THE IMPROVING

CHESS THINKER
DAN HElSMAN

BOSTON
© 2009 Dan Heisman
All rights reserved. No part of this book may be reproduced
or transmitted in any form by any means, electronic or
mechanical, including photocopying, recording, or by an
information storage and retrieval system, without written
permission from the Publisher.

Publisher: Mongoose Press

1005 Boylston Street, Suite 324
Newton Highlands, MA 02461
info@mongoosepress.com
www.MongoosePress.com
ISBN: 978-0-9791482-4-8
Library of Congress Control Number: 2009926552

In North America distributed by National Book Network

custserv@nbnbooks.com, 800-462-6420
For all other sales inquiries please contact the publisher.

Editor: Dr. Alexey Root

Proofreader: Jonathan Hilton
Cover Design: Kaloyan Nachev

First edition

Printed in China
Dedication

To Dr. Adriaan de Groot, whom I was amazed to learn had only

passed away recently! I had assumed that since he had performed
his experiments years before I had been born, he had passed away
some time ago. But he performed his pioneering work at the age of23
and lived into his 9 0 's. Thanks for the legacy. I hope I have carried it
forward in a worthy manner, at least in some tiny way.
Contents
Dedication 3
Foreword by Lev Alburt 5
Acknowledgements 7
Preface 8
Introduction 9

Chapter 1. The Exercise 15

Chapter 2. Thought Process Basics 24
Chapter 3. Class F 43
Chapter 4. Class E 54
Chapter 5. Class D 68
Chapter 6. Class C 81
Chapter 7. Class B 94
Chapter 8. Class A 110
Chapter 9. Expert and Above 127
Chapter 10. The Thinking Cap 143
Chapter 11. The Basics of Time Management 178
Chapter 12. What the Researcher Learned 187

Appendix A: Dr. Max Euwe Protocol of de Groot A 201

Appendix B: Computer Analysis of Positions 205
Glossary 210
Bibliography 219

4
Foreword
"'{ A Then a club player decides he wants to raise his rating to the next
V V class level, he typically looks to improve his chess knowledge in
some way. Perhaps he will learn a new opening or try to improve his
middlegame or endgame understanding. Often enough, however, the
true secret to improvement lies not in studying additional chess mate­
rial but in learning a more powerful method of thinking.
What is meant by a "more powerful" method of thought? We are
talking about an efficient, rigorous, and pragmatic search for the best
possible move in a given position. It is clear that those players who can
best apply their knowledge and intuition in over-the-board situations
rise to the top. Those whose thought processes contain flaws often reach
rating "plateaus" no matter how hard they study in their spare time.
What this book provides is an in-depth look into the minds of players of
all levels. Using the numerous "protocols" - recorded thought process­
es of real players - that NM Dan Heisman has compiled throughout
his extensive teaching career, The Improving Chess Thinker provides
examples of all types of thought processes, from very weak to powerful,
game-winning ones.
In each chapter, Heisman breaks down the thought processes of
each rating class and explains their mistakes - with an eye toward im­
provement. One of the reasons that those who can afford the luxury of
a trainer often experience significant rating growth is that a trainer is
able to teach his students better methods of thinking. However, within
this book lie all the materials necessary to begin discovering higher­
level chess thought processes immediately. Heisman has already done
all the work for the student: all that a student needs to do is to study
the written "protocols" of his peer group and to compare them with the
ones found in more advanced chapters. Learning to think in a more
logical, informed, and practical manner is necessary for a student to
unlock his full potential. If a student does not have a coach available to
"show him the way," this book provides the perfect solution.
Does reading a book dealing with the subject of "chess thought" and
the decision-making process really allow one to improve? Given that
every move in chess is a unique decision, it is actually possible to im­
prove one's entire game at once by improving one's thought process. In

5
fact, it is easy to argue that how well a player knows how to choose his
move is far more important in raising his level of play than his general
chess knowledge! Often, the difference between a Master and an Expert
comes down to something as simple as time management habits. Inde­
cisiveness, inability to discard a poor candidate move, and over-think­
ing in non-critical positions are all examples of game-losing habits this
book can help turn around.
Finally, this book is not only a guide to better thinking. It is also
a significant contribution to the realm of "scientific" chess literature.
Heisman's experimental methods replicate those of Dutch psychologist
Dr. Adriaan de Groot using everyday club players as test subjects. These
"protocol" records contained within the work will prove valuable for
anyone hoping to gain insight into "the chess mind." But by now, I am
sure you are eager to get started diagnosing what setbacks you might be
facing in your own chess thought process. Enjoy!

Lev A/burl

6
Acknowledgements
! would like to thank my publisher, Mongoose Press, for giving me
the incentive to get this material into public hands. I would also
like to thank Howard Goldowsky for putting me in touch with Mon­
goose and for Howard's help with the manuscript. I would like to
thank my editor, Dr. Alexey Root, and the proofreader, NM Jonathan
Hilton. I thank all the players who, over a period of more than forty
years, participated in the exercises contained in this work. Thanks to
Hanon Russell for allowing me to express my thoughts on chess instruc­
tion at the leading online magazine, Chess Cafe, via my Novice Nook
column. Finally, I would like to thank IM Jeremy Silman, who gave me
permission to include material from my Thinking Cap column, which I
had written for his web site.

7
Preface
he Improving Chess Thinker owes its origins to two previous
Tbooks:
• Thought and Choice in Chess by Dr. Adriaan de Groot
• Silman's Complete Endgame Course by IM Jeremy Silman
De Groot's work because I duplicated his experiment requiring sub­
jects to "think out loud" to find the best move in chess positions, and
Silman's because the layout of this book is similar to his, with each of
the middle chapters devoted to protocols (the transcripts of think-out
loud exercises) provided by players one class higher than the previous
chapter. The strong acceptance and approval of this format encouraged
me to seek out a publisher and make this book a reality.
A third book, Inside the Chess Mind by Grandmaster Jacob Aagaard,
contains almost entirely protocols. However, Inside the Chess Mind dif­
fers from The Improving Chess Thinker in several ways - the former
primarily covers Grandmaster and International Master level players
with little additional comment.
The Improving Chess Thinker attempts to fill in the gaps by inclu­
ding primarily non-master players of all levels. It also contains instruc­
tional comments and additional chapters on thought process.
Finally, a tip of the hat to one of my earlier books, The Improving
Annotator, whose title helped me decide upon The Improving Chess
Thinker.

8
Introduction
enjoy performing chess research in the following two areas:
I 1) How do players learn and improve?
2) How do players think during a game?
It may seem that these two are only remotely related, but they have
many common areas.
When someone is taught to play baseball, they don't primarily
learn about innings, outs, and bases. Instead they are taught how to
bat, throw, catch, and run the bases. But when players learn chess, the
only things they are usually taught - even by competent beginner books
- are the basic rules such as checkmate and draws, and how to move the
pieces. Then they are taught more and more about what the pieces can
do. A few principles like "Get all your pieces into the game," "At the start
of the game try to control the center," "For your first move, push a pawn
two squares in the center," "Castle your king early," or "Knight on the
rim your future is dim" are thrown in for good measure.
But chess is a thinking game, and few beginners are taught how
to think to find their move. No wonder everyone learns their "chess
thought process" in a non-systematic way that quickly leads to bad
thinking habits. A thought process represents all the generic "steps"
or sequences of logi c that go through a player's mind both during his
move and his opponent's move. A thought process does not include the
content of what is being thought.
For example, systematically searching for threats after an opponent's
move is part of a process, but deciding where to put a bishop in a par­
ticular position is content.
Beginners are first taught basics such as how to differentiate the
queen and king, how each piece moves, and that the players alternate
moves. It is not effective to immediately thereafter attempt to teach
thought process via suggestions such as "The first thing you do is exam­
ine your opponent's move to see its effect on the position . . . " Beginners
need time to assimilate basic concepts before attempting to implement
higher level ideas.
However, it is also true that to never teach someone the ba­
sics of thought process can lead to bad habits and, consequently, an

9
The Improving Chess Thinker

unnecessarily chaotic thought process. Some aspects of the thought

process are common among good chess players, and can be taught rela­
tively early in a player's development.
I have explored this phenomenon in many articles, notably online
via Novice Nook at Chess Cafe and The Thinking Cap at IM Jeremy
Silman's Web site, and books such as Everyone's Second Chess Book
and Looking for Trouble.
The Improving Chess Thinker explores the results of "think out
loud" exercises performed by subjects of differing levels of ability. These
exercises are based on those performed by the first serious researcher
into chess thought process, Dr. Adriaan de Groot.
In the late 1930's, Dr. de Groot, a professional psychologist and chess
master, recorded the thought processes of dozens of players of all levels
(mostly stronger players). His purpose was to scientifically discover
how chess players really think. Players were given a position and asked
to find a move while simultaneously verbalizing everything they were
thinking. Each subject's verbalization was recorded and transcribed;
these transcriptions were called protocols.
Dr. de Groot analyzed dozens of protocols and published his results.
His publication was a doctoral-type thesis that was eventually trans­
lated into English as the book Thought and Choice in Chess, now out of
print. Thought and Choice in Chess is not so much a "chess book" as it
is a psychology book about chess.
With the 1938 AVRO tournament nearby, Dr. de Groot garnered
some of the best players of the day to participate in his experiment:
Alexander Alekhine, Reuben Fine, Dr. Max Euwe, Paul Keres, and right
on down the line to a few class players. Most of the protocols were by
players of at least master strength. Dr. de Groot wanted to find out how
players arrived at a move in a typical tournament setting. He did not
show them "White/Black to Play and Win" positions, since the thought
process for problems is much different:
• In "Play and Win" problems the solver is not just looking for the
best move. He is required to find a move that leads to a forced
win in all variations. If he doesn't find a forced win with one
candidate, he tries another. In solving these problems one
doesn't give up until a solution is found, because it is given that
there is one!
• In normal positions such as those found in tournament games
and most de Groot exercises, one should attempt to find the
best move possible in the given time available. Weak players

10
 Introduction

often play more quickly than they should. For strong players
the opposite problem is often the case. Since proving the best
move on each move throughout the game often takes more time
than the clock allows, subjects should use time management
and practical alterations of their maximal thought process to
find a move in a reasonable amount of time.

Therefore Dr. de Groot selected interesting positions from some of

his games and asked the players to come up with a move in a reasonable
amount of time, as if they were playing an important tournament game.
He chose "rich" positions that would be interesting and would help dif­
ferentiate the abilities of the various subjects. Dr. de Groot used dozens
of positions, but most of his work was centered around three, which he
labeled Positions "A," "B," and "C."
As a result of this fascinating experiment, Dr. de Groot was the first
to prove that, contrary to popular belief, grandmasters do not think
deeper than strong club players, say those at United States Chess Fed­
eration (USCF) "Expert" level, rated between 2000-2199. What sepa­
rates Grandmasters from Experts is not just the incisiveness of their
analysis - they analyze much more accurately and efficiently - but also
the number of positions they already know how to play. Dr. de Groot
speculated that Grandmasters know how to play roughly 100,000+
positions, while masters know only 10,000+. It is much easier to know
something than to figure it out. I might add that Grandmasters are also
much better at evaluating positions than are experts, enabling them to
choose better moves even when their actual move-by-move analysis is
similar to that of lesser players.
Thought and Choice in Chess is a tough but fascinating read. I perse­
vered enough to plow my way through it in the late 1960's. Afterwards, I
began recording the thought processes of some local players. I remem­
ber doing this in college, but I believe I began while in high school.
Today I use de Groot's exercise as a tool in my full-time profession as
a chess instructor. Thus I have been administering his thinking process
"protocol" exercise for over 40 years! I hope it is not immodest to say
that I undoubtedly have given the "de Groot Exercise" to many more
players than did Dr. de Groot!
To administer the exercise I primarily used the same positions that
Dr. de Groot gives in his book. Immediately after the exercise, I debrief
the student and usually read him what Dr. Max Euwe or others thought
about the same position. I then compare the student's thought process
to that of the former World Champion's.

11
The Improving Chess Thinker

Many of my students have expressed the "eye-opener" effect of do­

ing the de Groot exercise, hearing my feedback, and comparing their
processes to that of Dr. Euwe or others. If I repeat the exercise with the
same student, I use different positions, occasionally ones from my own
games.
A little background on how exercise positions were chosen: there are
two types of positions with regard to the type of thought process best
used to find a move: analytical and non-analytical:
• Analytical positions are identified by the clash of the two forces,
and require the well-known "if I go there, then he has to go
there" thought process. These usually involve the kind of analy­
sis trees found in Thought and Choice in Chess or in Kotov's
Think Like a Grandmaster.
• Non-analytical positions usually lack tactical opportunities and
are more positional. These are found in the early opening or
sometimes in closed or locked positions, and contain little di­
rect clash of the armies. A non-analytical thought process relies
on a player's application of principles and less on analysis, and
thus requires more judgment. Non-analytical positions can be
considered judgmental.
This book, like de Groot's and Aagaard's, is heavily biased toward
analytical positions. Analytical positions are both much more condu­
cive to objective play and are usually more critical. Analytical positions
are easier (and, in a sense, more fun) to study because one examines
specific moves and the evaluations of the resultant positions. In con­
trast, in non-analytical positions, one relies on the general principles
of strategic play. That is not to say that non-analytical positions are not
important. However, at least at most lower levels of play, playing ana­
lytical positions well is the key to winning most games.

Chapter Overview
The first chapter describes the de Groot exercise and the second the
basics of thought process. The middle chapters each contain several
representative protocols made by players in a particular rating class,
starting with Class F (below 1000) and working up to Expert (above
2000) and Master (above 2200).
Each protocol includes the identification of the protocol position,

12
 Introduction

the subject's rating, his age ("adult" if 18 or over), and the estimated
total time taken. Subjects with Internet Chess Club (ICC) standard
ratings are rated approximately 150 rating points above their U.S. Chess
Federation (USCF) equivalents. So a subject with a rating of "ICC 1500"
is about 1350 USCF. All ratings are USCF unless otherwise noted.
To make the information more interesting and instructive, com­
ments are added after each protocol, such as what was done correctly or
incorrectly, or what could be learned from studying the player's thought
process.
At the end of each "class" chapter will be a summary of thought
process aspects characteristic of that level. Also highlighted are sug­
gestions that would help a player progress to the next class level. For
example, the section on "C" players (1400-1599 USCF) contrasts the
thought processes of C players with those of players in the1600-1799
(Class B) range.
The final chapters contain additional instructive material on thought
process including a separate chapter on time management and lessons I
have learned from administering the de Groot exercise.
Appendix A contains Dr. Max Euwe's protocol (recorded transcript)
of the "de Groot A" position, plus some observations. Appendix B con­
tains computer-aided analysis of each test position.

How to Use This Book

This book can be used in multiple ways:
• As a text supporting psychologists, demonstrating how chess
players think at various levels.
• As a "chess book" showing players how thought process works
in practice. This includes aspects such as which approaches
are effective and which are not, how much time players use in
determining a move, how much effort is spent on various tasks
such as identifying candidate moves, and so on.
• As a manual for improving one's chess thought process. This
can be accomplished via any or all of the following:
o Read the section representing your rating. Then read
the sections one and two classes above your level to see
what knowledge, process, and logic is applied by supe­
rior players to arrive at their move.

13
The Improving Chess Thinker

o Focus on the sections of the book which address

thought process and its cousin, time management,
found primarily in Chapters 2 and 10-12.
o Skim the protocols and focus on the comments after
each and the summaries for each section. Much of
the instruction in the book is contained in these com­
ments.
o Learn from the general principles involving thought
process. These principles are placed in italics outside
the protocols.
Tip: Because there are only six protocol positions, it would be
repetitious if we provided a diagram of the same positions before each
protocol. Therefore, if you are reading a protocol and want to see the
starting position, bookmark the protocol positions in Chapter 1 or make
a separate copy of these positions to keep them handy. To minimize this
concern, many protocols include a diagram of a key analysis position.

14
Chapter 1

The Exercise
he d e Groot experiment allows a researcher t o determine how chess
Tplayers find their moves during competitive play. This is clearly
different than how a player solves a puzzle. In a puzzle the solution is
guaranteed. Thus a player can adopt the attitude, "If this attempt does
not solve the puzzle then I will try something else. The solution has to
be the re."
In the de Groot exercise the players (or subjects of the experiment)
are asked to find moves just as they would in a tournament game. Dur­
ing a game there is no guarantee that there is anything good to find,
such as a mate or win of material. The position may contain no clear
ideas or candidate moves that lead to winning or even drawn positions.
In many practical positions there is no "best" move. There may be sev­
eral almost equally good alternatives.
Thus a researcher performing the de Groot exercise is interested in
how players find their moves when the goals are open and a clearly best
move may or may not exist. For the chess player, performing the exercise
and learning from its results can have enormous practical benefits.
Since chess is a mental sport - a thinking game - the process used
to find a move is of importance not only in examining the source of the
player's current strength, but also in determining his future possibili­
ties for improvement. If a player has a poor process for move selection
then his ability to increase his playing strength is impaired, even if his
other chess skills and knowledge improve. An exercise that can diag­
nose a player's process and expose him to a superior player's process is
a beneficial tool for both instruction and psychological study.
In Thought and Choice in Chess de Groot relied on one position for
much of his conclusions. He also included two other "primary" posi­
tions and a few supplemental ones as well. In giving this exercise to
hundreds of my students, friends, and acquaintances, I also utilized
mostly de Groot's first position, which he labeled "de Groot A". But I
also occasionally used the other two primary de Groot exercises ("de
Groot B" and "de Groot C") as well as a few of my own.
The following six diagrams are the ones discussed in this book. The
first three are from de Groot's book; de Groot "Shafritz" is from one of
my games; and the final two were from my students' games.

15
The Improving Chess Thinker

de Groot A: White to move

de Groot B: Black to move

de Groot C: Black to move

16
 The Exercise

de Groot Shafritz: White to move

de Groot Zyme: White to move

de Groot Ernie: White to move

17
The Improving Chess Thinker

All of the above positions are at least somewhat analytical. Among

the six, the de Groot Shafritz is the most "quiet" but it still depends
partly on analysis. The most commonly used position, de Groot A, is pri­
marily analysis. Thus the bias in this book is toward analytical thought
process. While it is possible to do a de Groot exercise with a purely non­
analytical position, it is not as instructive. So I always choose a position
requiring at least some analysis.
The best ways for a student to improve his thinking process in non­
analytical positions are the "traditional" best ways to improve chess
judgment: extensive slow game experience, reviewing one's games with
stronger players, reading annotated master games, learning principles
and how to apply them, and so on. My archived Novice Nook columns
at www.chesscafe.com deal with the best ways to implement these tra­
ditional learning methods.

Instructions
Before participating in the exercise, each subject was given the follow­
ing instructions:
• Pretend you are playing at the World Open, with the time limit
40 moves in two hours,
• Your game is important but not necessarily the final round
when you are playing for big money,
• You have plenty of time left on your clock - more than an hour,
but certainly not an unlimited time,
• Don't do more or less analysis than you would in the above situ­
ation; i.e. don't "show off' for the exercise,
• This is not a "play-and-win" problem. This is supposed to be
an interesting position from a real game, and so to look for a

"winning" (as opposed to "best") move may be a futile effort.

• Use algebraic notation. That is, don't say, "I will take that
pawn. " Instead use "liJxdS", and
• Use stream-of-consciousness verbalization. Try to say every­
thing you think; don't think quietly and then summarize. This
is important because I can only help you based on what you
say; if you think something and don't say it, I have to assume
you did not think it.

18
 The Exercise

Further, live subjects were given a clock to time their move and in­
structed to make their move and hit the clock when finished. For Inter­
net or over-the-phone students I recorded the starting and end times
so I could calculate their total thinking times. These subjects were in­
structed to conclude by stating their move choice and then saying "Push
Clock!" to indicate that they were finished.
During a student's verbalization, I tried to restrict my comments
and interruptions to the following:
• "Please speak up - I can't hear you," and
• ''You are not saying anything. I cannot use the exercise to help
you improve if you don't articulate what you are thinking."
As happened with Dr. de Groot, I occasionally had to remind a sub­
ject that the clock was running and they had used significant time. I
usually did not do this until 30 minutes or more had elapsed.
Although the purpose of the exercise was to see how players found
a move under "real" conditions, I let each subject know that there were
two reasons why their verbalization would take much longer than it
would to make an identical move at the World Open :
• No one can think out loud as quickly or as efficiently as they can
quietly, and
• Unlike a real game, players were given the position "cold" and
needed time to acclimate: figure out the material count, the
threats, etc.

I estimate that these two factors caused roughly a 2 : 1 ratio in the

length of time it would take to make a move, so if a player were to take
7 minutes for the same move at the World Open, it would take about 14
minutes to do it as a de Groot exercise. Therefore, subjects were advised
to divide by two to roughly estimate their equivalent over-the-board
thinking times. All times given in the book are the actual times.
Although some of my earliest protocols were tape recorded, all the
protocols in this book were recorded manually. This necessitated some
shortcuts - whenever a subject said something non-relevant or redun­
dant, it was usually ignored. For example, if a player said
"I can play knight takes the queen's pawn. That's capturing the pawn
in the center. Let's see what that capture does. I am threatening to win
a knight and go ahead a piece,"
... then I might have written

19
The Improving Chess Thinker

"I can play Nxd5 to win a center pawn; Let's see - I am threatening
to win a knight and go ahead a piece."
... or something similar.
The exception was a student on the Internet Chess Club who was
hard of hearing and, by necessity, typed his thoughts. That made his
protocols, A-1 and A-2 in Chapter 8, different and interesting, so they
were included in this book.
Just as the subject can't talk as fast as he can think, I can't write as
fast as they talk! There was no budget to hire a professional stenogra­
pher. I strongly believe that for the hundreds of protocols transcribed I
retained the essence of what the subject intended. But if a researcher is
trying to find how many times a player stuttered or used passive tense,
these protocols will be quite useless since they are certainly not verba­
tim. However, I do believe that the "shortened" protocols presented in
the book in no way detract from its chess use.
Chapters 3-9 present samples of actual protocols by ascending rat­
ing class. Each chapter represents a rating class and contains protocols
of players who were in that class at the time of the exercise. Analysis
diagrams will be provided for instructional purposes.
Occasionally multiple protocols will be presented for one subject,
each utilizing a different position. For ease of reading, multiple exer­
cises by a single subject are placed consecutively in the chapter repre­
senting his lowest (initial) rating.
There are two types of comments interspersed in the middle of the
protocol:
• Parentheses indicate outside actions or comments that reflect
the action of the player, such as "(silent)" if he paused for a
long time.
• Brackets represent my thoughts at the time. Since I did
not interrupt my students (with the two exceptions noted
above) these comments were not verbalized but were often
used during the post-exercise review. My use of [sic] usu­
ally means the subject is making a clear mistake in analysis or
visualization. The use of [ ! ] indicates the subject has made a
comment very insightful for their level of ability or a surprising
error considering the player's previous comments or his level
of play. A frequent note is " [No eval]" meaning the subject did
not try to evaluate which side stood better, by how much, and
why.

20
 The Exercise

After each protocol, there is a discussion of the interesting aspects

of that protocol. These aspects may include typical problems for the
protocol, identification of key steps that were lacking (such as count­
ing material or doing an evaluation), and/or good points that others at
that level may lack. The summary at the end of each chapter identifies
common strengths and weaknesses at that level, and discusses thought
process improvements that could be made by those players to get to the
next level.

Post Exercise Review

After each exercise, I reviewed the most instructive points with the stu­
dent, including both things they did well and particular areas to note for
improvement. There is a strong overlap between the comments found
after the protocols and these review sessions with the student, though
some parts of the post-exercise reviews were omitted to avoid repetition.
Often the review of the exercise took longer than the exercise it­
self. It was not uncommon for a de Groot exercise plus the review to
take more than a normally scheduled lesson of one hour. During the
review we would discuss the students' process, looking to see whether
they ever evaluated the exercise position, looked for possible opponent
threats, took a reasonable amount of time compared to the criticality of
the position, were thorough and systematic in their analysis, and any
other issue that was particularly striking or instructive.
As part of the review, I usually read Dr. Euwe's instructive protocol,
especially if the student had analyzed the same position: de Groot A
(See Appendix A). Before reading Dr. Euwe, I have found it very helpful
to suggest:
"Don't try to follow all of Dr. Euwe's analysis, which can be quite
rigorous. I want you to concentrate upon following his process - how
he was able to find his move; what logic he used."
I also took questions from the student, e.g. if they were curious about
the "truth" of the position (see Appendix B), desired analytical reasons
why their moves were not best, wished to pinpoint ways to improve
their processes, or wondered why Dr. Euwe approached the problem
the way he did.
For example, at the end of Dr. Euwe's protocol he concludes with
conditional comments like "Probably some more accidents will hap­
pe n" and "Much is still up in the air", so I often ask my student:

21
 Thought Process Basics

9. You calculate the position well, but misevaluate it. Misevalua­

tion means you come to incorrect conclusions about the "goodness" of
the positions at the end of your analysis. Thus, even if your analysis is
plausible, you may end up choosing a move which leads to an inferior
position. Result: Another wrong, possibly disastrous, move chosen.
10. You find a reason why your opponent made a move and then stop
looking for additional reasons. Result: After you move, your opponent's
reply reveals another reason he made his previous move. Oops! This
oversight is enough to lose another game. I devoted the chapter "Just
Because it is Forced" to this concept in Everyone's 2"d Chess Book.
11. When you consider a move, you don't assume your opponent will
make the best (or most dangerous) reply. Result: You play bad moves
and hope your opponent plays worse ones. When he makes a good reply
you think "Whoops" or "Darn ! " Cousin Problem: When your opponent
moves, you assume it is a good or safe move without any analysis.
Result: You are giving your opponent too much credit! By assuming
your opponent has played a good move, you never analyze that your
opponent's move allows you to mate, win material, or gain advantage in
some other way. While analyzing your move, assume that your oppo­
nent will make the best move against it. When your opponent actually
makes a move, assume it might be a mistake and check to see if you can
take advantage of it.
12. You don't play enough chess to be able to recognize common pat­
terns and gain experience. Result: Both the probability and the effect of
many of the previously noted problems are increased.
If you find yourself a victim of one or more of the above problems,
you are not alone! There are plenty of other players out there who are
nowhere close to master - or even expert - strength. There is likely
some reason for this besides just raw talent. You may think the reason
you are not as good as players in the classes above you is that they know
the Caro-Kann better. But your problems are more likely one of the
above.
By not properly implementing the basics of thought process, many
players end up making the game of chess much harder than it is. For
example, they might avoid piece trades because an advanced positional
text tells them not to trade when they have an isolated pawn. Yet by
blindly following that advice they overlook a simple trade that would
win material. Sound familiar? I see these "penny wise and pound fool­
ish" decisions quite frequently. It would be better for many players to

27
The Improving Chess Thinker

avoid extensive study of positional weaknesses until their ratings got

above 1400.
All of the above does not mean that chess is an easy game. Some
things in chess are extremely difficult. Let's list a few of the more dif­
ficult tasks:
1. Finding a combination which would make Shirov or Kasparov
proud. It may be 15 ply deep in the critical line and would take a grand­
master 20 minutes to find, if they could find it at all. These are the kind
of tactics featured in the advanced text The Magic of Chess Tactics by
Meyer and Miiller or even Nunn's Chess Puzzle Book. There is practi­
cally no limit to the difficulty of this part of chess.

White to move after 31 i.e2 •••

I am playing White and have just allowed Black to double attack my

Queen and Rook with 31 . . . i.e2 . My opponent is a good tactical player
and believes he has found a hole in my analysis. However, I found a
very pretty combination which for many players would fall under the
category of "difficult":
32.Y:rxcS! hdl This is forced, as 32.l"lxc8? 33.l"lxc8+ Wff8 would lose
to 34.l"lld 8! (Stronger than 34.l"lxf8 + , which also wins easily.) 33.Y:re6+
Always consider checks, captures, and threats, usually in that order of
descending force. 33 Y:rxe6 Black's moves are all forced, which is
•••

good ! 34.�xe6 !:ie8 Black might have also tried 34.l"lxf2, hoping for
the unecessarily complicated 35.l!lxf2 li:ld3+, though 35.l"lxdl instead
wins cleanly. 35.!:ixdl !:ixe6 36.!:ibl Removal of the guard is probably
the most underrated tactical motif; here White removes the guard to the
square d5. Black resigns, as the perfect geometry continues: 36 . . . l"le4
is not safe, and a knight move is met by 37.�d5 winning the exchange
and a pawn.

28
 Thought Process Basics

2. Deciding between two subtle evaluations; subtle but consequen­

tial. Spot the small but important difference between the two similar­
looking positions in the diagrams below.

Black to move # 1

Black to move # 2

In the first diagram Black can eventually draw with the subtle
1 ... lt>g7!
However, in the second diagram Black has no reasonable defense
to the threat of 2.l:ih8 followed by either promotion on a8 or a skewer
on h7, e.g. L.lt>d7 2.l'!h8 l'!xa7 3.l'!h7+ wins the rook or l . . . lt>f6 2.l'!f8+
promotes. If instead l...l'!al+ 2 .lt>g2, eventually Black will be faced with
the same unstoppable threat.
It is not often easy to spot subtle details, but sometimes it is criti­
cal. A little difference can sometimes make all the difference between
a winning and losing position. Subtle but critical distinctions happen
quite frequently in the endgame. To get it right requires skill, patience,
and a good eye.

29
The Improving Chess Thinker

3. Deciding on the right plan when either none look promising or

many look equally so. It takes quite a bit of experience and judgment
to find the right plan a reasonable percentage of the time. And, if you
go down the wrong track, it could prove to be the decisive mistake. Not
easy at all.
4. Winning a won game when the margin for victory is razor-thin
and the opposition is putting up optimum resistance. This is sometimes
the equivalent of finding a needle in a haystack. The ability to win a
not-so-easy-to-win won game is called technique. This is different than
the less-often discussed (because it is supposed to be "easy") ability to
win an easily won game.
So there are many difficult challenges in chess which give the game
its deserved reputation as mental challenge requiring skill. Players who
play too slowly think these critical situations come up almost every
move and, when facing easy decisions, treat these as way too difficult.
Don't make all chess play so hard. Sometimes playing reasonable chess,
especially in non-analytical positions, is relatively easy. Of course, if
you are not sure your move is non-critical, you must assume that it
may be critical - and play slowly and carefully.
Moreover, for every reader who plays too slowly there are likely
two who play too quickly and carelessly. Those who play fast are espe­
cially vulnerable when faced with a move that is critical. In those situ­
ations one loses immense strength unless time is taken and analysis is
done carefully. For those that cry "I didn't see it! " afterwards - of course
you didn't see it. When playing too quickly one misses a lot!
The moral of the story is clear. There are two main goals most play­
ers should work toward:
1. Becoming a better analyst, a strong component of which is in one's
tactical recognition and solving, and
2. Learning to recognize how much analysis is needed in each po­
sition. Improve at recognizing when your decision is critical. Critical
positions include not only the difficult situations illustrated by the dia­
grams, moves, and commentary above, but also key strategic decisions
such as where to place your king or whether to trade queens. There is
a reasonable amount of time you should take on each move. There are
often dire consequences if you play too fast or too slow.
If the position is critical, take your time and try to analyze carefully.
If you have a non-analytical position where tactics and analysis are not
required, it is probably best to use general principles and play relatively
quickly.

30
 Thought Process Basics

2.2 A Simplified Thought Process

The following is a five-step process based on the ideas expressed previ­
ously.
1. What are all the things my opponent's move does? In other words,
what are all the moves he can do now that he could not do before, what
are his threats, and did how did his move parry my previous threat?
Don't forget the important step of asking about his move, "ls it Safe?"
Also, don't stop when you find one reason for your opponent's move,
because the ones you miss may cost you the game.
2. What are all the positive things I want to do? This is the main
goal of planning, and your decisions should be based on both sides'
threats, strengths, and weaknesses. This step also includes identifica­
tion of potential tactics.
3. What are all the candidate moves which might accomplish one
or more of those goals? I once read the advice, "Don't look for the best
move; look for the best plan and the moves which accomplish that
plan." This advice describes a combination of Steps 2 and 3. The moves
identified in this step are initial candidates.
4. Which of those initial candidates can I reject immediately be­
cause they are not safe? In other words, are there any opponent checks,
captures, or threats which can quickly defeat an initial candidate? Once
you have eliminated these "unsafe" candidates, the remaining candi­
date moves are final candidates. Doing this step consistently I call "Real
Chess". As stated in Error List #l, not doing it is "Hope Chess". In the
protocols, the majority of players made their move without checking to
see if their opponent could reply with a decisive, forcing move.
5. Of the final candidate moves, which one is best I can find in a
reasonable amount of time? This is the hardest step by far; if you can
perform this step perfectly then you can play World Championship
chess. On the other hand, as will be seen in the Chapters 3-7, weaker
players don't usually compare final candidates. They often just latch
onto one idea and play a move they think meets Step 2, ignoring World
Champion Emmanuel Lasker's advice: If you see a good move, lookfor
a better one. For analytical moves, Step 5 involves both finding best
move sequences (searching for the Principal Variation; PV) and com­
parison of possible positions to see which one is the best. Assuming
the opponent plays his best moves, the move which leads to the best
position is the best move.

31
The Improving Chess Thinker

Interestingly, strong players usually perform Steps 1-4 in a very

short period and then spend the overwhelming majority of their time
on Step 5. In a sense, many "improvement" chess books (except those
on planning) are about performing Step 5. However, most weak players
omit one or more crucial steps in performing Steps 1-4, or, conversely,
spend way too much time on them! Consistently being able to complete
all the steps at least moderately well in a reasonable amount of time usu­
ally means that you are on your way toward becoming a good player.

2.3 De Groot's Four Phases

The following are the four phases of thought process proposed by de
Groot as a result of his study:
Phase 1: Orientation (Problem formation) - the subject assesses and
evaluates the position and tries to postulate what he is trying to achieve.
Phase 2: Exploration - the subject tries out a couple of lines in
various moves to see what is out there as possibilities.
Phase 3: Investigation - Deeper searches are made to try to
strengthen the case for one move or another. Here the analysis (for
stronger players) is more sharp and precise and oriented toward the
goals set in Phase 1.
Phase 4: Proof - the subject attempts to prove that either the move
found is the best available to meet the objectives, or at least better than
the others he has been able to find in a reasonable amount of time. (In
his work de Groot does not mention time, so this is my spin since the
best move cannot always be proven or found in a reasonable amount of
time.)
Part of de Groot Phase 2 and all of Phases 3 and 4 occur during what
I called "Step 5" in Section 2 . 2 . The reason is that de Groot worked pri­
marily with very strong players, and players at those levels do Section
2.2's first four steps very quickly and accurately, so that those four steps
are almost entirely contained within his "Orientation" phase.
However, I work with players of all classes, and have found that the
further players are rated below 2000, the more difficulty they have in
performing Steps 1-4 accurately or consistently, so these "basic" steps be­
come a much bigger part of the observed process for players at all levels.
Therefore, players who use a Hope Chess thought process (and those
who use Flip-Coin Chess - see Chapter IO for more on this) - that is
players rated under about 1600 USCF/FIDE - should concentrate more
32
 Thought Process Basics

on improving the first four steps of the process in Section 2 . 2 . Stronger

players looking to improve their processes should concentrate mostly
on Step 5: determining which of the final candidate moves is the best
they can find in a reasonable amount of time.
For example, in Step 5, when examining multiple moves in depth,
players should use what de Groot terms "progressive deepening." That
means a player should not spend all his time on one final candidate,
but instead look around and examine all the candidates at progressively
deeper levels. This is a much more efficient way of trying to either elimi­
nate candidates or identify one as clearly best.
Let's illustrate this benefit of progressive deepening. Suppose there
are three final candidates "A", "B", and "C". If a player initially spends
all of his analysis time just considering A, then he has no information as
to how good "A" is compared to "B" and "C". There may be information
he can derive from quickly examining "B" and "C" that could either:
• eliminate "A" - by making it clear "B" or "C" is clearly superior,
and spend the rest of his time examining only "B" and "C", or
• eliminate both "B" and "C" - in which case once "A" is estab­
lished as remaining safe, it could be played.
In each of those cases a player would save an enormous amount of
time because he would not have to examine "A" nearly as deeply. Keep
in mind that the goal is to find the best moue possible, not to identify
exactly how good "A" is. Therefore, progressive deepening by examin­
ing each move a little deeper and deeper in turn is much more efficient
than completely examining candidate "A" first.

2.4 Performing a de Groot exercise

When performing a de Groot exercise the subject is faced with three
primary differences between the exercise and a real game: the require­
ment for verbalization, lack of familiarity with the position, and lack of
knowledge of the opponent's last move. While verbalization primarily
slows down the process, it should not fundamentally change it. How­
ever, lack of familiarity does change the process by requiring players to
first und ergo the "Phase 1" step identified by de Groot, which we'll call
"familiarization."
The primary part of familiarization involves static evaluation. The
best way to begin is to count the material. The other primary static eval­
uation criteria I suggest are king safety, the activity of all the pieces on
both sides, and pawn structure. The major dynamic, which can make all

33
The Improving Chess Thinker

the difference, is whose move it is. In a de Groot experiment, it is always

the subject's move. Using prior experience, these factors are weighed.
Questions such as "Who stands better, and by how much?" can then be
answered.
This evaluation is important for two reasons:
1. It provides the basis for deciding on drawing maneuvers. A
player who is otherwise worse would be happy to force or play
for a draw, while the player who is better would almost never
do so.
2. In accordance with Steinitz's laws, the static evaluation, if
properly done, should be the optimum "goal" of the de Groot
exercise! For example, if White stands a little better, then, theo­
retically, his best move should leave him a little better.
Very few subjects are aware of #2 and, among those, even fewer use
this information. One subject, USCF Expert Jerry Kolker, as discussed
in the Introduction, proclaimed White to be better but could not find a
line where that was the case. Using his evaluation as a reason, he refused
to give up. Only after half an hour did he find a line to his satisfaction.
I give great credit to Jerry for taking this approach, one that I have not
seen duplicated in roughly 500 de Groot exercises.
Before starting his exercise, Dr. de Groot did not provide the oppo­
nent's previous move, and I did not either - although, if asked, I postu­
lated a reasonable one. Without this information, the subject has to be
extra careful to look for opponent's threats. Many weak players not only
skip the evaluation, but also, without the cue of the prior move, forget
to look for those threats.
Before the exercise, I did not cue any of the subjects as to how to
find their moves, other than the instructions listed in Chapter 1. Thus
each player - at least the first time they performed the exercise - used
whatever process they normally applied during tournaments, resulting
in a variety of methods and abilities.

2.5 Types of Errors Made in the Protocols

There are many types of thought process errors made by subjects during
the protocols such as errors in logic, assuming the opponent will make
silly moves, or the order in which they do the analysis. For example,
before examining candidate moves, it would make sense to count mate­
rial, do a static evaluation of the position, and then look for all of the

34
 Thought Process Basics

oppo nent's threats. Yet many of the players omitted one or more of the
above (especially evaluating the position). Even intermediate players
ofte n waited until halfway through their protocol to ask themselves if
anything was en prise.
Most of these thought process errors will be addressed in the com­
ments after the protocols, but it is worth discussing some of the most
generic errors.
There are three types of chess "vision" and it is easy to make errors
on all three types:
Board Vision : The ability to quickly and accurately recognize where
all the pieces are and assess what they are doing in the present chess
position. Examples of board vision errors: miscounting the material,
not seeing that a king is in check, or not noticing a bishop attack from
across the board.
Tactical Vision: The ability to quickly and accurately recognize
known tactical patterns and their likely consequences. Tactics includes
not only winning material and checkmate, but also tactics for defense,
such as preventing material loss or checkmate. Examples of tactical vi­
sion errors: missing that the opponent is threatening a back-rank mate
or overlooking an easy removal-of-the-guard to win material.
Visualization: The ability to keep track of where all the pieces are
(and "see" them as a position) as you move the pieces in your head
while analyzing future possibilities. One common visualization error,
as termed by GM Nicolai Krogius in Chess Psychology, is a "retained
image" error where one visualizes a piece as remaining on a certain
square even though that piece "moved" earlier in the envisioned ana­
lytical sequence.
Here are some examples:
White to move and mate in 2

35
The Improving Chess Thinker

You may at first spot that 1.Qh6 does not work since the knight
guards g7 and Black can always play 1.. .f6. However, the additional
mating pattern with the queen going to h8 is "on" and so 1.ti'f6 does
the trick. This is an example of using tactical vision.
Now let's look at the same problem in a harder setting, harder be­
cause the tactical part is the same but the board vision is more difficult:
White to move and mate

White has the exact same solution, 1.ti'f6, with the minor exception
that this time the problem is not "mate in two" because the black knight
can sacrifice itself on d4 and the queen on g2 before succumbing to the
mate.
If these two problems are presented to intermediate players (but
separated in time so they don't know it's the same pattern), most will
take longer on the second problem because there are extra, non-essen­
tial pieces on the board and because the bishop and queen are further
away from the vicinity of checkmate. These are board vision issues - the
tactical vision remains almost exactly the same.
Here is a position from one of my student's games:
Black to move

36
 Thought Process Basics

Fearful of back-rank mate problems, Black played 1 ... hS? I asked

him if he considered the forcing moves first (which always should be­
gin by examining checks). His only check results in a mate: 1 ... !:iel+
2.� MI+ 3.lt>hl (or 3.lt>gl) 3 .i.h3#. He replied that he did consider
••

L. J�el+ 2.lt>g2 but he did not "see" 2 . . . �fl+ . If he were truly not able to
visualize the bishop coming in from d3, this would be categorized as a
visualization error. However, if he saw the bishop on d3 but did not rec­
ognize the checkmate pattern as meaningful and thus did not consider
2 ... �fl+, then that would be a different error, a lack of tactical vision.
The most common thought process mistake of weaker players is to
play "Hope Chess," where one makes a move without looking to see if
any forcing move - check, capture, or threat - in reply by the oppo­
nent can be successfully met. Even when intermediate players (in this
case USCF -1300-1700) play very slowly, it is likely that they never ask
themselves "if I make move X, can my opponent then reply with a check,
capture, or threat which I cannot meet?" Further, many players spend
10+ minutes on a move and never spend one second to consider what
the opponent might do after that move! That's an important finding
for anyone wondering why these players do not have a higher playing
strength despite, in some cases, impressive chess knowledge. To avoid
the possibility of a quick loss, a player has to expend some effort each
move making sure that he can meet each check, capture, or threat after
the move he plans to play.
Players are often inefficient in their thought processes or confused
as to when, during that process, they should examine forcing moves:
checks, captures, and threats. One very effective way of improving a
player's thought process is to advise them that there are three distinct
times during the thought process that you should use the forcing move
selection :
1. During the "What are all the things m y opponent's previous
move does" step (Step 1 in 2.2). By using the null move process
(assume you skip your move), ask "If I do nothing, what could
he do to me next move?" and examine his checks, captures, and
threats,
2. During your candidate step (Step 3 in 2.2), it makes sense to at
least consider all your checks, captures, and threats, and
3. During your "Is my candidate move safe?" step (Step 4 in 2.2),
check to see if any forcing move defeats that candidate. This is
the Real Chess step.

37
The Improving Chess Thinker

On a note of efficiency, once a player does the "What are my oppo­

nent's threats?" step correctly, it makes it much easier to do the Real
Chess step, since the forcing moves the opponent has in the first are usu­
ally almost the same forcing moves he might do in the other. Of course,
the difference is determined by how your candidate move affects the po­
sition, and this might make all the difference between safe and unsafe.
Let's return to an earlier example but let the defender play first:
Black to move

If Black is performing the threat step correctly, he will find that

White is threatening 2.i&f6 with inevitable checkmate. Therefore, all his
candidates should use 2. i&f6 as a killer move and not allow the mate.
So if Black chooses the candidate 1...h5, he can ask the Real Chess
question "Does White have any forcing move that can defeat l.. .h5" and
he should examine first the killer move 2. i&f6 when afterwards he needs
to see if the mate can still be stopped. Hopefully, he will then notice that
2 . . . ltlh7 allows the rook to guard h8, and that keeps l...h5 alive as a pos­
sible final candidate. The difference in the position between the threat
step and Real Chess step was that l.. .h5 allowed the king move, which
in turn changed the safety possibilities after 2.i&f6.
If a player consistently plays Real Chess, then he has completed an
important prerequisite toward becoming a good player.
Quiescent Errors
Another very common type of error is the quiescent error: a player
stops analyzing too soon, and evaluates the position before all further
meaningful checks, captures, and threats are considered. In other
words, he thinks the position is "quiet," and thus will yield a reliable
evaluation, when it is not.
An example of a quiescent error from de Groot "Ernie":

38
 Thought Process Basics

de Groot "Ernie": White to move

Suppose you say to yourself, "My queen and bishop are both attacked,
so I need to move my queen and guard my bishop." So you examine
l.%Ve4. To stop there and think the move is safe just because the queen
is no longer attacked and the bishop is guarded would be a quiescent
error because there are further forcing moves for Black which must be
considered. Moreover, this thought process is also Hope Chess, since
1. %Ve4 is a candidate and you did not even consider what Black might do
to you after this move. After l.%Ve4? Black can play l.. .dS ! hitting both
the queen and the bishop at once, winning a piece.
Not all quiescent errors are also Hope Chess, although these two
thought process errors are clearly related, especially at low ply (the first
move and the possible reply).
The above is a "defensive" quiescent error since it misses that l.%Ve4
is not an adequate defense. Here is an example of an offensive quiescent
error:
White to move

39
The Improving Chess Thinker

Suppose White rejects 1.Yfe8+ because of 1. .. gxe8 2.gxe8+ .if8.

This would be a quiescent error because there are further lines to be
analyzed, specifically 3 ..ih6 setting up a mating net:

Black to move

As it turns out, the best Black has is 3 ... Yfal+ 4.mh2 and either
4.Vfe5+ 5.�e5 .ixh6 or 4 ...Vfg7 5..ixgl mxg7. In either case White has
a winning endgame as the rook can gobble up a queenside pawn or two.
Another very common thought process error occurs when a player
see an opponent's move and asks, "What does that move do?" instead
of, "What are all the things that move does?" For example:

White to move

White is ahead a piece but Black's pawns can become menacing.

White tries to counterattack and plays 1.�e5, threatening fl. Black
replies with 1 ... Yfd5.

40
 Thought Process Basics

White to move

It would be easy for White to think, "I know why he made this move
- Black had to defend fl," and then stop. If he did this, then he would
miss the second, even more important reason for this move, which is
that Black also threatens 2 . Wlhl # !
..

T o White's credit, h e did see the threat and decided t o play 2.�£J.
Unfortunately, a few moves later after 2 ... c3 3.Y!?e5? c2,

White to move

White played Hope Chess and quickly moved 4.fixd5? on general

principles (I call moving only on general principles in analytical posi­
tions hand-waving), apparently believing that when ahead in material
he should always trade queens. However, his sense of danger failed
him; he did not look to the next move and realize that it would be very
difficult to save the game after 4 . . . Ei:xd5 because of the threat 5 . . . Ei:dl.
However, White has one final resource. Can you find it?

41
The Improving Chess Thinker

White to move and not lose immediately

Thanks to Rybka, I found the correct plan: 5.li:Je5! l'lxe5 (5 . . . l'ldl

6.li:Jd3) 6.l'lcl l'lc5 7. W el and White will win the dangerous c-pawn with
some chances to save the game. Unfortunately White played 5.e3 and
after 5 .. J�dl resigned.
For more about errors and overall conclusions, see Chapter 12,
"What the Researcher Learned."

