.d 8 fxe4 (36 ... �b7 37 . .l::t.f8+) (the extra pieces only slow down the
37.l:1xa8 e3 38 . .l::t.a l �eS 39.tt::le2 �dS advance of the passed pawn) 33 ... .l::t.xg5
40. J::i.d l+ �c4 41..l:ld4#. 34.tbxg5+ �f6 3S. J::i.h l etc. But this is
But in fact the position is unclear. After not the main thing.
34 ...tt::l c4! (weaker is 34 ... tt::lf7 3S .<;t>g4 Instead of 3 l ...hxg5+, much stronger is
gxfS+ 36.exf5 tt::lh 6+ 37.�f4 tbxfS 3 1 ...fxg5+ 32.hxg5 tt::ld 6! after which
3 8 .tbc5! and the knight on fS has no it is hard to see a move for White. The
way to come to the rescue of its bishop) threat is to take on fS , and 33.g6+ is
3S.l:Id8 gxfS 36.exf5 �xfS 37.tt::lc 5 �b7 senseless because of 33 ... <;t>f6.
3 8 . .!:f.b8 tt::ld 6, the black pieces form a It seems his intuition did not deceive
united group, and coordinate well. the great Cuban and he could really
One can conclude that the move 29.g5 have saved the game.
has its place and poses Black a difficult 30. �3
practical problem, but one cannot say Necessary accuracy; after 30.hxg5
that it wins. hxgS+ 3 l .�f3 .l:Xh8 , Black is first to
29 ... g5+ seize the newly-opened file.
It was this move for which Capablanca 30 ... ttJb6 31 .hxg5 hxg5 32 . .l:.h3
criticised himself, considering that 29 ... Ignoring the proffered trophy. After
gxfS promised a draw, although he did 32 . .l::t.xd6 tt::l c4 33.l:ldl .l::th8 , Black gets
not give any variations. a breathing-space, at the cost of a pawn.
But Vladimir Grigorievich Zak provided 32 ... l:.d7
them in abundance. The main events, 32 ...tbc4 looks more natural, but
in his opinion, take place after 30.exf5 Reti found a forced win in this case:
(of course, not 30.gxf5 .l:lxg3 3 l .�xg3 33 . .l:Xh7+ �e8 34 . .l:lal ! �b7 (the rook
.l::te 8, and the white pawns are stalled) cannot be taken, and the intermediate
30 ... d5 3 l .g5 check on eS cannot change anything)
3S .tbc7+! �d7 36 . .l::t.xe7+ �xe7 37 . .l:Xa7,
and material losses are unavoidable:
37 ... .l::tb 8 3 8 .tt::la6, or 37 ... �c8 3 8 .tt::l 7d5+
and 39.tt::lx f6.
33. �g3 �e8 34. .l:.dh1 �b7
analysis diagram
3 l ...hxg5+ 32 .hxg5 fxgS+ 33.tt::lxg5+
�f8 34.f6 �a7 35.�e5 with dom
ination. But alas, the variation con
tains several inaccuracies. Instead of
200
35.e5! great psychologist, the first to bring this

White also wins with 35 . .l:lh6 �f7 element into chess. His opponent, David
36.l:!g6, but the text is sharper. It shows Janowski, was an inventive player, but
the correctness of the white plan, begun easily discouraged and with an unsta
as far back as move 1 2 . ble character. Put yourself in his place.
3 5. . .dxe5 36. t004 ttJc:I S 37. ltJSc5 After just thinking a draw was the most
�c8 38. �d7 bd7 39. l:.h7 l:.f8 he could hope for, suddenly he has the
40.l:.a1 <iltd8 41 . l:.a8+ �c8 42. ttJc5 chance to play for more! The chances of
Black resigned. his getting carried away and losing his
objectivity, in striving for the win, were
Emanuel Lasker very great. And then White in turn will
David Janowski get chances.
Paris 1 909 20 ... l:.e7 21 . �f4 ttJc6 22.g4 l:.he8
23. l:.e3
The alternative was 23.J::te l , intending
0
after 23 ... tLle5+ 24. �xe5 J::txeS to reply
25.Wf4. The rook d3 is freed from defen
sive functions and can swing across the
third rank to the kingside. However, the
variation needs to be continued: 25 ... c4
26.l:th3 h6 27.g5 (or 27. .l::th 5 f6 28.h4
Ilh8 29.Wf3) 27... hxg5+ 28 .\txgS l:i.xe4
29. Ilxe4 l:lxe4 30 . .l:lh7 �e3 3 I. .l:txg7
l:txc3. Without further analysis, it is
20.f5!? hard to say whose passed pawn is the
Even after seeing the two previous more dangerous.
examples (and the whole plan) , this 23 ...f6
move looks unexpected. Just as it Here it made more sense to exchange
seemed the black knight had no pros minor pieces: 23 ... tt:Je5+ 24. �xe5
pects, suddenly it is given the central �xeS. But Janowski, as often happened
outpost eS. In addition, White appears with him, rushes on at full speed.
needlessly to weaken his entire pawn 24.g5 ttJa5
structure, from the a-file to the h-. Again refraining from 24 ... tt:Je5+.
However, looking further into the posi 25.h4 ttJc4
tion, we can understand that Lasker's
move is not bad. Effectively, the second
World Champion is carrying out the
same plan as against Capablanca, five
years later. After waiting until his oppo
nent plays .. .f7-f6, White begins a pawn
push: g2-g4-g5 , h2-h4-h5-h6, trying to
open lines on the kingside.
One cannot omit to point out the psy
chological aspect also. Lasker was a
20 1
Black has achieved his aim. The knight

stands very attractively on the blockad
ing square c4, but what else? There is no
obvious plan to strengthen the position.
But White does have a plan: the advance
of the kingside pawns.
26. l:e2 .l:.f7 27 . .1:.g1 c;f;>d7 28.h5 tbd6
29.h6!
The game is effectively over. There fol
lowed:
29 ... fxg5 30. :Xg5 g6 31 .fxg6
hxg6 32. :Xg6 .l:.ef8 33 . .1:.g7 :Xg7 While Black runs o n the spot, White
34.hxg7 .l:.g8 35 . .1:.g2 tt::ea 36. �e5 flexes his muscles. In the diagram posi
�e6 37. c;f;>f4 1 -0 tion, everything is set for the storm.
The signal for the attack should be the
Emanuel Lasker move 3 l .'iVc3 ! . After 3 l ...'iVb4, a queen
Georg Salwe sacrifice is prepared: 32 .tbxf6! 'iYxc3
St Petersburg 1 909 33 . .l:hh7#. Mate follows after 32 ... gxf6
33 .'ifxf6+ Wg8 34J:rg4+. More tena
cious is 3 l ...tbeS , but then 32 .tbxe5 is
D
winning in all variations:
32 .. Jhe5 33 .tbxg7! �xg7 34.'iYg3+
�h8 3S.!lxh7+! (again, a cascade of
sacrifices leads to the win) 35 ... �xh7
36.'iYg6+ <;t>hs 37.'ifxf6+ �h7 3 8 . l::re 3!
l::rxe4 39.'iYg6+ �h8 40.'iYh6+ �g8
4L�g3+ �f7 42 .l:tg7+ c;t>fs 43 .'iYf6#;
32 ... dxe5 33 .'iYg3 ! (on 32 .. .fxe5 the same
move would follow) 33 ...'iVd6 34.'iVg6!
1 6.f5!? h6 (34 ... hxg6 35.tbxf6#) 35 .tbxf6!
The well-known, but always unex 'iYxf6 36 . .l::f.x h6+ �g8 37.'iVh7+ �f8
pected move. The black knight, cur (37... �f7 38.�xf6+ �xf6 39.'iVg6#)
rently stranded on the edge, gets the 38Jhf6+ gxf6 39.'iVh6+ �f7 40.'iVg6+
square e5. But here White's plan will be �f8 4I .'ifxf6+ etc.
somewhat different. Lasker is not plan However, White dallied with
ning to advance g2-g4-g5 . He leaves the 31 . .1:.g4
two files on the edge of the board free, Although he won 30 moves later, one
for a piece attack on the king. can say that the game went off its logi
1 6 ... f6 1 7. tbd5 :ea 1 8.c4 l:i:Jf7 cal course.
1 9.1i'c3 .l:le5 20. tbd2 c6 21 . 1:i:Jf4
1i'b6 22.b3 .l:.be8 23. 1i'g3 c;f;>h8 So, should Lasker be considered the
24. 1:i:Jh5 .l:.g8 25 . .1:.f4 1i'd8 26. 1:i:Jf3 founder of the plan with f4-f5 ? That
.l:.e7 27 . .1:.h4 1i'e8 28. 1i'f2 .l:.f8 is the generally-accepted opinion. But
29. 1i'd2 'ifb8 30. c;f;>h1 .l:.fe8 look at the following encounter.
202
Joseph Blackburne nothing decisive is to be seen: 23 .'iVf6+

William Steinitz (23.tbe4 klxe6) 23 ... �g8 24.tbe4 ti:Jds
London 1 883 or 24 . .l:ld3 .l:Ig7 2S .tbe4 ti:Jds .
1 8.f5! tbc8 1 9.e5!? 23. l:l.d3!

From Lasker, remembering his game There is no defence against the doubled
with Salwe, one could expect the major pieces on the h-file. Black played
unhurried bringing-up of the reserves, on to the bitter end:
with something like 19 . .l:i.f3 . Blackburne 23 ... l:l.xe6 24.:Z.h3 'ife7 25. 1Vh6+
proceeds more energetically. ..tg8 26. l:l.f8+! "it'xf8 27. 'ifxh7#
1 9 ... fxe5 20. ltJe6+ .b:e6 21 .fxe6
l:l.e7 One is reminded that Lasker once com
More precise is 2 l ....l:i.f6, preventing the pared Steinitz's heritage to a university
queen raid on gS , because of 22 ... .l:[xfl+. department. Maybe the second World
22. 'ifg5 1Ve8? Champion had studied this game and
This natural move is the decisive mis learnt the idea - even though it was by
take. Also weak is 22 ... 'iVb6+ 23 .�hl Steinitz's opponent this time?!
.!:f.xe6, because of 24.�g4 l'le8 2S .ti:Je4 The idea being that every weakness can
'iVc7 26.tbgs . But after 22 ... ti:Jb6, be a source of strength.
203
Chapter 22
Structure and Plan

When a chess player thinks about a plan, he thinks, among other things, o f changes
in the pawn structure. The opposite is also true - after noting a favourable change
in the pawn structure, the player forms his plan. These two components, structure
and plan, are inter-connected, and should not be separated.
What role in the formation of a plan is filled by the classical heritage (knowledge
of classical games, devices, typical pawn structures) ? It serves as a support. It is al
ways nice, and very useful, to know that somebody, sometime, carried out a similar
plan, especially when this 'somebody' is a well-known, strong player and especially
when the plan succeeded.
Knowing this, it is like the player has firm ground under his feet. It develops
confidence, and feels as though he has a helper sitting on his shoulder, an authori
tative grandmaster, giving him advice on what is good and what not.
He can play restrictively with ... c7-cS,

Boris Spassky preventing the advance d3 -d4. This is
Artu r Jussu pow the most solid plan. This is what hap
Linares 1 990 pened in the game Spassky-Timman
(Cannes 1 990): 1 2 ... tt.Jd7 1 3 . �el cS!?
• 14.b4 �b7 l S .bxcS dxcS 1 6 . �aS l::tac8
1 7.4.Jd2 �fe8 1 8 .tt.Jfl 'iYe6 1 9.4.Je3
tt.Jb8 ! ? (a truly grandmasterly move)
20.�c3 tt.Jc6 2 l .b3 tt.Jd4. Black has
comfortable play, although, admittedly,
he was not able to win the game.
There is also one more plan, the most
concrete. The lines of the bishop on
b7 and rook on f8 connect at a single
square, f3 , currently occupied by a
Black to move. He has practically fin white knight. It is tempting to carry out
ished his development, and connected the advance .. .f7-fS , opening the f-file
his rooks. This is the appropriate time and so harry this knight.
to think about a plan. He has three ideas White will also not sit still, with his
to choose from, all of them connected arms folded. Once the enemy knight
with a change in the pawn structure. moves from f6, to free the path of the
He can prepare the central break ... d6- f-pawn, White will try to react in the
dS. But this is risky; after the opening of centre with d3-d4. We are very likely
the e-file, the pawn on eS can become to reach a situation with two pairs of
a target. pawns in tension: d4/e5 and e4/fS.
204
Chapter 22 - Structure and Plan
Here, general considerations will not 1 6 ... .:Xf3!

