Grandmaster Opening

Preparation
By

Jaan Ehlvest

Quality Chess
www.qualitychess.co.uk
 First edition 2018 by Quality Chess UK Ltd

Copyright © 2018 Jaan Ehlvest

GRANDMASTER OPENING PREPARATION

All rights reserved. No part of this publication may be reproduced, stored
in a retrieval system or transmitted in any form or by any means, electronic,
electrostatic, magnetic tape, photocopying, recording or otherwise, without
prior permission of the publisher.

Hardcover ISBN 978-1-78483-053-3

All sales or enquiries should be directed to Quality Chess UK Ltd,

Central Chambers, Suite 247, 11 Bothwell Street Glasgow G2 6LY, United Kingdom
Phone +44 141 204 2073
e-mail: info@qualitychess.co.uk
website: www.qualitychess.co.uk

Distributed in North and South America by National Book Network

Distributed in Rest of the World by Quality Chess UK Ltd through

Sunrise Handicrafts, ul. Szarugi 59, 21-002 Marysin, Poland

Typeset by Jacob Aagaard

Proofreading by John Shaw
Edited by Ian Kingston and Colin McNab
Cover design by Andrés Wadalupe
Printed in Estonia by Tallinna Raamatutrükikoja LLC
 Preface
This book is about my thoughts concerning opening preparation. It is not a strict manual; instead
it follows my personal experience on the subject of openings. There are many opening theory
manuals available in the market with deep computer analysis – but the human part of the process
is missing. This book aims to fill this gap.

I tried to present the material which influenced me the most in my chess career. This is why a large
chapter on the Isolated Queen’s Pawn is present. These types of opening positions boosted my
chess understanding and helped me advance to the top. My method of explaining the evolution
in thinking about the IQP is to trace the history of games with the Tarrasch Defence, from
Siegbert Tarrasch himself to Garry Kasparov. The recommended theory moves may have changed
in the 21st century, but there are many positional ideas that can best be understood by studying
“ancient” games.

Some readers may find this book answers their questions about which openings to play, how
to properly use computer evaluations, and so on. However, the aim of this book is not to give
readymade answers – I will not ask you to memorize that on move 23 of a certain line you must
play ¤d5. In chess, the ability to analyse and arrive at the right conclusions yourself is the most
valuable skill. I hope that every chess player and coach who reads this book will develop his or her
understanding of opening preparation.

The book includes a lot of games which are historically significant, but my main focus is on the
opening phase of the game. Even so, usually I prefer to give the whole game, even if the final
unannotated moves are not strictly relevant to my theme. I wish any readers who are curious
about how the game ended to have the option to play through the remaining moves. Or if you
prefer, you may ignore the final moves and skip ahead to my next point. I am sure that a reader
who is especially keen may also find these games in other sources with comments on the phases
after the opening.

I would like to thank my opponents and other chess players who contributed to this book with
their games; they are definitely co-authors of this book. Finally, I would like to thank Jacob
Aagaard and Quality Chess for accepting me as an author.

Jaan Ehlvest
Tallinn, Estonia
April 2018
 Contents
Preface 3
Key to Symbols used 6

1 Introduction 7
Overview of the opening problem 10
Brief history 11
2 Evolution of the Isolated Pawn 19
Siegbert Tarrasch 23
Paul Keres 36
Anatoly Lein 48
Boris Spassky 56
Garry Kasparov 72
Application of classical ideas 95
Exercises 109
Solutions 113
3 Key Ideas and Positions 125
Opening concepts 135
The critical or key position in the opening 153
Critical positions in the Scheveningen 156
Sozin Attack 158
Keres Attack 168
Positional ideas 173
4 Computer Preparation versus Human Common Sense 187
5 Modern Trends in the Opening 207
6 Memorizing Opening Lines 211
7 Overview of Critical Positions 221
Some final thoughts about the opening 221
Critical positions for White after 1.e4 221
Sicilian Defence 224
Closed systems for White 226
8 How to Build Your Opening Repertoire 229
The beginner 229
The tournament player 235
Pre-professional level 236
Professional level 247
Psychological warfare in chess 252
Opening repertoire for women 258

Game Index 262

Name Index 266
Opening Index 271
 Key to symbols used
² White is slightly better
³ Black is slightly better
± White is better
µ Black is better
+– White has a decisive advantage
–+ Black has a decisive advantage
= equality
© with compensation
„ with counterplay
÷ unclear

? a weak move
?? a blunder
! a good move
!! an excellent move
!? a move worth considering
?! a move of doubtful value
# mate
Chapter 1
Introduction
In recent years the opening phase has become the most important subject in competitive chess.
The use of strong computer programs has offered a helping hand, and in some cases has made
chess understanding almost unnecessary for becoming a grandmaster. Many players think that
there is no longer a need for a coach: you just need to check the computer evaluation. There
are hundreds of opening books available for different levels. Most of them contain thousands
of lines where the only explanation given is that one move is good and another is not, based on
the computer’s evaluation. The reasons why some moves are wrong and why you should play a
particular line are not explained. How top players distinguish between good and bad lines and
moves remains a secret. What is going on in the head of World Champion Magnus Carlsen is not
available to the general public.

Opening preparation is the most secretive phase of a player’s chess preparation. One may
remember the scandals during the Karpov – Kasparov confrontations, when some of their seconds
were accused of selling information to their opponent’s team.

In the past we had adjournments in chess tournaments, when the sealed move was the most
valuable secret that one might want to know. Nowadays, any information about your potential
opponent, even if it’s only their first move, is already valuable. The main goal is to reduce the
risk of being surprised, to prepare yourself so well that you are not afraid of what your opponent
might do. Of course this is not totally achievable: it always depends on the strength of your
opposition.

In the Moscow Olympiad in 1994, I was playing on Board 2 for Estonia. Our team was doing
very well, and I was winning game after game. One day I was discussing my performance with
my good friend, and one of the best coaches I ever met, Alexander Yurkov. He pointed out that
everything depends on the level at which you are playing. Weaker opponents just do not test
your opening lines adequately, for example. This can mean that you do not need to start solving
problems at the very beginning of the game, and because of that you are less likely to get into
time trouble.
The next day I was Black against Vladimir Kramnik, and now I had a problem to solve: which
opening to choose. My idea was just to get away from his opening preparation as soon as possible.
8 Grandmaster Opening Preparation

Vladimir Kramnik – Jaan Ehlvest defend a slightly worse position. Instead I

posed him some questions very early in the
Moscow (ol) 1994 game.

1.¤f3 b6 
   
   
     
        
        
       
     
   
 
10.exd5 ¥xd5 11.¤e5 0–0 12.£h5
 
Here I did not have much of a choice. It
was a long time since I had played in super-   
tournaments, and my opening preparation was
limited to open events, where, being usually
   
one of the favourites, I often played risky lines    
anyway. Against Kramnik I did not want to
play some major classical line without having
  
studied the position at least for a week.     
   
2.e4
Kramnik was at that time rated 2725, and  
I did not believe that he had ever looked at    
1.¤f3 b6.

12...£d8!?
2...¥b7 3.¤c3 e6 4.d4 ¥b4 5.¥d3 ¤f6
I did not like 12...£g5 because of the
6.¥g5 h6 7.¥xf6 £xf6 8.0–0 ¥xc3
possible 13.£xg5 hxg5 14.f4, when White has
8...0–0 9.¤e2 feels slightly unpleasant for
an initiative in the endgame. White can also
Black.
just play 13.£h3.
9.bxc3 d5?!
13.c4
This move looks suspect, but this was my
13.f4 ¤d7 and Black is just in time.
plan: to make my opponent solve concrete
problems as soon as possible. Back then (and
13...¥b7 14.d5 £d6
still today), Kramnik was one of the best
The only move.
positional players around. I did not want to
 Chapter 1 – Introduction 9

14...exd5?? would lose immediately to 15.£f5 (22...£xc4 23.£f4† ¢g7 24.¦e7†+–) 23.¦e6†
g6 16.¤xg6+–. ¢f7 24.¦fe1+–

15.¦ae1 exd5 16.£f5 g6 17.£h3 ¢g7 20...¤d7 21.f5 ¦h8! 22.fxg6†

 22.£xg6†?! ¢f8! is fine for Black.

    22...¢g8 23.£f4 £xf4

   Now I offered a draw, which was accepted.
My bluff in the opening was successful this
    time, but you cannot rely on this very often.
    ½–½
    It was obvious after this game that I was not
   ready to play Kramnik-level players every
  day. I would need some opening preparation
to survive their pressure. The next year I was
    invited to play in a series of super-tournaments,
 and with the help of Yurkov my opening
18.¤xf7! ¢xf7 preparation improved drastically.
18...¦xf7 19.¦e6 £f4 (19...¥c8 20.¦xd6
¥xh3 21.¦xg6† ¢f8 22.gxh3±) 20.¦xg6† ¢f8 At that time I did not pay much attention
21.g3+– to opening preparation in general. I started
to think about it only in later years, when I
19.£xh6 ¦g8 became a coach myself. I think I tried to
 follow the same route as my coaches in the
Soviet Union: just concentrating on the moves
   or analysis without spending much time on
   opening philosophy. One thing, which was
taught to me by Yurkov, was clear – you should
    not come to any final conclusions too quickly.
    If you studied a position for at least a week
then you might have a valid opinion about
    it. Of course, this was the golden era, when
    computers had not yet interfered.
  Opening preparation is very different today. In
    some cases, as we know, there are still some very
 prominent chess theoreticians around who are
helping today’s top players. One of them is
20.f4?!
Vladimir Tukmakov, who has helped Anish
Until this point, Kramnik had played an
Giri for years and has probably contributed
excellent game.
a lot to his opening preparation. Some of the
players are themselves established authorities,
In time trouble he probably missed that after
like Boris Gelfand and Vassily Ivanchuk.
20.¦e3 d4 he has: 21.c5! £xc5 22.¥c4†! ¢f6
10 Grandmaster Opening Preparation
They share the common trait that they work
very hard on their openings and most likely
they have a very good professional memory.
Years ago, there was a joke about it. It was said
that the late Estonian GM Lembit Oll was one
of the best and he knew nearly everything,
but Vassily was a little bit better, the reason
being that not only did he know everything
that Lembit knew, but he even knew the
blitz games played by Dutch GM Jeroen
Piket.
This of course raises the question of what is
more important: to just remember all the lines
or to understand the subtle nuances of those
variations as well. I think it depends on your
level in chess. In this book I aim to present
material mostly from my own experience; I am
sure this is not the final truth. Everybody can
draw their own conclusions from the material
presented here – I am sure it has something
for everybody. Coaches might see something
in a new light, especially as many coaches lack
practical opening preparation themselves. I am
happy if my fellow grandmasters also find this
material useful.
The aim of this book is to give you the
knowledge of how to work on your opening
preparation.
The book is not only for players. I am
sure most professional seconds and coaches
will gain some value from my observations
and from the practical examples given. It is
not a scientific paper or manual, but rather
I have tried to present material which has
an emotional connection with my chess
career.
Overview of the opening problem
Opening theory stands out in chess because
you face it immediately when starting to
play. Your first introduction may be Fool’s
Mate. Then some other tricks are taught
and finally your coach introduces you to the
systems that he or she knows the best. The
young player advances on the rating ladder
and their openings develop accordingly. At
every stage there are some openings that
are good and some to which you need to
say goodbye. Finally, you reach the World
Championship match and you play only very
limited openings. What happened? Why can
you not play all the openings you would like
to play?
Chess, in this instance, is like life. You can
love anyone you like when you are young and
romantic, but you cannot marry just anyone
when you mature. This transformation, after
which you need to play more solid or classical
chess (I will explain later what I mean by
classical) cannot be ignored. Or if you do
ignore it, then you will probably not advance
high enough and would probably not be
reading this sentence now.
Opening theory is like a living thing, it has
its own evolution and development. There is
a huge amount of data: all the games played
up till today, and a certain number of possible
games which are hidden in the black box,
although we know they exist. Today we can be
sure that every second somebody is running
some opening line on their computer and is
probably hoping to notice some discrepancy
in known theory, or just preparing to make a
bluff.
The criterion for commonly accepted correct
lines today is the so-called computer proof.
This was not the case in the past – we just did
not have computers. Practice was the arbiter
of truth. It took months, years, even decades
to establish a final verdict. The grandmaster
was the solver of the secret. Well-known
theoreticians were the authorities in the
domain of opening theory.
 Chapter 1 – Introduction
Today there are a lot of very talented
youngsters, but not all of them become strong
grandmasters. One trait, however, is common:
they study opening theory a lot.
The problem is how to work with the data.
Even before that you need to have some
idea what you want from this data, which is
available using the ChessBase program and
hundreds of opening books. Finally, there are
also paid consultants available: private coaches.
At the very top level, there is a whole team of
consultants.
Brief history
Chess openings are not natural phenomena –
they are man-made creations. Some opening
ideas that were created hundreds of years ago
are still used today in top games. For example
the Italian Game has been played and analysed
since the 16th century. The Italian Game is
also known as the Giuoco Piano, which means
“quiet game” – an apt description. We shall see
later how World Champion Magnus Carlsen
plays this opening quietly yet effectively.
How were the secrets of this opening developed
in medieval times? Who told the players where
to develop their pieces? Chess logic gives some
hint. To attack the f7-square as quickly as
possible, developing the bishop and knight
for this purpose looks like a good plan. Chess
history tells us that at the end of the 19th
century the romantic style changed to a more
positional approach, when Wilhelm Steinitz
became the first official World Champion.
Steinitz’s new positional style prevailed. We
may say that in the 16th century, players could
not create a correct opening system because
of their lack of knowledge. Before Steinitz
there was Paul Morphy, but he was a romantic
player: quick development and attack at any
cost.
11
We can only speculate how the openings
evolved from the past. Did they have enough
data in the 16th century to use practice as
the criterion of truth? Most likely not, but
definitely other factors were present, among
them fashion and the master who played
the opening in a certain way. If we look into
human evolution using the chess paradigm,
then the human player from the 16th century
looks like some savage who probably could
not beat today’s schoolchildren. They were
not so sophisticated back then, but they had
already started to build opening theory, and
once again it is amazing to see that some of
their openings are still used today. Chess is, at
least in theory, a system with a finite number
of possible games. We do not need to calculate
the number of all possible chess games, but
opening theory makes a point: today, some
openings are listed as no good.
The history of opening theory as such does not
offer much of value to the practical game today,
but it gives you a lot of understanding behind
the lines. We will look at this in more detail
in the chapter on the isolated pawn. Nobody
cares today what openings were played in the
19th century. Professional opening preparation
and study started much later.
First there was Botvinnik, some may argue,
but it is commonly accepted that Robert James
Fischer was one of the first real professionals.
We can assume that from 1970 opening
preparation became the major part of the
top player’s everyday work on chess. Anatoly
Karpov, Viktor Korchnoi and Garry Kasparov
were the leading theoreticians for decades.
In the late 1980s another player stormed to
the top: the controversial, but very talented
and hard-working, Vassily Ivanchuk, a player
who can say about himself that chess is really
his life. It is a pity that Vassily has not written
about (or at least has not published) any of
his hidden path on the road to excellence.
12 Grandmaster Opening Preparation
I remember once at the Linares super-
tournament in 1991 when, after I lost a game
against Jan Timman, we ended up in Vassily’s
hotel room and went through the game. This
was a very disappointing tournament for me,
but Vassily beat both his rivals, Karpov and
Kasparov, and there seemed no doubt that
he would be the next World Champion. This
would never happen, but we did not know
that yet. However, he was definitely interested
in my game with Timman, and he graciously
recommended and showed me which
plan I should have followed in the critical
position.
Jan Timman – Jaan Ehlvest
Linares 1991
1.d4 ¤f6 2.c4 e6 3.¤f3 b6 4.g3 ¥b7
My countryman Paul Keres was a huge
expert on the Queen’s Indian Defence and he
played both 4...¥a6 and the text move. 4...¥b7
was just considered too solid, and this is why
4...¥a6 was introduced by Keres as a more
aggressive move against weaker opponents.
5.¥g2 ¥e7 6.0–0 0–0 7.¤c3 ¤e4

  

   
    
  
   
 
  

Timman was not doing well in the
tournament and I was not in the mood to
fight either. Suddenly I understood that he
was not going to offer me an early draw. This
lowered my mood, because I never played for
a draw, and if I had known that he wanted to
play, I would have played a different opening.
The point is that here we are playing for only
two possible results. Black can hardly push for
a win in this line of the Queen’s Indian, and in
my situation, where I could no longer fight for
a good overall result, this boring scenario led
me to play below my usual strength.
8.£c2!?
Somehow I was expecting 8.¤xe4 ¥xe4
9.¤e1 ¥xg2 10.¤xg2 d5 11.£a4 dxc4
12.£xc4 c5 13.dxc5 ¥xc5 14.¥e3 ¤d7 with
a quick draw.
8...¤xc3 9.£xc3 f5
This position was not new to me. I had
played it as White in training games against
World Championship Candidate and USSR
Champion, Andrei Sokolov. We had training
camps together during the Soviet Union
days.
10.d5!
A rare move. This was the position that
caught Ivanchuk’s eye.

  
 
   
  
   
   
 
   

10...exd5 11.¤e1 d4?
 Chapter 1 – Introduction 13

I was still under the illusion that we were 16...¤e7?

going to make a draw, and I could not force 16...£e7 was much better, with the idea of
myself to work on other moves. playing ...£e4 on the next move, but it just
was not my day.
After the game, Vassily showed me the correct
plan: 11...¤a6! 12.cxd5 ¤c5 17.£d3 d6 18.¦fd1 £c8 19.b4 £a6 20.£b3
 ¤g6 21.¦bc1 ¦e8 22.¦c2 ¤e5 23.¥d4
   ¤g4?! 24.h3 ¤e5
  
       
     
        
    
     
       
   
During the game I did not like how the
white pawn on d5 was killing my bishop on

b7. However, decades later I can see that the    
Houdini program also agrees with Ivanchuk.
We looked at this position for a while, and

25.c5!± bxc5 26.bxc5 dxc5 27.¥xc5 ¢h8
after that I went to the lobby bar. While 28.¤e3 ¦d7 29.¦xd7 ¤xd7 30.£f7 £c8
sipping a beer, I was still wondering at how 31.¥xa7 ¦d8 32.£c4 c5 33.¥xc5 f4 34.¥d6
good Ivanchuk’s opening preparation, or in £xc4 35.¤xc4 fxg3 36.¥xg3 ¦c8 37.¤e3
this case his erudition, was. ¦a8 38.¤d5
1–0
12.£xd4 ¥xg2 13.¤xg2 ¦f7 14.¥e3 ¤c6
15.£d5 ¥f6 16.¦ab1 Mikhail Botvinnik is cited by many as one
 of the first methodical chess researchers. He
   did this tedious job when preparing for his
world championship matches. Throughout
  his career, it seems that he always prepared
    certain openings against certain opponents.
As he mentioned himself, his weakness was
   his less brilliant tactical vision compared with
    his challengers. This handicap was perfectly
     compensated by a clever choice of openings.
This is why we can see in his repertoire
  openings such as the Caro-Kann and the
   French Defence. These are openings in which
general strategy is more important than some
 tactical shot.
14 Grandmaster Opening Preparation

Mikhail Tal was the complete opposite of


Botvinnik. He was famous for handling   
dynamic positions and spicing them up with  
sacrifices. When this was not possible, he    
sometimes lost his objectivity. I think this is    
why he had less impressive results against
players like Paul Keres, who could counter his
  
ideas, as in the following example.   
 
Mikhail Tal – Paul Keres   
Moscow 1959

23...¦a5 24.h6 ¤g6 25.£f3 ¦h5 26.¥g4
¤xf4 27.¥xh5 ¤4xh5 28.hxg7 £d7 29.¢g2
1.e4 c6
¤g4 30.¤d2 ¤e3† 0–1 Fischer – Keres, Bled/
Choosing the Caro-Kann against ambitious
Zagreb/Belgrade 1959.
and talented players who excelled in dynamics
was the perfect decision. Later in the same
2...d5 3.¤c3 dxe4 4.¤xe4 ¥f5 5.¤g3 ¥g6
year, Keres managed to beat Fischer twice
6.¤f3 ¤d7 7.h4 h6 8.¥d3
with the Caro-Kann in the Candidates
Some readers might ask why Tal did not
tournament in Belgrade. This game against
include the move 8.h5 before playing ¥d3. At
Tal was played in August, just before the
that time the positional approach, in which
Candidates.
White relies on the long-term advantage in
the endgame using the fixed pawns on the
2.d4
kingside and the weakness on g7, had not yet
One of the games mentioned in the previous
been worked out. Piece play prevailed.
note continued: 2.¤c3 d5 3.¤f3 ¥g4 4.h3
¥xf3 5.£xf3 ¤f6 6.d3 e6 7.g3 ¥b4 8.¥d2 d4 
9.¤b1 £b6 10.b3 a5 11.a3 ¥e7 12.¥g2  
  
  
    
       
      
     
  
  
    
 
12...a4 13.b4 ¤bd7 14.0–0 c5 15.¦a2 0–0 8...¥xd3 9.£xd3 ¤gf6 10.¥f4 £a5†
16.bxc5 ¥xc5 17.£e2 e5 18.f4 ¦fc8 19.h4 A good move, after which White cannot
¦c6 20.¥h3 £c7 21.fxe5 ¤xe5 22.¥f4 ¥d6 castle long unless he retreats the active bishop
23.h5 from f4. This is a common idea used today by
Caro-Kann aficionados.
 Chapter 1 – Introduction 15

11.c3 e6 12.0–0 ¥e7 13.¦fe1 0–0 14.¤f5 For some reason, after 1959 Keres only
¦fe8 played the Caro-Kann on a few occasions.
Keres makes the most solid move. He preferred his other pet opening, the Ruy
Lopez. If we look at the statistics, this was not
14...exf5 15.¦xe7 ¦ae8 was also possible. a great success. In his last match in the World
 Championship cycle against Spassky in Riga in
1965, his choice of the Ruy Lopez was one of
  the reasons he lost the match.
  22.¦e3 ¤e8 23.¦g3† ¤g7 24.¦dd3 f6
   25.¤g6 ¤xg6 26.¦xg6 ¢f7 27.h5 £a6
    28.b3 £xa2 29.d5 cxd5 30.cxd5 ¦xd5
     31.¦xd5 exd5 32.£d3 £a6 33.£xd5† £e6
34.£f3
   
       
      
   
15.¤xg7?!
Tal did not like grinding chess.    
After 15.¤xe7† ¦xe7 16.c4 ¦ee8 17.¥d6
    
White has some pull.   
   
15...¢xg7 16.¤e5 ¦h8 17.£h3 ¦h7 18.c4
¤f8 19.¦ad1 ¦d8 20.¥d2 £b6 21.¥c3 ¢g8     
 
34...¦h8 35.¥d2 ¦d8 36.¥xh6 £e1†
    37.¢h2 £e5† 38.¢g1 ¤f5 39.¥f4 £e1†
  40.¢h2 ¦d1 41.¢h3 ¦d4 42.¦g4 £d1
   43.h6 ¦d3 44.h7 ¦xf3† 45.gxf3 £xf3†
46.¥g3 ¤xg3
     0–1
    Tal was not a researcher. He noticed and used
    the best chess theory created by others. He did
   well when everything went smoothly his way.
He became World Champion in 1960, beating
    Mikhail Botvinnik convincingly, 12½–8½.
 A year later, however, he lost the return match
White has some compensation and the also convincingly, 8–13.
initiative might look dangerous, but this In both matches the Caro-Kann was
did not work out against players like Keres. Botvinnik’s main choice when defending
16 Grandmaster Opening Preparation
against Tal’s 1.e4. In the first match it was
something of a surprise for Tal, but not in the
second match. However, Tal had not worked
out a general strategy; in other words, he could
not present any real problems for Botvinnik
to solve. Instead he was just trying to surprise
his opponent with some sidelines, which did
not work out. It was not the Caro-Kann itself
which brought the victory to Botvinnik.
Here I just want to point out the differences
in opening preparation between the two
players. In the third game of the first match,
when Botvinnik switched from the French
Defence to the Caro-Kann, Tal came up with
an astonishing move.
Mikhail Tal – Mikhail Botvinnik
Moscow (3) 1960
1.e4 c6
Botvinnik had lost the first game of the
match with the French Defence and now
switched to the more solid Caro-Kann.
2.¤c3 d5 3.¤f3 ¥g4 4.h3 ¥xf3
This position was definitely not new to Tal.
Botvinnik had already had this position in
his match against Vasily Smyslov two years
previously.