42
Chapter 3

Class F and Below

his chapter includes players rated below 1000 USCF, or roughly
Tbelow 1150 Internet Chess Club (ICC) standard.
This beginning level is composed primarily of youngsters or adults
who have just started playing. Their thought processes (especially
youngsters) are usually way too fast, quite superficial, rarely systematic,
and contain little regard for safety.
For each protocol, I list the de Groot position (for example de Groot
A, de Groot B, de Groot C, etc.; see Chapter 1 for those positions); the
age of the subject; the subject's rating; and the time the subject spent to
choose his move.
Comments in parentheses indicate outside actions or comments that
reflect the action of the player, such as (silent) if he paused for a long
time. Comments in brackets are my thoughts. My frequent use of [sic]
means the subject is making a clear mistake in analysis or visualization.
In contrast, [ ! ] indicates the player has made a comment very insightful
for their level of ability or a surprising error. A frequent note is [No
eval] meaning the subject did not try to evaluate which side stood bet­
ter, by how much, and why.
Protocol F-1 (de Groot A; adult; 750 ; 13 minutes) :
1. il.h6 [no evaluation]. J . fiJ d7 no good. J . l"iJxf7. 1. 'il.fel doesn't help.
1. il.h6 to attack rook. 1. fiJxdS he can just recapture and get his rook
open. 1. il.xdS: no, that doesn't work. J . fiJ d7 attacks the rook but then
1 ... fiJxd7 2. il.xe7 fiJxe7 The knight on eS is safe. 1. Vf!aS does not make
much sense [?]. Hmm. 1 . fiJ d7 then l . . . fiJxd7 or 1 . . . il.xd7 J. fiJxg6 fxg6
- > 2. Vflh3 he takes my bishop. The knight onf6 guards h7, so I have to
remove that knight with 3.il.xj6 hf6. Doesn't work. 1. Vflh3 with the
idea of2. Vfih6. 1.il.xf6 hf6 no good. 1 . il.xj6 hf6 2. il.xdS exdS 3. fiJxdS
push clock.
Unlike many beginners, subject F-1 did take his time, which is good.
However, the density of his analysis per unit time is extremely low,
indicating he was spending a lot of time just trying to figure out what
might be possible. In the end, without verifying his analysis, he indicated
a " PV' (principal variation his guess for best moves for both sides)
=

which involved trying to remove the guard. In his PV, he forgot about
the bishop on c6:

43
The Improving Chess Thinker

de Groot A
Black to move after PV 1 ..ixf6 .ixf6 2.hd5 exd5 3.�xd5?

White's last move, the capture 3.�xd5, is a visualization error since

3 . . . .11.xd5 would win the knight.
Anyone can make this mistake in analysis. However, if you feel like
you are winning a pawn (or any amount of material), then that is a "red
flag" indicating you should be extra careful. Whenever there is a "red
flag" in your analysis, you should verify that analysis before commit­
ting yourself to the first move of a critical sequence. If the sequence
doesn't work, it can either cost you the game orjust the opportunity to
have played a much better move instead.
In addition, if the subject were trying to get in �d7, then he is
making the common mistake of sacrificing a piece to win the exchange,
and that is not good, either. We will see this "give up a piece to win the
exchange" mistake quite frequently among the de Groot A protocols of
weaker players.

Protocol F-2 (de Groot A; adult; 800; 4112 minutes)

1. fi.. h 6 [no eval] l . . . �f4 attacks the queen, so better is 1. Vfih3. Would
like the queen on h6. That leaves the pawn open on d4 for queen to
capture so 1. Vfih3 Vfixd4 not so great. How can I protect the d-pawn
and make an offensive move at the same time? He could not do that
[why?]. l. fi.. h 6 �f4 threatens queen then 2.hf4. So 1. fi.. h 6 is decent.
Still leaves pawn on d4 open - not it does not; Vfff d3 stays. He has the
bishop pair [sic]. So I could play l. �xc6 to win the bishop pair imme­
diately and he could capture with the queen or rook. Let's see - game
is tied [finally counts material]. OK. Pick off bishop pair with l . �xc6
push clock.

44
 Class F and Below

This protocol contains some typical "beginner" thought process er­

rors. The subject did not notice Black was threatening 1... \Wxb2 and did
not consider all capturing sequences. At least the reason for the move
_ to win the bishop pair - is a lot better than those of many intermedi­
ate players! Notice that the subject's one big idea is to checkmate the
black king on the dark squares. Therefore, he focuses on getting the
queen and the bishop over there. Many weaker players go right for a
checkmate pattern, no matter how hopeless or uncalled for by Steinitz's
Rules, which require a player to have - or be able to generate - some
advantage near the opponent's king before he can expect to successfully
attack there. However, weaker players often get into bad habits of mak­
ing otherwise counterproductive checkmate threats because their weak
opponents sometimes allow such checkmates! At least in this case the
subject properly saw that the checkmate attempt was hopeless. When
the idea failed, his Callback - the bishop pair - was reasonable.

Protocol F-3 (de Groot A; age 11; 950; 2 112 minutes )

1. fi.h6 [no eval] attacks the rook and then ... or 1. tiJxdS exd5 2. fi.xd5
fi.xdS no good. 1. fi.xf6 fi.xf6 something. 1. 'il,fel or something. 1 . tiJ a4 to
attack the queen. What other piece to move? 1. tiJ e4 tiJxe4 2. \Wxe4 fi.xg5
not good. I think that's it [what's "it"?]. I think that's it. 1. fi.h6 push
clock.
It's good that this subject considered the capture on d5. However
he rejected the capture just because it did not win material. In other
words, he treated the position as if it were a problem. This is a common
error among players rated below -1600.
The subject did not consider all his opponent's recaptures after
1.tiJxd5. But he is correct that, if you find one reply which does not suit
you (as he did, although for the wrong reason), then you don't have to
consider the others. That would be a waste of time.
As it turns out, the subject's chosen move - 1.fi.h6 - forces the rook
to a better position (d8). Therefore, unless there is a strong follow-up,
this kind of threat should usually be eliminated quickly as a candidate
move.

Protocol F-4 (de Groot A; age adult; 900; 13 minutes)

Piece Safety - protected vs. adequately protected. Checks, captures.
Knight takes pawn [?]. 1.fi.xf6 fi.xf6. l. \Wxg6+ takes the pawn. 1. tiJxd5.
l. tiJe4 to ward a knight fork. I had better make a move soon [7.5 mi­
nutes]. I can take the bishop on c6 [does not mention it wins the bishop
pair]. I could play 1. fi.h6. If I take the knight I am subject to a pawn

45
The Improving Chess Thinker

fork [?]. Running out of things to think. I haven't come up with a good
move. I am contemplating 1 . £iJ e4. l . £iJ e4 EiJxe4 2. Wfxe4 EiJxc3 [sic] with
discovered attack. I can 'tfind the helpful move. I will play 1. £iJe4.
The subject first looks at safety - good. Of course, counting the ma­
terial would help, too ! Then he considers checks and captures, again
good. Unfortunately, he does not consider all captures, and see how
good each re-capture is. This subject only had a few random thoughts
in the first seven minutes and was, by far, the least verbal subject (per
unit time) in the book!

Protocol F-5 (de Groot A; age adult; 850; several minutes)

1. Wfxg6+ no good [No evaluation]. Captures: l. il.xdS, 1.il.xf6 (lots
of silence). l. 'ilfel strengthens the center or 1. 'ilcel then 2. EiJg6 'ilfeB.
1. EiJe4 pressures f6 - protected. If 1... EiJxe4 2. Wfxe4 EiJxc3 [sic!].

de Groot A
Black to move after l.EiJe4 .!Cixe4 2.Y:be4
Retained image; Both F-4 and F-5 wish to play 2 .. EiJxc3 but don't "see"
.

the real free piece 2 ..hg5.

lf4 to support the bishop and knight. 1... £iJxf4 2. 'ilxf4 that's not bad.
So lf4 EiJxf4 2. 'il,xf4 if he ...If not l... EiJxf4 then I have the break move
f4-f5. I think that is my best move so far [good!]. l. iJ.. h 6 pressures the
rook so lf4 hit clock.
When Subject F-5 considered 2 ... EiJxc3 right before the diagram, he
was making the same retained image error made by Subject F-4. Both
"see" the discovered attack on the queen and want to capture the knight
on c3 which is no longer there, and neither notices the bishop hanging
on g5.

46
 Class F and Below

Despite some confusion, great credit to subject F-5 for mentioning

that a move was his best move so far! This use of the "King of the Hill" (see
Chapter 10.2) is very rare among weaker players. Yet it was not beyond
World Champion Euwe who, when considering 1.li:Jxc6, mentioned, "If
I don't find anything better I can always do this." This is the equivalent
of stating that the move is the King of the Hill (See Appendix A).

Protocol F-6 (de Groot A; age adult; 700 ; 17 minutes)

White's knight, bishop on gS both attackf6, the bishop is unprotect­
ed. The queen on d3 has OK diagonals. Black's knight on d5 is attacked
and defended. l"l.cl has an open.file; Black's rook also on c8; both sides
castled. The pawn on g6 is advanced - there is a hole at h6, so 1. fi..h 6
is a candidate and 1. fJ..xj6. 1. fi..h 6 candidate; if l ... liJg4 then 2. liJxg4.
1.1'.xj6 yield no advantage. There is an isolated queen's pawn; White
has three pawn islands; Black has two. White's bishop on a2 attacks
the knight on dS, defended by lesser pieces. 1. fJ..h 6 is best sofar. Let me
do a sanity check on 1. fi..h 6. How would I follow that? 2.liJg4 liJxg4;
2. liJxc6 - the bishop is protected - waste of time because the knight is
good. The queen would be good on h6. I could play Wih3-h6; have not
looked any farther. Get pieces from White's queenside to the kingside.
l. liJa4 attacks the queen, but then he plays 1 ... ha4. Looking at vari­
ous options - anything else? 1. liJxg6 hxg6 may be OK, but l .. .fxg6 is
good for Black. 1 . liJ d7 hd7 but the bishop protects the knight at dS.
1. liJd7 hd7 then dS is less protected: 2. liJxdS exdS 3. fJ..xdS is nothing.
1. '?!ff h3 and I can 't coordinate the queen and the bishop. 1. fi..h 6, 1.liJa4
are candidates. The bishop at e7 is weak, so 1. liJe4. After 1. fi..h 6 what
can happen? Can only be attacked. The rook is attacked - would move,
so is viable. Am I missing some other great things? 1. liJxdS (or l. fJ..xdS)
guarded three times - forget it - trying to take on c8. What is he
threatening? d4 is attacked by the queen. [15 minutes]f2 is safe, back
rank. 1... fJ.. bS attacks the queen and the rook; not good. 2. fJ..c4 l"l.xc4
but 2. liJxbS so 1. fJ..h 6. Sanity check: Seeing if . . . liJdS can do something
dangerous. Can move queen. 1 ... liJf4 2. fJ..xj4. So l. fJ.. h 6. push clock.
This protocol shows that not all beginning-level players move ex­
tremely fast. Subject F-6 waited until 15 of his eventual 17 minutes had
elapsed before considering the opponent's threats. Note the common
error of assuming that just because d5 is guarded more times than it
is attacked, capturing there is not good. Many weak players make the
common mistake of treating each chess position like a "play and win"
problem. They look for wins through forcing moves and, if the win is

47
The Improving Chess Thinker

not there, reject the forcing move. But in a real game there may be no
winning move and the best move can easily be a forcing move, such as a
capture, which does not win material. So to reject a capture just because
it does not win material can be a big mistake.

Protocol F-7 (de Groot A; age adult; 950; 20 minutes)

tugS[bishop] - Bishop[knight] on f6 backed by bishop at e7 pro­
tected by tudS. What is the lay of the land? 1. tua4 threatens the queen.
Is that good? If J . tu a4 what would the queen do? [1 ... ha4]. 1... '/fic7,
1... '/fia6 - would not go to c5 -free queen. 1... '/fibs is possible. Oh, OK,
J . tu a4 '/fibs is protected by the bishop. White's queen is protected by
the knight at e5. Then 2. 'lfixbS hbS and the rook's look at each other;
pulling pieces might expose something. Other options? Don't like
J.h:j6; don't like taking knights for bishops. J . hf6 hf6 now - do I
have something going on here? tud7 is possible. Wait a minute. There
is a bishop on a2; If I took... want tudS. If I take l. hdS and if J.. . hdS
2.hf6 hf6 3.tud7 attacking the queen on b6 and the rook onfB.

de Groot A:
Black to move after 1.hd5 hd5? 2 ..ixf6 .ixf6 3 .!i:ld7 •

The removal of the guard of d7 and then the fork: a major

attraction for tactics seekers

That's if he took - he would have to not see this. He can also take
back with the pawn on e6 [good!]: J. Si..xdS exdS - he can see this. The
knight on e5 can go to c4. J . tu c4 hits the queen. Then 1 ... '/fibs pins the
knight to the queen. Then the queen on d3 is not protected. So the
knight is stuck on c4 but the queen is unprotected [sic tLlxbS]. Well, let's
see ... There is something. The rook on ft is not developed. All pieces
are not developed. The queen on d3 is halfway protected - it can't be
moved out of the way. If Black were to drop... No; move the knight on

48
 Class F and Below

f6 so mewhere but that exposes the bishop on e7. The rook on.fl is unde­
veloped. The e-file is partially open. So 1. 'iJ.lel OK, now what else can I
do? The bishop on a2 is in a hole, protected by the knight on c3. Things
are not solidly protected. l ... t:Lixc3 then the bishop on a2 is loose. What
to do. Like J. t:Lic3-a4 thenfreebie! 1...ha4 [finally!]. Didn't see that at
first. Feel like leading candidate is 1. 'iJ.fel - oh, yes - there is - slick
willie - a pawn out b2 hanging out [finally!]. 1. 'iJ.fel '?!ixb2 OK. So
what do we do now? 1. '?!ic2 or 1. '?!id2 or 1. '?!ie2 or 1.b4 - gotta protect
the pawn. Lock knight onto '?!id3 - have to get into a defensive thing.
Need to protect the pawn - '?fi move or 1.b4. Anybody else in trouble?
Interesting. Just noticed 1. t:Lixc6 - lines me up 1. t:Lixc6 '?!ixc6 or l ... 'iJ.xc6
- need to protect the pawn. Don't see alternatives to giving that pawn
up. Gotta protect the pawn. Also 1. 'iJ.bl or 1. 'iJ.c2. 1. 'iJ.c2 allows me to
double rooks on the c-file. I need to be careful - then 1...1i.a4 but 2. t:Lia4.
1. 'iJ.c2 protected by the queen. 1...1i.a4 then I can exchange: 1.'iJ.c2 Ba4
2. t:Lixa4 'iJ.xc2 3. '?fixc2 then ... 'iJ.fB is not guarded. He would not want
to do that. 1. 'iJ.c2: pressure on the c-file - protects b2. No immediate
threat. Looking at his knight on d5. 1. 'iJ.c2 not threatened by the bishop
on c6 or the knight on dS. It helps develop the rook on.fl. The knight on
f6 is still pinned. OK. 1. 'iJ.c2 push clock.

A very interesting protocol! For a player rated under 1000, he not

only took his time, but saw the attraction of the fork at d7. What was
interesting at that point was his comment, "That's if he took - he would
have to not see this." I don't think he meant " .. .if he took" so much as
" ... if he took with the bishop on the first move." That's the part that is not
forced. Black has two other ways to recapture. He then properly notes
that l.Bxd5 exd5 is possible and drops the analysis. When I was rated
1900 I made that same evaluation mistake, too! But World Champion
Euwe (see Appendix A) evaluates much better and sees that l.. .exd5 is
bad for Black. We could hardly expect that of this subject. So he was do­
ing very well to get that far, much better than some much more highly
rated subjects in this book. After twenty minutes he finally plays "Hope
Chess" and moves l.'iJ.c2 without seeing if any check, capture, or threat
can beat it. In fact l.'iJ.c2 is a bad move because after the only answering
capture l...t:Lixc3 White would have to spot 2.t:Lixc6! to avoid disaster. To
not see this when playing l.'iJ.c2 is not only Hope Chess, but just lucky,
since not anticipating such replies is often enough to lose on the spot.

Protocol F-8 (de Groot A; age 13; USCF 800/ICC 1260; 1112 mi­
nutes)

49
The Improving Chess Thinker

There is a knight on eS. The bishop pins the knight onf6. [no evalu­
ation]. Umm. 1. '/f!fxg6+ No. Can 'tfork li:JeS-d7. J. li:ixdS no. 1 . fi.xf6 push
clock.
This seems to be choosing a move by random process of elimina­
tion. The player finds a couple of moves he does not like and figures he
will play the first one he sees which does not seem to lose. I would be
curious to know why he thought 1.li:Jxd5 was a "no" - perhaps because
it did not win material? If so, that would be the same error we have
already seen several times: rejecting a capture just because it does not
win material.

Protocol F-9 (de Groot A; age 9; USCF 600+; 4 minutes)

Interesting. Let's see. l. li:ixdS fi.xdS. Both pieces can capture. l. li:ixd5
Hmm. l. li:ixdS '!f!!xb2 never mind. There I can take. l. li:ixd5 '/f!fxd4 [l...
'!f!!xb2?] 2.li:ixe7+ is goodfor me. If l . fi.xdS exdS.

de Groot A
After 1.hd5 exd5
Good guess! Subject just assumes the move, correct in this case

l.'iJ.c2 and then 'iJ.lcl to pressure fB. l. il.h6 attacks the rook on fB.
Then there is, after I get rid of the bishop, li:id7. I have to get rid of the
knight onf6, so l . fi.xdS push clock.
Talk about making the right move for the wrong reasons! But this
player was trying to be analytical and, although his analysis was all over
the place, showed great promise for someone young and so low rated.
In fact, two years later, his USCF rating was up to 1019. Most young­
sters show great rating gain as their deductive logic capabilities im­
prove around the age of puberty. Notice how he came to his conclusion :

50
 Class F and Below

he scanned the board for various ideas and then, without really analyz­
ing the move actually played, just played it. That's better than playing
the first move you see, but still a good recipe for disaster. It's always a
good idea to check to see if the move you are going to play is at least
safe. In this case - a capture of a piece - it's likely safe because Black is
forced to recapture to maintain material equality.

Protocol F-1 0 (de Groot A; age 14; USCF 900+; 2 minutes)

Moving 1. Vfij3 but the knight onf6 is well protected. His queen is at­
tacking the pawn on d4 if 1 move the queen. Maybe 1. tLl e4 to put pres­
sure on the knight. Then Vfij3 would help. Jf he plays 1 . . tuxe4 2. Vfixe4
.

- not better - if the knight on d5 moves he has a discovery with the

bishop on c6. Let's see . .l should be doing something with the knight
.

on f6. J . tu e4 tuxe4 2. Vfixe4 - would that get me anywhere? The rook

is attacking the bishop on c6. Yeah I think I will go 1.tue4 push clock.
(PV=l. tue4 tuxe4 2. Vfixe4].
Subject F-10 omits evaluating the position and does not look for op­
ponent threats such as l.. .Vfixb2. Then, he not start his examination of
his candidate moves with checks, then captures, then threats. I guess
he felt that 1.tLle4 was a threat but the forcing moves l.tuxd5, 1.hd5,
l.tuxc6 were not even considered - quite the opposite of what Dr. Euwe
selected to consider first (see Appendix A). Another curious aspect of
this protocol was that subject F-10 saw that after 2.Vfixe4 that the queen
could be subject to a discovered attack after the knight on d5 moves
- and still chose the move despite clear and present danger (see the
diagram and discussion earlier with protocol F-5). Of course, besides a
discovery, Black can just play 2 . . . hg5, which was not mentioned.

Summary of Class F and Below

Characteristics of this thought process level:
• Although in general weaker players play faster, this chapter
contains examples of some very slow players. Nevertheless, as
will be seen throughout the remaining chapters, higher-rated
players usually play slower. Besides rating, there is also a cor­
relation between time taken and overall personality. Players
who are very patient in other endeavors usually play slowly and
carefully. On the other side of the spectrum, players who are
"in for the ride" often just play something quickly and see what

51
The Improving Chess Thinker

happens. Although the net result occasionally comes out the

same - or at least results in a similar rating - being careful is a
very good trait in chess. Mercifully, some of the younger "fast"
players end up slowing down as they become more mature in
their understanding of, and approach to, playing chess.
• One thing that sticks out about the protocols in this chapter
is their lack of focus and systematic consideration of alterna­
tives for either player. The players, understandably, don't know
how to find candidate moves or to analyze correctly. So their
approach is very scattershot, even among the slower players.
Rarely does a player stop and ask himself a basic question such
as, "What are all my opponent's checks, captures, and threats?"

or "If I make this capture, what are ALL the ways my opponent
can recapture and how good are each of them?" Instead, af­
ter a capture they just assume a recapture (or a reply) almost
randomly and try to make conclusions from that, even though
that is illogical. In defense of weak players, they just don't have
the tactical vision to see what is threatened and therefore what
is forced, so their move choices are much more random and
less critical to what actually is going on. The solutions are not
only developing better thought processes, but also solving basic
tactics problems to better identify safety issues and playing lots
of slow chess to improve board vision and visualization.
• Players at this level should think primarily about safety and
much less about strategy but, in practice, it often is the other
way around. They make moves without really trying to see the
consequences and sometimes they don't even care. This lean­
ing toward strategy over safety is true of many lower classes,
not just the lowest class. If a move is not safe then it does not
matter if the move is positionally or strategically desirable. An
unsafe move almost always has to be rejected as "not tactically
justifiable." The trait of choosing unsafe moves often comes
from reading too many advanced chess books, where subtle dif­
ferences are stressed but not the foundations of basic safety.
Improvements to the thought process that would help in getting to
the next level:
• Start the thought process by examining the opponent's threats
from previous move, asking "What are ALL the things my op-

52
 Class F and Below

ponent is trying to do?" In a de Groot exercise, one does not

know the previous move. But outstanding threats by the op­
ponent still have to be identified. If only one threat is missed,
that could easily be enough to lose the game.
• Consider as candidate moves all checks, captures, and threats.
• Play more slowly in slower time control games, especially for
critical moves.
• Do repetitious study of basic tactical patterns to recognize safe
and critical moves. Treat these patterns like the "multiplication
tables" of chess - the basis for your sense of chess safety.
• Don't just consider one possible opponent's reply, but stop to
ask "What are all the dangerous things my opponent could do if
I do this, and can I meet them?"
• Try to evaluate resulting positions and learn to pick out the
best one (assuming opponent's best replies) , rather than just
finding a sequence which looks promising and playing it to see
what happens.
• Practice visualizing pieces while looking ahead during slow
moves. While this is a skill at which some players are natu­
rally better than others, with practice everyone can improve
in this area.
There will be a fair amount of overlap between the needs of one
level and the needs of the next, but over several classes those needs will
change, sometimes dramatically.

53
Chapter 4

Class E
his chapter includes players rated 1000-1200 USCF, or roughly
T1150-1350 ICC standard.
For weaker players I primarily gave them de Groot A. Giving a more
complex line such as de Groot C or de Groot Ernie would be less in­
dicative of their overall abilities. Giving them a more strategic position
like de Groot Shafritz would not reveal their ability to analyze "simple"
capturing sequences, which is an early skill to try to master.
For each protocol, I list the de Groot position (for example de Groot
A, de Groot B, de Groot C, etc. ; see Chapter 1 for those positions); the
age of the subject; the subject's rating; and the time the subject spent to
choose his move.
Comments in parentheses indicate outside actions or comments that
reflect the action of the player, such as (silent) if he paused for a long
time. Comments in brackets are my thoughts. My frequent use of [sic]
means the subject is making a clear mistake in analysis or visualization.
In contrast, [ ! ] indicates the player has made a comment very insightful
for their level of ability or a surprising error. A frequent note is [No
eval] meaning the subject did not try to evaluate which side stood bet­
ter, by how much, and why.

Protocol E-1 (de Groot A; adult; 1000; Short time):

The material is even. There is pressure on the black king. There is
an idea ofplaying for 4Jd7. If I play J. 4Jxc6 I win the bishop pair.
de Groot A
Bravo! Simple move l.4Jxc6 wins the bishop pair

54
 Class E

Suppose I play J. tuxdS with the idea of getting in tud7. Then after
1. tuxdS exdS I am not getting anywhere. Suppose I develop the king 's
rook. Then I can wait and see what Black wants to do. My move is
J. 'Sfel."
There is some good promise in this protocol, but one can see why
this player is currently weak. He does take the time to count the mate­
rial (most weak players do not) but he quickly spots the removal-of­
the-guard idea with tud7. Moreover, he is one of the few players at this
level to clearly point out the opportunity to just procure the bishop pair
with 1.tLlxc6. It isn't the best move, but it's better than what he played.
A major mistake was to skip looking for Black's threats, so that 1.. .1&xb2
was not spotted. This also had the effect of making his move "Hope
Chess" since he did not bother to see what Black could do in reply to
his proposed candidate 1.'Sfel. As noted many times, you should spend
a fair amount of time seeing if the opponent can defeat your candidate
move. If you don't, he surely will spend some time trying to defeat it
and might succeed! Subject E-1 did not systematically consider all his
captures - failing even to consider the best move 1.hd5, and reject­
ing 1.tuxd5 apparently just because 1...exd5 does not lose material. He
made his move fairly quickly and pretty much "blind" - no analysis of
1.'Sfel at all, much less a principal variation.
The following is another example of weak players not always being
too fast:

Protocol E-2 (de Groot A; adult; 1170 ; 20 minutes)

Any threats? The bishop on gS threatens the knight on f6 - it is
guarded by the bishop on e7 and the knight on dS. The knight on e5
threatens the pawn on j7 and the bishop on c6. The bishop on c6 is
guarded by the queen on b6, the pawn on b7, and the rook on cB. The
knight at c3 threatens dS with the rook on cl behind it. There is then
a double attack on c6 with the knight on e5 and the rook on cl. The
bishop on a2 attacks dS, which is protected by the pawn on e6, the
knight onf6. If thef-pawn moves the king on gl is indirectly attacked
by the queen on b6 - I have to be careful.
The pawn on b2 is en prise to the queen on b6. Looking at the d4
pawn. The queen attacks it but it is protected by the queen on d3.
Do I have any good moves? 1.hf6 Ah! 1. �h6 threatens the rook.
Any combinations or sacrifices? l. tuxg6 but it is protected by the
pawns on h7 andj7. Can't get the queen in after the capture on g6. Any
way to loosen the position up? J.h4-h5 and then hxg6 hxg6 tuxg6fxg6

55
The Improving Chess Thinker

'?ffxg6 + @hB [waste of time to give White three moves in a row]. Then
il.h6 threatens mate. Look to see if Black can defend this combination.
For example after h5, ... h6 threatens the knightfor thefinal attack. But
makes it easier for the queen. Looking at the bishop on a2 attacking
gB. Too much blockage. One way is j2-f4-j5. Can be taken by f?Jd5.
'ilxf4 might be vulnerable.
The knight onf6 is guarded by two pieces and is OK. I am lulled into
thinking without verbalizing. Candidates lj4 or 1.h4 with the idea of
advancing to thefifth rank. l.1l.xj6 does not merit much; after 1...hf6
and there is no clear continuation. 1. f?ixd5 - there is only one way
to recapture as l . . . f?Jxd5 leaves il.e7 en prise [sic! A common error in
analysis of de Groot A] so he has to retake with 1...hd5 [sic]. J. f?ixd5
threatens the queen and bishop.
What type of attack can I get? Tactical shot. Looking at J. l?ixd5,
also attacked by bishop on a2, guarded by a pawn and two pieces
(quiet). Attack d5; can all recapture back? Win a bishop on e7? 1. f?ixd5
exd5 doesn't look too prosperous - just exchange bishop for a pawn.
Can force him to take with the knight or bishop.
OK, just noticed I have a possiblefork at the end: get knightfrom e5
to d7 to fork queen and rook. That might generate something. J . fixf6
removes a defender; l.. .. l?Jxf6 another defender. One knight can re­
place the other knight. So take d5 first - can capture with knight or
bishop. If he recaptures with the knight then 2.he7 wins [sic]. So
1 ... hd5 2. hd5 and then 3. f?id7. Is that accurate? On J. f?Jxd5 he can
just play l ...exd5. J. f?ixd5 exd5. I don't want to play 2. hd5; he can
take with knight or bishop. Back at J. f?Jxd5 exd5 then before 2. hd5
play 2.hf6 1l.xj6 3. hd5 fixd5 4. l?Jd7. That's the sequence. I make
1. l?ixd5 push clock.

There is a lot to be learned from studying this protocol! Let's begin

by examining his conclusions. The subject made the common mistake
among weak players, believing that he is winning a rook when actually
he is only winning the exchange (a rook for a bishop or a knight) :

See diagram top of next page

56
 Class E

de Groot A
White to move after 1.�xd5 exd5 2 ..ixf6 .ixf6

The subject's PV at this point is 3.hd5? hd5 (not forced by the

way; Black could consider 3 . . . heS first) 4.�d7 forking queen, bishop,
and rook.

de Groot A
Black to move after 3.hd5? hd5 4.�d7

But if Black just guards the bishop, say with 4 ... tfd6, then after the
desired 5.�Jd8 Black can just recapture with 5 ...tfid8 or 5 . . . !ht'B and
have two bishops for a rook and a pawn, a fairly sizable advantage.
Beginners are taught that knights and bishops are, on the average,
Worth three pawns, but that is only a rough integer value. Actually
knights and bishops are each worth - again on the average - more than
three pawns and a bonus for possessing the bishop pair is approximate­
ly an additional lf2 (reference IM Larry Kaufman's scientific method

57
The Improving Chess Thinker

for valuing pieces in "The Evaluation of Material Imbalances," Chess

Life, March 1999), so winning two bishops for a rook and a pawn is
an advantage of about a pawn, which is usually enough for a winning
advantage.
The capture on f8 has won the exchange (worth only about l1/4
pawns), not a rook (worth 5). That's a big error! What the subject has
done is to sacrifice a bishop on move three to win the exchange on move
five. Because he uses the word "rook" he thinks White is coming out
ahead. Over the years, probably a dozen subjects have made this exact
mistake when analyzing de Groot A. This "rook instead of exchange"
mistake provides several insights:
• Weak players often do not differentiate between the following
two situations:
When they capture a piece worth more than the cap­
turing piece and it can be captured back, and
When they make the same capture but there is no re­
capture.
For example, there is a big difference between capturing a
rook with a knight and the opponent capturing back the knight
("winning the exchange") vs. capturing a rook with a knight and
the opponent cannot capture the knight ("winning a rook").
I have seen this mistake made dozens of times, often with
disastrous consequences. In the previous diagram 5.Nxf8 wins
the exchange, not a rook. If the subject saw the recapture but
still misevaluated, then this is still incorrect.
One teaching trick I used was to perform the same exchange
with money instead of piece value. This often worked wonders:
"Would you give up $3.25 for the opportunity to exchange
a further $3.25 for $5? Would you do it if I gave you an extra
$1 (pawn)?"
I think most everyone understood when I phrased it this
way!
• The wording of your thoughts matters! The players say to
themselves the appealing "I am giving up a bishop but I am
winning a rook." More accurate would be, "I am giving up a
bishop and then I am capturing a rook but he gets my knight"
or "I am giving up a bishop but afterwards I am only winning
the exchange." With accurate wording, the trades don't look
so appealing any more. See Chapter 12 for other words which
mislead.
58
 Class E

• Weak players are often OK until they have to calculate a criti­

cal sequence, when often they get very sloppy and sometimes
make enormous errors. Yet they are often unaware they might
be making an error and don't double-check their analysis, es­
pecially analysis that contains several recaptures and precise
counting. Take more time for critical sequences, such as those
involving multiple captures.
This subject also thought he saw a winning line, but that did not
send up a "warning flag." The applicable principle in Alburt and Law­
rence's book Chess Rules of Thumb: Whenever you see a winning line
you need to stop, take time, and make sure it is right. Don't just plow
ahead. There are two reasons for this:
1. If the line is truly winning, then you probably don't need as
much time or effort to finish the game. So making sure it is win­
ning now is more important than playing out an easily winning
position, and
2. If the line is not winning (as was the case above), then you
would like to know why. The reason your candidate does not
win may be enough to convert your previously best move into
a move which isn't very good, and you should not play that for­
merly "winning" move.
In this case the subject had thought for a very reasonable twenty
minutes. Yet once he spotted the "winning" line he did not double­
check it or ask himself if it was truly winning. Instead he terminated the
analysis instantly - a potentially gigantic mistake.
A common mistake Subject E-2 made was to assume after 1.ltlxdS . . .

de Groot A
Black to move after 1.c!OxdS

59
The Improving Chess Thinker

... that the recapture 1 . . . li:lxd5 is impossible because of 2.he7 "win­

ning a piece". This common error is due to one of two possibilities:
1. The player assumes that by moving the pinned piece on f6 that
the piece behind it can be safely captured, since it's original
guard, the knight on d5, has been captured, or
2. The player does not attempt to visualize the position, and fails
to see that the knight on d5 was replaced by another knight on
d5, which is equally able to guard the bishop on e7.

As a result of this and other analytical problems, few players rated

under 1800 USCF ever analyzes the critical line 1.li:lxd5 li:lxd5 2 .hd5
(attempting to remove the guard on e7) 2 ... hg5 with complications.
Players who don't get that far are not getting to the heart of whether de
Groot A contains a winning tactic.
Notice that subjects E-1 and E-2 neither counted the material nor
looked for threats. Subject E-2 eventually did notice his b2 pawn was
attacked and not defended, but doing so near the beginning of the
thought process often saves a lot of analysis time!
Let's contrast E-1 and E-2. They are both rated in the same class;
although E-2 was rated higher at the time of the protocol, E-1 eventually
surpassed him. At the time of the protocol E-2 had been playing a long
time and had a sophisticated but rather flawed thought process. On the
other hand, E-1 had less tournament experience and did not know how
to deeply attack the problem. Once E-1 learned how to handle this type
of analysis, he slowed down and easily surpassed E-2 , who was "mired"
in his methods.
The next subject is very young and, at the time, probably somewhat
overrated. Consider his protocol as possibly belonging in the previous
chapter:

Protocol E-3 (de Groot A, age 6, 1150, less than a minute)

J. 'll,fel
I want it in the game. l. Vffh3. I will move 1. W!h3 Push clock.
-

There is no check for safety or even a reason why the subject likes
the move chosen. Perhaps he wanted the queen to head toward the
weak square h6? ! Of course, at this speed he did not consider any re­
plies, much less such potentially dangerous ones as 1 ... li:lxc3, 1 . . .Vffxd4,
or 1 ... Vffxb2. At least he considered more than one move and did give a
reason for his first candidate.
Lest you start to believe that haste is primarily an attribute of youth,
not all short protocols are youngsters:

60
 Class E

Protocol E-4 (de Groot A; adult, 1050, 2 minutes)

I am looking at the piece positions. l. 'i!.fel - just on general prin­
ciples . Hit clock. [no eval; no analysis; no PV]
Moving on general principles in positions with lots of checks, cap­
tures, or threats may be a fun way to spend an evening, but without
learning to recognize these positions and perform solid analysis on
them you won't become a good player!
This leads to the idea of criticality assessment. Criticality assess­
ment is the ability to discern whether the current move is likely to have
a major impact on the outcome of the game. A cousin of criticality
assessment is a sense of danger. This subject seemed to have neither
and thus moved very quickly. However, de Groot A is a fairly tense and
analytical position, so moving on general principles only works if you
are very lucky.
Protocol E-5 (de Groot A; adult; 1200; short time)
The b-pawn is hanging. The center is important. 1 . hdS l'iJxdS
2. l'iJxdS. J.l'iJxdS. Push Clock. ["What is your PV?"} My PV is J. l'iJxdS
l'iJxd5 2.hdS and after Black recaptures on d5, 3. he7.
Yet another case of assuming the opponent will make a move (like
1...l'iJxd5 after 1.b'.d5) without considering whether it is the best move,
or even the alternatives. Notice that subject E-5, like many weak play­
ers, considers a second candidate 1.l'iJxd5 and chooses to play that sec­
ond move, which he did not analyze, instead of the one he did analyze.
At least he failed to analyze his second candidate verbally, which is all
I can judge. That's why, after he pushed the clock, I asked him what his
PV was. His analysis in the PV is also a mistake, as after 1.l'iJxd5 l'iJxd5
2.b'.d5 - to remove the guard on e7 - he assumes Black must recapture
on d5 and lose the bishop on e7. Instead Black can make the forced
recapture 2 . . . hg5 .
de Groot A
Black to move after 1.l'iJxd5 l'iJxd5 2.b'.d5; Black has 2 . . . b'.g5

61
The Improving Chess Thinker

This is a key line that most players under 1800 never consider,
even though it is critical for determining whether 1.lll xdS is better than
1.b'.dS. See Euwe's analysis in Appendix A.

Protocol E-6 (de Groot A, ICC 1300, adult, 3-4 minutes)

Isolated queen 's pawn. The material is equal. Black's king is slightly
open. 1 ... Y9xb2 is a threat. Can 't push l.b4 - doesn't look quite right.
What else can I do? The rook is undeveloped, but the have to protect
the pawn [?!]. OK, I think I can ... hmm. Want to develop the rook atfl
but pawn under attack. After 1.b4 does he have any sacrificial attack
with l . . . lll xb4? Does not look like it. I get more space. Is there anything
else under attack? OK. l .b4 push clock.
Again the familiar "Hope Chess" idea of not seeing that after l.b4
if captures like l ... lll xc3 can be safely met. Also, the subject takes a
completely passive approach. His single-minded approach to b2 being
threatened: move the pawn from b2.

de Groot A
Black to move after 1.b4
Settling for moving the attacked piece

Instead, the subject could have taken Dr. Euwe's superior ap­
proach:"OK, b2 is threatened, but let's see if I can do something myself
first." Hopefully players of all levels would play a checkmate instead of
guarding a pawn, but that same logic applies to lesser initiatives as well.

Protocol E-7 (de Groot A; adult; 1050; 3 minutes)

J.fi.h6 1. lll d7 to remove guard. J.il.Jif6followed by 2. lll xdS... lll d7. In
between moves? 1. Y9 h3 so try to win material. 1. lll xdS hdS I win a piece
[?!]. Push clock. PV=l.il.Jif6 il.Jif6 2. lll xdS hdS 3. hdS exdS 4. lll d7

62
 Class E

Once again we have a case where the subject assumes the opponent
will make moves that help the subject (rather than moves that will help
the subject's opponent), even if they are not forced. For example, in his
pV Black could easily vary by replying to l.�6 with l ... ltlxf6 or, in the
given sequen ce 2 . exd5 allowing the bishop on c6 to remain guarding
. .

d7. Note that the subject was only winning the exchange (a knight for a
rook) in the PV, yet said he was winning a piece, which usually means
obtaining a knight or a bishop for a pawn or less.
This subject could have done several things better. As happened with
E-2, he thought he saw a winning move but then failed to re-analyze to
see if th is good fortune was correct. He should think, "Gee! I seem to be
winning a piece by force - is that really true?" and re-check his analysis.
Any time you see a line which seems to win, you should always be very
careful because there are only two possibilities: 1) You are correct, in
which case the game is virtually over and the extra time you take will
not cause future time trouble difficulties, or 2) You are incorrect, in
which case the move you are plan ning might not only not win, might
not even be the best move or, in the worst case, could be a monumen­
tal blunder. An example of the latter occurs when you see a move that
seems to checkmate with a queen, missing that a piece on the other side
of the board could just capture it for free. This happened in an Internet
game with a position similar to:

White to move
Not re-checking what appears to be a win

Desp ite having about an hour on his clock, White played 1.Yl?xg7+??
alm ost immediately, thinking it was mate, and then resigned after 1 ...
�g7. If White had not been in such a rush, he might have seen the
bishop o n al and would have avoided such a rash mistake.

63
The Improving Chess Thinker

The next protocol is by a seven-year old-who, in a couple of years,

ended up being one of the top players for his age group in the U.S. But
at this point he was very much an 1100 player, which is great for age
seven.

Protocol E-8 (de Groot A; age 7; 1100; 80 seconds)

1. �h6 'il,feB not enough to exploit dark square weaknesses. If1. l"iJxdS
to exchange pieces but not so good so less pieces to attack king. 1 . 'Il.fel
to go to e3 to h3 attacks h7 to attack or exchange. 1. hf6 to remove the
defender or 1. Wlj3 to hit both knights and that's what I'll do. Hit clock.
Unlike Subject E-7, Subject E-8 is all offense and no defense! He
does not look to see if Black is threatening the b2 pawn and, just as
bad, does not see that his planned l.'?'Jf3 abandons the d-pawn as well.
And those threatened knights? The one on dS is not less safe due to the
additional attack by the queen on f3; the knight on f6 is already guarded
twice as well. So any damage White can do after l.'?'Jf3 would take a lot
more analysis than was provided. In a few years he will do a quite a bit
better.
Subject E-8 jumps around until he hits on the idea l.'?'Jf3 and then
plays it immediately. This "jumping around and then instantly playing
upon seeing one attractive move" is indicative of a lot of low-rated play­
ers. Not only don't they look for a better move, but, more importantly,
they don't take much time to analyze whether the move they are going
to "bet the entire game" upon is as good as they think. Should Black
have a good forcing move in reply (say 1...'?'Jxd4), the chosen move then
turns out not do all that the subject's initial analysis promised.
In his classic work Think Like a Grandmaster, Kotov suggested
trusting your analysis and not re-checking your lines. But several strong
players have scoffed at Kotov's suggestion and advised that if you are
going to play a critical analytical line, it is worth checking to make sure
your analysis is correct. I might add the weaker you are, the more you
should not trust your initial lines. You should especially not trust the
lines for the move you are actually going to play! So reviewing critical
lines when you have plenty of time is an excellent habit.

Protocol E-9 (de Groot A; adult; 1100; 6 minutes)

Counting material - even. His king is a little less safe - not too
much. No imminent danger - no Black captures . . . l"iJxc3 not to o
scary . . . Wlxd4 covered. Where else can he take?. Oh! ... Wlxb2 is en prise .
Any tactics here? 1. �h6 hits the rook - not bad. But the knight onf6
is pinned - hard to evaluate. [Looks for threats] What about. . . l"iJf6

64
 Class E

move? �c6 m oves - opens c-file. Given that, I can 't see anything too
Jan ey - captu re d5. Take with bishop, pawn, or knight, . . . hd5 opens
c-.file. (silen ce). . . W!xb2 so that's possible. Don't want to take the knight
_ g ives up the two bishops [apparently 1. hdS]. Isolated queen's pawn
_ suppose to stay active and support the knight on dS. Protect b2 pawn.
J. !iJb5 shaky. J. 'il,c2; J. l'iJ b5 interesting. 1 . l'iJ b5 shaky. I'd be afraid to do
tha t. Is J.b4 safe? 1.b4 a5 2.bxa5 . . . OK. 1.b4 a5 2.b5 is interesting. 1.b4
is .. .J ... W!xb2 is my main concern. I. Wfc2 W!xd4 bad so I'm thinking J.b4
push clock.
Subject E-9 did a much better job than some Class E players, but
still was mostly passive. He did consider the capture on d5, but rejected
it just because it did not win material. Weaker players often treat nor­
mal positions as if they were problems with possible solutions. When
a capture fails to win, they reject it even if it turns out to yield a better
position that the moves they actually choose! And a capture is more
likely to be safe, as passive moves like 1. b4 require the player to analyze
replies such as 1 . . . l'iJxc3. A capture of a piece is much less likely to lose
material. We have seen this mistake in almost half the protocols in the
Chapters 3 and 4.
To his credit Subject E-10 did not completely play Hope Chess. Al­
though he did not consider the critical capture 1 . . . l'iJxc3 after his planned
1.b4, he did consider some replies such as 1.. .a5.

Protocol E-1 0 (de Groot A; adult; 1000; 8 minutes)

Look to activity on the kingside. What are Black's captures? c3
- can take back with a rook. White can take on dS - would not do
anything. Don't see any glaring weaknesses for Black so far. No piece
undefended. First move to come to mind is I. Wfh3. The knight defends
h7. 1. l'iJ xd5 hdS - not going to take with the pawn - that makes it iso­
lated. 1 . l'iJxdS hdS 2.hf6 and then 3. W!h3 - the bishop on g5 hangs.
l. l'iJxdS hdS 2. hf6 hf6 and I can 't do anything on W!h3. I need to
work out some way to do damage on the kingside. A pawn storm with
h4 to b reak open the g-file. 1.h4 h6 2. �d2 l'iJxc3 to prevent �cl 3. 'il,xc3
1b5 4. Wih3 'il,xc3 [4 . . . hfl?!}. I'll sacrifice the pawn 5.bxc3 push clock
(PV 1.h4 h6 2. �d2)
=

Wow! This is an aggressive, attacking protocol. Too bad Subject E­

l O 's analysis is
not rooted in reality. He does not mention the hanging
P awn on b2, nor does he consider Black's checks, captures, and threats
after l .h4. Of
course, after 1.h4 h6 retreating the bishop to d2 is com­
plete ly hazy.
Why would Subject E-10 assume that Black should play

65
The Improving Chess Thinker

l ... h6 when White can just play 2.hh6? Again we are seeing a player
assume an opponent reply that is at best not forced and at worst just
good for the thinking player instead of trying to find the move best for
the opponent. Maybe E-10 did not "see" 1.h4 h6 2.hh6 but that would
be a good reason for going slow and asking himself if each move were
forced, or at least reasonable, before continuing analysis. Many weaker
players make the mistake of analyzing moves at some depth in their
analysis tree that likely never would be reached for a variety of reasons,
such as this one. Since Subject E-10 bases his final move on this se­
quence, it would be good to re-check to see if it was plausible.

Protocol E-11 (de Groot A; adult; ICC 1220; 2 minutes)

Looking at piece positions. I would play 1. '8fel on general prin­
ciples. [No evaluation, no analysis, no PV] push clock.

de Groot A
U�fel - Just developing, with no analysis

One shouldn't use general principles to make a move in an analytical

position. Many weaker players make this mistake, but if they wish to
improve, they will have to roll up their sleeves and attempt the analysis.
It is true that most weaker players are not that proficient at analysis
and may not have as much natural talent as Kasparov, but they will get
better with practice.
I call making moves on general principles in an analytical position
"hand-waving" because one is waving his hands for effect instead of do­
ing the work. In positions like de Groot A, it's great to develop your
pieces, especially with one clearly less developed piece like the rook at
fl. However, you should analyze the forcing sequences first, as any pos­
sible tactic or safety issue takes precedence over activity issues.

66
 Class E

sum ma ry of Class E
Characteristics of this thought process level:
• The analysis is more concrete than Class F, but the subjects still
do a lot of hand-waving (using principles when only analysis
will do).
• The subjects make many assumptions about opponent replies,
but are either lacking a basis for these assumptions or are just
making plain old bad guesses. At least they are considering
these replies.
• The subjects are trying to see what moves are reasonable in
the position. However, not only are many of their attempts in­
correct, but are sometimes also unnecessary as forcing moves
should take precedence.
• The thought processes are all "Hope Chess" in that moves
are made without seeing if all replies of checks, captures, and
threats can be met. For example, in de Groot A, often a move
was chosen that allowed an advantageous 1 . . . ltlxc3. The conse­
quences of 1...ltlxc3 were not even considered, much less seen
as safe.
Improvements to the thought process that would help in getting to
the next level:
• More systematic selection of candidate moves (this is true of
most lower classes! )
• More checking to see if candidates are safe.
• Consideration of forcing moves first (for both sides)
• Taking more time to identify and evaluate the main candidates
and, even if satisfied, looking for a better move.

67
Chapter 5

Class D
his chapter includes players rated 1200-1400 USCF, or rough ly
T1350-1550 ICC standard.
For each protocol, I list the de Groot position (for example de Groot
A, de Groot B, de Groot C, etc.; see Chapter 1 for those positions); the
age of the subject; the subject's rating; and the time the subject spent to
choose his move.
Comments in parentheses indicate outside actions or comments that
reflect the action of the player, such as (silent) if he paused for a long
time. Comments in brackets are my thoughts. My frequent use of [sic]
means the subject is making a clear mistake in analysis or visualization.
In contrast, [ ! ] indicates the player has made a comment very insightful
for their level of ability or a surprising error. A frequent note is [No
eval] meaning the subject did not try to evaluate which side stood bet­
ter, by how much, and why.