help, and we need accurate calculation The key moment of the game. Clearly,
of variations. Jussupow did not come this far to settle
No doubt Jussupow went through this for a simple recapture (16 ... dxe5 or
same reasoning process. But he was also 1 6 . ..l2Jxe5). Instead, he initiates great
able to draw on knowledge, the expe complications.
rience of previous generations. As far 1 7.gxf3
back as 40 years earlier, grandmaster It is important that White does not have
Lilienthal carried out a similar plan in the zwischenzug 1 7.exd6. After 1 7...'iff6
much the same position. And, almost 1 8.gxf3 �xf3 the bishop on d2 deprives
mating Taimanov, he won the game in its queen of vital operating space. We
brilliant style. There were also other will return to this moment later.
games, in which the plan with .. .f7-f5 1 7 ... tbxes 1 8 . ..tf4
justified itself. Black would have retained a strong
1 2 ... ttJd7!? attack after 1 8 . .l:le3 'i¥g5+ 1 9.�h1 'ifh5
The move-order is also significant. or 1 9.�fl l:if8.
If Black begins with 12 ... �b7, White
advances his bishop: 1 3 . �g5 , and holds
up the advance .. .f7-f5 . Such a develop
ment of events took place in the game
Spassky-Short (Thessaloniki 1 988). In
addition, one must reckon with 1 3 .t2Jh4.
1 3 . .:1.e1 ..tb7 1 4.d4 f5!?
Both players should have got this far
in their preliminary calculations. The
forcing variation continues:
1 5.exf5
In the game Lengyel-Lukacs (Budapest 1 8 ... tbxf3+
1 993), White was clearly disturbed by Commentating on this game, Jussupow
something he imagined, and he played criticised himself for hurrying to regain
1 5 .dxe5. After 1 5 .. .fxe4 16.t2Jg5 'ifxe5, the material, giving his opponent a
Black had a healthy extra pawn. chance to make a draw in the endgame.
1 5 ... .:Xf5 1 6.dxe51 In his opinion, it was stronger to play
1 8 ... l::[f8 , and if 1 9.�xe5 dxe5 20.'ifd4,
then 20 ...'ifg5+ 2 1 .'i¥g4 'iff6.
Jussupow's decision is a pragmatic one.
Rather than an unclear middlegame,
with the initiative but an exchange
down, he prefers an endgame with a
guaranteed extra pawn. Anyone willing
to criticise him for this has never played
serious chess themselves.
1 9. 'ifxf3 'ifxe1 + 20. .:Xe1 .hf3
21 . .:1.e7 .:r.ca
205
It is too early for 2 I ... .l::i.f8 22.1td2 �f7, game, after 16.dxe5, Black essayed the
on account of 23 . .!::re 8+ .!::rf8 24. �e7, thematic exchange sacrifice 1 6 .. Jhf3 ,
repeating moves. but ran into the unpleasant zwischen
22.b4 h6 23.h4 .l:.f8 24 . .ie3 .l:.f7 zug 1 7.exd6. The problem is that after
25 . .:.Xf7 Cit>xf7 1 7 .. ."i¥f6 1 8 .gxf3 1txf3 , White, unlike
Only now can we say that the plan begun in the game Spassky-Jussupow, has the
on move 12 has ended. Play proceeds resource 1 9.'ii'd 2! and there is no mate:
for two results. The opposite-coloured 19 ...'iVg6+ 20.'iVg5 , or 1 9...'iVh4 20.'iVe3
bishops in the ending give White 'fi'xh3 2 1 .'ii'e 6+. That said, Black retains
enhanced drawing chances, but a pawn promising play after 1 8 ... tt::le 5! 1 9.l::i.xe5
is a pawn. At the board, Spassky could 'ii'xe5 20.dxc7 'fi'xc7 or 20 ... .l::i.f8 .
not cope with the problems and lost. Nevertheless, we can conclude that
Jussupow carried out the classical plan
And now a look at the source game. in an even more favourable form.
1 4. 'iVd2
Mark Taima nov Taimanov, like Spassky, seeks salvation
Andor Lil ienthal in the endgame.
Moscow 1 948 1 4 ... f5! 1 5.'ifg5 'iVf7 1 6. 'ifxf5 'iVxfS
1 7.exf5 bf3 1 8.gxf3 .:.Xf5
• One should not overstate Black's advan
tage. Yes, the white pawn structure is
in ruins, but a bishop is a bishop. After
1 9.ret>g2 tt::lc s 20.b4 tt::le 6 2 1 . 1te3 it is not
entirely clear how to free the rook from
the defence of a6 , and without the rook,
it is even less clear how to develop the
initiative.
1 9.f4 exf4! 20 . .l:.e7 lLleS
Black creates a mating net with mini
This example and the previous one mal forces.
are like Siamese twins. The small dif 21 . .:.Xc7
ferences are insignificant. White has
played h2-h3 , and Black ... 1tc8-b7.
Instead of 1tcl -d2 , White has played
.l::i:f l -e l . That is all.
1 3 ... llJd7! ?
In his notes, Lilienthal gives the vari
ation 14.d4 f5! 1 5 .exf5 �xf5 16.tt::lh 2
"ifh4 1 7. 1te3 .l::taf8 «with an irresistible
attack».
However, the move 1 6 .tt::lh 2 is not the
most principled. A more important var 21 ... .l:.e8!?
iation occurred in the game Janosevic Lilienthal gives this move an exclama
Messing (Cateske Toplice 1 968). In that tion mark. Mate is certainly close, as
206
confirmed by the vananon 22.l::[xa6 The unfortunate knight is not on b7, but
.l:tg5+ 23 .�fl tLlf3 . Even so, more accu d8 , which is slightly better, because it
rate was the prosaic 2 I ....l::f.g5+ 22 .'�fl frees some communications within its
tLlf3 23 .�e2 tLlgl + 24.�dl tLlxh3 . own camp. The knight is heading for f7.
Now, however, with the move White needs to act energetically, else his
22.h4 advantage will be reduced to nothing.
Taimanov deprived the rook of the 30.f4! f6
square g5 , and somehow managed to The break f2-f4 is not good, if after the
hold together his defences. The battle capture on f4, Black can quickly get his
was prolonged, but Black won at the knight to e5. In this situation, there is
first time control, move 40. no question of this. The variation 30 ...
exf4 3 1 .tLlxf4 is described as 'cheerless'
In the classical variation of the Spanish, by Karpov, who adds that 3 I ...ii.h6 is
the centre is often closed (pawns on e4/ not then possible, because of 32 .tLlxg6! .
d5 against pawns on e5/d6). The weight 31 .f5 gS
of the battle shifts to the flanks. Knowing what awaits Black, one wants
White often wins the battle for the a-file to look for some sort of alternative, such
in such cases. But that is not the main as ... tLld8-f7 (immediately or after cap
thing. The black knight, driven on the turing on f5).
route b8-c6-a5-b7, is dominated by the 32 . .tc2!
pawn on b4 and becomes a burden to The eventual aim is the square h5.
Black. White has practically an extra 32 ... .tf7 33. tbg3 t:Db7
piece. And if, having obtained a certain The last chance of activity was 33 ... h5!?
advantage on the queenside, he man 34.ii.dl h4.
ages then to transfer the battle to the 34 . .td1 h6 3S . .th5 'ife8 36. 'ifd1
other flank (beginning an attack on the t:Dda 37.J:.a3 <Jtfa aa . .:1 a2 <;t;>ga
king) , the extra piece in the attack can 39. tbg4
be a big factor.
The following game has long since
become an absolute classic.
Anatoly Karpov
Wolfgang U nzicker
N ice 1 974
One can only admire the positional

mastery of the young Karpov. All of
White's pieces are ideally placed!
The bishop on a7 prevents Black
fighting for the a-file. The rooks are on
a3 and a2 , so as to allow the queen, if
207
necessary, to go to a l . The other pieces The classical plan (tbg3 -f5 , \t>gl -h2,
are involved in the fight for the light g2-g4, klel -gl etc.) looks a bit cum
squares on the kingside. And they have bersome here. White needs something
almost won that fight. more concrete.
39 ... cif.?f8 28. l:[f1 !
Taking on h5 was not possible. Making it quite clear that the next
40. l2Je3 �g8 41 . .bf7+ tbxf7 move is going to be f2-f4. The counter
42. 'ifh5 ttJd8 play with taking on f4 and putting the
Karpov gives the amusing vananon knight on eS does not work: 28 ... .id6
42 ... tt:Jh8 43 .tbg4 '#Wxh5 44.tbxh5 \t>f7 29.f4 exf4 30 . .ixf4 .ixf4 3 1 . llxf4 �d6
4S . .ib6 l::tx a3 46 . .l:1xa3 lla8 47. l:i.xa8 32 . .l::f.cfl . One can suggest the exchange
tbxa8 48 . .id8 or 48 . .ia5 , and not one of queens: 28 ... tbc3 29.'#Wc2 'iVa4, but
enemy piece can move. Gligoric decides to meet the storm with
43. 'ifg6! �8 44. tLlh5 a board full of pieces.
Black resigned. 28 ... 'ifc7 29.f4 ttJd7 30. 'ifc2 tLlc3
31 .f5 ttJf6
In the first half of the 1 970s, Karpov
often reached Spanish positions, in
which the advance f2-f4 was an impor
tant element of the plan. It got to the
point where this device became known
in the press as 'Karpovian'.
We will examine another example on
this theme.
Anatoly Karpov
Svetozar Gligoric
San Antonio 1 972 32. l2Je2
Played in Botvinnik style. White carries
out his general plan, refusing to be dis
D
tracted by trifles, even if these are in his
favour. This is what Karpov wrote about
his 32nd move:
«It looks more solid to win a pawn with 32 .tbf3
and 33 . .id2, but it seemed to me that Gligoric
would then be able to hold up the kingside storm,
by establishing a blockade on the dark squares
(. .. .if8-e7 and ... tbf6-h7). Therefore, I decided
not to lose any time.»
Unlike the previous example, Black 32 ... tbxe2+ 33 . .be2 �d6
does not have a terrible knight on b7. Freeing a path for the king to the
On the contrary - on a4, the knight is queenside. The alternative was the
active, and threatens to land on c3 with above-mentioned blockade on the dark
tempo. Overall, the position is unclear. squares: 33 ... tbh7 34.tbf3 .ie7 35 .'#Wd2
208
Chapter 2 2 - Structure and Plan
'ifd8 , and if 36.tLlxe5, then 36 ... tt:Jg5 bring one of his knights to e5 quickly.
37.i.f3 i.f6 38 .tLlg4 tLlxe4. However, Here, this is impossible, whilst the
after 33 ... tLlh7, one must reckon with white knight on h2 obtains access to the
34.f6!? tt:Jxf6 35.i.xh6. White's attack is ideal square d4.
dangerous. 30. 'ifxf4 .rl.fe8 31 . tl::lf3 �g7
34.g4 �8 35.h4 �e7 36.g5 hxg5 The only choice is to rely on tactical
37.hxg5 tricks, based on the undefended knight
Karpov has fulfilled his plan, and has on e3. But the immediate 3 I ...f5 32.exf5
managed to seize space. But the black �xe3 is bad, because of 33 .'iYd4+ .!::f. 3 e5
king has fled in time. There is still a 34 . .!::f.xd6!
great deal of play ahead, and White 32. tt::Jd 4
finally broke his opponent's resistance
only on move 52.
But even Karpov had people from

whom to learn. Maybe it was from
Leonid Stein, a brilliant player, three
time champion of the USSR, who died
far too early.
Leonid Stein
Aleksandar Matanovic
Tel Aviv 1 964 32 ... f5
This loses, like everything else. The
threat was 3 3 . l::r.b 6 'iYc7 34.l::t aa6, win
D
ning the b5 -pawn. Then the knight will
come into c6 and Black's position is
hopeless.
33 . .rl.a8! 'ifxa8 34. :Xa8 :Xa8
35.exf5 :es 36. tl::lg 4 :aea 37.f6+
�7 38. tLlh6+ tl::lx h 6 39. 'ifxh6 �xf6
40. 'ifxh7
Black resigned.
«Stein's talent was fantastic!» Karpov.
-
29.f4!
«The signal for the attack. Despite the material But Stein was not the first to play f2-f4
equality, Black's position is objectively lost, as his in this structure. In the 1 920s and 30s,
pieces are badly placed» Stein.
- Akiba Rubinstein loved to defend such
29 ... exf4 positions as Black. He even agreed to
An incorrect choice. It was necessary to have the knight on b7. He would then
play 29 .. .f6 and, gritting one's teeth, to transfer it to d8 and f7 and grind his
resist with all one's might. opponents down in 1 0 0-move battles.
As we have already noted above, Black Those opponents, in turn tried various
can only afford to take on f4 if he can plans, including f2-f4.
209
George Thomas More promising is 2 1 .f5 , but this only

