CHESS

MIDDLEGAME
STRATEGIES
Volume 1

by

Ivan Sokolov

www.thinkerspublishing.com
Managing Editor
Romain Edouard

Assistant Editor
Daniël Vanheirzeele

Graphic Artist
Philippe Tonnard

Cover design
Iwan Kerkhof

Typesetting
i-Press ‹www.i-press.pl›

First edition 2017 by Thinkers Publishing

Chess middlegame strategies. Volume 1

Copyright © 2017 Ivan Sokolov

All rights reserved. No part of this publication 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.

ISBN 978-94-9251-001-3
WD/2016/13730/6

All sales or enquiries should be directed to Thinkers Publishers, 9000 Gent, Belgium.

e-mail: info@thinkerspublishing.com
website: www.thinkerspublishing.com
 TABLE OF CONTENTS

KEY TO SYMBOLS 5
PREFACE 7

1. UNUSUAL ROOK LIFTS 9

2. BACKWARD KNIGHT MOVES 39
3. BACKWARD BISHOP MOVES 45
4. MINORITY VS MAJORITY 53
5. VOLUNTARY PAWN STRUCTURE DAMAGE 77
6. OPPOSITE-COLOURED BISHOPS 89
7. DEFENDING WITH LIMITED RESOURCES 115
8. VOLUNTARY KING RUN 135
9. MATERIAL IMBALANCE 149

EXERCISES 217
SOLUTIONS 225
 KEY TO SYMBOLS

! a good move
? a weak move
!! an excellent move
?? a blunder
!? an interesing move
?! a dubious move
 only move
= equality
 unclear position
 with compensation for the sacrificed material
 White stands slightly better
 Black stands slightly better
 White has a serious advantage
 Black has a serious advantage
+- White has a decisive advantage
-+ Black has a decisive advantage
 with an attack
 with initiative
 with counterplay
 with the idea of
 better is
 worse is
N novelty
+ check
# mate
 PREFACE

The ‘work database’ collection for this book which I started about two years
ago meant the initial plan was for just one volume on Chess Middlegame
Strategies.

However, the way my work and material built up and progressed, it soon be-
came obvious that this ‘one volume’ would be approximately 500 pages long
– so, the decision was taken to split the material into two volumes. The second
volume is planned for the second half of 2017.

In this first volume I have annotated 37 games representing different aspects

(chapters) of ‘chess middlegame strategies’.

Readers will see that the vast majority of the games selected belong to the
works of ‘contemporary artists’.

Notwithstanding the ‘traditional/golden classics selection’, I preferred to use

the ‘contemporary learning examples’ feeling that: A) younger/middle genera-
tion reader will more easily associate with them and B) they perfectly mirror
the middlegame strategies used in modern top-class chess (i.e. today’s chess
at the highest level).

Opening play is in general not the subject of this book, although I did pay at-
tention to opening ideas in the cases where I felt I had something useful to say
on the subject (in the particular game selected).

The aim of this book is to offer the reader a spectrum of middlegame ideas/
strategies, thereby inspiring and improving their practical tournament play.

Besides this book’s ‘practical aim’, I hope that the reader will also gain purely
aesthetic pleasure from playing over the games selected (and variations given).

Enjoy!
Ivan Sokolov

December 4th 2016

 1. UNUSUAL ROOK LIFTS

The rook lift is a very common (most- SECOND EXAMPLE

ly attacking) theme/manoeuvre dealt SALOV-KARPOV
with in many books (including my
own ‘Sacrifice & Initiative’ 2013).
The aim of this example is to under-
The most common purpose of a rook stand the extra possibilities/strategies
lift is to bring the ‘heavy artillery’ to in hedgehog positions in which the
a flank and exert pressure on the en- white knight is developed to d2 (in-
emy there. stead of the more common c3 square).
This knowledge can be used in simi-
In this chapter, I deal with unusual lar positions in practical play. I un-
rook lift(s) — anyhow I (based on my derstood the full scope of the con-
knowledge/experience) have found cept only after deeply analysing the
them ‘non-standard’. game mentioned. I have also included
some general guidelines for the open-
So, let me give some explanation ing line here.
(one-by-one) as to why I chose them
for this book.

THIRD EXAMPLE
TOPALOV-CORI
FIRST EXAMPLE
SEIRAWAN-KARPOV
Topalov’s plan of attack in this game
can be used by White in a closed Cat-
An original ‘out-of-the-box’ idea by alan, KID attack (i.e. KIA as White),
Seirawan in a very popular variation. French (as in this game). The rook lift
The idea found no followers (only played by Topalov (non-standard in
three games — all played by Yasser). such positions), combined with (on
This should help the reader to also several occasions) a  g5 piece sacri-
think ‘out-of-the-box’ in similar po- fice — sacrificing a piece to open the
sitions and can also be used as open- h-fi le route to the enemy king — is
ing preparation (the idea is sound and very instructive!
the opening line is topical).
10 CHESS MIDDLEGAME STRATEGIES

FOURTH EXAMPLE 1
ALEKHINE-KMOCH
▷ Y. Seirawan
▶ A. Karpov
‘Chess from the last century’ ☺. In London, 1982 [D53]
standard isolated pawn positions
a rook lift is a ‘common thing’, while 1. f3  f6 2.c4 e6 3.c3 d5 4.d4 e7
a double rook lift is not. Using a dou- 5.  g5 h6 6.  h4 0–0 7. c1 b6
ble rook lift, Alekhine skilfully sus- Karpov goes for his favourite Bond-
tains the pressure until his opponent arevsky/Makagovov (or also in many
collapses. books called Tartakower) variation.

8.cxd5  xd5 9. xd5 exd5 10.  xe7

xe7 11.g3
FIFTH EXAMPLE
8 rsnl+ trk+
TOMASHEVSKY- 7 zp zp wqpzp
MCHEDLISHVILI 6 zp + + zp
5 + +p+ +
4 + zP + +
I was walking around the boards at 3 + + +NzP
the European Team Championships 2 PzP
in Reykjavik, saw this game, remem-
+PzP zP
1 + tRQmKL+R
bered Alekhine-Kmoch, and almost
instantly had a ‘dejà vu’ moment! a b c d e f g h
A standard isolated pawn position,
a double rook lift, skilful play by To- 11... e8
mashevsky. The other main move here is 11...
 a6 (at the top level Black has
In both these games (fourth and fifth) also tried here some other bishop
please notice that, contrary to ‘com- moves, such as 11...  b7 and 11...
mon knowledge’, the exchange of the  f5) 12.e3 c5 There is a very in-
light pieces did not help the side fight- structive game from Tigran Pet-
ing against the isolated pawn, as the rosian I would like to show here:
double rook lift provided the opposi- 13.  xa6 (13.dxc5 was the theoreti-
tion with extra attacking possibilities. cal subject in Korchnoi,V (2695)—
Karpov,A (2700) Merano 1981) 13...
I hope these five examples prove use-  xa6 14.0–0  c7 15.b3  ac8 16.  e1
ful (to give you ideas) in similar posi-  fd8 17.h4  e6 18. d3 f6 19. g2
tions in your practical play. cxd4 20.exd4  xc1 21.  xc1 Because
 1. UNUSUAL ROOK LIFTS 11

of his c-file control White has a small a pawn here and Black is doomed to
advantage. Black wants to neutralize a passive defence. 27...a5 28.b5 d7
this by active play. 21... f4 Black uti- 29.  f4 h5 (29... f8 30.  c6  g6+
lises tactics. White’s rook is attacked, 31. g4  b8 32.h5 does not improve
and if it moves Black will play ... g4 Black’s situation.) 30.e5!  f8 (30...
creating the threat of ... f4. Black’s  xe5 31.  xe5) 31.  b7 f6 32. c6
queen on f4 ‘obviously’ cannot be  g6+ 33. g3  d6 34.  xb6  e6
taken as Black will recapture with 35.  b8+  f8 36.  a8  e1 37. d8
his knight and, owing to a simple h7 38.b6  b1 39.b7  d7 40.  xa5
fork, White will ‘lose’ a pawn. 1–0 Petrosian,T (2640)—Portisch,L
Here comes a moment of superb (2645) Palma de Mallorca 1974.
judgement from Petrosian!!
8 + +k+ tr 12. c3!?
7 zp + +pzp
6 zp +n+ zp
8 rsnl+r+k+
7 zp zp wqpzp
5 + +p+ +
4 + zP wq zP 6 zp + + zp
3 +P+Q+NzP 5 +
2 P+
+p+ +
+ zPK+
1 + tR + +
4 + zP + +
3 + tR +NzP
a b c d e f g h
2 PzP +PzP zP
22.gxf4!! The ninth World Cham-
1 + +QmKL+R
pion allows Black to execute his idea!
In the resulting endgame White will a b c d e f g h
have lasting pressure worth more A very unusual rook lift (for such
than the sacrificed pawn. Simply a standard QGD position). White’s
great judgement by Petrosian! 22... idea has a tactical element, combin-
 xf4+ 23. g3  xd3 24.  c3  b4 ing the already existing c-fi le pres-
25.a3  a6 26.b4  b8 27. c7 sure with a timely  e3 threat (after
8 sn +k+ tr developing his queen to a4). With
7 zp tR
+pzp the  e3 move White also breaks
6 zp + + zp Black e-fi le pressure. The drawback
5 + +p+ + of this 12.  c3!? Idea? It costs time
4 zP zP + zP by delaying the white kingside’s de-
3 zP + +NmK velopment. It is interesting that this
2 + + zP + original idea from Yasser Seirawan
1 + + + + did not fi nd any followers (!). In
a b c d e f g h
my database all three games were
The resulting position of White’s played by Yasser himself. 12.  g2 is
pawn sacrifice idea. The active rook a common move here, seen in many
on c7 compensates for more than top games.
12 CHESS MIDDLEGAME STRATEGIES