Protocol D-1 (de Groot A, ICC 1350, adult, 5 minutes)

There is a fork 11J e5-d7. I am hitting the knight onf6 with the bishop
on g5 but the bishop on c6 covers dS. I don't see the knightfork work­
ing [No evaluation]. Look at . . . no tactics. Look at moving a rook, like
l . '!J.fel [no checking for opponent threats].
Hmm. Can't see any immediate checks or direct threats on his part.
He can take b2 pawn - maybe what to do with that. One solution is
J.b4. He wouldn't play J... 11Jxb4 to lose a knight for two pawns. That
might be good if I don 't .find a tactic. He can play l . . . il.b5 to skewer
queen and rook but then 2. 11Jxb5 so that's not really a threat.
Probably in this position I would play J.b4 to save b2. I can 't see
any better so J.b4 - longest I would have thought. Still looking at J.b4
hb4 2.axb4 . . . so J.b4 push clock.
This subject, like many others rated below 1600, does no initial
evaluation and thus has no goals or expectations to meet during the
analysis. One big reason to evaluate a position is to figure out what goal s
you have. For example, if you think you are slightly better, then you are
looking for a "best" continuation that leaves you at least slightly better.
Subject D-1 cleverly sees the fork possibility on d7. Yet he does not
logically follow with systematic analysis to discover whether any forcing
sequence can take advantage of the fork possibility. Once he reali zes

68
 Class D

that his pawn on b2 is in danger, he switches his focus exclusively to

trying to make the pawn safe, ignoring possibilities to make a forcing
move (check, capture, or threat) first. Many weaker players don't real­
ize th at if they can make a capture first, then defense of the pawn may
not be necessary. A series of forcing moves might keep the initiative,
and never allow the opponent time to capture the pawn. Instead they
tend to focus just on the defense of the pawn and exclude their own pos­
sibilities. Subject D-1 gets credit for spending some time seeing if the
opponent can defeat his final candidate move 1.b4. Most weak players
never spend any time seeing if their opponent can defeat the move they
are about to play, and thus play "Hope Chess."

de Groot A
Black to move after 1.b4

Unfortunately the subject concentrates only on the pawn moved, and

not on the square it had been guarding, c3. So he completely overlooks
whether Black can defeat 1.b4 with a sequence starting with 1...li'ixc3
and continuing with 2 . . . �bS or 2 . . . li'ie4 or 2 ... �e4. If White does not
examine these Black responses before playing 1. b4, then he is just lucky
if Black does not defeat him with a series of captures!
The same subject a few months later:

Protocol D-2 (de Groot Zyme, ICC 1400, adult, 10 minutes)

"Equal in pieces [sic! Bishop pair] - one more pawn; black king
exposed - may be about equal. Checks are 1. �c7+ or 1. �e7+. None
looks that good - loses a piece. Captures: 1. Vffxg 4 ; 1. hg4; 1. VffxaB in­
teresting - can 't recapture.
Th rea ts for White: Maybe get knight away on g4. Black's checks:
dang erous 1... Vffxc3+ wins rook. 1... V!Jxf2+ dangerous. 1... Vffxd6; 1...
C/jxfi. Lots of dangerous threats.
69
The Improving Chess Thinker

1. V!lxaB then 1 . . . V!lxc3+ 2. Wdl only move 2 . . . V!lxal+ orfork 2 . . . 0.xf2 +

[mate]. Dangerous to take 1 . V!lxa8. 2 . . . V!lxal+ and then . . . 0.xf2 loses
two pawns by taking the rook. So 1. V!lxaB probably not possible. 1 . �eS
pins the queen to the rook but 1 . . . 0.xeS. A safe move to cover the square
is 1. V!ld4 to protect bishop on d6, c3, and j2. If I don't .find anything
better I can do that [King of the Hill].
Can also play 1.f3 but that does not cover 1... V!lxd6 then 2. 'll. dl but
1... V!lxc3+ is not good. Hmm. Can 't find anything much better than
1. V!ld4. That threatens 2. �eS+ winning the queen so . . . I don 't think J
can.find anything better so 1. V!ld4 push clock."
The de Groot Zyme exercise is even more tactical than de Groot A,
and thus one would expect that weaker players should take more time
on this protocol. The subject gets credit for starting by considering his
checks, but his fatal flaw is not adding up two key pieces of information
which he properly gathered: l.�e7+ is guarded by l...V!ff6 and l.V!fxa8
fails to l.. .V!fxc3+ . The key to this problem is realizing that if White can't
take the rook because of the counterattack on c3, the cost to deflect this
attack, a bishop, is less than the gain, the rook on a8. Thus the subject
fails to realize that although each piece of information in itself seems
enough to prevent 1.�e7 + or 1. V!fxa8, together they indicate that playing
l.�e7+ first allows 2.V!fxa8.
This type of deductive logic is partly learned, partly genetic. Even
very brilliant people who have only been playing chess a short time
often miss easy chess logic because they are not familiar with the para­
digms of the game. Innate capabilities have to be cultivated with years
of good experience to become a strong player.
The move played, l.V!fd4, is not acceptable, as after l...V!fxd4 2.cxd4
the endgame is promising for Black due to his queenside pawns:
de Groot Zyme
Black to move after 1.Yfd4 Yfxd4 2.cxd4

70
 Class D

Unfortunately, although this was Subject D-2's chosen move, he did

not analyze the consequences of playing it. This is unfortunate because,
although D -2 just spent 10 minutes finding this move, none of that time
was devoted to investigating whether he could save the game after it
was made. If your intended move leads to a lost position, then should
search for a better one. Of course, if your candidate were the only legal
move or the only move to save the game (''forced"), then you would
just play it. Since here it White's move is not forced, the consequences
should be weighed against those of other moves.

Protocol D-3 (de Groot A, 1250, age 17, 16 minutes)

Black's dark squares and king weak - I could have a fork on d7.
J. hj6 hf6 2. hdS exdS fork at d7. J . hf6 hf6 2.hdS exdS 3. CiJxd5
fi.xd5 4. t:ud7fork wins the exchange but I gave up a piece for a pawn
so I am worse giving up two pieces for a rook and pawn.
J. fi.h6. Better to take knight on d5 first. J. CiJxdS hdS 2. hj6.
J. CiJxd5 exd5 2.hf6. J. hd5 exd5 hf6 hf6 Don 't want to give up
dark squared bishop. Don't like giving up knight on e5. 1. Vffj3 but he
has threats. 1.f4{6 and 2. Vffj3. J . hf6 hf6 2. Vffj3 not that good. 1. Vffj3
and 2. hf6 CiJ:if6 hits the queen.
1. CiJxd5 t:uxd5 2. hd5 exd5. 1. CiJxg6 does not look good. I. Si.bl never
mind. J. CiJxdS then Black can 't play 1 . . . t:uxd5 due to 2. he7 [sic!].
1. CiJxdS exdS does not accomplish anything on 2. fi.f6. 1. Vff h3 not much.
1.CiJxd5 exd5 J. fi.h6 is King of the Hill [but no PV]
1.'lk2 don't know if it makes sense. 1. '8fel with the idea of '8e3-h3
after 2. CiJxdS to remove knight. 1... Vffxb2 is a threat. Then 2. '8bl could
be dangerous for him: 2 ... Vffxa3 [2. . . Vffxc3] 3.'8b3 escapes - no good
for me. So 1. '8c2 protects pawn - allows doubling on thefile.
l.b4 with the idea of a4-b5 1. '8c2 is King of the Hill. 1. '8fel makes
sense but 1... Vffxb2. There must be something better than 1. '8c2
1. fi.h6 '8fd8 no follow-up. 1.b4 defends - that threatens b4-b5 to
hit bishop and then I can possibly get in t:ud7. J.b4 weakens c3 and c4.
l. '8bl helps advance - relinquishes file [not considering replies except
on captu res].
1. liJ xd5 exdS 2. '8fel hits bishop on e7 on discovery. That's King of
the Hill. 1 . CiJxd5 exd5 2. '8fel Vffxb2 no good. My PV is J. CiJxdS exd5
2. '8fel push clock.
Subj ect D-3 used less description and more "algebraic notation" than
most. Like almost all others at this level, he does not begin by counting
th� m aterial or making an evaluation. He begins by immediately identi­
fyi ng candidate moves without attempting to identify threats generated

71
The Improving Chess Thinker

by the opponent's previous move. Subject D-3 gets very high marks
for trying to make forcing moves work. But his approach is unsystem­
atic and he does not list all possible recaptures with the goal of finding
which recapture is best. This scattershot approach is not likely to find
anything except by accident. Credit also goes to this subject for using
the King of the Hill method, in which he identifies the best move found
thus far. However, it is not clear how he compares candidate moves
and their resulting positions to replace his King of the Hill. The process
almost seems to happen by magic.

Protocol D-4 (de Groot Zyme, 1200, adult, 11 minutes)

Black is ahead a pawn. White has the bishop-pair. The open cen­
ter is helpful; need to finish my development. Black's rook on aB is
hanging. What happens if 1. WfxaB? He can check onf2 - I move away.
1. WfxaB. . .My checks, captures, and threats? Don 't believe, so 1. �c7+
©xc7 ruins his chances of castling [sic] but I lose the bishop. Any
other checks? No [!]. 1. WfxaB he has 1... Wfxj2+ . How would I respond?
2.©dl ILJe3+ then king moves: 3. ©cl Wiel + chases me up the board.
4. ©b2 no more checks. ..4 ... ILJcl+ ©a2 and the knight is pinned to the
queen. 1... Wfxc3+ is a fork, so 2. ©dl Wfxal+ 3. ©d2 What can he do ?
[Note: does not matter much - things are already bad]. 3 ... Wfa2+
4. ©cl Wfal+ is a challenge. Can he bring anyone else over to help?
Not quickly. What else? 1 ... ILJxf2 then 2. '8gl can 't castle kingside (si­
lence). Going back looking at a check 1. �e7+ forks king and queen
but then 1 ... Wixe7 2. Wfxe7+ [??] 2 . . . ©xe7 is no good [!]. Threats are
1... Wixd6 and 1... Wfxc3+. If 1.0-0-0 ILJxf2. 1. Wfd5 protects the bishop.
Two ways to protect the bishop. I can play 1.f3 to threaten the knight.
1... Wfxd6 2fxg4 threatening 3. '/'!,dl [!]. He can never take the bishop
because of '8dl. ls that true? 1. WfxaB Wfxd6 2. '8dl is no good for him .
1. WfxaB Wfxj2+ - been through that. What moves are better? 1.0-0-0
is possible - he can 't check [?]. But there he has a fork - no good. So
1. WfxaB push clock.

Subject D-4 played 1.VBxa8 even though he saw the best reply
1.. .VBxc3+ . Just the line 1.VBxa8 VBxc3+ 2 .©dl VBxal+ should have been
enough to scare him away even though he (and many others drawn to
the hanging rook on al) missed the crushing . . .

72
 Class D

de Groot Zyme
Black to move after t.Y:Jxa8 �xc3+ 2.lt>dt

2 ... �xf'2#

Notice also how quickly Subject D-4 glossed over the best move,
1.ie7+ , which after the forced 1.. .i&xe7 now allows his actual move
safely 2.i&xa8. He listed the checks, incorrectly noting that 1.�c7+ lt>xc7
stops castling, failing to notice that the Black king is already on d8 and
therefore could not castle anyway. The better check 1.�e7+ is "seen"
as so bad (or not seen at all) that the answer to "Any other checks?"
is "No." This is probably an excellent example of a quiescence error:
the player "saw" 1.�e7+ without mentioning it, saw the move appeared
unsafe, but did not realize that the deflection of the queen plus the loose
rook on a8 added up to quite a reasonable move.

Protocol D-5 (de Groot A, USCF 1380; adult; 3112 minutes)

l. CiJxd5 [No eval] CLJxd5 2. hd5 hd5 3. CiJ d7. 1. /:iJxd5 /:iJxd5 2. hd5
hd5 3. /:iJd7 i&a6 4. '?f!xa6 bxa6 5. /:iJxfB wins the exchange, push
clock.

Subject D-S's brevity makes his protocol seem more like some of
those in Class E. The subject correctly decides to consider a capture
- after neither evaluating the position nor looking for the opponent's
threats - but then just assumes one of the three recaptures. Worse, he
an alyzes his only line incorrectly, as he needed to spot 2 . . . hg5:

73
The Improving Chess Thinker

de Groot A White to move after

1.�xd5 �xd5 2 ..ixd5 .ixg5

The subject does not consider any other recaptures on move one, nor
did he consider 2 ... exd5, leaving the bishop on c6 to guard d7. I guess,
like many other weaker players, he either expects his opponent to play
the moves which are good for him (White) or his unforced removal of
the guard of d7 is wishful thinking. Later, if Black plays a better move,
White's explanation will likely be, "I did not see it," rather than, "I did
not even consider it."

Protocol D-6 (de Groot A, ICC 1500, adult, 7 minutes)

Count material. Seven pawns to six [?!]. Down material [!]. Iso­
lated queen's pawn. Look at his checks, captures, and threats, then
mine. 1... Wfxb2 undefended. Queen attacks d4 - defended. Bishop at­
tacks a3, defended. Threats? 1... Wfxb2, h6 not a threat. So. How would
I defend the b-pawn or is that the right answer? My pawn structure is
OK. Three pawn islands. Probably protect the b-pawn and perhaps at­
tack on the dark squares double on the c-file. OK. I think I'll look at J .b4
then 1.. .aS trade there. 1. 'i'J.c2 with the idea of 2. 'i'J.lcl protects b-pawn.
I like that better. I can play 1.h4. Trade knight on c3 for knight dS. I
can trade 1. CiJxc6. I don 't. .. 1. fiJxdS is defended three times, attacked
twice. Does not work. Although fi.gS pins knight to bishop on e7. So
really only defended twice, and dS attacked twice. Like better 1. 'i'J.c2
with the idea of 2. 'i'J.lcl although ... If I double he could play ic6-a4 but
then CiJxa4. Going to play 1.'i'J.c2 push clock.
Subject D-6 did a great job by beginning with a material count. Un­
fortunately, he miscounted and thought he was losing. Of course, if th at

74
 Class D

were so, then a good player would know to look for aggressive moves
first (not a bad idea at any time) and not settle on passive moves which
just defend the b-pawn. As many low-rated players did, he dismissed
the capture 1.ti:lxd5 primarily because it is defended adequately, as if
this were a "play and win" problem. In most cases it is ironic that the
players who reject captures often choose a move (in this case Uk2)
which also does not "win" material. Apparently they more highly value
non-captures that don't win material over captures that don't win
material. Captures and non-captures should be weighed at least equally
to see which produces the better position.

Protocol D-7 (de Groot A, 1350; adult ; 112 minute)

I need to do J.b4 to protect b2, push clock.
This was perhaps the shortest protocol, but not the worst. At least
Subject D-7 took the time to determine the opponent's threats and
meet them. Unfortunately he did not follow Lasker's advice: "If you see
a good move, look for a better one," nor did he attempt to see if he had
anything better than passive defense. If he'd had an easy combination
to win, he would have missed it while concentrating on defending the
b2 pawn !

Protocol D-8 (de Groot A, 1300, adult, 5 minutes)

l. tll xc6. Messy. The bishop at e7 is vulnerable. 1. tll xd5 hd5; if in­
stead 1... tll xd5 then 2. he7 [wrong].

de Groot A:
White to move after 1.�xd5 �xd5
The recapturing knight on d5 now guards e7. Yet many weaker play­
ers either visualize this incorrectly or assume that, with the piece on f6
gone and the knight on d5 taken, the bishop on e7 must hang.

75
The Improving Chess Thinker

After J... (/JxdS 2. il.xdS exd5. Is that good? Isolating the d-pawn.
Other weaknesses? 1.b4 loosens c3. 1. (/Jxc6 wins the bishop pair; is the
knight stronger than the bishop? The bishop on c6 hits g2, so it is a
good idea to win the bishop pair. He can take back with the rook - no
problem. 1. (/Jxc6. Looking for better. 1. (/Jxc6.
Before giving the de Groot exercise to my students I almost always
make sure they understand the value of the bishop pair. According to
GM Larry Kaufman, who did scientific analysis to assess the values of
the pieces, the bishop pair is worth, on the average, about a lf2 pawn
bonus. In de Groot A one can win the bishop pair with l.(/Jxc6. Yet most
weaker players don't even mention this candidate, much less play it. So
a lot of credit to subject D-8, who, although he did not look for a better
move, at least chose one resulting in something positive. It would have
been much better to thoroughly analyze all the forcing moves to see if
one was better than L(/Jxc6. Like many other subjects, D-8 did not do
an evaluation at the start (Who stands better, by how much, and why?)
nor did he look for his opponent's threats. So l.(/Jxc6 may have been
"lucky" in that he did not notice that 1...VBxb2 - or anything else - was
threatened.
The following subject was a youngster who was just starting lessons
and USCF play. By age 16 he was a USCF expert and co-PA State High
School Champion.

Protocol D-9 (de Groot A, 1200+, age 11, 6 minutes)

The square h6 is weak. 1. (/Jxg6 fxg6. No [No evaluation]. 1. (/Jxd5
BxdS. 1. (/Jxc6 i'lxc6 2. (/Jxd5 Shows how little I think. 1. i'lfel develops.
1. (/Jxg6 no, so 1. i'lfel - no - 1... Wfxb2 is a threat. l. i'lc2 to double. What
does Black have? l... (/Jxc3 and skewer (on b5)? 2.bxc3 skewer [sic?]
and 2. i'lxc3 Wfxb2. The d-pawn is weak; maybe 1.b4. If I let him play
1 ... Wfxb2 then 2. i'lbl Wfxc3 so guard him with 1.b4. 1 . (/J c4 �b5? Guess
not: 2. (/Jxb5 Wfxb5 the knight is pinned; does not look good. So 1.b4 hit
clock.
Subject D-9 shows very early signs of becoming a good player. Al­
though there was no evaluation and the identification of threats came a
little late, he begins by looking at his aggressive captures, although he
cuts l.(/JxdS short does not mention winning the bishop pair on l.(/Jxc6,
and never considers the best move l.�dS. A big dose of Hope Chess as
he does not consider whether Black has any check, capture, or thre at
that can beat l.b4 - especially the dangerous 1.. .(/Jxc3. The analysis giv­
en, while sketchy and incomplete, is at least almost entirely accurate.

76
 Class D

Protocol D-1 0 (de Groot A, 1200, age 9, 1 minute)

1. fi.xf6 hf6 2.(/JxdS flxdS 3.flxdS exd5 4.(/Jd7 push clock
This protocol, wherein the youngster stops analyzing after seeing
a fork, encapsulates in one line what is wrong with many fast, weak,
and/or young players:
• They don't follow Lasker's Rule: "When you see a good move,
look for a better one."
• When they see a winning line they just play it instead of follow­
ing the advice (from GM Alburt and others): "If you see a win­
ning line, then that is a critical move and you should spend lots
of time." The idea is simple: If it really is winning then using
extra time won't hurt. If it is not winning, then the move may
not even be a good one, so it is important to look for a better
one.
• They play a line that is completely unforced, expecting the op­
ponent will gladly play bad moves for them. Of course, the op­
ponent is trying to find his best moves, so assuming they will
play the ones you want will not lead to accurate conclusions.
For example, here Black can vary on the very first move, re­
capturing on f6 with l . . .(/Jxf6 instead of l.. ..bf6. Or on move
two Black can retain his bishop as a defender of d7 by recap­
turing on dS with the pawn with 2 . . exdS instead of with the
.

anticipated 2 . . .bdS. When you have a winning line, it has to be

.

winning in all variations, not just the ones that you would like.
• They play very quickly as if to see what happens, rather than
trying to figure out what could happen if the opponent played
well. When you ask players who have this problem why they
played so quickly, you often get an answer like, "I thought the
line won," which, of course it did not. But even if it did, they
should have all the more reason to check it and make sure.
If you want to play quickly and superficially for fun, by all means go
ah ead. But if you purchased this book with the idea that it could make
You a better player, then avoiding the thought process displayed in pro­
tocols like this one is definitely a step in the right direction!
The next two protocols, done back-to-back, were by a normally very
slow player. The first one was uncharacteristically very fast. His haste
shows that not only do some players "show off' by moving uncharac­
teristically slowly for the exercise
(understandable, but not helpful in

77
The Improving Chess Thinker

diagnosing their "normal" problems) but the opposite is also occasio n­

ally true: slow players sometimes play too quickly.

Protocol D-11 (de Groot Zyme, 1200, adult, 2 minutes)

Looking at a mate in one - My bishop is attacked, so . . . Do I ha ve
any checks, captures, or threats here? No. I am going to save my
bishop and prevent mate with l. fJ.. g3 push clock (after moving: "Oh !
I lose a pawn ...) .
de Groot Zyme
Black to move after 1.fJ..g3

The subject not only did not count the pieces or do an evaluation,
but he did not look for all the threats, such as 1.. .�xc3 2.l!ldl lll xf2 # ,
but h e did not even see h e could possibly play 1.�xa8. Purely defen­
sive, and not much of a defense at that. Subject D-11 did exclaim, "Oh!
I lose a pawn," immediately after he moved when he belatedly spot­
ted 1.. .�xc3+ . Many players play like this. To get to the next level, they
have to consistently play more slowly and become more systematic in
searching for both sides' forcing moves. A wise move is to realize, "I am
a fairly weak player, so I should take my time, because when I move fast
I often make big tactical mistakes. "

Protocol D-12 (de Groot Ernie, 1 2 0 0 , adult, 2 minutes)

White is a pawn ahead, 6-5. I am looking at the board and trying
not to miss any pieces. The two rooks are on open.files. I have a cen­
tralized queen and a well-placed bishop. His rook attacks my bishop
and his knight attacks my queen. There are light-square holes aroun d
his king for my queen and bishop. I need to get in there. How do I

78
 Class D

save my bishop and queen at the same time? Umm. Let's consider the
queen .first: where can she move? Anywhere on the fifth rank? Looks
like the a, b, and e files. Umm. Or she could retreat to d4, d.2, or dl
(silen t). Can 't put. . . If I keep the queen on the a or b-fi.les I can protect
th e bishop, but the problem is that the pawns on a7 and b7 can chase
me - so tha t won 't do any good. I can move 1. 'ff e4 to protect the bishop
th at way - the attacked knight is covered by the black queen. Do I have
any th reats right now? Black's king is protected by the pawns at h6
and g7, which are both dark squares - I have a light-squared bishop
_ J can 't move the rooks or check, else I lose my queen. So ... (silent)
Umm. Need to move the queen to e4 to protect the bishop and get the
queen out of the attack with the knight. I think that's the best I can do
right now. No checks - can 't capture anything. l. hd7 is not good - he
just takes. My knight is too far away to do any good right now. So I
will play 1. 'f!e4 (hesitates; silent). 1. 'f!e4.
This is an excellent example of Hope Chess, as Subject D-12 never
considered what Black could do to defeat his intended move. Indeed,
after 1.'ffe4 the reply l. . .dS!

de Groot Ernie
White to move after 1.lre4? d5!

. . . hits the queen and attacks the bishop a second time, winning a
.
piece. An im portant tip: In analytical positions your opponent will
spend almost 100 % of his time trying to defeat your move, so to make
a mo ve a nd not spend ill!!l of your time trying to see if he would be
s� ccessful (and, if so, discarding your candidate move) is often quite
disastrous.

79
The Improving Chess Thinker

Summary of Class D
Characteristics of this thought process level:
• Good : A lot more analysis and less hand-waving than at the
Class E level. In accordance with those better actions, more
time is taken to make a move.
• Bad : The analysis is often slipshod, illogical, and/or unsys­
tematic.
• Hope Chess still reigns; hardly anyone spends any time seeing if
the opponent can defeat his selected candidate move. This leads
to the #l lesson I learned as a beginning tournament player:
It's not enough to be smart and be able to solve "problems"
better than your opponent. Some problems are insoluble and
some threats are unstoppable. So to allow such problems/
threatsjust loses instantly - they hav e to beforeseen and pre­
vented because to waitfor, "I didn't see that!" is too late.
Improvements to the thought process that would help in getting to
the next level:
• Less assumption of what the likely reply is by the opponent and
more consideration (or listing) of all the reasonable replies. For
example, in de Groot A after 1.liJxd5, it would be good to list
the three recaptures and figure out which one is best for Black.
1.liJxd5 is only as good as the best recapture allows it to be.
• Don't eliminate captures just because they don't win material.
They often keep the initiative. Chess is not always a "play and
win" puzzle.
• Take time to evaluate (who's winning, by how much, and why)
at the start of exercises. That provides a strong basis for decid­
ing what needs to be done.
• Take some time to see if the chosen move can be defeated by a
tactic. If so, discard the move.
• Search for all the opponent's threats before beginning the se­
lection of candidates. Don't take an inordinate amount of time,
but don't stop after you see the first threat either.

80
Chapter 6

Class C
his chapter includes players rated 1400-1600 USCF, or roughly
T1550-1750protocol,
ICC standard.
I list the de Groot position (for example de Groot
For each
A, de Groot B, de Groot C, etc.; see Chapter 1 for those positions) ; the
age of the subject; the subject's rating; and the time the subject spent to
choose his move.
Comments in parentheses indicate outside actions or comments that
reflect the action of the player, such as (silent) if he paused for a long
time. Comments in brackets are my thoughts. My frequent use of [sic]
means the subject is making a clear mistake in analysis or visualization.
In contrast, [!] indicates the player has made a comment very insightful
for their level of ability or a surprising error. A frequent note is [No
eval] meaning the subject did not try to evaluate which side stood bet­
ter, by how much, and why.
The first two protocols are by the same student a couple months
apart, as he was pushing past the upper end of the class:

Protocol C-1 (de Groot Zyme, 1570 + , adult, -10 minutes)

If U!JxaB then if 1... V!Jxd6+ OK. 1 . . V!Jxj2+ 2. ltidl liJe3+ does not add
.

anything. 1... V!Jxj2+ seems only reasonable. 2. ltidl liJe3+ don't see any­
thing good for Black after 2. ltid2 So what? Any others? 1. V!JxaB V!Jxj2+
2. ©dl everything else is guarded. No queen checks, Black can't castle.
1 li:Je3+ doesn't do anything. What am I missing? 1. V!JxaB V!Jxc3+ 2. ltidl
...

de Groot Zyme
Black to move after 1.Y!Vxa8 Y!Vxc3+ 2.�dl

81
The Improving Chess Thinker

2 ... Wfxal+ 3. lflc2 or lfld2. Then 3 . . . Ci:Jxj2 rook moves not so good
for White. 1. WfxaB Wfxc3+ 2. ltldl Wfxal+ lfld2 Black is up three pawn s.
No continuation for White. 1. WfxaB Wfxc3+ 2. ltldl Wfxal+ 3. lfld2 0ixj2
White is not threatening i.e7+. Bad for White. l. i.e7+ Wfxe7 2. Wfxe7+
ltlxe7 can't be goodfor White [sic]. J..i c7+ not decent. l.h3 hits knight
- bishop still hangs. 1. WfdS then still I . . . Wfxc3+ 1. Wfd4 stops the threats.
Black's king is in trouble.
1. Wfd4 Wfxd4 2.cxd4 not bad for Black. White is down a pawn . Not
particularly desirable. Three connected passed pawns and the knight
is not trapped. Can go tof6. Hmm. Don't like any of this. 1. WfxaB Wfxc3 +
2. ltldl Wfxal+ 3. lfld2 doesn't seem good. So I don't like 1. WfxaB. 1. Wf dS
then c3 is vulnerable. 1. Wfd4 holds but a bad endgame. Chasing knight
is no good - no other tactical threats. So choose between 1. WfxaB+ loses
pawn and bad position, interesting. But 1. Wfd4 is just sedately losing.
1. WfxaB is wild and wooly but bad 1. WfxaB Wfxc3+ 2. ltldl Wfxal+ 3. lfld2
Ci:Jxj2 push clock.
As we have seen with weaker players, they are so enamored with
the opponent's 1.Wfxa8 Wfxc3+ 2.lfldl Wfxal+ (which should be enough
to dissuade them from 1.Wfxa8) that they don't look for the better 2 . . .
Ci:Jxf2 # . What's interesting i s that since 2 ... Wfxal+ i s strong enough to
eliminate the entire line, the need to find 2 . . Ci:Jxf2 # is moot. The only
.

time it would not be moot would be the case illustrated by C-1, where
for some reason 2 ... Wfxal+ is not enough to dissuade the subject, and he
chooses that line anyway. In that case, finding 2 .. . Ci:Jxf2 # is important to
eliminate 1. Wfxa8.
Note that even at the Class C level the players don't always start by
evaluating the position. Subject C-1 had no idea that Black was ahead a
pawn at the start, and so just trading down was not necessarily good for
White. Even in a game, where players should have a better handle what
is going on, this same problem persists: weaker players trade down
even though it is bad for them because they don't stop to evaluate the
position. If your "best" move makes the resulting position easier fo r
your opponent to win, that is a great reason to question it.
Finally, kudos to C-1 for correctly evaluating that 1.Wfd4, whil e it
stops all Black's threats, is insufficient due to l.. .Wfxd4 2.cxd4 with a
good endgame for Black. It is not enough to stop all the threats - if th e
result is a likely loss then searching for something better is a good idea .

Protocol C-2 (de Groot C, 1590+, adult, 11 minutes)

Material: no pieces trapped, no tactics. White's king is ope n . Is

82
 Class C

Wh ite threatening? He threatens the e-pawn. l . . . eS 2. hcB with the

idea of 3.dxeS loses a pawn. Does not seem to work. Like to activate
the ligh t squared bishop. How about l...b6? Still leaves the e-pawn
ha ng ing. If2. he6+ he6 3. Wixe6 but knight onf3 hangs [does not see
check].
de Groot C
Black to move after 1 ...b6 2.he6+ he6 3.'ti'xe6+

So don't have to worry about it yet, so . . . EleB is possible. Then 1 ... EleB
2.d5 Wid7 doesn't work - no it does - it doesn't. How to defend? l . . . EleB
2.d5 exdS is OK. So l . . . EleB 2.d5 exd5 is goodfor Black. So l ... EleB with
the idea of2 . . e5. l . . . Ele8 2. /Ug5 what is the move after l... EleB? l... EleB
.

2 . . e5 3.hcB exd4 4. /Uxd4. Tough to think through. l ... EleB 2. ? e5 Oh,

.

l . EleB 2.any e5 3. hcB ElaxcB 4.dxeS dxeS 5. CiJxeS Wixhl 6. Elxhl ElxeS
..

7. 1/ffxe5 CiJd3 But what if he does not capture on eS? Not good. Oh bug­
ger! Complex so analysis is meaningless. So 1... EleB hit clock.
I must have mentioned something about evaluating the position
between protocols C-1 and C-2 because this time the subject counted
the material first. The difficult de Groot C position causes everyone
problems and he handled it passively. After a couple of initial attempts
to lo ok at ideas like 1.. .e5 (reasonable) and 1.. .b6 (strange), he honed in
on the passive 1...Ele8 and never really tried to find a better move. He
was trying to prove that his move was reasonable rather than making
any attem pt to show it was best.
. In many situations (especially in quiet positions orfaster games) it
15 okay to
just to find a "good" - or at least reasonable - move. But to
do s o in sharp positions is often a crucial mistake. In sharp positions,
the expected
result of the game after the best move may differ greatly
from that
of the second best move. For example, the best move might

83
The Improving Chess Thinker

be the only winner, while the second best move draws. Or the best move
might ensure an easy win whereas the second best move requires gre at
technique. In these protocols most of the positions are more critic al
than the average position, so at least some attempt to find a very good
move is justified.
Note that Subject C-2 terminates his thought process not because he
feels he has found the best move or has taken enough time, but rather
because, "Complex so analysis is meaningless." However, the right idea
is to play relatively quickly in non-complex (non-critical) positions so
as to save your time for analyzing complex, critical positions. If you
find yourself giving up because the analysis is too complex, but have
lots of time remaining which can be applied to this move, perhaps a
short break to get water might be better than picking a random move !
Remember, you may never be as good as Kasparov at analysis, but
that does not mean you can 't get better with practice. Practice may not
make perfect, but it sure helps, so practice as much as you can.

Protocol C-3 (de Groot A; 1500, adult, 11 minutes)

Hanging? Immediate attention knight c3 attacked by knight dS.
Pawn hangs at b2. His knight atf6 is under attack; the bishop at c 6
is attacked. Probably not going to worry too much about b2 - not an
immediate concern. I like 1. £i:Jxc6 to get the bishop pair [!] There a re
immediate tactics which come to mind. Fair amount points to king.
g6 weakens king. 1. £i:Jxg6 or something to make it better. 1...fxg6 not
exciting to follow-up. So 1.h4 is a candidate. I like it as a possibility.
J. Ei::i xc6 and l.h4 candidates - bothforcing enough that 1 . . . Wffxb2 is OK.
Think more. Prefer J. Ei:ixc6 less so J. Ei:ixc6. I am thinking he might lose
a pawn if. .. If J. 'ifrc6 can I win a pawn on dS? [!?]. No, f6 knight is on
it. So l ... !'lxc6 is possible. So think about l. £i:Jxc6 bxc6 what would be
the follow-up? Don't see anything exciting. Bishop pair. The knight on
e5 is in a good spot, but bishop c6 was pointing at my king. On the
whole, probably not my best move. Back to l.h4 Is hS a good follow?
No, because of the knight atf6. So what is thefollow? Want to play hS.
So I need to get knight onf6 to get off defense of hS. So 1. Ei:ixdS Ei:ixdS
but 1.h4 with the idea of 2. Ei:ixdS so I like h4 more. l .h4 response? l . hS . .

any follow-up? [not captures] 1.h4 h5 2. Ei:ixg6 fxg6 Wffxg6+ may be OK

So l...hS likely not good. What is good after 1.h4? Think ahead: l. h4
Wixb2 2. fibl is on g6, too. Not sure what the bishop on a2 does. lf I
am going after the king, then the bishop is better on bl. So 1.h4 Wixb2
2. !ibl. I don't know. He plays . . . ? 2 . . . £i:Jxc3 3. !'lxc3. It's all getting w ild

84
 Class C

a nd I do n 't see the consequences. My play is still on the kingside - l.h4

still looks good. I could set it up with l. fi.bl first but l.h4 looks more
in teresting. So l.h4 push clock (PV=l.h4 '?!fxb2 2.fi.bl . . ?)
.

subj ect C-3 did not count the material, but did survey the position
and found that b2 was attacked. He gets good marks for deciding not
to worry too much about b2 (yet) and noticing he can win the bishop
pair. However, he quickly fixates on trying to make a pawn break with
h4-h 5, which in many fianchetto positions is reasonable. In positions
like these with so many forcing moves, such slower considerations are
usually subservient to the immediate forcing moves: checks, captures,
and threats. Not considering any capturing sequence other than 1.liJxc6
is inefficient and lowers the probability of finding the best move.
Here are three in a row from another student. On the first he is al­
most Class C, but by the third he is almost above it!

Protocol C-4 (de Groot A; 1350; age 10; 2 minutes)

The knight onf6 is pinned by the bishop if you get rid of the knight
on d5 [No evaluation]. 1. fi.h6 weak dark squares. Can move open to
half-open.file. May need to protect b2. Can move l . liJ e4 - set knight up
to a good square. Don 't see anything e/se. 1. liJxdS...
A spotty protocol that was not completely finished, but that hap­
pens when you are ten years old, talented, and not so verbal. Sometimes
talking out loud is distracting, especially for a process that one usually
does silently. Subject C-4 did not count material but did do a rough as­
sessment before mentioning a candidate move. Also to his credit is that
he spotted the hanging b2 pawn and did not state that he had to defend
it - he properly noted that he may need to do so. Far more experienced
players have made the mistake of being too black or white on such is­
sues - keeping an open mind with "may" is a great idea.

Protocol C-5 (de Groot Zyme; USCF 1500; age 1 2 ; <5 minutes)
The rook on aB hangs. Black threatens f2 with the queen and the
knight. lj3. Wait he also threatens 1 ... '?!fxc3+ winning the rook. l. fi.e7+
to win the queen ? No, need to stop l . . . Qxc3+ by 'fl.cl, 'fl.a3, or c.!i d2. lf
there was not a knight on g4, then fi.eS skewers queen and rook. 1. '?ff d4
stops 1 '?!1xc3+. Could play J. fi.c7+ c.!ixc7 don't know if that's good.
.. .

The n - no 2 . 'fl.a6, no ha6. J. fi.e7+ '?!fxe7 2. 'fl.dl+ fi.d7 3. '?!fxaB+ and

2 ·· - � eB 3. '?!fxaB and no attack on c3 or f2. That's about it. J. fi.e7+ hit
clo ck.

85
The Improving Chess Thinker

de Groot Zyme
Black to move after 1 . .fl.e7 + !

The two years meant a lot! The subject still did not count material
but he quickly spots all the key tactical issues. Although at one point
he finds the passive 1.\Wd4 to meet all the threats, unlike Subject D-2
he properly looks for a better and more aggressive answer and surveys
his checks. In doing so he does not commit the quiescence error that
weaker players make and instead deduces the right idea 1..ie7+. Bravo!
I suspect this player will be well above 1600 in a few years . . .
Protocol C-6 (de Groot Ernie; 1550; age 1 2 ; 7 minutes)
White's queen is threatened by knight on f4. If 1. WleS Wlxc4 2.'8e4
wins the knight - no, then 2 . . . t/Je2+ or 2 . . . Wlxc3 - doesn't work. I. Wff cS
Wlxc5 2.bxc5 '8xf5 tripled isolated pawns and down a bishop - hope­
less. ]. Wld4 pressures g7 and allows '8e7. Wait! 1. Wld4 '8xf5 2. '8e8+ <j{h7
3. Wle4 pins the rook. Then .. .f6 [sic: ... g6] holds piece. 3 . . . g6 4. Wff e7+
wins. 1. Wld4 '8xf5 2. '8e8+ <j{h7
de Groot Ernie
White to move after 1.�d4! gxfl) 2.E!eS+ 'tt>h7
Subject wants to play 3.\We4 but 3.'8e7! is stronger.

86
 Class C

3. Wff e4 g6 4. Wff e7+ rook moves 5. Takes - checkmate so he can 't take
the bishop. 1. Wff d4 "iJ,fl 2. "iJ,eB+ "iJ,fB 3. "iJ,xfB#. 1. Wff d4 then l.. .d6 or l...dS.
J d6 2. Wffxf4 "iJ,xf5 White is up a pawn. I am not even in material, Ijust
••.

noticed. Play l . "iJ, eB "iJ,xeB or 1... Ci:ixdS. 1. Wffg B+ "iJ,xgB no tactics. Back to
J . Wie5 V!ffxc4 never mind (silent). 1. Wff d4 push clock. PV=l. Wff d4 d6; if
1. V!ff d4 "iJ,xf5 2. "iJ,eB+ <tlh7 3. Wff e4 g6 wins.
Again the youngster hits a quick bulls-eye with the computer's top
move! This time, though, he was quite a bit lucky because he played it
without correctly analyzing several of the main lines. If you see a move
like that, you need to work out some of the lines. If there is one big
defensive idea, that may be enough to cause that line to be discarded,
much less considered best. While the idea is to find the best move you
can, in volatile positions one analytical mistake can make all the differ­
ence between a great move and a terrible one. Still, an A+ for finding a
pretty difficult idea fairly quickly.

Protocol C-7 (de Groot A; 1550; adult; 5 minutes)

Black is threatening 1... Wffxb2. Nothing is en prise. He threatens to
capture J... Ci:ixc3. I can capture J . :il.xf6, or l. Ci:ixc6 gets the bishop pair,
but that trades a good knight for a bishop that is not active. Do I have
any tactics? No tactics. Count material; isolated pawn. He's weak on
the kingside dark squares. Candidate moves: 1.b4 - against the threat
l. li'i xdS Ci:ixdS 2. hd5 wins the bishop on e7 [wrong].

de Groot A
Black to move after 1.Ci:ixd5 Ci:ixd5 2.hd5
Missing that 2 ... b:d5?? 3.b:e7 is not forced; instead 2 . . . b:g5

l. li'i xdS hdS nothing that benefits [wrong]. J. "iJ,fel; l. Ci:ia4 ha4;
l . 'lff d2 g uards b2. I can double rooks with 1."iJ,c2. That's a candidate
- to double rooks. l. "iJ,c2 - push clock.
87
The Improving Chess Thinker

Subject C-7 begins with a decent assessment although indication of

who is better. Notice the strange logic: First Subject C-7 says th at if
l.li:lxd5 then if l...li:lxd5 2.i.xd5 wins the bishop. This is incorrect, but he
does not check it. Then he correctly tries to find a better reply for Black
and sees l.. .i.xd5 but then again incorrectly concludes that White has
nothing, missing the removal-of-the-guard and then fork idea 2.hd5
li:lxd5 3.i.xe7 li:lxe7 4.li:ld7.
At the end of his protocol, Subject C-7's process is pure Hope Chess.
He sees a move that does two things - it prepares to double rooks and
also guards the b-pawn - and plays it without checking to see if it is
safe. Players who play Hope Chess and then have to reply to l.. .ii:lxc3
usually think "Hmm! I wonder how I should take back - maybe I am in
danger . . .
"

Protocol C-8 (de Groot A; 150 0 + ; adult; 7 minutes)

J . hf6 or 1 . hdS [No evaluation] 1. ill xc6 but not sure if it's best.
What's going on? Material - White down a piece. No exchanges. No,
pieces even. White has an extra pawn - no, even. If J . hf6 ill xf6 is
best; after 1 . . . hf6 2. li:lxd5 hd5 then 3. li:ld7 is afork. �ut 1 . hf6 li:lxf6
is OK. Trying to affect ill d7. 1. hd5 ill xdS 2. ill xdS. �ack up. 1. ill xdS If
1 . . . hdS 2.hdS; l...exd5 2. hf6 hf6 3. ill d7 [missing 3 . . . hd7].

de Groot A
Black to move after l.ii:lxd5 exd5 2 ..txf6 .txf6 3.ii:ld7
C-8 makes a visualization error; because the bishop on c6 moved
in another line, thinks that it is no longer there.

1 ... liJxdS 2. he7 li:lxe73. li:ld7 [again 3 . . . hd7]. J. li:lxd5 hd52.hd5 .

lf 1. liJxdS liJxdS 2.hdS liJxdS and 3. ill d7. So 1. ill xd5 Hit clock.
For clarification, at this point I asked the subject, "What is your PV?"

88
 Class C

and he replied, "1.ltJxd5 hd5 2.hd5 ltJxd5 3.he7 ltJxe7 4.ltJd7."

subj ect C-8 makes a similar mistake as was shown in the diagram
for C-7. C-8 incorrectly assumes after 1.ltJxd5 that l...ltJxd5 is not pos­
sible due to the "win" of the bishop ofe7. As for the PV, he made another
common mistake by assuming he was winning and entering a critical
line without re-checking his analysis. In this case the mistake was a
retained image error, forgetting the bishop on c6 is still guarding d7.
Wh en you see a move which wins, jam on the breaks and make sure
_ do n'tjust play it and then say "Whoops!"
Protocol C-9 (de Groot A; USCF 1550; adult; 16 minutes)
Count material - White has more space. Isolated queen 's pawn.
Rook has open.file, queen is centralized, White's dark bishop is a "bad"
bishop, actively placed. Black has moved a pawn in front of his king,
has a rook on the open.file, the queen threatens . . . 1/?ixb2, the knights are
roughly centrally located. Black's dark squared bishop is not protected
{sic], so the knight onf6 is pinned. So White, with more active pieces,
should attack. Can White win anything? To calculate: l. ltJxdS to re­
move guard of bishop on d7: knight on f6 is protected twice. l. ltJxdS
il.xdS 2. ltJd7 ltJxd7 3.he7 just trade a bunch of pieces [?!]. ls there a
combination here? 1. hdS hdS 2. ltJxdS exdS - I don't see a way to win
material [?!}. l . hf6 f4!6 2. ltJxd5 or 2. hd5 hdS 3. ltJxdS exdS 4. ltJd7
forks bishop [rook?] and queen or ltJd7fork. 1 . hf6 hf6 2. hdS he5
is forced [sic] 3.dxeS. I don't think that is going to work. Now I am
just looking to improve the position. Black's plan? I don't know Black's
plan; . . . �bS with skewer, but the knight on c3 can just capture on bS.
1 ... ltJxc3 and I take with the queen to avoid the skewer, so if I take
the knight on dS, . . . �bS is troublesome, but perhaps not a real threat.
Need a plan. More space - pieces more active. I could challengefor the
c-file or attack his king. 1.'fk2followed by 2. 'i'!.lcl lines the rooks on the
c-file. He could. . . double his rooks also and end up trading. If l . . . ltJxc3
2.bxc3 �bS is kind of nasty [sic]. It's a threat I need to deal with. Also,
I have a pawn in take at b2; that's an issue that needs to be dealt with.
So maybe I just take the knight . . . The more I look at this, the more I
th ink it is a problem that needs to be dealt with. So I suppose I just
mov e l. Wi c2 or 1. 1/?ibl but that leaves d4 unprotected. So I see more and
more p roblems as I go. Do I move 1 . 1/?id2 to avoid all problems? 1 . 1/?id2
li:l xc3 2.bxc3 �bS, so avoid attack ofbS andfl, so l. Wid2 hit clock.

Subject C-9 is more verbal than most - he was able to put his con­
ce rns out loud more readily. That should not preclude one from doing
concrete analysis, and there he had several problems. He often had

89
The Improving Chess Thinker

trouble figuring out what was forced during the capturing sequences.
Subject C-9 had typical basic tactical vision problems seeing the dis­
covered attacks and figuring out the removal of the guards. Like alm ost
all players under 1800, he did not see the key line after 1.lll x dS, e.g. 1. ..
lll xdS 2.h:dS (to remove the guard on e7) 2 ... h:gS with complications
not unfavorable to Black. He never considered the best move 1.hdS
even though it is a major capture. So, like many C players, his analysis
was neither systematic nor correct. His final move l.Wld2 is defeated
decisively by l . . lll xc3 (which, to his credit, he did consider, so it was
.

not completely Hope Chess! ) but then, after his intended 2.bxc3, Black
can play 2 ... lll e 4:

de Groot A
White to move after 1.'f!Jd2? �xc3 2.bxc3? �e4

...which double attacks the queen and bishop and also makes a dis­
covered attack on the bishop on gS. White cannot meet all these threats,
and thus loses a piece.
So long as we are covering C-players who like 1.Wld2 :

Protocol C-1 0 (de Groot A; 1500; adult; 9 minutes)

Pawn structure; king safe; light square bishop long diagonal - not
that active. Knight on eS, bishop on g5 both attacking. Black's g7 is
weak - no bishop on g7 - !kB balances Elcl. Checks and captures for
Black: l... Wlxb2; 1... lll dSxc3 - might not exchange. If l . . . lll xc3 ope ns
�a2 diagonal. Black's lllfg5 [sic] attacks the bishop on g5. Protect th e
b2 pawn? l. Elbl or 1 . El c2 or 1.b4 gets space. Attacker 1 . �h6 is begin­
ning but weakens king - maybe combine by Qf6 later. 1. lllj3 prote cts
the king. Left with 1.h3 [White is not considering his checks, captures,

90
 Class C

a nd threats]. l. fi.e3 guards d4. Long range: Trade off Black's dark
squared bishop. Don 't like his knight on d5, so l. lLJxdS is possible.
Look at J.b4 or J. lLJxdS. Maybe exchange fi.a2 on d5 as well: isolate
his pa wns. Hanging pawn has to be dealt with. 1. lLJxdS attacks queen
bu t after the exchange the pawn is still attacked so that is not a good
idea[?!]. l . i!.c2: swing to e2 to attack the bishop on e7 eventually. I
remem ber that b2 is attacked. 1.h3 removes . . . lLJg4. Drop queen to d2:
g uards b2 and d4 and g5. if 1. Vfff d2 not a bad idea. Check 1. Vfffd2 a little
fu rthe r. Any tactical shots? So 1. Vfffd2 or 1.b4. Don't want to take the
rook off the c-file - converging a little on 1. Vfff d2 . Gives up the light
squares. 1 . . . fi.b5 skewers the rook but 2. lLJxbS. I like 1. Vfff d2 is what I'd
like to do, push clock.
Again a great example of Hope Chess, as Subject C-10 does not con­
sider the winning sequence in response for Black given in C-9. It is also
interesting that Subject C-10 spent so little time on his forcing moves,
briefly considering l.liJxdS but not systematically examining Black's
three possible recaptures to see if something good could be forced. No­
tice after first considering l.Vfffd 2 he properly asks "No tactical shots?"
but misses the glaring l.. .li:lxc3, removing the guard on e4. You can have
a good process but that does not mean your analysis will be good. On
the other hand, if you have a bad process then it is highly unlikely your
analysis will be consistently good.
The following protocol was by an accomplished Internet Chess Club
(ICC) player who was about to start playing over-the-board chess. His
first USCF ratings, around the time he did this protocol, were in the
"Class C" range. He later quickly rose into the "B" class, at one point
even getting his rating just above 1800.