 
 
   
   
   
  
 
 

5.gxf3?
At the time, what exactly happened here
was much debated. One theory was that Tal
wanted to show Botvinnik that he could play
against him as he pleased. In chess terms this
is a horrible move. But the winner is always
right, and we can only speculate that this really
shocked Botvinnik. Still, he came back in the
return match a year later, though at a time
when Tal had very serious health problems,
which helped Botvinnik a lot to win the match.
5...e6 6.d4 ¤d7 7.¥f4 ¥b4 8.h4 ¤gf6
9.e5 ¤h5 10.¥g5 £a5 11.¥d2 £b6 12.a3
¥e7 13.¥e3 g6 14.¤a4 £d8 15.£d2 ¤g7
16.¥g5 h6 17.¥xh6 ¤f5 18.¥f4 ¦xh4
19.¦xh4 ¤xh4
Black has a big advantage, but Tal was lucky
enough to somehow save the game.

  
 
 
   
   
   
   
   

20.0–0–0 b5 21.¤c5 ¤xc5 22.dxc5 ¥xc5
23.¥e2 ¥e7 24.¢b1 £c7 25.¦h1 0–0–0
26.¥g3 ¤f5 27.¦h7 ¦f8 28.¥f4 £d8
29.¥d3 ¦h8 30.¦xh8 £xh8 31.£a5 £h1†
32.¢a2 £xf3 33.£a6† ¢b8 34.£xc6 £xf4
35.¥xb5 £xe5 36.£e8† ¢b7 37.£c6† ¢b8
½–½
In his youth, Tal looked like he was unprepared
or just happy to play any position resulting
from the opening. In later years he started to
 Chapter 1 – Introduction 17

show a much more practical approach. His In the same year, Vladimir Yurkov visited
sudden reappearance at the top started with Tallinn in September, lecturing for a week on
the Riga Interzonal tournament in 1979. In the Tarrasch Defence and the Scheveningen.
my opinion, this was due to Kasparov. Garry My knowledge of this line is from that time.
enjoyed working on openings and he was
always very concrete. I do not know exactly

when they started working together, but in 
1980 they had training camps together and  
played some training games which are available
to the public. They both missed the correct    
move order in the following game from 1979.     
Mikhail Tal – Garry Kasparov   
    
Minsk 1979
 
1.e4   
In the USSR Championship in Minsk, Tal
had a mediocre result. Kasparov, however,

16...¥e6?
had just started to rise to stardom, sharing a
Seeing this move and the quick draw by
respectable third place.
repetition that followed, it is clear that both
players were satisfied with a quick draw.
1...c5 2.¤f3 e6 3.d4 cxd4 4.¤xd4 ¤f6
5.¤c3 d6
16...¢h8! is the correct move order.
The classical Scheveningen, an opening that
served Kasparov for years to come.
17.¦ad1
 17.¥h6! g6 18.¦ad1 offers White an
  advantage.

  17...¢h8 18.¥g5 ¤g8 19.¥e3 ¤f6 20.¥g5

    ¤g8 21.¥e3
     I am not against prearranged draws myself;
my point here is that it is obvious that Tal and
    Kasparov had already started working together
     at around that time.
½–½
 
  Very soon after this cooperation began, Tal
produced not just good results, but excellent
 ones. It might sound strange, but the young
6.¥e2 ¥e7 7.0–0 0–0 8.f4 ¤c6 9.¥e3 a6 Kasparov influenced him tremendously in
10.a4 ¦e8 11.¢h1 £c7 12.£e1 ¤xd4 opening preparation. Tal himself probably
13.¥xd4 e5 14.fxe5 dxe5 15.£g3 ¥d8 did not change his attitude much. He never
16.¥e3 became a researcher himself: he just assimilated
18 Grandmaster Opening Preparation

the available information and used it in the

best way. In 1979 he won the Riga Interzonal
and showed some tremendous ideas in the
openings.

In opening preparation there are two different

approaches. One is where you work and create
something of your own. The second style –
the easier one – is just to copy the best lines
available on the market. As a top player, you
always have some of your own ideas as well
as using other players’ ideas. There is always
some kind of mix, especially when coaches and
seconds are also doing a lot of work for the
player. Which approach prevails is connected
to the player’s personality. Some players show
excellent results when they have prepared very
well themselves. They can copy the opening
lines, but they also need to work it out to be
sure. Players like Kasparov, Gelfand, Ivanchuk
and Kramnik belong to this group of players.
Tal, Anand, Khalifman and Karpov belong to
the second group.
Chapter 2
Evolution of the
Isolated Pawn

There are lots of openings that involve an isolated pawn. In some cases it might occur as a sideline.
Most of the time, however, the whole opening is based on the theme of the isolated pawn, and
it is not an accidental occurrence. The opening or variation depends on the value of the isolated
pawn: it determines the whole strategy.

When we look at it from White’s side, it is quite straightforward, and because of the extra tempo
White should always have at least an initiative to compensate for having the isolated pawn.
Opening variations such as the Panov Attack in the Caro-Kann, and many positions from the
Queen’s Gambit Accepted are the first that come to mind.

White rarely has any difficulty keeping the balance. In the worst-case scenario there is always the
simplifying push in the centre, and after trading the isolated pawn a draw is usually the outcome.
White’s plan is to create a kingside attack using the space advantage in the centre afforded by the
isolated pawn, and the constant threat of pushing the pawn forward creates a lot of dynamics. The
term “dynamics” is used here to describe situations in which forced tactical lines are the biggest
factor to look for when evaluating the position.

Positional factors are just connected to the pawn structure, and these become important when
Black can comfortably blockade the isolated pawn and simplify the position. In this case the
dynamic factors do not prevail, and Black may take over the initiative and win the game. Anatoly
Karpov has many brilliant victories on this theme. In his 1987 Candidates match against Andrei
Sokolov, his choice as Black was the Caro-Kann and the Panov Attack occurred. With the white
pieces, Karpov played into the Queen’s Indian, where again isolated pawn positions developed,
this time from Black’s side. The match was all about the isolated pawn and how to handle it.
Sokolov could not get enough attacking chances as White, while with Black he got good positions
but Karpov outplayed him.
20 Grandmaster Opening Preparation
With the white pieces, as I have already
mentioned, having an isolated pawn is not a
particularly risky business. However, in this
chapter I will discuss the opposite situation:
when Black has the isolated pawn after the
opening phase of the game. There are two
main openings, the French Defence and the
Tarrasch Defence, in which right out of the
opening Black has an isolated pawn. In the
French Defence, after 1.e4 e6 2.d4 d5 3.¤d2
Black has the choice of playing a closed
position with 3...¤f6, or going for the isolated
pawn with 3...c5.
The Tarrasch Defence is usually about the
isolated pawn, although it can be avoided in
the main line after 9.¥g5.

 
 Tarrasch Defence can offer. But Tarrasch
 
   
   
    
   
 
  

There is a possible deviation with 9...c4,
which is covered in detail in Grandmaster
Repertoire 10 – The Tarrasch Defence by
Aagaard and Ntirlis. To argue here over which
move is better – 9...c4 closing the position and
avoiding the isolated pawn, or playing with the
isolated pawn after 9...cxd4 – is not relevant.
My purpose and task here is to show how
strategy and practice have evolved over time,
especially in the Tarrasch Defence, rather than
to determine which move is strongest or has
the best computer evaluation.
I do not discuss here the subtle nuances of
the isolated pawn in the French Defence.
although comparing the isolated pawn in the
French Defence with other similar positions
would also be valuable. Karpov’s games are
a must for every player planning to advance
to the top – in particular his match against
Viktor Korchnoi in 1975, where the isolated
pawn in the French Defence was put to the
test. However, the isolated pawn positions
in the Tarrasch have historically much more
interesting material.
The Tarrasch Defence was introduced, as the
name suggests, by the German master Siegbert
Tarrasch, who was one of the top players of the
late 19th and early 20th centuries. His creation
was not well regarded by his contemporaries,
who did not fully appreciate the activity the
himself was undeterred. Quoting here from
Wikipedia: “Tarrasch continued to play his
opening while rejecting other variations of
the Queen’s Gambit, even to the point of
putting question marks on routine moves in
all variations except the Tarrasch (which he
awarded an exclamation mark) in his book Die
moderne Schachpartie.”
What was the understanding of these positions
at that time? The main idea was obvious. The
compensation for the isolated pawn was free
development of the pieces due to the open
space around the isolated pawn. Tarrasch
could not find many followers. Other players
saw it as an unnecessarily weakening strategy.
Later, after the Second World War, Paul Keres
noticed the hidden resources of the Tarrasch
Defence and included it in his repertoire as
a surprise weapon for certain moments – he
managed to use it in 1959 against Mikhail
Tal. Other professionals noticed his successful
attempt, and the debate for and against the
isolated pawn took off. The testing of the
 Chapter 2 – Evolution of the Isolated Pawn
defence saw its culmination in the 1969
World Championship match between Tigran
Petrosian and Boris Spassky.
During the 20th century many players have
used it in very important tournaments and
matches. Garry Kasparov and later Alexander
Grischuk, among others, have played it.
Certain players, on the other hand, such as
Karpov, have always played only against the
isolated pawn.
Before beginning our study, I would like to
explain my thoughts as a coach. I am not sure,
but I probably have a slightly different view
of chess coaching and how to train compared
with some other well-known authorities.
One such topic is the Soviet Chess School,
which people view in different ways. There is
a story about an Indian chess player visiting
Moscow who insisted on being shown the
school and would not accept that it never
existed physically in some building in Moscow.
Another viewpoint is that of the famous
emigre from the Soviet Union coaching in
New York City, who markets his methods as
from the Soviet school. A third opinion is that
both these views are wrong and that there is
no such thing as the Soviet school – it is just a
myth made up to explain why the Soviets were
so good at chess.
I grew up in the Soviet Union (in Tallinn in
Estonia) and I attended the so-called chess
school twice a week after my regular school.
My first coach, Tonu Truus had graduated
from the Moscow Sports Institute and worked
as a professional coach. He was not a strong
player himself, and very soon I was stronger
than him.
Once he organized a training camp, inviting
Vladimir Yurkov from Moscow. Yurkov was a
stronger player then Tonu, but he realized early
on that he lacked the ability to become a top
21
player himself and started his coaching career
very early. He was coach or second to a new
wave of future grandmasters: Yuri Razuvaev,
Yuri Balashov and World Championship
Candidate Andrei Sokolov were all influenced
by Yurkov. He also coached some top women
players, such as Nana Ioseliani.
Tonu met Yurkov in Moscow during his
studies and they became friends. In 1977 he
invited Yurkov to Tallinn, where he conducted
some serious private opening preparation for
me. It took only a week. Half of the time he
explained everything at the board and half of
the time I just wrote everything down. One of
the topics was the Tarrasch Defence. I was 15
years old at that time and my main weakness
was my lack of opening preparation. After that
session I used the Tarrasch Defence with success
in subsequent years, and most importantly
I learned how to handle the isolated pawn
structure. I could say that I received this
knowledge from the Soviet school.
In this chapter I have added a lot of games that
Yurkov did not mention in our session. I have
tried to cover the subject as fully as possible,
so that after reading it, the chess aficionado,
coach or avid player has some knowledge of
the material. But what kind of knowledge and
at what level?
First, chess is not like mathematics.
Memorizing a good idea or good advice in a
certain position is not enough to master chess.
The concrete approach is very tempting and
there are numerous books that explain nearly
everything in chess: how to win this or draw
that, or how to win with or against some line.
After many years of being around chess and
being one of the top players, now in my mature
years I think I may draw some conclusions.
There are some areas of chess that it is essential
to cover and learn by heart. One of these areas
22 Grandmaster Opening Preparation

is theoretical endgame positions. Furthermore, In the third position the pawn is on a5.
it is not enough only to learn them, but also
to practise and repeat them from time to

time. There is no excuse for forgetting some     
theoretical position. The understanding of     
these positions may be described at three levels.
    
The first level of knowledge is giving a simple     
answer, which might be draw, Black wins,
White wins, or you do not know. In the case of
    
the last of these answers, you cannot proceed     
to the second level.
Let’s look at these positions. The difference is
  
in the square that the a-pawn is on. In the first     
position the pawn is on a2. 
 Can White win these positions? What
     difference does the square that the white pawn
is on make? If you know the correct answer to
     these questions you have passed the first level.
     The second level question is also very simple:
how? If you pass this – showing on the board
     how it works – then you can try the third
     level. You must not only have the level of
understanding about the position and how
     to accomplish the win or draw, but you need
  to be capable of explaining or teaching it to
     others.
This is not easy. Once, in a training camp
 with youngsters from Curaçao, I explained
In the second position the pawn is on a4. it on the first day, but on the last day of the
 camp, when I asked them to explain it to me,
they could not. The learning or mastering of
     chess is a little more complicated process than
     just accumulating the facts.

     Moving on from certain endgame positions

     that you should study, there is a certain
    amount of opening theory that, depending
on your rating, you must know. As Kramnik
     puts it, when they ask him why he plays the
   Petroff Defence, he answers that it is because
other openings just do not work. Fortunately,
     there are still some other lines which do work,
 the last time I checked. Kramnik’s point is
 Chapter 2 – Evolution of the Isolated Pawn 23

more that some openings are good in an open Curt von Bardeleben – Siegbert Tarrasch
tournament, but in a world championship
match you need to have other lines. Learning Leipzig 1888
openings is a must. So why not work on them
in such a way that you gain the maximum 1.d4 d5 2.c4 e6 3.¤c3 c5
result? A professional chess player, or someone 
who wants to become one, must find the time
to study some lines in a very tedious way, like 
Botvinnik.  
The first step is to understand how the line
   
evolved historically. There is no need to go    
back as far as in our example with the Tarrasch
Defence, but you need at least to find the
   
turning points – the critical positions. It does     
not make sense to memorize a lot of lines and  
then to just discover (or even worse have your
opponent discover) some critical positions or  
lines that make the whole opening not to your 
liking anymore. According to the database available to me,
In the past there was Chess Informant, which this is the very first game in which Tarrasch
delivered the novelties and the important used the early counterattack against White’s
games on a regular basis. Certain players are centre.
able to rely only on their memory, and they use
other people’s work, following modern theory 4.cxd5 exd5 5.¥f4 ¤f6 6.¤f3 ¤c6 7.e3 c4!?
and lines and implementing them in their own 7...cxd4 8.¤xd4 ¥b4 9.¥e2 ¤e4 10.¤db5!
practice. This is the practical player: Mikhail 0–0 11.0–0 ¥xc3 12.¤xc3 ¤xc3 13.bxc3 ¥e6
Tal was one of the players who mastered this and Black had a solid but passive position in
approach. Psakhis – Ehlvest, Lvov 1984.
Botvinnik was the complete opposite: he was
the researcher. The difference between practical

players and researchers is a very grey area in  
reality. Still, I would recommend that the
player who has yet to become a grandmaster
 
should first try the researcher approach. It is    
like being a detective to discover the critical    
moments and changes, and to enter the minds
of the great masters of the past.    
   
Siegbert Tarrasch
  
Let’s start by looking at Tarrasch’s games and  
how practice made improvements to how to
play the isolated pawn positions.

8.¥e2 £b6?!
24 Grandmaster Opening Preparation

Tarrasch for some reason liked this early Teimour Radjabov – Pavel Eljanov
aggressive move a lot.
Astrakhan 2010
Better is 8...¥b4! 9.¤d2 0–0, and Black has a
very reasonable position. 1.d4 ¤f6 2.c4 e6 3.¤f3 d5 4.¤c3 ¥b4
5.cxd5 exd5 6.¥g5 h6 7.¥h4 c5 8.e3 c4!?
9.£c1 ¥b4 
And later Tarrasch lost the game.
 
In this game the main idea of the opening is   
not apparent; it is a little bit chaotic. Still I like
the move 7...c4!?, because in my game with
    
Psakhis I was not satisfied with the outcome. I    
had not studied Tarrasch’s games at that time.
Yes, I admit it – I did not myself study the
   
openings in the way I am recommending here.    
One excuse might be the lack of the ChessBase   
program and good trainers. Another might
be how much time was available. You need  
to allocate your time to the many opening 
positions you need to study. The success of the Eljanov, when commenting on this game,
player is measured in practical tournaments. wrote here: “A rather rare opening variation. I
If you end up finding the right solutions in learned it when Alexander Morozevich tried it
certain openings, it might be reflected in your successfully against... me in the Tal Memorial
final result, but it might not. (blitz).”
The balance between opening preparation The idea or concept belongs to Tarrasch in
and other chess-related activities is very my opinion. The move order and the position
important. To accumulate new ideas in is slightly different, because the white bishop is
different opening positions is very useful and on h4, not on f4, which I believe is in Black’s
is up the player himself. Very rarely is someone favour.
else going to do this for the player. The coach
or second is often not as strong as the player 9.¤d2 g5 10.¥g3 ¥f5
himself. Constantly comparing and thinking
about opening problems should make the

difference between becoming just a good   
player or becoming something extraordinary.   
Tarrasch’s idea of the early ...c5-c4 has     
been implemented in other lines. It might   
be coincidence, and in chess nobody can
copyright ideas; still, we can see the similarity.    
In the following game, the early ...c4 was used     
with success.
  
 

 Chapter 2 – Evolution of the Isolated Pawn 25

Black’s strategy is to control the light squares, 14.¥g3

and sooner or later he needs to take on c3. This move was an improvement on the earlier
This plan or strategy was lacking in Tarrasch’s game Gauglitz – Dizdar, Halle 1987, where
opening preparation. I am not criticizing White opted for 14.¥a3, after which 14...£a6
Tarrasch, just pointing out that he discovered gave Black a clear advantage. Radjabov’s move,
the idea of an early ...c4 and this is used today however, does not refute Black’s set-up.
in many similar positions.

11.¥e5   
As mentioned above, Eljanov had an earlier  
experience as White in this line. In Eljanov –
Morozevich, Moscow (blitz) 2008, he played     
the modest 11.¥e2, which does not promise   
any advantage for White, and he went on to
lose the game.
   
    
11...¥xc3
White’s last move might be difficult to
  
understand, but actually there is a simple  
trick. Black cannot play 11...¤bd7? because 
of 12.¥xf6 and Black cannot recapture with 14...£b2!
12...¤xf6 because of the check 13.£a4† and Eljanov’s move forces the queen swap, which
the bishop on b4 is lost. After 12...£xf6 Black is the safe approach.
would lose the pawn on d5.
The computer gives the more complicated
12.bxc3 ¤bd7 14...£a5 15.£c1 0–0 with the slightly better
 game for Black.
   15.£c1 £xc1† 16.¦xc1 b5
  And Black won this endgame.
  ...0–1

   The early advance of the pawn to c4 is not

   of course a position with an isolated pawn,
but instead a different, very ambitious and
  complicated plan.
 
  In the normal lines, Tarrasch was severely
punished by the top players of that time. In
 the next game his opponent used a system
13.¥d6 £b6 that is a popular weapon against the Tarrasch
As in the Tarrasch game the black queen for players who like a very small advantage
moves to b6. Very important here is that the without any risk.
bishop on f5 controls the b1-square.
26 Grandmaster Opening Preparation

Emanuel Lasker – Siegbert Tarrasch 

  
Duesseldorf/Munich 1908
 
1.d4 d5 2.c4 e6 3.¤c3 c5 4.¤f3 ¤c6    
4...cxd4 is much stronger, but our aim here
is not to point out some early mistakes. This
  
just shows us the lack of opening theory at that    
time.     
5.e3 ¤f6   
   
  
16...¤c4 17.¥a1 ¥b8 18.¦e1 £d6 19.g3
  ¦fd8 20.¤b1
   Lasker gives credit to Black’s position and
    starts regrouping. There is no specific idea
behind this regrouping, but if you know
    Lasker’s philosophy in equal positions then
    you can understand it. His concept was that in
equal positions one should just outmanoeuvre
   one’s opponent, steadily weakening his position
  by making use of the opponent’s mistakes if he
fails to adjust quickly to the new situation on
 the board.
6.a3!?
Lasker had some ideas! He was definitely a 20...¤b6?
researcher. Now the best for Black is to use the While White can play as he pleases, Black
same strategy and wait with 6...a6. must be more alert. This move is not good
positionally. The knight on b6 is never good
6...¥d6 7.dxc5 ¥xc5 8.b4 ¥d6 9.¥b2 0–0 and in a few moves it goes back to c4.
10.¦c1
Lasker is again following the same strategy: Better was 20...¤e4, when White has nothing
he postpones the development of his king’s better than to return his knight to c3: 21.¤c3
bishop to gain a tempo. ¤f6 It is again up to White to show how he is
going to improve his position.
10...a5 11.b5 ¤e5 12.cxd5 exd5 13.¥e2
¥e6 14.0–0 £e7 15.a4 ¦ac8 16.¤d4 21.¥c3 ¦e8 22.£b3!
White has the blockade, but the weaknesses This is stronger than 22.¥xa5 ¦xc1 23.£xc1
of the b4- and c4-squares must give Black a ¤xa4.
reasonable game.
22...¤c4
 Chapter 2 – Evolution of the Isolated Pawn 27

 26...¤xe3! 27.¦xc8 £xc8 28.b6 ¥xb6?

  After 28...¤xd1! 29.bxa7 ¤xf2 30.¢xf2
  £c5† 31.¢g2 £xa7 32.£xd5 £b6 33.¤d2
¦d8 34.¤c4 ¦xd5 35.¤xb6 ¦d2† White
    would have to be careful, though it is still a
   drawish endgame.
   29.fxe3 ¦xe3 30.£xb6 ¦xf3 31.£xa5 £c4
    32.£d2 f5 33.¦c1 £g4 34.£xd5 f4 35.¤d2
    ¦e3 36.¦f1 ¦e6 37.¦xf4 £d1† 38.¢g2
b6 39.£d7 £e2† 40.¦f2 £h5 41.¤f3 h6
    42.¤d4 ¦e5 43.£d8† ¢h7 44.£f8
 1–0
23.¤xe6!
Suddenly it becomes clear that Black is not Aron Nimzowitsch – Siegbert Tarrasch
controlling the central squares, and with this
St Petersburg 1914
simple trade White gains a big advantage.
This is a very well-known historical game and
23...£xe6 24.¥xf6 gxf6
you can find it in many chess books. I include it
24...£xf6 25.¥xc4 dxc4 26.¦xc4 is also
here because it is one of the first games in which
hopeless.
Tarrasch actually used his system, playing the
aggressive 2...c5, which commonly leads into
25.¦ed1?!
the Tarrasch Defence. This early attack usually
25.¤c3 allows 25...¤xe3 26.fxe3 £xe3†
leads to a position with an isolated pawn for
27.¢g2 ¥a7 with counterplay.
Black, and yes, sometimes, you need to defend
your weak pawn on d5, but this of course is
However, after 25.¥f3 White could expect to
not the idea behind the Tarrasch Defence.
win soon.
In my opinion, Tarrasch was an avid
attacking player and he just liked to attack
25...¥a7 26.¥f3?!
from the very first move. In the beginning
 you can attack the centre, and only after that,
  when you have open diagonals and files due to
your control over the centre, may you launch a
  successful attack on the king. How to achieve
    such a position using a sound opening is very
well demonstrated in this game.
  