12... a6  xc8 16.  g2 c1+ 17. d2 c4 18. a3

Direct central play with 12...c5, aim- xa3 19.bxa3  f6
ing to take advantage of White’s de- 8 + + +k+
layed kingside development, looks 7zpl+ +pzp
perhaps the most principled response 6 zp + sn zp
for Black and was played by anoth- 5 + +p+ +
er QGD connoisseur Efim Petro- 4 +rzP + +
3 zP + +NzP
vich Geller: 13.dxc5 bxc5 (13... c6
2 P+ mKPzPLzP
14.cxb6 (14.  g2?  a6) 14...  b7
1 + + + +R
15.  g2 axb6 (15...  a6 16.0–0!  xe2
a b c d e f g h
17.  xd5) 16.0–0) 14. xd5  b7
15. d2  c6 16.  e3 (16.  g2  ad8 This is a tactically sharp end-
17.  e3! f8 transposes to the game.) ing; while the engine evaluation is
16... f8 17.  g2  ad8 18.  xe8 xe8 around 0.00, to a human it simply
19. e3 d4 20. xe8+  xe8  looks like ‘sharp play’. 20.e5  xd4+
8 + +r+k+ 21. c3  a4 22. c1  xa3+ 23. b2  a5
7zpl+ +pzp 24. xf7 (24.  c7  b5+ 25. a1 d4! )
6 + + + zp 24...  b5+ 25. a1; 12...  a6 13. a4
5 + zp + + leads to some White plus, for exam-
4 + sn + + ple: 13...c6 14.  e3 (14.e3!?) 14... d8
3 + + +NzP 15.  xe8+ xe8 16.e3 e4 17.  xa6
2 PzP +PzPLzP xf3 18.0–0  xa6 19. xa6
1 + + mK +R
a b c d e f g h
13. a4
21. d1 (21. xd4 may be an im-
provement on the game: 21...  xg2
8 r+l+r+k+
7 zp zp wqpzp
22.  g1 d5 White seems (to me) to
have a small advantage in this ending. 6 nzp + + zp
23. b3 (23. f5  xa2 24. d2  e6 5 + +p+ +
25. e3) 23...  b8 24. d2 c4 (24... 4 Q+ zP + +
 xb3 25.axb3  xb3 26.c2 This rook 3 + tR +NzP
ending should be a draw, but it is def- 2 PzP
initely better to sit behind the white
+PzP zP
1 + + mKL+R
pieces here.) 25.c1  xb2+ 26. c3)
21...  d8 22.e1  xg2 23. xg2  b8 a b c d e f g h
24. c1 (24.b3 c4=) 24... xe2+ 25. c2
d4+= 1/2–1/2 Seirawan,Y (2600)— 13...c5?
Geller, E (2575) Linares 1983; Facing a novel idea, Karpov goes
12...  b7 can lead to interesting for a tactical solution and definitely
complications: 13. a4 d7! 14.  xc7 overlooks something in his calcula-
 ac8 15.  xc8 (15.  xd7 e4!–+) 15... tions. 12. c3!? The idea will now (for
 1. UNUSUAL ROOK LIFTS 13

White) work to perfection! Facing the 8 r+l+R+ +

same position (against the same op- 7 zp zp +pzpk
ponent!) for the second time, the 12th 6 +r+ + zp
5 +p+p+ +
World Champion came better pre-
pared and went for 13...b5! removing
4 sn zP + +
3 + +P+NzP
the white queen from the a4-e8 di- 2 PzP mK zP zP
agonal in order to, after 14. a5, play 1 + + +L+R
actively (14. xb5?  b8) with 14... a b c d e f g h
e4! 15. d2!
8 r+l+r+k+ An unusual position with many ide-
7 zp zp +pzp as. I prefer White. 21.a3 (21. e1  xa2
6 n+ + + zp 22. c2) 21... c2+ 22. e3 c6 23.h4
5 wQp+p+ +
4 + zPq+ + a) 23.  g2  xb2 24. c1  b7 25.  xa8
3 + tR +NzP  xa8 26.  h3 (26. h4 b4!) 26...b4
2 PzP mKPzP zP 27.axb4  xb4 28.  f5+ g6 29. d7;
1 + + +L+R
a b c d e f g h
b) 23.  h3?!  xh3 24.  xa8 e7;
15… e6
(15...  b1 can lead to complica- 23...  b8 24. e2 b4 (24...  e6 25.  xb8
tions resulting in a peaceful ending:  xb8 26.  b1 ) 25.axb4 (25.  d1
16. xb5 c6 17.  g2 xh1 (17...xa2  xb2 26. g5+? hxg5 27.hxg5+ g6
18. b3 a5 19. a3 b5 20. e5 28.f4  g4!–+; 25.a4  xb2 26.  c1 b3!
 b8 21.  a1) 18.  xc6  d7 (18... 27.  xc6??  xe2+! 28. xe2 b2–+) 25...
 xg2?? 19.  xe8+  h7 20.  xc8+-)  xb4 26.  b1) 16...b417. e3 (17.  d3
19.  xh1  xc6 20.  xc6  b4 21.  c5 c5–+) 17... b1 18.  xe6 b2+ 19. d1
 xa2 22. e5  b4 23.  xd5  xd5  xe6 20. xa6 a1+ 21. d2 c3+
24.  xd5  ed8= This ending should 22. d1
be a draw.) 16.b3? The move Yasser 8 r+ + +k+
played in the actual game, but White 7 z p zp +pzp
simply does not have the time (here) 6 Q+ +l+ zp
for such a (positional) approach. 5 + +p+ +
Karpov takes advantage of White’s 4 zp zP + +
3 +Pwq +NzP
tempo loss and swiftly develops
2 P+ +PzP zP
a mating attack. (16.  e3! is defi-
1 + +K+L+R
nitely the move Yasser wanted to
a b c d e f g h
play when facing Geller a year later.
Play is sharp and White has reason- Obviously Black (here) wants
able hopes for an advantage: 16... b1 more than a draw. 22...  f5! 23.e1
17. a3  c6 18.  e8+  h7 19. d3+  b8–+ Another rook lift brings
xd3+ 20.exd3  b4 the black rook to the c-fi le, aiding
14 CHESS MIDDLEGAME STRATEGIES

the queen in attack and deciding 19...  ac8

the battle. 24. xa7 (or 24.e3  b6 19... b4+ 20. f1+-
25.xa7  c6–+) 24...  b6 25.e3  c6
26.  c4 a1+ 27. e2  b2+ 28. d1 20.  f3 c2 21.0–0  xb2 22. d1 d8
(28. f1  h3+ 29. g1 c3–+) 28... 23.d4  d7 24.c6 e8 25. xa7
dxc4 29. a8+ h7 30. xc6 c3 0–1  c7 26.a4 a8 27.  xd5  xa7
Seirawan,Y (2595)—Karpov,A (2700) 28. d8+ h7 29. d3+ f5 30. xf5+
Hamburg 1982; g6 31. e6 1–0

An immediate 13... e4? runs into

a nice tactical refutation: 14. d2! 2
 d7 15. xd7 xh1 16. a4!  and
Black loses his knight on a6. ▷ V. Salov
▶ A. Karpov
14. e3! e6 15. xa6 cxd4 Hoogovens Wijk aan Zee (7),
24.01.1998 [E15]
8 r+ +r+k+
7 z p + wqpzp 1.d4  f6 2.c4 e6 3. f3 b6 4.g3  a6
6 Qzp +l+ zp 5. bd2
5 + +p+ + This development of the white
4 + zp + + knight leads to different strategic
3 + + tRNzP possibilities compared to standard
2 PzP
(c3 development) positions.
+PzP zP In this game, the positions we will
1 + + mKL+R examine will be hedgehog-related.
a b c d e f g h After I once lost (a terrible) game to
Lembit Oll (mentioned later in the
16.  b3!+- comments) and later analysing posi-
White will evacuate his king and tions coming from this game (hedge-
his material advantage will prove de- hog-type play), I have formed the
cisive. opinion that aiming for ...d5 — and
(often) hanging pawns in the cen-
16...  f5 tre after subsequent Black ...c5 po-
16... c5 17. d3  f5 18. xd4 a5+ sitions is Black’s best. The reason is
19. d2 xa2 20.d4+- that the white knight developed on
d2 (compared to c3) is not pressur-
17.  g2 c2 18. xd4  xb3 19. xb3 ing Black’s centre (compared to a c3
Black is running out of threats and hitting the central d5 square). This
White has a material advantage. opinion, however, is personal and is
Bringing his king to a ‘safe place’ will a matter of taste.
mean the game has been decided.
 1. UNUSUAL ROOK LIFTS 15

5...  b7 9.h5 g6 10.h3!+-  d5 11.cxd5 exd5

5...d5 for example can (later) lead to 12.0–0 1–0 Oll,L (2600)—Guimaraes,J
‘hanging pawns in the centre’-type (2305) Oviedo 1992) 9.cxd5 exd5 (9...
positions.  xd2 10. a4+ d7 11.  xd2  xd5
12.  xd5 exd5 13.0–0 with Black’s
6.  g2 c5 king in the middle of the board,
White has a winning advantage here.
8 rsn wqkvl tr 13...f6 14.  fe1 fxe5 15.  xe5+ e7
7 zpl+p+pzpp
8 r+ wqk+ tr
6 zp +psn + 7 z p +nvl zpp
5 + zp + + 6 zp + + +
4 +PzP + + 5 + zpptR +
4 Q+ zP + +
3 + + +NzP 3 + + + zP
2 PzP sNPzPLzP 2 PzP vL zP zP
1 tR vLQmK +R 1 tR + + mK
a b c d e f g h a b c d e f g h