Protocol C-11 (de Groot A; 1500+; adult; 10 minutes)

What's going on? Equal? Count material: equal. Isolated queen 's
pa wn. More of a position to protect. I need to defend d4. I see l. li:lxd5
or . . . li:lxc3. Which is better for me? . . . V!ixd4 J. li:lxdS li:lxd5 2. fi.xe7 is
good? Then 2 ... lLJxe7 ls that good for me? Would simplify the game.
He has nothing better than 2 . . . li:lxe7. ls that good for me? l. li:lxdS
li:lxd5 2. he7 li:lxe7 - worried about his rook going to dB to pressure
d4 . Los t opportunity to play fi.e3. I don't think it is good to exchange
beca use d4 becomes more exposed. What else can I do? . . . li:lxc3 2. i!.xc3
� r 2.bxc3. If 2. i!.xc3 is an open.file - not terrible. Am I blatantly miss­
in g so meth ing ? Do I lose a piece? Position looks equal. I want to better
my positio n. Consider l. i!.fdl to strengthen d4. If 1. i!.fel fi.a4 2. li:lxa4.

91
The Improving Chess Thinker

Well, alright I am waiting for a move to jump at me. 1. 'Hfel push clock
[then 1 . . . V!!xb2?]

This protocol contains some very interested pros and cons. Subject
C-11 never considered all of Black's threats. My note about 1...V!ffxb2 at
the end expressed my surprise at his omission, although 1.'Hfel V!!xb2
can be met by 2.lll c4.

de Groot A
Black to move after U!fel '9'xb2 2.lll c4
Even Grandmasters never mentioned 1...V!ffxb2 2.lll c4

No Grandmaster who performed this exercise with Dr. de Groot

mentioned the possibility of 2.lll c 4, but that is not an omission since it
was irrelevant in the PV for the best move 1.hd5.
Subject C-11 was very conscious of evaluation ("Is that good for
me?") which other players at this level sometimes lack. As players im­
prove, this conscious concern about evaluation grows.
Note his extensive use of the "null move" technique to find what
would happen after ... lll xc3, which makes his missing . . . V!!xb2 all the
more puzzling. Although this subject seemed unaware of the principle
"When you have the isolated queen's pawn avoid trading all (or most oO
the minor pieces" he still came to the correct conclusion that trading for
trading's sake would be bad for him.
Another good trait of Subject C-11 was his excellent use of compari­
sons, asking "Which is better for me?" This kind of question, which ofte n
directs the analysis, is usually lacking at lower levels. Unfortunately , his
focus was primarily defensive, and he did not systematically consider
his captures, a big mistake. His actual move 1.'Hfel was made with very
little analysis.
92
 Class c

summ ary of Class C

Characteristics of this thought process level:
• On the average, C players are much more detailed and analyti­
cal than Class D. C players do a much better job of looking at
specific sequences and trying to figure out how far to look be­
fore making an evaluation.
• C Players seem to have trouble listing all the opponent's recap­
tures, and also have trouble figuring out which ones are forced
(these two issues are related).
• By looking deeper, C players make more visualization mistakes
since they are trying to hold those positions in their heads. They
actually visualize better than Class D players, but the additional
lines being visualized results in more total errors.
• Players at this level often miss hanging pieces at the end of
analytical lines, or come to "big error" conclusions because of a
missed capture, or visualization error.
Improvements to the thought process that would help in getting to
the next level:
• More frequently ask about analysis moves "Is that forced?" or
"Is that all of the opponent's possibilities?" Frequently players
of this level either jump to conclusions or don't stop to ensure
they have considered all the key lines.
• Double-check key lines to make sure the analysis is correct, or
that the line is really quiescent. Often the players make a key
mistake and don't check it, or stop too soon.
• Do lots of basic tactics problems to work on visualization and
quiescence. As players get near 1600 they think they are too good
for basic tactics. But this chapter shows that is clearly not so!
• Be more consistent in the process. Start by evaluating the situ­
ation, then find the threats, then assess candidate moves, etc.
Sometimes C players know what to do, but skip steps in the
excitement of the position, or too quickly gravitate to one idea
without searching for all the good ones.
•
Make sure to understand the situation. The candidate moves
have to fit the needs of the position. When the player doesn't
correctly identify those needs, candidate move selection is
more random. If you are losing you want to play more aggres­
sively and avoid trades. If you are winning you usually want to
avoid complications and play simply.
93
Chapter 7

Class B

his chapter includes players rated 1600-1800 USCF, or rough ly

T1750-1950 ICC standard.
For each protocol, I list the de Groot position (for example de Groot
A, de Groot B, de Groot C, etc.; see Chapter 1 for those positions); the
age of the subject; the subject's rating; and the time the subject spent to
choose his move.
Comments in parentheses indicate outside actions or comments that
reflect the action of the player, such as (silent) if he paused for a long
time. Comments in brackets are my thoughts. My frequent use of [sic]
means the subject is making a clear mistake in analysis or visualization.
In contrast, [ ! ] indicates the player has made a comment very insightful
for their level of ability or a surprising error. A frequent note is [No
eval] meaning the subject did not try to evaluate which side stood bet­
ter, by how much, and why.
The first two protocols are by the same student roughly six months
apart:

Protocol B-1 (de Groot A, 1600, adult, -2 minutes)

Isolated queen's pawn. Not down. Opponent OK. All major pieces
are on the board. Afake pin atf6 and e7. I can take on d5 with knight
on c3; b2 hangs. 1 . . . Wixb2 (hmmm). Rooks developed. d4 attacked.
So . . . .I can take bishop pair but now. Looking for something to catch
my eye. All pieces except b2 safe. I can play l. &LJxc6 to get the bishop
pair. Then I can protect b2. So 1. &LJxc6 push clock.

I can never fully condemn the "Let's grab the bishop pair" idea in de
Groot A Subject B-l's thought process is not very thorough for such a
complicated position and does not follow the maxim "When you see a
good move, look for a better one," so I can't fully endorse it, either. At
this level we see more frequently that the players see the pawn on b2 is
hanging, but understand that it does not necessarily need immediate
attention.
Subject B-1 makes an interesting use of the term "fake pin" I have
-

used a similar term, "phantom pin" so perhaps that's the same thing.

94
 Class B

Protocol B-2 (de Groot Ernie, 1630, adult, 6 minutes)

(coun ts material) six pawns to five pawns. Bishops and knights
e ven; my rooks have open files. Black's a-rook and his bishop are
trapped. His knight onf4 is near my king. He threatens the queen. The
queen protects the c-pawn. The bishop is protected by the queen and
in ta ke. Queen also guards c4. Where can I put the queen to guard the
c-pawn and the bishop? 1. Wib5 allows 1 . . . a6 so might not be too good.
Let's see. Make sure to guard bishop because I am up a pawn anyway.
J . Wff e4 protects both after all. 1. 1/?ie4 let's see. 1. 1/?ie4 opponent's options?
Hmm. The knight doesn't have . . . l . . . CiJe6 or l . CiJ h5. Ifl take the knight. . . .
opponent can't push g5 and pin the bishop to the queen but that is no
good. So . . . the knight can stay onf4 guarded by the queen. It is good to
make the queen guard the knight. No threats besides those two. Want
to keep bishop on bl-h7 diagonal. I have used six minutes; that is good
enough. 1. Wie4 push clock.
The players who want to use the following logic: "1) My queen is
attacked, 2) My bishop is attacked, and 3) My queen needs to move
and guard the bishop, so 1.1/?ie4 is a great idea" are playing Hope Chess.
That is, they don't attempt to see if their main candidate (1.1/?ie4) is safe
before playing it. Of course, give half those players the Black pieces and
say "Black to play and win" . . .

de Groot "Ernie"
Analysis: Black to move after 1.Y!Je4?

.. . and they will find 1...d5! in a minute or less. That means their
�n alysis, in a nutshell, is Hope Chess. Their moves can be defeated eas­
ily by forcing moves (check, capture, or threat) which they don't even
attem pt to find. Instead they are surprised (unhappily) when their op-

95
The Improving Chess Thinker

ponents do. To see this in a Class B protocol tells you that even B players
sometimes play this way. This partially explains why they may know
so much about chess but aren't Experts. Subject B-2 makes the sa me
mistake as Subject D-12.

Protocol B-3 (de Groot A; 1650; adult; 9.5 minutes)

Even material. White has a little more space - good. The knig ht
has an outpost on e5; I don't want to trade that. The rook has an op en
file - I'd like to get my rook on fl to el. Umm - the bishop on a2 is
pretty good. Possibly some tactics later at p. Black has dark square
weaknesses on the kingside. He's got a bad bishop - no, dark square
weaknesses so getting rid of his dark squared bishop is good. Moves:
Exchanges. I can capture the knight on dS: l. li:JxdS He's got that square
guarded three times so there is no way to win anything on d5. IfI don 't
take him and he takes me l ... li:Jxc:3 then that strengthens my pawns by
bringing a pawn to the center. My d-pawn is isolated on an open.file,
so I would want to win in the middlegame, possibly with a kingside
attack. It is not apparent that there is any way to take advantage of
tactics, so l. '8fel. Hmm. OK. Another is l. li:Jxc6 because it gives me . . . no.
Bishops are better than knights. But in this position the knight is prob­
ably better. 1. �h6 with the idea off4. Hmm. Need a moment (silent).
Not seeing any tactics - how to improve the position? J. li:Jxc6 gives
the two bishops but costs my best minor piece. Moves like J.f4 don't do
anything. Don't see a way to weaken his dark squared bishop. l. '8fel
and l. �h6 are best - the only moves to improve the position. The idea
of l.b4-b5 to hit the bishop on c6. So l.b4 is a candidate. Hmm. Yeah
I might play l. '8fel followed by Qh3 so li:ixp with the idea of Vffxe6.
No, that's not even playable. Alright, l. li:Jxd5 exdS and my pawn is
no longer isolated [sic]. 1 . . . hd5 2. hdS exdS then I no longer have
an isolated pawn - the e-file is open. So 3. '8fel. l. li:JxdS li:JxdS 2. il.xd5
hdS. l. li:JxdS No, all those exchanges on dS would be bad for me. So
1. '8fel push clock.
Subject B-3 started with a good evaluation. Nevertheless, except for
the length, this seems more like a C player protocol as Subject B-3 never
searched for his opponent's threats and did not spot the b2 pawn hang­
ing. Once you spot it hanging, you either conscientiously have to ign o re
it while looking for something offensive (as World Champion Eu we
aptly noted in Appendix A) or look for moves to guard it. Otherwise, if it
turns out that the b-pawn cannot be taken, then you are just lucky!
Subject B-3 first dismissed capturing on d5 because it did not wi n

96
 Class B

anything and then later just because "all those exchanges would be bad
for me ." When you have a typical isolated queen's pawn position, equal
trades of minor pieces generally favor the player attacking the isolated
pawn. Although White has the isolated pawn, the trades are not "equal."
Ironically such principles take a back seat to concrete forcing lines, and
it turns out the "bad" trade of minor pieces is actually the right idea!
Playing more by general principles than by solid analysis and evalua­
tion can be very dangerous in analytical positions.

Protocol B-4 (de Groot A; USCF 1600 ; adult; 12 minutes)

Plusses/advantages: isolated queen's pawn. My queen is better; the
d-pawn is pressured, blockaded. I have more room on the kingside; his
black bishop is out of place. He threatens . . . V!ixb2. I'd like to double
on the c-file, so Elc2 is a candidate. The bishop at gS is undefended. If
J. E!c2 E!fdB I can defend easily. J. fi.h6 or 1. V!ih3. J . £iJ e4 drops a bishop.
J. liJxc6 wins the bishop pair but gives up the centralized knight. 1. ELJxdS
but I don't like exchanges - that helps the player playing against the
isolated pawn - I would probably avoid without a tactical reason.
1. liJxdS ELJxdS 2. hdS

de Groot A
Black to move after 1.£iJxd5 £iJxd5 2.hd5
A most common problem: missing the forced capture on gS.

2 .. hdS 3. he7. J. £iJxdS. With the rook on el I would have an attack

.

o n . . . get rid of the isolated pawn because his piece is hanging. J. CLJxdS
Ci:ixdS 2. hdS £iJxd7 3. he7 ELJxe7 4. CLJxc6 ELJxc6 5.dS EiJeS 6. V!ie4 and
the pa wn is hanging: ... V!ixb2. 1 . E!c2 is best for if exchanges then the
�a wn han gs. 2. E!fcl. 1. fi.h6 with the idea of CLJj3- £iJ g5. J . hf6 with the
idea of 2. £iJe4 fi.e7. J . £iJ e4 hangs the bishop. 1 . f4{6 hf6 2. £iJe4 with

97
The Improving Chess Thinker

the idea of going to cS or g5. 1 . hf6 t:/Jxf6 is not as good - it opens the
diagonal for the c6 bishop. l. 'i1.c2 push clock.
Subject B-4 was all over the place. (I don't know what he inte nd ed
for Black's second move in the line 1.0ixd5 0ixd5 2 .hd5 0ixd7 - oc­
casionally I transcribe moves incorrectly and occasionally my subj ects
"misspeak.") After jumping around, suddenly he returned to his origi­
nal idea and played it without even checking to see if it was safe ( Hope
Chess ! ) Jumping and guessing would seem indicative of a player closer
to 1600 than 1800. On the positive side, his 1600+ traits included know­
ing that trades usually help the side attacking the isolated pawn (see the
comments to B-3). Also, to his credit, Subject B-4 tried to have an evalu­
ation at the end of most analytical lines, which players at his level rarely
do, although sometimes his conclusions were not very clear.
The following subject was one of the country's top eight-year-old
players, and soon after hit "A" level. He is now an Expert and rising.
Here are his two protocols:

Protocol B-5 (de Groot A; USCF 1600; age 8; 2.5 minutes)

Force something on c-file. Pressure his knight on d5 is supported by
the bishop at c6 and the knight atf6. His dark squares on the kingside
are weak. The knight on f6 cannot be protected by pawns - ... 1/!ixb2
is a threat. 1. 'i1.fdl; l. 'i1.fel to try. Not l. 'i1.bl to protect pawn. So l.b4 is
good. l.b4 hit clock.
Subject B-5 provides another purely defensive protocol, with the
main intention to safeguard the b2 pawn. It could be worse, but at age
eight there are always some big deficits to one's overall approach. On
the other hand, to be 1600 at age 8 means many of those deficits will
soon be overcome! For example, at age 10 this same player would prob­
ably take Dr. Euwe's approach and think "OK, I can guard the pawn,
but first let's see if I can do something offensive instead." He would also
play more slowly.

Protocol B-6 (de Groot C; 1600+; age 8; a few minutes)

White's queenside pawns are up the board. l.. .d5 because c4 is
pinned. The queen and rook can attack j3. l ... t:/Jfd5 with pressure on
thej3 knight. J... fi.d7 - the ... e6 pawn is attacked. l...d5 with the idea of
. . . t:/J e4. l ... dS 2. t:/JeS is bad for Black. J . . . t:/J e4 with the idea of a qu een­
side fork. J.. . t:/J e4 2. t:/Jg5 no good: 2 ... d5 exd5. 2. 'i1.hel d5 3. t:/Je5 1/!i b6 .
No good for Black; 3 ... 1/!ia6 4. t:/Jxc7. 3 ... 1/!ib6 4.cS 1/!ia6 pins queen, so
t:/Jxc7 1/!ixe2 with the idea of ... c6. J . . . t:/J e4

98
 Class B

de Groot "C"
White to move after 1 ... ctle4

�
·� � �s -­
r � :t
� 1 ()'r..&%
� n
• ()'..&%
..� i• ,�
��- ·j· �­
0

ft � ft ��- �
� ,� ��
,ft �r.
�.���.r:i ��
� ��
2. 'Shel d5 3. 0i e5 Vff b 6 4.c5 Vff a 6 push clock.
Subject B-6 is quite a bit hasty and judgmental in such a complicated
position! But again youth will be served. Overall he exhibits a decent
tactical feel, and the final selection of l.. .0ie4 is commendable. Several
moves are reasonable and worthy of consideration in this difficult exer­
cise. However, the logic that led to l.. .0ie4 is at best questionable, and some
of the moves in the lines not as forced as Subject B-6 seemed to think.
Protocol B-7 (de Groot A; 1600; adult; 7 minutes)
Middlegame. The knight on f6 is pinned [no evaluation]. If I can
remove the guard I can apply more pressure to that knight. So ifI cap­
ture on dS there are two pieces covering dS and possibly I can fork on
d7. Looking at l. fi.. h 6 - no purpose in that. The rook on the c-file would
be nice to win. 1. tfJxdS if 1... tfJxdS then bishop attacks g5. 2. hdS hg5
3. hc6 and 3 . . . 'Sxc6. No, I think - I think I need to think quiet [?!}.
1. tfJxdS tfJxdS 2. hd5 hg5.
de Groot A
White to move after 1.ctlxd5 ctlxd5 2 . .ixd5 .ixg5
Avoiding a most common problem: catching 2 . . . fi..xg5.

99
The Improving Chess Thinker

OK, so I think [no conclusion!] I will play J . hdS. That way th e

knight. ..J.hdS then if l ... 0ixd5 2. f:LJxdS to attack the queen. Then if
2 . . . hdS 3. he7, so J.hdS [PV=l. hdS exdS gives him a bad pa wn
and then maybe 2. Yfij3].
This is an unbelievably good, if lucky, protocol from a 1600 player.
Without fully analyzing or evaluating the position he hits on exceptio n­
ally good and key points. For example, he does a good job of discussing
why l.f:iJxdS might not work, but then skips to 1.hdS without showing
how it does. He completely overlooks the key evaluation after 1.hdS
exdS until I ask him about his PV, and even then it's at best shaky. It's
not really a case of a blind squirrel finding an acorn, but this does hap­
pen. There really was a lot missing, so this is not a great analysis. How­
ever, it was like great analysis in one sense: it came to some of the same
conclusions and moves!

Protocol B-8 (de Groot Zyme; 1640 ; age 10; 4.5 minutes)
I can take 1. '/!fxaB but then 1... '/!fxc3+ wins the rook on al. So do n 't
allow 1... Yfixc3+. Move such as l. ikl looks good - then the rook on aB is
trapped - can 't move. So far 1. 'il.cl looks best. Oh! He has l . . . Y!ixf2+. So
he has two threats! So I. Yfij3 stopping both threats is a candidate. But
if 1 . . . Yfixj3 I don 't have a pleasant choice. 2.exj3 and I have a wide open
queenside. Anything besides I. '/!fj3 ? Whew! I can distract with J. il.e7+
'/!fxe7 2. '/!fxaB - that looks best now. Is there anything better? He is
attacking the bishop now. Alright. Let's see. Alright, well. My bishop is
not really threatened. If 1... '/!fxd6 then 2. 'iJ.dl but then 2 . . . Yfixdl+ w i ns a
queen for a bishop and a rook - but then
de Groot Zyme
Checking a threat: White to move after 1 ...ti'xd6 2.!::!dl ti'xdl +
Finding out the bishop on d6 might be hanging after all.

100
 Class B

3. \flxdl 1£ixf2+ and 4 . . . 1£ixe4. Is there anything better than J . .te7+ ?

Do n 't see. Let's evaluate. White has.five pawns, Black has six - other­
wise the same: Black is ahead a pawn. So winning the exchange would
help, but not that much. I like J. .te7+ The check is forcing - he can 't
decline. Alright J. .te7+ (hits clock).

subject B-8 is obviously a talented player, but some additional think­

ing discipline would make him much more efficient. First, if he had first
counted material then, knowing he was down a pawn, he could have
made further evaluations more efficiently. Second, after an early iden­
tification of his opponent's threats, if he had then systematically con­
sidered his checks, captures, and threats, he would have found t.ie7 +
more quickly. The capabilities are all there; he just has to organize them
better to be more efficient.

Protocol B-9 (de Groot A; USCF 1600 ; adult; 21 minutes)

Material is equal. Is there anything of White which is hanging ? The
bishop at a2 is loose but not attacked. The d4 pawn is OK. The dark
square bishop is hanging but not attacked. The b2 pawn is the only
one loose. I am double attacked on c3, but no problem. Black's knight
on f6 when moved can discover attack on g5. No checks or mating
threats for Black. Who is better? Does Black have anything hanging ?
No. Who's in better shape? White king is a little safer. Both queens
are OK. White's light squared bishop is better, dark squared bishop is
better. The knight on e5 is better than the one onf6; Black's knight on
d5 is better than the one on c3. White has more space. So the position
is roug hly equal, maybe slightly better for White.
Either save the b2 pawn or sacrifice itfor something better. l . .th6
attacks the rook onfB. 1... Yffxb2 2. hfB !£ixc3 3. Elxc3 Yffxc3 4 . .th6 No!
4. ixe7 Yffxd4 5. hf6 What a mess. I have captured a rook, bishop,
a n d knight and he has captured a rook, knight, and two pawns. Not
correct. J. .th6 Yffxb2 2. hfB. None of this works because the queen is
protected so Black loses the exchange for a pawn. J. 1'.h6 Yffxb2 2. 1£ixc6
i'lxc6 3. Yffxa2 he7 4. l£idS [1£idSxe7?]. White has captured a rook and
bishop ; Black has captured a pawn and bishop, so 1. 1'.h6 is decent
so what would happen? Black will not take the pawn so l . . . ElfdB to
overp ro tect 1£id5, Black's minor piece 2. 1£ixdS to remove the hit piece.
2 ·-- 0i xd5. Is there some way to take advantage of the dark squares
a �o u n d Black's king ? Is there any way to get rid of his dark squared
� hop? 1. l£i xd5 Let's see what happens. Black isforced to recapture 1...
d5 beca use the knight is pinned [sic].

101
The Improving Chess Thinker

de Groot A: White to move after 1.0ixd5 0ixd5

Another subject who visualizes this incorrectly or assumes that with
the piece on f6 gone and the knight on dS taken,
then the bishop on e7 must hang

2.hdS exdS isolates both pawns. I've got to protect b2, say 3.b4 but
the bishop is still hanging. The immediate 1.b4 gets to the same place
- it's a complicated position. 1. flh6 '8fd8 2. 0ixd5 The knight can't take
on dS because of the pin [same mistake as in the diagram] . 2 . . . hd5
or 2 . . . exd5. Is there any way to get another piece in? Yes, 3. 0id7 wins
material - or does it? [sic '8xd7] 1. flh6 '8fd8 2. f:LJxdS hdS: Black can
trade rooks or take with the pawn. Trading rooks doesn't help. 2. tuxd5
hd5 3.hdS exdS 4 . 0i d7 attacks the rook [sic] and pinned knight. Oh!
The rook has moved! That does not work. But J. flh6 '8fd8 2. 0ixd5
flxd5 3. hd5 exd5 4. '8xc8 '8xc8 5. iiJd7 Hmm. 5... Wff dB and I can 't play
6. 0ixb7 [sic?]. I should move the queen back to the.first rank. I get a
knight and bishop for a knight and bishop, pair of rooks. White h as
won a pawn on b7 - is this a drawish position? Ifelt the position was
equal, so I am not terribly excited. Don't see a winning combina tion
anywhere. The knight on c3 is unprotected. 1. iiJxdS hdS 2. flxdS exd5
3.b4 and. . . (silent) Having a rough time figuring out a better move.
This is all very messy. I think I like the idea of getting my bishop to h 6
even if it leads to an equal position. So I would play 1. flh6 '8fd8 2.b4. 1
have improved my position slightly. So J.fi.h6 push clock. (PV J . fih6
=

'8fd8 2.b4)
Subject B-9's protocol exhibits several interesting points:
• There are two primary ways to keep track of material duri ng
a series of exchanges: A) keep a running count (the metho d I
use the majority of the time) and B) visualize the final posi tio n

102
 Class B

and count the material. This subject uses a third, slightly dif­
ferent method: he recounts the material captured during the
sequence at the end. I rarely see this third method employed as
it seems slightly less efficient and possibly more prone to error.
If you go through a series of exchanges and are not sure of the
material count, it is important to double-check, either with the
same method or a different one.
• The key to this subject's entire plan was to safely establish
the bishop on h6. He never really asked himself if getting the
bishop to h6 at the expense of forcing the Black rook to the
center, where it attacks the key d5 and d4 squares, was worth it.
Instead he took significant time justifying 1.�h6 on the tactical
idea that the 1. . .1/!llxb2 counterattack did not work! So he was
assuming that if he could justify 1.�h6 and force the rook on f8
to move, he would have accomplished something positive. Yet
when his combinations with lt:ld7 failed to work because either
d7 was covered by a rook or the rook was not available on f8
to be forked, he never put one and one together and deduced,
"Hey! If I don't force that rook to d8 I can do lots of extra stuff
after lt:Jxd5. So instead of 1.�h6 !l:fd8 2.lt:Jxd5 I should be look­
ing at just 1.lt:lxd5 first!"
• Subject B-9 did an extraordinarily good job of attempting to
evaluate the position before analyzing. In this book very few
players rated below 1800 (and even below 2000) try to come
to some "static" conclusion of who stands better, by how much,
and why before trying to figure out what to do. This evaluation
sets a "bar" for analysis. If you stand better, then you would
not want to force perpetual check, or even settle for a line that
is dead even. Although the subject was wrong to think the ini­
tial position were even (see Appendices A and B), his logic in
comparing how good his best line is with how good he thinks
the position should be is excellent. This initial evaluation and
comparison with the outcome of lines is rarely found among
players at this level (or even a level above) !

Protocol B-10 (de Groot A ; FIDE 1700; adult; 10 minutes)

Assess position: White has the initiative on the queenside. The bish­
op would be better on g7. The knight on d5 protects the bishop on e7.
Ma terial: nothing off except two pawns. My move - can he do checks,
a tta ck the queen? . . . ltJf4 loses to the bishop capture, . . . ltJ b4 loses - can

103
The Improving Chess Thinker

attack with the bishop... fi.b5 0ixb5 or V!fxb5. Can attack with the q ueen.
My rooks can 't be attacked. What's undefended? Every piece but the
bishop on g5, which is undefended but is subject to discovered attacks
with the movement of the knight onf6. All Black's pieces defended. Is
my bishop on g5 in danger? If the knight on f6 retreats, then he7;
. . . liJg4 he7 - prefer not to swap bishops. Potential discovery on c3
with the bishop on c6. Keep dark-squared bishop so 1 . fi.h6 maybe with
h4-h5 with the idea of using the queen on the kingside. His initia tive
is on the queenside: . . . V!fxb2 is undefended {!}, threatens to swap off
on c3. Attention to b2 pawn. Two things: 1) Defend with a piece - not
good, or 2) b4 - I like this better. It leaves the bishop open on the a2-g8
diagonal but the c3 knight is not protected by a pawn so l.b4 is pos­
sible. Quickly check: Can he do nasty knightforks? ... liJdS retreats is not
a bother; going tof4 and e3 loses; . . . liJg4 loses, and l... liJxc3 2. l"txc3 and
don 't think he ... 1 ... liJxc3 2. Elxc3 fi.b5 pins queen to rook and can't take
with knight - not so good. 2. fi.c4 or 2. liJc4 attacks queen. 1.liJc4 attacks
queen and guards b2. Not too attractive: 1.liJc4 clogs the c-jile. How
am I going to defend the b-pawn or sacrifice itfor a kingside attack? If
he takes with the queen, is the queen in trouble? a3 would be hanging
- don 't want to sacrifice that (a3) pawn. Honest, I would have a long
think - I'm not sure what to do. I like 1.b4 - worried about l . . . liJxc3

de Groot A
White to move after 1.b4 �xc3
Analyzing the reply: that's NOT Hope Chess !

2. Elxc3 fi.b5. I could play 1. V!fd2 - it defends b2 and removes the

skewers on the b5-fl diagonal. I don 't like 1. V!fbl or l . V!fc2, the latter
putting the queen on the same line with the rook on cB. So 1. V!fd2 o r
1. V!fe2. The queen on e2 can still get skewered so 1. V!fd2 (silent). Wha t
am I thinking? I quite have begun to like 1. V!fd2 liJxc3 2. Elxc3 - ha ve
104
 Class B

th e op tion off4 or h6 - can 't attack d2 by ... l'i'ie4 because of l'i'ic3xe4.

Qu ickly. J. Yfid2. Ah, l'i'ixe4 both the knight and the queen defend g5.
Th a t re inforces the only hanging piece on g5, so l. Yfid2 push clock.
See the comments to Protocol C-9 for the refutation of 1. Yfid2. Amaz­
in gly, Subject B-10 saw both the idea of 1 ... l'iJxc3 and ... l'iJe4 but did not
put one and one together and realize that the combination of the two
refutes his move outright.
Interestingly, the subject mentioned l.l'iJc4 to guard b2, but did not
find that 1.. .Y!ixb2 would not have been possible anyway due to 2.l'i'ic4
trapping the queen. (As mentioned earlier, though, even the Grand­
masters in de Groot's book missed this possibility.) The diagrammed
position is noteworthy also: the subject did not just look at 1.b4, but
tried to see if it was safe after 1...l'iJxc3. I believe Subject C-9 was the
lowest-rated player in the book to find the refutation of l.b4 systemati­
cally. This is consistent with my conjecture that Hope Chess is practiced
by most players until their rating rises to 1600-1700.

Protocol B-11 (de Groot A; 1660; adult; a few minutes)

The material is even. White has a free hand. Black is weak on d7
andf6. My worst piece is the bishop on a2 and Black's best piece is his
knight on d5, so let's consider l . hd5. l . hd5 l'i'ixd5 (slow) ... wins a
piece, push clock.
This is the classic case of playing the right move for mostly the wrong
reasons. Here the Subject B-11 is happy to win a piece on 1.hdS l'iJxdS
but does not even consider the other recaptures 1.. .hdS or 1.. .exdS. If
he had looked at 1...hdS and found the eventual knight fork and con­
sidered 1 ... exdS and decided it was good for White on positional grounds
a nd had considered other moves and seen that indeed 1.hdS led to the
best position, then his analysis would have been perfect. Moreover, if
you see a winning move, the next thing you should do is ask yourself
whether or not this good fortune is really true and proceed slowly and
carefully. Many weak players do exactly the opposite and play the "win­
ning move" immediately, as in this protocol.
Wh en I bought de Groot's book I was a Class A player and, to the best
of my recollection, I also saw that 1.hdS l'iJxdS won a piece but, unlike
Subject B-11, I also saw that 1.. .hdS lost the exchange. Unfortunately,
I th ink my reaction to 1.hdS exdS was:
"Darn! If l ... exdS had Jost material too, I would be winning, but
l ... exds is OK for Black, so that doesn't seem to be my best move."
Dr. Euwe evaluated 1...exdS much better! (See Appendix A.)

105
The Improving Chess Thinker

Protocol B-12 (de Groot Ernie; 1800 ICC; adult; - 2 0 minutes)

Material - White has an extra pawn - it's doubled. Threats? No
harmful checks. The queen is en prise and can't take the knight. I need
to move the queen. 1. Y!ip 'i1xj7? 2. 'ii e B+ will mate. So that is possible
- threatens - can play l . . . liJe6 but then 2. 'i1xe6 removes the guard, so
I probably have to move the rook. What else can I do? Better is 1 ... 1/!idB
but then 2. 'i1e7 threatens mate - no, 2 ... 'i1xj7 so 1. Y!ip YfidB threatens
the queen. Is there anything better? 1. Y!ieS d6 and �lack threatens the
bishop on JS. If the queen moves off the fifth rank then the bishop is
loose. 1. Y!id3 is not possible so 1. Y!ie4. Hmm (silent). What else? 1.1/!i e4
then what would he do? l ... d5 attacking the queen and double attack­
ing the bishop. liJot so appealing. 1. Y!ie4 d5 anything else? liJot really.
So 1. Y!ie4 is not a good idea at all. 1. YfieS probably fails to l...d6. l.1/!ic5
or 1. Y!ia5. 1. Yfic5 Yfixc5 and the bishop goes. 1. Y!ia5 Y!ixa5 2. 'i1xa5 covers
the bishop but then a removal of the guard with ... b6. Umm. Can move
the rook to b5, d5, or eS. d5 is no good. 1. Y!i a5 Y!ixa5 2. 'i1xa5 b6 3. 'i1b5 or
'i1e5 but 3. 'i1e5 d6 or 3.�b5 a6 4. 'i1xb6 'i1xf5 5. 'ii e B+ lflh7 and I am down
a piece although Black's kingside is messy, but eventually he will free
his game so I'm not too keen on that. Back to 1. YfiP. Atfirst 1 ... Y!idB is the
answer. The white queen is short of squares. Tactics? Well, 2. CfJeS 'i1xP
3. liJxp+ wins the queen. 1. YfiP Y!i dB 2. liJeS threatens 3. liJg6+ too - it
looks more promising than the others. Anything else besides 1... Y!idB?
l ... 'ii g B 2. CfJe5 is stronger so 1... YfidB looks strongest. Then 2.CfJeS What
has Black got? (silent) Hmm. I suppose something like 2 ... d6 3. liJg6+
liJxg6 4. 1/!ixg6 hf5 wins the bishop. 1. Y!ip YfidB 2. liJeS d6 3. hcB Y!ixcB
then the queen's en prise and 4 ... 'iixP 5. liJxp+ no longer hits the queen.
Is there anything I can do? liJg6 exchanges the knight. What does that
leave me? Just a pawn up. 1. Y!ip Y!i dB 2. CfJeS d6 3. hcB Y!ixcB 4. liJg6+
liJxg6 5. Y!ixg6 Yfixc4 but then 6. 'i1e7 Y!ixc3 attacks al and guards the
mate on g7 - I may have to protect the pawn then. Anything better?
(silent) Don 't think I've . . . (silent) Hmm. Anything else I can do to pro­
tect the .. Alternatives to the 1. YfiP sequence? 1. Y!id4 'i1xf5 2. 'ii e B+ @ h7
then pin the rook to the king. Wondering what options Black has if I
do that? (silent). 1. Y!ieS YfixeS 2. YfixeS d6 looked at that before (sile n t).
The best line I can see is 1. YfiP - looksfairly even from White's point of
view. Wait a minute. 1 . liJ e5 liJxd5 2. liJp+ ltlgB works (silent) ["Plea se
do your analysis out loud if possible. "] OK, Umm. OK, so looking at
l . liJ eS liJxd5 2. liJp+ ltlgB is just winning the queen. So I do have to
move the queen. 1. Y!ip YfidB . . . (silent) . . .No, I can 't see anything better
than that. Maybe I am missing something ... (silent) 1. YfiP push clo ck.

106
 Class B

Wow! This analysis shows the difference between being a great ana­
lyzer and a decent one is not that big. Let's consider the three key lines
of Subje ct B-12's analysis and identify the mistake in each :
A) First, his PV of 1.ti'f7 ti'd8 2.�e5 d6 3.hc8:
de Groot Ernie
Black to move after 3.hc8

Here Subject B-12 only considered the move 3 . . . �xc8 and concluded
that after 4.ti:lg6+ the position was more or less equal. But suppose Black
makes the "other" capture 3 ... dxe5! Then White still has his queen and
bishop attacked and loses a piece. Probable cause of error: Not consid­
eri ng all the forcing moves at a critical node of the analysis.

B) His rejection of the second best move 1.'9'e5 was based upon the
key removal of the guard idea 1 ... Y:Jxe5 2.!ll:e5 d6

de Groot Ernie
White to move after 2 ... d6

107
The Improving Chess Thinker

But here White has the forced 3.hc8 dxe5 4.hb7 when White
already has more than enough for the exchange (two pawns) plus both
the a7 and e5 pawns are vulnerable and the c-pawns are strong pas sers.
Probable cause of error: Quiescence error - assuming that after 2 d 6 . . .

that the case was closed and Black was doing well, when in fact the
exchange sacrifice was good for White.
C) Finally, in the best line he rejected White's play after 1.ti'd4! !:ixf5
2.ges+ �h7
de Groot Ernie
White to move after 2 •.• �h7

... noting correctly that White could pin the rook with 3.Wle4. But
even better is to threaten mate with 3.ge7. For a full analysis of this
see Appendix B. Probable cause of error: A combination of quiescence
and not considering all the forcing moves, as there was no further in­
vestigation of how good 3. Wle4 was, nor realization that 3.l"1e7 had to be
investigated to properly evaluate the strength of 1.Wld4!
Subject B-12's analysis, in general, was very good. But that one little
missed move here or there, or misevaluation, was just enough to cause
another move to be played. Just these subtle errors caused Subject B-
12 to misevaluate each line: the move chosen (Line A) loses, whil e the
moves rejected (B and C) are both good for White. Contrast this near­
miss analysis to that of much weaker players, who can't find the ma in
forcing moves and don't get anywhere near as close to the truth as di d
Subject B-12.

108
 Class B

sum mary of Class B

Ch aracteristics of this thought process level:
• The first sign that Hope Chess is disappearing! For example
subject B-10 considered l.b4 and then tried to determine if that
was playable after l . . . ti:lxc3 !
• The players in this class start to show the characteristics of
much stronger players, but they all seem to have a fatal flaw or
two: one is too young to do all the deductive logic, a second as­
sumes too much in opponent replies, another knows too much
and uses general principles where only analysis will do. Finally,
we have one subject who gets the right move for the right rea­
son even if he didn't go through the entire proof process in a
way that would make it repeatable.
• Players are not as defensive and are more aggressive, looking
for wins and not settling for purely "safe" or "saving moves."
Improvements to the thought process that would help in getting to
the next level:
• Like all lower levels, consistency is a key. "B" players have the
rudiments of better play but jump around too much and don't
consider all the key possibilities in a reasonable order.
• "B'' players have a tendency to play a winning idea as soon as
they "think" they see it! Better to follow GM Lev Alburt and Al
Lawrence's advice in Chess Rules of Thumb, "If you see a move
which seems to win that is a critical move; a critical move is one
about which you should think long and hard ! " Many players get
excited when they think they find a winning line and translate
that excitement into quick play. It would be better for them to
take that excitement and use it to generate extreme caution
against believing something too good to be true! Be more cau­
tious when you have more to lose - it is worse to throw away
a win and lose than to throw away a draw.
•
More than players above them, "B" players seem to analyze
various candidates and then return to their final move without
double-checking to make sure it is good, or at least safe. If you
are going to play a move, then your opponent is going to spend
100% of his effort trying to defeat it. So spend at least a decent
percentage of your time making sure the move you play is not
easily defeated.

109
Chapter s

Class A

his chapter includes players rated 1800-2000 USCF, or roughly

T1950-2150 ICC standard. At this level the average reader should
start to focus on what these players do correctly and how to best emu­
late them.
For each protocol, I list the de Groot position (for example de Groot
A, de Groot B, de Groot C, etc.; see Chapter 1 for those positions) ; the
age of the subject; the subject's rating; and the time the subject spent to
choose his move.
Comments in parentheses indicate outside actions or comments that
reflect the action of the player, such as (silent) if he paused for a long
time. Comments in brackets are my thoughts. My frequent use of [sic]
means the subject is making a clear mistake in analysis or visualization.
In contrast, [!] indicates the player has made a comment very insightful
for their level of ability or a surprising error. A frequent note is [No
eval] meaning the subject did not try to evaluate which side stood bet­
ter, by how much, and why.
In the following two Internet protocols the subject was hard of hear­
ing. So he typed his "thought process" in real-time. There is something
lost in doing this, but there is also something gained. It is very inter­
esting to compare these written protocols to the others, which I tran­
scribed, so I made sure to include them.
Protocols A-1 and A-2 are edited only for readability.

Protocol A-1 (de Groot A; adult; 1800 FIDE; 30+ minutes)

Ok, look at the material situation . . . it's even . . . No checks. . . but b2
pawn is loose . . .
Potentially g5 bishop is loose, subject to discovered attack, bu t
knight atf6 can't go somewhere else to attack bigger game like a queen
or king. OK, let's carry on with identify ing threats to White . . . b2 p aw n
is loose, anything else? Black could play knight takes c3 . . . ah, if bxc3
then . . . ha3.
Thinking about l. CiJxc6, as that would attack Black's bishop at e7.

110
 Class A

Bu t after l... Elxc6 Black still has his threat on b2 plus he ca n double
on c-.file. Ok, what happens on 1.b4 or even 1.b3? If 1.b4 hb4 2.axb4
tt:i xb4 forking queen and bishop, but queen d2 or e2 defends the bishop
on a2. [Perhaps the subject doesn't see the knight on c3 already lends
so me protection to the bishop on a2.]
So far J.b4 seems reasonable. I'll hold that and see if there are any
other moves ...
1. b3 no good as 1 ... ha3. 1. Rbl another possibility. Thinking of
1. xc6 now... 1 . . .Rxc6 in reply doesn't get me anywhere and Black
tt:i
looks better... Perhaps 1. CiJxdS? Sort offeeling some possibility of tac-
tics based on knight attacking c6, bishop g5 on e7, knight c3 on dS.
Thin king J. CiJxdS hdS 2.hdS CiJxdS 3. he7 CiJxe7. Hmm, then 4. CiJd7
forking queen and rook atfB.

de Groot A
White to move after 1.�xd5 .ixd5? 2 . .ixd5 �xd5 3 .ixe7 �xe7 .

Removing the guard for the 4.CiJd7 fork is the Holy Grail
of many subjects in de Groot A

But [going back to move J . CiJxdS] bishop at c6 doesn't have to

m ove, gu arding d7 square. Let's try again, 1. CiJxdS maybe 1 ... exd5. No,
doesn 't seem to get anywhere for me . . . IfI have to make a move now, I
Would do 1 . 'ifbl. Check this.first to see if safe and keep that as reserve. . .
l . . 0i xc3 2. bxc3 uncovers attack on queen ... Queen ca n go a6 or a 5 but
.

looks OKfor Black. Don't want to move knight on e5-outpost.

b
Le t's try again, 1. CiJxdS. . . Getting complex for me. . . don't like b4
eca use of a5 in reply ...

111
The Improving Chess Thinker

de Groot A
White to move after 1.b4 a5

l. 'Hbl still my best move . . . Do I want to grab the bishop on c6 with

the knight? No, remember CiJ eS is in quite a nice position. l. fi.h6? 'flfdB
in reply, covers d7 square so knight can't go there... 1. 'Hc2? Can it be
attacked? For example, by Black's knight on dS? Hmm, I like 1. 'Hc2 and
maybe 2. 'Hfcl to follow. Check for tactics ... Looks OK, 1...ha3 2.bxa3
and knight can'tfork queen and rook on b4 because of the pawn at a3.
And I don't mind knight atf6 moving as I have he7 in response.
Prefer l. 'Hc2 to l. 'Hbl as it's more active. Ah, what about 1 . . . CiJxc3 and
then 2 ... fi.a4? Well, I can do b2-b3 in reply or do 'Hxc3 instead of bxc3
after ... CiJxc3. 1. 'Hc2 is my no. 1 move so far. Thinking 1...ha3 and
possible Black's queen zooming down b-filefor a double attack ... no it's
OKfor White. Anything better than 'Hc2? No, I'll cut off here and play
'Hc2. l. 'Hc2 push clock.

[The rest is summary after the protocol] :

Dan: Let's take a look at your PV. After 1.'Hc2 0ixc3, what were you
planning to play? [The subject moves 2.'Hxc3.] Yes, but what about 2 . ..

'!Wxb2?
Subject: Argh!
Dan: [My student continued playing White and I, with help fro m
Rybka, played Black: 3.'Hc2 '!WbS 4.'!WxbS hbS 5.'Hfcl 'Hxc2 6.'Hxc 2 E!d8
and my student admitted that Black was pretty much winning.]
Subject: I've lost one pawn already, so yes.
Dan: But I can't say you played "Hope chess" because you did tr)'
to find this. You did try to make sure 1.'Hc2 did not lose to any check,
capture, or threat.

112
 Class A

su bject: Well, I couldn 't see anything better and my mind was
g etting fogged up with all the possibilities .. .
D an: Note that you saw Black's idea of ... �xb2 but planned to recap-
tu re with the rook on c3 anyway. This should raise a flag. Your earlier
" King-of-the-Hill" 1.li:lxc6 just wins the bishop pair, so it is much better
then Uk 2. Half a pawn is not bad.
Subject: True, but I lost the e5-outpost knight.
Dan: Well, although li:lxc6 was not the best move, all the grandmas­
te rs who analyzed this position said it was a good move. Outpost knights
are great for things like winning the bishop-pair! Anyway, back to the
start. You did count material at the beginning - you counted the pieces
but stopped there. Much clearer would be to evaluate the position stati­
cally before you analyze: Who stands better, by how much, and why?
Dan: For example, suppose 1.li:lxg6 led to perpetual check by 1.li:lxg6
hxg6 2.�xg6+ i> h8 3.�h6+ lt>g8 4.�g6+ i>h8 5.�h6+ . If so, would you
do it?
Subject: Depends on if l felt Black were better.
Dan: Exactly, but you never let me know! Most grandmasters at the
start try to get a feel which side is better (statically). That gives them a
"hoop" to jump through during analysis.
Subject: Yes, you're right I didn't do a static analysis. [evaluation]
Dan: For example, if they feel White is a lot better they are looking
for a move which leaves them a lot better. If they feel the position is
equal, then they are thrilled if they can force a line with a slight advan­
tage, showing they were wrong. Theoretically, if your static evaluation
is perfect, then your PV's evaluation, if also perfect, should match it.
Does that make sense?
Subject: Yes. If you make no mistakes, your position can't get
worse.
Dan: Let's move on. You did pretty well when it came to picking
out a " King of the Hill". When you saw a good move, you looked for a
better on e. The main problems were: 1) Your analysis did not examine
all checks, captures, and threats before moves like Uk2, and 2) When
You did make a capture you didn't systematically attempt to find your
oppone nt's best recapture.
Subject: Yes, I found it hard work, so I probably held back.from
exa min ing every possibility.