Akiba Rubi nstein became the fashion many years later,
Liege 1 930 after the analysis of games like this.
Although, it should be said, in this case
too, Black's defensive resources are far
0
from exhausted: 2 I ....ixg5 22 .'ifxg5 h6
(but not 22 .. .f6 23.tt::lh 5+!) 23.tt::lh 5+
�h8 24.tt::lxg6+ fxg6 25 .'ifxg6 .ie8 .
The attack comes t o nothing.
White is not obliged to sacrifice the
knight, although after 23.'ife3 , there
is 23 ...'iVc5 , forcing the exchange
of queens, whilst after 23.'ii'd 2 .l:If8
24.�fl 'iVd8 25.tt::lf3 tt::lc s, Black holds
Tempting is 20.tt::lh f5+!?. If 20 ... gxf5 , everything.
then 2 1 . .ixf6+, mating: 2 I ... .ixf6 21 ... t2Jxe7 22. <�h2 tbg8 23 . .1:.11 16
(2 I ...�xf6 22.'ii' h 6#) 2 2 .tt::lh 5+ �g6 24 . .1:.12 .1:.18 25. l:.a11 .l:.ae8 26.1xe5
23 .exf5+! .ixfS (23 ... �xh5 24.g4+ dxeS
�h4 2S .'ifh6#) 24 . .ixf5+ �xfS Bad is 26 .. .fxe5 2 7.tt::lh 5+! gxhS
2 5 .'ii' h6 .ih4 (25 ... 'ife7 26.tt::lg3#) 28.'ii'g 5+, winning the exchange.
26.tt::lg 7+ �e4 27.f3#. Stronger is 27. 'ti'e3 tbd6
20 ... .ixf5 2 1 .exf5 tt::lc S, and nothing
decisive is obvious, although after
2 2 .f4 or 22 .b4, White does have the
initiative.
In 1 930, Sir George Alan Thomas was
close to his 50th birthday. As befits an
Englishman, he played in a solid style,
not liking to throw his pieces around.
In the diagram position, he thought up
a different plan:
20.14!?
The idea is that after 20 ... exf4 2 1 .'ifxf4, The structure has again solidified.
White (for example, in reply to White has not achieved a great deal; he
2 I ...tLlc5) would play 22.�fl . establish has more space, but that is all.
ing unbearable pressure on the f-file. Alas, after a few more moves, Rubin
And 2 l ...�f8 is not possible, because stein lost on time. A few years later,
of 22 . .ih6+, whilst after 2 l ...�e8 the he ended up in a psychiatric hospital,
sacrifice 22.tt::lhf5+ gains many times in from which he never emerged until his
strength. death.
Rubinstein (as he often did in this struc As for the tournament at Liege, it is
ture) retreats behind his long-prepared remembered chiefly not for Rubinstein's
barricades: failure (he was 9th) , and not even for
20 ... ltJg8 21 . he7 the result of the winner (Tartakower
210
took first prize confidently). It was in anyone. In slow, trench warfare, he was
Liege that the incomprehensible Indian in his element.
star Sultan Khan was first widely seen. Understanding this, White forces the
Six wins in a row at the start, then four pace:
defeats in a row at the finish... but this 21 .f4!?
is quite another story. A rare case, where the break f2-f4 is
used not to search for an advantage, but
in order to maintain equality.
Mario Monticelli 21 ...exf4 22 . ..bf4 �d7 23. 1i'f3 �g7
Akiba Rubi nstein 24. Aad1 l:l.f8 25. �h1 :aea 26. l:l.e2
Budapest 1 926 �c8 27.l:l.de1
Black does not have a bad knight on b7 By tying the enemy forces down to
and in general, has no bad pieces at all. stopping e4-e5, White prevented them
Having completed his development, pursuing their own plan.
and placed his pieces in the best pos Rubinstein subsequently did not
sible way, he will probably prepare one manage to find a way to untangle. He
of the breaks ... b5-b4 or .. .f7-f5 . If the played .. .f7-f5 , but this only led to the
game developed quietly, in manoeuver exchange of rooks on the e-file and a
ing style, Rubinstein was not inferior to draw.
Chapter 23
Torture to any Taste

In chess books, one can see many references to the 'Spanish torture'. The analogy
is as follows. In many variations of the Closed Spanish, Black emerges from the
opening with a cramped and difficult position. And then, over the course of 20, 30
or even more moves, he fights for a draw, and has not the slightest chance to seize
the initiative. This difficult, even despairing experience creates associations with
mediaeval Spanish torture, such as the infamous 'boot', which crushed the toes of
the foot. Even if by chance the victim was allowed to go free after this torture, he
would be unable to walk.
Something similar occurs in certain lines of the Spanish Opening. Imagine that,
after all his trials, Black manages to draw. He is likely to be so shattered after this
desperate battle for half a point that he is unable to play at his normal strength in
his next few games in the tournament. The 'Spanis
h Boot' leaves its marks.
We will now speak about one of these variations.
vian plan, seen in the previous chapter.

That plan assumes that the main route
for the attack will be the f-file. The plan,
as we have seen, has demonstrated its
effectiveness.
So if Black's prospects are so bad, why
does he go in for this position? The
answer to this question can only be
given from a historical perspective. We
need to make a small excursion into the
past.
In the first third of the 20th century, the
genius of Rubinstein dominated world
chess. It was his play that gave rise to
the concept of the 'long-term plan'. It
The pawn structure gives White the turned out that this grandmaster's plans
advantage on both flanks. On the queen lasted not 2-3 moves, or even 5-6, but
side, he can become active with a2-a4 far more. Sometimes the plan would be
and b2-b4; opening either file is to his one for the whole game, from opening
advantage. On the kingside, his pros to endgame.
pects are even brighter. He can choose Rubinstein played in a style shorn of
between two attacking plans: the clas unnecessary spectacle. He loved posi
sical (g2-g4, lt:Jd2-fl -g3 -f5 , �gl -h2, tions where the pieces could manoeu
llel -gl etc.) and the modern, Karpo- vre slowly, frequently changing places.
212
Chapter 23 - Torture to any Taste
The contours of such positions would - combine threats on both flanks to tie
remain unchanged. But here is a strange up the enemy forces, force favourable
thing: with each new exchange, the exchanges and seek chances in the end
position of Rubinstein's pieces would game.
somehow slightly improve, and that
of his opponent, slightly deteriorate. Such a plan is easier to describe than to
Finally, there would be a collapse of the implement. But let us see how Rubin
enemy position. When one analysed the stein did it.
game, it would appear that all of Rubin
stein's moves, so apparently random at George Thomas
first sight, were united in one single, Aki ba Rubi nstein
surprisingly deep, plan. The opponent, Baden·Baden 1 9 25
without realising it, had been bent to
Rubinstein's will.
0
Rubinstein's contemporaries were
unable to copy his style. Far from all
were able to put up any opposition to
him, either, only the very strongest.
Akiba Kiwelowicz was somewhat ahead
of his time. Only in the subsequent
generations did players emerge, who
could construct plans which matched
Rubinstein's in depth and power. They
learned from his games and took the For convenience sake, in this and the
best from them. next few examples, we will pick up the
Let us return to the last diagram. Black's notation from move 14. The first 1 3
position is difficult, but he is not losing moves are given above.
by force. Rubinstein worked out the fol 1 4. ltJf1 tLle8 1 5.a4 .:tb8 1 6.axb5
lowing plan: axb5 1 7.g4 g6 1 8. lt:)g3 lt:)g7 1 9. '1th1
- connect the rooks. This is done by f6 20. .:tg1 ltJf7 21 . 'it'f1 �d7
... g7-g6, ... tLlf6-e8-g7; .. .f7-f6, ... tLld8-f7; The first stage of the plan is com
... �c8-d7, clearing the back rank; pleted.
- secure himself against a mating attack. 22. �e3 .:ta8! 23. 'it'g2 :Xa1
With this in mind, wait for the opening 24. :Xa1 'it'b7 25. '1th2 :as 26. 'it'f1
of the a-file and exchange all the rooks; :as 27. ttJd2 'it'a8 28. :Xa6 'it'xa6
- once his own king is safe, go over The second stage is also completed. (It
to active operations. The queen, uti sounds like the controllers of a space
lising the absence of enemy rooks, rocket programme, announcing the
can threaten (or pretend to threaten) ejection of successive parts of a rocket!)
to penetrate along the open queenside 29. ltJb3 lt:)g5 30. '1tg2 h5!
file(s). On the kingside, exploiting the Black starts on stage three.
advance g2-g4 (which is almost bound 31 .h4 ttJf7 32.gxh5 gxh5 33. '1th2
to be played), prepare the counterblows 'it'c8 34. 1i'g2 Wf8 35. ttJd2 f5!?
... h7-h5 and .. .f6-f5; 36.exf5 bh4
213
Clearly, he is talking about the varia

tion 41 . �xh5 ti:lxhS 42.'ifxh5 �g4
43.'iVg6 Vi'fs 44.'ifxf5 �xfS . It is a pity
that Lasker does not show concretely
how Black can turn the dS-pawn into 'a
convenient object of attack'. Maybe by
transferring the knight via f7-h8-g6-e7?
This takes time, and White has pre
pared a drawing scheme: 45 .\t>g2 , then
�g2-f3 , g3-g4 and ti:ld2-e4.
41 ... '1txf7
37.f6 Thus, Black has emerged from the
One of the few moments of the game, complications with an extra pawn, but
when the commentator rightly offers a there is still a long way to go to win the
concrete variation. By means of 3 7. ti:lxhS! game. White only acknowledged defeat
ti:lxhS (37... ti:lxf5 3 8.ti:lf3 !) 38.Vi'g6 ti:lf6 on move 63.
(weaker is 3 8 ... ti:lf4 39. �xf4 exf4 40.f6
�xf6 41 .Vi'xf6) 39.ti:lf3 White could This game was played in round two, and
insure himself against the risk of defeat. already in round four, Thomas repeated
The best Black can do is the position the whole variation. Evidently, he could
after 39 ... �e7! 40.ti:lxh4 Vi'h8 41 .'it>g3 not believe that with such an advantage
ti:lxdS 4L�e4 ti:lf6 43 .f3 , with a prob in space and numerous attacking plans,
able draw. White was in any danger of not win
37 ... .hf6 38. 1i'f3 i.h4 39.i.g6 ning, let alone losing.
Not immediately 39.ti:lxh5 �g4. This time his opponent was the young
39 ... .hg3+ 40.fxg3 Viennese player Ernst Griinfeld. The
Here it is hard to understand why White players repeated Thomas-Rubinstein up
refused 40.'it>xg3 �fS 4L�xh5. One to move 1 8 . Then White deviated with
gets the impression that he had lost the 1 9.'it>h2 (instead of 1 9.'it>h1). This 'nov
thread of the game. By taking with the elty' made not the slightest impression
pawn, he presents Black with the tempo on Griinfeld, who continued to play
move of the bishop to g4 in several var exactly as Rubinstein had done:
iations. 19 .. .f6 20 . .l::tg 1 ti:lf7 2 1 .'iffl �d7 22.�e3
40 ... ..if5 41 . .hf7 .l:f.a8 ! 23 .Vi'g2 'it>h8 24.ti:ld2 .l::tfb 8
«White prefers to play for the attack, rather than 25 .'i:Vf3 'iVc8 26.'iVe2 .l:Ixa1 27. .l::tx a1
regain material equality with 41 . �xh5 - per l::f.a8 28.llxa8 'ifxa8 . The first two stages
fectly correctly, since in the latter case, after the of Black's plan have been carried out.
exchange of the bishop on hS, Black would obtain Then there begun a phase of manoeu
an attack with ... �g4 and .. .''iffs . In the end vering, which lasted some 50 (!) moves.
game, however, the pawn on dS will be a conven Only on move 78 did Black decide to
ient object of attack and the eS-pawn a powerful play .. .f6-f5 . The 'Spanish Boot' oper
weapon» Lasker.
- ated against White, as the exhausted
Let us try to put the second World Thomas soon missed a simple tactical
Champion's comments into notation. blow and resigned:
214
1 4.a4 l:.b8 1 5.axb5 axb5 1 6. tt:lf1

tbe8 1 7.g4 g6 1 8. �h6 tt:lg7 1 9. tbg3
f6 20. 'ifd2 tt:lf7
89 .. .'iYxf2! 90.'iVxf2 tt:lxe4+ 9 l .�e1

tt:lxf2 etc.
Meanwhile, the variation continued to 21 . �h2 tbxh6 22. 'ifxh6 l:.f7

develop and Rubinstein refined his plan. Only here does Black step aside from
For theory, important games are his two the general plan of exchanging all the
against Bogoljubow (Breslau 1925 and rooks on the a-file.
Berlin 1926). At the same time, Akiba 23. 'ifd2 �h8 24. tt:lg1 f5! ? 25.f3
switched his attention to other variations �h4
of the Spanish (see the games against Soon Rudolf Spielmann won.
Thomas and Monticelli above). All this
naturally led to the whole plan involving The years passed. White players refined
the knight marches to f7 and g7 to be their move-orders, honing their plans.
regarded as authored by Rubinstein. For example, the manoeuvre �e3-a7!?
But we should restore historical jus was found, so as not to permit Black to
tice. Back at the beginning of the 20th exchange all the rooks along the a-file.
century, a game was played, in which But the bishop can only come to a7 if
the ideas of the variation were demon it is developed to e3, rather than d2 or
strated very clearly: h6. Thus the original idea was replaced
by an accurate move-order in one line ..
Curt von Bardeleben Gradually, the Rubinstein system came
Rudolf Spielmann to be regarded as difficult for Black.
Dusseldorf 1 908 Maybe it really is, I will not argue. But
the variation has its merits, the main
one being that it is very good against
players who are not well versed in
positional manoeuvering. Such players
usually lose their way in long-winded
battles. They themselves do not have a
clear plan and do not know which plan
their opponent intends.
In this respect, I will permit myself to
draw attention to the game Hussein-S.
215
Guliev (Teheran 1 992). The game can question of what Black did wrong, so
be found in any computer database and that you do not repeat his mistakes in
the reader can familiarise himself with your own games.
it, if he wishes. Black played in the clas Finally, it is important to know typical
sical way. The game was long, lasting combinations. And here it is quite unim
62 moves, and was hard. But the stra portant whether the players' names are
tegical outline was simple enough: the famous ones. All that is important is
manoeuvres ... g7-g6, ... ti:Jf6-e8-g7; .. .f7- how typical the sacrifice is. If you plan
f6 , ... ti:Jd8-f7; ... �c8-d7; the exchange to include the defence in your reper
of all four rooks; new manoeuvres; the toire, then you need to know how the
penetration by the queen on the open opponent plans to attack your king.
files; new exchanges; transition into The most typical sacrifice in this
an endgame with an extra pawn and variation (indeed, in the whole classical
finally, victory. All as in Thomas-Rubin Spanish) is the knight sacrifice on
stein. fS .
To refuse such a variation is like killing Boris Vlad i m i rov