   1.d4 d5
   There is a so-called classical approach in chess.
Players like Paul Keres and Anatoly Karpov
     were classical players. The classical approach
   usually means following standard mainstream
concepts that have been developed over time.

28 Grandmaster Opening Preparation
It was discovered long ago that the centre is
very important, and in practice everybody tries
to play towards the centre; it looks very weird
to start your game by pushing your a-pawn.
White can use the first-move advantage to
start building up the centre, so Black has two
basic choices. One of them is just to ignore this
and counterattack the centre later in the game,
while the other is the classical approach: to try
to do the same as your opponent. The problem
is that, because of the first-move advantage,
Black is forced at some point to defend the
centre. Still, at move one he can freely place
his pawn on a central square.
2.¤f3 c5 3.c4!
The correct counterattack. A passive
approach where the centre is concerned is
never good. Losing the initiative is one thing,
and one can never become successful in chess
by playing timidly in the opening.


 
    
   
   
   
 


3...e6
This is the situation mentioned earlier. Black
needs to defend his centre because of White’s
first-move advantage.
Opening the game with 3...dxc4 4.d5 or
3...cxd4 4.cxd5 £xd5 5.¤c3 are not good
options.
4.e3
White of course does not really need to
defend the centre and could just go for the
normal Tarrasch tabiya after 4.cxd5 exd5
5.¤c3.
4...¤f6 5.¥d3
This game was played more than a hundred
years ago, and some subtle opening nuances
were not yet known. I need to stress here that
the whole approach to the game of chess itself
was very different. There was little ready-made
material to study. The only criterion of truth
was practice, not computer power. Access to
the available chess information was still limited
and therefore the subtleties of move orders
were not so important – the man sitting at the
board was. This is why personality played a
much bigger role at that time.
White’s last move is not good because Black
could now transpose to a favourable version
of the Queen’s Gambit Accepted with 5...dxc4
6.¥xc4 a6.
5...¤c6 6.0–0 ¥d6 7.b3 0–0 8.¥b2 b6

 
  
  
   
   
 
  
 

9.¤bd2
Where to put your knight in this kind
of pawn structure is very important. The
knight on d2 is slightly passive: it works well
 Chapter 2 – Evolution of the Isolated Pawn 29

as a defensive piece, but from c3 the knight


can jump to a4 in many lines, attacking the    
c5-pawn. Again, these subtle differences had
not yet been worked out.
 
  
9...¥b7 10.¦c1 £e7 11.cxd5 exd5 12.¤h4   
g6 13.¤hf3 ¦ad8
   
   
      
    
   
    Black had achieved an excellent position
     from the opening and won in style.

  16.¥xc6 ¥xc6 17.£c2 ¤xd2 18.¤xd2 d4!

   19.exd4 ¥xh2† 20.¢xh2 £h4† 21.¢g1
   ¥xg2 22.f3 ¦fe8! 23.¤e4 £h1† 24.¢f2
¥xf1 25.d5 f5 26.£c3 £g2† 27.¢e3 ¦xe4†
 28.fxe4
14.dxc5?
While the 9th move was a little passive, this

aggressive move is definitely a mistake. This    
type of position was frequently employed in
tournament practice by Paul Keres, and later
   
by other modern grandmasters like Vladimir    
Malaniuk. The most famous historical game is   
that between Paul Keres and Vasily Smyslov in
Zurich 1953. The symmetrical pawn structure    
makes a mirror effect, but usually White has    
this position instead of Black in our game.
  
The plan here should be 14.¦e1! with the    
idea of playing ¥f1 and g2-g3, cementing the
kingside against possible attack on the b8-h2

In this position Black had a mate in three,
diagonal, and only after that should White but instead prolonged the game to give mate
start looking at how to attack Black’s central with the bishop. It was argued that he did this
pawns. This plan or idea is known from Black’s on purpose. As I mentioned before, you need
side, along with how to defend against the to destroy the man at the board, so this might
possible attack on the centre. be a true assumption.

14...bxc5 15.¥b5 ¤e4 28...f4†

30 Grandmaster Opening Preparation
28...£g3† 29.¢d2 £f2† 30.¢d1 £e2# was
the mate in three.
29.¢xf4 ¦f8† 30.¢e5 £h2† 31.¢e6 ¦e8†
32.¢d7 ¥b5#
Alexey Selezniev – Siegbert Tarrasch
Gothenburg 1920
1.d4 d5 2.¥f4
This innocent move, avoiding forcing lines
and leading to the London System, is popular
nowadays. Usually White starts with 2.¤f3,
but the text move sometimes allows him to
play f2-f4 after trading the bishops on d6.
During Tarrasch’s time, this was just a move
without any deep back-up.
2...c5 3.e3 £b6?!


 
    
   
    
    
 


A very interesting moment. As we saw in
previous examples, Tarrasch was an aggressive
player and he liked this early queen move in
similar positions. When we are dealing with
non-standard plans in openings you need to
check things very carefully, especially when
playing with the black pieces; the margin for
error is not very wide. With the white pieces
you have more freedom to try some so-called
interesting ideas without being punished, but
first you need to master and know the standard,
classical plans and ideas in the openings and
play accordingly; only after that should you
take the liberty of playing more interesting
stuff. I have seen many very talented players
who behave like spoiled kids, avoiding the
normal lines. Yes, some top grandmasters do
play interesting opening ideas, for example
players such as Alexander Morozevich and
Richard Rapport, but they also know how
to play the normal moves. The best way is to
balance things.
4.¤c3! e6 5.¦b1?
Too timid – a typical mistake in the opening
phase for many players. They sometimes force
themselves to find a good aggressive plan and
they make the move that is dictated by the
position, but then they pull back from it on
the next move.
As Kasparov put it when describing the
situation when playing against a computer,
there is always a moment at which you can
no longer play a safe, positional game, and
you have to switch to playing tactically. This
was exactly the situation in this game. White
punished his opponent for the too early
3...£b6 with a nice move, 4.¤c3, but then
he suddenly went back to defensive mode.
Instead 5.¤b5! was promising.
This is a typical situation when a weaker
player is facing a much stronger player. The
strategy of playing safe against a stronger
opponent works from time to time, and I
have heard from my GM friends the typical
complaint that the “patzer” dried out the game.
The patzer of course has the white pieces, and
there are unfortunately a lot of drawish lines
against nearly any opening.
There are two strategies against it. One
is to keep playing classical openings and to
try to outplay your opponent from an equal
position. This is usually the approach of the
Soviet Chess School, and the best exponent of
this style was Anatoly Karpov.
 Chapter 2 – Evolution of the Isolated Pawn 31

The other approach is to play some very


aggressive lines from the beginning of the   
opening phase. Players like Rapport and
Shabalov are the most aggressive players
 
around, and their opening choices are difficult    
to foresee. They also dare to play these lines   
against very strong opposition.
I mixed both styles during my career. In my    
opinion, the choice of opening or the style itself    
is not paramount to succeeding in the long run.
You just need to study the lines very deeply,

especially if they are not positionally very sound.  
Everybody has at some point been in a must-win
situation playing with the black pieces against a

13.¤g3 ¥g6 14.¦e1 ¤bd7 15.£c1 ¦fe8
very solid opponent. It is then that research into 16.£b2 ¤b6?!
not-so-correct opening lines may pay off. In the long run it is a good idea to transfer
 the knight to a4, but this is too early: the
 knight was controlling the e5-square, so it was
better to play 16...b6.
 
    17.¦e5 ¤fd7 18.¦xe8† ¦xe8 19.¤f1 ¤f6
20.¤e3 ¤e4 21.¥xe4!
    The typical drying-out move by someone
     who is playing for a draw.

     21...¥xe4 22.£a3 ¤a4

  
   
  
5...c4
Here we come. Tarrasch certainly    
remembered his earlier games, where he had    
used the same idea.
 
6.e4 ¥b4!     
Just in time – the position now resembles
something from the French Defence.
 
   
7.exd5 exd5 8.¥e2 ¤f6 9.¥f3 ¥xc3† 
10.bxc3 £c6 11.¤e2 0–0 12.0–0 ¥f5 23.f3 ¥g6 24.¤xd5!
Black has an excellent game, controlling the We see that the plan of transferring the
light squares and having a good blockade in knight to a4 has not worked. White was just in
the centre. time to attack the d5-pawn.
32 Grandmaster Opening Preparation

24...£xd5 25.£xa4 f6 26.£b5 £xb5 This I also do not like. The plan of how to
27.¦xb5 b6 play in similar positions was worked out
½–½ later by Mikenas and Keres.
14.¤xc6?!
Emanuel Lasker – Siegbert Tarrasch 14.£c1 ¥xb2 15.£xb2 offers a comfortable
advantage. The trade of bishops definitely
Berlin 1918 favours White.
14...bxc6 15.£c1 ¥xb2 16.£xb2 c5
1.d4 d5 2.¤f3 c5 3.c4 e6 4.cxd5 exd5 5.¤c3 Black had equalized in Bogoljubow –
¤c6 6.g3 ¤f6 7.¥g2 Tarrasch, Berlin 1920.

  8.0–0 0–0 9.dxc5 ¥xc5 10.¥g5 d4 11.¤e4
¥e7 12.¥xf6 ¥xf6 13.¦c1 ¦e8 14.¤e1
  
    
     
        
        
     
       
  
7...¥e7
In the following game, Tarrasch deviated   
from this normal move order: 
7...¥e6 8.0–0 ¥e7 9.dxc5 ¥xc5 10.b3 0–0 The plan of playing for the blockade on the
11.¥b2 £e7 12.¤b5 a6 light squares was clearly a surprise for Tarrasch,
12...¤e4 should be the automatic answer, and in the next few moves he just blundered.
and after 13.¤bd4 ¦fd8 Black is close to Lasker’s approach against Tarrasch was a stroke
equality. of genius. Against a super-aggressive player,
13.¤bd4 ¥a3 you want to dry out the dynamics. The most
 difficult thing in chess is to find the right plan.
   Tarrasch was more like a gambler, and he did
  not have enough patience to figure out the
plan (meaning where to put his pieces). He
  started the Tarrasch Defence, but he never
    found the correct plan. However, others picked
     up his idea of counterattacking the centre
   very early in the opening and developed it to
  perfection.
   14...¥f5

 Chapter 2 – Evolution of the Isolated Pawn 33

This impulsive move is not a mistake, but In the opening, everybody has had a situation
something goes wrong on the very next move. in which they have misplayed or forgotten
something – it does not matter which. We
14...¥e7 15.¤d3 ¥f5 16.£b3 £b6 and have all had lost positions just out of the
Black has a satisfactory position. His only opening. In this case you need to push yourself
concern might be that it is difficult to create to find some practical chances. The other thing
any counterplay; Black can hardly win this is that you cannot afford to think for too long.
position. This does not suit everybody, and Time trouble and a lost position are not a good
Lasker’s choice was psychologically very mix. Even Viktor Korchnoi, who was a famous
unpleasant for Tarrasch. time-trouble player, sometimes when finding
himself in a lost position started to play very
15.¤c5 fast.
 The last trick is to change the position
dramatically. This means to change the
  character of the position – the material balance
  if possible, or the dynamic of the position.
It does not make sense to try to defend a
    position a pawn down when your opponent
    has everything under control, and all possible
     future endgames promise no drawing chances.
This means that you can sacrifice another pawn
     without too much worry. In muddy waters
  you still might catch something.

   In this game, Tarrasch gave up too early.


15...¥c8? 16.¤xb7 ¥xb7
15...£b6 16.¤xb7 ¤e5 17.£b3 ¥e7 and 
suddenly the knight on b7 is misplaced. Black
has enough compensation.
 
 
The general advice is that even if you think
you have blundered something, it is not a
   
good idea to go into passive defensive mode.     
If the position is lost, an evaluation of minus     
2 or minus 4 is not relevant: it is still lost. One
should seek some counterplay, and not lose     
without a fight. It is very easy to recommend  
this, but in practice it is usually hard to make
a choice between two options that both look
  
terrible. My advice is just to calm down and 
try to figure out which one offers the best 17.¦xc6!
practical chances. Passive defence usually does Possibly Tarrasch just missed this move.
not work.
34 Grandmaster Opening Preparation
17...£a5 18.a3 ¦ab8 19.¦c2 ¥a6 20.¤d3
¥xd3?
It is tempting to go for a position with
opposite-coloured bishops when you are a
pawn down, but here this only helps White.
21.£xd3 ¦ec8 22.¦fc1 ¦xc2 23.¦xc2 h6
24.¥e4 ¦e8 25.b4 £e5 26.¥f3 ¥d8 27.¦c5
£e7 28.¦b5 £f6 29.¦f5 £b6 30.¦b5 £f6
31.a4 £d6 32.£c4 ¥f6 33.¥d5 ¦e7 34.¦c5
White has an extra pawn and a kingside
attack. A typical plan is to push the passed
pawn in conjunction with the attack.

   
   
    
   
  
    
   
    

34...¦d7 35.a5 g6 36.h4 ¥e5 37.¢g2 ¢g7
38.¥e4 £e7 39.¥c6 ¦d8 40.¥d5 £f6
41.¦c6 £f5 42.¦c5 £f6 43.¥f3 d3
Black seeks some counterplay, but it will not
be enough to save the game.
44.exd3 ¥d4 45.¦c6 £e5 46.¦c7 £f6 47.b5
g5 48.hxg5 hxg5 49.£a2 ¦h8 50.¥d5 ¦f8
51.b6 axb6 52.a6 g4 53.a7 £h6 54.¢f1
£d6 55.¦xf7† ¦xf7 56.¥xf7
1–0
Richard Reti – Siegbert Tarrasch
Bad Pistyan 1922
This game was played in 1922, by which time
Tarrasch could have had enough time to figure
out how to play the system. However, the game
shows only his weakness in understanding the
isolated pawn position. He allows White to
quickly obtain an ideal set-up where Black has
no counterplay. Which pieces to exchange and
Black’s ideal set-up remained a mystery to him.
We cannot really praise Tarrasch for his own
system. He just discovered it and had some
tactical ideas – mainly the aggressive early
attack with ...c7-c5 – but that was it. He was
not capable of developing it further. However,
other players studied and followed what he was
doing, and there was a long history ahead for
the Tarrasch Defence.
1.d4 d5 2.c4 e6 3.¤c3 c5 4.cxd5 exd5 5.¤f3
¤c6 6.g3 ¤f6 7.¥g2 ¥e6
Why Tarrasch chose this move order instead
of the more logical and normal 7...¥e7 is
difficult to understand.
8.0–0 ¥e7 9.dxc5 ¥xc5 10.¤a4 ¥e7 11.¥e3
Even in those days, Reti was most likely well
prepared for the game.

  
 
  
   
   
   
 
  

 Chapter 2 – Evolution of the Isolated Pawn 35

11...b6?? 15.£b3 0–0 16.¦fd1

This is not how you want to play the Tarrasch Now White’s pieces are all well placed, and
Defence. The correct plan is 11...¤e4 12.¦c1 Black has a difficult, passive defensive task
0–0 13.¤c5 ¥xc5! 14.¥xc5 ¦e8 15.¤d4 £f6 ahead.
16.¤xe6 £xe6! and the position with two
knights against two bishops is fine for Black.

This idea of trading the bishops for the knights    
was discovered and brought into practice in  
the 1960s.
   
A few months later, however, Tarrasch repeated    
this plan with ...b6 in an almost identical
position:
    
1.d4 d5 2.c4 e6 3.¤c3 c5 4.cxd5 exd5 5.¤f3    
¤c6 6.g3 ¤f6 7.¥g2 ¥e7 8.0–0 0–0 9.dxc5
¥xc5 10.¤a4 ¥e7 11.¥e3 b6
 
Tarrasch was somewhat stubborn and    
repeated his mistakes. From the game against 
Reti he did not draw the right conclusions. 16...¤e8 17.a4! ¥c5?! 18.¥xc5 d4 19.£a3
12.¤d4 ¤xd4 13.¥xd4 ¥b7 14.¦c1 £d7 bxc5 20.£xc5
15.¤c3 ¦fd8 White is winning.

   20...¥b3 21.¦d2 £c7 22.£xc7 ¤xc7
 
        
   
       
         
      
      
    
16.£d3
Stronger was 16.£b3 ¤e4 17.¦fd1.   
16...¤e4 17.¦cd1?! £e6     
Tarrasch succeeded in defending this position
in Teichmann – Tarrasch, Teplitz Schoenau 
1922, but from the opening perspective this 23.¤b5 ¤xb5 24.axb5 ¥d5 25.¦xd4 ¥xg2
was a clear disaster again. 26.¦xd8 ¦xd8 27.¢xg2 g6 28.¦xa7 ¦d2
29.b3 ¢g7 30.¦a4 ¦xe2 31.b6 ¦e7 32.¦b4
Returning to Reti – Tarrasch: ¦b7 33.¢f3 ¢f6 34.¢e4 ¢e6 35.¢d4 ¢d6
36.¢c4 ¢c6 37.¦b5 f6 38.¢b4
12.¤d4 ¤xd4 13.¥xd4 £d7 14.¤c3 ¦d8 1–0
36 Grandmaster Opening Preparation
Paul Keres
Paul Keres experimented with a lot of openings
in his early years. Throughout his career he
never had a long-term second. In the Soviet
Union the socialist system provided some help
to players, but this was not for Keres; he was
not a party man. Most of his consultations
took place between friends. Keres was one of
the most introverted personalities around and,
unfortunately, he did not write any books
about himself. Most of his work was about
pure chess, especially the endgame manuals.
My interest is in how he ended up playing the
Tarrasch Defence. I can only gather the data.
When I saw his game against Mikenas, I could
imagine some kind of scenario. The game was
played in a small Estonian resort town. These
summer tournaments were always friendly
affairs where some legendary happenings
were recorded. On one occasion, Keres, a
keen tennis player, had just finished his tennis
match and was asked who he was going to play
in that day’s round. He answered something
like, “Who cares?” He really did not care –
unlike Lev Psakhis in 1996, who, when asked
by someone before the round whom he was
going to play, answered that he could not
remember the name, but while preparing he
noticed that his opponent was very strong.
The opponent was yours truly, and we drew
our game that day.
Yes, Keres had some flaws in his opening
preparation. Botvinnik put it this way:
“...first was his slight uncertainty when he had
to navigate himself in the opening schemes.
He preferred, on the whole, obsolete opening
systems. That was why he had a taste for open
play.” This was a quote from 1947. It was
true, but Keres was definitely learning from
his mistakes. At the Absolute Championship
in 1941 he lost his game against Botvinnik
because of the opening, when he trusted an
earlier game played in the same line between
Mikenas and Botvinnik. He could not read
Russian and missed the improvement available
to Botvinnik, which had been played in a game
between Simagin and Belavenets.
Paul Keres – Mikhail Botvinnik
Leningrad/Moscow 1941
1.d4 ¤f6 2.c4 e6 3.¤c3 ¥b4 4.£c2 d5
5.cxd5 exd5 6.¥g5 h6 7.¥h4 c5 8.0–0–0?
Keres was repeating how Mikenas had played
against Botvinnik.

 
  
    
   
    
    

 

8...¥xc3!
The refutation, unknown to Keres, which
was first played by Belavenets against Simagin.
8...0–0? 9.dxc5 ¥xc3 10.£xc3 g5 11.¥g3
¤e4 12.£a3 ¥e6 13.f3 ¤xg3 14.hxg3 £f6
15.e3 ¦c8 16.¢b1 ¤d7 17.¤e2 ¦xc5 18.¤d4
was played in Mikenas – Botvinnik, Moscow
1940. White has a clear positional advantage
because the isolated pawn on d5 is blockaded
and Black does not have enough counterplay
to compensate for this.
9.£xc3 g5 10.¥g3 cxd4 11.£xd4 ¤c6
12.£a4 ¥f5
 Chapter 2 – Evolution of the Isolated Pawn 37

Black is already winning – a complete Paul Keres – Vladas Mikenas

opening disaster.
 Parnu 1955

   1.d4 d5 2.c4 e6 3.¤c3 c5 4.cxd5 exd5 5.¤f3

   ¤c6 6.g3
Mikenas was a Lithuanian born in Estonia;
    he moved to Lithuania in 1931. He was fluent
   in both languages and had a good relationship
with Keres. They were both representatives
    of a small country occupied by Soviets. The
     Tarrasch Defence was used by Mikenas on
  a few occasions, but he usually tried to play
like Tarrasch: playing some sidelines and for
  tricks.
 
13.e3 ¦c8 14.¥d3 £d7 15.¢b1 ¥xd3†
16.¦xd3 £f5 17.e4 ¤xe4 18.¢a1 0–0

19.¦d1 b5 20.£xb5 ¤d4 21.£d3 ¤c2†  
22.¢b1 ¤b4    
0–1
   
In 1947 Keres lost again to Botvinnik at the     
Chigorin Memorial in Moscow. This time
he could not handle Botvinnik’s Stonewall
   
system in the Dutch Defence. The next year,   
at The Hague/Moscow World Championship
tournament, he was also well behind in opening
 
preparation compared to Botvinnik, who 
easily won the event. There are controversial 6...c4
claims that most likely the winner was decided After 6...cxd4 7.¤xd4 Mikenas
in Moscow, and Keres was just forced to let experimented with both 7...£b6 and 7...¥c5
Botvinnik win. Keres’ opening play in that without success.
tournament against Botvinnik was very
dubious. Only the last game between the two, 7.¥g2 ¥b4 8.0–0 ¤ge7 9.e4! 0–0 10.exd5
when Botvinnik had already secured first place, ¤xd5 11.£c2!?
was won by Keres. Botvinnik was a researcher, This is a typically modest move by Keres;
and to compete with him one needed to do he liked small queen and rook moves.
one’s own research. Theoretically this is not best.

Keres landed on the Tarrasch Defence and 11...¥g4?

most likely the following game helped to hook 11...h6! is fine for Black.
him.
38 Grandmaster Opening Preparation

  
 
   
   
  
   
 
   

12.¤g5 ¤f6 13.¤d5 g6 14.£xc4 ¤xd5
15.¥xd5 £e7 16.¤e4 ¥h3 17.¥g5 £c7
18.¦fc1 ¦ac8 19.¥h6 ¦fd8 20.¥xf7† ¢h8
21.¥f4 £a5 22.d5
1–0
It is very common to obtain or borrow opening
ideas from your opponents. This is why, when
you are preparing against a certain opponent,
it is useful to check all his games with both
colours. Strong players notice new ideas used
by their opponents, and quickly start using
them themselves.
In our case, Keres most likely discussed and
analysed the Tarrasch Defence with Mikenas.
The above game was played in 1955, but
already in 1951 Keres had had an encounter
with the Tarrasch Defence in the same location
against Igor Bondarevsky.
Paul Keres – Igor Bondarevsky
Parnu 1951
1.d4 d5 2.c4 e6 3.¤c3 c5 4.cxd5 exd5 5.¤f3
¤c6 6.g3 ¤f6 7.¥g2 ¥e7 8.0–0
This training match game was played eight
years before Keres beat Mikhail Tal with the
Tarrasch at the Candidates tournament in
1959 – a long time by today’s standards, but
I am sure Keres became interested in the
Tarrasch Defence in those days when he faced
it against Bondarevsky.
The story starts here in the small Estonian
seaside resort town of Parnu and continues
until the match between Spassky and Petrosian
in 1969, when Spassky used the Tarrasch
Defence against Iron Tigran. He won the
match and became World Champion. One of
his long-time coaches and seconds was Igor
Bondarevsky.
Paul Keres rarely published or commentated
on his own games after 1940. He was a
very quiet man and I can only speculate as
to whether he helped Spassky to prepare
the Tarrasch or not. Spassky beat Keres in
their Candidates match in 1965, but later
lost the first World Championship match
to Petrosian. One thing is sure – Keres was
influenced by the potentially venomous
Tarrasch Defence and adopted it as his surprise
weapon.
8...0–0 9.¥g5 ¥e6
This move was the custom back then,
and only in later years did two other main
alternatives, 9...cxd4 10.¤xd4 h6 and 9...c4,
become more popular. The text move is just
too passive.