=
Dan: For example, you would have to make sure Black's best reply
er l.li:l xc6, 1.li:lxd5, 1.ixd5, and 1.ixf6 led to a good position for Black
fo re settling for Uk2. And after 1.li:lxd5 instead of saying, "Now Black
has th ree r
ecaptures: 1...exd5, 1...li:lxd5, and 1...hd5; which would he

113
The Improving Chess Thinker

do?" you first assumed 1. Bxd5 and later 1...exd5, never mention ing
... lll x d5. Now that would make sense if 1.lll xd5 exd5 was so good for
Black that you needn't consider 1...lll xd5 but you did not make that
clear.
Subject: Yes, I need to be more systematic in looking along analysis
trees.
Dan: I would guess from reading in between the lines that you liked
Black's position better after 1.lll xd5 exd5 than you did after 1.E:c2 lll xc3 '

right?
Subject: Er, I can't quite recall.
Dan: Otherwise you would never play 1.E:c2 !
Subject: E:c2 felt safer to me . . .
Dan: But that's exactly what you need t o d o t o decide t o play U'lc2
- make sure your position after Black's best reply is better for you than
it would be after any other candidate and his best reply, right?
Subject: Yes, do the analysis tree.
Dan: So therefore 1.E:c2 lll xc3 is better for White than 1.lll xd5 exdS,
I would guess. The problem is that your process in showing this was
not clear - it did not look as though that was what you were deciding,
when that is exactly what you are supposed to be weighing: Do I want
the position after line A or line B? Does that make sense? Of course,
that only applies to an analytical position like this. In a non-analytical
position you play much faster and by general principles, usually. I am
going to email you World Champion Max Euwe analyzing this position.
As a math professor, Dr. Euwe is very systematic - compare what he did
with what you did and it should be very helpful in pointing out clearer
ways to approach the problem.
Subject: Okay, I will look forward to that!

Now, here is the same subject a couple of months later:

Protocol A-2 (de Groot Ernie; adult; 1800 FIDE; -30 minutes )
OK, White has 6 pawns vs. vs. Black's 5 pawns, else mate rial is
same. The knight atf4 is attacking white's queen. White doesn 't ha ve
an immediate mate, also no safe checks. Can't ignore knight attack on
queen, as have nothing to seriously threaten Black with. See that the
rook atfB is loose but can't see easy way to take advantage of it. OK,
some possible initial candidate moves 1 . V9 d4, er ... 1 . V9 d2, 1 . V9 cS (don 't
like it), hmm, 1. lll e5 a possibility . . OK, take each in turn ...
.

1. V9d4, then ... 1 . . E:xj5! Forgot that the bishop at j5 was under a t-
.

114
 Class A

ta ck by the rook, so have to hang onto it. New initial candidates 1. Vf! e4,
J . "lfff eS (loses c4 pawn?), 1 . Ci'i eS(??)
Look at Ci'ieS first. 1 . . . Ci'ixdS, 2. Ci'ig6+ @ gB and that's it. Can't do
li)e7+ so forget Ci'i eS possibility. Just remembered, ought to really get
a feel for who is better here. Looks like White has an advantage one
pa wn up, queen centralized, both minor pieces out, both rooks on open
files. Need to make the most of it. 1. VfleS Vflxc4, l . . . 'IJ.e4. 1. Vf!eS Vf!xeS
gxeS that looks goodfor White as can do 'iJ.ael...
-

Now look at 1. Vfle4. What about 1...dS in reply? Hits the queen and
uncovers an attack onj5 bishop by the bishop at cB ... Doesn't look good.
I see that dS or d6 is a possibility to consider for my VfleS initial move.
So .. 1. Vf!eS Vf!xeS 2. 'iJ.xeS and d6 . . . Not goodfor White ... Hmm, not only
.

I have to think about the knight atf4 attacking White's queen, but also
... d6 or ... dS. Might even have to think about l.VflcS... NO, that's no
good as after 1... VflxcS 2.bxcS 'IJ.xj5. So the position is proving a little
tricky ...
Looking at possibility of knight forks or bishop skewers or rook at
back rank but can't see at moment any way of luring Black's pieces
to their doom . . . At the moment, my best shot seems to be 1. Vfl e5 Vf!xeS
2. 'axeS d6 3.hcB dxeS 4. hb7... At the end of that I have two pawns
for the exchange .. but need to make sure position is quiescent.. . Hmm,
after 4.hb7 'iJ.bB, 5. 'iJ.xa7 getting another pawn? Can't see a better ini­
tial try than 1. VfleS as 1. Vfle4 fails to dS, 1. VflcS to 1 . . . VflxcS and 2 . . . 'IJ.xj5...
think some more... Think about 1. Vfl eS d6... but that's OK after 2. Vflx;f4
'8xf5 queen moves away ... So my king of the hill is 1. VfleS (ugh) VflxeS
2. '8xe5 d6 3. hcB dxeS 4. hb7 'iJ.bB 5. 'iJ.xa7. At the moment no other
likely sounding first move comes to mind. . . recheck my best line .. .
Thinking 1.'iJ.eB, with 1... 'iJ.xeB in reply ... no, not getting anywhere .. .
OK, recheck.first line ... 1. VfleS VflxeS 'iJ.xeS d6 hcB dxeS hb7... Yes, I'll
go for 1. VfleS and press clock.

[The rest is summary after the protocol]

Dan If we can improve your analysis skills you become a better player,
and positions like this can force you to stretch yourself. Okay, let's
look at some key points:
l) You did
not explain why a simple move guarding the bishop like
l . Y>\1a5 was bad (or good).

2) You did not look at the interesting threat 1.Vflf7.

115
The Improving Chess Thinker

de Groot Ernie. Black to move after 1.Y!Jf7

3) After 1.%\ld4 Elxf5 you didn't look much to see if you could get away
with sacrificing the bishop - you seemed to assume that it was just
pretty bad.
On the other hand, you get and A for your analysis of 1.%\leS and,
indeed, the computer says that your PV is the 2nd best line for White
and does retain an advantage. Take a look at 1. %\lf7 and tell me what you
think.
Subject: Yes, I was looking at 1. %\!fl (and 1. %\laS)... That is, when
you mentioned these and not before! 1. %\!fl, clearly 1 . . . Elxfl loses to
2. EleB+ as the bishop at j5 covering escape square at h7. So 1. %\!fl, look
at 1... %\!dB but 2. %\lg6 looks good. Looks like 2. . . Elxf5 forced. So 1. V!ij7,
1 . . . %\ld6 coversfB and g6 . . . but 2. EleB puts the kibosh on it.. . Try l . . . CUe6
in response to %\lfl. . . 1 . . . Elxe6 looks strong . . . Yes, 1... %\!dB looks the best
defense for Black for 1. %\!fl... Thinking about 1. %\!fl %\ld6 2. EleB CUe6. . .

Protocol A-3 (de Groot A ; adult; 1850 ; moderate time)

Static features: isolated queen's pawn, material even [but no evalu­
ation]. . . . %\lxb2 is a threat. Black has dark-squared weaknesses. I would
exchange dark squared bishops. 1. tiJxc6 Elxc6 wrecks the queenside
pawns. Minor pieces OK; the knight on e5 is the best. The bishop on a2
is posted but hits granite. At some point Elfel to overprotect eS. 1. hfo
to remove d7 defender. 1.hdS hdS 2. �6 and then tiJd7 wins the ex­
change. J.hdS tiJxdS 2.he7 tiJxe7 allows 3. tiJxc6 %\lxc6 does not work .
. . . tiJxdS is a good antidote. 1.hdS tiJxdS 2. tiJxdS hdS 3. he7 win s th e
exchange with tiJd7 [sic]. 1.hdS tiJxdS 2. tiJxdS exdS 3. he7 and ex­
change the bad bishop with b4 or b3. Double check. l.hdS push clock.
Another "right move for the wrong reason" protocol. It is true th at
1.hdS is the best move, but not only did Subject A-3 fail to con si der
116
 Class A

Black 's best reply 1.. .exd5, but after 1.hd5 li:Jxd5 2 .li:Jxd5 ixd5 3.ixe7
be said he wins the exchange when it is "twice" as good: he is winning
a piece. Full credit to A-3 for double-checking a line where he thinks
he is wi nning - but then he didn't find any corrections ! For if 1.ixd5
is indeed winning you would want to make extra sure - if so, the game
is basically over and if not, then the idea that you are missing might
mean th at the candidate is not even a good move! If for some reason he
thought 1.ixd5 exd5 were bad for Black (and I did not when I was his
level), then he certainly did not say so, or why. And if it were bad for
Black, the part of the "proof' where you show your leading candidate
is better than any other move, say 1.li:Jxd5, is missing as well. Compare
this protocol to how Euwe found 1.hd5 best in Appendix A. Euwe was
convincing; Subject A-3 less so.
Protocol A-4 (de Groot A; adult; 1840 ; 20 minutes)
Material: White is one pawn up [sic]. Weakness on dark squares
for Black's king. If the knight on dS is off, then the bishop is the only
one guarding d7, and lll d7 ca nfork rook and queen. The bishop on c6
and the knight onf6 which are guarding d7 can possibly be exchanged
on dS. I have a ha nging pawn on b2. Now consider moves. I have
no checks or attacks on the king. Consider J. li:JxdS, J . hdS, J . hf6.
Another is to get in lll d7 in some combinations. Another is J. li:Jxp.
Candidates: captures on d5, f6 with the idea of lll d7. So 1 . hdS then
l .. /:iJxdS or 1 ... hdS. 1 ... §J.xdS 2. hdS 3. lll xdS he7 with the idea of
.

12i d7. 2 . . . exdS 3. hf6 with the idea of 4. lll d7 so I am good. Two other
op tio ns: l ... li:JxdS 2.he7 lll xe7 and I cannot fork. I can consider the
sacrifice onp then ... 'fl,xp. If J . hdS hdS then I am happy - can get
in liJ d7. 1.hdS lll xdS no refutation so far.
de Groot A
White to move after 1.hd5 tll xd5
Missing 2.li:Jxd5 winning a piece.

117
The Improving Chess Thinker

He's not going to capture with the bishop. He can capture with the
pawn. On l . l°iJxdS he can capture with the bishop, knight, or pawn .
l. l"iJxdS I don't see anything good over there. Don't see any captures
on dS with the idea of l"iJd7 too appealing. So l. l"iJ e4 hangs a pa w n o n
b2 - don't see anything. I. l"iJ e4 l"iJxe4 is OKfor Black. 1 . hf6 If1 . . hf6 .

2. l"iJ e4 not really threatening anything. Maybe hdS. Not good. Over­
all - I am a pawn up still - I can go for simplification [!]. Can I just
play l.b4? My position is better on the kingside. Maybe just simplify
- Oh! I am NOT a pawn up, so there is no reason to simplify (sile nt).
Kind of stuck but I have to play a move. So l.b4 to gain space a n d
potentially bring the knight to cS. Does Black have a threat? l . . . l°iJ xc3
helps me double rooks [?!]. He has a queen and rook aiming at that.
0K, so Ijust now don't like l.b4 because of 1 . . . l"iJxc3. Need to protect the
pawn. 1. l!,c2 maybe? Can double rooks on the c-file. That's an option.
Can't find tactics in exchanges over there. l. l°iJxc6 so knight on e5 is
good. No reason to exchange. Any tactics? l. l°iJxc6 can play l . 'fl.xc6 ..

or l...bxc6. 1... l!,xc6 2. l°iJxdS can play 2 ... exdS. OK. Idea : l. l°iJxc6 'fl.xc6
2. l°iJa4 If capture with queen can't play 2. l°iJa4 Yeah. 1 ... V9xc6 2. hdS
exdS 3. l°iJe4 and the queen on the sixth rank protects the knight. Have
to make a move. b2 is annoying. Want an improving move. Go back
to l.b4. l.b4 l"iJxc3 seems bad to me. Can move my queen to f3 on the
diagonal but the d4 pawn hangs. Let's see l. fi.. h 6 then l . . . l°iJxc3 or 1 ...
rook on f-file moves. Already .fifteen minutes. Back to l.b4 - protects
pawn and gains space - c5 an outpost 1 . hf6 but l... l"iJxf6. 1.hf6 ffi.x/6
2. l°iJ e4 don't want to let his bishop get to g7. What about 1. V9h3 with the
idea of fi.. h 6? But d4 hangs. OK. Hmm (silent). The line I'm thinking of
is 1. hd5. 1 . hdS hd5 is no good. 1.hdS l"iJxdS I'm OK: 2. he7 0ixe7
3. l"iJ e4 - can have a strong knight. I don't have a dark squared bishop
- my knights are OK. 1 . hdS l"iJxdS 2. 0ixe7 he7 then b2 hangs. I can
play 3.b4 and then play 0i e4 later. So 1. hd5 push clock.
At the end this protocol is remarkably similar to that of Subject A-3:
Subject A-4 did not consider 1.hd5 exd5 and he also played 1.� d 5
0ixd5? 2. he7 or 2.0ixe7 missing 2.0ixd5 ! winning a piece. Yet d esp ite
these lapses 1.hd5 emerged as clearly best. In chess it is not better to
be lucky than good, although on the way to being good one gets lucky
more and more often !
Notice that when Subject A-4 considered 1.l°iJxdS, he next stated: " . ..
he [Black] can recapture with the bishop, knight, or pawn." This is the
way to begin the kind of systematic analysis missing in almost all p r oto­
cols in previous chapters. Unfortunately Subject A-4 then dismisses the
entire line without further analysis!

118
 Class A

Without mentioning the bishop pair, he carefully considers 1.li:lxc6

does consider the various captures. Careful analysis takes
and this time
tim e , and it is noticeable now that the "A" players' average protocol
le ngth is longer than those in the preceding chapters.

Protocol A-5 (de Groot Ernie; adult; 1834; 34 minutes)

Pa wn structure? Two islands to three. Black is worse. White's more
developed. White's in a predicament. The bishop at f5 is attacked - is
it really? Back rank thing prevailing. Let's see (silent). Well, back rank
comes into play 1. V!Jj7 "Bxfl? 2. "BeB+ and leads to mate. 1. V!Jj7 then the
bishop on cB can't develop so he must play 1 . . . V!JdB or 1 . . . "BgB. Usually
not good to have to move 1 . . . "BgB. That gives White has a free move,
so to speak. 1. V!Jj7 V!JdB . . . 2.queen moves: 2. V!Je7 V!Jxe7 3. "Bxe7 "Bxf5 and
Black is doing well. 4. "BeB+ @ h7. 1. V!Jj7 V!JdB Black can't develop the
queenside. Maybe . . . hmm. As White needs a way to stop him from de­
veloping his queenside. Let's see (silent) ["Do out loud'1. 1. V!Jj7 V!JdB
hard to continue for White. The bishop onf5 is a problem; the bishop
is somewhat a problem. There are cases where lll h 4 comes to mind
- a lot of possibilities, but not right away. So . . . (silent). If White just
moves his queen, wants to move to attack the knight onf4 to prevent
... d6,for example. Let's see. Umm. Candidates 1. V!Jj7, 1. V!Je4 What else?
(silent). Looking atforcing moves like 1. "BeB - trying toforce a conclu­
sion. May be some type offorceful conclusion - Black is awkward.
Candidates 1. V!Jj7, 1. V!Je4. Two moves to look at first. If not good for
me, then analyze other moves. 1. V!Jj7 V!JdB and doesn't seem like 2. V!Je7
'l!ixe7 3. "Bxe7 "Bxf5 4. "BeB+ @ h7. . . question there - not a good variation
- Black can play . . . b6 and �b7. If 1. V!Je4 then 1 . . . dS even if it sacrifices
a pa wn which it doesn't. 1. V!Je4 d5 2.exdS � that's obviously losing
(chuckles). Not much other possibilities for moving the queen - others
drop the bishop onf5. Interesting. Well, let me think (silent). Back to
1 . '/!ip VfidB. The queen's attacked. Doesn't seem to be a way to . . . well.
One more possibility. 1. V!ie5 attacks the queen on c7 then 1... V!J moves
2. V!ixf4 So 1 . V!ie5 either 1 ... tll e6 or J... V!ixe5. 1 . . . V!Jxe5 2. "Bxe5 d6 What
else? OK. 3. lll h4 dxe5 4. lllf4 covers g6. 1. V!ie5 V!Jxe5 Ha Ha Ha. Other
mo ves. Is Black really threatening l . . . "Bxf5? Can I lure him into that?
Maybe 1. V!id4 is interesting [15 minutes]. 1 . . . "Bxf5 2. "BeB+ @ h7 and now
3· Vfi e4 pins the rook and threatens g4. Kind of artificial. Well. Hmm.
This is my normal thing - this position will suck up a lot of my clock
[sho rt inte rlude
to discuss time management]. 1. V!Jj7 V!JdB White may
be . n trou
z ble. Come up with all the fancy moves 1 . V!Jj7; l. V!ie5. 1. V!ie5
Vfixe5 2. "Bxe5 d6 is annoying - it attacks both the rook and the bishop.

119
The Improving Chess Thinker

de Groot Ernie
White to move after 1.WleS WlxeS 2.!heS d6
Correctly finding the exchange sacrifice

Probably forced to play 3. hcB dxeS 4. hb7 '8b8 5. '8xa7. White

does get quite a bit of compensation for the exchange - light squared
bishop is a strong piece [Subject did not realize he was ahead a pawn
to begin with, and thus underestimates the result]. 1. WIxeS WIxeS 2. '8xe5
d6 3. hcB dxeS 4. hb7 maybe 4 . . . '8ad8 5. li'JxeS. White is OK - has
passed queenside pawns. 5. . . '8d2 li'J d6 OK. l. WieS WixeS d6 3. hcB dxe5
4. hb7 '8ad8 Umm. Want to play 5. li'JxeS - eliminate Black possibility
to put rooks on the second rank - probably good for White. Difficult
for Black to manage pieces to stop White's queenside pawns. Now that
I found that continuation - good for White - double check: is there
anything better?l. Wid4 If1 . . . '8xf5 2. 'f!.eB+ 'iklh7 3. Wie4 attacks the rook
and threatens g4 - now can Black play . . . g6 to defend the rook? Then
4. Wfe7+ with at least a draw. Any line where you have at least a per­
petual is worth looking at [!]. 1. Wfd4 '8xf5 2. 'f!.eB+ 'iklh7 3. Wfe4 d5 same
situation but missed it - have 4.cxdS and the threat for g4 is still on.
Now 4 ... g6 loses to 5. '8e7+. Wow! Seems like a pretty good contin ua­
tion. Also li'J h4. Hmm. All kinds of stuff - variation gets pretty com­
plicated there. Black can take on c3 and threaten Wixal - the n 'f!.el .
Very complicated. Break pin with li'Jg6? Let's see there. ls h4 and h5
a possibility? Also li'J d4 with h4 - . . . h4 what happens? (silent). Yea h,
cause 1 . Wid4 "i1.xf5 2. 'fl.eB+ 'iklh7 3. Wfe4 dS 4.cxdS li'Jg6 5.h4 with the idea
ofpushing g4, hS. But g4 then '8f4 so that does not work: '8f4-> 'fl.xg 4
so h4 is not effective. I take that back - 6.hS 'fl.xh5 7. 'f!.ael with the idea
of li'J eS. The whole thing is that Black's queenside can't get developed.
Time consuming. 5.h4 - like to get my queen to gB and cB. 1. Wfd4 'fl. x/5

120
 Class A

2. El eB+ cJ;h7 3. Wle4 dS 4.cxdS li::i g 6 - don 't see anything else for Black.
4 , .. /1'J g6 5.d6 with the idea of 5... Wfxd6 6. Wic4 threatening cB+, WigB#.
What can Black do? Anything? 6 ... li::i e7 holds both. Also 6 ... '8f8 that
doesn 't work. 5.d6 ... Wlxd6 this variation does not seem enough for
Wh ite - check one last time [30 + minute warning]. Last variation
so mewhat good for White. 1 . WieS W!xeS 2. '8xe5 d6 3. hcB dxeS - Can
Black take on cB? - 4. hb7 '8ad8 5. li::i xeS. I really like 5 ... '8d2 6.j3 §1.fB
and my minor pieces are good. Can't get checkmated. Then push c­
pawn with bishop on b7 controlling the c-pawn march c4-c5-c6. So
th at variation is good. 1. Wies hit clock.
This is a good example of "long analysis - wrong analysis." It is not
so much that Subject A-5 made many mistakes in his longer lines (al­
though everyone make some!), it is that he overlooks that he is trying
to find the best move, not figure out everything that is going to hap­
pen. The subject would be much more time efficient if he would hone
in on 1.Wie5 and 1.Wld4 and then simply ask himself "Based on the main
lines of each, which do I think is better for me?" and then just play
that move. (See the "progressive deepening" discussion in Section 2.3.)
For too long he tries to get into the intricacies of 1.Wld4 when much of
that information may not be needed. Although for 1.Wld4 quite a bit of
information is needed, there is always a practical point of diminishing
return, especially with the clock running. Either the main lines gener­
ate sufficient pressure to make that candidate more promising than the
exchange sacrifice of 1.Wfe5 or they don't. That's really the main issue.
When World Champion Dr. Max Euwe finds Ui.:xd5 (see Appendix A)
he doesn't look more than two ( ! ) moves ahead in the principal varia­
tion and decides he has enough pressure. Right before moving Euwe
states, "Much is still up in the air... " and then quickly moves. Instead of
deciding what is up in the air, he properly leaves this for future moves.
Euwe understood his goal is to find the best move possible, not to
fig u re out exactly how good it is. Once you determine the best move,
you are done. Any further analysis is just a waste of time. Of course,
sometimes you have to look very deeply to see if a move is good, but
that is not always the case. See Protocol A-7 for more discussion on this
concern.

Protocol A-6 (de Groot A; age 16; 1800; 8 minutes)

1. fti xdS hdS 2.hdS exdS 3. '8xc8 '8xc8 no good. 1 . hf6 hf6 2f4
il.d7 [?]. 1.b4 I don't like it. lj4 li::i xc3 2. '8xc3 '8fd8 3. '8fel No. J.fi.xdS
!iJxdS 2. he7 ftixe7 Hmm. J.hf6 §t.xf6 2. li::i e4 §t.g7 Hmm. 1.b4 li::i xc3

121
The Improving Chess Thinker

2. 1'lxc3 !'lfdB 3. !'lfel fi.e4 No, that works: good for Black. 1 . l:iJxc6 Wfxc6 -
never mind - no goodfor Black. I /4(?] l:iJxc3 2. 1'lxc3. l /4 Wfxb2 2. hdS
l:iJxd5 3. he7 l:iJxe7 no good for White. 1. l:iJxd5 l:iJxdS 2.he7 lll xe7
kinda like itfor White. 1. l:iJxdS push clock [PV l. l:iJxdS hdS 2.hf6].

There are two general ways to arrive at a move: by principle and

by analysis. Most positions require a little of both, but some are mu ch
more one way than another. Obviously Subject A-6 believes de Gro ot A
is a completely analytical position and he is mostly correct. He gave al­
most no descriptive clues as to why he was analyzing specific candidates
or why he thought certain replies were forced or best. He just went into
lines as if they were obvious or forced. The evaluations are there, but
not explained - good players do that sometimes in their hurry to beat
the clock.
Subject A-6 initially rejects 1.i?JxdS when only considering the reply
1.. ..il.xdS, but later decides he likes 1.i?JxdS if Black replies 1.. .i?JxdS. Then
he concludes that he will play 1.i?JxdS based on the latter! Where's the
logic in that? If 1.i?JxdS .il.xdS was not good for White, that why would
Black choose 1.. .i?JxdS - and why would White choose to play 1.lll xdS?
Subject A-6 did not even consider 1.i?JxdS exdS, which may be bad for
Black, but you would not know it from this protocol. While it is far bet­
ter to use 100% analysis on positions like this than 100% principles and
hand-waving (which would never work), this protocol is an excellent
example of how just examining moves without thinking about objec­
tives, principles, or descriptive evaluations can be dangerous, especially
if you are not a top-notch analyst. The subject also failed to assume the
opponent will play his best move.
Another interesting point is that, after 1.i?Jxc6 Wxc6, Subject A-6
states "never mind - no good for Black" and never considers it further.
But since he is White, did he mean " - no good for White"? If it were truly
no good for Black, that would certainly merit further investigation !
The next protocol is one of my favorites because it is very instruc­
tional. It represents a unique type of problem exhibited by players at
this level.

Protocol A-7 (de Groot A; adult; FIDE 1800; time unknown)

The position is about equal; White has the initiative but a poten­
tially bad endgame. After J. hd5 I think l ... l:iJxdS is OK. J. hd5 lll xd5
2. l:iJxdS (takes a while and then realizes White is winning a piece). 2 ...

!'lxcl [sic]. I think l. hd5 i s winning b u t no, Black can play l...exd5.

122
 Class A

de Groot A
White to move after 1 .b:d5 exd5
.

Correctly realizing this recapture is forced

When you see a good move, lookfor a better one. Now 1. hd5 exd5 is
forced. Then/ canplay2. V!!j3 Then 2 . . . cJ;g7isforced[Can the bishop on c6
move?]. 3. hf6 hf6 4. /:LJxd5 hd5andlcan win the d-pawn . . . [The sub­
ject spentadditional time examining the continuation of1. hd5 and then
played it; I am not sure ifI have the entire protocol.]
I will never forget the conversation after this protocol. I suggested
to the student that he could improve by keeping in mind that his goal
was to find the best move and it is not always necessary to spend a lot of
time figuring out exactly how good a candidate is. I read him the Euwe
protocol (see Appendix A) and emphasized the final paragraph, where
Euwe mentioned that "much was up in the air" but played the move
immediately thereafter anyway. There was a stark contrast between this
Subject A-7 and Euwe. Subject A-7 had correctly pinpointed 1.hdS as
the key move - and for the right reasons - but had continued exploring
it as no student has, before or since. After I explained how keeping in
mind the "finding the best move" would help him, my student protested
that without extensive analysis of 1.hdS he would not be able to find
out how good it was. Thus he could not determine whether it was the
best move without extensive analysis. I replied that extensive analysis
is sometimes necessary, but not in this case. Even if extensive analysis
were needed, that would require the player to periodically go back and
co m pare how good the positions were after the 1.hdS lines to other
li n es to see if indeed it were the best. Only ifyou have a choice between
two (or more) strong candidates would it make sense to analyze exten­
sively before playing one.

123
The Improving Chess Thinker

For example, suppose you have examined all the reasonable ca n di­
dates and know that the best you can possibly do with any of them , ex­
cept one, is to get a slight advantage. But with that candidate you get at
least a bigger advantage; it might even be winning. You can confid en tly
make that move because it is the best possible, even though you do n't
know exactly how good it is.
I pointed out to my student that none of the Grandmasters in de
Groot's book felt it was necessary to look so deeply into what might
happen after 1.hd5. It wasn't necessary. But instead of helping him,
this made my student defensive (something I never want to happen).
He became indignant and refused to accept my advice. I felt bad about
his reaction, but to this day I also feel like I did try my best to help him.
The pain is all the more since this student obviously had impressive
talent for an 1800 player: to identify the key line so quickly and so well.
Yet something was retarding his progress ...

Protocol A-8 (de Groot A; adult; FIDE 1800; 5112 minutes)

The pieces and pawns are equal. The knight on e5 can take the
bishop on c6 . ... g6 weakens Black's kingside. 1. l:iJe4 l:iJxe4 2. he7 then
2 ... l:iJxe7 protects the knight at e4, so no good.

de Groot A
White to move after 1.�e4 �xe4 2 .b:e7 �xe7•

The bishop on c6 guards the knight, although 3.i?Jxc6, removing the

guard, is worth investigation

The c-jile is blocked. Maybe 1. l:iJxc6followed by 2. l:iJ e4 to stop Bla ck

from guarding the knight at e4. 1. �xc6 then l...bxc6 or l ... '8xc6 or J ...

Wlxb2 attacks the rook at cl a n d the bishop at a2. I could play l . � h6

to attack the rook atfB. Then l . . . 'i!fdB attacks d4. b2 is loose in so m e

124
 Class A

va riatio ns. So I can play 1.b4 - that strengthens the outpost on eS. If
I don't do anything what would he do? 1 ... 'i'J,fdB anyway. l.b4 'i'J,fdB
How does that do? I can do 2. l:LJj3 to protect my d4 if I have to. 1. l:LJc4
allo ws 1 . . . Wia6 so I think my move is l.b4 to threaten l.b5 or. . . l.b4
Push clock.
This is quite a bit of Hope Chess from a player rated so highly. Sub­
ject A-8 does not consider whether Black has any forcing move (such as
1. .. l:LJ xc3) which might be difficult to meet after his chosen move l.b4.
subject A-B's overall feel for the position was not bad. After seeing that
i . l:LJ e4 probably wasn't safe (few players who considered 1.l:LJe4 were able
to see that Black's 2 ... l:LJxe7 recapture would allow the bishop on c6 to
guard e4), he deduced that 1.l:LJxc6 would make a later 2.l:LJe4 safer. This
type of good deductive analysis is not usually found in players rated
under 1700. The player had been close to FIDE 2000 but had semi­
retired from chess during his middle age. Subsequent play on the ICC
will probably show this player's strength to be somewhat below 1800
FIDE.

Summary of Class A
Characteristics of this thought process level:
• While "B" is the class where the "Real Chess" thought process
first appears, at the "A" level it is more prevalent, although not
universal.
• It is clear that players at this level, on the average, put a lot
more thought into their moves. The reason is not just because
they know to take more time, but also because they understand
how to use that time to proper effect. It is one thing to play
slowly. It is another to be able to use that time wisely, analyz­
ing lines that have a high percentage chance of affecting the
quality of the move. For example, unless one can move purely
by eliminating all the other moves by deductive logic (some­
times possible in certain endgames or when facing a check or
other severe threat), then it is imperative to spend a certain
amount of your thinking time on the move you actually end
up playing. This key trait is seen more and more as the players
approach Expert and Master level.
•
As each higher level is reached, the players become better and
better at seeing what the opponent will do. This creates a side

125
The Improving Chess Thinker

benefit in that they analyze lines more likely to occur - a nd

thus are more likely to accurately evaluate the position. We aker
players are not as proficient at deducing what is forced and th us
their final evaluation also suffers.
Improvements to the thought process that would help in gettin g to
the next level:
• It was once said to me: "Experts think to get out of trouble·
Masters think to avoid trouble." Yet this also applies to Class A
players trying to become Experts.
• I think Experts and Masters question themselves and their con­
clusions more honestly than do class players. As GM Jonathan
Rowson wrote in his book Chess for Zebras, when he analyzes
with Grandmaster, the Grandmaster often mentions several
possible ideas and plans, while his 1700 level students are often
sure there is only one idea in the exact same positions. This
conclusion apparently applies to "A" players as well.
• I think "A" players seem to spend less time on their final deci­
sion than do higher rated players. In the protocols of this chap­
ter, the subjects often consider several ideas but, when they
switch back to the final one, it is often for a short time. I think
stronger players return to their final move and try to do the de
Groot Phase 4 "proof" that it is the best move they can find.

126
Chapter 9

Expert and Above

his chapter includes players rated Expert (2000-2199 USCF or
TFor
roughly 2150-2349 ICC standard) and above.
each protocol, I list the de Groot position (for example de Groot
A, de Groot B, de Groot C, etc. ; see Chapter 1 for those positions); the
age of the subject; the subject's rating; and the time the subject spent to
choose his move.
Comments in parentheses indicate outside actions or comments that
reflect the action of the player, such as (silent) if he paused for a long
time. Comments in brackets are my thoughts. My frequent use of [sic]
means the subject is making a clear mistake in analysis or visualization.
In contrast, [!] indicates the player has made a comment very insightful
for their level of ability or a surprising error. A frequent note is [No
eval] meaning the subject did not try to evaluate which side stood bet­
ter, by how much, and why.

Protocol E-1 (de Groot Ernie; adult; 2150; 19 minutes)

Don't recognize the opening - backward d-pawn. His knight is
attacking the queen . Must move the queen. Tactics to delay it? 1. VJJp
is interesting! He can't take the queen because 2. :1'1.eB+ leads to mate.
I. V!Jp is a candidate. Others? 1. V!JcS V!JxcS 2.bxcS nothing. Material? Up
a pawn to boot! Don't have to go infor major tactics?! 1. V!JeS is worth a
look: 1. YfieS YfixeS 2. :1'1.xeS The rook attacks the bishop atf5 - that nar­
rows the candidates. 1. V!fb5 a6 Yeah, that's the case. Not 1. V!Ja5 maybe
1 . Wia5 b6 same problem 2.YfieS V!Jxe5 3. :1'1.xeS d6 attacks the rook and
bishop - doesn't look like it works. 1. Yfie4 looks playable. If 1. Y!ip does
not work can bail out with 1. V!Je4 - look at l.. .dS. Maybe I can't bail
o ut with 1. Yfie4. So 1. VJJp is the candidate move. 1. VJJp then he moves
1 .. :1'1. g B or 1 ... :1'1.dB or protect with 1 ... Yfid6 or l... l1Je6. No, 2. :1'1.xe6 and
.

he can 't take back. So l . . . l1Jg6 is ridiculous - 2. fi.xg6. 1. Y!ip Yfid6 or

1 .. Vfi d6; l... Y!id6 2. :1'1.eB is good - sufficient to win. 2 ... l1Je6 at worst
.

I exch ange the pieces and I am up a pawn - OK. J. Y!iP Wid6 2. :1'1.eB
ltle 6. Yea h,forced. 3. :1'1.xfB White is ahead in development with a much
bette r ga me. ... l1Je2+? 1. Y!ip Yf1 dB looks best - rule out 1 . . :1'1.dB 2. :1'1.eB+
.

'il,xeB 3 . V!fxe8# 1 . . . :1'1.gB again : 2. :1'1.eB my best - not as certain. 2 . . . l1J c6?
1:'0· Maybe 2 ... g6. 1. WiP :1'1.gB 2. :1'1.eB same problem. Same threat;
in most positions it(the defense) does not work. So 1. YffP V!JdB only.

127
The Improving Chess Thinker

Problem: both my queen and bishop are attacked. The queen has no
good squares (groan). This does not look very good now. 2. Vfie7 no. It
all looks bad. Starting to think I may be in trouble here. Any tricks ?
1. 'l'l,eB? Doesn't look very good. Re-evaluate. Anything tactical I missed
somehow? 1. V!ie4 d5. Maybe I should be looking at something else al­
together. 1. V!ip V!idB 2. V!ie7 V!ixe7 3. 'l'l,xe7 'i'J,xf5. Down a piece. 4 . l'U d4 ?
Can ! take advantage of the bishop being undeveloped? If l. Vfie5 d6 no
good. 2. V!ixf4. 1 . . . V!ixe5 2. 'l'l,xe5 d6 3. hcB dxe5 4. hb7 doesn't lo ok so
bad - picking up the bishop on cB and the pawn on d7. Pretty sure J
have to play 1. V!ie5. Only used 12 minutes (equivalent to six at a two to
one ratio for thinking out loud). 1. V!ie5 V!ixc4 is annoying but I am not
in danger of losing a piece. 1. V!ip was so pretty, too! But not against
1 ... V!idB. 1. V!ia5 V!ixa5 2. 'l'l,xa5 b6 Oh! I have 3. 1'..e4.

de Groot Ernie
Black to move after l.Vfia5 V!ixa5 2.i::!xa5 b6 3 ..ie4
Subject E-1 spotted this key idea but I didn't!

1. V!ia5 b6 anything worth ... ? So I can induce . . . b6 - is that helpful?

It allows fi.e4 with tempo or allows him to develop the bishop. 1. "l!i a5
b6 2. V!ie5 V!ixe5 3. 'l'l,xe5 not that concerned. Oh, I do - it hits the bishop
on f5. Looks like ... b6 is not helpful (to me). 1. V!ie5 V!ixe5 2. 'l'l,xe5 d6
3. hcB exd5 4. hb7 rook moves. Maybe then 5. 'l'l,xa7 - starting to look
very promising. I. V!ie5 V!ixc4 2. 'i'J,e4 ill e2+ doesn't win the knight. 3. ltlfl
ill g3+ + hits rook, so 3. Whl then maybe nothing. Still it looks pro m is­
ing. I. V!ie5 V!ixc4 'i'J,e4 ill e2+ 3. Whl - no, 3 ... V!ixc3 4. 'l'l,xe2 so 1. V!ie5 "/!ixc4
others? No, maybe 2 . . . d6 but I'm certainly not worse. He mig ht g et
equality but I should be OK. No alternatives to 1. V!ie5. 1... V!ixc4 or J . ..
V!ixe5. 1 ... ill e 6? No, 2.he6. 1 . . . move queen then 2. V!ixf4. Gotta mo ve
the queen unless J.'i'J,e8 'l'l,xeB cute but nothing. Alright, J. V!ie5, J. CiJ h4,

128
 Expert and Above

J . 11J e5. So rt of Philidor's Legacy - not enough tempos. Almost ready

to play 1. V!ieS V!ixeS 2. 'iJ.xeS d6 3. hcB dxeS 4. hb7 rook moves. I can't
play 5. 'iJ.xa7 if he threatens a back rank mate, but I have sufficient
co mpensation/or the exchange. OK, 1. V!ieS.
S ubj ect E-1 got off to a rough start, taking quite a long time for a
pl er of his caliber to notice that 1.Vfl'f7 is refuted by 1.. .Vfl'd8. However,
ay
the re mainder of his protocol is a model of good analysis which would
be acceptable to many master-level players. In the diagram he spotted
a key idea which I had missed when I was shown this position: luring
the pawn to b6 allows a counterattack with �e4. Moreover, Subject E-1
fully weighed the various possibilities and chose a line where an ex­
change sacrifice gives excellent winning chances. While he missed the
computer's best line, it was not that he saw a good move and did not
look for a better one. After 19 minutes he found a line which seemed to
win, and Trigger 2 (see Chapter 11) indicated that result was about as
efficient as he was going to get.

Protocol E-2 (de Groot A; adult; 2000; 12 minutes)

Material even, White has an isolated queen's pawn. The knight at
e5 is a decent outpost. Black blockades the isolated queen 's pawn well.
Threats: 1 ... V!ixb2. White can 't actually win a piece: J. hdS hdS does
not win a piece. Feels good to get a knight to e4, e.g. J . fa.xj6 fa.xj6 2. l1Je4
but l.fa.xj6 l1Jxf6 stops that. Like in most isolated queen's pawn games,
White should play for a kingside attack - in the middlegame. How do I
do that? Got to open things up on the king side. A pawn advance? 1.g4.
I want to get inf4 andj5. Need the pawn at g4 - may need a buildup to
make that happen: g4,f4, double rook's on thef-file. 1.g4 weakensf3 and
h3. Can Black take advantage? What would Black do ifI played 1.g4?

de Groot A: Black to move after 1.g4?!

129
The Improving Chess Thinker

1 ... Viixb2? Suppose 1.g4 'il.fdB. Are there any other Black candida tes ?
Don't see [!]. Black threatens l ... Ci:Jxc3 followed by . . . Vfixb2 and is a
piece up [?}. l.g4 V!ixb2 2. 'il.f2 or 2f4 Ci:ixc3 3. 'il.xc3. So g4 is prema tu re
or not good at all, so I need to deal with b2 (silence). l. 'il.c2 - I h ave
more active pieces or l.b3 or J.b4 possible; fibl is possible ijb3 but n ot
likely. l.b4 or l. 'il.c2 possible. l.b4 weakens c3 badly - don't believe he
can take advantage of that: J.b4 Ci:ixc3 2. 'il.xc3 Ci:J e4 big problem. So
no l.b4. How about l. 'il.c2? It guards the b2 square - tidy and tigh ;
- no major threats that will kill me here. 1. 'il.c2 Ci:ixc3 with the idea
of ... 1;.a4. What other possibilities for Black? l. 'il.c2 'il.fdB 2. 'il.fcl iss ue:
2 ... Ci:ixc3 - never mind - it's OK. Would I lose the d-pawn ? No. J.'i'k 2
'il.fdB 2. 'il.fcl and my development is completed with a typical isolated
queen's pawn game. I rate the position dynamically equal [good]. How
do I move forward after that? l. 'il.c2 'il.fdB 2. 'il.lcl...? (silent). I don't see
any major threats for Black orfor White - I can move forward. White
moves l. 'il.c2. Blunder check and make l. 'il.c2, push clock. (PV=l. J;l,c2
'il.fdB 2.'il.lcl with approximate equality).
A very strange protocol for a 2000 player. Subject E-2 hardly con­
sidered any of his forcing moves, not 1.Ci:ixdS or even 1.Ci:Jxc6 winning
the bishop pair. To spend so much time on the very strange 1.g4? ! with­
out investigating the more forcing captures is original, but not indica­
tive of what most players at this level would do. The best way to meet
1.'il.c2 is 1...Ci:Jxc3 when White needs to find 2.Ci:Jxc6! to stay in the game
(although Black is still slightly better). Instead, if 2 .bxc3 then 2 . . . �e4 is
a winning skewer. If 2.'il.xc3 then 2 ... Vfffxb2 is very good for Black and,
finally, if 2.Vfixc3 then 2 ... Ci:Je4 wins but not 2 . . ..b:g2?? 3.Vfffxc8, a line
missed by several players who assumed the discovery on the queen
would automatically win.
The next subject woke up early to provide a protocol.

Protocol E-3 (de Groot Ernie; adult; 2040; 8 minutes)

White is better - on general principles: Black is not developed mu ch.
Black is attacking my queen. Unless I have a check or threat. l. B eB
flashed in my mind: l. 'il.eB Ci:ixd5 2. 'il.xfB#. Maybe he has l . . . Ci:Je6: l. B eB
Ci:J e6 - let's see: three attacking, one defending. l. 'il.eB Ci:J e6 2.he6 J;l,xeB
- so l can't do 2. he6. l. 'il.e8 'il.e6 2. 'il.xf8+ Ci:JxfB and. . . Well, let's see. I'm
a pawn up. He's got a crummy position. What would I do? Fun th en
after 2 . . . Ci:JxfB; can play something like - looking at 3.Ci:JeS and th en
ti:ip+ with general havoc - he could play ... Ci:J e6 again. I would go with
1. 'il.eB ... l. 'il.eB is a total disaster. l. 'il.eB 'il.xeB.Itis too early in the morning !

130
 Expert and Above

So J ca n't
play 1. 'il.eB - I like spectacular moves. So I have to move
my qu een. What about 1. W!.fl using the 'il.eB idea? 1. W!.fl threatening
2. 'l,ffxjB +. He has no checks to discombobulate that. Defend the rook or
move it. 1 ... 'il.dB 2. 'il.eB+ finishes that. So he would have to play 1 ... W!d6
2. Y!. eB atta cking the rook twice - not a bloody thing he can do about
that - tha t's about it. It looks like 1 . . . 0i e6 'il.xe6 pretty much.finished off
the kn ight. Yeah, I am going to play 1. W!.fl. Push clock.
Unfo rtunately, it was probably too early for Subject E-3. As an Ex­
pert you would not expect him to miss both 1.'il.e8?? 'il.xeB - which he
eventu ally did find, and also 1.�f7? �dB! which he did not, but Subject
E-1 did.

de Groot Ernie
Analysis: White to move after 1.�f7? �dB!

Surprisingly, l. ..Wid8 ! is not just a defense - it's a winner. For ex­

ample, the best try 2.tll e5!? still loses to 2 d6! 3.hcB dxe5!, win­
•. .

ning a piece as the queen and bishop are both attacked, and 2 .W!e7 W!xe7
2.l'!xe7 'il.xf5 3.'il.eB+ ©h7 is not nearly sufficient compensation for the
piece for White. I told Subject E-3 to keep in mind that if he thinks he
has found a winning move, slam on the brakes and take extra time to
make sure it is really winning: if so, then the extra time spent will not
hurt him ; if not, then the reason why (in this case l...V!id8 ! ) might turn
out to be so important that the seemingly winning move might not even
be good. Even though he thought 1.V!if7 won, Subject E-3 did not follow
that advice, or Lasker's Rule to look for a better move. If you have time
and don't do either, you had better be sure the win is there! After the
� ercise, Subject E-3 felt a little sheepish at not doing so well (partly
om not getting enough sleep), and asked to do it again:

131
The Improving Chess Thinker

Protocol E-4 (de Groot Shafritz; adult; 2 040; 38 minutes)

Reminds me of a position I played as Black in Canada in the 1 970 's
- some kid took me to school. I thought Black was threatening bu t he
nullified it with e3 and f4. Wow! So I go back to things I re me mber
Funny how you do that. First move I think of .. no pawn on e2! I consid�
er myself to have some problems. Black has a king side demonstratio n.
He has . . . e4 and .. f4. Oddly enough, my.first impression is a defensive
one. Let's see. I don't like .. . look at pawn moves. Don't so anything too
great. If lf4 then l.. .e4 and that might not be too bad - stops all the
crap. Then something like 2. ltJ bS with the idea of 3. f;..c3 to get rid of
his pesky bishop. That's something. Oh, my, I could try 1.j3 to stop that
altogether. Just pawn moves - what are the other candidate moves?
Pretend to Think Like A Grandmaster (laughs). 1.j3 or lf4
What if I do 1. ltJ b5 first? He can't chase it. If l . . . e4 then f4 is avail­
able to the other knight. If 1 .. f4 then I just win a pawn. Sometimes I
just see ghosts. 1. ltJ b5 f4 then 2.gxf4 and then . . . oh boy, he would not
take onf4. J . ltJ b5f4 2.gxf4 . . . If I were Black - White can 't take on e5
to give that square to the bishop and knight. Maybe 2 . . . Wff6 with the
idea of getting the bishop to h3 or to get the rook tof5 - he has all sorts
of things. Seems like a very good position [for Black then]. I can bring
the knight back to c3 to get into the e4 square. He can play ltJef6 to dis­
courage that, or 1 . ltJ b5f4 2.gxf4 Wfg6 3. ltJc3 ltJdf6 then 4.fxe5 dxe5. Oh
my! I can play 5. ltJg3 and I am happy as a clam - a pawn up. So l . f4 .

is not playable and so 1. ltJ b5 f4 2.gxf4 Vfig6 3. ltJ bc3 ltJef6 then instead
of 3 ... ltJ df6 to stop White.from exchanging on e5. He wants to keep e5
open. Now White can 't take on e5, so 4. <tlhl with the idea of5. '8.gl with
my own little counterattack. Pretty happy with that - don 't want to
see Black start to attack, so I don't think 1 . ltJ b5f4 is so hotfor Black.
After 1 . ltJ b5 if1. . . Vfih5 - very "Dutch-like" - trying to.figure out what to
do (short break) 1 . CiJ bS W!h5, so l look at the reason ltJ b5 is to get �c3 to
oppose the diagonal and stop pawn attacks: 2j4 e4 3. ic3 - he doesn 't
want to exchange. So 3 . . . ltJdf6 and I don't want to allow . . . ltJg4 so I
have to play h3: too many weaknesses for White: ... <tlhB and . . . '8.gB to
attack the g-pawn. So maybe we consider 1 . ltJ b5 or lf4. Pawn mo ves
create weaknesses. I can move the queen and he can move to the king ­
side - good spot. I can put the knight on e2 somewhere.
Maybe 1 . ltJ cl looks . . . Then if J . . j4 Umm. I have three defenders
. . . fJ. 1 . . f4 2.j3 V!ig6 then 3. ltJe2 right back where I started however
3 .. jxg3. Would almost rather be Black here; White is holding on here.
Then . . . ltJdf6 to open the bishop. Then . . . ltJ hS. So I have to come up w ith

132
 Expert and Above

so m eth ing. Might want to play the bishopfrom e3 tof2 and be OK. e.g.
!tl hS if2 . Then White is kinda OK. Don 't care for this position - I
�ould rather be Black here. Umm (silent) I don 't know.
It's kind of a closed position but I prefer the [l f4 e4 liJdl} 2 ... liJdf6
3.ltle3 have a n ice blockade.

de Groot Shafritz
Black to move after 1.f4 e4 2.liJdl liJdf6 3.liJe3

Then Vf!a4 or Vf!a3 and then b4 and do that. I don 't know. He has
to play ... lf4 e4 2.liJdl liJdf6 3. liJe3 hS. I can hold that position with
4. Vf!a3. If he plays 4 ... liJg4 then 5.b4 and Ifeel like I am doing some­
thing - he can 't take me. He can play 5 ... id4. Having ... Umm - Trying
hard to imagine the position (silent). Well, I'm looking at the other
alternatives - they were too fluid and helped Black. Here I may have
a chance and block the attack: 4. Vf!a3 id4 - two pieces attacking e3,
two defending - can 't take the c-pawn because of the bishop on d4.
Then I've got to get rid of the bishop with 5. liJxd4 cxd4 then got to
move the knight somewhere like 6. liJ c2. Then he has those pawns - I
ca n blockade and try stuff like c5. He has to defend the d-pawn and
can 't do that. I think I am OK here.
Go back to lf4. I think I would play l f4 then l . . . liJ df6. No, he can 't
- then 2.fxe5 so maybe l . . . liJ ef6 (silent). The d-pawn is not doing
a nything so 2. liJ b5. If he goes something like 2 . . . Vf!e7 that is bad - the
�night can get to c7, so l. .. liJef6 2. liJ b5 liJ e8 to protect cl and d6 so liJ b5
18 good against liJ 8f6. lf4 liJef6 2. liJg5 liJeB 3.jxe5 liJxe5 then 4. liJf4
both sides have nice postsfor their knights. My knights are well-placed
- that 's good for White - the knights have good squares. Try to get
b4 i n - ne ither White nor Black can be successful [in this line] on the

133
The Improving Chess Thinker

kingside. The plan is that after 4. liJf4 the idea is to tie up the king side
- since he [Black] is stronger than I am there. Then I have the idea of
Vfia3 and b4, exchanging pawns - I am making some queenside move­
ment there. 1J4 - he can't move the knights. 1 J4 V!ihS 2.fxe5 Now I ca n
get squares for my knight if 2 . . dxe5: If he plays .. J4 then I ge t /:iJ e4
.

and if he gets ... e4 then liJf4. So those pawns are just hanging there o n
e5 andj5. lf4 f;.h6 - can 't take him [pin on bishop on d2]. Hm m. So
he has to go back to 2. Vfic2 [At this point I gave a 35 minute warning.]
Don 't know. What would he play here? So many moves. Comfortable
blocking thef-pawn - to be honest, I don 't see. l . liJ b5 seems good, but
I can always go liJ b5 so I would play lf4 push clock.
Well, perhaps feeling chastised by playing a little too superficially in
E-3, the Expert more than made up for it in E-4. I think protocol E-4 is
somewhat a case where a player was trying too hard to look deep in some
lines analyzing how good a move was, without keeping in mind that he
only has to find the best move (as was discussed in the comment for
Protocol A-5). True, sometimes you need to look deeply to know exactly
how good a move is. But other times you don't really have to know to
that degree, and the extra depth of search wastes time. This admonition
is especially true when the moves you are considering are not forced.
Many of the lines that Subject E-4 considered for Black were "tries"
rather than forcing sequences that were immediately dangerous. It may
be enough to see that they are not dangerous and save the time for later
in the game, when truly dangerous lines appear on the horizon.