the goose that lays the golden eggs. And Alexander Ryskin
if the reader is interested in the Rubin USSR 1 974
stein Variation and wishes to include
it in his repertoire, then I will offer
0
several pieces of advice.
Firstly, it is useful to study purposefully
the games of top players who play the
variation. For White, that means Karpov
(games against Spassky, Moscow 1 973,
and Unzicker, Nice 1 974) and for Black,
Smyslov (games against Schmid, Havana
1 967, and Geller, Palma de Mallorca
1970). By analysing these games, you
will learn to see the positions and prob 1 4. tt:Jf1 tt:Jea 1 S.g4 gs 1 6. tt:Jga tt:Jg7
lems facing each player, through the 1 7. cJi'h2 f6 1 8 . .l:.g1 tt::lf7 1 9 . ..td2 cJi'h8
eyes of world champions. 20. 'ife2 ..td7 21 . .l:.g2 tt:Jgs 22 . .l:.ag1
Secondly, it is also handy to examine tt:Jxf3+ 23. 'ifxf3 .l:.f7 24.h4 .l:.af8
games in which White achieves his aims
by purely positional means. Thus, in
the game Keres-Vidmar (Bad Nauheim
1 936), White managed to break through
his opponent's defences on the kingside,
without the aid of sacrifices. And in the
game Stein-Ivkov (Amsterdam 1 964) ,
White broke through o n the queenside
with enviable ease. In analysing these
games, you should try to answer the
216
25. tlJf5!? gxf5 26.gxf5 l:lg8 27.h5 forced win: 32.hxg7+ �xg7 (32 .. Jhg7
By sacrificing the knight, White opened 3 3 . �h6 I;lxg2+ 34.'ii'xg2 with the
the g-file for his major pieces. It is not threat of mate on g8 or g7) 33. �xe8
mate, nor even a forced win, but his �xe8 34.�g4, threatening 3S .'i¥xh7+!
activity promises a lasting initiative. �xh7 36 . .l:i.h4+ �h6 37. l::rx h6#, whilst
White will most likely regain the piece. after 34 ...�f8 , the move 3S.I!h4 wins
For example, after 27 .. J�ff8 28.�h6 without any beauty.
�f7 29 . .l:lg3 �d8 30.�g2 �f8 3 1 . �e3 32 ... 'ifxe8 33.l:lg4 'iff7 34. l:l1 g3
and then h5 -h6. Black shows his 'ife8 35. �h1 l:lc7 36. �g2 aS
willingness to part with his extra mate 37. 'Wh4 'ifd8 38. �3 b4 39.l:lg6
rial immediately: l:lf7
27 ... �f8 28.h6 �e8 The last few moves were obviously
But White does not wish to take back played just to get past the time con
the knight: trol in one piece. We would point
29. 'Wh3 out that the rook cannot be taken:
Such indifference to material is also 39 ... hxg6 40.fxg6! (but not 40.hxg7+
typical of this variation. After 29.hxg7+ �xg7 41 .l::i.xg6+ �f7 42 .�h7+ �g7) ,
l::rfxg7 White would have to exchange and Black can only meet the threat of
rooks and would have nothing left with mate by returning all his extra mate
which to attack. rial: 40 ... tt:Jfs 41 .exf5 Ilh7 (the threat
29 ... l:ld7 30. �d1 'Wd8 31 . ..th5 was 42 .g7+ �h7 43 .�h5) 42 .gxh7
3 1 .f4!? was worth considering, but �xh7. In this position, White wins
White hopes to break the defence with with 43 .�xg8 �xg8 44.�h5 and
out opening additional diagonals. 4S .�g6(+).
31 ... 'fle7 40. 'iti>e2 'ife8
The time control is passed. White can
quietly calculate variations and find the
forced win.
32. be8
Evidently both players were in
time-trouble at this moment (also typ
ical for this variation - a complicated
manoeuvering battle takes up a lot of 41 .hxg7+
thinking time) and from this results the It is striking that White finally decides
inevitable mistakes. to take back the knight only after I 5
A move earlier, Black had to exchange moves of manoeuvering battle.
on hS. Now, however, White misses a 41 ... l:lgxg7
217
On 41 ..Jlfxg7, White mates by means ever, let us not be too hasty in con
of 42 .Vi'xh7+ .l::i.x h7 43 Jhg8# or demning Black's pawn thrust.
42 ...�xh7 43 . .l::f.h 6#, and after 4l... itxg7,
by means of 42 .�xh7+ �xh7 43 . .l::th 3+
ith6 44J�hxh6#. The text move also
does not save the game:
42 . ..th6 bxc3 43. :Xg7 ..txg7
44 . ..txg7+ :Xg7 45. 'ifxf6
Black resigned, since the pawn ending
after 4S ...Vi'f7 46.Vi'xg7+ Vi'xg7 47. .l::r.xg7
�xg7 48 .bxc3 is completely hopeless
for him.
It is also extremely useful to know how 23.fxe5 dxe5

dangerous it can be for Black to play the But this was definitely weakly played.
freeing break .. .f7-fS (instead of .. .f7-f6 One senses that, although the black
«a Ia Rubinstein») prematurely. Although player is a very strong, titled player, he
dangerous does not necessarily mean is not entirely familiar with all the sub
bad. tleties of this pawn structure.
There is no question that the only cor
John N u n n rect move is 23 .. .f4!.
N igel Short White has four possibilities in reply:
Brussels 1 986 24.exd6 Vi'b6+ 25 .\it>hl itxg3 26.eS;
24. itxf4 dxeS 2 S . itxeS 'lli' b 6+ 26. itd4;
24.e6, and finally, 24.tbfS !?.
0
The game could take on a rather irra
tional character, but in all cases, Black
has his chances.
24.exf5!
The correct reaction. White tries to give
his opponent the so-called 'hanging
pawns'. Additional factors, also favour
ing White, include the insecurity of the
black king and the possibility of the
1 4. tt::lf 1 tt::le8 1 5.a4 l:b8 1 6.axb5 white rook penetrating down the a-file.
axb5 1 7.b4 c4 1 8. tt::lg 3 g6 1 9. tt::lh 2 All variations after 24 ... gxfS 2S .tbg4!
tt::lg 7 20 . .l:.f1 !? ..td7 21 .f4 'iit>h8 (2S ...tbf7 26.tbxfS) 26.ith6! are
For readers who have already seen in White's favour. For example, 26 .. .f4
the games Karpov-Unzicker, Karpov 27. itxg7+ '1t>xg7 28 .'iVe4 l:f.h8 29.tbhS+
Gligoric and Stein-Matanovic, this plan 'iit>f8 30.tbxeS with a crush.
is not a novelty. Also bad is 24 ... tbxfS 2S.tbxfS gxfS
21 ... ..th4 22. 'iff3 f5! ? 26.�hS .
More usual would b e ... f7-f6, either at Black tries, by means o f a n intermedi
once or after the exchange on g3 . How- ate exchange, to depart from the road
218
down which White is leading him, but Perhaps the most instructive example of
in vain: this is a game played almost 90 years
24 ... .hg3 25. 'i!fxg3 llJxfS ago.
Now not either 25 ... gxf5 26 . .th6, or
25 ... �xf5 26.tbg4! and if 26 ... .txc2, Frederick Yates
then 27.tbh6+ \itlh8 28 . .l:!.xf8#. Lajos Asztalos
26. 'iff2 ! London 1 927
A move with many plans. From here,
the queen is looking at a7 and f8 at once.
26 ... lLlb7 27. lLlg4 hS 28. Aa6!
The decisive breakthrough.
28 ... hxg4
The sixth rank cannot be closed:
28 ... tbbd6 29.tbxe5 .
29. :Xg6+ lLlg7
1 4. lLlf1 lLle8 1 5.g4 g6 1 6.lLlg3 lLlg7

1 7.rJi'h2 f6 1 8 . ..te3 lLlf7 1 9. Ag1
..td7 20. l:g2 rJi'hB 21 . 'i!fe2 l:g8
22.lLld2 'ifca 23. Aag1 as 24.f3 b4
25.c4 'i!Vb7 26. rJi'h1 Aaf8 27.h4
'i!fc8 28.h5 gS
30. :Xg7+!
Black resigned, without waiting for
3 0 ... �xg7 3 1 . .th6+! mating in sev
eral ways: 3 1 ...\itlxh6 32 .'iVh4+ �g7
3 3 .�h7# or 3 1 ...\itlg8 32 .'iVxf8+ l::rxf8
3 3 Jhf8#.
Finally (if you have seriously decided

to include the Rubinstein Variation in
your opening repertoire), it is extremely
important to understand the degree of Black only needs to play ... a5-a4, and
solidity in Black's kingside position. the fortress is ready. White can continue
Quite often we get so-called 'fortress' manoeuvering and prepare some sort of
positions. Black does not equalise, but sacrifice, but with accurate defence, all
sets up the barricades on the second these attempts should be in vain. The
and third ranks, and waits for the draw. position is a draw.
White stands better, but cannot break 29.h6!?
through the fortress, unless Black inad A good chance. Although after
vertently helps him. 29 ... tbxh6 3 0 . .l:lh2 tbf7 3 1 .\itlg2 h6
219
3 2 . l:Ighl \th7 nothing decisive is evi

dent, why should Black give the white
major pieces an open file? After all,
Black's extra pawn is never in any
danger of being realisable.
The only shame is that these considera
tions were discarded after a few moves.
29 ... tbe8 30. ttJf5 �d8 31 . l:[h2 .l:.g6
32. l:[g3 ..bf5
Black decides to grab a useless extra
pawn after all. Why? After 32 ... a4
33.tt::Jf l �a6 , all he needs to show is Maybe it was at this point that Black
patience and calmness. began to regret gorging himself on the
Now, however, the battle flares up obviously toxic h-pawn.
anew. 48 ... h6
33.exf5 :Xh6 The threat was 49.�xh7+! .
He could decline the Greek Gift with 49. �f7 �f8
33 ... .l:tgg8 , but even so, the position after The last chance of saving himself was
34. �a4 tt::Jc 7 3S.tt::Je 4 is not what it was. 49 ... l::!.h 7. The analysis of the endgame
The threat of a sacrifice on the squares after SO.�xg8 (weaker is SO.�g6 �g7
cS or d6 (after appropriate preparation) S l. .l:ixh6+ tt::Jxh6 52 .�xh6+ .l:;lh7, and
is a real one. he has to agree to a repetition of moves)
34. :Xh6 ttJxh6 35. 'ifh2 ttJg8 SO ...�xg8 S l .�g6 �f7 52.�xf7 .i:txf7
36. l:[h3 l:[f7 37 . ..b4! �e7 38. tbe4 does not give a clear answer to the ques
White's minor pieces are ideally placed. tion of whether White has a forced win.
Now he has to find a way to break Play could continue 53.�xc5 !? dxcS
through. 54. l:r.xh6+ �g7 (not 54 ... �g8 SS .d6
A tempting idea is a2-a3 . But, before �f8 because of 56.tt::Jxf6+ �g7
starting on this plan, White tries 57. .l:Ig6+ and 58.l:!.g8#) SS . .l:tg6+ \tf8
another idea, which is almost a trap. He 56.d6 axb3 5 7.axb3 tt::Jd 7 or 55 ... \th7
transfers the queen to hS, threatening 56.d6 �d8 57.tt::Jxc5 axb3 58.axb3 tt::Jd 7
in some variations to bring his bishop (58 ... tt::Jc 8 59.tt::Je6) 59.tt::Jxd7 (59.tt::Je 6
via c4 and e8, to the square g6! �b6) 59 ... l::txd7 60.c5 .l:ta7 6 l .c6 l::ta 2+
Unexpectedly, this idea works, but only 6H�;>g3 l:f.c2 63 .c7 �xc7 64.dxc7 llxc7
because Black manoeuvres his pieces 6S . .l::txf6 l:!.c3 66.�e6 l::txb3 67. l:!.xe5
extremely badly. �g7, and Black is still alive.
38 ... 'ifd8 39. <i�Jlg1 l:iJc7 40. 'iff2 ttJa8 Now, however, it is all over:
41 . l:[h1 ttJb6 42. �c6 �f8 43.b3 l:[g7 50 . ..bg8 :Xg8 51 . 'iff7 �g7
44. 'ifh2 �e7 45. 'ifh5 a4 46. 'ittg 2 52. 'ifg6 l:[f8 53 . ..bg5!
'it'b8 47. �e8 'ifd8 48. �g6! Black resigned. He can only meet the
The black position is deteriorating mate threat after 53 .. .fxg5 54.tt::Jxg5 by
quickly. giving up the queen.
220
Chapter 24
I ndirect Borrowings
Borrowing - it is a very subtle and delicate topic. Who first developed this or that
idea? After whom should an opening be named? Chess historians argue themselves
hoarse. Sometimes, such disputes become rather dramatic.
Borrowing is not always direct; sometimes it is more indirect. After all, players
do not only analyse concrete variations. They also analyse pawn structures and piece
set-ups. It sometimes happens that a set-up which works well in one pawn forma
tion is transferred to another. Or a structure which has done well in one opening
is transplanted to another.
Here is a typical example. In the English the French Defence, White has his task,
Opening, after the moves and in the English, he has a different
1 .c4 tbf6 2.tbc3 e6 3.tbf3 b6 4.e4 one. In this concrete position above, the
i.b7 e4-pawn is hanging at move 4. Boleslavsky
... back in the 1 970s, the following idea long ago pointed out that S.eS is not good
became popular: because of S .. .t2Je4. The move S.d3 could
5.'ife2!? easily be one that the white player would
prefer not to play. Black replies S ... d6 (but
not S ... dS 6.cxd5 exdS 7.e5 t2Jfd7 8.d4),
and sets up a Hedgehog with ... �f8-e7,
... 0-0, ... c7-c5. The advance of the pawn
to d4 cannot really be dispensed with
forever, in which case White will end up
having lost a tempo.
Thus was the idea of S.'i:!Ve2!? born.
White defends the pawn on e4 and sets
up the threat of e4-e5, since now the
One of the first games with this was black knight does not have the square
Stein-Smyslov (Moscow 1 972). In their e4. Slightly later, Lev Polugaevsky and
book of Stein's best games, Gufeld and Oleg Romanishin, independently of
Lazarev give the following note to the each other, began developing a differ
diagram position: ent idea: S.�d3 !?. A new structure was
«One is struck by the similarity between Stein's born, where the pawns go to e4 and
idea and Chigorin's famous invention against the c4 immediately, and the d4-pawn only
French Defence: l .e4 e6 2.'ii'e2!?. Stein's play later, after the bishop has retreated to c2 .
often betrays the influence of the Russian school of The appeal to 'the influence of the Rus
chess, from Chigorin to Alekhine.» sian school' in the above quote is also
Speaking honestly, there is not such an questionable, if not downright out of
obvious similarity in the ideas of the place. When one faces a specific prob
above position and the Chigorin line. In lem, there is no time to refer to tradi-
22 1
tions. On the other hand ... maybe Stein, 7.h4!? and many other lines typical of
in coming up with S.�e2 , really did the French Defence.
think of Chigorin's line, from almost I am joking, of course. The real issue
a hundred years earlier and in a com is not traditions, but the similarity of
pletely different opening. Who knows?! general ideas between one opening and
Let us leave this digression to the pundits another. If Black parts with his king's
and return to the game Stein-Smyslov. It bishop at an early stage, and White has
developed in a very interesting way and not yet castled, then the h2-h4-h5 -h6
it would be a shame to break off from march becomes relevant. It aims to
it so soon. underline (after either ... gxh6 or ... g7-
5 ... .tb4 g6) the weakness of the dark squares on
The alternative was S ... cS 6.e5 tLlg8 the black kingside and the fact that their
7.d4. In the 3rd game of the Kortchnoi main defender has been exchanged.
Petrosian match (Odessa 1 974), there This is what White's play is directed at.
followed 7... �xf3 ! ? 8.'iVxf3 tLlc6 9.d5 This is his plan. For this, he does not
tLlxeS 1 O.�g3 d6 l l .�f4 tLlg6 1 2.dxe6 mind spending tempi, nor (as a result)
fxe6 1 3 .0-0-0 tLlxf4 14.'iVxf4. White falling behind in development, or even
obtained substantial compensation for a breakthrough in the centre. The play
his small material deficit. on the kingside dark squares will make
6.e5 lbg8 7.d4 d6 up for all of these inconveniences.
In the Candidates' final match Karpov 1 o ... lbd7 1 1 .h5 ..bf3 1 2. 'ifxf3 dxe5
Kortchnoi (Moscow 1 974), Karpov Smyslov picks up the gauntlet. How
twice played 7... tLle7 with the idea of ever, he did not have a great deal of
... d7-d5 , rather than ... d7-d6. In one choice. Sharp play with 1 2 ...h6 1 3 .'iVg4
game, there followed 8 .�d3 dS, and in lLlfS leads to an unpleasant transition
the other 8.�d2 0-0 9.0-0-0 dS. into the endgame after 14. �d3 dxeS
8.a3 ..bc3+ 9.bxc3 �7 1 0.h4!? l S . �xfS exfS 16.'ifxg7 (less is promised
by 1 6.'ifxf5 0-0 1 7.0-0 .l::re 8) 16 ... �f6
1 7.�xf6 lLlxf6 1 8 .dxe5 tLle4, in which
White has an extra pawn, albeit a weak
one. But he also has excellent prospects,
with his bishop, which dominates the
knight, and play on both flanks.
1 3.h6
If one wishes to talk about traditions