16.  ae1 (16.  g5? strangely enough,

Black is aiming for a hedgehog-type lets it slip 16...b5!= 17.xb5  b8 18.c6
position and here we often get plenty  b6 19.a4  b4 20.c6  b6 1/2–1/2
of ‘direct play’ lines also. 6... e7 7.0–0 Goldin,A (2535)—Smirin,I (2490)
0–0 8.b3 d5 9.cxd5 exd5 as seen in many Moscow 1989) 16...0–0 17.  xe7  f6
games, leads to a strategically different 18.  g5 d6 19.  1e6+-) 10. a4+
kind of play, where White (as already e7 (10... c6 11. xe4+-; 10... d7
mentioned above) needs to prove that 11. xe4 dxe4 12.  h3 is hopeless for
his knight developed on d2 brings him Black as 12... c8 loses to 13. c6!  b8
advantages (again compared to the 14. xf7!+-) 11. xe4 dxe4 12.dxc5
‘standard’ c3 development). e8 (12...d5 13.  f4 f5 14.  g5+ 1–0
Skytte,R (2416)—Nikcevic,N (2429)
7.e4! Budapest, 2014) 13.c6  xc6 14. xe4
White wants to take advantage of  xe5 15.  xb7+  f6 16.0–0  b8
Black’s loss of time with 4...  a6 and 17. e4 g6 18.  e1  d6 19. h4+ 1–0
then 5...  b7 and so opts for tactical- Chernin,A (2605)—Podolchenko,E
ly-based, dynamic central play (aim- (2463) Ohrid 2009.
ing to take space).
8.0–0!
7...cxd4 Arguably the most challenging
7... xe4?? is a blunder and should move for Black to face! White com-
lead to an opening disaster for Black: pletes his development, maintaining
8. e5 d5 (8... c3 loses a piece to the tension. 8.e5
 CHESS
MIDDLEGAME
STRATEGIES
Volume 2
Opening meets Middlegame
by

Ivan Sokolov

www.thinkerspublishing.com
Managing Editor
Romain Edouard

Assistant Editor
Daniël Vanheirzeele

Proofreading
Bernard Carpinter

Graphic Artist
Philippe Tonnard

Cover design
Iwan Kerkhof

Typesetting
i-Press ‹www.i-press.pl›

First edition 2018 by Thinkers Publishing

Chess middlegame strategies. Volume 2: Opening meets middlegame

Copyright © 2018 Ivan Sokolov

All rights reserved. No part of this publication 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.

ISBN 978-94-9251-020-4
D/2018/13730/2

All sales or enquiries should be directed to Thinkers Publishing, 9000 Gent, Belgium.

e-mail: info@thinkerspublishing.com
website: www.thinkerspublishing.com
 TABLE OF CONTENTS

KEY TO SYMBOLS 5
PREFACE 7

CHAPTER I. SICILIAN STRATEGIES 9

1.1. Rauzer Ideas 9
Game 1. J. Klovans — M. Tal (1981) 10
Game 2. V. Anand — V. Kramnik (2000) 15
Game 3. T. Baron — Z. Kozul (2010) 20
Game 4. E. Prandstetter — Z. Kozul (1988) 29
Game 5. E. Geller — B. Larsen (1966) 35

1.2. Velimirovic Piece Sacrifice 38

Game 6. D. Velimirovic — U. Andersson (1982) 40
Game 7. D. Velimirovic — L. Gutman (1988) 48
Game 8. D. Velimirovic — M. Suba (1987) 65

1.3. Fortified Knight on e5 69

Game 9. J. Polgar — V. Topalov (2006) 70
Game 10. D. Ghizdavu — B. Spassky (1974) 78
Game 11. A. Karpov — B. Spassky (1982) 82

CHAPTER II. SPANISH PIECE SACRIFICE 87

Game 12. J. Polgar — B. Spassky (1993) 88
Game 13. H. Nakamura — V. Golod (2009) 93
Game 14. A. Rasmussen — I. Sokolov (2017) 96
Game 15. J. Polgar — I. Sokolov (2003) 109
CHAPTER III. THE CARO-KANN DOUBLED F-PAWN 121
Game 16. A. Khalifman — Y. Seirawan (1991) 122
Game 17. E. Torre — V. Kortschnoj (1978) 131

CHAPTER IV. BENONI STRATEGIES 141

4.1. Positional Exchange Sacrifice 141
Game 18. T. Petrosian — N. Rashkovsky (1976) 142

4.2. Kingside Structure Transformations 150

Game 19. E. Magerramov — G. Kasparov (1979) 151
Game 20. E. Bukic — M. Tal (1967) 166

4.3. Two Typical Sacrifices 173

Game 21. L. Le Quang — Saleh Salem (2015) 174
Game 22. V. Kortschnoj — G. Kasparov (1982) 179

CHAPTER V. CATALAN SACRIFICES 191

5.1. Knight Sacrifice on f7 191
Game 23. V. Mikhalevski — A. Rabinovich (1999) 192
Game 24. A. Giri — A. Morozevich (2012) 204

5.2. Pawn Chain Break with the b2-b3 Push 208

Game 25. B. Spassky — L. Zinn (1962) 208
Game 26. Ding Liren — W. So (2016) 222

CHAPTER VI. KNIGHT TALES 233

Game 27. A. Karpov — J. Timman (1982) 234
Game 28. A. Karpov — Z. Azmaiparashvili (1983) 241
Game 29. A. Karpov — M. Taimanov (1977) 249
Game 30. I. Sokolov — A. Miles (1989) 256
 KEY TO SYMBOLS

! a good move
? a weak move
!! an excellent move
?? a blunder
!? an interesing move
?! a dubious move
 only move
= equality
 unclear position
 with compensation for the sacrificed material
 White stands slightly better
 Black stands slightly better
 White has a serious advantage
 Black has a serious advantage
+- White has a decisive advantage
-+ Black has a decisive advantage
 with an attack
 with initiative
 with counterplay
 with the idea of
 better is
 worse is
N novelty
+ check
# mate
 PREFACE

In modern chess practice, the theory of the opening often stretches into the
middlegame, sometimes quite deeply. When players choose particular opening
variations, they are in essence deciding on the type of middlegame in which
they would like to fight their opponents.

In this second volume of my “Chess Middlegame Strategies” series I have de-

cided to focus on this connection between the opening and the middlegame.
I have selected different themes from popular openings and examined the
resulting middlegames, trying to offer a sort of user manual to guide readers
through these middlegames. I have also offered some ideas on the theory of the
different openings. The 30 games in this book cover a wide range of different
middlegame types, so I hope there is something for everyone and that study-
ing this book will enhance readers’ understanding of middlegame strategies.

Only one chapter in this book, “Knight Tales”, is not related to a particular
opening. It concentrates on the middlegame theme of knights versus bishops.
Originally I had planned to present much more pure middlegame material,
but sometimes in the work process we adjust the original plan and, well... this
book is the product!

Naturally I still have this pure middlegame material in my work database and
I do not exclude the possibility that volume three may arrive around the end
of 2018. Meanwhile, I hope you enjoy this book and the game selection.

Ivan Sokolov
November 2017
 CHAPTER I. SICILIAN STRATEGIES

In modern chess, opening theo- 1.1. RAUZER IDEAS

ry meets the middlegame. When
choosing specific opening varia- 8 + + + +
tions, players aim for particular
types of middlegame positions in
7 + + +p+p
6 p+ zppzp +
the hope of understanding these po-
5 +p+ + +
sitions better than their opponents.
4 + +PzP +
Opening preparation will not usu- 3 + + + +
ally lead to a clear advantage for 2 PzPP+ +PzP
either side. If the computer evalu- 1 + + + +
ation of the early middlegame is a b c d e f g h
not “0.00” then it will usually be no
more than plus or minus “0.30” and Understanding this typical Sicilian
the computer will say “=”. So know- Richter-Rauzer position is a must
ing and understanding the typical for any 1.e4 player, any Sicilian play-
plans and motifs in the chosen type er and also any player keen to im-
of middlegame comes to the fore prove his or her general level. I have
and will prove crucial in over-the- focused on dynamic decisions, in
board play. positions with an undefined or mo-
bile pawn centre. Black has doubled
In this chapter I have selected f-pawns and potentially the more
a number of popular opening po- exposed king, but as compensation
sitions, hoping to help readers im- he has the advantage of the bishop
prove their understanding of such pair.
positions and ultimately become
better players. I have selected five games trying to
delve into the concepts involved in
this complex position. Different ide-
as often combine with one another
here. The material is divided into
two parts.
10 CHESS MIDDLEGAME STRATEGIES

8 r+-+kvl-tr king, then opening up the position

7 +-+-+-+- will only benefit the black bishop
6 p+lzppwQ-+ pair. In the Baron-Kozul game we
5 +pwq-+-+p
see Black dominating in a pawn-
down endgame! Endgames are gen-
4 -+-+P+-+
erally good for Black here.
3 +-sN-+-+-
2 PzPP+-+PzP