Summary of Expert
Characteristics of this thought process level:
• I once read about an experiment where researchers performed
EEG's on players of different strengths to determine which
rating levels used which amounts of analysis vs. memory to
determine their moves. The researchers made the distinction
by the activity in various parts of the brain. Their conclusion?
The players that do the most analysis are Experts! Playe rs b e­
low Expert tended to do less and less analysis. Players ab ove
Expert relied more and more on their memory. In other wo rds,
Experts know how to analyze and do it quite intensely. As th ey
get more expertise and rise in rating, they recognize sim ilar

134
 Expert and Above

positions they have either analyzed before or seen in literature.

Then they play those ideas more from memory. GM Michael
Rohde once wrote in Chess Life that there are about five moves
per game where Grandmasters don't know what to do and have
to think, International Masters eight moves per game, FIDE
masters ten, and National Masters fifteen.
• As shown by the Expert protocols, sometimes within all that
analysis resides the one big error that one "could drive a truck
through." That's why it is so important to at least double-check
the analysis on the move you are going to play before playing
it - especially if the position is complicated and the move is
critical! Interestingly, this same type of error happened in the
World Championship game the same day I edited these lines:

2 0 08 World Championship
Kramnik-Anand
White to move

Kramnik had 23+ minutes remaining until the first time con­
trol and had thought several minutes on this move (the unofficial
Internet Chess Club clock went down from 23:45 to 16:11 for this
move) before deciding on the critical line of sacrificing a piece with
29. � xd4? Anand thought for only about 2 112 minutes before see­
ing the hole in Kramnik's play and, after Black's move 29 ...Y!?xd4!,
both players rattled o ff the remaining moves very quickly: 3 0 .l::!d l
�f6 31.fui:d4 �xg4 32.gd7+ Not best but already too late. 32 ...
lt>f6 33.gxb7 gel+ 34..ifl �e3! The fly in the ointment. 35.fxe3
fxe3 and Kramnik resigned:

135
The Improving Chess Thinker

2 0 08 World Championship
Kramnik-Anand
White resigns

- which shows that although these type of "big" analysis errors occur
less and less as the level of play increases, they never entirely disappear.
• Experts are usually more thorough and accurate than "A" play­
ers. They know where to extend their analysis and have a better
feel for which replies are critical. While this is true for any set
of levels, the lack of consistency and thoroughness for levels
below "A" makes the contrast more clear between Expert and
"A" since both know to take their time to look for the meaning­
ful lines and do it fairly consistently.
Improvements to the thought process that would help in getting to
the next level:
• Expert is the level where aspects such as opening knowledge
and evaluation start to matter much more. By doing whatever
it takes to improve their evaluation function (such as analyzing
with strong players, comparing their evaluations of positions
to those of computer engines, reading appropriate books, etc ),
experts would be able to take some of that extensive an alysis
and provide more meaningful conclusions. Result: higher play­
ing strength.
• Experts often get into time trouble due to their intense analysis,
so practicing good time management would help many at this
level.
• The observed difference in planning and judgment between
expert and master is high. Therefore, one of the best things an
136
 Expert and Above

expert can do is study many annotated master games or, even

better, hang out with masters, play chess with them and, best of
all, analyze games with them.
• Getting to the next level becomes much more difficult once you
are above 2000!

Master and FIDE Master

Protocol M-1 (de Groot A; adult; 2300; 5 minutes)
EiJd7 is possible. Tons of trades you can do. No real mate threat.
That's not what I would look at. l. EiJxdS hd5 then 2.ixj6 hf6 3. EiJ d7
wins something. 1. EiJxdS then Black must play l... EiJxd5 2.hdS threat­
ens 3. he7 so 2 ... hg5

de Groot A
Black to move after l . .!Lixd5 .!Lixd5 2.hd5
The master easily deduces that 2 ... hgS is forced .

... then t/J d7 is possible. So then the knight is trapped and fJ.xfJ. is
possib le. So J . t/J xdS t/Jxd5 2.hdS hg5 3. E!xc6 gives the d7 square. I
sort oflike that idea. 1. t/JxdS exd5. I guess he would take back with the
pawn ifI took 1. t/JxdS. Ah! Then maybe 2. V!Jj3 and he has trouble hold­
ing the knight onf6. So the only other trade is l.ixj6. I don't see how
tha t leads to a whole lot. So I would probably do 1. t/JxdS push clock.
[PV= J. t/Jxd5 and I didn't see a good continuation for Black].

Arou nd the same time he performed this protocol, Subject M-1 beat
GM Hikaru Nakamura in a tournament game - a huge upset! In this
protocol he easily finds 2 . . . hgS after 1.t/JxdS t/JxdS 2.hdS - in my
exp erience almost no player rated under 1800, in a total of hundreds
of subje cts, has ever calculated this critical move. However, after that
sharp find his analysis is a little vague. Somehow he reaches the conclu-

137
The Improving Chess Thinker

sion (maybe a visualization error, not seeing that the bishop on c6 is

still guarding d7) that this line is good for White, possibly after 3. !'lx c6 .
He likes the idea so much that he never looks for a better move, er­
roneously saying the only other trade is 1.ixf6 when, in fact, there is
also 1.iLx:dS, which happens to be the best move. For a 2300 player, not
considering a major capture is a big oversight, probably explained by
his satisfaction with finding 1.tf:lxdS to be better than it actually is.
Subject M-1 did not verbalize an evaluation, but it's a good bet he
could tell you the material count, which side he felt was better, and why.
When a player gets to that level, board vision becomes so proficient
that this base information is usually quickly assimilated and escapes
the necessity for conscious verbalization.

A friendly FM did two protocols (M-2 and M-3) for me:

Protocol M-2 (de Groot C; adult; 2300; 14 minutes)
Material equal. Queen in a funny spot. Strange squares. White
knight onj3 a target. Behind in development but my king is safer. I can
put a piece on e4 or play . . . �d7. Umm. I see various things not worth
comment. l . . . �d7 loses a pawn but compensation: l . . . �d7 2.he6+
Bxe6 3. V!fxe6+ ©hB with compensation. tf:lg5 a worry - rather annoy­
ing. 4 . . . !'leB 5. tt:ip+ ©gB. l . . . tf:le4 but 2. l!hel but 2 . . . tf:lc3. I suppose he
would have to play 2. ©b2 with the idea of !'!hel. Doesn't look all that
useful. 1 ... V!fe4 leaves me weak on c7 - only good if it was somethin g .
2. !'!hel. lf no follow-up, then it only helps White. 1 . . V!fe4 2. !'lhel V!f g 6
.

3. tf:lxc7 goodfor White. Love to do something to tf:lj3, but doesn't seem

possible. I am more drawn to l . . . �d7 or l . . . tf:lh5 but 2. tf:lg5 makes it
look silly. Oh! l . . . e5 I had not noted.

de Groot C: White to move after 1 e5 ...

138
 Expert and Above

Wou ld take advantage of king-position and pin. Looks like a good

move. l ...eS 2. hcB 'il.acx8 3. CiJ a7 I want to work: 2 ... 'il..fxc8 3.dxe5 CiJ e4
with th e idea of CiJ c3. No, wait a second. Is 4. CiJfd4 a winning move
there? I don't see it if 3 . . . dxeS 4. CiJxeS. Hmm. If it get my queen out of
the ta ngle I am not too optimistic. Can I prepare l . . . eS? Something like
1 ... l!fhB. Now 2. he6 is impossible. So maybe 2. fi.g2 but CiJfJ has no
pla ce to go. 2. fi.g2 Vfie4. No the c-pawn still hangs. Hmm. Strange posi­
tion (silent). 1 ... fi.d7 weird because it runs the queen out ofplaces to go.
If J l!l hB 2. 'il.hel what did that do? Sort of cramped. Back to 1 . . . CiJ e4.
•••

Wait! l ... CiJ e4 2. l!lb2 CiJxj2 could make a huge difference. What does he
do? Oh wait! CiJ b5 defends c3 - embarrassing with CiJ c3. l... CiJ e4 just
looks silly. Hmm. Now I like the position less and less. I want to play
J eS - positively I almost have to do it. 1 ... eS 2. hcB 'iJ..fxcB 3.dxeS does
.•.

not look so good. Is the position so bad as to be worth sacrificing a

pawn? Yeah,frustrating - no way out of it. But I have to make a move.
1 .. @hB probably wastes time and does not make it better. How about
.

1 ... h6? Probably a decent waiting move. l.. .e5 2. hcB 'iJ..fxcB 3.dxeS
and I see no move for Black so I return to general principles so I play
J... fi.d7, push clock.

The de Groot C position gave even the Grandmasters difficulties.

After I gave this position to the computer (Appendix B) I had a difficult
time determining why its best lines were superior. Therefore, congratu­
lations to Dr. de Groot for finding such an unclear position and kudos to
those subjects who successfully waded through the murky lines.
In the E-4 protocol the subject begins by stating that the position re­
minds him of something he has seen before, and Subject M-2 begins by
noting that some pieces are placed strangely. For strong players this ca­
pability to "compare notes" with previous situations is a powerful tool.
Another similarity is length - good players are very careful to take their
time in positions where they are not sure what they are doing - their
sense of the "red light of danger", based on their criticality analysis (see
Section 11.1), gets very keen as a player improves to this level.
Let's list the candidate moves in protocol M-2 : 1.. .Vfie4, 1...CiJe4,
1.. . id7, 1...CiJhS, 1.. .eS, 1... W h8, and 1.. .h6. Subject M-2 was working
hard trying to find something he felt satisfactory. Good players are not
satisfied to play "just something" and see what happens. They are chess
"control freaks" who wish to find something where the continuation
Will be comfortable, and doing the work to find that comfortable line is
often fun for them.

139
The Improving Chess Thinker

Protocol M-3 (de Groot B; adult; 2300; 14 minutes)

Count material: two bishops for a rook and three pawns - ordi­
narily not enough. Second set of rooks helps Black. Vague kingside
attacking chances.

de Groot B: Black to move

Advanced knowledge. FM knows: 1) If there are two rooks vs. one, the
side with one does not usually want to trade off.
2) When behind in material look to attack, especially the king.

Can we open the h-file? Give rook...So 1 ... h5 or 1... Wg7. Hard for
White to organize his position. 1.. .h5 most dynamic. Hardfor White to
ignore that. Can I recapture with the pawn and go to h3? l ... h5 2.gxh5
©g7 3.hxg6 '8h8+ 4. Wgl not too impressed by that for Black. Hmm.
Strange position. 4 ... !J.. h2 + 5. ©fl !J..f4 and it's hard for White to do
anything. 4 ... !J..h2 + 5. ©fl !J..f4 what does White do? He can just push
pawns in the absence of a follow-up. My guess is that White is faster:
1.. .h5 2.gxh5 gxh5. Incredibly hard to stop the bishops. Hmm. Hard
for White to find anything to do at all. 1.. .h5 2.gxh5 gxh5 3.c4 'fl.dB so
4. '8adl. Doesn't look that great: 4 ... '8d4 White can play ©gl and meet
h3 with g3. Another idea 1... '8b8 White guards and then lift rook to
kingside. l . . . '8 b8 2.b3 '8b5. Even there, doesn't seem I am going to ge t
anything concrete. l ... '8b8 2.b3 '8b5 complicated. 3.c4 '8g5 4.d4 '8xg4
5.d5 - quite awkward, actually. IfI can 't get a concession on the atta ck
on the g-file. If White gets in d5 things will be bad for Black. 1 . 'f!. bB
. .

2.b3 '8b5 3.c4 '8g5 4.d4 does not look goodfor Black. Not so easy for
White either. If pawns advance they can be blockaded more easily . I
do win a pawn after all. ... ©g7. White has to commit himself to a p la n.
Best case ... '8g5 ©hl '8h2. Then White has a problem. Then g3. No t so
easy: '8g2+ and ©fl so what? Oh, the idea is ... '8hl# . ... '8h2 .f3. Wha t

140
 Expert and Above

ha ve I gained from this venture? I can push the g-pawn down. This
takes 1-2-3-4. One free move for White without captures and two more
to push pawn to g4. Implies he will play c4, d4, and dS plan and bishop
is dead. With the h-file attack rather do 1... <bg7. Only other thing is
J f!,bB 2.b3 f!.bS. Doesn't matter - can 't stop l . . . f!.bS-gS. Threats to a7
•..

n ot relevant since I can play ... as. ... �d7? . . . hg4 f!.e4 wins bishop. I
get thefeeling White is slightly better overall.

de Groot B: White to move after 2 ... gbs

High-level players are careful to make evaluations and let this
information guide their planning.

1 ... f!.bB 2.b3 f!.bS 3.c4 f!.gS 4.d4 I can at least cause some trouble - dan­
gerous for Black to abandon the back-rank away from the queenside.
Well, I think I prefer White here, but so far it looks like 1 . . . f!.bB. I
might start and wait for some en passant play. Another weird idea:
march the king to grab g4. l ... f!.bB makes things worse there, so
1 ... © g7. It's a nice idea; it's not so easy. White would love to sac the
exchange but he can 't afford to do it. 2.c4 f!.dB 3. f!.adl f!.d4 4. <bgl
un rav els the king. 4 ... �d6 S.j3 �cS. Hmm . ... hS still presses upon the
kingside. So l . . . <bg7 gives both ideas . . . hS or . . . f!.f6 . . . <bgS so more
flexible than l . . . f!.bB. Choosing between 1 ... <bg7 and l ... f!.bB. Well, ah
- this is basically a race. Eventually White will mobilize the pawns so
I would play 1... <bg7. Last check: 1... <bg7 no other hot ideas. Sharpest
is 1 ... hS 2.gxhS gxhS 3.c4 f!.adB 4. f!.adl f!.d4 maybe that's better than I
tho ug ht. l...hS 2.gxhS gives up d4. Maybe equally or more significant.
... h4 d4 f!.dB c4 h3 dS and I'm out of commission. Hmm. 1. .. <bg7 I guess
�.d4 is White's move. 2. . .hS 3.gxhS f!.hB 4.c4 f!.xhS+ S. <bgl with the
idea of d.S. Black is the tiniest bit too slow. So still 1 . . . <bg7 or 1... f!.bB so
l . .. © g7. Push clock.

141
The Improving Chess Thinker

One error I noticed in the FM's analysis was that he often bec ame
fascinated with his apparent main line and looked really wide an d deep
within it, at the expense of looking wider earlier. Interestingly, I am
reading GM Andrew Soltis's great book The Wisest Things Eve r Sa id
About Chess, and # 112 is "It is more important to look around th an to
look ahead."
Notice that Subject M-3 feels that he is worse in de Groot B and thus
knows to try to use his bishop pair for a kingside attack. This type of
general knowledge, when used properly, will create a "mini-plan" which
will drive the analysis. In other words, he is first figuring out what Black
needs to do and then trying to find moves to meet those needs.

142
Chapter 10

Th e Thinking Ca p

10 . 1 I ntroduction
This chapter addresses various aspects of the chess thought process:
1. What it is
2. What are the component parts
3. Why it is important
4. How it varies from individual to individual
5. How it is learned incorrectly by almost all
6. How it can be re-learned to improve your chess play
7. How it is measured
8. How it relates to other important issues, like Time Manage­
ment
9. Exercises you can do to practice a good process, etc.
Let's start at the beginning. Chess is a thinking game, so obviously
your thought process is an extremely important part of your chess pro­
ficiency. Yet almost no one is initially taught a good thought process, so
almost all beginners develop bad habits that must be overcome if they
want to become stronger players.
Normally, when someone learns chess, all they are taught is:
1. How to set up the pieces
2. How the pieces move
3. Basic rules such as checkmate and some draws (like stalemate
or insufficient mating material), and
4. Some tips, like "Keep all your pieces safe," "In the opening,
don't move your Queen out too early," and "If you see a good
move, look for a better one."

143
The Improving Chess Thinker

This is all good and adequate to help a beginner start playing and
enjoying the game. However, it does postpone the important questio n :
"Once your opponent makes his move and it is your turn, what is the
process you should use to efficiently and effectively find and make your
move?"
Great players do not all use the same process. For example, GM Vic­
tor Korchnoi is known as a meticulous calculator, while World Cham pi­
on Mikhail Tai would play the same positions primarily on instinct and
judgment. But all good processes possess common basic elements, such
as not allowing one's opponent to make a one-move threat which could
not be parried. If they did not have this element, then you or I could oc­
casionally beat a Korchnoi or Tai just by threatening a checkmate that
they could not defend against. However, they would never allow us to
make such an unstoppable threat unless in severe time pressure - and
probably not even then !
Moreover, a good thought process must be subconscious, because if
one has to think about how he is thinking, it interferes with the process!
For example, when you first learned how to walk, your brain spent a lot
of processing time trying to figure out how to prevent you from falling.
But once you got the hang of it, you no longer consciously thought about
keeping your balance. Now when you walk across the room you don't
think about which foot to move next or which muscles need to be used
in order to do so.
But when you develop bad habits in your chess thought process and
become aware of them, for a time you need to adjust your process. This
requires conscious effort. At first this intrusion into your chess thinking
is awkward, and possibly counterproductive. However, once you play
many slow games and have a more effective and efficient process down,
you think about the improved process less and less until you just do it,
with markedly improved results. More on this as we proceed.
Before we go any further, let's introduce some definitions, since
there are no standard ones and chess authors tend to use some of these
terms differently:
Analysis The part of the thinking process in which you say to
-

yourself, "If I go there, what is he going to do and then what I am going to

do in reply?" It is the part of the process that creates the mental "tree of
moves" so to speak. Some players call this calculation, but I usually do n 't
use that tenn. If pressed for a definition, I would say that calculation is
the part of analysis that deals with forced sequences, such as tactics.

144
 The Thinking Cap

Evaluation - Looking at a position and deciding who is better, by

how much, and why. Static evaluation is when you evaluate a given
position without trying to move the pieces. Dynamic evaluation is done
at the e nd of each analysis line, after you have determined a potential
seque nce of moves. Note: When someone says "Evaluate this move,"
they are really saying "Evaluate the positions that would result from
this move - assuming each player is trying to make his best move."
Planning - What you do with the information of why someone
is better (evaluation of strengths and weaknesses) and what might be
done about that in future moves. It is how you will try to exploit op­
ponent weaknesses, negate yours, use your strengths, and negate his.
It is the way you are going to try to achieve some general short and
long-term goals.
Threat A move which, if left unmet, could do something harmful
-

(win material, checkmate, damage the position) on the next move.

White to move

1.Y!Jc5 threatens checkmate with 2.Y!Jc8#

Attack To move a piece so that it can capture a piece on the next

-

move. An attack on the King is called a check. (Another definition of "to

attack" is to play aggressively, keeping the initiative.) Note that not all
threats are attacks (a threat to checkmate is not an attack, nor is a threat
to co ntrol an open file), and many attacks are not threats. A queen move
Which attacks a guarded pawn is not usually a threat since taking the
Pawn next move usually results in loss of material.

145
The Improving Chess Thinker

White to move

White is attacking the pawn on g6, but not threatening it.

Candidate move - a reasonable move a player might/should con­
sider.
Killer Move a move which would refute most potential moves
-

and prevent them from becoming candidate moves.

White to move

For example, in the diagram White is considering his moves, but any
move that would allow Black to play 1. "e'h3 with a further unstoppable
••

2 "e'g2# on the next move is refuted. Therefore, 1 "e'h3 becomes a

•.• •••

killer move.
Time Management The process of managing one's clock ti me so
-

as to find the best moves possible in the given time limit. (For mo re on
time management, see Chapter 11.)

146
 The Thinking Cap

10 . 2 King of the Hill

The re is a children's game called "King of the Hill." The idea is to find a
mound of dirt and all the kids try to climb to the top. The one who gets
there first is the "King of the Hill" and the others try to knock him off
and take his place.
Computer programmers have something similar. It is called the first
pass of a "straight sort." Assume there exist a set of numbers and one
needs to find the largest (or smallest). Programmers start by assigning
the value of negative infinity to the King of the Hill and then proceed to
look at each number in the set. Each time they find a number larger than
the King of the Hill, it becomes the new King of the Hill and they continue
examining the next number. Once they have examined all the numbers,
the one left as the King of the Hill is the largest number of the set.
In chess we have a saying, "If you see a good move, look for a better
one - you are trying to find the best one you can ! " When it is your move,
the best move you have found so far is the King of the Hill. When you are
finished analyzing, the move you should play is the "final" King of the
Hill . Every chess player should have something similar in his thought
process when considering moves during slow games, but many don't for
various reasons.
Take the following diagram from a recent event. A student of mine,
rated 1680 USCF, had White and was on the move:

White to move:

Wh ite was sorely tempted by the smothering move 1.liJf7+ winning

the exchange, and played it. However, after 1. llxf7 2.l'!xf7 axb3 he
••

so o n ran into some trouble (if 3.lll xb3 1/!Vb2) and was losing at one point,
al tho ugh he later pulled out the win.

147
The Improving Chess Thinker

When I reviewed the game with him, we reached the diagram m ed

position and I did not know which move he had chosen. I said,
"Well, you can win the exchange with 1.li:lf7+, but when yo u see
a good move, look for a better one ... Suppose you play 1. lh!S , wh at
then? If Black plays 1 .hfS then 2 .fl.f6+ li:lxf6 3.Yfxf6+ ig7
••• •

4.Yfd8+ .fl.t"B 5.Vfxt'8# and if 2 .fl.g7 then (I paused for a secon d or

•••

two) 3.c!ll f7 is a smothered mate! So yes, I would play 1.gxt'8."

Jumping on desirable moves like 1.li:lf7+ without fully analyzing the
consequences and looking for better alternatives is a common mistake
among weaker players. A player gets enamored with an idea or a type of
attack and pursues it without regard to other, possibly better, options.
Therefore, a good thought process needs to contain an efficient way
to search for the best move possible in a reasonable amount of time.
There are many ways to do this, but they mostly boil down to picking
out the reasonable moves (perhaps the moves which implement your
most plausible and effective plans), analyzing each, assuming the op­
ponent will make the best responses to each, finding the resulting posi­
tions, and then evaluating those positions. There is no way around it
- in most positions this takes time, which is why the best players are
usually the last ones to finish at Open events.
Suppose the first move you look at is OK, but you feel the position
promises more. Then you should feel that a further search is worthwhile
and that you might be able to find a better move than your King of the
Hill. If the first move you consider results in a bad position and you feel
your position is not bad to begin with, then you might not even assign
that move as King of the Hill, knowing that a reasonable candidate is
sure to emerge.
Suppose you find a move that exceeds all expectations. The tempta­
tion might be great to play the move immediately, but maybe you have
underestimated your position and can get even more. Our example
position above covers this case. White thought he would be better with
1.li:lf7+ and stopped his search, but if his King of the Hill expectations
had been higher, he might have kept looking.
A computer chess program will show one or more analysis lines th at
look something like this:
11 ply (1/42) + 0 .83 19.ghel d5 2 0 .cxd5 exd5 21.ib4.
If there is more than one line of sequences (the default setting for the
Fritz engine is to show the expected variations of the top three moves) ,
then the top line contains the "King of the Hill" - in this case 1 9. Elh el
- and the line of expected moves on that first line (the best moves for

148
 The Thinking Cap

e ach side) is called the Principal Variation (PV). The "11" on the left
in dicates how many half-moves deep the computer is currently search­
in g, the "(1/42)" means that it is analyzing its best current move out of
the 42 legal possibilities, and the "+0.83" means the computer thinks
Wh ite is better by about 0.83 pawns in the PV.
But once the engine finds a move it considers better, the first move
of the sequence changes:
11 ply (7/42) + 0 .98 19.ll!adl d5 2 0 .cxd5 ll!xd5 21 .tt'2
•

... and that means the new King of the Hill is 19.l!adl. Note that the
new evaluation, in this case 0.98, must be higher than 0.83 or it would
not have changed its King of the Hill.
Humans don't think in this manner, but their intent should be simi­
lar: Consider reasonable moves, assume the opponent's best replies,
evaluate what will happen, and then compare this evaluation with the
one you estimated with your current King of the Hill. Replace your King
of the Hill if the new move results in a superior position. Given the situ­
ation and the time control restraints, continue for a reasonable amount
of time and then play your King of the Hill. (For more, see the discus­
sion about ''Trigger Two" in Chapter 11.)

10 .3 Is It Safe?
While a good thought process is flexible, there are common elements. A
checklist you might use after your opponent makes a move is:
1. Is his move legal?
2. Am I in check?
3. Can I checkmate him by a series of forced moves? (If so, noth­
ing else usually matters . . . ! )
4. I s his move safe (for him)?
5. What are all the things his move does? (In other words, how
does his move change the position, and what can he do now that
he could not do before? What are his threats for next move?)
6. What are my candidate moves?
.. . and so on.

. For now we are just going to consider the fourth item on this list: "Is
his m ove safe for
him?"
• While reviewing my students' games, after some moves I blurt out,
He ca n't do that - it's not safe ! "

149
The Improving Chess Thinker

That means the move leaves the piece in immediate jeopardy. I t is

not usually subject to a combination, but rather one of three si mple
possibilities:
1. The piece is en prise - that is, it is attacked and not defended,
2. The piece is subject to capture and there is a sequence of cap­
tures on that square that loses material, or
3. The piece is instantly trappable - usually that means it ca n be
attacked by a piece of lesser value and has no safe way to re­
treat.
As an example, take the sequence 1.e4 c5 2.�fJ �c6 3 . .ib5 a 6
4 . .b4??

Black to move

The bishop is not safe because Black can perform a simplified "No­
ah's Ark" pattern trapping the bishop with 4 ...b5 5 . .ib3 c4.
In many cases where a clearly unsafe move was made in my students'
games, neither my student nor his opponent realized that the piece was
not safe! They were too busy asking themselves other questions, like
"What does that move do?" or "What are his threats?" or "What are my
candidate moves?" When I show a position to a student and ask, "Black
has just moved X. What should you do as White?" the first thing they
usually do is look for candidate moves. This may work if it is a "Wbite
to play and win" problem but, in a game, looking for candidate moves
before you check for the safety and purpose of your opponent's move
can be disastrous. If a piece has just been made unsafe, I almost al ways
spot that first, but my student sometimes sees the problem only afte r
I bring the issue to his attention. The fact that sometimes they d o s ee

150
 The Thinking Cap

it immediately means that their thought process is faulty, because they

obviously can see the move was unsafe; they just don't always consid­
ered the possibility. Therefore, the immediate safety issue is automatic
_ an d important - to me, but not automatic for them.
Many students are almost as likely to spot an unsafe move by their
opponent in a speed game as they are in a slow game! The reason is
simple: in a speed game they expect their opponents to make silly, ma­
te rial-losing moves and are alert for those kinds of mistakes. In a slow
game, they assume their opponents will not make such bad mistakes,
and so they just skip that safety check, even though their opponents
might be weak or playing fast and may just be giving them material.
A common, related mistake is for a player to remove his own piece's
guard ! He takes a piece that is defending another and moves it, leaving
th e one that was adequately defended unsafe. For example, consider
th e position after 1.e4 c5 2.c3 �c6 3.d4 cxd4 4.cxd4 d6 5.�fJ g6
6 . .ic4 .ig7 7. 0 - 0 �f6 8.�bd2 0 - 0 9.�g5?

Black to move after 9.�g5?

After this error Black often becomes so fixated on the new "threat"
to fl that he fails to notice that the knight that moved to g5 was is no
longer guarding d4. Since fl is adequately guarded and 9.�g5 has
made the pawn on d4 unsafe, Black should just calmly capture it with
9 �xd4. But instead he often plays an unnecessarily passive move
•••

like 9 . . e6? From this example, you can see that you should not only ask
.

Yo urself if the moving piece is safe, but if the moving piece is leaving
so mething else unsafe.
Another mistake many players make is to break off analysis in the
mi ddle of a forcing sequence (captures, checks, threats) and to evaluate
the positi on before reaching the end of the sequence. This is called a

151
The Improving Chess Thinker

quiescence error because they think the position where they stop and
evaluate is quiet when it is not. Consider the following example:

White to move

Black has just played U'!d8-e8. White might calculate the sequence
of captures on d5 and think that Black has removed his own guard as in
the previous example, but the d5 pawn is still safe because if 2.�xd5??
�xd5 3.llld5? l'!el#. For White to stop his analysis after 3.l'!xd5 and
conclude that he is up a pawn is a quiescence error because he is not
searching one move further to see the checkmate on el.
Sometimes players assume an opponent's piece is safe, but they don't
consider all the key lines and see that the entire capturing sequence is
good for them. An excellent example of this occurs in the Ruy Lopez af­
ter 1.e4 e5 2.�:f3 �c6 3 . .ib5 a6 4 . .ia4 �f6 5.0-0 .ie7 6.l'!el 0-0?
7..ixc6 dxc6 8.�xe5 Y:Yd4 9.�:f3 .ig4??

White to move

152
 The Thinking Cap

I have shown this position to students of many levels, even letting

the m kn ow that 9 ... il.g4?? is a mistake and that it is now White to play
and win. Yet even after they are aware that White can win, many con­
tinue to make the error of assuming that their knight is pinned and
cannot move, and so look to just preserve their extra pawn with a move
like 1 O .d3? But when you see a good move, look for a better one. White
has a much better one by moving that pinned knight with 1 0 . �xd4
_b:dl 11.!b:dl. This wins a bishop since Black's queen was unguarded
and White's queen was not. Therefore Black's move 9 ... il.g4?? did not
make his queen safe, as one should readily find if asking the right ques­
tio n: "What happens if l still capture the Black queen?" If White misses
1 0 .�xd4, this type of error is similar to a quiescence error because
White wrongfully assumes the position is quiet - and bad for him - if
he ever moves his knight and allows his queen to be captured. The "tac­
tic" that wins material after 1 0 . �xd4 I simply call counting. Counting
is the ability to determine if any sequence of captures can win mate­
rial, and becoming proficient at it is one of the most important parts of
learning to play well.
None of the above mistakes involve difficult combinations; all are
basic safety issues. Even good players can make these kinds of simple
mistakes occasionally - especially in time trouble. Therefore, unless you
are playing a computer or a very high level of competition, don't assume
that your opponent's move is safe. Even if your opponent's move is safe
on 99% of his moves, that means once every 2-3 games you are going to
be given a gift! Asking about each move "Is it safe?" is important part of
the thought process.

10 .4 Three Levels of Chess Thinking

I have studied players' thought processes for about 40 years. I am con­
vinced that one way to differentiate levels of chess play is to separate
playe rs into three categories, depending on how they deal with threats
when playing slow chess:
1. Does not pay attention to all (or sometimes even any!) of the
threats generated by the opponent's previous move.
2. Does pay attention to all the threats generated by the opponent's
previous move, but, before making their current move, does
not check to make sure that all checks, captures, and threats by
the opponent on the next move (in reply to the player's move)
can be safely met.
153
The Improving Chess Thinker

3. Not only deals with opponent's threats from the previous move
but, before making their move, also makes sure that the oppo­
nent will not have any checks, captures, or threats that cann ot
be met after that move, and does this check on every move pos­
sible/necessary.
Several years ago I published this thesis and dubbed these th ree
levels of thinking Flip-Coin Chess, Hope Chess, and Real Chess res pec­
tively. Defining these levels in relation to threats was published in my
book Looking for Trouble.
Let's take the "killer move" example from 10.1 and apply it to the
thought process definitions.

White to move

Suppose White sees that Black is threatening his knight. That means
he is not playing Flip-Coin Chess since he did see Black's threat. But now
suppose he is playing Hope Chess and decides to save the knight with
1.<tlxe4? Then after 1 Y:rh3 White would be unable to stop the threat­
•••

ened checkmate on g2. If White were playing Real Chess, he would see
that Black is not only threatening his knight, but also 1 Y:rh3, and would
•••

try to find all the moves which would stop both. As the reader may al­
ready have seen, the best way to do that is 1.gds+ Y:rxdS 2.<tle6+ with
a good king and pawn endgame. To reject 1.gds+ because it is not safe
would be a quiescence error.
Perhaps Hope Chess was a bad choice for the middle level, because
many readers confused Hope Chess with the common "hope" problem
of making a threat and hoping that your opponent will not see that threat,
or, similarly, making a bad move and hoping the opponent makes a wo rse
one. I have dubbed these mistakes (they are not thought processes as
much as they are bad habits within a process) as just Bad Chess or p er-

154
 The Thinking Cap

haps even Hopeful Chess, but not Hope Chess, as defined in #2 above.
Flip-Coin Chess describes the thought process used by most young­
sters right after you teach them how to play. They haven't developed the
ab ility to anticipate their opponents' moves, and the winner is usually
the one who makes more, or larger, threats that are duly ignored. At this
level of play, threatening checkmate, no matter how bad the move is
otherwise, is often rewarded - which of course leads to bad habits.
Hope Chess is practiced by 99%+ of the adults who do not play in
tournaments, and by almost all tournament players rated under -1600
USCF. I have run into several players rated - 1300 who tell me that they
h ave read my material on Real Chess and are now happy to announce
they no longer play Hope Chess. Unfortunately, upon testing them, it
turns out they still play Hope Chess (else their rating would not remain
at 1300). When I find evidence of Hope Chess in their play and explain
the concept of Hope Chess to them, their most common answer is, "Oh!
So that is what Hope Chess is! I thought it was ... " and then they go on to
describe something else, often Hopeful Chess.
I have had a few strong players tell me that my theory is wrong be­
cause they do not use what I describe as a Real Chess thought process.
But upon inspection it almost always turns out that they actually do,
or at least they incorporate the minimum criteria for Real Chess. For
example, if you don't play Real Chess, then you often allow unstoppable
threats by your opponent. Strong players rarely allow such threats
and, therefore, must use this aspect of Real Chess to reject candidate
moves that allow them. However, they may not realize they are using
this process because they have been doing it so automatically for a long
time and are not consciously looking for all upcoming checks, captures,
and threats. Otherwise, any 1300 player could occasionally beat a very
strong player (say an IM or NM) who allows such an unstoppable
threat. Unless the strong player is in severe time trouble, this happens
so rarely that we can easily conclude that strong players avoid unstop­
pable threats, consciously or not.
It becomes natural for strong players to think, "Suppose I make
move X, then what will he do? Suppose he then plays Y, threatening Z,
what can I do?" If the answer is "I have no defense against Z, so I would
lose," then they discard X as a candidate. Of course it takes good board
vision and analysis skills to quickly recognize all of the forcing Y's that
the opponent can do to you, and, further, to figure out whether the re­
sulting threat Z is stoppable or can be allowed. There is a great amount
of skill involved, which is one small reason why the range of Real Chess
Players is so very large (approximately 1600 to 2800).

155
The Improving Chess Thinker

Bottom line: if you don't use a Real Chess thought process, you
probably will never be a strong player. However, even if you do practice
Real Chess, there is no guarantee that you will be a strong player either!
You still have to learn about all the other important concepts that mo st
players study: openings, endgames, pawn structures, planning, tactical
patterns, etc.
One key to graduating from Hope Chess to Real Chess is checking
for upcoming danger on every move, and notjust most of the time. For
example, suppose you "only" play Real Chess on 95% of your moves but
on the other 5% allow unstoppable threats. Assuming the average game
is 40 moves, twice each game (5% x 40) you open yourself up to an im­
mediate loss. If you allow these two oversights each game, your rating
will be much lower than if you play Real Chess on every move. After all,
it only takes one bad move to lose a game! If you play 1700 strength for
38 moves but on two moves play at only a 500 level, what do you think
your average playing strength will be for the entire 40 moves?
It is not a mystery that players who read many books and accumulate
decent chess knowledge often lose to players with much less knowledge.
The "well-read losers" can attribute their losses to talent or luck (usu­
ally the latter! ), but often they are simply not playing Real Chess. This
also explains why 1900 players with relatively little chess experience
(but who are "good game players") can easily beat 1500-1600 players
with far more experience - the two main reasons are tactical ability and
a better thought process.
What does it take graduate from Hope Chess to Real Chess?
1. The knowledge of what Real Chess requires.
2. The desire to do it (if it is fun, you will).
3. The opportunity to practice it (slow games at 90 minute or
more per side are helpful).
4. Practicing it until you don't have to think about it - this is actu­
ally one of the easier parts, since at first you will think about
your process and it will be distracting, but if you play enough,
checking for danger becomes subconscious, like walking.
5. Practicing Real Chess consistently, persevering move after
move, game after game. There is a fine line between desire an d
being careful - no doubt, naturally careful players have an ad­
vantage, just as players who are naturally good with any chess
skill also have an edge.

156
 The Thinking Cap

Im ple menting a good thought process - combined with having the

opportunity to hone this process through many slow games against
st rong o pposition - is a key to unlocking that "barrier" that separates
weaker players from those at the next level - or two!

10.s Checks, Captures, and Threats

chess players use analysis primarily to determine what might happen
if a certain move is played. This creates a "tree" of analysis with moves
branching at each ply. However, the number of legal possibilities is very
large, so humans "prune" their tree of moves to only include ones they
think might be (at least at first glance) reasonable. If this is not done,
then the tree can grow astronomically! For example, if each side has 30
legal moves, then to look at even 2 moves deep (4 ply) would require
30 x 30 x 30 x 30 sequences, or 810,000! Hardly possible for a hu­
man in a year, much Jess in a game requiring 40 moves in two hours.
Similarly, pruning the tree is necessary even for computers, which use
" full-width" searching to examine all moves, but alpha-beta (and other)
cutoffs to avoid looking at the entire tree.
One way to help categorize what happens when a human prunes his
analysis tree is to break the type of moves into different categories:
1. Checks
2. Captures
3. Threats
4. Moves that increase your pieces' activity or decrease the oppo­
nent's pieces' activity
5. Forced moves, like "the only legal" move, or necessary recap­
tures (note that forced moves are almost the opposite of forc­
ing moves like # l-3, although it is possible for a forced move to
be a forcing move! )
6. Other moves
One way to control the tree - and, happily, to find the best move
- is to try and maximize the situations where the opponent's moves are
forced. In other words, if your opponent has to play a forced move and
does not have the flexibility to do whatever he wishes, then the tree is
not only smaller, but more under your control. If you can consistently
fo rce your opponent to reply to your forcing moves we call this pleasant
situ ation having the initiative.

157
The Improving Chess Thinker

For this reason, when you are searching for your best move or co n­
sidering your opponent's best reply, you want to start with (or assume)
the most forcing moves, which are almost always checks, captures, and
threats. Although most checks are more forcing than most captures and
most captures are more forcing than most threats, this is not always the
correct order of how forcing your potential moves may be. For example,
a threat to mate-in-one on the next move is often more forcing th an any
capture, and even some checks. Therefore, when searching for forci ng
moves, always consider mate threats near the top.
The Real Chess thinking process requires a player to see if his candi­
date move is refutable before considering it further, much less playing
it. At a minimum, this usually means asking:
"Suppose I make this move. Then what are all my opponent's forcing
moves in response, and can I safely meet all of them?"
At this point you consider each of your opponent's checks, captures,
and threats, making sure you have an adequate answer to each. If you
do, then the move is plausible and it may remain a candidate move.
However, not all forcing moves are good moves, and some bad­
looking moves are often actually good moves because of their forcing
nature. Let us consider two examples. The first is problem #225 in the
first edition of John Bain's excellent book on beginning tactical motifs,
Chess Tactics for Students:

White to move and Win

The answer is 1.!3xe7 @xe7 2 . .ib4+ winning a piece. But let's ask
a different question: Excluding checks and captures (which sometimes
can also produce threats), how many White moves are threats? Th is
includes only immediate threats, not long-term threats such as to create
an outside passed pawn. Hint: not all threats are good moves; they ju st
have to be able to do something harmful next move. To find a th reat,
158
 The Thinking Cap

cons ide r a move, skip the opponent's reply, and then see if the next
m ove can gain something.
The answer is 5: 1.i.a6 and 1.i.c6 (they both threaten 2 .hb7),
1.ib4 (threatens to capture on e7), while 1.h5 and 1.i.e5 threaten
the removal of the guard tactic 2 .hc7+ lt>xc7 3.l'!xe7+. On the other
band, 1.g4 does not threaten 2 .g5 since after 2.g5 li:lh5 holds g7. Notice
i.ia6, 1.Sl.c6, and 1.Sl.b4 are terrible moves that allow a capture, but are
legitimate threats.
However, not all bad-looking threats are actually unplayable and a
good player needs to keep an open mind and investigate. The second
example is the famous position Bernstein-Capablanca 1914:

Black to move

Black's move 1 Y:Yb2! at first looks like a typographical error. How­

•••

ever, it not only contains two strong threats, 2 . . . 1/;\lxe2 and 2 ... 1/;\lxc3,
but the combination of both is unstoppable . White resigned. White
.

clearly cannot play 2.1/;\lxb2 due to 2 . . . l'!dl# If White had played the
tricky 2.l'!c8, Black would play 2 . . . 1/;\lal + 3.1/;\lfl 1/;\fxfl + 4.lt>xfl l'!xc8. Of
course not 2 ... 1/;\lxe2?? 3.l'!xc8# nor 2 . . . l'!xc8?? when 3.1/;\lxb2 guards cl
and stops the mate. White can try 2.l'!c2, but then 2 . . . 1/;\lbl + 3.1/;\lfl 1/;\lxc2
wins. Or if White tries 2 .1/;\lel then 2 . . . 1/;\lxc3 !
These examples show that it pays to consider all forcing moves, in­
cluding all threats. These include any moves your opponent may be forc­
ing you to consider by his previous move, your current forcing moves,
plus (importantly) his possible forcing replies to your candidate moves.
If you have the time, include the moves that may appear silly at first.
In su mmary, when analyzing, the candidates that should be of most
concern are the forcing moves: checks, captures, and threats. On of­
fense, if you can successfully continue to make forcing moves, you have
the initiative. Defensively, if you can successfully identify and plan to
159
The Improving Chess Thinker

meet forcing moves by your opponent (or avoid them altogether!), then
you will not be surprised, can usually stay in the game, and will oft en
have good chances to win - or at least draw - against formidable op­
ponents. Therefore, learning to identify and deal with forcing moves is
an important part of becoming a proficient chess analyst!

10.6 Space: A Means to an End

I am often asked questions about criteria used to evaluate chess posi­
tions. This is a subject I find very interesting: my first book, Elements
of Positional Evaluation, was written in 1974 on a typewriter and ad­
dressed this subject extensively.
One interesting criteria is "space". In order to discuss space, we
should first attempt to define it:
Space is the amount of area (roughly measured in squares) be­
tween your pawn chain and your first rank (including the first rank,
but excluding the squares of the pawns, which are inaccessible to your
other pieces).
Let's test this definition: As the pawns disappear, the amount of
space becomes more amorphous and, as seems reasonable, disappears
entirely in a pawn-less endgame. That intuitively makes sense; there­
fore so far, so good.
Having more space is usually good for all those advantages we desire:
• Allows more room for your pieces to maneuver,
• Allows less room for your opponent's pieces to maneuver, and
• Gets your pawns closer to promotion, which also makes sacrifi­
cial promotion combinations possible.
But now the important part; consider the following pawn structure:

160
 The Thinking Cap

It is fairly safe to state that, by any measure, White has much more
space. Further, if we place the pieces as follows (admittedly an extreme
exam ple ! ) , White's pieces are dominant:

A great triumph for White's space advantage!

But suppose we leave the "space" unchanged (and even the king po­
sitions) and change the positions of the other pieces:

Now Black's pieces have managed to "get around" and penetrate the
"White space", resulting in an enormous reversal of fortune.
You may think this example is far-fetched, but such penetration oc­
curs fairly frequently. For example, consider an Open Sicilian where
White castles kingside and correctly pushes his pawns for a kingside
attack. But suppose that attack is misplayed, fizzles, and Black breaks
thro ugh in the center. Then White's advanced pawns often leave him
With an exposed King in the middlegame, and this infiltration can be
the de ciding factor in Black's favor.