  
 
  
   
    
   
 
  

10.¦c1
 Chapter 2 – Evolution of the Isolated Pawn
10.dxc5 ¥xc5 11.¥xf6 £xf6 12.¤xd5 £xb2
13.¤c7 ¦ad8 14.£c1 £xc1 15.¦axc1 b6
16.¤xe6 fxe6 17.e3 has been played in many
games. In this endgame White has a risk-free
advantage, which might suit players who like
the style of Petrosian or Karpov. Of course,
neither Spassky nor Kasparov, when playing
the Tarrasch Defence against Petrosian and
Karpov respectively, allowed this endgame line.
10...cxd4
After this, the opening battle is lost for Black
and White has safe control over the central
squares, which guarantees a long-lasting
advantage.
11.¤xd4 h6 12.¥f4 ¤xd4 13.£xd4 £a5
14.¥d2 ¦fd8 15.¦fd1 £a6 16.¥f1 £c4

  
  
   
   
   
    
  
  

17.£xc4!
Transforming the advantage. It is now very
difficult for Black to defend his queenside
weaknesses.
17...dxc4 18.¥g2± ¦ac8 19.¥e3 b6 20.¤b5
¦xd1† 21.¦xd1 a6 22.¤d4 ¤g4 23.¥f4 ¥c5
24.¤xe6 fxe6 25.¥h3! h5 26.e3 ¦f8 27.¦c1
b5 28.b3 ¥a3 29.¦c2 ¦c8 30.f3 ¢h7
39

   
   
  
  
  
 
  
    

31.bxc4
But not 31.fxg4?? cxb3 32.¦xc8 b2 and
suddenly Black is winning.
31...g5 32.¥xg5 ¦g8 33.¥f4 e5 34.¥xg4
hxg4 35.¥xe5 ¦c8 36.fxg4 bxc4 37.¢f2 ¥b4
38.¥c3 a5 39.¢f3 ¢g6 40.h4 ¢f7 41.¢e4
¢e6 42.¢d4 ¥d6 43.¥xa5 ¥e5† 44.¢e4
¥xg3 45.h5 ¥h4 46.¥c3 ¥g5 47.¦b2 ¦e8
48.¦b5 ¥f6 49.¥xf6
1–0
I do not have any information about what kind
of relationship the players had. As they played
the training match, they must have had some
kind of common interest – was it the Tarrasch
Defence? How deeply they worked together
on the Tarrasch Defence remains unknown,
but the next game, played by Bondarevsky
later that year, definitely did not go unnoticed
by Keres.
Nikolay Novotelnov – Igor Bondarevsky
Moscow 1951
1.c4 e6 2.¤c3 d5 3.d4 c5 4.cxd5 exd5 5.¤f3
¤c6 6.g3 ¤f6 7.¥g2 ¥e7 8.0–0 0–0 9.¥g5
¥e6 10.dxc5 ¥xc5 11.¦c1 ¥b6 12.¤a4
I think this game was the prelude to Keres’
spectacular win over Mikhail Tal in 1959.
40 Grandmaster Opening Preparation

A very similar idea or plan is demonstrated in
this game.
12...d4
Here I do not want to argue with any chess
aficionado who blindly follows the computer
evaluation. The whole concept was developed
here – it is not dangerous to give up your
bishop in positions with an isolated pawn.
It even favours the knight over the bishop in
many situations. This is why it is better to have
the isolated pawn when you have knight versus
bishop, because with a symmetrical pawn
structure the bishop is usually much better
than the knight.
Black has an aesthetically very beautiful
position. The bishop on f4 has no job and
Black’s space advantage and control over the
light squares are overwhelming.
18.¦c2 £d5† 19.f3 ¦e7 20.¤d3 ¦ae8
21.¥c1 £b3 22.¦e1 ¤d5 23.¤f2 ¤e3†
This active move is not the best – we do not
want to exchange the passive bishop on c1.
23...¦e6 24.£d3 £a4 was much stronger
and White needs to stay put with moves like
25.£d1.
24.¥xe3 ¦xe3 25.¦d2 £b5 26.¤d3 ¤a5
27.¢f2 ¤c4 28.¦c2

 
     
    
       
        
       
      
    
      
 
28...¦3e7 29.£c1 ¤e3 30.¦c8 £h5 31.¢g1
13.¤xb6 axb6 14.a3 h6 15.¥f4 ¥d5 16.¤e1
£h3 32.¦xe8† ¦xe8 33.¤f4 £d7 34.£d2
¥xg2 17.¢xg2 ¦e8
g5! 35.¤g2 ¤c4 36.£c2 £e6 37.¢f2 b5
 38.¦d1 £h3 39.¢g1?
  White blunders just before the time control.

   39...£xg2†

    0–1

     Keres’ opening repertoire was not very wide

     throughout his career. The Queen’s Indian and
Nimzo-Indian were his main choices as Black
     in the closed systems. However, he prepared
   the Tarrasch Defence and used it as a surprise
   weapon in very important tournament
situations.

 Chapter 2 – Evolution of the Isolated Pawn 41

Tigran Petrosian – Paul Keres 13...¥b6 14.¤xd5 ¦c8 15.e3

Petrosian wrote here: “This game played an
Amsterdam/Leeuwarden 1956 important role: in the event of a win I would
have gained good chances of second place in
1.d4 d5 2.c4 e6 3.¤f3 c5 4.cxd5 exd5 5.g3 the tournament.”
¤c6 6.¥g2 ¤f6 7.0–0 ¥e7 8.¤c3 0–0
This game was played in the last round of 15.¤xb6?! £xb6 16.£d4 ¦c2 gives Black
the tournament. Vasily Smyslov had already counterplay.
secured first place. Keres needed a draw to
secure second place, which was useless in itself, 
but did place him in chess history as forever   
second. Chess historians have always argued
that most of the time the Soviets made some
 
pre-tournament and pre-game arrangements,    
and it was no secret that their trump card
for this tournament was Smyslov. Was this
   
last round game a prearranged draw or not?    
It is difficult to say. Keres picked his surprise    
opening line for some reason.
  
9.¥e3 ¥e6?!   
The text move is too passive, and 9...¤g4 is
also a shot in the air.

15...¥xd5 16.¥xd5
9...c4 is the best move here.
Petrosian wrote here: “Now further loss of
 material is inevitable for Black. On 16...¤xf2
   White simply takes with the king. Naturally,
I declined the draw offered here by Keres.”
  The reason for Keres’s play may be that he just
   wanted to finish the tournament as quickly as
    possible.

     16...¦c7 17.£f3?
    It is rather strange that White missed the
simple win with 17.¥xf7† ¦cxf7 18.£xg4,
  and now Black cannot play 18...¦xf2 because
   after 19.¦xf2 ¥xe3 there is a crucial check
20.£e6† and White wins.

10.dxc5 ¤g4 11.¥d4 ¤xd4 12.¤xd4 ¥xc5 17...¤e5 18.£g2 £e7 19.¤d4 ¦d8 20.¦ad1
13.¤b3 g6 21.b3 ¦cd7 22.e4 ¦c8 23.¢h1
This double attack wins a pawn. It looks A draw offer from Keres was now accepted.
strange that Keres has misplayed the opening Petrosian had a very good sense of danger and
so badly. On the other hand, it did not matter I am sure he realized that Black’s initiative after
much – the tournament was effectively over. the simple 23...¥xd4 24.¦xd4 ¦c2 would put
42 Grandmaster Opening Preparation
White in serious trouble. Keres, as I mentioned
before, did not care anymore.

  

   
   
   
   
  
 

½–½
Nowadays, players vary their opening choices
every month, but at that time, work done one
day might only be used a few years later. This
probably happened with Keres. He waited
a long time to show his homework and the
chance came in 1959.
Mikhail Tal – Paul Keres
Bled/Zagreb/Belgrade 1959
1.¤f3
This game was played towards the end of the
Candidates tournament when it seemed clear
that Tal was going to be the sole winner. Keres
still had some theoretical chances, but first he
had to win this game.
Keres was a classical player whose opening
repertoire consisted mainly of the solid Queen’s
Indian Defence and the Ruy Lopez. He never
fancied the King’s Indian. However, the solid
Queen’s Indian is no good if you need to win.
The Tarrasch was a good choice against Tal.
In virtually any opening White can force some
very dry, equal line. However, in the Tarrasch
Defence, while the pawn structure is fixed
in the opening, it remains dynamic. White
cannot claim a drawish position from the
opening.
At the beginning of the 20th century some
players treated the Tarrasch Defence as a
slightly inferior, but solid, defence to play
for a draw. Their argument was that it is very
difficult for White to win the isolated pawn.
They accepted passive defence and the hidden
dynamic resources of the Tarrasch Defence
were never revealed.
Keres, of course, was an outstanding
player, and in this game he uncorked one of
the greatest positional ideas concerning the
isolated pawn. The game itself is very pleasing
for its aesthetic qualities and was nominated as
the best game of the tournament.
My focus, however, is on the positional
approach to the game. Unfortunately, I do
not have any information about how much
time Keres spent on his 11th move. The move
itself is not the computer’s first choice, but
Keres could play it because of his tournament
situation. Did he prepare it at home? Keres did
comment on the move, but did not answer
these questions. One thing is clear – he played
an outstanding game and without the plan
starting at move 11, this would never have
happened.
1...d5 2.d4 c5 3.c4 e6 4.cxd5 exd5 5.g3 ¤c6
6.¥g2 ¤f6 7.0–0 ¥e7 8.¤c3 0–0 9.¥g5

 
 
   
   
    
   
 
  

 Chapter 2 – Evolution of the Isolated Pawn
9...¥e6
This move was frequently played at the time.
The more modern approach of 9...cxd4
10.¤xd4 h6 11.¥e3 was probably developed
by Spassky’s second and coach Bondarevsky.
However, it is a little counterintuitive, because
from Black’s perspective we would prefer not
to take on d4.

 
  
   
   
    
    
 
  

This became the tabiya during the Spassky –
Petrosian match in 1969. Here White has full
control over the blockading d4-square. Even
so, Black has a lot of play, and this position has
been tested in several high-level games.
10.dxc5 ¥xc5 11.¤a4!?
Tal was a fighter, and if you are leading the
tournament with a 2½ point margin with five
rounds to go, you can behave like a gentleman.
After 11.¥xf6 £xf6 12.¤xd5 £xb2 13.¤c7
¦ad8 14.£c1 £xc1 15.¦axc1 b6 16.¤xe6 fxe6
17.e3 White can hardly lose. This has caused
most players to abandon the passive 9...¥e6 in
favour of the more dynamic 9...cxd4.
In 1950s the idea of squeezing in chess – when
someone tries to win a slightly better endgame
– was not a very popular approach. Chess
professionalism started later – some may argue
that it started with Bobby Fischer. He was
probably the first one-hundred-percent chess
43
player around. The academic style was, in my
opinion, developed by Petrosian, who liked to
play risk-free positions. Then came Karpov,
whose endgame technique and especially the
ability to use every small chance, made him
one of the best tournament players ever. The
difference was clear: the older generation were
gentlemen who did not care to play these kinds
of endgame positions.
Of course, all this had a tremendous impact
on opening choices. Suddenly the equal
endgames which arise from some openings did
not look so drawish anymore. I can cite one
of Kasparov’s seconds during his match against
Karpov. Supposedly Kasparov was demanding
that the analyses of some opening lines should
not end with a conclusion that it’s a drawish
endgame, but should be analysed to the very
end.
11...¥b6

  
 
  
   
   
   
 
  

This was the bombshell! As I have already
mentioned, this is not the computer’s first
choice, and if you show it as a good move
to your students, they might doubt your
competence. The move is a move like any other
chess move; the concept, however, was new.
I am sure that the great masters from the
past already knew circumstances when the
knight is better than a bishop. All this was
44 Grandmaster Opening Preparation
very clear when we were talking about closed
positions. However, nobody had defined it
clearly when we were talking about positions
with an isolated pawn. The Tarrasch Defence
did not have a lot of followers. Positions with
an isolated pawn from the French Defence
were also very rare guests in tournament chess.
There were not enough games to look at.
Of course, masters were aware that the knight
is a better blockader and the well-known pawn
structure from the Queen’s Gambit Declined
occurred from time to time in tournament
games. All this was relevant from White’s side.
Keres managed to open a new way of thinking
about how to handle the pieces in the Tarrasch
Defence and this game had a huge impact on
the development of these ideas in the 1960s.
Soviet masters picked it up and players like
Anatoly Lein and (of course) Spassky did a fine
job. I am sure that if Keres had not played this
game, the Tarrasch Defence would not have
had such a boost.
When talking about the great players and
their input to chess history, in my opinion
this was the moment for Keres. He was a
romantic player at the beginning of his career
in the 1930s, culminating in winning the
AVRO tournament in 1938 – the tournament
where all the best players were participating.
However, he never became World Champion
and is regarded as the eternal second-best.
These are his accomplishments in chess from
the competitive aspect.
There is an opening line in the Sicilian that
bears his name and his research in endgames is
also very well known. But for some reason the
full impact of his game against Tal in the 1959
Candidates tournament is not fully recognized
as discovering a new method of handling
positions with an isolated pawn.
12.¤xb6 axb6 13.¤d4 h6 14.¥f4
Tal did not go for 14.¥xf6 £xf6 15.e3, with
a drawish position. If you want to become
World Champion, you cannot play chess like
that.
14...£d7

  
 
  
   
    
    
 
  

15.a3?!
White should play: 15.¤xe6! £xe6
(15...fxe6 16.e4! ¦fd8 17.exd5 exd5 18.£d3
and White has a big advantage, because the
bishops are better in the open position) 16.a3
¦fe8 17.¦e1 d4 (17...b5!?) 18.¦c1 g5! 19.¥d2
¦ad8 Black has more space in the centre, where
the two knights are better than the bishops.
The position is roughly equal.
15...¥h3
Now Black has full control over the light
squares in the centre. The white bishop on f4
is also without any role.
16.£d3 ¦fe8 17.¦fe1 ¥xg2 18.¢xg2 ¦e4
19.¤f3 ¦ae8 20.¥d2 d4
Now the best chance for White was passive
defence. The usual strategy when you control
the centre is to start an attack on the king,
but here it is not easy for Black to push the
kingside pawns because the white knight is
controlling the h4-square.
 Chapter 2 – Evolution of the Isolated Pawn 45

 22...g5! is a computer suggestion and it is

   strong. The idea is to kick the white knight
  from f3 and to take on d4 with the knight. But
this is too complicated for a human to find,
    because you must foresee a crucial tactic on
     the 24th move: 23.h3 h5 24.¥xg5 (24.¢g1 g4
25.hxg4 hxg4 26.¤h4 ¤xd4) 24...¤e5!!
    
     
     
         
   
21.e3?    
21.¦ac1 £e6 22.b4! defends the e2-pawn  
with a counterattack. Now Black cannot take    
the pawn: 22...¦xe2? (22...£d5! and Black still     
has unpleasant pressure) 23.b5 ¤a5 24.£xe2
£xe2 25.¦xe2 ¦xe2 26.¥xa5 bxa5 27.¤xd4

25.dxe5 £xd3 26.¥xf6 ¦xe1 27.¦xe1 ¦c8
With an equal endgame.
Black should win in the long run.
21...£d5!
23.¦xe8† ¤xe8 24.£e2 ¤d6
Now the situation becomes critical for
White. 
22.exd4
   
22.¢g1 ¤g4 23.exd4 ¦xe1† 24.¦xe1   
    
      
       
   
       
      
       
     
     This position is much better for Black,
 and in a practical game it is very difficult to
24...¤ge5! and Black is winning. defend with the white pieces. White has no
pawn weaknesses, but Black’s dominance
22...¦xd4 in the centre is so important that White
Keres makes a simple move where he is sure needs to defend very carefully not to lose
to have some advantage. immediately.
46 Grandmaster Opening Preparation
Nowadays we can quickly check the
evaluation of any position with a computer,
but back then you needed to rely on your
understanding. Keres was one of the best
players in open positions, which is why he had
good results against Tal. Keres did not believe
Tal’s bluffs and refuted them with his perfect
calculation. In this position, Tal does not even
have the chance to bluff.
25.¥e3 ¦d3 26.¢g1 ¤c4 27.¤e1 ¦b3
28.¦c1 ¤xe3 29.fxe3 £e5
29...£e4! was more precise, but in time
trouble Keres wins a pawn, which is a very
understandable practical solution.
30.¤g2
30.¤f3 would have given White more
chances to defend in the endgame, but Tal
does not like to suffer.
30...¦xb2 31.£d3 £e6 32.¤f4 ¦b3 33.¦c3
¦xc3 34.£xc3 £e4
The position remains the same: Black still
dominates in the centre, but now he also has
an extra pawn.

   
  
   
    
   
    
    
    

35.£b3 b5 36.£xb5 £xe3† 37.¢f1 £f3†
38.¢g1 £e3† 39.¢f1 g5 40.¤e2 ¤e5
41.£xb7 ¤d3 42.£c8† ¢g7 43.£f5 £d2
44.¤d4 £e1† 45.¢g2 £e3!
Queen and knight is the most powerful
combination of pieces, especially when
attacking the king. Keres keeps playing at the
highest level in this part of the game as well.

    
   
    
   
    
   
   
    

46.£d5 £f2† 47.¢h3 £f1† 48.¢g4 ¤f2†
49.¢f5 £d3† 50.¢e5 ¤g4† 51.¢d6
£xa3† 52.¢c7 £e7† 53.¢c8 ¤e3 54.£b5
£e4 55.£b2 ¢g6 56.£b6† f6 57.¤e6
¤c4 58.£a6 ¤e5 59.¤c7 £c2 60.£d6
£xh2 61.¤d5 £f2 62.¢b7 £xg3 63.£xf6†
¢h5 64.£e6 ¤g4 65.¤e7 £f3† 66.¢c8
¢h4 67.¤f5† ¢h3 68.¢d8 h5 69.£g6
¤e5 70.£e6 ¤g4 71.£g6 ¤e5 72.£e6
£d3† 73.¤d4† ¤g4 74.£d5 ¤f2 75.¢c8
h4 76.£e5 £e4 77.£f6 £f4 78.¤f5 ¤e4
79.£e6 £g4
Finally Tal resigned.
0–1
Leonid Stein – Paul Keres
Moscow 1966
Keres did not have any more real chances to
become World Champion after the 1959
Candidates. This, however, did not mean
that he quit working on his chess. He used
the Tarrasch Defence until his sudden death
in 1975. He did not lose a single game in the
Tarrasch and used it mostly just to make a
draw. This next game is a typical example.
 Chapter 2 – Evolution of the Isolated Pawn
1.c4 ¤f6 2.¤c3 e6 3.¤f3 c5 4.g3 d5 5.cxd5
exd5 6.d4 ¤c6 7.¥g2 ¥e7 8.0–0 0–0 9.dxc5
¥xc5 10.¤a4 ¥b6
Against a young and ambitious Ukrainian
player, Keres repeats the famous line. His
opponent was undoubtedly very well aware
of his game against Tal. Stein played more
carefully, but could not create real problems.
11.b3 ¥f5 12.¥b2 ¥e4

  
 
   
   
  
  
 
  

Usually the bishop is not good in the centre.
It needs a helping hand, which is usually a
pawn to protect it – in our position the d5-
pawn. Anyway, it is not the normal way of
handling the bishop problem. Our next hero
of the Tarrasch Defence – Anatoly Lein – was
undoubtedly following this game and making
some notes.
13.¤xb6 axb6 14.£d2 £e7 15.¦fc1 ¦fe8
16.£f4
If you try to guess the next move, you may
be astonished.
16...h6!
This move is also the computer’s first choice,
in contrast to when Keres played the surprising
11...¥b6 against Tal!
After 16...¤d7 17.£c7 ¦ab8 18.¦d1 ¦ec8
47
19.£f4 ¦e8 20.¥h3 White has a strong
initiative, because the position remains
relatively open for his bishops.
17.¥f1
A similar plan is often used in Queen’s
Indian-type positions when White tries to
fight for the e4-square, but without allowing
the trade of bishops. However, now the bomb
explodes again. This was definitely a moment
that the Soviet masters in Moscow did not
miss.

 
  
   
   
   
  
  
   

17...¥xf3! 18.exf3!
18.£xf3?! is weaker, because of 18...d4
19.¦d1 ¦ad8 20.¦ac1 £e6, and despite the
fact that White has two bishops against two
knights, Black has firm control over the central
squares. It is very difficult for a computer to
evaluate these positions, because the doubled
pawns and the two bishops should give a huge
advantage for White, but in reality White is
only slightly better.
18...¤h7
This was not the best move – it is too slow.
Better was: 18...d4 19.¦d1 g5!? (19...¤d5
20.£d2 ¦ad8 21.f4 ¤c7 22.¦e1 £d6 23.a3
and White has the better position because
Black can only wait and defend his weak
48 Grandmaster Opening Preparation

pawn on d4) 20.£d2 ¦ad8 21.¦e1 (21.¥b5 26.¦d3 ¤a7 27.¥a4 b5 28.¥b3 ¤c6
£c5 22.a4? ¤e5 and suddenly Black is much 29.¦c1 b6 30.¦e1 ¤f6 31.¥d1 ¤d5 32.¥f3
better) 21...£d6 22.¦xe8† ¤xe8 23.¦e1 ¤g7 ¢f8 33.¦c1 ¤de7 34.¢f1 ¦8d7 35.h4 ¤d8
24.¢g2 36.h5 f5 37.¦e1 ¤dc6 38.¦ed1 ¢f7 39.¥e2
 ¦d5
    
       
      
    
        
   
       
       
    
White is better. This is easy to assess with
computer help, but in a real game White   
would need to show great patience. 
The computer overestimates White’s
19.¥b5 position. This is because even winning the
19.h4!, restricting Black’s knight, was more d-pawn would not promise a lot of winning
precise. chances due to the reduced material. The
computer “thinks” incorrectly that at
19...¤g5 20.£g4 £e6 21.£xe6 ¦xe6 22.f4 minimum it might reach a good rook endgame
¤e4 23.¦d1 ¦d6 24.a3 ¦ad8 25.b4 d4 a pawn up. However, we humans know that
 such a rook ending may be a dead draw,
    though it might take a while for a computer to
understand this.
  
    40.¦3d2 ¦d8 41.¥f3 ¦5d6 42.¥e2 ¦d5
43.¦a1
    ½–½
   
Anatoly Lein
    
     Yuri Anikaev – Anatoly Lein
    Grozny 1968

The endgame is much better for White, but 1.d4 d5 2.c4 e6 3.¤c3 c5 4.cxd5 exd5 5.¤f3
much better in the endgame can often just ¤c6 6.g3 ¤f6 7.¥g2 ¥e7 8.0–0 0–0 9.dxc5
mean a draw. In this game Keres defended the ¥xc5 10.¤a4 ¥e7 11.¥e3 ¥g4
position without too many problems.
 Chapter 2 – Evolution of the Isolated Pawn 49

 the time. 18.¥xf3 ¤e4 19.¤f4 d4 20.£a4

   ¦cd8 21.a3 £f6 22.¤d3 With an unpleasant
  position for Black.

    b) 17...¥f5! 18.¤f4 ¥e4 19.£d2

    
    
      
   
     
      
   
12.¥c5  
This plan is not dangerous for Black. White     
does not have enough time to complete his
plan, which is to trade all of his bad pieces.