1 +K+R+L+R 1
a b c d e f g h
▷ J. Klovans
The first three games of this Sicil- ▶ M. Tal
ian subchapter examine the ideas Riga (9), 1981
related to Black having his f6-pawn
removed — in games one and three 1.e4 c5 2. f3 d6 3.d4 cxd4 4. xd4
even by the means of a pawn sacri-  f6 5.c3 c6 6.  g5 e6 7. d2 a6
fice — so that Black’s dark-squared 8.0–0–0 h6 9. e3  xd4 10.  xd4
bishop can take over the a1-h8 diago- b5 11.f3  b7 12. b1  e7 13.  xf6
nal. In games one and two Black has gxf6 14. d3 b6 15.f4
a mobile pawn centre, with the dark-
squared bishop well placed on the 8 r+-+k+-tr
long diagonal, and he castles queen- 7 +l+-vlp+-
side. Tal achieved this position by 6 pwq-zppzp-zp
sacrificing a pawn, while in Anand- 5 +p+-+-+-
Kramnik the material is equal.
4 -+-+PzP-+
3 +-sNL+-+-
Advice for White: exchange one of
2 PzPPwQ-+PzP
the black bishops! Black’s two bish-
ops are superior to White’s knight 1 +K+R+-+R
plus bishop but once one of the black a b c d e f g h
bishops is traded, Black’s initiative
will diminish or vanish altogether. Through a slightly unusual move or-
der we have reached a typical Rich-
Once the black king is safe (here on ter-Rauzer Sicilian pawn structure.
the queenside) Black is fine! White Opening theory is not our concern
needs to be energetic and catch the here so I will just mention that Black
black king in the centre, as in Kasp- often keeps his king in the centre in
arov-Hracek or Sengupta-Edouard. these positions. Tal decides to castle
If White cannot embarrass the black queenside.
 CHAPTER I. SICILIAN STRATEGIES 11

15...0–0–0 16.f5 18...f5!

White challenges Black to push The logical 18...d5, which is also the
...e6-e5 and surrender the d5-square, first choice of the computer en-
a transaction which in this line is gines, leads to a balanced game after
often less beneficial for White than 19.exd5  xd5 20.  he1. White threat-
may initially appear the case. Tal ens 21. f4.
decides to keep his pawn structure
flexible. (20.  hf1 also looks about equal)

16... b8 17.fxe6 fxe6 18.e2 20...  g5 21.g3 e5 22.c3 with un-
clear play. White will exchange the
8 -mk-tr-+-tr light-squared bishops, making the
7 +l+-vl-+- black king potentially vulnerable.
6 pwq-zppzp-zp
5 +p+-+-+-
19.exf5 e5
4 -+-+P+-+
For the small price of just one pawn,
3 +-+L+-+-
Black has gained a rolling pawn
2 PzPPwQN+PzP
centre and opened the h8-a1 diago-
1 +K+R+-+R nal for his dark-squared bishop. In
a b c d e f g h such positions, for the defensive side
(in this case White) it is clever to ex-
White plan’s is clear. He wants to change one of the opponent’s bishops
jump e2-  f4, hitting e6 and hop- as quickly as possible, as Anand did
ing to provoke ...e6-e5. Then he would against Kramnik in our next game.
aim to dominate the light squares,
while the black bishop e7 would be 20. g3
badly placed in such a scenario. Tal
decides to sacrifice a pawn in order The f6-square is a perfect spot for
to activate his dark-squared bishop Black’s bishop that is now on e7, so
and get his central pawns rolling White hopes to prevent that move
forward. In a matter of just a couple by placing his knight on h5.
of moves Black is to unleash the dy-
namic power of his bishop pair and 20...h5!
his central rolling pawns, obtaining
a winning attack. Tal’s method is White here underestimates the dan-
highly instructive and can be used ger, losing a couple of important
in a number of similar positions. tempi to grab the irrelevant h-pawn.
12 CHESS MIDDLEGAME STRATEGIES

8 -mk-tr-+-tr [25.  f1  df8 (25...d4?! 26.  xe4!

7 +l+-vl-+-  xe4 27.f4+)]
6 pwq-zp-+-+
5 +p+-zpP+p
25...  hg8
4 -+-+-+-+
(25...  df8 26.  g4)
3 +-+L+-sN-
2 PzPPwQ-+PzP
26. xh4  xg2 Black definitely has
1 +K+R+-+R enough for his minimal material
a b c d e f g h deficit, however the position is start-
ing to simplify and a draw is becom-
21. e4? ing the likely result.

As mentioned earlier, White needed 21...d5 22.  f3 h4 23. h5 e4 24. e2

to exchange one of Black’s bishops,
reducing the power of the enemy’s 8 -mk-tr-+-tr
initiative. For example: 21.  he1! 7 +l+-vl-+-
h4 22.e4 d5 23. g5 Black is now 6 pwq-+-+-+
forced to part with his bishop pair: 5 +p+p+P+N
23...  xg5
4 -+-+p+-zp
3 +-+-+-+-
(23...e4?! does keep the bishop pair,
2 PzPPwQL+PzP
but also sacrifices an exchange:
24.  f7  hg8 25.  xd8  xd8 1 +K+R+-+R
26.  f1  f6 27. h6 and Black does a b c d e f g h
not have enough compensation for
the sacrificed material.) 24...d4!

24. xg5 e4 Black is getting a huge number

8 -mk-tr-+-tr of tempi for the price of that little
7+l+-+-+- pawn on h4!
6 pwq-+-+-+
5 +p+p+PwQ- 25. f4+ d6 26. xh4 e5
4 -+-+p+-zp
3 +-+L+-+-
2 PzPP+-+PzP
It seems both sides got what they
1 +K+RtR-+-
wanted. White has collected his
a b c d e f g h pawn and is now two pawns up,
Black has advanced his central
25. e2 pawns and placed his dark-squared
 CHAPTER I. SICILIAN STRATEGIES 13

bishop on a perfect position. By rook on h1 are all perfectly mis-

grabbing the h-pawn White has placed)
placed his knight totally out of play
and it will take quite some time and 28...d3! 29.cxd3 a5! 30. d2
effort for the knight on h5 to enter
the fray. White’s king has been left (30.e1 exd3 31.  xd3  xb2–+)
without defenders. The attackers-
defenders ratio favours Black, who 30...e3 31. e1 d4! and Black wins
is ready to push ...d3 opening the h8- with a mating attack. Again, pay at-
a1 diagonal for his queen-and-bish- tention to the misplaced knight on
op battery to deliver a deadly blow. h5.

8 -mk-tr-+-tr 27... c5–+

7 +l+-+-+-
6 pwq-+-+-+ Sensing that his opponent has no
5 +p+-vlP+N
counter-play Tal takes his time in
finishing things off. The threat is
4 -+-zpp+-wQ
28...d3 29.cxd3 b4.
3 +-+-+-+-
2 PzPP+L+PzP
The immediate 27...d3 28.cxd3 d4
1 +K+R+-+R
8 -mk-tr-+-tr
a b c d e f g h 7 +l+-+-+-
6 p+-+-+-+
27.  hf1?! 5 +p+-vlP+N
4 -+-wqp+-wQ
3 +-+P+-+-
White keeps neglecting his king 2 PzP-+L+PzP
position and gives Black a relative- 1 +K+R+R+-
ly easy kill. However, Black has too a b c d e f g h
many attacking motifs and White
simply does not have a good de- was also winning as after 29.  d2
fence. For example, 27. e7, trying Black has plenty of time to bring
to bring the queen to the aid of its his remaining forces into the at-
king, does not help either: 27...  he8 tack: 29... c8 30.  fd1 d5 31.a3 c7
28. b4 32.dxe4

(28. a3 d3! 29.cxd3 f2! 30.  g4 (32.f6  hc8 33.dxe4 xd2–+)
 xd3 31.  xd3 exd3–+ and Black
wins in a few moves while white’s 32...  xe4+ 33. a1
knight on h5, bishop on g4 and
14 CHESS MIDDLEGAME STRATEGIES

(33. a2  d5+–+) 8 -mkr+-+-tr

7 +l+-+-+-
6 -+-+-+-+
33... c1+–+. 5 zp-+-vlP+N
4 -zp-wq-+-+
28. c1 3 +R+Pzp-+-
2 PzP-+L+PzP
28. f2 1 +K+-wQR+-
a b c d e f g h
8 -mk-tr-+-tr
7 +l+-+-+-
6 p+-+-+-+ 34.a3
5 +pwq-vlP+N
4 -+-zpp+-+
(or 34.  f3  xh5! Black keeps his
3 +-+-+-+-
2 PzPP+LwQPzP
bishops pair! 35.  xh5  d5 36.e2
1 +K+R+R+-
a4–+)
a b c d e f g h
34... d5 35. d1 and now Black has
loses to 28...e3! 29. e1. Black proba- enough time to bring his rook into
bly has several ways to win this, but the attack and decide the battle. 35...
the following line nicely illustrates  h7–+ Black’s queen-bishop battery
the domination of the black bishop is still operating at full force. 36.f6
pair and White’s helplessness. Pushing the f-pawn is White only
counter-play. 36...  hc7 37.f7  xb3
(29.  h4 d3! 30.  xd3  xd3 38. xb3 c2 39. xc2  xc2 40.f8 +
31.  xd3 — or 31.cxd3  c8 — 31...  c8 41. xc8+  xc8 42.  c1+ d8
 xh5 32.xh5 d4–+) 43. c2 b3 44. c3 h4–+.

29...d3 30.cxd3 28...d3!