161
The Improving Chess Thinker

Although the above diagrams are extreme examples, they help prove
an important point! Space in and of itself is not an inherent a dv a n ­
tage; it is a means toward an advantage. The real advantag e of hav­
ing space is that, if utilized correctly, it allows your pieces to do m ore
than your opponent's pieces ("better a rmy activity"). No more, and no
less. If a spatial advantage does not incur an activity advantage , then it
can be - and likely is - meaningless.
Another common example is a king-and-pawn endgame with both
kings centralized and the pawns locked, with one side having much
more space (you can even use a similar pawn structure to the p revio us
examples). Then the side having more space is often at a disadvantage
since the enemy King has gotten "behind the lines" and can be used to
"elbow out" the King whose side enjoys the space advantage. Take the
following position with Black to play:

Black to move

Black, with the space advantage, loses because White has the oppo­
sition, but also because Black's pawns are so advanced and vulnerable.
So, on the average, having more space lli an advantage, but the ad­
vantage usually has to be in the form of more activity or the potential to
promote pawns. The real end goal is more active play, not space.
The same argument could be made about time/tempos as about
space. For example, in a given position one side can be granted free ex­
tra tempos, but if those tempos are used unwisely to place pieces on less
effective squares, then the "extra" time would not be helpful. Therefore,
time, like space, takes a back seat to the real goal: having more - an d
better - things to do with your pieces.
In Elements of Positional Evaluation: How Chess Pieces Get Their
Power, I define the term mobility as the number of squares to which

162
 The Thinking Cap

a
piece can move. Mobility is extremely important since piece value is
highly correlated with mobility - a queen is more valuable than a rook
]Jecause it can also move like a bishop. But the value of mobility in a
real position can, like space, also vary in that a piece can have many
usel ess moves. Therefore the real goal is to have a "good" piece (or, in
total, army) - that is, activity. No matter how you define activity, when
pieces have many good things to do - they are mobile, flexible, attack
key points, and so on - that is the real advantage.
In the aforementioned book, I call space a pseudo-element. That
means space exists as an understandable and useful concept, but is
not really an elemental basis to evaluate a position. So what are good
measures of positional, static evaluation? Total piece activity covers
the useful byproducts of space and time. The four most important static
evaluation criteria, in order of importance:
1. Material
2. King Safety (you can make this #1 if the King is extremely un­
safe ! )
3. Total Piece Activity
4. Pawn Structure
The fifth, "non-board" consideration is the clock time factor, which
can become the most important one during severe time pressure. Time
remaining is always a big factor in faster time control games.
It is not that pawn structure is unimportant - it is, or it would not
be on the list. For example, in common positions where the material
is even, both kings are safe, and both armies equally mobile, pawn
structure can easily be the deciding factor. Many weaker players greatly
overrate pawn structure and think they are winning when their pawn
structure is somewhat better, even in the face of clearly more active
enemy forces! Similarly, sometimes players strive for more space and
get it, only to find that active and flexible opponent pieces make their
space advantages rather moot! This happens in many openings; Open
Sicilians and the Modern Defense readily come to mind.
To arrive at the proper dynamic evaluation requires analysis to
re ach positions worthy of evaluation: quiescent and speculative. These
are covered in the next section.
In sum mary, next time you consider pushing pawns to make space,
evalu ate the likely result to make sure you are providing an advantage
fo r your army. If so - and there is nothing better to do - go for it!

163
The Improving Chess Thinker

10. 7 What Makes One Move Better Than Another?

One question I ask chess students is "What is your main goal each
move?" Sometimes I get reasonable, but erroneous answers like:
"I am trying to win" or
"I am trying to make my position better."
... but the correct answer is "I am trying the play the best move" or
more accurately, "I am trying to play the best move I can find, given th �
time constraints."
The interesting follow-up question is:
"OK, but then what makes one move better than another?"
We touched upon this question in "The King of the Hill," but now we
are going to focus on what exactly it encompasses. This is important,
for "What makes one move better than another?" is at the very heart of
what it means to play good chess!
First, let's eliminate the plausible but out-of-bounds answer "The
move that best follows the correct plan." While this may be an excellent
way to help choose your candidate moves, it does not answer the ques­
tion directly.
Let's list common (but not correct) answers:
1. "It leads to a win of something."
2. "It does more."
3. "The position after the one move is better than the position af­
ter the other."
The last answer is almost correct! In theory it is correct, but for hu­
mans it is usually impossible to apply (as will be shown below), so it
won't do. However, this answer does address the fact that we evaluate
positions rather than moves, so if one move is better than another, that
means it must lead to a position that is better than another. However,
which positions should we use?
The problem with using the position immediately after the candi­
date move is that the position may be clearly non-quiescent. For exam­
ple, suppose you compare a "nothing" move like >t>hl with a move l ike
'Wxh7+ , capturing the h-pawn. It may seem better to win a pawn with
check than just move the king, but what if your opponent's next move is
. . . >t>xh7, winning the queen for the aforementioned pawn? So we must
look further to see what happens after 1Wxh7+ , or else the evaluation
that White is ahead a pawn is meaningless.
Generally, we stop our analysis at a point where we can evalu at e as

164
 The Thinking Cap

foll ows:

1. The position is quiescent. There are no more meaningful checks,

captures, or threats which would change the evaluation. Most
non-analytical positions are quiescent immediately after candi­
date moves.
2. The position is speculative. Then we use our judgment - for
example, we sacrifice a piece to expose the enemy King, and
even though we cannot practically analyze to quiescence, we
judge whether or not the exposed King is worth the sacrifice.
3. A sacrifice fails since the further possibilities cannot possibly
give us a return equal to or greater than our sacrifice. For ex­
am pie, if you sacrifice your queen and see that you might later
win back a rook, there is no sense in analyzing further to see
whether that is true, since the risk is greater than the reward in
any case.
The first of these three is the most common.
However, knowing where to stop analyzing a line still leaves the
question: With a large tree branching from each move, which positions
are the ones we want to evaluate? In other words, which branches of the
tree are meaningful? The answer is that we must assume - to the best of
our ability to judge - the best moves for each side:
A move is only as good as the positions that will be reachedfrom it,
assuming best m oves for both sides!
If you are not good at judging what the best moves are during
analysis, you will arrive at the wrong positions and reach the wrong
conclusions when you evaluate. This weakness, along with an inability
to accurately evaluate positions with approximately equal material and
the inability to recognize quickly and accurately the common tactical
motifs, are the three of the biggest "thinking" problems of weaker play­
ers !
So now we have our answer. Suppose we have analytical moves
"A" and "B" and we want to know which is better. First we attempt to
determine the best sequence after A and B, only going as deep as neces­
sary to evaluate, for example:
My move A followed by his best move A' followed by my best move
A" leads to quiescence (graphically):
A - A' - A" vs.
My move B followed by his best move B' followed by my best move
B", his best move B"' and my B"" leads to quiescence:

165
The Improving Chess Thinker

B - B'- B" - B"' - B""

This leaves a position after A" we feel we can evaluate. Let's call this
position A* and the evaluation of this position E(A*).
The same holds for B: the position after B"" we call B* and we eval u­
ate it as E(B*).
Then we compare the evaluations and decide which positio n we
like better. IfE(A *) is superior to E(B*) then we like move A better!
Good players spend a great deal of time doing this when th ey are
thinking!
Finally, we apply the King of the Hill method discussed in Section
10.2, and apply this to all our candidate moves in order to find the best
one.
There are shortcuts to this method. Depending upon the position,
non-analytical moves are made more on judgment than the above type
of analysis. Good players don't consciously go through all this, but this
is basically what should be happening when we think, or else we will
end up with a less than optimum move. For example, many beginners
don't even follow the guideline When you see a good move, look for a
better one - you are trying to find the best one!
Let's see how this comes out when a computer does it. From Section
10.2, we know an engine's analysis window displays something like:
11 ply (1/42) + 0 .83 19.!::!h el d5 2 0 .cxd5 exd5 21.i.h4 .
According to "game theory", your position cannot get better - or
worse - if both you and your opponent make the best moves from a
given point in a game like chess. Since playing the best move keeps the
current evaluation from the previous ply, the engine assigns the value
from the final position back to the original move! So if the position at
11 ply deep is evaluated at + 0.83, then it judges 19.Rhel to be "+0.83"
as well and assigns it that value. The lines below the PV reach positions
that are evaluated as not as good and will be assigned lower values.
Next time you are trying to find the best move, keep in mind what
you are really trying to do: find the likely optimum position(s) that arise
from this move and compare them to similar optimum positions that
arise from each of the other candidate moves. Assume best play on both
sides - don't wish that your opponent will make bad moves! The move
that, in your judgment, leads by best play to the position you like the
most is the move you should generally play.
Of course, analyzing in practice is difficult (and many books are filled
with practical advice on doing so) , but it helps to start by understanding
what, in theory, should be happening. I hope laying this groundwork
helps!

166
 The Thinking Cap

10 .8 Only Part of the Analysis Tree Changes Each Move!

one fact that makes chess "easier" is that only two things change since
the previous time you had to decide on a move:
1. Your previous move

2. Your opponent's response

That means if you perfectly understood what was happening on the
board last move - and that's a big "if' - then, unlike a problem in a
book, you don't have to start from scratch to figure out what is happen­
ing this move - you only need take into account the consequences of the
two moves that just occurred.
An easy example: suppose your rook on dl was safe on the previous
move, and no direct or indirect consequence of either new move af­
fected dl. Then the Rook must still be safe.
That forms the basis for guidelines such as "Each time your opponent
makes a move, ask yourself, "Wlzy did my opponent make that move?
What are all the things my opponent can he do now that he could not
do before? "Wlzat are all the legal moves my opponent can make now
that he could not make before?"
These guidelines often overlook the other consequence of the op­
ponent's move, which is, "Wlzat was my opponent's piece doing before
his move which it is not doing now? This idea of checking to see what
a piece is "no longer" doing is even more important when considering
your move: If I move this piece, then what was it doing before that it
will no longer be doing now? I have many students who occasionally
lose material because they are so interested in what a piece will be doing
after their move that they forget that it was doing something important
before, like guarding material or preventing checkmate.
Interestingly, the extra knowledge about only taking into account
recent changes works better for stronger players. Why? Because stron­
ger players have better board vision and more seasoned logic to account
for the changes since the previous move. Weaker players, due to lack of
board vision or even carelessness, often do not take into account all the
aspects of a position and move hastily or with incomplete knowledge.
In addition to the changes since the previous move, there are "carry­
over" issues from the previous move. Carry-over issues are primarily
due t o three factors: threats that had been purposely ignored, zwischen­
ZU.gs which left primary threats on the board, and blind spots/blunders
Wh ich overlooked a threat on the previous move. For example, suppose
a pl ayer does not notice an exploitable weakness in his position which

167
The Improving Chess Thinker

his opponent mistakenly does not exploit. Since his opponent's m ove
does not exploit the weakness but rather does something else, the n if the
player thinks, "What can he do now?" based solely upon the oppon ent 's
previous move, he will continue to overlook the leftover weakness an d
will again allow it to be exploited on his opponent's upcoming move.
When this oversight happens to both players, it often leads to a "co m­
edy of errors" where both sides continue to miss important issues that
should have been addressed several moves ago.
This leftover problem does not occur nearly as much with stronger
players, and not just because of their superior board vision. Stronger
players also face stronger opponents, who do not often leave "unattend­
ed business" lying about the board. In other words, if a strong player
overlooks something, then his opponent usually exploits this error, and
the overlooked problem is resolved in the opponent's favor. But with
weaker players, once both players have missed an idea, that leftover
problem may remain around several moves until either one player "ac­
cidentally" notices it and either fixes it (if that player was the culprit) or
takes advantage of it (if the noticing player is the opponent), or one of
the players unknowingly makes a move which changes the position so
that the leftover problem disappears.
An example from a slow game played recently on the Internet:

White to move after 33 ...hc4

White, seeing the loss of his pawn on c4, completely overlooks th at

Black is also threatening the pawn on a2 and too quickly plays 34.f.J,
allowing Black to win another pawn with 34 . . . exf3 35.hf3 b:a2 . In­
explicably, Black also misses this and plays 34 . . .f5, instead guard ing
the e-pawn. White finally realizes the leftover problem on a2 and plays
35.fxe4 fxe4 36.a3 saving the pawn.

168
 The Thinking Cap

One interesting side effect of weaker players allowing leftover

p roblems is the following: Even if they are taught to look only for the
differences in the position since the previous move, they often adjust
their thought process to include a search for leftover problems. While
this widening of their thought process is understandable, it is terribly
inefficient; it causes a weaker player to approach positions more like a
problem, in that he has to find and solve everything each move, and not
just carry over from the previous move the pertinent safety information
and only account for the differences. This extra work often becomes a
burden for the weaker player who, probably correctly, does not fully
trust his own analysis.
What can a weaker player who wishes to play efficiently, but is afraid
of overlooking a "leftover" problem, do? There is no perfect answer, but
here are some suggestions:
1. Look for leftover problems during your opponent's move. But
if you find one, don't slap your head and yell, "Oh, what did I
miss? ! " That tends to tip your opponent off, and he may use
his move to identify and fix - or take advantage of - what you
found. Some players make exclamations like this purposely as
a ploy to lure their opponent into a trap. However, that ploy
is also counterproductive because if your opponent is smart
and you alert him - even as a ploy - he will be less likely to
fall into the trap, since the main two possibilities are that you
blundered or that you are trapping him, and he will figure out
which one it is !
2. Work on improving your board vision so you miss less in your
analysis. This can be done by doing board vision puzzles or just
playing appropriately slowly in many slow games.
3. Take your time and be careful on each move. If you are thor­
ough, then you are likely missing less. That means you can be
more efficient next move, and not have to worry about leftover
problems nearly as much.
4. Practice an efficient and consistent thought process. If you have
a good thought process (such as one that includes the ques­
tion "If I make this move, what are all the checks, captures,
and threats he can make in reply, and can I safely meet them
all?"), then you will find most, if not all, of the key problems in
the position.
There is one instance where a leftover threat is normal and must
not be overlooked. This occurs when an opponent's threat is met with

169
The Improving Chess Thinker

a zwischenzug - an in-between move - or counterattack, which tem ­

porarily ignores the threat. In that case one purposely allows the origi­
nal threat to continue, so remember it next move or it could cost the
game !
Let's summarize what should occur during analysis: A player cre­
ates a mental "analysis tree" of possible moves. Each time a move is
made, he can trim that tree to include only the moves that were actually
made. If the opponent makes a move that was not on the tree, he should
see how that move affects the position, and generate new branches. If
his opponent's move was already on the tree, he should verify previous
analysis of that move and add new analysis. In general, any new analy­
sis is most efficient when it focuses on the changes in the position made
by the previous move, including both the creation of new possibilities
and the elimination of old ones.
From a practical standpoint, using guidelines such as "What are all
the reasons my opponent made that move?", "What are all the things
he can do to me now and could not do before (and can no longer do)?",
and "What are his new threats, if any?" are very helpful and powerful
shortcuts for promoting thought efficiency. However, these shortcuts
are much more effective once a player is strong enough to avoid most
leftover problems. Therefore, play slow games and practice being thor­
ough and careful.

10.9 Don't Waste Time on Lines that Won't Happen!

The main principle t o guide your analysis time:
Spend most of your time analyzing forcing moves in analytical po­
sitions. Don 't spend as much time analyzing non-forcing moves; use
your judgment in non-analytical positions.
However, from listening to hundreds of amateurs analyze, I have
learned that they spend inordinate time analyzing lines that could never
happen, primarily because:
1. The player does not understand or attempt to deduce which
moves are forced. Therefore he assumes moves that would
never happen.
2. The player wishfully thinks his opponent will not make the best
move and assumes a bad reply, resulting in the overestim ati o n
of his own move.
3. The player does not know when to stop analyzing a particular
line. He reaches a position where he should stop and evaluate,
170
 The Thinking Cap

but instead continues unnecessarily into lines that not only

may not happen, but are irrelevant to the process of choosing
the move. One common example is when a player proves that a
move is not best (at any point in a line), but nevertheless con­
tinues because he does not realize that analyzing past such a
move is irrelevant, since the analysis almost always will not be
played.
There is no instant cure for wasting time analyzing moves that can't
- or shouldn't - happen, and even strong players do this occasionally.
But it is important to continually improve your ability to identify what is
relevant and/or critical so you don't waste time on unnecessary lines.
What can be done? From a practical side, playing lots of slow games
against stronger players and then analyzing with them afterwards is a
great step. Stronger players are better at weeding out what is irrelevant
and, when they analyze with you, often show or state why. Even watch­
ing strong players analyzing their games can be quite an eye opener. If
you cannot play stronger players, at least play reasonable competition.
As you and your opponents improve, the moves you face will, more and
more, feed back to you whether you were anticipating something rel­
evant or not and your skill at differentiating relevant lines will slowly
improve.
It is very helpful to first look for the forcing moves - checks, cap­
tures, and threats for both sides. An example of how to find one type of
forcing reply would be to ask, If I move the piece there and he simply
attacks it with a piece of lesser value, would I have anywhere safe to
mo ve ? Looking for forcing moves is especially helpful if the reason you
waste time analyzing lines that could never happen is because you have
no idea which moves to analyze!
If in response to a candidate move there are no forcing replies by
your opponent, then likely you can cease analysis of that line and evalu­
ate the resulting position. If your opponent does have forcing moves,
it can be wasteful to assume he won't play one and instead analyze a
non-fo rcing move. If you are consistently surprised by opponents' tac­
tics then systematically considering your opponent's forcing moves as
pos sible replies will likely be very helpful.
Another tip is to ask, "Would my opponent really do this?" or "Why
Would my opponent make this move?" If the answer is he would not make
this move - because it is helpful to your cause or irrelevant to his - then
it is likely he will not. Analyzing such moves just to see how good they are
Wastes valuable time. Remember that your opponent is trying tofind his
best mo ve, so assume he will make his move with that in mind.

171
The Improving Chess Thinker

My students say they often consider opponent moves which are pos­
sible but, with a little analysis, can clearly be eliminated. If that is so
then your opponent will likely come to the same conclusion and, whil �
he may also consider such lines, likely will not play them. You do need
to analyze any plausible reply at least lightly to see if it really might be
playable for the opponent. However, once you determine that a move is
not playable, further analysis is not necessary.
You can also waste time by analyzing lines that are possible but not
relevant. Suppose you determine the best move, but instead of maki ng
it, continue to analyze to determine future possibilities. For example,
suppose your opponent puts you in check and you have only one legal
response, but afterwards the game gets complicated. The reasonable
choices are to make the move immediately or to resign. Looking ahead
to see what will happen is not only unnecessary, but helpful to your
opponent, because in that circumstance he can think about his next
move with perfect efficiency on your time. While the "only one move
to get out of check" is the most extreme case, the same problem can
happen in less extreme cases: players often find the best move but don't
make it out of curiosity. Even if a player delays playing the best move to
triple-check whether it is really best, the time spent can get diminished
returns, especially if the move is non-critical.
Even if you wanted to be as efficient as Dr. Euwe (see Appendix
A), you would have to develop his superior board vision, analysis, and
evaluation skills. Improving your skills - especially your thinking pro­
cess - will also improve your efficiency so that more of your analysis
will be relevant. Do too little analysis and you risk Hope Chess - your
opponent will continually surprise you with moves. Do too much and
you confuse yourself and get into time trouble. The goal is an efficient
and happy medium.
Go "wide" in your analysis before you go "deep". There is no sense
looking 10 ply (half-moves) down a line where the third or fourth reply
is improbable. Instead consider:
1) other candidate moves for yourself (first ply) or,
2) other reasonable replies for your opponent (second ply).
The deeper you analyze, the less likely the line you are analyzing will
occur. For example, suppose a line is not forcing and the chance that a
specific move will be played at each ply is 40%. Then when searching
three moves ahead (6 ply), the chances that this line will occur is only
0.4 to the sixth power, or about 1 in 250. It is more efficient to consider
other moves at the first or second ply.
172
 The Thinking Cap

The bottom line: Follow #322 in GM Alburt and Lawrence's Chess

R u les of Thumb: "Think along the top of the variations. Before you
go into a jungle of deep variations, search for different opportunities
for yo urself, and for your opponent, on the very first moves." Good
advice !
I n The Seven Deadly Chess Sins, G M Rowson advises that you always
ask yourself, "What great things does this move do for my position?" If
you can't find "great things", then it may be efficient to eliminate that
move. Similarly, if you are considering an opponent's reply that does
not do "great things for his position," then further analysis of that move
is likely unnecessary. Analyzing lines that are likely to occur results in a
better chance of finding the best move efficiently - and that's the goal.

10. 10 Thinking on Your Opponent's Time

Using your opponent's thinking time efficiently is an important part of
managing a chess game. Your opponent's time is the best time to drink,
eat, go to the restroom, and stretch your legs. It is beneficial to eat and
drink during long games, and stretching your legs once in a while keeps
your blood circulating and your body limber.
Although these activities take a certain percentage of your oppo­
nent's thinking time, you should keep these extraneous activities down
to a moderate amount. For example, suppose you get up and stroll
around after every move, checking your friends' games and so on, and
return to your table only after noticing your opponent has moved. Even
if you could be perfectly efficient and return each time just as his moves
(in practice virtually impossible), then you still waste 10 seconds each
move getting back to your board. In a 40 move game, this amounts to
400 seconds, or almost 7 minutes ! That wasted time can adds up, in
addition to the lost opportunities to use the opponent's thinking time
more wisely. Therefore, wandering too much is wasteful. But if you are
at the board, what should you be doing?
The classic suggestion is to think specifics and tactics during your
m ove and generalities and strategy during your opponent's move
- and this is good advice. For example, on your move you might ana­
lyze specific lines such as "If I make move 'A', will he reply move 'B',
and what is my evaluation of the position if he does?", while on his
turn consider ideas such as "With this pawn structure, should I push
my queenside pawns and if so, which ones first?" or "What would be
the ideal square for this knight, and should it be the next piece that I

173
The Improving Chess Thinker

need to reposition?" If you learn to do this efficiently, you will be mak­

ing good use of your opponent's time, and that should result in better
performances.
Another important part of general strategy that can be performed
while your opponent's clock is ticking is time management. A great tip
is:
While your opponent is thinking, try to determine whether you are
playing too fast or too slow.
Base your consideration on such factors as :
• the complexity of the position,
• how many moves the game is likely to last,
• the time control,
• the current move number, and
• how much time remains on your clock.
A periodic adjustment is very helpful early in the time control, and ad­
justments should be made frequently as time becomes more of a factor.
Although considering general strategy is the most common activ­
ity that occurs on your opponent's time, it is not the only one. Decid­
ing which activity is the most efficient depends on factors such as how
much time is left on the clock and how efficiently you can predict your
opponent's next move.
As was discussed in Section 10.9, suppose your opponent only has
one legal move, but for some reason he (erroneously) is waiting to make
it. You can predict your opponent's move with perfect efficiency so, in­
stead of thinking in generalities, assume that move will be made and
begin to analyze your reply, just as if it were your turn.
With regards to your opponent's likely move, there is an enorm ous
grey area between "forced" and "wide open," when his move is com­
pletely discretionary. In this grey area, the more you can antic ip ate
your opponent's move, the more you can think concretely on his time.
For example, suppose your opponent has two crucial replies and is
taking some time to decide between them. You should also figure out
that he has two choices, assume the one you believe more dangerous
and analyze what you would do if he made that move. If it turns out he
makes the other move, you have lost very little - especially if you were
correct that the one you were analyzing was better!
An example:

174
 The Thinking Cap

Yehl-Heisman, Liberty Bell 1968

Black to move after 16.i.b5+

In this complex position Black has four legal replies: 16 . . .W'xb5,

16 ... lll c 6, 16 ... id?, and 16 ... @xt7. However, both sides quickly saw that
16 ... W'xb5?? fails to 17.lll d 6+ winning the queen and 16 . . . @xt7?? allows
16 ... W'f6#. So the real issue was whether Black should play 16 . . . lll c 6 or
16 ... i.d?, and I was taking a while to decide what to do.
White should take advantage of this by thinking along with Black.
Since there are only two reasonable replies, White can efficiently assume
the one he will calculate as best and start to work on his planned reply.
Moreover, White may have already calculated that 16 ... lll c6 was bad
due to 17.hc6+ bxc6 18.lll d 6+ l!ld8 (18 ... l!ld7 19.W'xh7+ wins) 19.ia5+.
If so, White could concentrate all his efforts efficiently on 16 . . . . i.d?. I
also calculated that 16 ... id? was best but in my inexperience decided
to calculate whether it was winning for me or not (it was). I spent so
long doing calculating 16 ... i.d? that I forgot why 16 . . . ©xf7?? was bad,
changed my mind, played it, and was instantly mated! This example
shows that the advice to play the best move once you determine it is
ve ry important!
A common practice among stronger players is to assume your prin­
cipal variation (PV) is going to be played. Since knowing the PV usually
involves finding your opponent's best move, you can assume that move
- at least initially - and see if your intended reply still holds. Quite
ofte n when using your PV it turns out that although your opponent's
next move may be optimum, your originally intended reply to it may
? nly be sufficient. For example, you might think "I will play 'X', which
is clearly the best move, and if he plays move 'Y', then I can at least re­
ply with move 'Z'." But that usually only proves that your intended next
move Z is sufficient to meet his expected move Y, not that Z is necessarily

175
The Improving Chess Thinker

your best reply to Y. Therefore, while.finding such a sufficient Z is usu­

ally necessary to play X, it does not mean that Z is what you should
play if Y is actually made. Therefore, while your opponent's clock is
running, you can think, "Alright, I planned Z if he plays Y, but suppo se
he does plays Y - can I verify that Z is sufficient and possibly find a
move better than Z?"
By cleverly using your opponent's time to determine if move Z wo uld
truly be best then, if your opponent does play Y as expected, you won 't
have to rely on Z as simply a basis for new analysis, but will have much
more information to start your turn.
While you are thinking during your opponent's turn, you may - un­
happily - refute your intended move Z, and need to find another move
that is at least playable. Hopefully that scenario won't happen too of­
ten, but when it does - especially when you are short on time and such
analysis errors are occurring more frequently - starting your search for
a sufficient move on your opponent's time rather than waiting and then
finding out Z is inadequate on your time may be the difference between
winning or losing! It has happened to me more than once.
As it turns out the next Yehl-Heisman game is an example!

Yehl-Heisman, Keystone State 1968

White to move after 13 . . . .ixg4

White has sacrificed his kingside pawns intending 14.hh7 + ci>xh 7 1 5 .

lll g S+ ii-moves 16.\Wxg4 with attack. However, ifWhite had been th in k­
ing on the expected 13 ... .bg4, he might have realized that 14 . .ixh ?+ is
refuted by 14 . . . iih8 ! when Black wins the knight on f3. Unfortunately
for White, this realization came too late and he took quite some tim e
before admitting his mistake and playing 14.ie2.

176
 The Thinking Cap

After that I could have won prosaically with 14 . . 1/NeB ! , but instead
.

was bent on revenge and played the complicated 14 . . . .ixf2 + ! ? The

game concluded in spectacular fashion: 15.©xt"2 llie5 16 .ig5! Y!Yb6+
•

t7 .ie3 Y!Yf6! 1s.gh1 .ht'3 19 . .ht'3 llig4+! 2 0 .©el llixe3 21.llixd5

•

€'J 7xd5 22 . .hd5+ ©h8 23.Y!Yd2 llixd5 24.Y!Yxd5 Y!Yt"2+ 25.©dl

!!!ad8, White resigns.
Sometimes when analyzing your opponent's possible reply Y, you
realize Y is not best (your PV was not correct) and he would be making
a mistake to play it. Then you should no longer worry about Z, but can

spend your energies seeing which replies are more likely than Y and
what you can do about them.
Sometimes time trouble prevents using sufficient time to find a PV.
Then it is especially important to make optimum use of your opponent's
time to calculate concrete variations in case they are played. Sometimes
your opponent, if he has quite a bit more time, may move quickly to
stop you from doing so. Ironically, this is usually a big mistake on his
part, because he is negating his time advantage. During your opponent's
thinking time, you can't think as efficiently about what your opponent is
going to do as your actual opponent can, so it would be a better strategy
for him to play slowly - and for you to use that time as best as pos­
sible.
The major exception to playing slowly when your opponent is in
time trouble is when a player is clearly winning but is very short on
time. Then it is correct for the losing player to play quickly, even if he
has adequate time.
I read that world-class GM Michael Adams had a habit of taking
a stroll after almost every move. However, he became an even better
player once he mastered the art of using his opponent's thinking time
efficiently. Therefore, next time you play a slow game, see if you can
implement some of these strategies to improve your results, too. Good
luck!

177
Chapter 1 1

The Basics of Time Management

" ext to winning, one of the most important goals of a chess g ame
N should be to use as much of your time as reasonable to maxi­
mize your thinking resource."
One of the controlling features of thought process is time manage­
ment. Because all serious chess games are played with a clock, it is an
important skill to manage your time wisely. Time management is fur­
ther divided into two main skills:
• Micro Time Management - The ability to allocate your time
for each move, giving more time to moves which need it (i.e.,
critical moves where investing extra time is helpful), and less
time to moves which don't, and
• Macro Time Management - The ability to pace yourself during
a game to use almost all your time without getting into unnec­
essary time pressure.

1 1 . 1 Micro Time Management

Micro Time Management depends upon:
• The "normal" amount of time one should spend on a move (I
call this Trigger 2),
• The possibility of being 100% sure you have found a "best"
move before reaching Trigger 2 (I call this Trigger 1),
• The awareness of how much time you have spent thinking about
the move.
Breaking this down further, the normal amount of time one sh o uld
spend on a move (Trigger 2) is based upon:
• the time control (the slower the time limit � more time can be
spent per move),
• your time remaining (more time remaining � more tim e per
move),

178
 The Basics of Time Management

• how many moves remain in the time control (more moves to

play � less time per move), and
• Criticality analysis: the ability to identify critical moves, i.e.,
how important playing a best or near best move can affect
the outcome of the game (higher criticality � more time per
move).

Finally, examples of what is meant by a critical move:

• The important strategic decisions: trading queens, where to
place the king, whether to open the position, where and when
to attack, trading into a king and pawn endgame, whether the
creation of an imbalance (such as a sacrifice) is worthwhile,
etc.
• Most moves in complicated positions,
• Moves that make key changes in the pawn structure, especially
in the center,
• When the best move may be clearly better than the second best
move (excluding trivial recaptures), and
• The first move each game where you are no longer 100% certain
you are in your opening "book" (as mentioned by GMs Alburt
and Soltis)

The time control is a major factor. One could have the same deci­
sion on the same move number, but the proper amount of time could
be vastly different based on the time control. For example, the same
position on move 28 at 40 moves in 2 hours (40/2) would suggest much
more time than it would on move 28 of a G/30 game.
A key point is that reaching Trigger 1 or Trigger 2 should be the
only two reasons you should halt your thought process and make a
move! Either you have found a move which cannot be bettered or you
have spent a reasonable amount of time to find a move. Once that rea­
sonable amount of time is reached, further search will yield diminish­
ing retu rns, and thus you should play the best move found up to that
ti me (The King of the Hill - see Section 10.2). There is no other proper
re aso n to ever stop your thought process and make a move. If Trigger
2 is hit and the King of the Hill is completely unacceptable, then likely
Trigger 2 was too short and it has to be extended.

179
The Improving Chess Thinker

What is the best way to know that one has reached Trigger 1? I call
this the point where you are willing to argue with a room full of Gran d­
masters that they can 't.find a better move. A couple of examples:

Example 1: A trivial recapture. Black has 1 hour and 57 minutes

left on his clock to complete the game after 1.e4 eS 2.<tl:f3 d6 3.d4 f6
4.dxeS dxeS 5.§'xdS+ :

Black to move after 5.§'xdS+

Arguing with a roomful of Grandmasters (1)

Here Black has plenty of time on his clock and thus Trigger 2 might
be 30 seconds or more, but to take that much time is wasteful. Black
should reach Trigger 1 quickly and play the trivial recapture S ©xd8•••

because the only alternative, 5 ©f7, is not worth further thought. I

. . .

don't think anyone in that roomful of Grandmasters would argue that

5 lt>f7 is a better move.
...

This is also a great example of a non-critical move. While it is critical

that Black play 5 lt>xd8 instead of the alternative, it is trivially critical
• • •

since the alternative doesn't make sense. Therefore, any time it is easy
for players of all levels to see that there is only one reasonable move ,
by definition that is a non-critical move. Taking more time than neces­
sary to play such a move (e.g. to see what White will do after 5 lt>xd8)
••.

is counterproductive. Forced moves, once properly determined, shou ld

always be played as quickly as possible to preserve time for moves wh ere
there are choices.

Example 2: White has 1 minute remaining on his clock to com plete

the game:

180
 The Basics of Time Management

White to move
Arguing with a roomful of Grandmasters (2)

In this position White has only a minute remaining to complete the

game and spots the basic "stair-step" checkmate 1.�b3+ ©cl 2.�c2#
At this point it is reasonable to just play the forced mate using Trigger 1.
However, suppose after the game someone steps forward and says "Ah!
But if you had just taken more time you would have found the better
move 1 . .icl#, since mate in one is better than mate in two."

Would it be reasonable to go in front of a roomful of grandmasters

and argue that 1.�3+ cannot be bettered? Sure it would! There are
two reasons for this :
• You can't do better than win a game, and a forced mate in two
will not result in a worse result than mate in one, and
• Suppose you search for a mate in one and then, before you find
it, your clock falls? That would be far worse!

So you have a strong argument that there can't be a better move than
l .�3+. A win is a win is a win. Is it possible some Grandmasters may
argue against you? Of course! But just as many might argue in your
favor - the point is that you are willing to argue with the room full of
Grandmasters because you have a good case.
With this arsenal in place, it is now easy to define what is meant by
playing too fast and too slow!

• One plays too fast if making a move in less time than Trigger 2
without reaching Trigger 1, and
• One plays too slow if taking more time than Trigger 2.

181
The Improving Chess Thinker

For example, suppose you are playing a one-minute game, no

increment or time delay, and have only five seconds left on your clock to
make the last several moves. Then Trigger 2 becomes extremely sm all '
so it is virtually impossible to play too fast in that situation.
Lasker's Rule: When you see a good move, look for a better one
(and, if you see a better one, look for an even better one because you are
trying to find the best one you can).
But how long should you look? How do you know when you are done?
Searching for better moves for a long period is often impractical.
In many situations your clock will fall long before you can prove you
have found the very best move. Instead, you should look for a better
move until you have hit either Trigger 1 (the move about which you are
willing to argue with Grandmasters), or "t", the Trigger 2 time that is
the reasonable amount to spend on that move. In time trouble Trigger
2 is always short, but the principle is the same, since in that situation
thinking for only a short time is reasonable and necessary.
Average time per move
To calculate the average time to spend on a move for a game, do the
following:
• Assume a conservative 40 moves to the game (the median
number of moves in a game is less, but some games last much
longer)
• Divide 40 into the number of minutes for the game.
• Add any time delay or increment
During the game, adjust accordingly if the game looks like it will
become much shorter or much longer than 40 moves.
Example: Suppose you are playing a 90 minute game (G/90) with a
five second time delay.
90/40 =2Y. which is 2 minutes and 15 seconds. Add the five sec­
ond time delay, and you have 2 minutes and 20 seconds for each move
in a 40 move game. Therefore, an optimum strategy is to take clearly
less than 2 minutes 20 seconds for a non-critical moves and take more,
sometimes much more, for critical moves.
When one examines the issue of the optimum time to analyze critical
lines, one important question is "How long should I analyze before I get
diminishing returns and thus waste time?" Besides Trigger 2, there is a
very helpful answer for this question with regards to sacrifices (whose
consideration often requires a large amount of time! ) :

182
 The Basics of Time Management

Continue to look at a sacrificial line so long as the reward remains

g rea ter than the possible risk, and adequate time is available.
For example, suppose you examine a line in which you sacrifice a
bishop. If all you see is that you might later win a pawn, it is not worth
tim e to see if that can be forced, since the reward is less than the risk.
However, if you see you can possibly win a rook, then further analysis
is worthwhile. Whenever the reward is checkmate, then any sequence
of sacrifices is worth examining so long as there remains a reasonable
ch ance for checkmate.
Finally, if the position remains unclear and forcing sequences can­
not be resolved to quiescence (at least not in a reasonable amount of
time), then your judgment becomes involved, but the overall time con­
sequences remain the same. In this case you can probably save clock
time, since - to paraphrase GM Rowson in The Seven Deadly Chess
Sins although your analysis gets better with additional thinking time,
-

your judgment probably doesn't.

Here is another tip to help you decide in unclear situations:
Ifyou are playing chess to improve, then always choose an unclear
line over a clearly equal line.
There are several reasons to do this. One is that you are trying to
improve your evaluation skills, and if you never play unclear lines they
will always remain unclear to you. When you play unclear lines, you
will start to develop a feel as to which "unclear" positions are good and
which are not. Secondly, most non-titled players are much better on
offense then they are on defense, so if you make an unclear sacrifice to
get an attack (or some similar compensation), the defender often goes
wrong and your practical results will improve.
Many players play too quickly, and one reason for this is that they
consider few sacrifices because they think they are "losing" material.
Those players tend to do well on tactical problems, but not so well on
th ose same tactics when they occur in games ! Section 2 .5 has a discus­
sion on this problem, which is usually due to quiescent errors.

1 1.2 Macro Time Management

M acro time management is the ability to use almost all your time every
ti me a game goes "full course."
For example, if you are at the World Open and the first time control
is 40/ 2, then you want to finish your first forty moves with at most a few
mi nutes left on your clock. If you reach move 40 and have 37 minutes

183
The Improving Chess Thinker

left, it is very likely you could have benefitted from some extra ti me
thinking on some of your moves. On the other hand, if you reach move
21 and have only 9 seconds left on your clock (as one of my stude nts
once did ! ) then, even with a five second time delay, you are in severe
time trouble. You likely have taken much too long on some non-criti­
cal moves, when trusting your judgment sooner would have been much
more effective.
How important is macro time management? Here is one way to look
at this question: I estimate that if there is a typical middlegame posi­
tion, and one side has only 5 minutes remaining and the other has 15
minutes (without a time delay or increment), that advantage is worth
about 200 rating points. For example, suppose both players have the
same rating and the position is even but the time is 15-5. Then the player
with 15 minutes left is roughly a 3-1 (75%) favorite.
It is possible to have excellent time management and have plenty of
time left on your clock at the end if one of the players makes a gigantic
mistake and resigns, thus greatly shortening the expected length of the
game.
Interestingly, if one practices good micro time management then
that usually results in good macro time management, but the reverse
is not necessarily true. It is possible to pace yourself to use almost all of
your time while misallocating more time to non-critical moves and less
time to critical moves.
I once had a student who practiced exactly that reversal. He was too
slow and thoughtful on close, but non-critical decisions: Where does
the bishop go? Which rook to move to the middle first? Then, when the
game finally got complicated and he had the extremely critical decision
whether to capture a bishop with a pawn or a rook with a knight, he took
the rook almost immediately, explaining, "That kind of calculation is
too difficult for me, so I just used the principle Take the highest valued
piece." This "principle" does not exist, and taking the rook lost material,
while taking the bishop would have won material. Thus, down material,
he returned to slowly, but surely, losing his lost game, eventually us­
ing almost all his time. His macro time management was good, but his
micro time management was abysmal.
However, if you allocate time correctly to each move, then your mac­
ro time management is likely good. It is also possible to practice g oo d
micro time management and run into reasonable time pressure: Sup­
pose the game become critical early. Usually at some point the sm oke
clears and one side or the other emerges ahead, or the game beco me s
less complex. But sometimes, when two good players play, the gam e

184
 The Basics of Time Management

stays critical and complex all the way through, as attack and defense
both play equally well. Then it is entirely possible that the players will
take more than the average amount of time on many early moves, and
then have much less time than desirable to complete the game. In this
case one or both players will get into time trouble, and this occurs even
with strong, well-intentioned players.
Here are some tips to help your macro time management:
• Periodically check your time when your opponent is thinking.
Ask, "Am I playing too fast or am I playing too slow?" and ad­
just accordingly.
• Write down your time remaining (in minutes) after each move
to make yourself more aware of your time situation.
• Before the game, write "milestones" on your scoresheet for how
much time you think you should have remaining at specific
points in the game. This works best for time control that aren't
"sudden death", such as 40 moves in two hours.
Botvinnik's Rule: World Champion Mikhail Botvinnik suggested
that for "normal" openings one should not take more than 2 0% of the
first time control to make the first 15 moves. At a 40/2 time control,
20% x 120 minutes 24 minutes, so one should have about 1 hour and
=

36 minutes left on move 15.

Knowing the average time per move is not only helpful in micro time
management, but also in macro time management. For example, sup­
pose you are playing 40/2 . Botvinnik's rule would have you reach move
15 with 1 hour and 36 minutes left on your clock, leaving 96/25, or a
little less than four minutes per move for the rest of the game, which
should contain most of the critical moves. But if you have been playing
the non-critical moves in the opening way too slowly, you will exceed
Botvinnik's rule limit by quite a bit. Let's assume you reach move 15
with only 65 minutes left. You have 25 moves remaining, so your aver­
age has decreased from the original 3 minutes per move to only a little
over 2.5 minutes. If you don't adjust and speed up a little now, you will
likely find yourself in time trouble.
I often get asked by slow players, "But if I play my non-critical
moves more quickly, won't I either lose more games in fewer moves
and/or have a worse position when I get to critical moves?" This is a
good question, but it overlooks the fact that for every game you would
lose more quickly, there should be many games where you win or draw
later on instead of losing due to time trouble. No one is suggesting that

185
The Improving Chess Thinker

you race through the non-critical moves. But taking 90 seconds instead
of 5 minutes for moves that require judgment and not analysis should
not have a strong detrimental affect on your play. Again, paraphrasi ng
GM Rowson, your judgment does not necessarily get better as you take
longer. So save the time for analytical moves, where the extra time is
much more helpful !
A few years ago I was visiting the National Chess Congress, and all
the sections were playing in one large room. The round started at u
AM and was 40/2, so first time control would be reached at 3 PM. At
one point I stopped to rest next to a pillar in the middle of the room.
To my left were all the high-rated sections; to my right were all the Iow­
rated sections. I looked at my watch and it was 1 PM - exactly halfway
through the first time control. To my left almost every game was still in
progress - the only seats unoccupied seemed to be players who were
walking around and watching other games while their opponents were
thinking. To my right, games were still being played on only about 25%
of the tables! This was not coincidence: the weaker players, on the aver­
age, had been playing much too fast.
The moral: following good macro time management is very impor­
tant. You don't want to beat yourself; you want to make your opponent
beat you. Players who play much too fast or much too slow (these total
about 80% of my intermediate students ! ) at some point are making
unnecessarily quick decisions, and that has a very strong detrimental
affect on their play. Correcting your macro time management will have
a much better affect on your playing strength than learning more moves
in the Caro-Kann.

186
Chapter 12

What the Resea rcher Lea rned

have been administering the d e Groot exercise fo r over forty years, a

I total of hundreds of exercises. What have I learned?
I often chose de Groot A as the exercise position and afterwards read
Euwe's protocol (see Appendix A) to the student as a pristine example
of how to process the position. Result: I can almost recite Dr. Euwe's
15-minute protocol by heart!
My principal findings:
1. Players rated below 1600 USCF /FIDE almost all practice "Hope
Chess." That is, they make moves without considering whether they
can successfully deal with all the threats their opponents might make

next move. To put it another way, they often make a move without at
all considering the consequences of what might happen when they do.
They even make analytical moves without a Principal Variation (PV)
- not looking to see if their move allows an opponent a forcing (check,
capture, or threat) reply that might win the game immediately. This was
not a surprise; it is very difficult to have a rating that low if you have
reasonable chess knowledge and play "Real Chess."
One student e-mailed me:
I identified myself as a "Hope Chess" player when previously I had
fooled myself into thinking I was largely past that phase. However,
as with the weakest link, if I play hope chess on any move, the game
is hope chess. Loosely analogous to 12-step programs, thefirst step to
recovery is to admit that I am a hope chess player.
Yesterday's walk through my game where I had ignored many
relatively simple responses (and only won because my opponent was
equally lax) was a terrific eye-opener, as was the de Groot exercise.
2. For analytical positions I advise students to first consider their
most forcing moves as the first step in identifying candidate moves.
However, many fail to realize how important this advice is, and don't
systematically attempt to list these moves. Weaker players' analysis is
sporadic and non-systematic, rarely covering all the pertinent lines,
su ch as considering all their checks, captures, and threats, or possible
op ponent recaptures after a capture.

187
The Improving Chess Thinker

After a capture that allows multiple recaptures students often just

assume a recapture without asking "What are all the recaptures avail­
able, and which one would my opponent choose?"
Rarely did a weaker player consider all his captures, and rarer s till
did one list all the recaptures, e.g. the three recaptures (l.. .liJxd5, 1 . . .
hd5, l...exd5) after 1.liJxd5 in de Groot A:

de Groot A: After 1.liJxd5

The mistaken assumption that the opponent will play the only re­
capture considered often leads to conclusions that prove very little. In
mathematics there is a theorem that states, Start with a false assump­
tion and you can prove anything. A corollary would be, Start with a
false assumption and you are very likely to reach a false conclusion; if
you reach a true conclusion, you are just lucky! Making bad assump­
tions in chess analysis leads to similar failures. Strong players don't
make this kind of mistake, especially in clear positions where, with a
little effort, they can work out all the lines. Implementing a more sys­
tematic (but not rigid) approach would definitely benefit many weaker
players.

3. While weak players analyze much worse than strong players, it is

the evaluation skills that help separate the 1900-2 200 players from the
international players. Grandmasters see immediately which positions
are good or bad, and which are worth further consideration. 1900-2200
players analyze fairly well, but arrive at much less accurate conclusions !
For example, it was common that players in this range (including my­
self! ) rejected 1.hd5 in de Groot A:

188
 What the Researcher Learned

de Groot A: After 1.hd5

. . . because Black can play l. .. exd5 which, unlike the other recap­
tures, does not immediately lose material. However to see that l. .. exd5
is very good for White is much more difficult. Almost all the grandmas­
ters evaluated this correctly:

de Groot A: After 1.hd5 exd5

White is better

Players rated 1900-2 200 usually used similar deductive logic as

Grandmasters to decide what is forced and what is not, and what is
likely to happen. Their analysis results are not as consistently good, of
course, but many times their logic is usually good and often they see the
same lines as do the Grandmasters. But when they get to the ends of
the same lines, their evaluations are often faulty. For example, when a
G ran dmaster might conclude, "That is good for me - if I don't find any­
thing better I will certainly be happy with that!" the 1900-2200 player

189
The Improving Chess Thinker

might look at the same position and say, "I am not winning any mate ri al
and I don't see anything special, so I will probably play another move."
Players rated 24 00+ are clearly better, but not extremely better
in analysis than the players in the 190 0-23 0 0 range. (Don't get m�
wrong; top players are more accurate i!!ld less error-prone! ) However
there is a relatively larger difference in evaluation abilities, especiall ;
evaluation of "even material" positions. Therefore Grandmasters and
International Masters are much more likely to be able to choose the
best continuation among several alternatives when the most differen­
tiable evaluation criteria of king safety and material are not big fac­
tors. Moreover, weak players almost always value pawn structure above
initiative and the entire army's activity, while strong players don't care
nearly as much about the pawn structure if they can maintain a clear
initiative.
Note: Although players in the 2200+ range evaluate much better,
there is still considerable difference - and thus much room for improve­
ment - in the analysis skills of lower players, so even "Real Chess" play­
ers in the 1600-2 200 range can greatly improve their playing strength
by improving their analysis skills as well. Some of the analysis mistakes
in this rating range were quite striking - and that was only on one move.
The cumulative effect of making this type of mistake on multiple critical
moves can add up to several hundred rating points.

4. Weaker players often evaluate "potential" tactical moves primarily

on material, but non-tactical moves on positional grounds. This seem­
ingly minor nuance often turns out to be a big mistake. For example,
they will reject a capture because "it does not win material", but instead
play a quiet move that often fails to force play, allowing the opponent
to seize the initiative. Apparently they feel that when a capture fails to
win material it does not succeed in its primary purpose, and for that
reason alone assign it a lower evaluation than a non-capture, even if the
resultant material is even in both cases. This is faulty logic. It is entirely
possible that the capture leads to a better position than the non-cap­
ture. A part of chess skill is seeing how captures which don't change th e
material balance might prove favorable by trading your bad pieces for
his good ones, eliminating key enemy defenders, improving the relative
pawn structure, keeping the initiative, etc.
The percentage of players who make this "capture does not win ma­
terial so it must be inadequate" mistake is very high. My conclusio n is
that at least some weaker players must confuse finding chess moves in
games with doing chess problems. These players are always looking fo r

190
 What the Researcher Learned

forcing moves which win (as always happens in play and win problems).
Many can't "switch gears" and realize that in game situations a forcing
capture that does not win anything is possibly the best move, or at least
a ve ry good move worthy of their consideration.