and associations between the English
Opening and the French Defence, then
this is the time - one only has to think
of the variation l .e4 e6 2 .d4 dS 3 .tLlc3
�b4 4.e5 cS S .a3 �xc3+ 6.bxc3 tLle7
222
Chapter 24 - Indirect Borrowings
White is close to his aim. After 1 3 ... g6 ting in the bishop on f1 , was tried by
14.dxe5, both 14 . ..lbxe5 1 5 .'ii'f6 , and Chigorin against the French Defence.
14 . ..tbf5 1 5 .g4 tt::lx eS 1 6 .'ii'e4 are bad. Stein took it (deliberately or not, we
However, there is a third possibility: will let the historians dispute) and used
14 ... c6!? 1 5 . �g5 'ii'c 7, although in this it in the English Opening, successfully,
case too, it is hard to believe that White as we have seen.
will not find some way to develop his
initiative. Now we will examine another exam
1 3 ... gxh6 1 4. hh6 ple, also taken from the play of the
14.d5 !? deserves attention. three-time USSR Champion.
1 4 ... exd4
Allowing the bishop into g7. Black played Leonid Stein
more strongly in Ikonnikov-Meier Ervin Haag
(Hockenheim 2006) : 14 ... tt::lf5 . A tac Tallinn 1 969
tical whirlwind ensued: 1 5 . �g5 f6 1 .e4 c6 2.d3 d5 3.tt:'ld2 g6 4.g3 Ag7
16.d5 fxgS 1 7.dxe6 0-0 1 8 . .l::!.d l tt::ld6 s. Ag2 es 6. tt:'lgf3 tt:'le7 7.0-0 o-o
19.'ii' h 5 tt::lf6 20.'ii'xg5+ �h8 2 1 . l:ld3 8.b4!?
'ii'e 7 22.�xh7+ 'ii'x h7 23 . .ld.h3 'ii'xh3
24.gxh3 l:i.ae8, from which White had
certainly not emerged the winner; two
rooks and a knight are more than suffi
cient compensation for the queen.
1 5 . .i.g7 l:l.g8 1 6. lbh7 tt:'lf5 1 7. hd4
cs
Grandmaster Stein played in a style

which can, with every justification, be
called modern. He created problems
from the first move, forcing his oppo
nent to think for himself. He loved
non-standard positions, in which the
standard orientation points were miss
1 8.g4!? ing. In such positions, there is perhaps
Stein does not relax, but keeps on ratch less harmony, but in return, there is
eting up the tension. Soon, Black could lively piece play.
not withstand the pressure, committed He could accept pawn weaknesses in his
the decisive mistake and had to admit own camp, or personally weaken his
defeat. own pawn chain, if in return he got the
Here is an example of indirect copying initiative. Against the Caro-Kann and
in the opening. The piece set-up with King's Indian Defences, Stein invented
the queen on e2, temporarily shut- a system which was far removed from
223
the classical standards. This variation place in this line is on c7, from where
became one of his calling cards. she defends the eS-pawn.
The idea of an early b2-b4 is not new. Possible variations: 1 0 . .ib2 d4 1 l .a4
It has been used successfully in the 'ti'c7 1 2 .c3 dxc3 1 3 . .ixc3 l:!.d8 14.tLlc4
King's Indian and Griinfeld Defences; (Stein-Hort, Los Angeles 1 968) or
in a word, in openings where the black 1 0 ...'ifc7 1 1 .'iYe2 d4 1 2 .c3 cS 1 3 .cxd4
bishop on g7 invites the white bishop to cxd4 14.a4 (Stein-Khodos, Kiev 1969). In
come to b2 . both games, White had no complaints
What is the advantage of b2-b4 com about the outcome of the opening.
pared with the usual b2-b3 ? If White 1 0. �b2 "ffc7 1 1 . 'ffe2
follows up with a later pawn advance The immediate 1 1 .a4 has also been seen,
on the queenside, the move b2-b4 saves and then: 1 1 ...h6 1 2 . l:!.e 1 d4 1 3 .c3 cS
a tempo. In addition, in variation lines 14.l::i.c l dxc3 1S . .ixc3 �a6 1 6 .tLlb3 �c6
of the King's Indian structure, the pawn 1 7.aS . White had seized the initiative
on b4 stops the advance ... c7-cS . in Stein-Coho Arteaga, Havana 1 968,
Now let us return to the diagram posi whilst the a2-pawn, which by all rights
tion. It is unlikely that Black plans an should be a weakness, had advanced to
early ... c6-cS , or that White plans an aS and ties down the enemy forces.
early queenside pawn advance. So the
advantages of an early b2-b4 are not to
be seen. But the most unpleasant thing
of all is that in reply to b2-b4, there can
follow ... a7-aS , and now White does not
have either the reply b4-bS , or a2-a3
(the rook on a1 hangs). It follows that
the only possible reaction to ... a7-aS is
the exchange on aS . But this means that
White voluntarily breaks up his pawn
chain, agreeing to (and even initiating)
the creation of an isolated pawn on a2. 1 1 ... d4
For what?! The threat was to take on dS, with an
We hope that this analysis of Stein's attack on eS. After 1 1 ...f6 the reply 1 2 .d4
games will partly answer this question. is unpleasant, whilst after 1 1 ...tLld7 there
8 .. 85
. follows 1 2 . Ilfe1 , and Black again has to
Of course! decide how to defend the eS-pawn.
It is not principled to play 8 ... tLla6 The simplest solution is to put paid to
9.a3 tLlc7. After 1 0 . .ib2 d4 1 l .c3 .ig4 the threat by exchanging on e4 oneself,
1 2 .'iYc2 tLle6 1 3 .cxd4 .ixf3 14.tLlxf3 but after 1 1 ...dxe4 1 2 .dxe4 White has
tLlxd4 1 S .tLlxd4 exd4 1 6.f4 (Stein lively piece play.
Hartoch, Amsterdam 1 969) White has 1 2.c3 dxc3
a noticeable advantage. Conceding the centre. After 1 2 ... cS, as
9.bxa5 :XaS we have seen in the examples quoted
More natural is 9...'ifxaS , although the above, White chooses between 1 3 . l::rfc1
position is almost the same. The queen's and the immediate 1 3 .cxd4 cxd4 14.a4.
224
Chapter 2 4 - Indirect Borrowings
1 3 . .bc3 :&4 1 4. llJc4 all is that he never once had to com

The plan is clear. By attacking the pawn plain about the breaking up of his pawn
at eS, White will sooner or later force structure, undertaken on move 8 !
the move .. .f7-f6, after which he will
advance d3-d4. Let u s see how the same idea works in
Stein has won the opening phase. the Leningrad Variation of the Dutch
Realising this, his opponent initiates Defence.
tactical complications, so as to change
the course of the game in his favour. Evgeny Bareev
1 4 ... b5! ? Vlad imir Malaniuk
Groningen 1 997
1 .d4 f5 2.g3 tlJf6 3 . ..tg2 g6 4. tlJf3
..ig7 5.0-0 0-0 6.b4!?
Not I S .tt:JcxeS because of 1 5 .. .f6, and

the knight is trapped in the middle of
the board. But White finds a way to
strengthen his position still further. Far from the most popular continuation
This is not surprising - in a tactical (usually, White plays 6.c4 or 6.b3), but
battle, Stein always felt like a fish in perfectly plausible.
water. Of course, this is not the Stein idea. It
1 5. 'ifc2! is really nameless, although someone
The knight is defended and the threat must have played it first. Maybe histori
to eS renewed: 1 6.ttJcxe5 f6 1 7.�b3+. ans will establish who. But does it really
Black should have maintained the ten matter?
sion with 1 5 ... ttJd7, but he prefers to A practical player, during a game, is
resolve matters: only interested in three things: the
1 5 ... ..te6 pawn structure, the piece set-up and the
There followed: concrete variations. All other questions
1 6.llJcxe5 f6 1 7.llJc4 .bc4 1 8.dxc4 are academic, historical, and, above all,
:Xc4 1 9.ltJd2 unnecessary.
White has emerged from the complica White plays b2-b4 (and not b2-b3)
tions with the two bishops and a cen because he wants to mount a queenside
tral preponderance. Black's pieces are attack with his pawns. In some varia
rather scattered and his king position tions, he also insures himself against
weakened. White won in just ten more ... c7-c5, and after 6 ... a5 , he has the reply
moves. The most surprising thing of 7.b5 .
225
6 ... tZ:le4 he does not object t o exchanges. Or,