Now the typical move when you control the
centre, or you are trying to take control over it,
12...¦e8 13.¦c1 ¤e4
is in order. One should not be afraid to weaken
This is not the best move order to counter
the king’s position: 19...g5! 20.¤d3 ¥f5 With
White’s plan.
a balanced position.
More accurate is 13...¥xc5 14.¤xc5 £e7
14.¥xe7 £xe7 15.h3
15.¦e1 ¦ac8 16.¤d3 h6 17.h3 and now:
15.¤d4!? is why the move order chosen by
 Lein was not the most precise. Now Black still
  has the bad bishop. Even so, after 15...¤xd4
   16.£xd4 ¥xe2 17.¦fe1 £f6 18.£xf6 ¤xf6
    19.¤c3 ¥c4 20.¦ed1 (20.b3 ¦xe1† 21.¦xe1
    d4=) 20...¦ad8 21.b3 ¥a6 22.¤xd5 ¤xd5
23.¦xd5 ¦xd5 24.¥xd5 ¢f8 the endgame is
    drawish.
 
  
     
  
a) 17...¥xf3? It is not good to give up the bishop
now. Why? Because there is no blockade on
   
d4 and White’s other knight is heading to f4.    
After that, Black is forced to move the pawn  
to d4 and White has an advantage on the light
squares. It’s not enough to blindly follow the   
standard plans in a complicated positional  
game – you need to make adjustments all
  

50 Grandmaster Opening Preparation

15...¥xf3!? 16...¦ad8 17.¥g2

I am sure the Moscow masters at that time 
were keenly following Keres’ games in the
Tarrasch Defence and had some very good   
ideas about how to play it. Now, however,  
Anikaev makes a grave mistake. As Stein had
already demonstrated in these positions, one
   
should seriously consider recapturing on f3    
with the pawn.   
16.¥xf3?    
We are lucky that Anikaev made this  
positional mistake as we can enjoy the
masterpiece demonstrated by Lein.   

16.exf3! is much stronger, and suddenly Black stands well in the centre. The bishop
White has an advantage. Psychologically it is on g2 is passive, because the knight on
a difficult decision, because nobody wants to e4 dominates the bishop. The d5-pawn is
double their pawns voluntarily. 16...¤f6 17.f4 protecting the knight on e4, and if White takes
 on e4 with his bishop or knight the d-pawn
  advances to e4 and the space advantage grows.
Now Black follows the classical route – you
  can successfully launch a kingside attack only
    when you control the centre.
   
    17...h5! 18.h4 g6 19.¤c3 £e5!
Centralization.
   
   20.e3 ¢g7!
   Improving the position also includes the
 king.
White has a dynamic advantage, meaning
that he has more control over the position 21.£a4 ¦e7 22.¦cd1
and Black lacks active counterplay, but even 
after winning the isolated pawn (or perhaps     
we should say passed pawn), it might be still
difficult to win the game. Nowadays, when we   
rely on computer chess, every novice player will   
see the computer evaluation here and happily
announce to everybody that White is better.
  
Back then, however, it was difficult to prove   
this. I am sure the verdict was that White has
enough compensation for the ruined pawn
    
structure, but not more.   
  

 Chapter 2 – Evolution of the Isolated Pawn 51

22...f5! This was the idea of Black’s last move, but in

Avoiding simplification. White’s position time trouble both players missed some tactics.
would be fine without rooks or queens.
Right now the passive rooks and Black’s space 27...¤e5 28.¦fd1 ¦f7 29.£b2 ¢h7 30.£d4
advantage make White’s play difficult. Yes, and White just needs to wait to see whether
computers still give the evaluation as roughly Black can find a way to break through. The
equal, but the initiative is fully in Black’s computers still give a lot of zeros, but in
hands. practice, in mutual time trouble, White’s task
is very difficult.
23.¤xe4?!
23.£a3! and the position is still balanced:

23...d4 24.exd4 ¦xd4 25.¤xe4 fxe4 26.£e3     
¦a4 27.¦d2 ¤d4 28.¦e1 ¤f5 White still    
needs to be very careful not to allow the ...e3
breakthrough. 29.£b3! ¦d4 This computer   
line gives zeros in its evaluation chart. This is  
the chess of the 21st century, but one should
only travel so far after carefully following the
   
classical route developed decades ago by the     
great masters.
  
23...fxe4 24.b4    
Desperately seeking counterplay. 
28.£xf5?
24...a6 25.b5 axb5 26.£xb5 28.£b2! £f6 29.exd4 ¤xd4 30.¦e1 ¤f3†
 31.¥xf3 ¦xd2 32.£xd2 exf3 33.¦xe7† £xe7
     34.£c3† £f6 35.£c7† £f7 with a draw.

    28...gxf5 29.exd4 ¤xd4 30.¦b2 b5

   
      
        
         
   o o
     M 
   +  
26...£f5!
I like this move a lot. Only really strong
R + 
players can execute such a move.    

27.¦d2 d4?
52 Grandmaster Opening Preparation

Black is still controlling the central squares


and White needs to find good moves to save     
the game.     
31.¦c1 ¢f6 32.¥f1?     
Wanting to activate the bishop is  
understandable, but this was not the moment
to do it.    
     
       
       
     
36...f4
  Another breakthrough!
   
     37.¥xb5 fxg3† 38.¢g1 ¦xh4 39.¦gb2 ¦he4
40.¥d7 ¤e2† 41.¢g2 ¤f4† 42.¢f1 ¦f3†
    43.¢g1 ¤e2†
    0–1

 Lev Polugaevsky – Anatoly Lein

32...e3!
The space advantage is visually attractive, Alma-Ata 1968
but to win the game you need to create
weaknesses in your opponent’s camp. After 1.c4 e6 2.¤f3 d5 3.g3 ¤f6 4.¥g2 c5 5.0–0
this breakthrough White’s position becomes ¤c6 6.cxd5 exd5 7.d4 ¥e7 8.¤c3 0–0
critical. 9.dxc5 ¥xc5 10.¤a4 ¥e7 11.¥e3 ¦e8
12.¦c1 ¥g4 13.h3
33.fxe3 ¦xe3 34.¢h2 ¦g8 
The black knight on d4 dominates White’s
rooks and the weakness of the g3-pawn decides  
the game in a few moves.  
35.¦g2 ¦g4 36.¦b1
   
   
  
  
 
  

13...¥xf3! 14.¥xf3 £d7 15.¥g2 ¦ad8
 Chapter 2 – Evolution of the Isolated Pawn 53

16.¤c5 avoid the Queen’s Gambit Accepted, among


  

   
   
    
   
 
  

16...¥xc5!
By exchanging this bishop, Black establishes
his knight in the centre.
17.¥xc5 ¤e4 18.e3 £f5 19.¥d4 ¤g5
20.¢h2 ¤e6 21.¥c3 d4
Usually we want to keep the position closed
when our opponent has the bishops. Here
the situation is different, because opening the
position makes the dominating knight on e4
with the active rooks on the d- and e-files so
dangerous for White – suddenly his queen is
in danger of being trapped.
22.exd4 ¤exd4 23.¥xd4 ¤xd4
And a draw was agreed.
½–½
Mikhail Podgaets – Anatoly Lein
Ivano-Frankivsk 1971
1.c4
In the opening there are many different move
orders that finally end up in the same position.
Against a player whose opening repertoire is
very limited, sophisticated opening specialists
can dig up the most unpleasant move order.
The aim of the first move 1.c4 is usually to
others. The drawback is that the English
Opening itself does not promise much. On the
other hand, Black needs to do a lot of work if
there is a possibility of White playing 1.c4 and
has the English Opening in his repertoire.
At the very top level, 1.c4 has been very
popular throughout chess history and was
used by players such as Tigran Petrosian,
Viktor Korchnoi and others. The move is also
very popular among professionals who make a
living playing in Open tournaments – players
such as Mikhail Gurevich, Evgeny Bareev
and Daniel Fridman. Their purpose with this
move order is to just play a slow game against
less experienced weaker players and not get
involved in some early theoretical discussion.
1...¤f6 2.¤c3
Of course, after 2.d4 Black can hardly get
into the Tarrasch anymore, but White had
already chosen not to play this move on move
1!
2...c5 3.g3 e6 4.¤f3 ¤c6 5.¥g2 d5 6.cxd5
exd5 7.d4 ¥e7 8.0–0 0–0 9.dxc5 ¥xc5 10.b3
This is another possible move beside the
more direct 10.¤a4.
10...¦e8 11.¥b2 ¥g4 12.¦c1
Now something unexpected happened. Lein
voluntarily gave up not one, but both bishops!

 
 
   
   
   
  
 
  

54 Grandmaster Opening Preparation
12...¥xf3!?
I am not here to make any critical remarks
about the move itself; final chess truth is not
the aim of this book. I just want to uncover
the process of how some of the positional
ideas evolved in the Tarrasch Defence. I met
Anatoly Lein when he was living in the United
States in the 1990s. At that time I had already
studied his games and for me everything was
clear. Unfortunately, I did not ask him how he
worked on his Tarrasch Defence.
To complete my detective work today, I
would need to go through all the available
chess material published in the Soviet Union
starting from the 1959 Tal – Keres game,
until 1971 when this game took place. It
might still only be of academic interest,
because I am sure that my readers can already
draw conclusions as to how Lein got the idea
of giving up his bishop for the knight.
13.¥xf3 ¥d4 14.¥a1 ¥xc3!? 15.¦xc3 £d6
16.¦d3 ¤e5 17.¦d2 ¤xf3† 18.exf3 ¦ad8
19.¥xf6 £xf6 20.¦xd5 g6
White has won a pawn, but does not have
any real winning chances here. Podgaets even
managed to lose the game.

  
 
   
   
    
  
   
  

21.f4 ¢g7 22.£f3 ¦xd5 23.£xd5 ¦e2
24.a4 b6 25.¦d1 £b2 26.£d4† £xd4
27.¦xd4 ¦b2 28.b4 a6 29.h4 ¢f8 30.¢g2
¢e7 31.¦c4 ¢d6 32.¦d4† ¢c6 33.¦e4
¢d6 34.¢f3 ¦b3† 35.¢e2 ¦a3 36.b5
axb5 37.axb5 ¦b3 38.¦e5 f6 39.¦e8 ¦xb5
40.¦h8 h5 41.¦g8 ¢e6

   
    
  
  
    
    
   
    

42.¦xg6?? ¢f7
White is losing his rook.
0–1
Soviet masters picked up the Tarrasch Defence
from Paul Keres. As we saw, Spassky’s coach
Bondarevsky was hooked on the Tarrasch
as well. It is more or less known who helped
Spassky, although how much help or advice he
got from Paul Keres is actually not clear. Back
then it was not considered good practice to
help other World Championship Candidates
if you were still a contender yourself.
Paul Keres, however, had a very friendly
relationship with Boris Spassky. Who would
not? At least before the latter became World
Champion. According to Alexander Roshal,
the title of World Champion spoils everybody,
and even Spassky was affected by the venom.
Still, he was an easy-going gentleman with
whom I had the chance to play chess and
tennis. Keres was a much better tennis player,
but he lost his last Candidates match to Spassky
in Riga 1965, which was his final attempt to
fight for the crown. It was the closest match
for Spassky and he went on to play the World
 Chapter 2 – Evolution of the Isolated Pawn 55

Championship match against Petrosian in
1966. The Armenian prevailed, but Spassky
was back in 1969.
The 1969 match was decided in the Tarrasch
Defence. It might seem a little counterintuitive
to play the Tarrasch Defence against Tigran,
who was famous for his very safe approach to
chess. For example, he never took risks at the
end of a tournament, making draws and just
patiently awaiting the outcomes of his rivals,
who on many occasions cracked under the
pressure.
In chess one must have a feeling for danger,
meaning that suddenly during the game there
is a moment at which, instead of having a
winning position or just a better game, there
is a danger that everything will turn around.
It does not mean that you must be super-
careful or timid in your choices, but nobody
plays 100% perfect chess all the time, and
you cannot win every position in which you
have an advantage – but you also do not
want to lose them. The best example of how
this kind of sudden death can happen was
demonstrated in a Korchnoi – Karpov game in
Baguio City.
Viktor Korchnoi – Anatoly Karpov
Baguio City (17) 1978

    
  
   
    
 
   
  

    

White had held a nearly winning advantage
for most of the game. In the diagram position,
in time trouble, Korchnoi decided that taking
the pawn on h7 was the simplest solution.
37.¦xh7?
37.¦e7! was the correct move. After
37...¤cd2 38.¦a3 ¦f6 39.¦a1 White is
winning.
37...¤cd2 38.¦a3 ¦c6
Suddenly there is no win. Not feeling the
danger, Korchnoi made the next move without
any hesitation.
39.¦a1??

    
   
   
    
  
    
  
    

39...¤f3†
A cold shower. Instead of winning, White is
getting mated in two more moves.
0–1
Feeling this moment of when to back off
was Petrosian’s trademark. Later, Vladimir
Kramnik became famous for having the same
sense. Of course, being super-cautious all the
time limits your chess achievements a lot; this
happened to the very talented Hungarian GM
Peter Leko.
In concrete situations, finding a forcing way
to back off, like simplifying into a drawish
56 Grandmaster Opening Preparation
endgame, is the way to go. The Tarrasch
Defence seems innocent and very safe if you
look at it from White’s perspective. However,
there is no forced draw available. We are not of
course talking about the situation where one’s
aim is to draw with the white pieces. White
always has some kind of theoretical opening
advantage in the Tarrasch, but there is no way
to leave the battleground whenever you please.
Black can force White to play on even when
standing slightly worse.
Boris Spassky
Spassky’s first experience with the Tarrasch
Defence in an important sporting situation
was during his match against Efim Geller in
1965.
Efim Geller – Boris Spassky
Riga (7) 1965
1.d4 d5 2.c4 e6 3.¤c3 c5 4.cxd5 exd5 5.¤f3
¤c6 6.g3 ¤f6 7.¥g2 ¥e7 8.0–0 0–0 9.¥g5
¥e6
In his Candidates match against Geller,
Spassky was still following in the footsteps of
Paul Keres. Later, against Petrosian, he did not
play 9...¥e6.
10.dxc5 ¥xc5 11.¤a4

  
 
  
   
   
   
 
  

Now Spassky did not play the Keres move,
11...¥b6. The reason is simple: Keres created
the concept, but not a forced line which you
should blindly follow.
11...¥e7 12.¥e3 ¥g4!
Spassky follows the concept. The bishop
should be traded for the knight.
13.¦c1 ¦e8
Another deep move, which Spassky also used
later in his match against Petrosian in 1969.
When preparing against a certain opponent or
opening, one should notice these subtle moves.
There are common plans and ideas in similar
positions and openings. This rook move, which
looks like a waiting move, is actually very
useful, and if you are familiar with it you can
easily find it in similar positions. This is why
the study of different openings is very useful,
even if you do not play them. However, it can
be difficult nowadays to find a good opening
book that does not just follow the computer’s
top lines, but explains the moves and
ideas.
14.¤c5

 
 
   
   
   
   
 
  

14...¥xc5!
Spassky follows the Keres concept. If you
know it, you do not need to calculate other
 Chapter 2 – Evolution of the Isolated Pawn 57

moves like 14...£b6!? 15.¤e6 £a5 16.¥d2 24...¥h3 25.¥h1?! ¥f5

£b6, although this is also satisfactory for 25...£g4! 26.¦xd5 ¦xd5 27.¦xd5
Black. 
  
15.¥xc5 ¤e4
15...£d7! was more precise.
  
   
16.¥e3 £d7 17.£a4    
17.¤d4! ¥h3 18.£a4 and White has a slight  
pull, but not 18.¥xh3 £xh3 19.¤xc6 bxc6   
20.¦xc6? ¦e5! which gives Black a winning
attack on the h-file.
  
   
 
  27...¤xf2! 28.£xg4 ¤xg4 Black obtains a
 better endgame.

    26.¤d4 ¤xd4 27.¥xd4 b6 28.¦e3 £g6?

    28...£h6! 29.¥g2 ¦c8 with a dynamic
equilibrium.
 
    
    
      
    
17...h5!? 18.¦fd1 ¦ad8 19.¦d3   
19.b4 a6 20.£b3 h4 leads to a complicated   
game.
    
19...h4 20.¦cd1 hxg3 21.hxg3 £c8 22.£a3   
22.¦xd5 ¦xd5 23.¦xd5 ¤xg3! is dangerous
for White.
  

22...£e6 23.£b3 ¦d7 29.¦c1
The game is approximately balanced. Spassky 29.¥xb6! axb6 30.¥xe4 ¦xe4 31.¦xe4 dxe4
has obtained a complicated dynamic position 32.¦xd7 ¥xd7 33.£xd7 e3 34.¢g2! gives
from the opening, meaning that he has won White an extra pawn in the endgame.
the opening battle.
29...¦ed8 30.¦c6 £h7 31.¥g2 ¥h3 32.£c2
24.£a4 ¥xg2 33.¢xg2 £h5 34.¦f3 f6 35.¦c8 £e8
In mutual time trouble some mistakes were 36.¦xd8 ¦xd8 37.¦e3 £d7 38.¦d3 ¦c8
made by both sides, and the game ended in a 39.£d1 ¢f7 40.£h1 £g4 41.£h3
draw. ½–½
58 Grandmaster Opening Preparation
The theoretical preparation for the 1969 match
took place at several training camps sponsored
by the Soviet sports authorities. Who exactly
came up with the idea of playing the Tarrasch
Defence remains a secret. Spassky himself was
not very hard-working; for example, before the
match against Fischer his favourite place was
the Jurmala resort in today’s Latvia (back then
in the Soviet Union).
Nikolai Krogius took part in this training
camp as well. He was not a strong GM
throughout his chess career, but in the 1980s
he became one of the highest Soviet chess
officials. He also had a degree in psychology.
However, I do not really believe that the
opening preparation against Petrosian or
Fischer involved any high-end psychology.
Bondarevsky was no longer in Jurmala, but he
played a big role in the match against Petrosian.
Also there were other training camps where
all the Soviet Olympiad team members took
part. Probably it was only at these camps that
Spassky and Keres exchanged some of their
opening knowledge. Keres did not influence
Spassky’s opening choice directly and I suspect
that it was Bondarevsky who recommended
the Tarrasch, because against Fischer, when
Bondarevsky was no longer around, the
Tarrasch Defence, which may have actually
suited Spassky, did not appear.
Of course, the last word in making the
opening choice is for the player himself. I
believe this was the case with Spassky as well.
Since he was not a hard-working professional
like some other famous World Champions,
especially Botvinnik before him and Fischer
after him, most likely it was easy to talk him
into playing the Tarrasch.
At the Jurmala camp the occasional morning
cognac helped the creative atmosphere, but it
cannot be recommended – or at least there
is no evidence that it helps your opening
preparation. I also cannot speculate on how
significant the impact of opening preparation
was against Petrosian. Spassky was younger
and fitter than his rival, and as we like to say, it
was just his time.
The Tarrasch Defence served Garry Kasparov
for years to come in his title matches, but
I think the match between Spassky and
Petrosian was more or less the crucial test for
the Tarrasch Defence, and no more really fresh
ideas were found in subsequent years. There is
a book about the 1969 match where two of
the seconds of the match (Igor Bondarevsky,
who helped Spassky, and Isaac Boleslavsky,
who helped Petrosian) annotated the games
separately, not knowing his counterpart’s
comments. Later these comments were
published together in a book which was like a
bible for my first coach Tonu Truus, who rarely
lent his copy of the book to a third party. To
my knowledge this book is not available in
English, so I include the comments of the two
seconds here.
Tigran Petrosian – Boris Spassky
Moscow (4) 1969
“After this game, when Spassky again played
the Tarrasch Defence, it became clear that
Petrosian was going to face this opening
throughout the match. However, when
preparing for this game Petrosian and his
coaches were already forced to come up
with some improvement compared with the
second game of the match, when Black had no
difficulties levelling the game.” – Bondarevsky
1.c4 e6 2.d4 d5 3.¤c3 c5
“When in the second game Spassky played
the Tarrasch Defence, we had no clue whether
it was a thoroughly prepared system for the
whole match (not of course the only system,
but an opening system which would occur
more than once) or just a one-off, hoping for
 Chapter 2 – Evolution of the Isolated Pawn
a surprise effect. Petrosian did not have a very
high opinion of the Tarrasch Defence and was
inclined to the second view, but this game
made it clear that the Tarrasch Defence was
not finished yet.” – Boleslavsky
4.cxd5 exd5 5.¤f3 ¤c6 6.g3 ¤f6 7.¥g2
¥e7 8.0–0 0–0 9.¥g5
Petrosian was ready for the Tarrasch this
time, and most likely made this move hoping
that Spassky was going to repeat his game
against Geller.

 
 
   
   
    
   
 
  

9...cxd4!
We saw 9...¥e6 above in Geller – Spassky,
Riga (7) 1965.
The other main line here starts with
9...c4, although I never played this myself.
The traditional idea of the Tarrasch Defence
– to get dynamic counterplay which usually
compensates for the isolated pawn – is not
present after ...c5-c4.
10.¤xd4 h6 11.¥e3 ¥g4?!
Later on during the match, the more flexible
move 11...¦e8! was found.
“Petrosian was very sceptical of this move and
was taken aback that Spassky introduced it
twice. Still, he needed to take his time now
59
to figure out a better plan than in the second
game. Anyway, the next two moves made
by Petrosian in the second game were fine.”
– Boleslavsky
12.¤b3
Petrosian’s camp was not ready for the
Tarrasch and it took some days to realize that
here 12.£a4! is very strong. This move was
used in the 12th game of the match, where
White got a clear advantage but was still not
able to win the game.
12...¥e6 13.¦c1

  
  
  
   
    
   
 
  

13...¦e8!
This small move changed the theory of
the Tarrasch to the correct path. Very soon
after this game, Spassky’s team realized that
the correct move order is to play this on the
11th move. I still admire Spassky’s intuition
in making such a move without home
preparation. Calculating a forced line is much
easier then uncorking some complicated move
order during a practical encounter.
“Now we have reached the critical position.
It did not make sense to play 14.¤b5 again,
after which Black obtained a comfortable
position in the second game of the match.”
– Boleslavsky
60 Grandmaster Opening Preparation
14.¦e1
This quiet move protects the e2-pawn just
in case, and also defends against the possible
...£d7 followed by ...¥h3.
14...£d7 15.¥c5!
The World Champion is also alert. He of
course knows which pieces he needs to trade.
We already know that the more active 15.¤c5?
is a positional mistake.

 
 
  
   
    
   
 
   

15...¦ac8
“Black could avoid the bishop exchange
with 15...¥d8, but this move disconnects the
rooks and does not suit Spassky, whose first
commandment is piece harmony.” – Boleslavsky
16.¥xe7
“White could postpone this exchange,
because it is not convenient for Black to take on
c5. An interesting option was to play 16.¤b5
increasing the positional pull.” – Boleslavsky
“After 16.¤b5 Black had 16...¥h3 and if
17.¥h1 (after 17.¥xe7 ¤xe7 it is not possible
to avoid the exchange of bishops, which
weakens the king’s position) then 17...a6
18.¥xe7 £xe7 19.¤5d4 ¤e4.” – Bondarevsky
16...£xe7 17.e3
“If 17.¤b5, with the idea of transferring
it to d4, or immediately 17.¤d4, Black has
17...£b4. On the other hand, Black is planning
to play 17...¦ed8 followed by ...d4. Because
of that, White is once more strengthening the
d4-square and opening the e2-square for his
queen, at the same time freeing the d-file for
his rook.” – Bondarevsky
17...¦ed8
As always in the Tarrasch, White has a small
plus, but here is the situation I have previously
described. The position is alive, and you
constantly need to make decisions. Here, for
sure, Petrosian did not yet sense any danger.
The position looks very safe for White. He
did not realize in time that Spassky was not
playing for a draw, which in match practice is
the common trend, but instead was looking to
take any chance to take over the initiative.