(30.  xd3  xd3 31.  xd3 and the The queen-bishop battery decides.
greedy knight on h5 comes to
a sticky end: 31...  xh5–+) 8 -mk-tr-+-tr
7 +l+-+-+-
30...  c8 31.  c1 d4 32.  c3 b4 6 p+-+-+-+
33.  b3 a5! The full triumph of 5 +pwq-vlP+N
Black’s strategy that started with
4 -+-+p+-wQ
18...f5! and made maximum use of
3 +-+p+-+-
the h8-a1 diagonal. Not having any
2 PzPP+L+PzP
counter-play, White is tied to an
electric chair waiting for the execu- 1 +KtR-+R+-
tion! a b c d e f g h
 CHAPTER I. SICILIAN STRATEGIES 15

29.cxd3 d4 30. c3 b4 31.  b3 d5 12.e5!?

0–1
Direct and radical! We will see
a similar decision in Geller-Larsen
2 later in this chapter.

▷ V. Anand 12...d5 13. b1

▶ V. Kramnik
Wijk aan Zee (4), 2000 This feels a bit timid...

1.e4 c5 2. f3 d6 3.d4 cxd4 4. xd4 a) 13.f5 looks more consequent,
 f6 5.c3 c6 6.  g5 e6 7. d2 a6 leading to a dynamic balance: 13...
8.0–0–0  xd4 9. xd4 e7 10.f4 fxe5 14. xe5
b5 11.  xf6 gxf6 8 r+lwqk+-tr
7 +-+-vlp+p
8 r+lwqk+-tr 6 p+-+p+-+
5 +p+pwQP+-
7 +-+-vlp+p 4 -+-+-+-+
6 p+-zppzp-+ 3 +-sN-+-+-
5 +p+-+-+- 2 PzPP+-+PzP

4 -+-wQPzP-+ 1 +-mKR+L+R
a b c d e f g h
3 +-sN-+-+-
2 PzPP+-+PzP
14...0–0 (Black cannot play 14...
1 +-mKR+L+R  f6 15. g3  b7 because White
a b c d e f g h did not waste time on 13. b1 and
is now faster: 16.fxe6 fxe6 17.  e2)
A typical Richter-Rauzer pawn 15. e4 f6 16. g3+  h8 17.fxe6
structure. Once upon a time, this  xe6.
was a pet-line of Kramnik when
playing Black. Understanding the b) Improving the bishop with
dynamics of these positions helps 13.  e2 before opening the cen-
players to understand other similar tre is probably a critical idea for
Sicilian positions. White is ahead in assessing this position: 13...  b7
development and eager to open files 14.f5! fxe5 15. xe5  f6 16. g3.
towards the black king. To execute Now with the white bishop already
this strategy White often has to be developed on e2, Black cannot al-
ready to sacrifice material. Black is low fxe6 so easily as in Kramnik’s
banking on his bishop pair, should game, as  h5+ will be coming.
the position open up.
16 CHESS MIDDLEGAME STRATEGIES

8 r+-wqk+-tr 21.a3! A surprisingly calm move,

7 +l+-+p+p eliminating the ... b4 threat. The
6 p+-+pvl-+
5 +p+p+P+-
black monarch cannot find shelter,
4 -+-+-+-+
while White develops a winning
3 +-sN-+-wQ- attack. Computer engines sup-
2 PzPP+L+PzP port this evaluation. It is interest-
1 +-mKR+-+R ing that even without knights on
a b c d e f g h the board Black still does not have
a suitable defence! 21...h5 22.  hf1
16...  b8 17.  h3! Keeping the e5 23. d3 g5 24. d7+-  xg2
queens on the board is here 25. de1 e7 26.  f5 h4 (26...  h3
a consequent decision, involving 27. e8+  xe8 28.fxe8 +  xe8
a piece sacrifice. For such deci- 29.  xg5+-) 27.  fe5  d8 28. c7
sions a player needs a brave heart  xd3 29.cxd3 1–0 Sengupta,D
and a good feeling for the initia- (2558)-Edouard,R (2636) Hastings
tive, or superb preparation! (17. 2011.
fxe6 leads to an ending that is fine
for Black: 17... xg3 18.exf7+ f8 13...  b7
19.hxg3 d4 20. b1  xg2=) Now
17...d4 effectively forces White As in Sengupta-Edouard above,
to sacrifice a piece. [After 17...b4 Black is in a hurry to catch up in
Black probably did not like 18. a4 development.
(The piece sacrifice played in the
game, 18.fxe6, is now less effective 13...b4 actually chases the white
for White because the d-file is not knight away to a good square:
open: 18...f4+ 19. b1 bxc3) 18... 14.e2 a5 15. g3 f5 16. h5
 c6 19.c5 and White has good 8 r+lwqk+-tr
attacking prospects.] 18.fxe6! f4+ 7 +-+-vlp+p
19. b1 dxc3 20.exf7+ f8 6 -+-+p+-+
5 zp-+pzPp+N
8 r+-+-mk-tr 4 -zp-wQ-zP-+
7 +l+-+P+p 3 +-+-+-+-
6 p+-+-vl-+
2 PzPP+-+PzP
5 +p+-+-+-
1 +K+R+L+R
4 -+-+-wq-+
a b c d e f g h
3 +-zp-+-+Q
2 PzPP+L+PzP
1 +K+R+-+R
A white knight is almost always
a b c d e f g h well placed on h5 in those positions!
Kasparov,G (2785)-Hracek,Z (2625),
Yerevan 1996, continued 16...  b8
 CHESS
MIDDLEGAME
STRATEGIES
Volume 3
Strategy Meets Dynamics
by

Ivan Sokolov

www.thinkerspublishing.com
Managing Editor
Romain Edouard

Assistant Editor
Daniël Vanheirzeele

Proofreading
Bob Holliman

Graphic Artist
Philippe Tonnard

Cover design
Iwan Kerkhof

Typesetting
i-Press ‹www.i-press.pl›

First edition 2019 by Thinkers Publishing

Chess Middlegame Strategies. Volume 3: Strategy Meets Dynamics

Copyright © 2019 Ivan Sokolov

All rights reserved. No part of this publication 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.

ISBN 978-94-9251-060-0
D/2019/13730/12

All sales or enquiries should be directed to Thinkers Publishing, 9850 Landegem, Belgium.

e-mail: info@thinkerspublishing.com
website: www.thinkerspublishing.com
 TABLE OF CONTENTS

KEY TO SYMBOLS 5
PREFACE 7
CHAPTER I. KARPOV’S KING IN THE CENTER 9
Game 1. G. Kamsky — A. Karpov 10
Game 2. A. Shirov — A. Karpov 20
Game 3. A. Shirov — A. Karpov 27
Game 4. V. Topalov — A. Karpov 34
Game 5. L. Portisch — A. Karpov 38
Game 6. V. Kramnik — V. Anand 42

CHAPTER II. GELLER/TOLUSH GAMBIT PLANS & IDEAS 49

Pure Geller/Tolush Gambit 49
Game 1. E. P. Geller — V. Smyslov 50
Game 2. E. P. Geller — W. Unzicker 60
Game 3. G. Kasparov — M. Petursson 67

Geller/Tolush — Similar Positions 72

Game 4. A. R. Salem — F. Caruana 72
Game 5. L. Ding — J. Duda 81

CHAPTER III. ANTI MOSCOW GAMBIT TYPICAL PLANS & IDEAS 93

Game 1. I. Sokolov — A. Dreev 93
Game 2. H. Nakamura — L. Ding 101
Game 3. A. R. Salem — V. Fedoseev 112
Game 4. V. Kramnik — V. Anand 120
Game 5. I. Sokolov — I. A. Novikov 131

CHAPTER IV. SPACE VS FLEXIBILITY 141

Game 1. S. Ganguly — E. Iturrizaga 142
Game 2. N. D. Short — E. l’Ami 147
Game 3. M. Vachier-Lagrave — V. Fedoseev 151
CHAPTER V. POSITIONAL EXCHANGE SACRIFICE 157
Game 1. Y. Seirawan — Z. Kozul 157
Game 2. P. Svidler — V. Bologan 164
Game 3. J. Lautier — V. Topalov 168
Game 4. G. Kasparov — U. Andersson 174
Game 5. A. Karpov — U. Andersson 184
Game 6. V. Topalov — P. Svidler 195
Game 7. J. H. Timman — G. Kasparov 199
Game 8. G. Kasparov — A. Shirov 206
Game 9. A. Karpov — B. Gelfand 210
Game 10. V. Ivanchuk — V. Kramnik 216
Game 11. N. Mitkov — S. Rublevsky 222

CHAPTER VI. OPEN FILE 229

Game 1. E. Sutovsky — B. Jobava 229
Game 2. I. Sokolov — R. Vazquez Igarza 235
Game 3. N. Vitiugov — I. Bukavshin 244
Game 4. M. Adams — Y. Wang 250
Game 5. A. Firouzja — A. Lenderman 256
Game 6. L. Aronian — W. So 265
Game 7. Yu Yangyi — Wen Yang 271

CHAPTER VII. G-PAWN STRATEGIES 277

Game 1. T. Radjabov — S. Vidit 277
Game 2. L. Portisch — A. Karpov 288
Game 3. J. E. Granda Zuniga — Wesley So 294
Game 4. E. Bacrot — A. Sokolov 298
Game 5. J. Smeets — I. Sokolov 307
Game 6. A. Grischuk — S. Mamedyarov 316
 KEY TO SYMBOLS

! a good move
? a weak move
!! an excellent move
?? a blunder
!? an interesing move
?! a dubious move
 only move
= equality
 unclear position
 with compensation for the sacrificed material
 White stands slightly better
 Black stands slightly better
 White has a serious advantage
 Black has a serious advantage
+- White has a decisive advantage
-+ Black has a decisive advantage
 with an attack
 with initiative
 with counterplay
 with the idea of
 better is
 worse is
N novelty
+ check
# mate
 PREFACE

Dear reader,

The book you are now holding is my third and final in the series on “Chess
Middlegame Strategies.” In this book I have taken strategic themes from spe-
cific openings and combined these with standard strategies such as the open
file, space or flexibility. Indeed, the subtitle for this book is “Strategy Meets
Dynamics” and that is a good definition of what this book is about.