It was worth considering 6 ... 4Jc6 7.a3 to put it another way, White uses his
dS!? changing the structure. In the game activity to induce his opponent to make
Anand-Nakamura (Wijk aan Zee 2 0 1 0) exchanges. Then, as the board empties,
there followed 8 . i.b2 ltJe4 9.4Jbd2 the advantage in pawn structure looms
i.e6 1 0.e3 aS l l .bS 4Ja7 1 2.'ife2 a4 more clearly.
1 3 . .l::tfcl c6 14.bxc6 bxc6 1 S .c4 tbc8
1 6 . i.c3 4Jcd6. Black managed to hold The outcome of the game was decided
the balance. deep in the endgame.
7.c4 e6 a . ..tb2 b6 9.tlJc3 ..tb7
1 0 . .:c1 aS
D
Grandmaster Malaniuk is one of the main
specialists in the Leningrad Variation,
but here he plays passively, following in
the footsteps of others' plans and ideas.
Instead of the move in the game, it was
worth considering 10 ...4Jc6 l l .bS 4Je7,
then ... h7-h6, ... g6-g5 and ... 4Je7-g6.
1 1 .b5 d6 1 2. tZ:le1 ttJxc3 13 . .bc3
.bg2 1 4. ttJxg2 0KJ7 1 5.tl:\f4 'ife7
1 6.d5 tlJcS 1 7 .dxe6 ttJxe6 1 8 . .bg7 Thus far, White has not made any
'1t>xg7 1 9. 0KJ5 changes to the pawn structure, fearing
piece counterplay, but now there are
hardly any pieces left on the board. It
is time!
29.f4!
Of course, not 29.e4 because of 29 .. .f4!.
29... �7
The pawn ending after 29... �e6 30.e4
l::txe4 3 1 . l:Ixe4 fxe4 32 .�f2 \t>f7 33 .�e3
�e6 34.�xe4 is hopeless for Black.
More tenacious is 30 .. .fxe4 3 1 .�f2
�f7 32 .�e3 �f6, but after 33 .g4 I;le8
The position has clarified. White has an 34.�d4 hS 3S .h3 hxg4 36.hxg4 White
advantage that is not great, but is stable. also has winning chances.
He has the better structure, with his 30. '1t>f2 '1t>e6 31 .e4 '1t>d7 32. '1t>e3 l:[f7
three queenside pawns holding the four 33. '1t>d4 fxe4 34 . .:Xe4 h5 35.a4 .:ta
black pawns. If he manages to advance 36.h3 l:[f7 37.g4 hxg4 38.hxg4 l:[h7
e3 -e4 and exchange the e- and f-pawns, It is impossible to stop the advance of
White will practically have an extra the f-pawn in any event. For example,
pawn on the kingside, which could 38 ... .l:tf8 39.�d5 Ilf6 (39 .. Jlf7 40. Ile6)
become passed. 40.f5 gxfS 41..l::f.f4 etc.
The peculiarity of this structure is that, 39.f5 gxf5 40.gxf5 l:[f7 41 . l:[f4 '1t>e7
although White is the more active side, 42. '1t>d5 '1t>f6 43. l:[f1 l:[e7 44. '1t>c6
226
Chapter 24 - Indirect Borrowings
The advantage has grown notably. In 1 8 ... tbas 1 9. .l::i.xc7 'iYd8 20 . .l:i.cs d6.
the end, after a series of manoeuvres Even three pawns could prove insuf
(mostly undertaken to gain time before ficient compensation for the piece.
the time control) , White sacrificed the Meanwhile, if 1 6 .e4, then Black has a
passed fS-pawn, activated his rook and pleasant choice between 1 6 .. .f4 1 7.eS
won the pawn on c7 and then the one l:f.h6 1 8 .'iVf3 'iYgs and 16 ... eS!? (which
on b6. Faced with the threat of a new is probably even stronger) 1 7.dS fxe4
queen appearing, Black resigned on 1 8. �xe4 .l:i.h6 1 9.'iYe3 tbas .
move 6 1 . 1 6.f4!
Stopping both .. .fS -f4 and ... e6-eS .
The blockading move f2-f4 (before 1 6 ... tba5 1 7. 'iff3 d6 1 8. l:e1 'ild7
playing e2-e4) brings to mind the fol 1 9.e4 fxe4 20. 'ifxe4
lowing classic example.

Savielly Tartakower
New York 1 924
White has managed to open the posi

tion and thereby improve his chances.
Just as in any other sphere of human

act1v1ty, the process of acquiring
In a closed or semi-closed position, the knowledge in chess is objective and
knight could prove stronger than the individual. That which is invented by
bishop. White needs to open the posi one person becomes the property of
tion. For this, he has two ideas: c4-cS many. Ideas, refined through practice,
or e3 -e4. lose the personality of their inventor.
But on 1 6 .cS bxcS 1 7. .l::rb 7 White must One should recognise this and kneel
reckon with 1 7... c4! ?, and if 1 8 . �e2, before all those who contribute to the
then 18 ... dS, whilst if 1 8 . �xc4, then development of chess art.
227
Afterword
From Ecclesiastes to Eliskases

Finally, we will speak a little about the few, but nonetheless real practical examples
of so-called, 'knowing too much'.
Of course, one cannot say that knowledge is a bad thing. That would be absurd.
But sometimes it has a negative psychological effect, as it sometimes interferes with
one's thinking. The eyes remain closed to a possibility, the brain seizes up. Knowing
that in similar positions (especially in the endgame) such and such a great player
lost, a player can say to himself, 'Well, if even X could not save the game, then what
chance do I have ? ' . But then it turns out that the position is very slightly different,
and has its own nuances, and it is necessary to look not at the source game, but to
think with one's own head.
Here is a textbook example of this theme.
64.'iii>e 3 (after 64.'iii>d3 :h3+ 6S.'iii>d2

Emanuel Lasker :h8 ! - but not 6S ... .:Xb3 66.:£1 ! -
Aki ba Rubi nstein White still has to play 66.'ite3 , return-
St Petersburg 1 91 4 ing to the main variation) 64... :h3+
6S.'iii>f4, winning in all variations, as
• see:
6S .. Jlxb3 66.cj_)g4 .l:[c3 67.f7 .l:[c8
68.f8'ir+ .:Xf8 69. .:Xf8 'iii>c s 70.<Ji>f3 ;
6S ... :h8 66.'iii>gS d4 67.f7 :£8 68.'iii>f6
<Ji>ds 69.�g7 .:as 70.fs'it' .:Xfs n.:Xfs
d3 n . .:bs;
6S ... 'iii>e 6 66.f7! :h8 (66 ... �xf7
67.'iii> g4+) 67.'iii>gS :£8 68.'iii> g 6 d4
69.'iii> g7 .l:[b8 70.f8'ir .:Xf8 7 1 ..:Xf8 , etc.
60 ... .l:[f7 61 .f6 �d6 62. �d4 �e6
A rook ending, with equal material and
so few pawns left, would in most cases
be a draw. But in this concrete situation,
other factors matter. Firstly, the position
of the black rook is exceptionally unfor
tunate. Secondly, Black cannot go into a
pawn ending, by exchanging the d- and
f-pawns, as the resulting pawn ending
is lost.
Unfortunately, he cannot cure the first
factor by activating his rook: 60 ... 'iii>c 6
61 .'iii>d4 'iii>d 6 62 .:£2 :h6!? 63.f6 :h4+ 63 . .l:[f2!
228
Afterword - From Ecclesiastes to Eliskases
Preparing the transfer of the rook to a6; similar ending against Rubinstein. There, the only
much weaker is 63 .�c5 l::tc7+ 64.�xb4 difference was the absence of h-pawns.
�f7, and Black starts to have hopes. During the interval, I managed to establish that
63 ... �d6 the absence of these pawns changed the assessment
But now the second factor comes into of the position. I did not find this at once - I was
play - the impossibility of going into a hindered by knowing that Rubinstein himself
pawn ending. After 63 .. Jhf6 64 . .r!xf6+ had not been able to save the position.
�xf6 65 .�xd5 Black loses. I was looking at the board in desperation, when
64. :S2! .l:.c7 65.:S6+ 'iftd7 66 . .l:.b6 the leader of our delegation came into the room.
Black resigned. He knew that I stood badly, but with hope in his
The variation 66 .. J1c3 67. l::txb4 l:lf3 voice, he said:
68.�e5 �c6 69.l:r.f4 l:Ie3+ 70.�d4 - Mikhail Moiseevich, maybe you will find a way
shows that this decision was in no way to save it after all?
premature. Then I looked at the position again, without pre
conceived notions, and understood the secret of this
endgame.»
M i khail Botvi n n i k 41 . 'ifte3 'ifte5 42 . .l:.c2! c3 43. 'iftd3
M a x Euwe .l:.d8+
Groningen 1 946 The move 43 ... .l::rc7, analogous to
Lasker's play, is doubly useless in this
position. Firstly, because Black does not
0
threaten the manoeuvre .. J1c7-h7-h3
(the pawns h5/h4 are in the way), and,
secondly, because White can simply go
into the pawn ending: 44 . .l::f.xc3 ! �xc3+
4S .'i£txc3 'i£txe4 46.'i£tc4 �f4 47.'i£td4
'i£tg4 48.�e5 �xh4 49.'i£tf6 'i£tg4
SO.'i£txg6 h4 S l .'i£tf6 with a draw.
The position is very similar to the pre

vious one, with colours reversed. But
only similar. The addition of an extra
pair of pawns on the h-file deprives
the stronger side of the chance to
manoeuvre his rook around the side
(see 64Jh2 and 65 Ji:a6+ in the above
example) , and on the other, renders the
pawn ending drawn, after the exchange
of c- and e-pawns. 44. 'itte3!
It looks so simple. But listen to Not 44.'i£txc3 �xe4.
Botvinnik: 44 ... .l:.d4 45 . .:.Xc3 .:.Xe4+ 46. 'iftf3
«In this position, I sealed a move. Of course, both .:.Xh4 47 . .l:.c6 .l:.f4+
my opponent and I knew that Lasker had won a Or 47... 'i£tf5 48 . .l::tc 5+ �e6 49J :tc6+.
229
48. �e3 l:.e4+ 49. �3 �5 50.l:.f6+ 60. �g5