  
  
  
   
    
   
  
   

18.£e2
“White is very aware that without the rook
on d1, the d5-pawn – the only weakness Black
has – is out of danger. This is why the queen
needs to leave the d-file, but it is constrained
in its movements and there is no better square
for it than e2.
It might seem that Black can use the fact
that the white rook is not yet on d1 and the
white queen has crossed to the e-file, and play
 Chapter 2 – Evolution of the Isolated Pawn
18...d4 to get rid of the isolated pawn. In
reality, however, after 19.¤xd4 ¤xd4 20.exd4
¦xd4 21.¤b5 ¦xc1 22.¤xd4! ¦xe1 23.£xe1
£d7 (if 23...b6 then 24.¥h3 winning the
pawn) 24.¤xe6 fxe6 25.£e5 ¢f7 26.£b8
Black must fight for a draw in a slightly worse
endgame. Spassky of course prefers a more
complicated game.” – Boleslavsky
18...¥g4!
“There was no sense getting rid of the
isolated pawn with 18...d4. The simplification
after 19.¤xd4 ¤xd4 20.exd4 ¦xd4 21.¤b5
¦xc1 22.¤xd4 works in White’s favour. The
concept of playing the Tarrasch Defence is not
to think about the weakness of the isolated
pawn on d5. Primarily Black’s consistent play
should make use of the power of this pawn.” –
Bondarevsky

  
  
   
   
   
   
 
    

“The last move deserves an exclamation
mark because of its psychological impact.
In chess terms it is good also, not allowing
White to conduct his plan conveniently. If the
white rook were already on d1, then White
would have only one option; now he must
make a choice between pushing the f-pawn
or temporarily worsening his queen’s position
(19.¥f3 cannot be recommended because
the exchange of bishops weakens the king’s
position).” – Boleslavsky.
61
19.f3
Petrosian wants to control the situation as
much as possible and avoids more complicated
lines like 19.£f1 ¤e4 20.¤e2 ¤g5 21.¤ed4
£b4, when Black’s activity looks dangerous.
“This move is difficult to criticize. White is
pushing the bishop back and reliably securing
the e4- and g4-squares. The e3-pawn is easy
to defend and also it is not difficult to activate
the bishop. Still, the f2-f3 move is adding
some sharpness to the game, which the World
Champion usually tried to avoid. The queen
retreat looks more flexible and it must move
to f1. Withdrawal from the f1-a6 diagonal
would be a gain for Black. To provoke the ...a6
move [after 19.£b5] does not make any sense,
because after 19.£f1 ¤e5 20.¤d4 £b4 21.£b5
White has the better endgame. The plan after
19.£f1 is to push back the black bishop to
h5 with h2-h3, and to threaten afterwards to
transfer the knight to f4. After 19...¤e4 White
has only one correct move: 20.¤e2! The white
queen is temporarily immured, but it looks
like there is no way for Black to exploit this.
20...¤b4 is no good because after 21.f3 ¤d2
22.¤xd2 £xe3 23.£f2 £xd2 24.¦cd1 White
retains the material advantage. The exchange
of the knight on e2 does not solve Black’s
problems; after that it is difficult to defend
the d5-pawn. It is possible that after 19.£f1
Black must call back his attack and retreat
the bishop to e6, accepting passive defence.”
– Boleslavsky
“19.£f1 was possible, but the queen is
misplaced and after 19...¤e4 White cannot
continue without f2-f3 anyway.” – Bondarevsky
19...¥f5
“Retreating the bishop to f5, Black is
tempting White to chase the bishop further
with 20.g4, but there is no need to make this
weakening move now.” – Boleslavsky
62 Grandmaster Opening Preparation
20.¦cd1 ¤e5
This move is not correct in positional terms
– the knight on c6 controls the d4-square. But
Spassky was trying to create as unpleasant a
situation as possible for Petrosian to play.

  
  
    
  
    
  
 
   

21.¤d4
“Here White could help Black to get rid of
the isolated pawn by playing 21.¤xd5 ¦xd5
22.¦xd5 ¤xd5 23.e4, after which Black has
a choice between 23...¥xe4, leading after
24.£xe4 ¤c6 25.£xe7 ¤dxe7 26.f4 ¢f8 to an
endgame where White has just a ‘better draw’,
or sacrificing a pawn for the initiative after
23...¤d3!. Three white pieces are under attack.
24.exf4 is not possible in view of 24...¤xe1. In
the case of 24.exd4 £f6 the pawn on b2 will
be lost. The best option for White is 24.£xd3
¤b4 25.£b5, but even here after 25...¥e6
Black has a good game.” – Bondarevsky
21...¥g6
21...¥d7 was possible, but after that White
has a forcing line. The strategy used by Spassky,
however, was not to allow his opponent
concrete play. 22.f4 ¥g4 23.fxe5! ¥xe2 24.exf6
£xe3† 25.¢h1 £f2 26.¦xe2 £xf6 Despite the
material unbalance, according to the computer
the position is equal.
“21...¥d7 does not work because of 22.f4 ¥g4
23.fxe5! ¥xe2 24.exf6 £xe3† 25.¢h1 and
White has three pieces for the queen and the
better position.” – Boleslavsky
22.¥h3 ¦c4!?
An active move which is not really
threatening anything, but just forcing White
to calculate some tactics.
“With his last move Black is creating further
sharpness, taking a risk of ending up in a
difficult situation, as the hanging position of
his rook may undermine some combinational
motifs. The rook does not have a good square
on the c-file, because after 22...¦c7 White
has 23.¤cb5, and after 22...¦c5 White has
23.¤a4, but Black could just play 22...¦b8.
White’s main threat – regrouping one of the
knights to f4 – is difficult to achieve because
of the weaknesses of the d3-square. If White
uses his bishop to defend the d3-square, then
Black’s rook can return to c8. In general,
the position after 22...¦b8 is about equal,
but now Black is hanging by a thread. What
prompted Spassky to take this kind of action?
It is possible that the loss of the first game of
the match encouraged him to equalize the
score as soon as possible. Also he could draw
extra courage from the second game, where the
situation was favourable for him, and he made
this puzzling retreat (or rather attack) with
his rook speculating on his opponent’s future
mistakes.” – Boleslavsky

   
  
   
   
   
  
  
   

 Chapter 2 – Evolution of the Isolated Pawn
23.g4!?
Spassky’s bluff has worked, and Petrosian is
playing for an advantage.
“White’s last move is not aesthetically pleasing.
White is shutting down his own bishop and
weakening his kingside. The move itself does
not suit Petrosian’s style at all, but most likely
Petrosian realized that this move has to be
made. The game has entered a phase in which
one cannot restrict oneself only to general
understanding.
After 23.g4 the threat is f3-f4 followed by
¤xd5, which means that the black rook on c4
needs to continue its journey. The question
is, could White create the same threat just by
moving his queen to f1? After 23.£f1 ¦b4
24.b3 ¥h5! 25.g4 (25.£e2 ¤e4 is unpleasant)
25...¥g6 26.£e2 ¤c6 27.£d2 a similar
position could arise to that in the game.” –
Boleslavsky
“After 23.£f1 Black could play 23...£b4.” –
Bondarevsky
Most likely the World Champion spent a lot
of time on the combinational possibility: 23.f4
¥h5 24.£f1

   
  
    
  
   
   
   
  

It looks very promising, as 24...¥xd1 does
not work because of 25.fxe5.
Also not satisfactory for Black is 24...¦xd4
25.exd4 ¤f3† because of: 26.£xf3! £xe1
63
27.¦xe1 ¥xf3 28.¦e7 b5! 29.b4! (after
29.¤xb5 ¦b8 or 29.¦xa7 b4 30.¤a4 ¦e8
Black’s rook penetrates White’s position)
29...¥xg4! 30.¥xg4 ¤xg4 31.¦c7! ¦e8 32.h3
¦e1† 33.¢g2 ¢f6 34.¤xb5 White has good
winning chances.
“Still, the position does not promise much
because of 24...¤c6! 25.g4 ¤xg4 26.¤xd5
¦xd5 27.£xc4 £h4! 28.¥xg4 (28.£f1 ¤xe3!
29.¦xe3 ¥xd1 favours Black) 28...£xg4†
29.¢h1 ¦xd4 and draw by perpetual check.”
– Boleslavsky
23...¦b4
This was too active, but as Emanuel Lasker
pointed out, to win an equal position one first
needs to weaken one’s own position.
More solid was 23...¦c7 24.f4 ¤c4 25.¥g2
¦cc8 26.f5 ¥h7 27.h3 and White is better
because of the misplaced bishop on h7.
“In case of 23...£c5 24.f4 ¤c6 25.¤b3
£b4 White could achieve a better game
after 26.¥g2!. Now after 26...¤e7 White
has 27.f5 ¥h7 28.¤d4, and if 26...¥e4 then
27.¤d2 ¦c5 (27...¥xg2 28.¤xc4 ¥h3 29.¤e5
¤xe5 30.fxe5 ¥xg4 31.£g2) 28.a3! and it is
not possible to play 28...£xb2 because of
29.¤dxe4 £xe2 30.¤xf6 gxf6 31.¤xe2.”
– Boleslavsky

   
  
   
   
   
  
  
   

64 Grandmaster Opening Preparation
24.b3
“This move was heavily criticized by many
commentators during the match, but they
probably did not have sufficient basis for their
statement. The critics’ point was that White has
already played 23.g4 with the idea of shutting
down the black bishop, and now suddenly he
makes a move 24.b3 giving up the original plan
for no reason. The question is how good the
pawn push really is? It is true that after 24.f4
¤c6 25.f5 ¥h7 the black bishop is cut off from
the battle for a very long time, but the white
bishop on h3 is also passive, just protecting
the g4-pawn. White has weakened his central
squares and Black at some point might manage
to break the pawn phalanx on the kingside and
free his bishop, with devastating effect. Added
to this is the fact that Petrosian has already
committed himself to weakening the kingside,
which is contradictory to his customary play.
He naturally did not want to push his pawn
further, creating more vulnerable squares in his
camp. Petrosian continued with his intended
plan and focused on Black’s rook instead of the
bishop.” – Boleslavsky
“The tempting 24.f4 did not bring any benefits
after 24...¤c4 25.b3 ¤d6 26.f5 ¥h7 27.¥g2.
In my opinion after 27...£e5 White needs to be
alert about both the breakthrough threats ...g6
and ...h5. There is no way we can talk about
any advantage for White.” – Bondarevsky
With the luxury of computer analysis, one can
see that Bondarevsky’s assessment is inaccurate
due to 28.¤xd5! ¤xd5 29.¥xd5 £xd5 30.¤c2
followed by ¤xb4, ¦d4 and ¦ed1 when White
is close to winning.
Boleslavsky, on the other hand, was right to
give 24...¤c6! as Black’s best. In that case,
White has little to no advantage.
24...¤c6 25.£d2 ¦b6

   
  
  
   
   
 
   
   

26.¤ce2
Now it is obvious that the g3-g4 plan did
not work, or perhaps it would be better to say
that White failed to make it work. The game
has taken a lot of energy from both players and
now the slow shifting piece play following the
turmoil is understandable.
More active was 26.¤a4 ¦a6 27.¥f1 ¤xd4
28.exd4 ¦e6 29.¤c5 ¦xe1 30.¦xe1 £c7 with
an equal game.
“White’s last move deserves a question mark,
and not only because it is a bad move. We
have already mentioned that the knight
should go to f4, and if Black’s rook was on
b8 then playing the knight to e2 would be
possible.
However, here White did not continue the
plan which promised him an advantage. White
should play 26.¤a4!. There is always some
kind of tactical miscalculation behind the
decision to abandon an active continuation.
Tired by the previous play, Petrosian did not
see that after 26.¤a4 Black could not play
26...¤xd4 27.exd4 ¦e6, because after 28.g5
Black is losing the exchange. After 26.¤a4!
what options does Black have? Only 26...¦a6,
but after that White has a pleasant line:
27.¥f1 ¤xd4 28.exd4 (weaker is 28.£xd4 ¦e6
29.£xa7 h5! and Black has good counterplay
 Chapter 2 – Evolution of the Isolated Pawn
for the pawn) 28...¦e6 29.¤c5 ¦xe1 30.¦xe1
£c7 31.¦e5 White has a positional pull.
After 26.¤ce2 the worst for Black is
over. Also, the World Champion, who was
already in slight time trouble, started to play
uncertainly.” – Boleslavsky
Petrosian did not follow up with the rook
hunt, and probably could not find anything
concrete. 26.¤a4 certainly gave him some
initiative, but the rook remains safe. For
example: 26...¦a6 (26...¤xd4 is not possible,
because after 27.exd4 ¦e6 28.g5 White wins
an exchange) 27.¥f1 (after the more solid
27.¦c1 then 27...¤xd4 28.exd4 £a3 is
possible, or 28.£xd4 h5 with counterplay)
27...¤xd4 28.exd4 (or 28.£xd4 ¦c6 29.£xa7
h5) 28...¦e6 29.¤c5 ¦xe1 30.¦xe1 £c7
Black has a solid position without weaknesses”
– Bondarevsky).
26...¥h7
“A tricky move. Black moves his bishop in
time and implies that after 27.¤f4 he is going
to play 27...g5.” – Boleslavsky
“Facing the knight move to f4, Black moves
his bishop away in good time.” – Bondarevsky

   
 
   
   
   
 
  
   

27.¥g2
65
“In my opinion this is the critical moment
that changes the course of the game. White’s
initiative on the kingside has dried up and he
has ended up with weaknesses and a passive
bishop. This situation definitely had some
negative psychological impact on Petrosian’s
subsequent play. In general, the position is
still around equal.” – Bondarevsky
White gave up the 27.¤f4 idea mistakenly,
because 27...g5 28.¤fe2 h5 29.gxh5
¤xh5 30.¤f5 favoured White. Also, when
withdrawing the misplaced bishop from
h3 then it should only go to f1. After the
f2- and g3-pawns moved forward the bishop
had no active future and it was better to try
to exchange it. For example, 27.¥f1 ¤xd4
28.¤xd4 ¦d6 29.¥d3 with a good game for
White.” – Boleslavsky
27...¦e8 28.¤g3
Again not the best move. It is clear that
Black will not allow the knight to move to
f5, and after exchanges on d4 the knight on
g3 is misplaced. White should play 28.¤f4,
after which White still has the more active
position.” – Boleslavsky
28...¤xd4

  
 
    
   
   
  
  
   

66 Grandmaster Opening Preparation
Now the exhausted Petrosian took with the
pawn. There was nothing wrong with 29.£xd4
¦e6 30.¢f2, but again the position is still alive.
There is no way White can force a draw. This
was the outcome of the opening choice. To
win a game against an equal opponent, perfect
opening preparation is not enough. One
needs to create the right battleground with
the opening choice, where one can outplay the
opponent later in a middlegame that might
not be the opponent’s cup of tea.
“Most likely it was better to take with the
queen on d4, but in time trouble Petrosian was
seeking more exchanges.” – Boleslavsky
The point is that he was looking to trade the
rooks on the open c- and e-files.
29.exd4 ¦e6 30.¦xe6 £xe6
“Black has got rid of his unfortunate rook,
which took a circular route from c8 to e6 and
claimed the e-file. White in the meantime has
transferred his knight from c3 to g3, where it
stands no better, and has moved his bishop to
g2, where it is stuck behind his own pawn. It
is obvious that the positional manoeuvring
has benefited Black. Still, White’s position is
very solid and the draw outcome looks most
probable.” – Boleslavsky
31.¦c1

  
 
   
   
   
  
  
    

“More precise was 31.¢f2. White could
wait with the rook move to c1 and make it
only after Black has moved his rook to c8. It
looks as though White is still dreaming about
an advantage, but this kind of stubbornness
is fraught with dangerous consequences.”
– Boleslavsky
In this innocent-looking position, Spassky
found a fantastic resource. The timing was
also perfect, just before the time control in
Petrosian’s time trouble. And of course there
are no easy situations or games in world
championship matches – the tension is
incredible.
31...¥g6!
“Black vacates the h7-square for his knight
with the idea of strengthening the position
with ...¤h7 and ...£f6, and then routeing the
knight via g5 or f8 to e6. This set-up threatens
the d4-pawn and creates pressure on the weak
dark squares on the kingside.” – Bondarevsky
32.¥f1
“This move is already a serious mistake.
There is no time to activate the bishop. 32.¢f2
was a must. The text move, removing the
bishop from g2, makes it clear that White did
not understand the idea behind Black’s last
move.” – Boleslavsky
This move itself is not a mistake. It is difficult
to decide at which moment Petrosian lost the
feeling of danger. The position is still equal,
but some accuracy is needed.
32...¤h7
Switching the knight might look like a long
shot, but Black is objectively already very
slightly better.
“Only now after this excellent move is White’s
weak kingside in danger. Black’s battery of queen
 Chapter 2 – Evolution of the Isolated Pawn 67

on f6 and knight on g5 is very uncomfortable
for White. If only the bishop had remained on
g2 White could play 33.h4, but now this does
not work because of 33...£f6 34.h5 ¤g5 (but
not 34...£xf3 35.¦c3.” – Boleslavsky
33.£f4?!
This pseudo-active move is a real mistake.
“Here the queen is sooner or later attacked by
the knight, and also the exchanges on the e-file
(after ¢f2) are no longer possible. On the
other hand, there was no other way to prevent
33...£f6.” – Boleslavsky

  
 
  
   
   
for Petrosian, who was in serious time trouble.”
– Boleslavsky
34.¦c5?!
34.£e5! £d7 35.£c7 £e6 36.£e5 and
White can save the game. In the endgame
there is no danger at all.
“After 34.£e5 the best solution for Black
is 34...£xe5 35.dxe5 ¤e6 36.¢f2 d4 (or
36...¦d8), but after 37.¤f5! White only had
to make the remaining moves until the time
control and Black would have had very few
chances to win the game.” – Boleslavsky
34...¥b1!
With his last move White defended against
34...£b6, but he missed the other idea found
by Spassky.
“Spassky immediately uses his chance to
activate his knight with tempo.” – Boleslavsky

   

      
      
    
33...¤f8    
Stronger was 33...£b6! 34.£d2 ¤g5 35.¥g2    
¤e6 36.¤e2 h5, with the advantage.
  
“After this move White could exchange    
queens, which most likely would have secured
the draw. A much more energetic move was
  
33...£b6 34.¦c5 ¤g5. The threat of 35...¤e6 
forces White to take the pawn with 35.¦xd5, 35.a4
but then there is 35...£c6 36.£d6 £c3, and With 35.£d2 White could remove his queen
White is defenceless against multiple threats. from the knight attack, but that would then
Spassky had enough time to calculate this allow 35...£f6. After 36.¦xd5 £xf3 White is
line, but most likely at this moment he did losing a pawn. The text move was relatively
not consider his position as winning and did best.
not look for a forced win. The move he made,
raising the tension, was also very unpleasant 35...¤g6 36.£d2
68 Grandmaster Opening Preparation

Playing safe is no longer an option here. When 37...¤f4!

Petrosian had enough time to think, he avoided Most likely Petrosian was more worried
some concrete lines. Now, however, switching about 37...¤h4?, but after 38.¥e2 White is
from safe mode to active, concrete play was a fine.
must. He could still save the game after 36.£c1
¤h4 37.£xb1 £e3† 38.¢h1 £xf3† 39.¢g1 38.a5
and Black has only a perpetual check. This move is just panic in time trouble,
when your only concern is not to lose on time.
36...£f6 37.¢f2
“Despite all the pressure from Black and After 38.¦c3 ¥g6 Black has a big advantage,
the small inaccuracies made by White, the but is not winning immediately.
position was defendable. Petrosian, defending
the concrete threat in time trouble, forgot that 38...¥d3!
he cannot allow the knight to land on f4. The “You do not see this kind of move very often
question is: can he simultaneously defend the in world championship games. These moves
pawn on f3 and the f4-square? Not directly, are possible only when your opponent has
but in chess there are tactical motifs to solve no time left to see even elementary tricks.”
this kind of problem. White had 37.¦c1, – Boleslavsky
moving the rook back to the first rank and
attacking the bishop on b1, which cannot 39.¤f5
move back because of the knight on g6. White “Also no good was 39.£c3 £h4 40.¢g1
is not winning the bishop, but it is enough to ¥xf1 41.¦c8 ¥b5!. In this line too, the move
force the black knight to move to h4, where 38.a5 plays the fatal role.” – Boleslavsky
it is much less dangerous than on f4. After
37...¤h4 38.¥g2! (not 38.¥e2 because of

38...£xf3) 38...¥g6 39.¢f2 Black does not   
have any positional advantage left, but White   
needs to be careful here.” – Boleslavsky
    
“After 37.¦c1 ¤h4 38.¥g2 (if 38.¥e2 then   
38...£xf3) 38...¥g6 39.¢f2 ¢h7 Black has a
positional advantage.” – Bondarevsky
   
  
       
      
    
39...£g5!
    “This is even stronger than 39...¥xf5
   40.£xf4 g5 41.£d2 ¥xg4.” – Boleslavsky.

   40.¤e3 £h4† 41.¢g1 ¥xf1

     In all lines there is a simple win for Black.
   0–1


 Chapter 2 – Evolution of the Isolated Pawn 69

This was a very heavy blow to Petrosian’s playing the Tarrasch when you need a draw,
confidence. The fourth game was more or less compared with other opening lines, is that
crucial for the overall result of the 24-game here you have to make only dynamic (tactical)
match. There were a few more games in the decisions – assuming, of course, that you know
Tarrasch Defence in the match, but during the how to handle the positional play.
second part of the match, Spassky used other
solid systems to neutralize Petrosian’s attempts 12.¦c1 ¥f8 13.¤b3 ¥e6 14.¤b5
to even the match score. The Tarrasch had Typical Petrosian: trying to obtain maximum
already done its job. control of the position. In this position, he
The only significant game was the 18th wants to guard the d4-square as much as
game, where Spassky introduced the ...¦e8 possible.
idea on move 11, which became the standard
main line in the Tarrasch and was successfully 14...¥g4 15.h3 ¥f5 16.¤5d4 ¤xd4
used later by Kasparov, Grischuk and 17.¤xd4
others.

Tigran Petrosian – Boris Spassky  
Moscow (18) 1969
  
    
1.c4 e6 2.d4 d5 3.¤c3 c5 4.cxd5 exd5 5.¤f3
¤c6 6.g3 ¤f6 7.¥g2 ¥e7 8.0–0 0–0 9.¥g5
  
cxd4 10.¤xd4 h6 11.¥e3     
    
   
     
    
    17...¥d7!
A typical situation when White has the d4-
     square: Black tries to control the other squares
     around d4.

  18.£b3 £a5 19.a3 ¥d6 20.£d3 £d8

  

11...¦e8!
An excellent move, and again at the right
moment. It was 18th game of the match and
Spassky was ahead. Now he used the Tarrasch
Defence not to play for a win, as in Game
4, but just to keep it safe. The advantage of
This is not the best plan. Better was
20...¦ad8.
21.¦fd1 £e7 22.¥d2 ¤e4 23.¥e1!
Petrosian would never look at the following
kind of line: 23.£b3 ¦ad8 24.£xb7? ¥xg3!
25.fxg3 ¤xd2 and Black is winning.
23...¥e5! 24.£b3
70 Grandmaster Opening Preparation

 27...£f6
  Suddenly White’s major pieces are misplaced,
  and he is forced to take on e4.