5. Some players spend a ton of time looking at lines that are not
forced and almost never could happen. They don't deduce "Suppose I
do this - would he really do that? And if so, what would I likely do?"
Analyzing de Groot A, one of my college chess teammates looked an
astounding 40 ply ( ! ! ) or so ahead with perfect visualization, saying,
"Suppose I do this and then he does that, then I will probably do this
and suppose he does that and then I do this . . . " However, the ten minutes
he took to do so were a complete waste of time because not only was
the initial move he was contemplating not necessarily best, but none
of the subsequent moves were either. He made no attempt to show that
the moves under consideration were best or forced, or why he or his
opponent would play them. I would estimate the chances of that entire
line occurring as less than one in a trillion - completely worthless! It is
much better to spend time analyzing moves and evaluating lines that
occur early/shallow in the search and that might take place, rather
than spend time analyzing moves deep in the search that a lmost never
could happen - and even if such deep, non-forced lines did occur, you
could always analyze them during later moves.

6. Most students love the de Groot exercise and consider it most

revealing. A small minority find it hard to take for the exact same rea­
son: because it so pointedly shows what they are doing wrong. As noted
earlier, I once gave the de Groot exercise to a Class A player who made
a clear thinking process error that was easy for me spot due to its clear
difference from what the Grandmasters did. This serious error in his
thought process may have been enough to prevent him from becoming
an Expert. However, after he was finished and I explained what he did
and how that differed from Grandmaster protocols, he became indig­
nant and vociferously defended his process. Even when the exercise
works and is insightful, it is sometimes unfortunately incite-Jul!

7. Weaker players don't statically evaluate unfamiliar positions. I

define evaluation as looking at a position and determining who is bet­
ter, by how much, and why, and static evaluation as evaluation done
before analysis. Many weaker players mention that one side has an
isolated pawn or a weak square - one might call this an assessment

191
The Improving Chess Thinker

- but don't conclude with who is better, by how much, and why. Play­
ers rated below 1800 rarely include anything similar to the following:
"The material is even, the kings are about equally safe. White has a
better pawn structure, but Black seems to have more total piece activity.
Since it is Black's move, I think he can take advantage of that activity, so
I like Black much better."
When playing a real game, the knowledge of prior play removes the
necessity of making an evaluation before each move. However, when
starting a de Groot exercise, it is helpful to begin with an evaluation.
Instead, weaker players often start by either making a general assess­
ment with no conclusion or, worse, immediately searching for candi­
date moves. Many do not even count the material. That does not make
much sense because, without an evaluation, how do you know what you
might be looking for?
As a trivial example of how an evaluation is helpful, suppose you
find a forced draw - would you take it? If you think that otherwise you
are losing, you would probably be very happy with the forced draw.
But if you thought you were winning, why would you settle for a move
that forces a draw? So knowing which side you think is better provides
meaningful goals for your analysis.

8. Another thing weaker player should do, but don't always, is assess
the threats generated by the opponent's previous move. Since in a de
Groot exercise you are not given the previous move, then one needs to
look at all threats (in a real game you can often shortcut this process
by primarily considering the new threats identified by the move played
- see Section 10.8). The way to identify opponent threats is to ask, "Sup­
pose it was not my turn, but again my opponent's, then what would he
do?" Asking this also helps you find all the opponent's "killer" moves,
which are very strong threats that cannot be ignored (see Section 10.4 ).
Killer moves can eliminate your candidate move from consideration if
that candidate move does not meet the strong threat.

9. Does a player who plays too quickly slow down because he:
A) Acquires additional chess knowledge and has more to think
about, or because he
B) Knows he can play a lot better if he plays slower?
The de Groot exercises revealed that, for many, "A" is not the pri­
mary reason, although both usually apply. I rate the weight betwee n
these two as approximately 35%-65%, with reason "B" predominatin g.

192
 What the Researcher Learned

For example, suppose I teach someone "X" things they should think
about every move, but they still play so fast that they could not possibly
be thinking very much about those X. Then surely adding "Y" additional
factors to consider so they now have "X +Y" will not slow them down - it
might intimidate or confuse them so they are even less likely to think
about the X! It is clear that for players who do "X" that Reason B is
correct - they need to "buy" into the correct thinking process in order
to slow down. Teaching players who cannot do the basics, like "When
you see a good move, look for a better one" additional things to consider
during their moves gets severely diminishing, if not negative, returns.
A player's motivation to slow down often depends on how much fun
the extra or correct thinking is, and how much he wants to improve and
is willing to do the work. Once you are aware of what is involved, it is
not magic to begin practicing a good thought process. Doing a minimal
amount every move religiously is required for high level slow play.
Peer pressure is a strong motivating factor for players who play too
quickly. Over-the-board players seem to learn to slow down better than
Internet-only players because they go to strong tournaments and see all
the good players taking their time! This sometimes works wonders, as
most players don't want to be the first one done each round, and thus
learn to "imitate" the time management of those around them.

10. Even intermediate level tournament adults don't always follow

the advice, "If you see a good move, look for a better one - you are trying
to find the best one." Even in critical, analytical situations, often they
just calculate briefly to see if their intended move is reasonable and, if
so, immediately make that move. This is a big mistake. On most moves
that require analysis, the goal ofyour thought process is to prove that
you have found the best move, not to show that a move tha t attracts
you is reasonable! Proving that a move is reasonable is only an efficient
way to find a good move in non-analytical positions, and doing so in
analytical positions is a primary reason why some players play too fast.
In Thought and Choice in Chess de Groot identifies four phases of
th e thought process of stronger players and calls them 1) Orientation
to Possibilities, 2) Phase of Exploration, 3) Phase of Investigation, and
4) Striving for Proof (see Section 2.3). Weak players rarely go through
all - or sometimes any - of these phases. And intermediate and weaker
pl ayers almost never strive for proof, "proof' meaning they have sys­
tematically gone through the process of showing that the move they are
about to play is at least as good as any of the other candidates (that is
lea ds to a equal or better position, by force, than the others do).

193
The Improving Chess Thinker

11. AB one would expect, as one gets stronger, the average length
of time per move also increases, but this has to stop at the opti mu m
time limit. In other words, very weak players generally - but not always
- play much too fast, while almost all stronger players take roughly the
same amount of time, which is all the time allowed. However, how the
time spent is not always the same. Not all levels of players spend the
same percentage of their time doing analysis.

12. The words that players think to themselves often make a big dif­
ference ! In protocol E-2 the subject sacrificed a piece to win "a rook"
when all he was winning was the exchange. Weaker players often make
bad, aggressive moves instead of "retreating" when the retreat is simply
moving back to the only safe square. Many weaker players don't want
to trade queens because they don't want to "lose" their queen ! Another
misconception is that trades lead to drawish or uncomplicated games
when, in many cases, trades might lead to much less drawish positions.
Some of the most complicated positions are simple-looking endgames
where the right sequence is buried amidst a minefield of danger.

13. The weaker the player, the less they understand how to end their
thought process. Rather than going through de Groot's fourth stage
of Proof, which is indicative of a strong player's process (see Section
2.3) or because of Trigger 2 (see Section 11.1), the weaker players often
ended their analysis in seemingly random ways. While I could usually
tell when a strong player was nearing the finish (by his/her "closing in"
on the prooO, weaker players often surprised me by making their move
"out of the blue" for no apparent rhyme or reason. On occasion very
weak players would choose moves they had not previously mentioned !

14. Over the years, I have found that many players who do determine
whether their candidate moves are safe, which is Step 4 in Section 2.2 ,
do so much too late. For example, suppose a player goes through an en­
tire thought process and then, during a sanity check just before makin g
the move, first asks, "Is my move safe?" and then discovers it is not.
The good news is that he has caught himself in time and avoided a
blunder. The bad news is that he wasted quite a bit of time and is now
back to "square one" (actually Step 4).
Therefore, it is a major finding that one way to improve thou gh t
process is to teach players to first determine the safety of their m ove
before trying to compare safe (final) candidate moves to see which o ne
is best. Doing it in any other order is very inefficient. In other wo rd s,

194
 What the Researcher Learned

don't combine Step 4 and Step 5 as best possible, perform Step 4 in its
-

entirety before moving to Step 5.

15. I learned the differences between a player thinking quietly dur­

ing a tournament game and having to "perform" by thinking out loud.
A) Some players are naturally verbal and can perform almost as
well during a de Groot exercise. For others it is like pulling teeth
to try to convince them to say something. Take each protocol
in this book and calculate the number of lines per minute; the
average will vary greatly from quite low to quite high.
B) Some players could not help "showing off' for the exercise by
clearly taking longer (i.e., doing more analysis) on this exercise
than they do in actual play. Very few took less. Many of my stu­
dents who consistently play too fast used remarkable patience
when put under the microscope! Moreover, this phenomenon
is not limited to a de Groot exercise. During lessons, students
who normally play too fast often take considerably more time
to do a puzzle or find a move when it doesn't count than they do
when it counts! When that happens, I usually ask, "Should you
spend more time calculating a move in a lesson than you would
in a tournament game?" The implication is that they can take
their time, so doing it when it counts should be the norm.

What players learn from the exercise

In addition to the lessons I learned from administering de Groot exer­
cises, what should the average player learn?
After doing a de Groot exercise, students who listen to Dr. Euwe
analyze the same position always get an eye-opener! But, despite the
imposing depth of Dr. Euwe's thorough analysis (see Appendix A), it is
not terribly difficult to emulate his process. Thus everyone who hears
this process can, and should, strive to do something similar. The harder
part is not emulating the process, but picking up all the extra knowledge
and skills that allow one to analyze and evaluate well. It takes years of
good practice and judgment refinement to be able to evaluate a posi­
tion well, and more years to be able to do so with anywhere near the
sophistication of a Grandmaster, even if you have the capability. It does
not take nearly as long to learn how to analyze well, but even dramatic
i mprovement in that skill is almost always measured in years, and not
weeks or months, as so many players would wish.

195
The Improving Chess Thinker

The amount of work it takes to analyze well is much higher th a n

most players realize, or possibly even find fun. After students listen to
all the work Dr. Euwe did to find the best move, some of them wonder,
"Do I really want to do all that? ! " That is a reasonable reaction: If you
do not.find extensive analysis and delicate evaluation.fun in positions
that demand it, then you probably won't do it now or ever. However,
unless you change your preference, your chances of ever becoming a
very strong player are likely nonexistent.
Listening to Dr. Euwe also causes many average players to realize
that that the gap between him and top players is larger than previ­
ously imagined. Many players fool themselves into thinking that if they
studied more openings and endgames for a few years that they could or
should or would eventually move up 1000 rating points or so. I think
doing the de Groot exercise is an epiphany that shows that there are
more important things to do to get really proficient than just learning
some new moves in the Caro-Kann or rook-and-pawn endgames. Not
to say that opening and endgame study isn't important, but how many
players do you know that have played 10+ years, read 100+ books, can
quote chapter and verse on book knowledge, and yet are still rated 1500
or not much higher? Without testing them, I can tell that these players
have poor thinking processes, and will never get much better until they
correct these deficiencies.

How other Instructors can use the exercise

I highly recommend that all serious chess instructors consider helping
their students' thought processes via the de Groot exercise. Any tools
which are designed to diagnose and aid thought process and/or time
management have a powerful impact on improvement, as these two
non-traditional areas are often overlooked by many instructors.
To teach the exercise, an instructor needs the following:
• A series of positions. The best positions are those with fre­
quently occurring analytical or strategical decisions which are
not easy - but not impossible - for average players.
• The ability to hear the student via live or phone lessons. Doing
the exercise via email or typing on a server would be far less
efficient.
• The ability to record or quickly copy down the studen t's
thoughts.

196
 What the Researcher Learned

• Some samples (such as Euwe's in Appendix A) of a good thought

process.
I recommend administering the exercise no earlier than the third or
fourth lesson, except possibly with very strong students, when it may
be used sooner. The reason to do the exercise at that time is because
an instructor wants to spend the first few lessons trying to diagnose a
student's needs and weaknesses, make sure they have the fundamental
background which would make the positions chosen meaningful, and
make them comfortable with his teaching style. The de Groot exercise is
not the most interactive and is fairly intense, so performing it within the
first two lessons may be doing too much too fast. On the other hand, the
lessons from the de Groot exercise are so diagnostically helpful that wait­
ing until well past the third or fourth lesson is probably putting off a good
thing too long! Therefore, doing it about that timeframe seems best.
Before administering the de Groot exercise, be careful to give the
instructions as stated in Chapter 1. Failure to include even one of the
instructions can be quite detrimental. For example:
• Suppose you forget to tell your student to use algebraic nota­
tion and he says "Suppose I move the knight over there" or "I
could take the pawn" - these could mean any of many possible
moves, and later it will be quite confusing to try to sort out what
they meant.
• Suppose you are giving an internet lesson over the phone and
forget to tell the student to say "Push clock" at the end. The
student says "I think I will play �e4" and then the instructor
initiates the post-exercise review. Later the student might say
"Yes, but I really was not finished I was just verbalizing that I
-

might play �e4 and then you interceded."

After these problems arose, I adjusted my instructions to include
items that would preclude them from recurring. Future instructors can
learn from my mistakes!
At the end of the exercise, I recommend that an instructor com­
ment on the time taken for the verbalization, in comparison to the
ti me control and the criticality of the position. Since roughly 80% of
intermediate and below students either play too fast or too slow, the
time taken (roughly twice as long as it would take to do silently) should
help indicate how well the student is using his time. For example, if a
student took 20 minutes to make his move, that would equate to about
10 minutes in a real event, and that may be too slow or too fast.

197
The Improving Chess Thinker

Players who move too quickly have a tendency to slow down fo r de

Groot exercises. This is both bad and good. But I do tell everyone to not
"show off' for the exercise by doing more - or less - analysis than they
would in a real World Open 40/2 game.
I recommend repeating the de Groot exercise - with different posi­
tions - every few months with each student. That way one can t rack
progress in the areas of weakness exposed in the earlier exercises.
Between de Groot exercises I often perform what I call a "DATSCAN"
with students. This is a similar exercise, but it is "Dan - assisted think­
ing" where the student and I find a move together. That way the student
has the benefit of not just hearing Euwe analyze, but doing it "live" with
my help. In order to facilitate a DATSCAN, I find a recent position from
a game which neither my student nor I have seen. For this purpose I
often use the "finger live" command on the Internet Chess Club (ICC)
to get a list of recently played grandmaster games, and then my student
randomly picks a game and chooses a move number.
For example, we may pick "Linares09" and, if there are 30 games in
the ICC library list, my student picks a number in that range. Once the
game is selected, he may want to find White's 23rd move. Therefore, I
forward 43 ply to Black's 2 2nd move. We reveal Black's 22nd move and
together try to find White's 23rd. At the finish - after we have methodi­
cally chosen our move - we see what White actually played, play over
the rest of the game and, finally, give White's 23rd move to a strong chess
engine to compare how we - and the Grandmaster - did.
Both the de Groot and DATSCAN thought process exercises are very
popular with my students and get great feedback "ratings" as instruc­
tive tools.

Another view of the thought process

After administering a de Groot exercise I often am asked "What is a
minimally correct thought process that can be applied to a typical slow
game "analytical" position?" The student is usually looking for some­
thing less imposing then the detailed process I presented in a Chess
Cafe (www.Chesscafe.com) Novice Nook titled A Generic Thinking Pro­
cess. The following reflects my answer and will hopefully not conflict,
but rather augment, the information basic thought process discussion
in Chapter 2.

1. After your opponent's move, ask yourself "What are all the things
that move does?" and "What are all the moves he can do now which h e

198
 What the Researcher Learned

could not do before?" Concentrate upon opponent's moves and ideas

that can really hurt you. Obviously, if he made a check you need to get
out of it, and if he made a capture you likely need to recapture, pos­
sibly next move after a zwischenzug (in-between move). However, if his
move is not a check or capture, look for the threats it created. Threats
can be found by asking, "Suppose I pass and my opponent just moves
again. What could he do to me that I would not like?"

2. To begin looking for your move, consider moves that meet your
opponent's threats and also forcing moves: checks, captures, and
threats. If there are none of consequence, plan to make your army more
active, e.g. identifying your piece which is doing the least and finding a
move or plan which makes it do more or, conversely, moves that restrict
your opponent's activity. Another approach is to find moves that take
advantage of opponent weaknesses or your strengths. Don't waste time
on grandiose plans that are not, to paraphrase IM Jeremy Silman, both
feasible and effective. Discard potential threatening moves that are
easily met and leave your position worse than before. The reasonable
moves you generate are called candidates.

3. Find the checks, captures, and threats that your opponent could
reply after each candidate. If he can make even one reply that you can­
not survive, then your candidate should likely be discarded and is not a
final candidate.

4. For each of the remaining candidates, assume your opponent will

make his best reply and try to figure out what (short) sequence is likely
to occur. Visualize the position at the end of that sequence and evaluate
it. In order to evaluate a position, it usually should be a quiet one and
not in the middle of a checking or capturing sequence. For unclear sac­
rifices you just have to use your experience and judgment. Do not make
the common mistake of evaluating the position immediately after the
candidate move, ignoring opponent's replies, and failing to anticipate
thre ats! If your sequence is reasonable (for both sides), the evaluation
of the position at the end of the sequence will reflect how much you like
that candidate move.

5. If you see a good move, look for a better one! After performing
#4 for each candidate, compare the evaluation of the resulting position
with the evaluation of the best position you have found so far, the "King
of the Hill". If the new move's position is even better, it becomes the
new King of the Hill.

199
The Improving Chess Thinker

6. Once you have finished evaluating all your candidates (or as many
as possible within reasonable time constraints), your move of choice
is the final King of the Hill! The sequence of moves you found for that
"best" move is called the Principal Variation (PV). A PV is the sequence
that ChessMaster, Fritz, Rybka, and other chess engines display as their
top analysis line. If you see that the PV wins by force, then your current
position must be winning!

7. Do a sanity check. When playing over-the-board, write down your

move, close your eyes, and/or take a deep breath. Re-examine your
move with fresh eyes. Is it just crazy? Does it leave a piece en prise?
Miss a mate for you or allow one for your opponent? You should not try
to redo your entire analysis of the move. If the move is crazy, erase it
and reconsider, starting with the "second-best King of the Hill." If it is
not crazy, make the move, hit the clock, and record your move and your
time remaining.

Conclusion
Chess is a thinking game; however, although there are thousands of
books written about chess knowledge such as openings and endgames,
there are very few about the thinking process itself. The de Groot exer­
cise provides an excellent insight into a chess player's thought process.
The results of the exercise not only provide a roadmap into that process,
but can help pinpoint areas of improvement for players of all levels.
Hopefully The Improving Chess Thinker has helped illuminate not only
the processes used by typical players, but also provides a representative
selection of my many recordings of these exercises.

200
Appen d ix A

Dr. M ax Euwe's Protocol of de G root A

he following is excerpted from Thought in Choice in Chess by Adri­

Taan D. de Groot, published by Basic Books, Inc in 1965. It has been
translated from descriptive notation into algebraic notation, with a
couple of explanatory parenthetical notes added:

de Groot A: White to move

GM Dr. Max Euwe's Protocol of deGroot A

Time: 15 minutes December 15, 1938
"First impression : an isolated pawn; White has more freedom of
movement.
"Black threatens . . . 'ltlxb2 ; is it worthwhile to parry that? It probably
is. If he takes, then a3 is also attacked. Can White then take advantage
of the open file? Does not look like it. Still again 2.ti:lxd5 and then by
exchange the pawn at a3 is defended by the Queen, indirectly in connec­
tion with the hanging position of the Knight at f6 and possibly because
of the overburdening of the Bishop at e7.

"But wait a moment. No, . . .'lt!xb2 is rather unpleasant after all be­
cause the Bishop at a2 is undefended. Can I do something myself?
I nvestigate that first: the pieces at f6 and d5 are both somewhat tied
down . Let us look at the consequences of some specific moves:
"1. ti:lxd5, possibly preceded by 1.ti:lxc6. Then (after 1.ti:lxc6) l...:1'1xc6
is probably impossible because of taking on d5. Black has a number of

201
The Improving Chess Thinker

forced moves; there may be a possibility to take advantage of that. It is

not yet quite clear.
"Let us look at other attacks:
"l. .ih6 in connection with fl - but I don't really see how to get at it.

"l. b4 in order to parry the threat. But then exchanging at c3 will give
some difficulties in connection with 2 ... .ibS - oh, no! that is not correct,
one can take back with the Queen.
"So far a somewhat disorderly preliminary investigation. Now, let's
look in some more detail at the possibilities for exchange: l.li:Jxc6 or
l.ii:JxdS or maybe 1.hdS or maybe first 1.M6.
"l. li:Jxc6 E:xc6 2.capture on dS; for instance 2.ii:JxdS exdS wins a
pawn, but there may be compensation for Black on b2. But better is
2 ... ii:JxdS, then 3.hdS E:xcl is nearly forced ... no, it is not, he can play
3 . . . hgS as well. I see no immediate advantage. l.. .bxc6 is not forced
therefore, and even if it were forced you couldn't be quite certain to win.
It's happened before that such a position proved less favorable than it
seemed to be. The point dS is reinforced by it; that is a disadvantage.
(Let's look at) taking on dS. l.li:Jxc6 at any rate gives the pair of bishops;
if l don't find anything better, I can always do this.
"l.ii:JxdS hdS, is that possible? d7 is free then. 2 .M6 M6 3.ii:Jd7
%Yd8 can then be done. l.ii:JxdS hdS 2.ixf6 ixf6 will probably yield
something. l...ii:JxdS is also possible, maybe better. Then 2 .hdS �gS
and now there are the possibilities to take on c6 or to play something
like f4; once again :
"l. li:JxdS li:JxdS 2.hdS hgS - no, nothing then, 3.E:xc6 is a cut e
move but at the end of it all everything remains hanging. Something
else: 2.he7 - he just takes back. l...exdS is very favorable (for me) ; he
won't do that; it needn't be investigated.

"l. li:JxdS li:JxdS remains. 2.hdS hgS 3.hc6 hcl is then possible.
No, (I) can find no way to make anything out of this. l...ii:JxdS 2.�h6
Elfd8 3. %Yf3 with some threats; if Black now has to play his Bishop back
to e8, then one gets a good position.
"1.hdS: this must be looked into. Does that make any difference?
1.hdS hdS is again impossible because of ii:Jd7. That is to say, we will
have to look out for . . . �c4, but that we can possibly cope with; the worst
that can happen to me is that he regains the exchange, but then I h ave

202
 Dr. Max Euwe's Protocol of de Groot A

in any case some gain of time. 1.hdS ti:lxdS . . . same difficulty as before.
No! That is now impossible. 2.ti:lxdS wins a piece!
"1.hdS hdS 2 ..bf6 .bf6 3.ti:ld7 �d8. Let's have a closer look at
that: 4.ti:lxdS exdS and I'm an exchange to the good - very strong. l..b:dS
exdS is therefore forced. But that's good for White. The Knight on f6 is
weak and the Bishop at e7 hangs - and the Bishop on c6 stands badly.
On positional grounds one could already decide on 1.hdS.
"Is there some immediate gain? 1.hdS exdS looks bad for Black.
Probably some more accidents will happen. Much is still up in the air.
One plays, for instance, 2.�f3. Defending the Knight on f6 is not so
easy; 2 . . . l!lg7 looks very unpleasant. . . . . . Yes, I play 1.hdS."

Comments
Dr. Max Euwe was the World Chess Champion from 1935 through 1937.
In Thought and Choice in Chess, thanks to the proximity of the famous
AVRO tournament in 1938, de Groot was able to obtain and study the
protocols of several leading grandmasters, including Alexander Alekh­
ine, Reuben Fine, and Paul Keres, mostly on this position A. But Dr.
Euwe, the math professor, has the most logical and instructive protocol.
For that reason I have used his assessment as a shining example of how
to think in analytical positions. Some highlights:
• Dr. Euwe's order is classic: evaluate the position, evaluate the
opponent's threats, and then start figuring out if there are any
forcing lines in his favor.
• He clearly uses the "King of the Hill" device, as per his state­
ment " ... if l don't find anything better I can always do this."
• His choice of the order to examine the forcing moves is sys­
tematic and logical. l.ti:lxc6 wins the bishop pair, l.ti:lxdS is a
fair trade of knights, 1.hdS gives up the bishop pair, and lastly
l..bt"6 not only gives up the bishop pair, but also trades off
the bishop of the color on which his opponent is weak. So the
chances are that the first one he examines will be the best, and
that is usually the most efficient order to analyze.
• Not only does Dr. Euwe consider l.ti:lxc6 as his first recapture,
but he also makes a point to conclude that it is favorable to win
the bishop pair and that move is worthy of strong consideration.

203
The Improving Chess Thinker

Although I am careful to ensure that almost all my students un­

derstand the value of the bishop pair before they do a de Groot
exercise, most don't even mention winning the bishop pair, and
no one has begun the "candidate" portion of his analysis with
something like "I can always start by playing 1.'ll xc6 to win the
bishop pair, which is worth about a half pawn on the average .
Now let's see if we can find something better . . . " (For IM Larry
Kaufman's article about his scientific experiment to compute
the average value of the pieces, see: www. danheisman.home.
comcast.net/-danheisman/Articles/ evaluation_of_material_
imbalance.htm)
• After finding a forcing move that is better than any other, he
does not need to consider non-forcing moves. Euwe under­
stands that it is virtually impossible that a non-forcing move
can be better than a move which successfully forces the oppo­
nent to make a major concession.
• At the end, Dr. Euwe makes his move even though several of his
statements indicate the uncertainty as to how good that move
is. This shows that Dr. Euwe understands he is trying to find
the best move, not to determine (i.e., waste time) figuring out
how much better the best move is than the second best move.
Dr. Euwe has calculated that 1.hdS is best, but does not take
the time to figure out if it is actually winning, saying, "Much is
up in the air. . . probably some more accidents will happen."

2 04
Appen d ix B

Computer Analysis of Positions

ost of the following analysis is provided by Rybka 3 multiproces­

M sor 32-bit, currently the top rated program in the world.
de Groot A: White to move

1.hd5 According to a deep search by Rybka, the second best move

- believe it or not - is l.%Ye2, e.g. l.. .%Yxd4 2.hdS li:JxdS 3.!'lfdl with a
slight advantage.
Let's consider the more common candidate l.li:JxdS. It is amazing
that almost no player (among hundreds ! ) rated below 2000 reached
the critical position that is necessary to show that l.ii:JxdS is not best:
1.. .li:JxdS! 2.hdS attempting to remove the guard 2 ... hgS! For some
reason it was very difficult for intermediate players to deduce that this
line was forced for Black after 1. li:JxdS - perhaps they missed the discov­
ery idea on gS after the attempted removal of the guard on dS.
In the hundreds of de Groot A exercises performed, no one has ever
noticed that if White does not move and Black plays l.. .%Yxb2, then
2.li:J c4 traps the queen! So l.. .%Yxb2 is not really a threat at all. Even the
Grandmasters in de Groot's book failed to mention 2.li:Jc4 - not that it
was that relevant.
1 ... exd5 After l.�dS this pawn capture is pretty much forced:
1. . .li:JxdS? 2.ii:JxdS wins a piece as the queen and e7 bishop are both

2 05
The Improving Chess Thinker

attacked, and l.. .ixdS 2 ..hf6 .hf6 3.ll'id7 picks up the exchange. 2 .
!!fel Apparently equally effective is 2 .%Vf3 with play against the dark
squares f6 and e7, e.g. 2 . . . %Vd8 3.!!cel lt'ie4 4.fi.h6 removing the guard on
fl and White is pretty much winning. 2 Vd8 3.Ve2 !!e8 4.VfJ with
•••

a winning position, though not an easy win, for White, e.g. 4 . . . a6 5.h3
@g7 6.!!e3 gf8 7.gcel.
Against many of the passive defenses ofb2 preferred by weaker play­
ers (1.b4, i.gc2, l.%Yd2) the capturing sequences started by l...lt'ixc3 are
at least even and at best better for Black. For example 1.b4 lt'ixc3 2.!!xc3
.Abs 3 . .Ac4 hc4 4.1Lixc4 %Ya6 with queenside play for Black. This again
shows that it is better to punch first than to wait until your opponent
punches you !

de Groot B: Black to move

Without computers, de Groot was handicapped in his analysis of

most positions. He wrote about position B, "An objective solution to
this position cannot be given." In both B and C there were important
errors in his possible "truth" of the position. For example, here he
presented a main line of play as l gbs 2.b3 gbs 3.c4 !!g5 4.fJ h5
•••

5.gxh5 !!xh5+ 6.'ttgl .Ae5 7.!!adl .Ad4+ 8.@fl !!hi+ 9.'tt e 2 !!h2 .
But White has a couple of important improvements retaining a winning
advantage. For example, White can play 4.d4 instead of guarding g4
with 4.fJ?! , with the idea that 4.d4 gxg4 5.d5 is worth a pawn to both
protect against the kingside attack, block the bishop out of the game,
and get the pawns rolling. Even after the less effective 4.fJ h5 5.gxh5
!!xh5+ 6.'ttg l .Ae5 White can sacrifice with 7.!!xe5 !!xe5 8.'ttf'2 , al­
though this line is less clear.

206
 Computer Analysis of Positions

de Groot C: Black to move

In position C de Groot suggests Black should play 1 e5 2 . .txc8

•.•

gaxc8 with the two main lines of continuation 3.�a7 and 3.dxe5.
Rybka suggests that Black has at best a fight for equality with 1 �e4
•••

or possibly the passive 1 h6 or 1 ... ges. The key contention occurs in

••.

the long forcing line after 1 e5 2 . .txc8 gaxc8 3.dxe5 '9e4 4.ghel
.••

'9xe2 5.�e2 �g4 6.exd6 ht'3 7.d7 gds s.ges+ MS 9.�xc7 �xf2
Here de Groot bases his defense on White playing 10 .l:ixfS+ but the
computers greatly prefer 1 0 .gd4 with a nice White advantage, and so
1 e5 is not as good as he thought.
•••

de Groot Shafritz: White to move

Unlike most of the other exercises in this book, de Groot Shafritz

does not feature much friction between the forces, and thus is more
strategical than analytical. I played Black and had fallen behind in

207
The Improving Chess Thinker

development. During the game I realized that White would be better

if he found the only dangerous continuation l.f4 e4 2.g4! , taking ad­
vantage of the excellent rook position on the kingside. No other plan
would do, as waiting even one move would allow Black to guard g4 with
a move like l...ti:ldf6 and the danger would be passed. Interestingly, it
seems like this problem is a good one to separate Experts from Mas­
ters, as in my small sample the Experts did not see that 1.f4 followed by
2 .g4 was the panacea, while the Masters did! In the game White played
the cautious l.ti:ldl(?), and after 1...ti:ldf6, Black was almost equal and
went on to win in fine style:

Shafritz, Arnold (19 0 0) - Heisman, Dan (2285)

Main Line CC Champ G/75 (rd.3), Oct 22, 2002
1.d4 li:lf6 2.c4 g6 3.li:lc3 .!.g7 4.e3 0-0 5.g3 d6 6 .!.g2 li:lbd7
•

7.li:lge2 e5 8. 0 - 0 c6 9.ti'b3 ti'e7 1 0 .!.d2 li:lb6 11.d5 c5 12.a4 li:le8

•

13.e4 f5 14.a5 li:ld7 15.exf5 gxf5 16.l::!a el ti'f7 de Groot Shafritz occurs
here. 17.li:ldl 17.f4! e4 18.g4! and White is much better, e.g. 18 ... �d4+
19.lt>hl 1/!ig7 2 0.�cl 1/!ixg4 21..b:e4. 17 li:ldf6 18.lt>hl 18.f4! is still best
•••

but not as powerful as before. 18 ti'h5 19 .!.:f3 White is still a little bet­
••• •

ter after 19.ti:le3 or 19.f3. 19 li:lg4 2 0 .hg4 fxg4 21.li:lgl? White can
•••

maintain equality with 2 1.ti:lec3 1/!if5. 21 li:lf6 22.ti'e3 .!.f5 23.li:lc3 l:o!f7
•••

24.ti'g5? Better is 24.ti:lge2 when Black is better but not winning. 24 •••

ti'xg5 25.hg5 i.d3 26 .!.xf6 .!.xf6 27.li:le4 .!.e7 White resigns.

•

One possible continuation would be 28.b3 �b8 29.f4 b5 30.axb6 �xb6

31.ti:ld2 hfl 32.�xfl exf4 33.�xf4 �xf4 34.gxf4 �a6 etc.

de Groot Zyme: White to move

Here Black has strong threats, e.g. 1.1/!ixa8?? 1/!ixc3+ 2.lt>dl ti:lxf2 #

208
 Computer Analysis of Positions

(a move often missed by weaker players who instead spot l.. .�xal+ first
and never look for a better move). Whenever the opponent has a threat
of a forced mate, one must either mate the other player first or stop the
mate, so that sharply narrows White's possibilities. Passive defenses to
both c3 and f2 are possible, e.g. l.�d4 or l.�f3. However, the proper
defense is the forcing deflection 1.i.e7+ ! Y:Yxe7 when White can now
safely capture the rook with 2.�xa8 or, even better, throw in the zwisch­
enzug 2.!:Ml + 'it>eS 3. Y!YxaS with a much better game. In order to play
1.i.e7+ it is not necessary to figure out whether 2 .!'ldl+ or 2.�xa8 is bet­
ter since either leads to a game far superior to White's alternatives on
move 1. Therefore the proper thought process strategy is to determine
that Lie?+ is the correct move and, at that point, play it, and then next
move decide the best way to take advantage of the situation.

de Groot Ernie: White to move

According to a Rybka deep search, the best line for both sides is
1.Y:fd4! !1!xf5 2.!:o!eS+ 'it>h7 3.!1!e7! 3.�e4?! d5! with counterplay. 3 •••

!1!f7! 4.!:o!ael 4.!'lxf7?? ti:le2+ wins the queen. 4.!1!xe7 Also insufficient
is 4 . . . ti:lh3 + ! ? , a computer desperation move to clear f4 for the queen:
5.gxh3 �f4 6.!'lxf7 �xf7 7.ti:lh4 with a big initiative for White. 5.!1!xe7
�e6 6.Y:Je4+! 'it>gS 7.Y:Jxe6 fxe6 S.!1!xc7 leads to a clearly winning
endgame for White.
Other reasonable tries are 1.Y!Ye5 Y!Yxe5 2.!1!xe5 d6 3.hcS Forced.
3. dxe5 4.hb7 with a nice advantage for White, and 1.Y:fa5 Y!Yxa5
2. !1!xa5 b6 3.i.e4 (the move I missed when presented with this inter­
esting position) and again White retains a nice advantage. One common
error among weaker players was l.�e4?? to save the queen and guard
the bishop, but then l.. .d5! and Black wins a piece.

209
G lossa ry
Activity The amount of beneficial things a piece (or pieces)
can do in a given position
Amateur In chess, a non-master. At the US Amateur, mas­
ters cannot play. At the US Amateur Team tour­
naments, the team has to average below master
rating. Note : in chess, amateurs can win money,
sometimes quite a bit at tournaments like the
World Open.
Analysis The part of the thought process where you gener­
ate the move tree, e.g. "If I go there, what would
he do?"
Analytical Position A position where analysis is required to figure out
which move to be played. This would include all
positions with potential tactics, dangerous forc­
ing moves, etc. See Non-analytical position.
Back Rank The rank where a player sets up his major pieces
(1'' for White; s•h for Black)
Bishop Pair (The Advantage of. . . ) This is when one player has
two Bishops and the other does not.
Book Besides the kind with a spine, a "book" move is
one that a player has learned to play in a par­
ticular position in the opening (from a "book"
or other media) without the need to "calculate".
All the book moves in a certain opening before a
major deviation is a tabiya.
Blunder A bad move; primarily a move that turns a win
into a loss or draw, or a draw into a loss. Note
that ANY bad mistake is a blunder - not just a
counting mistake, or falling into another tactic.
Break move A pawn move attacking an opponent's fixed pawn
(thus forcing the possible "break up" of his pawn
structure)

210
 Glossary

Calculation The part of analysis involving forced sequences.

Candidate (move) A move under consideration (during the thought
process). Here are some special types of Candi­
dates
• King of the Hill - the best candidate found so

far.
• Initial Candidate - any move which does

something positive
• Final Candidate - a candidate which is safe
(cannot be easily defeated by a forcing move)
Closed File A file with pawns of both sides
Closed Position A position without any open or semi-open files.
CM Candidate Master - A FIDE title for those with
FIDE ratings between 2200-2299
Coordination The ability for various pieces to work together
harmoniously to achieve a goal.
Counting A basic tactical idea; the ability to calculate that
no sequence of captures would win material on
any square.
Critical move Where the best move(s) are enough better than
the second best to make a difference (win to draw
or loss; draw to loss; easy win to difficult win).
The exceptions are "only" recaptures where only
one piece can recapture which are, by definition,
non-critical. Critical moves often include compli­
cated decisions, trading pieces, or inflexible plans
which cannot be changed. Critical moves should
be played carefully and slowly.
Criticality Assessment The ability to differentiate the criticality of a
move.
De Groot exercise A "think out loud" exercise where a player is
given a position and asked to find a move as if
he were playing a normal slow game, except ver­
balizing his thoughts. De Groot positions are not

211
The Improving Chess Thinker

"problems" with a defined goal, but rather just

normal chess positions from games where there
may be no clear "solution".

Discovered Attack An attack by a piece that was opened up via an­

other piece's move. (also "Discovery")

Discovered Check A discovery where the discovered piece also

makes a check.

Discovery See Discovered Attack

Double Attack An attack on two (or more) pieces by a single move

Double check A discovered check where both the discovering

and discovered piece both deliver check. In that
case, the opposing king must move.

Doubled Pawns Two pawns of the same color on the same file as
a result of a capture.

En Passant Capturing a pawn that moved 2 spaces with a

pawn that could have captured it if it had only
moved 1 space, on the next turn only.

En Prise "In take" - able to be captured for free. A piece is

en prise if it can be captured but is not guarded .
Pronounced "aan-preez".

Evaluation The part of the thought process performed at

analysis nodes (moves that generate positions)
which asks, "Who stands better, how much, and
why?"

Exchange (the) Trading a Bishop or Knight for a Rook is win­

ning the Exchange. Losing a Rook for a Bishop or
Knight is losing the Exchange. NOT the same as
trading pieces.

Expert Someone with a US Chess Federation rating be­

tween 2000 and 2199

Fianchetto To develop a Bishop on a long diagonal (b2 or g2

for White; b7 or g7 for Black).

FIDE International Chess Federation

212
 Glossary

FIDE Master (FM) Someone with the lowest International

Chess Title
File The rows of a chessboard going up and down, let­
tered a-h Oower case), with "a" always on White's
left (and Black's right)
Flexibility The capability of a piece or entire army to do
multiple good things. A lack of flexibility is rigid
and generally very bad.
Flip-coin chess A type of thought process used by beginners
where they make their move without even check­
ing the threats of the opponent's previous moves.
The results of games between flip-coin players
are somewhat random, like flipping a coin.
Forcing moves Checks, captures, and threats
Fork A double attack, usually by a Knight or Pawn
(thus looking like a "fork" in the road).
Grandmaster (GM) Someone with the highest Interna-
tional Chess Title
Hole A weak square on the opponent's side of the
board.
Hope Chess When you make a move without considering
whether a possible opponent reply of a check,
capture, or threat can be met.
Hope Chess is NOT these other "Hope" possibili­
ties:
• When you make a threat and hope your oppo­

nent does not see it, nor

• When you make a bad move on purpose but

hope your opponent makes a worse one

• You make a unanalyzed move quickly and

hope the move/idea/maneuver works anyway

I ncrement time (in seconds) added to your clock for each
move. For example, in the Team4545 League the
time control is 45 45 so you get 45 minutes for

213
The Improving Chess Thinker

the game plus 45 seconds added for each move.

As opposed to time-delay
Initiative When your opponent is constantly responding
to your forcing moves (checks, captures, and
threats)
International Master Someone with the "intermediate" International
(IM) Chess Title
Internet Chess Club - a leading chess server and owner of the
(ICC) ICC Chess.FM radio station
Isolated Pawns Pawns that have no other pawns of the same color
on adjacent files.
Judgmental Position See "non-analytical position"
I.asker's Rule If you see a good move, look for a better one (be­
cause you are trying to find the best move you can
in a reasonable amount of time).
Master Someone with a US Chess Federation rating be­
tween 2200 and 2399 is a national master (NM).
Material Piece value - when you win a pawn, a piece, the
Exchange, you are winning "material" On the av­
erage a Knight and Bishop are worth 3.25 pawns,
a Rook 5 pawns, a Queen 9.75 pawns, and the
King has a fighting value of about 4 pawns. Give
yourself a bonus of 0.5 pawns if you have the ad­
vantage of the Bishop pair.
Mobility The number of moves a piece has, broken up
into:
• potential mobility - the number of moves a

piece would have from a given square if the board

was empty
• actual mobility - the number of moves a piece

has in a any given position

• global mobility - the number of future squares

a piece can land upon in any given position

Non-Analytical Position A position where judgment, rather than analy­
sis or calculation, can be used to select a move.

214
 Glossary

Theses are "quiet" positions such as those that

occur early in the game where the opponent has
no dangerous forcing moves to consider in reply.
Also known as a "judgmental position."
Notation The recorded moves of a game. Note that you
don't "notate" a game - you record it.
Open File One with no pawns of either color
Outpost A piece guarded in a hole (a weak square on the
opponent's side of the board).
Pawn chain A series of pawns on contiguous files, usually
where most of the pawns protecting each other

Pawn structure How the pawns for one side are currently placed
on the board.
Pin An attack by a Rook, Bishop, or Queen, on a piece
that cannot/should not move because a piece be­
hind the attacked piece along the line of attack is
worth even more (if the piece behind is a King,
this is an "absolute" pin) and the piece is not
allowed to move, or it would put the King into
check, which is illegal.
Ply A half-move, or the move of one player. When
both players move, that is two ply, or one full
move.
Promote What a pawn does that reaches the other side of
the board. The moving player replaces it with a
Queen, Rook, Bishop, or Knight on the promot­
ing square. Therefore, one can have nine Queens,
maximum.
Rank The rows of a chessboard going sideways, num­
bered 1''-Sth starting from White's side as 1''.
White's pieces are initially on the first rank;
black's on the s•h.
Rating A measure of skill. USCF Ratings range from
roughly O (basically impossible to get this low - no
one ever has) to 3000; most scholastic beginners
start around 400. Even if you lose all your games

2 15
The Improving Chess Thinker

in your first few tournaments you are still usually

about 200. Rating is different than a ranking,
which places players in some sort of order. For
example, a rating of 2800 would usually have a
player ranked first.
Real Chess The highest generic type of thought process: For
every move you consider whether a possible op­
ponent reply of a check, capture, or threat can be
met. Compare to the lower Hope Chess and the
lowest Flip-Coin Chess.
Removal of the Guard A chess tactic where the defender is captured or
forced to move so that it is no longer defending.
The two main types are deflection and over­
worked piece.
Rybka The best PC-based chess playing program in
2 006-09.
Semi-Open File A file containing only pawns of the opponent.
Skewer/ X-ray Tactical Motif: Sort of an inside-out pin. A move
that attacks a piece of value, and there is a piece
behind it along the line of attack of equal or lesser
value that will be attacked anyway if the attacked
piece moves.
Slow chess Chess played at long enough time controls where
good moves are generally more important than
the time remaining.
Tactics A forcing set of moves that involves piece safety
and checkmating. In advancing level of complex ­
ity, this includes:
• En prise (can take a piece for free)
• Counting (is each piece safe for potential cap­
tures on its square?)
• Motifs (pins, double attacks, removal of the
guard, etc.)
• Non-sacrificial combinations of motifs
• Sacrificial combinations

216
 Glossary

Tempo The "time" it takes for one of the players to make

one move. A "turn".
Threat A move which can win material, checkmate, or
make progress next move if the opponent does
not stop it. Attacks on under-defended pieces are
an example of a threat.
Time Control How much time one has for a game. "G" means
all the moves and "SD" is sudden death, mean­
ing that time control requires all the remaining
moves must be played in a specified period.
In USCF over-the-board notation one uses a slash
for moves/time: 40/2 ; SD/1 is 40 moves in 2 hrs
followed by remaining moves in 1 more hr; G/30
all moves in 30 min
Online the two numbers are minutes per game
and seconds incremented. So 60 5 is one hour for
the game and 5 extra seconds per move.
Time Delay The preferred way of using a clock at a USCF
tournament; a digital clock is set to NOT run for
the first N seconds on each move. Time is not
added, as it is with a time increment.
Time Increment Adding time for each move made. For example,
a time increment of 5 seconds adds five seconds
to the clock each time a move is made. Same as
"increment"
Time Management The skill where, when done correctly, you try for
two goals:
• To spend almost all your time for the game
("Macro time management") and
• Where you spend more time on moves that
require it (critical, complicated, etc.) "Micro
time management" See The Two Move Trig­
gers
Trade To exchange one piece for another of the same or
similar value. NOT the same as "the Exchange"

217
The Improving Chess Thinker

Trigger One Finding a move such that there can't be a better

one (thus triggering a player to make that move)
Trigger Two A reasonable amount of time to make a move,
given the position, the time control, the times on
the clock, and possibly the move number.
Two Bishops Short for "The advantage of the two bishops" -
See "Bishop Pair"
USCF United States Chess Federation
Weak square A square which can no longer be guarded by a

pawn.
Zugzwang When you have to move, but any move is
bad for you. Note: Some contend it is not true
zugzwang unless your opponent could not win
without this compulsion (in other words, if you
could pass but your opponent can still win, then
although any move is bad for you it is not a true
zugzwang).

218
Bi bliogra phy
Books
Aagaard, Jacob, Inside the Chess Mind, Everyman Chess, London,
2004.
Alburt, Lev, and Lawrence, Al, Chess Rules of Thumb, Chess Informa­
tion and Research Center, New York, 2003.
Bain, John, Chess Tactics for Students, Learning Plus Inc., Corvallis,
1993.
De Groot, A.D., Thought and Choice in Chess, Basic Books, Inc., New
York, 1965.
Heisman, Dan, Elements of Positional Evaluation. 3n1 Edition, Chess
Enterprises, Moon Township, 1999.
Heisman, Dan, Everyone's Second Chess Book, Thinker's Press Inc. ,
Davenport, 2005.
Heisman, Dan, Looking for Trouble, Russell Enterprises Inc, Milford,
2003.
Heisman, Dan, The Improving Annotator, Chess Enterprises, Moon
Township, 1995.
Kotov, A. , Think Like a Grandmaster, Chess Digest Inc., Dallas, 1971.
Krogius, Chess Psychology, Alfred Kalnajs, Chicago, 1972.
Lasker, Emmanuel, Common Sense in Chess, Dover Publications, New
York, 1965.
Meyer, Cluas Dieter and Muller, Karsten, The Magic of Chess Tactics,
Russell Enterprises Inc. , Milford, 2002 .
Nunn, John, John Nunn's Chess Puzzle Book, Gambit Publications
Ltd., London, 1999.
Rowson, Jonathan, The Seven Deadly Chess Sins. Gambit Publications
Ltd., London, 2000.
Rowson, Jonathan, Chess For Zebras, Gambit Publications Ltd., Lon­
don, 2005.

219
Silman, Jeremy, Silman's Complete Endgame Course, Siles Press, Los
Angeles, 2007.
Soltis, Andrew, Finding a Good Chess Move, B. T. Batsford Ltd., Lon­
don, 2 004.
Soltis, Andrew, The Wisest Things Ever Said About Chess, B. T. Bats­
ford Ltd., London, 2008.

Articles/Columns
Heisman, Dan, "Novice Nook", www .chesscafe.com.
Heisman, Dan, "The Thinking Cap", www.jeremysilman.com
Kaufman, Larry, "The Evaluation of Material Imbalances", Chess Life,
March 1999.