�xg5 51 . :Xg6+
Draw.
In the following example, lack of

knowledge of the classics did the Eng
lish GM good service.
Jonathan Speelman
M ichael Trauth
Berlin 1 980
60 ... �d8
0
In the variation 60 ... �dl 6l .�f4 �e2
62.�f3 �xf3 63 .�xf3 �b7 Black, as
usually happens, is one tempo short
of a draw: 64.�f4 �xb6 65 .�g5 �c6
66 .�xh5 �d7 67.<;t>g6 �e7 68 .�g7.
61 . -teB! ..b6 62. bh5 �e7 63 . .tf3
�7 64.h5 .icB 65.h6 �gB 66 . .te4
..ta6 67.�6 .tea 68. �e7 ..b6
69. �d6 � 70. �c7 �6 71 . .td3!
The drawing tendencies of this ending Black resigned.
are obvious. The square h8 is of 'the Great was Speelman's surprise, when
wrong colour', as chess players say, his colleague John Nunn pointed out
and so Black only needs to sacrifice his the existence of a predecessor:
bishop for the b-pawn and get his king
to h8 , in order to draw. In fact, after Erich Eliskases
taking the b-pawn, Black only needs to Jose Rau l Capablanca
get his king to f7 or f8 . Semmering 1 937
Even so, Speelman won, and, at first
glance, did so very easily.
0
52 . .td5
The pawn ending is lost for the time
being, as can easily be seen.
52 ... .tf1 53 . .tc6 .te2 54.b5 �dB
54 ... �xb5 55.�xb5 �d8 is hopeless,
because of 56.�d6, and White can
keep the enemy king from h8: 56 ... �c8
57.�e7 �c7 58.h5 �c8 (58 ... �b6
59.�f7!) 59.�f6 �d8 60.�g6 �e7
6 l . �c4! . 61 . ..b6+ �c6
55. �d6 .i.c4 56.b6 ..b6 57. �e6 A natural move, but at the same time, the
.te2 58. �6 �cB 59. �g6 h5 losing mistake. After 6 1 ...�b8! 62.�b4
230
Afterword - From Ecclesiastes to Eliskases
(6V�r>b5 �fl+) 62 ... �b7! it is impossi «The battle of the bishops for the diagonals is
ble for White to win. After 63 .\t>bS hS! instructive. White drives the enemy bishop from
or 63. �xb7 \t>xb7 64.\t>cs h5! the pawn the fl -h3 diagonal to the c6-a6 diagonal, but the
ending turns out to be drawn, whilst latter has too few squares for it» - Averbakh.
after 63.�e2 �g2 64.\t>cs \t>b7 there 66 ... �f3 67. �f1 �b7 68. �h3 <ite7
is no way to drive the king away from Or 68 ... \t>cs 69. �g4 \t>c4 70. �e2+
the b6-pawn, and the attempt to sacri �c5 7 Lili.a6 �f3 72.�c8 \t>c6 73 .\t>a6,
fice the pawn also fails: 65 .\t>d6 �xb6 winning.
66.�e6 \t>c6 67.\t>f6 \t>d6 68 .\t>g7 \t>e7 69. <itb5 <itd6 70. �g4 <ite7 71 . <itc5
69.\t>xh6 \t>f8 etc. �g2 72.�c8 <itd8 73. h6 �f3
62. �c8! 74. <itd6 �g2 75. �c4 <itc8 76. �d5
Now White is winning, although there �f1
are still difficulties lying ahead. Only now can White begin his march
62 ... �f1 to the kingside.
The threat was 63 .\t>a6. 77. <ite6 �e2 78. <itf6 <itd7 79. <itg6
63. �g4 �d3 h5 80. <itg5 <itd6 81 . �f7 <itc6
Alas, Black is too late after 63 ... \t>b7 82 . .hh5
64. �f3+ \t>b8 65 .\t>b4 �a6 66.\t>cs Black resigned.
�b7 on account of 67.�xb7 \t>xb7
68.h5! Isn't it the case that, after seeing this
64. �f3+ <itd6 65. �b7 �e2 difficult ending, full of nuances, one
wants to go back to the previous exam
ple and look at it afresh? How come
Speelman won so easily? And could he
have done so, had he been burdened
with the knowledge that the endgame
was so full of subtleties and needed
such accurate handling, so as not to fall
into one of the traps?
«For in much wisdom is much grief; and he that
increaseth knowledge increaseth sorrow.»
Ecclesiastes was right.
66. b6! And so was Eliskases.
23 1
BI BLIOG RAPHY
<<Viacheslav Ragozin». Edited M.Yudovich. Moscow, 1 984.
<<Tournament of Stars Montreal 79>>. Edited V.Chepizhny. Moscow, 1 979.
<<Chess: Science, Experience, Mastery» Collection of essays edited B.Zlotnik. Moscow, 1 99 0 .
Alekhine A. <<On the Road t o the World Championship». Minsk, 1 9 8 2 .
Alekhine A. <<New York 1 924». Moscow, 1 9 89.
Anand, V. <<My Best Games of Chess». London, 200 1 .
Averbakh Y. <<Chess Endings: Knight and Bishop». Moscow, 1 9 8 0 .
Averbakh Y . <<Chess Endings: Rook». Moscow, 1 9 84.
Boleslavsky and Bondarevsky: <<World Championship Match Petrosian-Spassky». Moscow, 1 9 7 0 .
Bondarenko F. <<Pawn Endgame Studies». Moscow, 1 9 7 3 .
Botvinnik M. <<Analytical and Critical Works Vol. 1 ». Moscow, 1 9 84.
Botvinnik M. <<Analytical and Critical Works Vol. II». Moscow, 1 98 5 .
Bronstein D. <<Ziirich 1 953». Moscow, 1 9 8 3 .
Chebotarev, 0. <<Lessons in Chess Strategy». Moscow, 1 9 8 1 .
Damsky, Y. <<Grandmaster Polugaevsky». Moscow, 1 9 8 2 .
Dvoretsky, M, Yusupov, A . <<Secrets o f Opening Preparation». Kharkov, 1 9 96.
Dvoretsky, M, Yusupov, A. <<Methods of Chess Study». Kharkov, 1 997.
Fischer, R. <<My 60 Memorable Games». Moscow, 1 9 7 2 .
Gligoric, S <d Play against Pieces». Moscow, 1 9 8 3 .
Globev, A . , Gutzeit, L. <<744 Games o f Bobby Fischer>>. Moscow, 1 9 9 3 .
Golombek, H. <<Richard Reti's Best Games». Moscow, 2002.
Gufeld, E . , Lazarev, E . <<Leonid Stein». Moscow, 1 98 0 .
Hajtun, ]. <<Grandmaster Portisch». Moscow, 1 9 7 7.
Hort, V. , Jansa, V. <<The Best Move». Moscow, 1 9 7 6 .
Ivanchuk, V. <<The Linares Super-tournament». Kiev, 1 9 9 2 .
Karpov, A. <<Selected Games». Moscow, 1 97 8 .
Karpov, A . , Podgaets, M. <<The Caro-Kann Defence, Panov Attack». Moscow, 2006.
Kasparov, G. <<Two Matches». Moscow, 1 987.
Kasparov, G. <<My Great Predecessors. Vols. I-ll». Moscow, 2 0 0 3 .
Kasparov, G. <<My Great Predecessors. Vol. III». Moscow, 2004.
Kasparov, G. <<My Great Predecessors. Vol. IV». Moscow, 2005.
Keres, P. << 100 Games». Moscow, 1 9 6 6 .
Kmoch, H. <<Defence i n Chess». Moscow, 2005.
Koblents, A. <<Lessons in Chess Strategy». Moscow, 1 9 8 3 .
Kotov, A. <<Alekhine's Chess Heritage. Vols I-II». Moscow, 1 9 8 2 .
Kotov, A. <<Think Like a Grandmaster». Moscow, 1 9 8 5 .
Kramnik, V. , Damsky, Y. <<Breakthrough». Moscow, 2 0 0 0 .
Marovic, D. <<Secrets of Positional Chess». London, 2 0 0 3 .
Matsukevich, A. <<The Comet Sultan Khan». Moscow, 2003.
Mednis, E. <<How to Beat Bobby Fischer». Moscow, 1 9 8 1 .
Negyesi, G . <<Champion's Friend, Friendship's Champion: Lilienthal's Hundred Best Games». Kecskemet, 20 0 1 .
Neishtadt, Y. <<The First World Champion». Moscow, 1 97 1 .
Neishtadt, Y. <<Siegbert Tarrasch». Moscow, 1 9 8 3 .
Nesis, G. <<Alexander Khalifman». S t Petersburg, 2 0 0 0 .
Nimzowitsch, A. <<My System». Moscow, 1 9 84.
Panchenko, A. <<Theory and Practice of Chess Endings». Moscow, 2 0 0 6 .
Petrosian, T. <<Strategia Nadyozhnosti». Moscow, 1 9 8 5 .
Razuvaev, Y. , Murakhveri, V. <<Akiba Rubinstein». Moscow, 1 9 80.
Reinfeld F., Soltis, A. <<Morphy Chess Masterpieces». New York, 1 9 74.
Sakaev, K. <<Linares 2002». St Petersburg, 2002.
Slutsky, L . , Shereshevsky, M. <<Contours of the Endgame». Moscow, 1 9 89.
Smyslov, V. «In Search of Harmony». Moscow, 1 979.
Speelman, ]. <<Analysing the Endgame». London, 1 9 8 1 .
Suetin, A. <<Grandmaster Boleslavsky». Moscow, 1 98 1 .
Vainshtein B . <<Myslitel». Moscow, 1 9 8 1 .
Vasiukov E . , Narkevich.A., Nikitin A. <<Mikhail Chigorin». Moscow, 1 9 7 2 .
Verkhovsky L. <<Karl Schlechter>>. Moscow, 1 984.
Zak, V. <<Lasker». Moscow, 1 9 6 3 .
Zak, V . <<Paths t o Self-Improvement». Moscow, 1 9 8 8 .
232
BIOG RAPHY
Sarhan Babash oglu (in Russian: Babashovich) Guliev was born in I 9 6 8 in the
town of Ashagy Molly, Azerbaijan, and today lives in its capital, Baku.
In I 9 8 9 , I 99 I and I 9 9 7 he became the chess champion of Azerbaijan. Before
that, he also won the national junior title three times. In the I 990s, Guliev won
many tournaments and he represented his country in international competitions.
In I 99 S he earned the Grandmaster title.
In I 99 2 , Guliev graduated in engineering and economics. In I 9 9 6 he gradu
ated as a sports instructor at the Russian State University of Physical Culture and
Sport in Moscow. From 2 0 0 0 onwards he has worked as a chess coach, for, among
others, the national teams of Iran, Turkey, and the United Arab Emirates. In 20 I I
he was awarded the title of FIDE Senior Coach, and in 2 0 I 3 he received the title of
Honoured Coach of Azerbaijan.
Guliev has written I I chess books, of which Tutorial Chess Endings was translated
in 8 languages. At this moment he is a teacher at the Azerbaijan State Academy of
Physical Culture and Sports.
233
I ndex of Games
Viswanathan Anand Evgeny Bareev Linares 1 9 93 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Robert Fischer Pal Benko New York 1 9 63 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Akiba Rubinstein Eugene Znosko-Borovsky St Petersburg 1 909 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
William Steinitz Adolf Anderssen Vienna 1 8 7 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Alexander Morozevich Smbat Lputian Dagomys 2 0 0 7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 0
Anatoly Karpov Garry Kasparov Leningrad 1 9 8 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I I
Veselin Topalov Krishnan Sasikiran Sofia 2 0 0 7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 2
Mikhail Botvinnik Viacheslav Ragozin Bolshevo 1 947 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 4
Emanuel Lasker Jose Raul Capablanca Moscow 1 9 3 S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I S
Fedor Duz-Khotimirsky Akiba Rubinstein Lodz l 9 0 7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 7
Alexander Alekhine Frederick Yates Hamburg 1 9 1 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 7
Mark Taimanov Robert Fischer Buenos Aires I 9 6 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 8
Jose Raul Capablanca David Janowski NewYork 1 9 1 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 8
Ruslan Ponomariov Levan Aronian Lausanne 2 00 I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I 9
Anatoly Karpov Ljubomir Ljubojevic Monaco 1 9 96 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 9
Arnaud Hauchard Ruslan Ponomariov Belfort 1 9 98 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 9
Vlastimil Jansa Efim Geller Budapest 1 9 7 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 0
Veselin Topalov Alexander Beliavsky Linares 1 9 9S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1
Veselin Topalov Krishnan Sasikiran Sofia 2 0 0 7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 2
Yuri Averbakh Semen Furman Odessa 1 9 60 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 3
Veselin Topalov Antonio Antunes Candas 1 9 92 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 4
Robert Fischer Tigran Petrosian Buenos Aires 1 9 7 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 6
Jules Arnous de Riviere Paul Morphy Paris 1 8 6 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 7
Veselin Topalov NN Frankfurt (simultaneous) 1 9 9 7 . . . . . . . . . 2 8
Veselin Topalov Michael Adams Dortmund 1 99 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 9
David Bronstein Tigran Petrosian Amsterdam/Leeuwarden 1 9 S 6 . . . . . . . . . . 3 0
Boris Gelfand Veselin Topalov Novgorod 1 9 9 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 1
Jaan Ehlvest Garry Kasparov linares 1 9 9 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 2
Vladimir Krarnnik Sergey Tiviakov Sochi 1 9 90 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 3
Mihail Tal Mikhail Botvinnik Moscow 1 9 60 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 4
David Bronstein Isaak Boleslavsky Moscow 1 9 S O . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 4
Mikhail Botvinnik Efim Geller Moscow 1 9 S 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 S
Alexander Baburin Imad Hakki Yerevan 1 99 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 7
Elmar Magerramov Tony Molina Dubai 2 0 0 7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 9
Siegbert Tarrasch Richard Teichmann S an Sebastian I 9 1 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
JanTimman Alexander Chernin Montpellier I 9 8 S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 1
Richard Reti Frederick Yates New York 1 9 24 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
Florin Gheorghiu Anatoly Karpov Luzern 1 9 8 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
Veselin Topalov Vladimir Krarnnik Wijk aan Zee 1 99 9 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4S
Richard Reti Akiba Rubinstein Karlsbad 1 9 2 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4S
Alexander Alekhine Aaron Nimzowitsch San Remo 1 9 3 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4S
William Winter Alexander Alekhine Nottingham 1 9 3 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
Robert Fischer Boris Spassky Belgrade 1 9 92 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Anatoly Karpov Boris Spassky Montreal 1 9 7 9 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Veselin Topalov Artur Jussupow Dortmund 1 9 9 7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . S O
Mikhail Gurevich Alexei Shirov Sarajevo 2 0 0 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . S I
Jose Raul Capablanca Alexander Alekhine Buenos Aires I 92 7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . S 2
235
Boris Spassky Tigran Petrosian Moscow 1 9 6 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 3

Vladimir Kramnik JanTimman Wijk aan Zee 1 99 9 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 4
Mikhail Gurevich Veselin Topalov Sarajevo 2 0 0 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 5
Alexander Khalifman Stellan Brynell Leningrad 1 9 8 9 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 6
Michael Adams Vladimir Kramnik Linares 1 99 9 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 7
Lajos Portisch Vladimir Bagirov Beverwijk 1 9 65 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 7
Hector Rossetto Mikhail Tal Portoroz 1 9 5 8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 8
Garry Kasparov Vladimir Kramnik London 2000 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 9
Jose Raul Capablanca Alexander Alekhine NewYork 1 9 24 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
Mikhail Botvinnik Salo Flohr Leningrad 1 9 3 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 1
Alexei Shirov Viswanathan Anand Moscow 2 0 0 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 1
Veselin Topalov Mainhard Moroder Frankfurt 1 99 7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 2
Anatoly Karpov Boris Spassky Leningrad 1 9 74 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 3
Emanuel Lasker Harry Nelson Pillsbury Paris 1 9 00 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 4
Alexander Khalifman Thomas Luther Leningrad 1 9 8 9 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 4
Tigran Petrosian Borislav Ivkov Bugojno 1 9 8 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
Tigran Petrosian Florin Gheorghiu Moscow 1 9 6 7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 5
Veselin Topalov Vladimir Epishin Dos Hermanas 1 9 94 . . . . . . . . . . . . . . . . . . . . . . . . . . 6 8
Erich Eliskases Grigory Levenfish Moscow 1 9 3 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 9
Veselin Topalov Claus Henrici Frankfurt 1 9 9 7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
Robert Hubner Miguel Cuellar Leningrad 1 9 7 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
Boris Spassky Bruno Parma Havana 1 9 6 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 1
Attila Groszpeter Andras Adorjan Hungary 1 9 8 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 2
Gyula Breyer Sieg bert Tarrasch Gothenburg 1 9 20 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 4
Richard Reti Vladimir Vukovic Vienna 1 9 2 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 5
Ruslan Ponomariov Vasily Ivanchuk Linares 2 0 0 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 6
Vasily Smyslov Isaak Boleslavsky Leningrad/Moscow 1 94 1 . . . . . . . . . . . . . . . . . . 7 7
Boris Spassky Jan Hein Donner Leiden 1 9 7 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 7
Louis Paulsen William Steinitz Baden-Baden 1 8 7 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 8
Vladimir Krarnnik Veselin Topalov Dortmund 1 99 9 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 1
Vlastimil Jansa Raymond Allen Weinstein Helsinki 1 9 6 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 3
Sieg bert Tarrasch Alexander Alekhine Bad Pistyan 1 9 2 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 4
Jeroen Piket Veselin Topalov Madrid 1 99 7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 5
Alexey Shirov Viswanathan Anand Linares 2 0 0 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 6
Mikhail Botvinnik Ilya Rabinovich Leningrad 1 9 3 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 7
Viktor Kortchnoi Anatoly Karpov Moscow 1 9 74 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 8
Carl Schlechter Emanuel Lasker Vienna 1 9 1 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 9
Robert Hubner Viswanathan Anand Dortmund 1 9 9 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
Evgenij Ermenkov Lajos Portisch Skara 1 9 8 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 1
Detlef Heinbuch Aloyzas Kveinys Bonn 1 9 9 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 2
Vasily Panov Vladimir Simagin Moscow 1 943 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 2
Abram Khasin Leonid Stein Tallinn 1 9 65 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 4
Francisco Briffel Alexander Khalifman Moscow 1 9 8 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 5
Oleg Korneev Veselin Topalov Cala Galdana 1 99 9 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 6
Octavio Troianescu Tigran Petrosian Bucharest 1 9 5 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 8
Samuel Reshevsky Tigran Petrosian Ziirich 1 9 5 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 0 0
Alexander Areshchenko Jan Timman Germany 2 0 0 7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 0 1
Tigran Petrosian Nukhim Rashkovsky Moscow 1 9 7 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 0 2
Shakhriyar Mamedyarov Elmir Guseinov Baku 2000 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I 04
.
Ruslan Ponomariov Viswanathan Anand Mainz 2002 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 0 5