     
     
      
      
      
       
     
24...¥xd4!?   
This looks like a very straightforward way to
play for a draw.
    

Black neither wanted nor needed to play the 28.¥xe4 ¦xe4 29.¦xe4 dxe4 30.¥c3 £f5
riskier 24...¦ad8!?. Although this would work 31.¦d1 ¦e8 32.¦d6
well after 25.£xb7? ¥xd4 26.¦xd4 ¤xg3!–+ 32.£xa7?? would lose to 32...e3!.
or 25.£xd5? ¥a4 26.£xe4 ¥xd1 27.¦xd1
£f6µ, White could keep an edge with 25.e3. 32...f6 33.¦d4 a6 34.¢h2 £g4 35.a4 ¢f7
36.¢g1 ¦e5 37.¦d6 ¦f5 38.b3 £h3 39.¥e1
25.¦xd4 ¥c6 £g4 40.¥c3 £h3 41.¥d4
This is passive, but if Black badly needs a 
draw then he can just try to defend on the light
squares. Without any black pawn weaknesses,     
White can hardly win. I am not recommending   
this passive approach, as practice shows this
kind of strategy does not work for everybody.
  
I did use this approach once against Kramnik    
with success, but if possible one should avoid it.   
26.h4 £e5 27.£e3?   
After 27.e3 White can play on forever using    
his small advantage. It is not only a question of
the two bishops – in isolated pawn positions     
the two bishops are not very useful. The point 
here is that Black cannot open the kingside, 41...h5
because White would then create an attack on 41...¦f3!? 42.exf3 exf3 43.£e6† would lead
the dark squares by trading the knight on e4. to a drawn opposite-coloured bishop ending.
This means that White can push his pawns
slowly on the queenside, making Black’s 42.¥c3 ¢g8 43.¦d8† ¢h7 44.¦d6 £g4
defence difficult. Black can only wait. 45.a5 £g6 46.b4 £f7 47.¥d4 £c4 48.¥c5
 Chapter 2 – Evolution of the Isolated Pawn
¦e5 49.¥d4 ¦f5 50.¥c5 £c2 51.£d2 £b3
52.£d1 £b2 53.£d2 £a1† 54.£d1 £e5
55.£d4 £e8 56.£c4 ¥b5 57.£e6 £xe6
58.¦xe6 ¦d5 59.¦d6!
Avoiding the last trap: 59.¦xe4?? ¦d1†
60.¢h2 ¥c6–+
½–½
The Tarrasch Defence has been used with
success by Kasparov, Grischuk and yours truly.
It was an opening that worked very well for
me when I was the underdog. Because of the
seminar with Yurkov I was very aware of where
to put my pieces and it also helped me to avoid
getting into time trouble. Opening preparation
and time trouble are often linked – lack of
opening knowledge forces you to spend too
much time in the opening phase of the game.
The Tarrasch Defence has one drawback: it is
very easy to prepare against with White. There
are always opening lines where either White
or Black can choose to follow the main line
or deviate from it, but in the Tarrasch, most
of the time it is White who can choose the
line. This makes it easy to prepare, and you
particularly do not want to allow this situation
when your opponent is weaker than you. This
does not mean that the Tarrasch Defence is
somehow passive and may be used only to play
for a draw; quite the contrary, as we saw in the
historical games presented here.
There are some openings that should be played
just to develop one’s chess understanding.
The isolated pawn from the defensive side is
difficult to master, and studying and playing
the Tarrasch Defence helps the young chess
student to develop to a new level. There are
advantages to knowing the ins and outs of the
isolated pawn from Black’s side. In my practice
I used my knowledge of Lein’s games against
none other than Vladimir Kramnik, whom I
have already compared to Tigran Petrosian.
71
Spassky did not play the Tarrasch Defence
against Bobby Fischer in Reykjavik in 1972.
This shows one thing: he was afraid of Fischer,
but not Petrosian. The specific opening choice
reveals your intentions very well. The Tarrasch
Defence is an opening where Black is ready
to play the slightly worse position that arises
from nearly every line, but relies on his or her
middlegame mastery to outplay the opponent
in a complicated positional game that on many
occasions becomes a dynamic battle.
Seeing Fischer on the other side of the board,
who had demolished other great players such
as Bent Larsen and Tigran Petrosian, might
lower your confidence level. Spassky and/or his
team decided that the Tarrasch would not be
a good opening against Fischer. This is a pity,
because in my opinion it is actually a good
choice against a player who likes very concrete
play, as Fischer did.
For some reason, nobody picked up the
Tarrasch for a long period after the Spassky
– Petrosian match. I was introduced to the
Tarrasch in 1977, but it took many years for
me to play it in big tournaments. Only in 1983,
when I played in the USSR Championship
qualification tournament, did I use the
Tarrasch Defence, and with success, drawing
all my games with it. As I was a player who
was much more dangerous when handling
the white pieces, I managed to qualify for the
1984 USSR Championship.
In the Soviet Union, players were not
ranked only by their FIDE rating, as today,
but another quality system was used. Among
chess players there was the following hierarchy.
The grandmaster title itself was not given any
authority as evidence of being a good player.
The greatest value a player could achieve for
his chess résumé was participation in the
USSR Championship final. The legend goes
that when Gufeld gained the GM title he
greeted Korchnoi by saying, “Hello, colleague.”
72 Grandmaster Opening Preparation
Viktor looked around and, seeing the
Mongolian grandmaster Myagmarsuren,
pointed to him and said, “That is your
colleague.”
For me, qualification for the USSR
Championship at the age of 22 was a major
success. At the time I was still relying on the
openings taught to me by Vladimir Yurkov,
but he was helping and working with his own
students, mainly Andrei Sokolov, who won the
1984 Championship. My opening preparation
or repertoire was good enough to qualify for
the Championship, but I ended up in the
middle of the table.
Garry Kasparov
The Tarrasch Defence was picked up by Garry
Kasparov, although the secrets of his opening
preparation were never revealed. I played
him for the first time in 1977, and I lost as
White in the Caro-Kann. Despite being from
the same age group, we met only occasionally
in team competitions. I won the European
Junior Championship in 1983, but Kasparov
had already made his meteoric rise to the very
top and was playing in another league. We met
again years later in 1988 in the World Cup
series of tournaments.
The top players are usually the ones who show
how to play the openings or what to play.
Kasparov started with the Caro-Kann when,
at the tender age of 12, he won the USSR
Junior Championship against players who
were up to 18 years old. Later he switched
to the Scheveningen in the Sicilian Defence.
The King’s Indian Defence was always his best
opening. When qualifying for the Candidates
cycle in 1982 he badly needed another more
solid opening, as an alternative to the King’s
Indian. That was when he started to play the
Tarrasch Defence.
In the Soviet Union, preparation was not only
done by the player and his second or coach;
usually the Sport Committee in Moscow put in
every effort to help the player. Also the player
might have local help – in Kasparov’s case the
Baku officials, where he was a citizen at that
time. This help consisted not only in covering
the everyday needs of the player, but more
importantly in providing the resources to buy
valuable chess information: inviting specialists
in certain openings to the training camps.
The World Champion at that time was
Anatoly Karpov and, for example, he could
afford to have in his training camp the East
German grandmaster Wolfgang Uhlmann,
who was a French Defence specialist. Karpov’s
preparation was aimed at Viktor Korchnoi,
who was a keen French Defence player.
The two most talented players of the 1970s
and 1980s, Karpov and Kasparov, had all the
people and material resources they needed
available. Eventually they met each other
for the World Championship, and their
confrontation overshadowed other contenders
for the crown.
Kasparov first needed to defeat the other
contenders, which he did with ease. First he
won the Moscow Interzonal in 1982, and then
he faced Alexander Beliavsky. Both players
were Communist Party members, so their
opening preparation was presumably backed
with the maximum resources available.
Kasparov used the Tarrasch Defence in
his Candidates quarter-final match against
Beliavsky in 1983 – which took a bad turn for
Beliavsky in Game 2.
Alexander Beliavsky – Garry Kasparov
Moscow (2) 1983
1.d4 d5 2.c4 e6 3.¤c3 c5 4.cxd5 exd5
5.¤f3 ¤c6 6.g3 ¤f6 7.¥g2 ¥e7 8.0–0 0–0
 Chapter 2 – Evolution of the Isolated Pawn 73

9.¥g5 cxd4 10.¤xd4 h6 11.¥e3 ¦e8 12.£a4 20.¤b3

¥d7 13.¦ad1 ¤b4 14.£b3 a5 Beliavsky could have played 20.¤db5!?
Beliavsky was not ready for the Tarrasch ¥xb5 21.¤xb5 ¦xa2 22.£b3 and White
and, as in historic games from the past, he was should not be in danger of losing.
unable to spot the turning point and make a
draw in time. 20...¥c6! 21.¥d4 ¤e4 22.¤xe4 dxe4 23.¦a1?!
 
   
    
        
        
        
       
   
       
 
15.¦d2?! 23...¥d5! 24.£b1 b6! 25.e3 ¤d3 26.¦d1
At that time I also had practical experience 26.¥xb6? hopes for 26...¥xb3? 27.¦xd3!÷,
from this position: 15.a4 ¦c8!? 16.¤c2 b5 but is refuted by 26...¥b4! 27.¦c2 ¥xb3 and
17.¤xb4 bxa4 18.¤xa4 ¥xb4 19.¤c3 ¥xc3 Black wins.
20.bxc3 a4 21.£a2 £e7 22.¦fe1 ¥c6 23.¥d4 26.¤c1 may be White’s best try, but
¤e4 Black had an edge and eventually won an 26...¤e1 keeps an edge for Black.
ending in Majorovas – Ehlvest, USSR 1983.
26...b5 27.¥f1
15...a4 16.£d1 a3 17.£b1 ¥f8 18.bxa3 27.¤c1 would now be met by: 27...¤xc1
¦xa3 19.£b2 £a8! 28.¦xc1 b4!µ
Black has the initiative.

  
     
      
       
       
       
        
   
    

74 Grandmaster Opening Preparation
27...b4 28.¥xd3?! exd3 29.£xd3 ¦xa2
30.¦xa2 £xa2 31.¤c5 ¥f3 32.¦a1 £d5
33.£b3! £h5 34.¤d3 ¥d6 35.¤e1 ¥b7
36.¦c1 £f5 37.¦d1 ¥f8 38.£b1
White lost on time.
0–1
This game itself did not decide the match, but
it shows that whereas in tournament chess you
can get away with weak opening preparation,
this is not the case in a match. You not only
have to guess what openings your opponent
might play, but you also need to have some
openings or lines in your own repertoire that
are not played very often at the top level.
A lack of homework by opponents is regularly
exploited by players today, when it is more and
more difficult to surprise your opponent in the
opening. Rapport and Morozevich are players
who like to experiment with rare opening lines.
This is a risky approach, but if the opponent is
lazy and unfamiliar with a not-so-correct line,
it works in the researcher’s favour.
Beliavsky had many tournament successes,
twice winning the USSR Championship, but
he was never a researcher – meaning that in his
opening preparation he relied on well-known
lines created by others, and he did not have
much original preparation of his own.
Kasparov used the Tarrasch once more in
this match, drawing the game, and then
suddenly switched back to his main weapon
– the King’s Indian Defence. His reasoning
was very simple: Kasparov was ahead, and
Beliavsky might feel that his last chance was
to win the 8th game of the 10-game match.
In the situation of being a point ahead with
three games to go, Kasparov’s choice was not
to defend, which you need to do most of the
time in the Tarrasch Defence, but to play for
counterattack from the very first move. This
tactic works in very tense, nervous situations,
because it is much more difficult to defend
when you are under enormous pressure. The
risk taken by Kasparov worked perfectly, and
most likely also had a surprise effect. Kasparov
took the initiative in the opening and won a
nice game.
The next barrier for Kasparov on his route
to Karpov was Korchnoi – a formidable
opponent. And for Kasparov the Tarrasch
Defence was no longer a secret weapon.
Viktor Korchnoi – Garry Kasparov
London (2) 1983
1.d4 d5 2.c4 e6 3.¤f3 c5 4.cxd5 exd5 5.g3
¤f6 6.¥g2 ¥e7 7.0–0 0–0 8.¤c3 ¤c6
9.¥g5 cxd4 10.¤xd4 h6 11.¥e3 ¦e8 12.a3!?
This move, made by Viktor with the idea of
hiding the queen on a2 in some lines, is very
poisonous and was picked up by Smyslov,
who used the line against Kasparov in their
Candidates final match later in the cycle.


  
   
   
    
    
  
  

12...¥e6!?
Probably best.
Black might be tempted to play 12...¥g4
first, with the point that after 13.h3 ¥e6 White
will not have ¥h3 in some lines. However,
after 14.¤xe6 fxe6 White has 15.f4, and now
Black cannot jump with his knight to g4.
 Chapter 2 – Evolution of the Isolated Pawn 75

Another problem with 12...¥g4 of course Here White should try 17.f4 ¦ad8 18.¢h1 –
is just 13.£b3, and after 13...¤a5 White has the plan Smyslov later tried to execute against
14.£a2, which was the original idea of 12.a3. Kasparov in their Candidates final match.

13.£b3 17...£e7 18.e3 a6 19.£h4 ¦ac8 20.e4 d4

This move is now useless, and Black obtains 21.¤e2 e5 22.¥h3 ¦c7
a good game after it.

Much stronger was 13.£c2 ¦c8 14.¦ad1 ¥f8    
with a complicated game. Later on, Karpov    
did not like the idea of 12.a3, but I still think
this is one possibly critical line for Tarrasch   
aficionados.     
13...£d7 14.¤xe6 fxe6 15.¦ad1 ¥d6?
   
This is too slow. The position still requires    
some precise action from Black.
   
The correct idea is: 15...¦ad8! 16.¥c1 ¥c5!   
This is the difference – the bishop belongs on 
c5 rather than d6. 23.¥g5 ¢g8 24.¥xf6 £xf6 25.£xf6 gxf6
26.¤c1
16.¥c1 ¢h8 White has simplified the position in time.
Black’s task is to be alert to the e2-e4 No doubt Viktor had studied the Tarrasch
breakthrough. thoroughly and knew the dangers.

   26...¤a5 27.¤d3 ¤b3 28.¥f5 a5 29.¢g2
¢g7 30.¢h3 ¦ee7 31.¤c1
   ½–½
   Korchnoi could not crack Kasparov’s Tarrasch
    Defence, and he ended up losing the 6th game
     as White. He felt obliged to try and achieve
something against the Tarrasch Defence and
    introduced an interesting plan at a very early
   stage.
   Viktor Korchnoi – Garry Kasparov

17.£a4? London (6) 1983
A typical situation in the Tarrasch Defence.
White has the better position, but this cannot 1.d4 d5 2.c4 e6 3.¤f3 c5 4.cxd5 exd5 5.g3
last forever without doing anything concrete. ¤c6 6.¥g2 ¤f6 7.0–0 ¥e7 8.¥e3!?
76 Grandmaster Opening Preparation

The plan with ¥e3 was already known, but


having the white knight still on b1 makes it  
more complicated.  
    
     
      
       
     
       
    
  12.¦c1!
  12.£xb6 axb6 13.¦c1 ¤b4 14.¤c3 ¥e6
15.a3 ¤c6 16.¤b5 ¦fc8 was given as fine for
 Black by Kasparov. However, after 17.¤d3
8...c4 9.¤e5 0–0 10.b3 White has some advantage.
This was White’s idea.
12...£xb3 13.axb3 ¤b4! 14.¤a3
The usual move is 10.¤c3, just transposing Much stronger was the more natural:
into normal channels: 10...¥e6 (in my 14.¤c3!
opinion 10...¦e8! is the most precise move) 
11.¤xc4! dxc4 12.d5 ¤xd5 13.¤xd5 ¥f6 This
has been played in numerous games. White
 
has achieved good results after 14.¤xf6†  
14...£xf6 15.¥xc6 bxc6 16.£d4, but it looks     
quite drawish, and I think 14.¦c1!? ¥d4    
15.¥xd4 ¥xd5 16.e4 ¥e6 17.¥c5 is more     
interesting.    
10...cxb3 11.£xb3 £b6!?
  
11...¤a5 was one of Kasparov’s suggestions     
here, but it seems that instead of 12.£a4, 
as given by Kasparov, White can just play 14...¦d8
12.£d3 and it is difficult to find counterplay 14...¥e6? 15.¤b5 a6 16.¤c7 ¦ac8 17.¤xe6
for Black: 12...¤e4!? 13.¤d2 (13.¥xe4 fxe6 18.¥h3±
dxe4 14.£xe4 ¦e8 is not good for White) 14...¥f5 15.g4! ¥xg4 16.¤xg4 ¤xg4
13...f6! 14.¤ec4 ¤xd2 15.¤xd2 b6 17.¤xd5 ¤xd5 18.¥xd5 ¤xe3 19.fxe3 ¥g5
16.¦fc1 ¥e6 17.£b5 White still has some 20.¢f2 ¦fe8 21.¦c3 ¦e7 22.e4±
pressure. 14...a6 15.¤a4! ¥f5 16.¦c7 ¥d6 17.¦xb7
¤c2 18.¦c1 ¦ac8 19.¤b6 ¤xe3 20.¤xc8
¥xc8 21.¦b6! ¤f5 22.e3, planning 23.¥h3±.
15.¤b5 a6
 Chapter 2 – Evolution of the Isolated Pawn 77

 Kasparov gives an exclamation mark to this

  move. I do not think it is a good move, though
  it is probably the only move still to play for an
    advantage. Korchnoi was a player known for
   his fighting spirit but, like many players before
him, he was also trapped into feeling that
     you cannot lose a game against the Tarrasch
    Defence.
  
     Much safer was 16.¤c2 ¤xc2 17.¦xc2 ¥f5
 18.¦cc1 ¦fc8 19.¥g5 ¦xc1† 20.¦xc1 ¥a3
16.¤c7! 21.¦c7 ¥d6 22.¦c1 ¥a3, with a draw, but
Here Kasparov gave only 16.¦c7, but after Viktor’s stubborn attitude did not allow him to
16...¥f8! both 17.¤xf7 ¦d7 and 17.¦xf7 make an early draw, even in a match situation.
¤c2 are unclear.
16...¦b8 17.¥d2 ¥f5 18.¦a4 ¤c2 19.e3 ¥d6 16...¥xb4 17.¤d3
20.g4 With the plan of continuing with 18.¤c5.
Black’s position becomes critical. This is easy 
to figure out nowadays with the help of powerful
software, but back then this just appeared to be   
a complicated endgame. Korchnoi made a very  
good choice and outsmarted Kasparov in the
opening, but he did not have enough energy
   
to work the details out afterwards.    
     
    
    
         
    
     17...¥d6?!
Kasparov is critical of this move and prefers:
    17...¥xa3 18.¦xa3 ¥f5
   But here Kasparov ignored the most logical
19.¤f4! ¦fc8 20.¦c5 ¥e6 21.h4 h6 22.b4
     g5 23.hxg5 hxg5 24.¤d3 when White still
 has better chances, and gave instead:
14...a6! 15.¥d2 ¦b8 19.b4 ¦bc8 20.¦ac3 ¦xc3 21.¦xc3 ¦c8!
With the idea 16...¥f5ƒ (Kasparov). Black is 21...¥xd3?! 22.exd3 ¦b8 23.¦c7 and White
now safe – all the squares are protected. is better.
22.¦xc8†
16.¥xb4?! 22.¤c5 b6 23.¤a4 ¦xc3 24.¤xc3 ¥e6
25.¤a4 ¤d7 and Black is fine.
78 Grandmaster Opening Preparation

22...¥xc8 23.¤f4 ¥e6


With an equal game.    
18.¤c2 ¥g4?!
  
Better was 18...¥e6 19.¤e3 ¦fc8 20.b4! and   
White still has some initiative.   
19.¢f1     
    
      
      
    
In this position White still has some
    advantage, because with the fixed pawn
    weakness on d5 the bishops are weaker than
White’s bishop and knight, which can attack
   the pawn on d5 twice. This is a very general
  assessment, and because of other factors the
    position is still close to equal. In time trouble
Korchnoi overestimates his chances and makes
 a nervous move forward.
19...¥f5?!
Kasparov was never much of an endgame 27.b5?!
player, and here he shows some anxiety. Better Better was 27.¦ba2.
was 19...¦fc8 20.¤e3 ¥e6.
27...a5 28.b6 ¦c6 29.¦b5
20.¤c5 ¦fc8 21.¤e3 ¥e6 22.b4 ¢f8 23.¦c2 29.¦a4!? ¦a6 30.¦b5 ¦cxb6 31.¦axa5 ¤e4
¢e7 was still equal.
23...b6? is a blunder of course: 24.¤xe6†
fxe6 25.¦xc8† ¦xc8 26.¥h3! ¢f7 27.¦xa6 and 29...a4
White is winning.

24.¢e1 h5    
A typical Kasparov move, with the idea of
...h4 or ...¤g4.
  
  
Not so good was 24...b6 25.¤xa6 ¦xc2  
26.¤xc2 ¦a8 27.b5 ¥d7 28.¦b1 and White
has the better game.    
   
25.¦b2 ¦c7! 26.¤d3! ¦a8!
Black gets ready to meet b4-b5 with ...a5.
  
    

 Chapter 2 – Evolution of the Isolated Pawn 79

30.¤xd5†?! A stronger idea was 35.¦c1, the point being

Viktor was famous for grabbing pawns that Black cannot push 35...a3? because of
and relying on his defensive skills. Here the 36.¢c2 trapping the rook. If instead Black
compensation for the pawn is obvious and he pushes 35...b4 then White has 36.¦b5 keeping
most likely regretted his decision very soon. the balance.

30.¤b4 was a safer move, but after 30...¥xb4† 35...¦c8 36.g4?

31.¦xb4 a3 32.¦b3 a2 33.¦b2 ¢d6 34.¦bxa2 A mistake in time trouble, but I think the
¦xa2 35.¦xa2 ¦xb6 the position does not mistake that effectively cost him the game was
promise any winning chances. made earlier, and this was mainly his attitude
– to win at any cost.
30.¤c5? ¥xc5 31.dxc5 and now Kasparov
wrongly gave only 31...¤d7? 32.¤xd5† ¥xd5 36.¦g5! g6 37.h5 (Kasparov), but after
33.¥xd5 ¦xc5 34.¥xb7 ¦xb5 35.¥xa8 ¤xb6, 37...gxh5 38.¦xh5 a3 Black should win.
again with an equal game.
Instead 31...¦cc8! 32.¦b4 (32.¦a3 ¥d7) 36...a3
32...¦xc5 33.¦axa4 ¦xa4 34.¦xa4 ¦b5 sees More precise was 36...¦c4.
Black emerge with an extra pawn.
37.f4 ¦cc3?!
30...¤xd5 31.¥xd5 ¥xd5 32.¦xd5 ¦xb6 More precise was 37...¦c4 and Black just
33.¦xh5 gradually pushes his pawns, for example 38.e3
It seems strange that a player of the calibre of b4 39.¦b5 ¢d7! or 38.¦d5 ¢e6 39.e4 ¦cc3
Korchnoi would make such a decision, going 40.¤c5 ¥xc5 41.dxc5 b4.
after the h5-pawn which has no relevance to
the situation on the board. The reality is that 38.¦d5 ¢e6 39.¦h5
now Black has two passers, along with a strong
bishop.