The book contains 43 annotated games divided among seven chapters. Within
these pages you will find strategic plans related to typical pawn structures
arising from the Geller/Tolush Gambit or the Anti-Moscow Gambit. By tak-
ing the time to study those chapters you should come away with a deeper un-
derstanding of the subtle nuances in pawn structures and the dynamics that
arise. In addition, you will learn different strategic/dynamic ideas and this will
strengthen your practical play in such positions.

The chapter on ‘Karpov’s King in the Center’ is where you will be surprised at
how often the 12th World Champion, playing Black no less, deliberately gave
up the right to castle. Not only will you gain perspective on when you might
wish to leave your king uncastled you will also see how Karpov arranges his
pieces so that each move is in harmony with those before and after the king
move. Lastly you will improve your knowledge, from both sides of the board,
of this particular variation of the Caro-Kann.

The open file theme is one that has been written on extensively. What I have
tried to do here is show some modern approaches to this concept.

Perhaps the one dynamic theme that often amazes players is the exchange sac-
rifice. I have included no less than 11 annotated games in this chapter sorted by
the different objectives that the player is trying to accomplish. There is plenty
of knowledge to be absorbed by studying these games.

Finally, the work concludes with the daring, some may even say swashbuck-
ling, aggressive g-pawn push! This aggressive move is used as a means to seize
the initiative. I have laid this chapter out based upon opening variations and
you may find it useful as opening preparation as well as ideas for dynamic play.

My aim in this book, as well as its two predecessors, is not only to help you
improve your middlegame understanding but to give you new and different
ideas/concepts to employ in your own play.

In addition to the chess improvement that is offered I also hope that you will
simply enjoy playing over the selected games.

Ivan Sokolov
Lelystad, Netherlands, August 7th, 2019
 CHAPTER I. KARPOV’S KING
IN THE CENTER

Anatoly Karpov was the 12th World As with all chess games nothing is
Champion. His reign lasted from one-sided. While Karpov did con-
1975–1985 and then from 1993–1999 nect his rooks, Kamsky respond-
during the split championship pe- ed with a principled knight move
riod. With a peak Elo rating of 2780 which included sacrificing a pawn.
and 102 months as the number one As a result he was able to remove
rated player in the world we can Karpov’s dark-squared bishop in
learn a thing or two from him. His the transaction gaining counter-
classical style, endgame technique, play against the black king.
deep understanding of positions
and harmonious piece play is some- A careful analysis of the following
thing one can wish to emulate. games will reveal how White play-
ers try to restrict Black from cas-
At some time around 1993 he came tling with either g4 or h4 only
up with an idea which I will para- to be caught off guard when they
phrase as “Castling is overrated”! suddenly realize that their queen
may be trapped and they must sur-
Competing in the Dortmund super render several tempos to relocate
tournament in 1993 Karpov intro- the queen.
duced an idea whereby he connect-
ed his rooks and made an imme- The first four games, with further
diate threat against the opponent. Karpov games given in the notes,
Karpov was playing none other than come from Karpov’s favorite Ca-
Gata Kamsky who had challenged ro-Kann variation. Thanks to Kar-
him for the world champion’s crown pov this variation was very popu-
in an 18-game match held in Elista lar back in the 1990s when masters
(Kalmykia) in 1996. Karpov won the contemplated positions instead of
20 game match and retained his title switching on an analysis engine.
with a score of 10.5–7.5 (+6=9–3). By I was very skeptical of this concept
the time of the Dortmund meeting when Dutch GM Jan Timman first
both players knew each other’s style showed it to me during an analysis
quite well. session. After additional reflection
10 CHESS MIDDLEGAME STRATEGIES

on how many times my own queen

(placed on either g4 or h4) became 1
a target and how difficult it was to
really attack Black’s seemingly ex- ▶ G. Kamsky
posed king I began to warm to this ▶ A. Karpov
idea. Dortmund 1993

Game five shows Karpov applying 1.e4 c6 2.d4 d5 3.d2 dxe4 4. xe4
the same idea in the Queens Indian d7 5. g5  gf6 6.  d3 e6 7. 1f3
while game six shows Vishy Anand  d6 8. e2 h6 9.  e4  xe4
connecting his rooks, by applying 10. xe4  f6
‘Karpov’s King in Center’ idea while
scoring an important win in his 8 r+lwqk+-tr
2008 World Championship match 7 zpp+-+pzp-
versus Vladimir Kramnik. 6 -+pvlpsn-zp
5 +-+-+-+-
A few important observations:
4 -+-zPQ+-+
3 +-+L+N+-
a) Karpov’s king in the center is
2 PzPP+-zPPzP
safer than it looks.
b) White players beware! White’s 1 tR-vL-mK-+R
queen hastily brought to g4 or a b c d e f g h
h4 can easily become a target!
c) Karpov’s idea can easily be ap- Th is pawn structure gives White
plied in similar positions. more space for the time being while
Black will seek to liberate his game
I hope you enjoy the chosen games by pushing his pawn from c6 to c5.
that illustrate this concept. White seeks to restrict Black’s de-
velopment and places his queen on
1. Kamsky-Karpov, 1993 a seemingly aggressive square.
2. Shirov-Karpov, 2001
3. Shirov-Karpov,2002 11. h4
4. Topalov-Karpov, 1997
5. Portisch-Karpov, 1996 After this game 11. e2 became the
6. Kramnik-Anand, 2008 main line for White.

It looks as if Black will have prob-

lems bringing his king to safety.
Should he castle kingside he runs
 CHAPTER I. KARPOV’S KING IN THE CENTER 11

the risk of a strong attack by White g5 13. h3 c5 also looks good for
which may end up in checkmate for Black.) 13.g5 hxg5 14. xg5 cxd4.
the black king. The idea of castling 8 r+lwq-+-tr
queenside is simply not attractive. 7 zpp+-mkpzp-
Karpov comes up with an ingen- 6 -+-vlpsn-+
ious concept. 5 +-+-+-wQ-
4 -+-zp-+-+
3 +-+L+N+-
11... e7! 2 PzPP+-zP-zP
1 tR-vL-mK-+R
8 r+lwq-+-tr a b c d e f g h
7 zpp+-mkpzp-
6 -+pvlpsn-zp White is already in serious trouble
5 +-+-+-+-
15. d2?! (If White is pawn grabbing
he should do it at once with 15.xg7
4 -+-zP-+-wQ
although Black has a better game af-
3 +-+L+N+-
ter 15...  g8 16.h6 a5+ 17.e2  d7
2 PzPP+-zPPzP
18.  g5  e5) 15... d7 16. xg7  g8
1 tR-vL-mK-+R 17. h6 c6 18. e2 c7 Black had
a b c d e f g h a large advantage in De Firmian,N
(2590)-Epishin,V (2635) Newark
Black keeps his king in the center 1995.
creating the threat of 12...g5! White’s
aggressively placed queen is now b) A relatively calm solution for
a liability. White was 12.  f4 when Black
gets an equal game after 12...  b4+
Black’s king, although looking a bit 13. d2  xd2+ 14. xd2 (14. xd2?!
strange on e7, is quite safe. Realizing looks like a dubious pawn sacrifice
what Black is up to Kamsky takes 14...g5 15.g3 xd4) 14... a5+ 15.c3
a pragmatic approach. c5=.

12.e5! 12...  xe5!

White sacrifices a pawn in order to Karpov correctly judges that the

win Black’s bishop pair, gain a lead pawn sacrifice must be accepted.
in development and try to target the
black king. If 12...g5 White plays 13. g3 and has
a comfortable initiative.
a) 12.g4? looks rather reckless and
plays into Black’s hands 12...c5! (12... 13.dxe5 a5+
12 CHESS MIDDLEGAME STRATEGIES

8 r+l+-+-tr position may be an improvement

7 zpp+-mkpzp- over the game.
6 -+p+psn-zp
5 wq-+-zP-+-
14... xe5+ 15. e3 b6 16.0–0–0 g5
17. a4 c5 18.  he1
4 -+-+-+-wQ
3 +-+L+-+-
2 PzPP+-zPPzP
8 r+l+-+-tr
7 zp-+-mkp+-
1 tR-vL-mK-+R
6 -zp-+psn-zp
a b c d e f g h
5 +-zp-wq-zp-
4 Q+-+-+-+
14.c3
3 +-zPLvL-+-
2 PzP-+-zPPzP
White’s pawn on c3 will later limit
his queen’s mobility thus he might 1 +-mKRtR-+-
have been better off leaving it on c2. a b c d e f g h