Lev Polugaevsky Boris Gelfand Reggio Emilia 1 99 2 . . . . . . . . . . . . . . . . . . . . . . . . . I 0 6
236
Index of Games
Miguel Najdorf Viacheslav Ragozin Saltsjiibaden 1 948 . . . . . . . . . . . . . . . . . . . . . . . . . . . . I 0 7

Judit Polgar Anatoly Karpov Hoogeveen 2 0 0 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 09
Emanuel Lasker Johann Bauer Amsterdam 1 8 8 9 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 09
Aaron Nimzowitsch Siegbert Tarrasch St Petersburg I 9 1 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . I I 0
Alexander Alekhine John Drewitt Portsmouth 1 9 2 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I l l
Boris Gelfand Vladimir Kramnik Munich 1 9 94 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1 2
Mikhail Botvinnik Alexander Konstantinopolsky Moscow 1 9 5 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 2 3
Surya Shekhar Ganguly Ruslan Ponomariov Szeged 1 9 94 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 24
Hikaru Nakamura Boris Gelfand Moscow 2 0 1 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 24
Hans Kmoch Akiba Rubinstein Semmering 1 9 2 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 2 5
Hans Kmoch Alexander Alekhine Kecskemet I 92 7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 2 6
Anatoly Kremenietsky Evgeny Vasiukov Moscow 1 9 8 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 2 7
Garry Kasparov Anatoly Karpov Moscow 1 9 8 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 2 8
Samuel Reshevsky Reuben Fine Hastings 1 93 7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 2 8
Ruslan Ponomariov Holger Griind Rimavska Sobota 1 99 6 . . . . . . . . . . . . . . . . . . . . . 1 29
Louis Paulsen Siegbert Tarrasch Nuremberg 1 8 8 8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 3 0
Jose Campos Veselin Topalov Palma de Mallorca 1 9 9 2 . . . . . . . . . . . . . . . . . . 1 3 1
Evgeny Alekseev Ruslan Ponomariov Artek 1 99 9 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 3 2
Mikhail Botvinnik Nikolay Sorokin Moscow 1 9 3 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 3 2
Robert Fischer Ratmir Kholmov Havana 1 9 65 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 3 3
Vladimir Bagirov Boris Spassky Leningrad 1 9 6 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 3 4
Alexander Meek Paul Morphy NewYork 1 8 5 7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 3 5
Levon Aronian Sarhan Guliev Istanbul 2003 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 3 5
Oleg Chernikov Anatoly Karpov Kuibyshev 1 9 7 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 3 6
Henrique Mecking Vlastimil Hart San Antonio 1 9 7 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 3 6
Garry Kasparov Miguel Najdorf Bugojno 1 9 8 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 3 7
Paul Keres Reuben Fine Ostend 1 9 3 7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 3 8
Lev Polugaevsky Mikhail Tal Moscow 1 9 69 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 3 9
Alexander Khalifman Przemyslaw Gdanski Leningrad 1 9 8 9 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 42
Alexander Alekhine Irnre Konig Vienna 1 9 2 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 42
Lajos Portisch Svein Johannessen Havana 1 9 66 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 43
Yuri Averbakh Oscar Panna Buenos Aires 1 9 5 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 46
Evgeny Bareev Gennadi Zaichik Kharkov 1 9 8 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 4 7
Vladimir Kramnik Rainer Knaak Dornnund 1 9 9 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 48
Vladimir Kramnik Alberto Mascarenhas Rio de Janeiro 1 9 9 1 . . . . . . . . . . . . . . . . . . . . . . . . . 1 49
Veselin Topalov Francisco Jimenez Morales Seville 1 99 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ! 5 0
Jose Raul Capablanca Vera Menchik Moscow 1 9 3 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ! 5 0
Szymon Winawer Rudolf Charousek Budapest 1 89 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ! 5 3
Alexey Shirov Veselin Topalov Sarajevo 1 9 99 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ! 5 4
Vasily Byvshev Konstantin Klaman Riga 1 9 5 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 5 5
Mikhail Botvinnik Vladimir Alatortsev Leningrad 1 9 3 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ! 5 6
Vladimir Kramnik Peter Toth Rio de Janeiro 1 9 9 1 . . . . . . . . . . . . . . . . . . . . . . . . . ! 5 7
William Steinitz Augustus Mongredien London 1 8 63 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 5 9
William Steinitz Emanuel Lasker New York 1 8 94 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 60
Alexander Khalifman Ian Rogers Groningen 1 99 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 64
Boris Gelfand Igor Khenkin Riga 1 9 8 7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ! 6 6
Akiba Rubinstein Georg Salwe Lodz 1 90 8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 6 8
Vasily Smyslov Garry Kasparov Vilnius 1 9 84 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 69
Tigran Petrosian Boris Spassky Moscow 1 9 69 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 7 0
Alexander Alekhine Jose Raul Capablanca Buenos Aires 1 9 2 7 . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 7 1
Yuri Razuvaev Ivan Farago Dubna 1 9 7 9 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 7 1
237
Alexey Aleksandrov Jaan Ehlvest Bled 2002 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 7 3

Miron Sher Elmar Magerramov Helsinki I 9 9 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I 7 3
Lajos Portisch Anatoly Karpov Linares 1 9 8 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 7 5
Vladimir Kramnik Michael Adams Las Vegas 1 99 9 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 7 6
Ljubomir Ljubojevic Ulf Andersson Wijk aan Zee 1 9 7 6 . . . . . . . . . . . . . . . . . . . . . . . . . . 1 7 7
Robert Fischer Ulf Andersson Siegen 1 9 7 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 8 1
Mark Taimanov Artur Jussupow Moscow 1 9 8 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 8 1
Julio Garcia Soruco Robert Fischer Havana 1 9 6 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 8 2
Louis Paulsen Paul Morpy NewYork 1 8 5 7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 8 2
Paul Morphy AdolfAnderssen Paris 1 8 5 8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 8 3
Paul Morphy Alexander Meek NewYork 1 8 5 7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 8 3
Peter Leko Wolfram von Alvensleben Budapest 1 9 90 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 84
Mikhail Botvinnik Khrisogon Kholodkevich Moscow 1 9 2 7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 8 6
Nikolay Zubarev Akiba Rubinstein Moscow 1 9 2 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 8 7
Esteban Canal Akiba Rubinstein Carlsbad 1 9 2 9 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 8 8
Jacques Mieses Mikhail Chigorin Barmen 1 9 0 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 8 9
Mikhail Gurevich C. Santos Pereira Breda 1 9 9 8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 9 0
Svetozar Gligoric Pal Benko Budapest I 948 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I 9 2
Garry Kasparov Judit Polgar Tilburg 1 9 9 7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 94
Robert Fischer Wolfgang Unzicker Siegen 1 9 7 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 9 6
Emanuel Lasker Jose Raul Capablanca S t Petersburg 1 9 1 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 9 8
Emanuel Lasker David Janowski Paris 1 909 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 0 I
Emanuel Lasker Georg Salwe St Petersburg 1 9 09 . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 0 2
Joseph Blackburne William Steinitz London 1 8 8 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 0 3
Boris Spassky Artur Jussupow Linares 1 99 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 0 4
Mark Taimanov Andor Lilienthal Moscow 1 948 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 0 6
Anatoly Karpov Wolfgang Unzicker Nice 1 9 7 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 0 7
Anatoly Karpov Svetozar Gligoric San Antonio 1 9 7 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 0 8
Leonid Stein Aleksandar Matanovic Tel Aviv 1 9 64 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
George Thomas Akiba Rubinstein Liege 1 9 3 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1 0
Mario Monticelli Akiba Rubinstein Budapest 1 9 2 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1 1
George Thomas Akiba Rubinstein Baden-Baden 1 9 2 5 . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1 3
Curt von Bardeleben Rudolf Spielmann Dusseldorf 1 9 0 8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1 5
Boris Vladimirov Alexander Ryskin USSR 1 9 74 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1 6
John Nunn Nigel Short Brussels 1 9 8 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1 8
Frederick Yates Lajos Asztalos London 1 9 2 7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1 9
Leonid Stein Ervin Haag Tallinn 1 9 6 9 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 2 3
Evgeny Bareev Vladimir Malaniuk Groningen 1 9 9 7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 2 5
Jose Raul Capablanca Savielly Tartakower NewYork 1 9 24 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 2 7
Emanuel Lasker Akiba Rubinstein S t Petersburg 1 9 1 4. . . . . . . . . . . . . . . . . . . . . . . . . . . 2 2 8
Mikhail Botvinnik Max Euwe Groningen 1 946 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 2 9
Jonathan Speelman Michael Trauth Berlin 1 9 80 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 3 0
Erich Eliskases Jose Raul Capablanca Semmering I 9 3 7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 3 0
238

Guliev - Winning Chess Manoeuvres (2015)

Uploaded by

Copyright:

Available Formats

Guliev - Winning Chess Manoeuvres (2015)

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Guliev - Winning Chess Manoeuvres (2015)

Uploaded by

Copyright:

Available Formats

What is the book about?

What is the book about?

What information is provided on the contents page?

What information is provided on the contents page?

Winning Chess Manoeuvres

Winning Chess Manoeuvres

Strategic Id eas thatMasters NeverFail toFind

New InChess 2015

Published by New In Chess, Alkmaar, T he Netherlands

All rights reserved. No part of this book may be reproduced,

Cover design: Volken Beck

Have you found any errors in this book?

Chapter 1 The Janowski Incident or Griefout ofWit ....................18

Chapter 2 Non-Routine Exchanges .................................26

Chapter3 Connecting the Endgame with the Opening ..................3 5

Chapter4 The Battle ofthe Major Pieces .............................43

Chapter5 Dances with the Knights .................................63

Chapter6 The Shuttle Manoeuvre ..................................75

Chapter7 Positional Sacrifices (Part One) ............................81

Chapter8 Positional Sacrifices (Part Two).............................90

Chapter9 Standing on the Shoulders ofthe Classics....................109

Chapter10 Fancy some Solving? ...................................113

Chapter 11 The Pawn Prod .......................................123

Chapter12 Doubled Pawns are Cool! ................................128

Chapter13 Alekhine's Nail .......................................13 7

Chapter 14 Attack and Counterattack with a Closed Centre ...............146

Chapter15 g2-g4, and White wins .................................154

Chapter16 h2-h4, and White wins .................................159

Chapter 17 Overcoming the Blockade ...............................166

Chapter18 With the Mirror's Help .................................175

Chapter19 Fischer's Hedgehog and Morphy's Needles .................. 181

Chapter 20 March ofthe Tank Columns ..............................185

Chapter 21 The Dialectics ofWeakness............................... 196

Chapter 22 Structure and Plan .....................................204

Chapter23 Torture to any Taste ....................................212

Chapter 24 Indirect Borrowings. . ..................................221

Afterword From Ecclesiastes to Eliskases .............................228

Biography .............................................. 232

8 ± White stands slightly better

The great Indian talent Viswanathan

Robert Fischer Akiba Rubinstei n

The forcing variation has ended. White William Steinitz

Repeating combinational schemes,

So why did he compare his game with

A large separate topic is the copying of

1 1 .b5 lt::le4 1 2. i.b2 :ea 1 3.a4 :es

29 ...'iVxc2 30.tt:Jxc2 tt:JgS 3 1 .f3 �f8

1 7. .I:f.c2 .l:!.e7 1 8 .J::i.fcl l:tec7 1 9.t2Jdl g6 M i khail Botvin n i k

22 .�c2! �xc2 23.�xc2 �c8 24.tbel Kotov's commentary in Think Like a

The following pair of twins look like Alexander Alekh ine

Fedor Duz-Khoti m i rsky

I sincerely hope you will enjoy this book!

The Janowski I ncident, or

Mark Ta imanov I would like to believe that the young

Despite his young age (he was still

Ruslan Ponomariov 90. i.12

Rook endings are the most complicated

Chess is an area where there are as

Black to play. If you do not know the Winning.

The method employed by Ponomariov

25 ... h5!? Black to move. His pawn on c7 is hang

I suspect that Donner, not only a grand

1 6 ... exf5 1 7 .exdS cxdS Here we can draw preliminary con