14.  d2!? may be an improvement 18... d7

on Kamsky’s play. 14... xe5+ 15. e3
b6 (15...xb2 looks rather risky for When computers first started play-
Black as White obtains strong coun- ing chess they were materialis-
ter play via 16.  c5+ e8 17.0–0 or tic and would not hesitate to grab
16.0–0.) 16.0–0–0 a pawn. In this position such pawn
8 r+l+-+-tr grabbing would be bad for Black af-
7 zp-+-mkpzp- ter 18... xh2?! White gets a strong
6 -zpp+psn-zp attack with 19.g3 g2 20.b4!
5 +-+-wq-+-
4 -+-+-+-wQ
19. a3
3 +-+LvL-+-
2 PzPP+-zPPzP
1 +-mKR+-+R
There is no immediate threat against
a b c d e f g h the queen so Black continues his de-
velopment.
16...g5 (16...  d8 17.  he1 a5 18.a3
with good compensation for the sac- 19...  hd8 20.g3
rificed pawn in Golubev,M (2492)-
Drozdovskij,Y (2627) Odessa 2010.) Here White had an interest-
17. a4 c5 with this pawn still on c2 ing possibility with 20.b4!?. The
White’s queen, once placed on a3, threat is  e3xc5 which forces 20...
would have more mobility and the cxb4 21.  xb4+ d6 (21... e8??
 CHAPTER I. KARPOV’S KING IN THE CENTER 13

22.  xb6+–) 22. xd6+  xd6 re- Now Black’s queen needs to move.
sulting in the following position.
8 r+-tr-+-+ 20... c7
7 zp-+l+p+-
6 -zp-mkpsn-zp 8 r+-tr-+-+
5 +-+-+-zp-
4 -+-+-+-+
7 zp-wqlmkp+-
3 +-zPLvL-+- 6 -zp-+psn-zp
2 P+-+-zPPzP 5 +-zp-+-zp-
1 +-mKRtR-+- 4 -+-+-+-+
a b c d e f g h
3 wQ-zPLvL-zP-
2 PzP-+-zP-zP
Although it is White to move, he
does not have a deadly discovered 1 +-mKRtR-+-
check. This is likely a position of a b c d e f g h
‘dynamic balance’ where play may
continue 23.  e4+ d5 24.h4  g8 White has enough compensation
25.hxg5 hxg5 26.c4  ac8 27.  d4 f5 here but needs to open lines to-
28.  c2!? (28.  xd5 exd5 29.  xd5+ ward the black king by removing
c7= leads to a draw) 28...f4 29. d2. the pawns which shield the king.
8 -+r+-+r+ White’s queen on a3 is out of play
7 zp-+l+-+- and can only be useful if White
6 -zp-mkp+-+ manages to either push b2–b4 or
5 +-+n+-zp- use the a3–f8 diagonal pin to place
4 -+PtR-zp-+
his bishop on d4. Kamsky tries to
3 +-+-+-+-
2 P+LvL-zPP+
make use of the pin.
1 +-mK-tR-+-
a b c d e f g h 21. d4

Now Black must march with his 21.f4 Trying to immediately open
king! 29... c5! 30. de4  f6 31. e5+ the position looks like another logi-
now Black can retreat with 31... d6. cal try. After 21...d5
(Black’s other option is to choose 8 r+-tr-+-+
to be ‘brave’ or ‘greedy’: 31...  xc4 7 zp-wqlmkp+-
32.  b3+ d4+ 33.b2 and the com- 6 -zp-+p+-zp
puter says this position is 0.00” but 5 +-zpn+-zp-
4 -+-+-zP-+
it looks like a mess to human eyes.)
3 wQ-zPLvL-zP-
32.  b4+ c7= White has compen- 2 PzP-+-+-zP
sation for his sacrificed pawn but 1 +-mKRtR-+-
not more. a b c d e f g h
14 CHESS MIDDLEGAME STRATEGIES

22. d4! looks like the critical line pawn defense crumbles leading to
here. a big advantage for White.

a) 22.  d2 is artificial: 22...gxf4 22. b1

23.c4  e3 24.  xe3 fxe3 25.  c2
c6 26. xe3  xd1+ 27.  xd1, now 22. e5 does not bring anything to
Black can be ambitious with 27... White after 22... c6.
 g8 (27...  d8 is easy equality)
28. xh6 e5. With his central- 22... d5
ized queen, only Black can be bet-
ter here. Karpov goes for the most logical
plan. He doubles his rooks on the
b) 22. fxg5  xe3 23.  xe3 hxg5 and d-fi le planning to exchange them
the engine gives around 0.00 but it (starting to force exchanges in gen-
is obvious that White is trying to eral) — a good plan when you are a
keep the balance here. pawn up.

22...f6 is very sharp. The engines The engine is witty enough to point
give many 0.00 variations, but it out that 22... c6!? now works with
is easy for a human to go wrong in White’s king on b1. A repetition of
such complex situations. position occurs after 23.  e5  b7
24.  a6  e4+ 25. a1 c6 26.c4
21... e8!  g4 27.  b5  b7 28.  a6. Karpov
may well have seen this but he is up
8 r+-trl+-+ material and not happy with a draw.
7 zp-wq-mkp+-
6 -zp-+psn-zp 23.f4  ad8 24. c2
5 +-zp-+-zp-
4 -+-vL-+-+
8 -+-trl+-+
3 wQ-zPL+-zP-
7 zp-wq-mkp+-
6 -zp-+psn-zp
2 PzP-+-zP-zP
5 +-zpr+-zp-
1 +-mKRtR-+-
4 -+-vL-zP-+
a b c d e f g h
3 wQ-zP-+-zP-
2 PzPL+-+-zP
21... c6? is a mistake due to 22. e5
d7 (22... b7?? drops the queen 1 +K+RtR-+-
to 23.  a6+–) 23.b4! now the black a b c d e f g h
 CHAPTER I. KARPOV’S KING IN THE CENTER 15

Now Karpov wants White to make pieces, but his king is now some-
a decision regarding his d4-bishop. what exposed. The position is one
of dynamic balance.
24...  5d6
28... e7
This move forces a few trades.
28... g7 was probably safer 29. e3
24...  c6!? was a good alternative f6 (29... d8!? leads to an unclear
since after 25.fxg5 hxg5 White lacks position after 30. e5+ f6 31. e4.)
a clear plan of attack and after the 30.h4  g6 31.  xg6  xg6 32.hxg5
natural 26.  f1  g4 the advantage d7 a draw is the most likely re-
goes to Black. sult after 33. f3 e5 34.  h1 f5+
35. xf5+ xf5=.
25.  xf6+
29. e3 f6 30.h4 gxh4
White gives up his bishop pair but
gains some threats against the black 8 -+-+l+-+
king. 7zp-wq-mk-+-
6 -zp-trpzp-+
25.  e5  xd1+ 26.  xd1 d7 looks 5 +-zp-+-+-
better for Black.
4 -+P+-+-zp
3 +-+-wQ-zP-
25...  xf6 26.fxg5+ hxg5 27.  xd6
2 PzPL+-+-+
 xd6 28.c4
1 +K+-tR-+-

8 -+-+l+-+ a b c d e f g h
7 zp-wq-+p+-
6 -zp-trpmk-+ 31.gxh4
5 +-zp-+-zp-
31.  h6?? is a blunder because
4 -+P+-+-+
Black’s king runs to safety on the
3 wQ-+-+-zP-
queenside 31...hxg3 32. g7+ d8
2 PzPL+-+-zP
33. xf6+ c8.
1 +K+-tR-+-
a b c d e f g h Once again, the silicon brain says
the position is easy ‘0.00’. To hu-
Opening the third rank to the white mans those positions are difficult
queen. Black has traded quite a few to play, and we are now going to see
16 CHESS MIDDLEGAME STRATEGIES

several big mistakes. Even for great this might be an easy 0.00 position
players it is easy to err in complex to the computer but to humans this
positions. is very unclear and easy for either
side to win or lose.
31... d7
32...e5
a) 31... d8? running to the queen-
side does not work for Black due to 8 -+-+l+-+
simple 32. f3. 7 zp-+qmk-+-
6 -zp-tr-zp-wQ
b) 31...  f7 can lead to a draw by 5 +-zp-zp-+-
32.h5  xh5 33.  h1 Black now needs
4 -+P+-+-zP
to find the tactical defense: 33... d7!
3 +-+-+-+-
34.a3 e8 35.  g1 f8 36.  h1 e8=.
2 PzPL+-+-+

32. h6 1 +K+-tR-+-

a b c d e f g h
a) 32.h5? is a blunder here as after
8 -+-+l+-+ 33.h5?!
7 zp-+qmk-+-
6 -zp-trpzp-+ Kamsky correctly understood his h-
5 +-zp-+-+P pawn to be a ‘power’ here but this
4 -+P+-+-+
move is not precise.
3 +-+-wQ-+-
2 PzPL+-+-+
1 +K+-tR-+-
33. g7+ was White’s best and af-
a b c d e f g h ter 33... e6! (33...  f7 34.h5; or 33...
d8 34. h8 both look very risky
32...  xh5 33.  h1 Black is a tempo for Black as White’s h-pawn is very
up compared to our 31...  f7 line strong.) White can make an en-
and White does not have a threat. tertaining draw with 34.  f5+ (34.
33...e5 g4+ e7 is a move repetition,
since 35.  f5? loses to a nice tacti-
b) 32.a3!?  is a computer engine cal shot due to his back rank prob-
proposal. White has now removed lems: 35...  g6!–+.) 34... xf5 35.  f1+
all of Black’s back rank  d1+ mo- e4 36. g2+ d4. Black’s king is
tifs and White has, due to Black’s his most active piece but consider-
unsafe king, good compensation for ing the long-term prospects White
his sacrificed pawn. As a reminder should be happy with a draw.
 CHAPTER I. KARPOV’S KING IN THE CENTER 17

8 -+-+l+-+ 33... g4! 34. h7+

7 zp-+q+-+-
6 -zp-tr-zp-+
5 +-zp-zp-+-
8 -+-+l+-+
4 -+Pmk-+-zP 7 zp-+-mk-+Q
3 +-+-+-+- 6 -zp-tr-zp-+
2 PzP-+-+Q+
5 +-zp-zp-+P
1 +K+-+R+-
4 -+P+-+q+
a b c d e f g h
3 +-+-+-+-
2 PzPL+-+-+
37.b3! (37.  d1+??  xc4–+.) Now
Black has several possibilities, but 1 +K+-tR-+-
all of them lead to no more than a b c d e f g h
a draw. 37...  h5 is the riskiest
choice. In the heat of the battle Karpov
makes a losing blunder!
[37... c3 38. g3+!  d3 (38... b4
39. e1+ a3 40. c1+=) 39. e1+ 34... d8?
d4 (39...  d2?? 40.  f3+ b4
41. c1+–) 40. f2+ c3=] Karpov’s desire to bring his king
closer to safety on the queenside
38. b2 now it is Black who must is easy to understand. Howev-
fi nd the only move to draw: 38... er, White’s h-pawn is now terribly
g4! strong and fast.
8 -+-+-+-+
7 zp-+-+-+- 34...  f7! was winning for Black as
6 -zp-tr-zp-+ White must deal with his back-rank
5 +-zp-zp-+l problems. The pride of White’s po-
4 -+Pmk-+qzP
sition, the h6 pawn, will either be
3 +P+-+-+-
2 PmK-+-+Q+
removed or stopped as 35.  h1 is
1 +-+-+R+-
forced.
a b c d e f g h
[35.h6 xc4! 36.b3 (36.  b3 d3+
39. f2+ d3 40. c2+ e3 41. e1+ 37.xd3  xd3 38.  h1  d8–+) 36...
f3 42.  f1+ (42.e2+ going for the g4 37.  h1 d8–+]
black queen does not make much
sense for White as only Black can be 35...  f3! (35...  xc4 36.b3 e2
better after 42... g3 43.  g1+ xh4 37.  g6  d1+ 38.  xd1 xd1+ 39.b2
44.  xg4+  xg4) 42... e3=. d2+=) 36.  g1 xh5–+
18 CHESS MIDDLEGAME STRATEGIES

35.h6!+– d2 Although a pawn down Black is not

worse thanks to his passed e-pawn
8 -+-mkl+-+ e.g. 42.  xg6+ xg6+ 43. a1 c2
7 zp-+-+-+Q 44.a3 d1+ 45.  xd1 xd1+ 46. a2
6 -zp-+-zp-zP e4=.
5 +-zp-zp-+-
37.  xf5
4 -+P+-+q+
3 +-+-+-+-
Securing the advance of the h-pawn,
2 PzPLtr-+-+
this ending is lost for Black.
1 +K+-tR-+-
a b c d e f g h 37... d7

36. f5? 37...  h2 offered more resistance but

not enough to change the outcome
36.  h8!+– pinning Black’s bish- after 38.h7 e7 39.  g1  f7.
op and promoting the h-pawn was 8 -+-+-+-+
easily winning for White. After 7 zp-+-mkl+P
36... e7 (36...  xc2 37. xc2 xc4+ 6 -zp-+-zp-+
38. b1+– and White will queen 5 +-zp-zpL+-
4 -+P+-+-+
his h-pawn.) 37.h7  h2 38.  f1 g5
3 +-+-+-+-
39. g8 h6 40. e4 the threat of h- 2 PzP-+-+-tr
pawn promotion has tied up Black’s 1 +K+-+-tR-
queen and rook and now after 40... a b c d e f g h
 h4 41.  g1+– the weak position of
Black’s king proves decisive. Now White needs to be clever and
win a tempo. The correct move is
36... xf5? 40.  g7!! (40.  g8 allows Black to
draw by building a fortress 40...
Being a pawn up and having a weak  xh7 41.  xh7  xg8 42.  xg8 f5;
king (and likely little time on clock) Another error would be 40.  g3? due
Karpov decides a queen swap makes to the bad position of the white king
sense. True...in general it does...but Black replies with 40...  e6–+.) 40...
now White’s h-pawn is impossible f8 41.  g3! when  h4 is arguably
to stop! the best defense but it does not help.
(41... e7 now loses immediately be-
36... h4! saved the game. After cause White improves his king with
37.e4 xh6 38.a8+ e7 39.xa7+ 42.c1! and his h-pawn queens with
 d7 40. xb6  g6 41. xc5+ f7. the help of  h3 or  h3: 42...  e6
 CHAPTER I. KARPOV’S KING IN THE CENTER 19

43.  h3 or 42...  h4 43.  h3 or 42... queens 39...  xc4+ 40. d2  d4+
f8 43.  h3.) 42.b3 e7 41. c3+–.
8 -+-+-+-+
7zp-+-mkl+P 38...  h2 39.h7 e7
6 -zp-+-zp-+
5 +-zp-zpL+-
4 -+P+-+-tr
8 -+-+-+-+
3 +P+-+-tR-
7 zp-+lmk-+P
2 P+-+-+-+ 6 -zp-+-zpL+
1 +K+-+-+- 5 +-zp-zp-+-
a b c d e f g h 4 -+P+-+-+
3 +-+-+-+-
White will now improve his king
2 PzP-+-+-tr
position to the point where the
bishop against two pawns ending 1 +K+-tR-+-
does not allow Black the previous a b c d e f g h
fortress. 43. b2 f8 (43...  e6 al-
ways loses to 44.  h3.) 44. c3 Black The last move before the time con-
is doomed to only making rook trol often turns out to be ‘the fatal
moves 44...  h6 (44...e7 45.  h3–+) one’!
45.  d3! g7 46.  d7  h5 47.  e6
 xh7 48.  xf7+ g6 49.  xh7 xh7 40. d3?
50. d3 g6 51. e4+–.
8 -+-mk-+-+ A losing blunder. 40.  g1 was still
7 zp-+l+-+- drawing after 40... e6 (40...f5?! al-
6 -zp-+-zp-zP lows White to promote his pawn
5 +-zp-zpL+- and could only bring Black trou-
4 -+P+-+-+
ble after 41.  g5! f6 42.  h5 and
3 +-+-+-+-
2 PzP-tr-+-+
now Black has to look for a ‘study
1 +K+-tR-+-
like’ solution 42...  xg5 43.h8 
a b c d e f g h  h1+ 44. c2  c6! hoping to make
a draw.) 41.b3 f8 42. c1
Most likely in time trouble Kamsky 8 -+-+-mk-+
makes a terrible blunder! 7 zp-+-+-+P
6 -zp-+lzpL+
38.  g6?? 5 +-zp-zp-+-
4 -+P+-+-+
3 +P+-+-+-
38. c1! was winning on the spot. 2 P+-+-+-tr
After 38... d4 39. e4! the h-pawn 1 +-mK-+-tR-
a b c d e f g h
20 CHESS MIDDLEGAME STRATEGIES

Black is still the one who needs to roll, and White cannot create any
be precise to reach a draw: 42... g7! counter threats.
43.  f5+ f7 44.  g6+ g7 45.  f5+
f7. White can attempt to win with 8 -+-+-+-+
46.  g8  xf5 47.h8   xh8 48.  xh8 7 zp-+-mk-+P
reaching the following position. 6 -zp-+l+-+
8 -+-+-+-tR 5 +-zp-zpp+-
zp-+-+k+-
7
4 -+P+-+-+
6 -zp-+-zp-+
5 +-zp-zpl+-
3 +-+L+-+-
4 -+P+-+-+ 2 PzP-+-+-tr
3 +P+-+-+-
1 +K+-+-tR-
2 P+-+-+-+
1 +-mK-+-+-
a b c d e f g h

a b c d e f g h
42.  g7+ f6 43.  xa7 e4 44. e2 f4
Black’s passed connected pawns 45.b3 f3 46. d1  f5 47. c1  xh7
provide enough counter play to 48.  b7 e5 49.  xb6  xa2 0–1
hold the balance 48...  g4 49.  a8 f5
50.  xa7+ f6 51.  a6 f4 52.  xb6+
f5 53.  b8 e4 54.a4 e3 55. c2 (55. 2
a5?? f3–+) 55... e4 56.  e8+ d4
57. d8+ e4 58. e8+=. ▶ A. Shirov
▶ A. Karpov
40... e6–+ Linares 2001

40...f5 was also an easy win: 1.e4 c6 2.d4 d5 3.c3 dxe4 4. xe4
41.  xe5+ f6 42. d5  xh7–+. d7 5. g5  gf6 6. d3 e6 7. 1f3
 d6 8. e2 h6 9.  e4  xe4
41.  g1 10. xe4

41.  g6 does not help as White’s 8 r+lwqk+-tr

h-pawn eventually falls to 41...f5 7 zpp+n+pzp-
42.  xe5 f6 43.  e1  d7 44.  d1 6 -+pvlp+-zp
 xg6 45.  d6+  xh7 46.  xd7+ 5 +-+-+-+-
g6 47.  xa7 f4–+.
4 -+-zPQ+-+
3 +-+L+N+-
41...f5–+
2 PzPP+-zPPzP

The rest of the game does not need 1 tR-vL-mK-+R

comments. Black’s pawns simply a b c d e f g h