     
       
      
       
       
      
       
        
     
39...b4?
 Kasparov is correct in pointing out that this
33...¦b3! 34.¢d2 b5 35.h4 was a big mistake. Allowing the white rook to
White’s plan is to push the pawns on the return to the game was a bad idea.
kingside, but this does not work.
80 Grandmaster Opening Preparation

39...¦xd3† 40.exd3 ¥xf4† 41.¢e2 ¦b2† ¢f5 48.¦b6 ¦c2† 49.¢e3 ¢xg5
42.¢f3 ¥d6 and now instead of 43.¦h8 ¢d5
44.¢e3 ¦g2!–+, given by Kasparov, much

stronger is 43.d5†! ¢d7 44.¦h8 b4 45.¦a8     
and suddenly there is enough counterplay.    
39...¦c4! was the correct route to the win, as
on earlier moves.     
   
40.¦a5 ¦xd3†
40...¦b2† 41.¤xb2 ¥xf4† 42.e3 ¥xe3†     
43.¢e2 axb2 44.¦b1 would also lead to a draw    
after either 44...¥c1 45.¦b5 b3 46.¦e5† ¢d6
47.¦f5 or 44...¥xd4 45.¦xb2 ¦e3† 46.¢d2
   
¦g3 47.¦b1 ¥c3† 48.¢c2 ¦xg4 49.h5.     
41.exd3 ¥xf4† 42.¢e2 ¦c3

50.d5
 Also good enough for a draw was: 50.¦b5† f5
     51.d5 ¢xh5 52.d6 ¦c5 53.¦xb2 ¦d5 54.¦b6
    g5 55.d4 ¢g4 56.¦c6 (56.¢d3? ¢f3–+)
56...¢g3 57.¦c5!=
   
     50...¢xh5 51.¢d4
51.d6? loses to 51...¦c6!.
   
    51...g5 52.¦b8
Instead 52.¢e5! was a relatively easy draw:
    52...g4 53.d6 ¦c5† 54.¢f4 ¦d5 55.¦xb2 ¦xd6
     56.¦h2† ¢g6 57.¢xg4 With a simple draw.
 52...g4 53.d6
43.g5?!
More precise was: 43.¢f3 ¥c1 (43...¦c1 
44.¦1xa3!= or 43...¥d6 44.¢e4 b3 45.¦5xa3
b2? 46.d5†!+–) 44.¢e4 ¢d6 45.¦a6† ¢d7
    
(45...¢c7 46.d5 ¢b7 47.d6! is dangerous    
for Black, though 47...b3 should still draw)     
46.¦a7†=
   
43...¥c1 44.h5?    
44.¦a4! b3 45.¦1xa3 ¥xa3 46.¦xa3 b2
47.¦a6† ¢f5 48.¦b6 and compared to the    
position reached in the game, here the g-pawn    
is still defended.
44...b3 45.¦5xa3 ¥xa3 46.¦xa3 b2 47.¦a6†
    

53...¦c6!?
 Chapter 2 – Evolution of the Isolated Pawn 81

One last try to win the game.
53...¦c8 54.¦xb2 (54.¦xc8? b1=£ 55.d7
£b6† 56.¢c3 £a5† 57.¢c4 £a4†–+)
54...g3 55.¢e3! ¢g4 56.¦b4† ¢h3 57.¦b5
¢h2 (57...¦d8 58.¢f3) 58.¦h5†= is a draw.
54.¢e5 ¦c5†

It is obvious that Korchnoi missed this move
when playing 55.¢f6.
Instead, 59...¢f4 60.¢e7 ¦g6 61.¦g2! ¢e4
62.¢f7 ¦g4 63.¢f6 is a draw.
60.¢e7 ¦d5
60...¦e6† 61.¢f7 ¦h6 62.¦g2 ¢f4 63.¢e7
¦h2 64.¦xg3 ¢xg3 65.d5 is also a draw.

     61.¦d3!
    61.¦g2? ¢f4 62.¦d2 ¢e4 63.¦g2 ¢f3
64.¦d2 g2 and Black wins.
    
 T K  61...¢f4 62.¢e6 ¦g5
  +  
        
         
        
     
55.¢f6     
55.¢d4 was a draw of course.
   
55...g3 56.¦xb2 ¦d5 57.¢xf7 ¦xd6 58.¦d2     
¢g4 59.d4
59.¢e7? loses to 59...¦d4 60.¢e6 ¢f4.
    
 
63.d5?
     When you are lacking energy, you can miss
    things at the end of the game. This last mistake
was probably due to time trouble, or perhaps
     the exhausted Korchnoi just thought that
     every move would make a draw and he missed
    Black’s reply.

     There is still some play left after the correct

     63.¦d1, although White’s subsequent moves
would be almost automatic, not leaving
     much chance for any further error. The way
 the young Kasparov pulled out victory at the
59...¢f5! last moment reminds me of Fischer, who also
fought to the last pawn and won many drawish
82 Grandmaster Opening Preparation

positions. Vasily Smyslov – Garry Kasparov

63...¦g6†! Vilnius (2) 1984

First pushing the king away.
1.d4 d5 2.¤f3 c5 3.c4 e6 4.cxd5 exd5 5.g3
64.¢e7 g2 65.¦d1 ¢e5!–+ 66.d6 ¦e6† ¤f6 6.¥g2 ¥e7 7.0–0 0–0 8.¤c3 ¤c6
67.¢d7 9.¥g5 cxd4 10.¤xd4 h6 11.¥e3 ¦e8 12.a3
 ¥e6 13.¢h1!?
This was the second game of the match, and
     facing a novelty – and a deep one – was very
    unpleasant for Kasparov. The plan is to prepare
f2-f4, and in some lines the bishop from e3
    can retreat to g1.
     We saw 13.£b3 above in Korchnoi –
Kasparov, London (2) 1983.
     
      
      
      
    
67...¦xd6† 68.¦xd6 g1=£ 69.¦e6† ¢f5
70.¦d6 £a7† 71.¢d8 ¢e5 72.¦g6 £a5†     
73.¢d7 £a4† 74.¢e7 £h4† 75.¢f8 £d8†
76.¢f7 ¢f5 77.¦h6 £d7† 78.¢f8 ¢g5
    
0–1   
This game effectively decided the match.
 
Korchnoi immediately lost the next game and 
the match was over. As in 1965, when Keres 13...£d7?!
lost his match against Spassky in Riga, this was Facing a novelty, Kasparov could not find
the decisive match in finding the next world the correct remedy.
championship challenger. The final match,
against the much older Vasily Smyslov, was 14.¤xe6 fxe6 15.f4 ¦ed8!?
just a formality for Kasparov. Kasparov decides that his rooks belong on
c8 and d8.
In the Candidates final match, Kasparov was 15...d4? 16.¤e4! is good for White.
not afraid to repeat the Tarrasch. Smyslov had
time to prepare, and repeated the line with 16.¥g1 ¦ac8 17.£a4 ¢h8 18.¦ad1
12.a3, coming up with a concrete plan. White has a clear advantage, because Black
cannot prevent the positional threat of e2-e4.
Kasparov was lucky to save the game on move
41.
...½–½
 Chapter 2 – Evolution of the Isolated Pawn 83

Kasparov’s team found a counterplan against 14...¥h5

13.¢h1. It may have taken a few days, because
in the next two games with Black Kasparov
avoided the Tarrasch, but he was back and
ready in Game 8.
Vasily Smyslov – Garry Kasparov
Vilnius (8) 1984
Kasparov repeated this move in the next
game as well, but the simple retreat to d7
gives Black a satisfactory position without
any hassle. I believe this is why Karpov’s team
dismissed this line entirely.
15.¥g1 £d7!
Black has a good game here.

1.d4 d5 2.¤f3 c5 3.c4 e6 4.cxd5 exd5 5.g3 16.£a4

¤f6 6.¥g2 ¥e7 7.0–0 0–0 8.¤c3 ¤c6 16.¤xc6 bxc6 17.¤a4 was played in Game
9.¥g5 cxd4 10.¤xd4 h6 11.¥e3 ¦e8 12.a3 10, but Black did not have any problems there
¥e6 13.¢h1 ¥g4! either.
I do not know who came up with this, but
it is a genius idea. Black has just developed 16...¥c5! 17.¦ad1 ¥b6 18.¦fe1 ¥g6!
the bishop to e6 and is now moving it again.
Moreover, my computer likes this move as 
well, but back then (to my knowledge) there  
was no computer analysis being used.
 
   
     
      
       
      
      
     
   Black has an excellent position. The game
  was drawn on move 28.
...½–½

14.f3!?
I do not believe that Smyslov’s team analysed
Kasparov’s novelty at all, which is why Smyslov
decided to continue with his original plan. By
the next game, he had had time to grasp the
point of Black’s last move at home, but he still
made the same move. Other possible moves
are 14.£b3 and 14.h3.
84 Grandmaster Opening Preparation

Vasily Smyslov – Garry Kasparov 

  
Vilnius (12) 1984
   
1.d4 d5 2.¤f3 c5 3.c4 e6 4.cxd5 exd5 5.g3   
¤f6 6.¥g2 ¥e7 7.0–0 0–0 8.¤c3 ¤c6
9.¥g5 cxd4 10.¤xd4 h6 11.¥e3 ¦e8 12.a3
   
¥e6    
     
    
    
   
    16...a6?!
Kasparov makes a big mistake here, probably
     underestimating White’s potential. The
     position looks closed, but White has some
threats with f4. Much stronger was to play
   immediately 16...¤a5! 17.f4 ¤c4 18.¥c1
   £b6, and Black is in time to take over the
initiative.

In this 12th game Smyslov suddenly took 17.f4
on e6. The plan with 13.¢h1 did not work White should have played 17.£h4!.
as hoped for, and no more new ideas about
how to beat the Tarrasch had been created.

He was already losing the match 4-7, and   
usually at these moments the main concern    
is how to push yourself to play the next game  
– preparing a concrete opening line is no    
longer so important. Nobody wants to lose
again, but Smyslov tried to beat the Tarrasch
    
by just playing the position; this quickly     
backfired.   
 
13.¤xe6?! 
13.£c2!? I like this move. 13...£d7 14.¤xe6 Black has serious problems here. The attack
fxe6 15.¦ad1 In the game White transferred on the light squares is very strong, and the
his queen to the kingside via a4, but here the white queen has a fine square on h3 in many
queen is placed very well, controlling the weak lines.
b1-h7 diagonal.
17...¤a5! 18.f5 b5!
13...fxe6 14.£a4 ¦c8 15.¦ad1 ¢h8 16.¢h1 Other moves are weaker: 18...¤c4? 19.¥c1±
or 18...¦c4? 19.£c2 e5 20.£d2! d4 21.¥xh6
 Chapter 2 – Evolution of the Isolated Pawn 85

gxh6 (21...dxc3 22.¥xg7†±) 22.£xh6† ¢g8 Instead of 22...¥xc1, Black must try 22...¤f6,
23.g4 and White’s attack is decisive. but here also after 23.£d4 ¥xc1 24.¦xc1 ¤xb2
25.¤xb5 White has a big advantage.
19.£h4
 20...¤c4 21.¥c1 ¥g5!
Very well played. Even in the middlegame,
   pushing the pawns too early is sometimes not
     a good idea. The difference between the moves
17.£h4 and 17.f4 becomes obvious now. The
   point is that without pushing his pawn further
  by f4-f5, White could not transfer his queen.
     This is why 17.f4 was a mistake. This kind of
mistake is easy to understand: Smyslov needed
     to play for a win and an active move like f2-f4
   is easier to make in such a situation than a slow
move like £h4.
  
  
19...¤g8!
After 19...¤c4? 20.¥xh6 ¤h7 21.¥xg7†     
¢xg7 22.£g4† ¢h8 23.fxe6 White has a
strong attack.
  
 
20.£h3    
More active and much stronger was 20.£g4!
and after 20...¤c4 21.¥c1 ¥g5    
   
 m  
+  
 +  22.fxe6 ¥xc1 23.¦xc1
+V  23.¦xd5?! £b6 24.¦xc1 ¤xb2 leaves White
 m q+ under pressure.
+ 
23...¤e3 24.¤xd5!
  24.¦g1 ¤f6µ ends any White counterplay.
B
 24...¤xf1
22.fxe6! White is close to winning. Kupreichik After 24...¦xc1? 25.¦xc1 ¤xd5 26.¦d1
in his comments missed that after 22...¥xc1 White regains the piece with advantage, for
23.¦xd5 £b6, instead of 24.¦d7? White has example 26...¤gf6 27.e4+– or 26...¤e3
a stronger move: 24.¦f7 and Black is helpless, 27.¦xd8 ¦xd8 28.¢g1+–.
as after 24...¥g5 White has 25.¦xg5 hxg5
26.£h5 ¤h6 27.£g6 and mate in a few moves. 25.¦xf1 ¦f8
86 Grandmaster Opening Preparation

 This hastens the end, but White was lost in

   any case.
     36...¤e7 37.b4 ¢h7
   Zugzwang!
   
         
       
     
     
   
26.¤f4?!
26.¦xf8! £xf8 27.¥f3 ¤e7 28.£h5 and    
White may survive, but this was the last chance   
for Smyslov to complicate the match situation;
a draw was not really an option.
    

26...¤e7 27.£g4? g5! 28.£h3 ¦f6! 29.¤d3 38.¢h2 ¦d8 39.e5 ¦xd3 40.¥xd3† £xd3
¦xf1† 30.¥xf1 ¢g7 31.£g4 £d5† 0–1
 In 1984 the world championship match
    between Karpov and Kasparov took place
     in Moscow. I was a student at Tartu State
University at that time and I was commenting
   on some of the games in a local newspaper. I
   was studying psychology, but I did not pay too
    much attention to the psychological aspect of
the match – the concrete moves on the board
    were the main subject. Now Kasparov was
    facing a really formidable opponent compared
with his opponents in the Candidates matches.
   The first test of the Tarrasch Defence failed.

32.e4
32.¥g2? is quickly mated: 32...¦c1†!
33.¤xc1 £d1† 34.¥f1 £xf1#

32...£d4 33.h4 ¦f8 34.¥e2 £e3 35.¢g2

35.hxg5? loses immediately to 35...h5!.

35...¤g6 36.h5?!
 Chapter 2 – Evolution of the Isolated Pawn 87

Anatoly Karpov – Garry Kasparov 

 
Moscow (7) 1984
  
1.d4 d5 2.c4 e6 3.¤f3 c5 4.cxd5 exd5 5.g3     
¤f6 6.¥g2 ¥e7 7.0–0 0–0 8.¤c3 ¤c6 9.¥g5
cxd4 10.¤xd4 h6 11.¥e3 ¦e8 12.£b3
  
This move, prepared by Karpov, refuted    
the Tarrasch, at least at world championship     
level. Kasparov tried it once more, losing again
in the same line, and then abandoned it for 
good. From a theoretical point of view, there    
is nothing wrong with Black’s position, but on
the other hand White also has a lot of forcing

15.¤xe7†
lines. This removes the uncertainty that, as we
In their next game Karpov played 15.¥d4!?.
saw earlier, is the main concept for Black –
playing a slightly worse position, but keeping
15...¦xe7 16.¦ad1
White in tension.
Black’s position here is not so bad, and the
 reputation of the move 14...¦c8 as not very
 good might be just because Kasparov lost the
game with it.
  
    16...£e8! 17.h3 ¥h5 18.¥xd5
    Karpov was considered a very strong
positional player. He of course successfully
     played all kinds of positions, but actually he
    was a very concrete player.

  

     
   
12...¤a5     
Forced; there is no good way to defend the
d5-pawn.
  
    
13.£c2 ¥g4 14.¤f5 ¦c8?    
Neither during this game nor after working
with his team, did Kasparov find the alternative  
14...¥b4! very attractive. This is the critical   
move and has become the main line.

18...¥g6
88 Grandmaster Opening Preparation
18...b5! 19.a3 ¥g6 20.£c1 ¤xd5 21.¦xd5
a6 and the inclusion of ...b5 and a2-a3
dramatically weakens the light squares. Black
has more than enough compensation for the
pawn. 22.¥xh6!? ¦xe2 (or 22...gxh6 23.£xh6
£f8 24.£g5 ¢g7÷) 23.£d1 ¦e6 24.¥c1 f6©
19.£c1 ¤xd5 20.¦xd5 ¤c4 21.¥d4
White has an extra pawn, but the opposite-
coloured bishops make the position difficult to
win.
21...¦ec7 22.b3 ¤b6 23.¦e5 £d7 24.£e3
f6 25.¦c5 ¦xc5 26.¥xc5 £xh3 27.¦d1

  
   
   
    
    
  
  
   

White gave back the pawn, hoping for an
initiative in the centre. Kasparov was probably
somewhat relieved and made his trademark
move, which in this case is unfortunately a
significant mistake.
27...h5?
Once Josif Dorfman, one of Kasparov’s
many seconds, was talking about Kasparov. He
admired Kasparov’s talent, but he pointed out
some of the characteristics of Kasparov’s style.
One was that, from the very beginning of his
chess career when Garry was very young, he
liked to push his rook pawns. The idea is to
gain space for a future endgame, which might
be very valuable, especially in rook endgames.
I do not know how many times Kasparov
really pushed his a- or h-pawns early in the
middlegame, but in this game it backfired.
Instead, 27...¥e8! was a good option. The
bishop needs a role and transferring it to c6
was the correct plan.
28.¦d4 ¤d7

  
  
   
   
    
  
  
    

29.¥d6!?
29.¥xa7? ¤e5 30.¤d5 ¤g4 31.¦xg4 £xg4
32.¤e7† ¢h7 33.¤xc8 £xc8 is drawish of
course. When using a computer to check the
lines, one needs to look to the very end. A line
that seems to promise a good advantage might
lead to a position which is actually very close
to a theoretical draw. Years ago, when I was
playing against a computer for fun, I always
found that lines where it had an extra pawn
in a rook endgame would be overestimated by
the computer.
29...¥f7 30.¤d5 ¥xd5 31.¦xd5 a6
31...¤f8 32.¦c5 b6 33.¦xc8 £xc8 34.£e7
a6 and Black is slightly worse. Kasparov
was never a very good defender of inferior
positions without counterplay, where there is
no immediate solution to clarify the situation.
32.¥f4 ¤f8 33.£d3
 Chapter 2 – Evolution of the Isolated Pawn 89

 40...g5 41.£a8 ¢g7 42.£xa6

   1–0
    Anatoly Karpov – Garry Kasparov
   
   Moscow (9) 1984

     1.d4 d5 2.c4 e6 3.¤f3 c5 4.cxd5 exd5 5.g3

  ¤f6 6.¥g2 ¥e7 7.0–0 0–0 8.¤c3 ¤c6 9.¥g5
cxd4 10.¤xd4 h6 11.¥e3 ¦e8 12.£b3 ¤a5
   13.£c2 ¥g4 14.¤f5 ¦c8
     As mentioned above, 14...¥b4! is critical.
 This move had already been seen in practice
during 1984, but the strong Hungarian GM
33...£g4?
Lajos Portisch had won two games as White,
33...¤e6! 34.¥e3 £g4 and Black has enough
and taking into account that the Hungarian
counterplay, because now there is no f2-f3, as
chess school was famous for its opening
played in the game.
preparation, this may have led to Kasparov’s
team rejecting the move. After this game the
34.f3 £g6 35.¢f2 ¦c2?
Tarrasch dropped out of Kasparov’s repertoire,
This pseudo-active move is another mistake.
at least at world championship level.
35...£xd3 36.¦xd3 ¢f7 was a simple draw.
15.¥d4
In the previous game, we saw that the
36.£e3 ¦c8
opening was not the reason for Kasparov’s
36...¦xa2? 37.¦d8 and White is winning.
failure. This blockading move is better than
15.¤xe7, one reason being that we prefer
37.£e7 b5 38.¦d8 ¦xd8 39.£xd8 £f7
knights when dealing with isolated pawn
40.¥d6
positions.
Black in now in zugzwang and he is losing all
his pawns on the queenside. 
  
      
        
      
      
         
   
      
     
 15...¥c5
90 Grandmaster Opening Preparation

This is the best move and it does not look This endgame is much better for White,
good, because we do not want to trade bishops. but it is very difficult to win because winning
This is the main reason I do not like the line the isolated d-pawn is not enough on many
with 14...¦c8. occasions. Black always has some counterplay
on the c-file. Also, rook endgames have drawish
16.¥xc5 ¦xc5 17.¤e3! tendencies even when a pawn down.
Now White has achieved a nice regrouping,
and it is one of Karpov’s best games because of 27.h3 h5 28.a3 g6 29.e3 ¢g7 30.¢h2 ¦c4
the fantastic endgame he won. I give the rest 31.¥f3 b5 32.¢g2 ¦7c5 33.¦xc4 ¦xc4
of the game here with just a few comments, 34.¦d4 ¢f8 35.¥e2
because this game has been thoroughly
annotated already in numerous books. 
     
      
    
      
       
       
       
     
    
 Now here is the interesting moment. The
computer approves of Kasparov’s choice in
17...¥e6 18.¦ad1 £c8 19.£a4 ¦d8 20.¦d3 the game. However, keeping the rooks on the
a6 21.¦fd1 ¤c4 22.¤xc4 ¦xc4 23.£a5 board might be the correct human decision. As
23.£b3! £c6 24.e4 d4 25.e5 is also good I have already mentioned, the rook endgame a
for White. pawn down is not lost yet, and it is much easier
23...¦c5 24.£b6 ¦d7 25.¦d4 £c7 26.£xc7 to defend in practice.
¦dxc7
 35...¦xd4 36.exd4 ¢e7 37.¤a2 ¥c8
38.¤b4 ¢d6 39.f3 ¤g8 40.h4 ¤h6 41.¢f2
    ¤f5 42.¤c2 f6 43.¥d3 g5 44.¥xf5 ¥xf5
   45.¤e3 ¥b1 46.b4
Karpov was most likely familiar with the
   following game, in which Fischer demonstrated
    how to play this kind of endgame. Karpov
definitely prepared against Fischer when
     their match was in the negotiation phase,
     although unfortunately the match never
  materialized.

   

 Chapter 2 – Evolution of the Isolated Pawn 91

Saidy – Fischer, New York 1964 contender should know. Once the move is
 made on the board, everybody can see the idea,
     which is to keep the route open for White’s
king, avoiding a blockade.
   
  47...hxg3† 48.¢xg3 ¢e6 49.¤f4† ¢f5
  50.¤xh5
    Black’s king cannot prevent the entry of the
   white king because of the threat of 51.¤g7†,
when the knight would invade Black’s camp.
   
     50...¢e6 51.¤f4† ¢d6 52.¢g4 ¥c2
 53.¢h5 ¥d1 54.¢g6 ¢e7 55.¤xd5†
38.¢e2 f4 39.¥f2 ¤g7 40.h3 ¤f5 41.¢d3
g4 42.hxg4 hxg4 43.fxg4 ¤h6 44.¥e1 ¤xg4

45.¥d2 ¢f5 46.¥e1 ¤f6 47.¥h4 ¤e4 48.¥e1     
¢g4 49.¢e2 ¤g3† 50.¢d3 ¤f5 51.¥f2 ¤h4
52.a5 ¤xg2 53.¢c3 ¢f3 54.¥g1 ¢e2 55.¥h2
    
f3 56.¥g3 ¤e3 0–1   
  
It is of course difficult to find exactly the
same position in the database, and the Saidy –     
Fischer game had some slight differences.    
46...gxh4?
    
Kasparov definitely missed White’s next move.    
47.¤g2!

55...¢e6
 White has achieved the maximum from the
     position, but it is still not clear how to win
after the more stubborn defence of 55...¢d6
     56.¤e3 ¥xf3 57.¢xf6 ¥e4.
    56.¤c7† ¢d7 57.¤xa6 ¥xf3 58.¢xf6 ¢d6
  59.¢f5 ¢d5 60.¢f4 ¥h1 61.¢e3 ¢c4
     62.¤c5 ¥c6 63.¤d3 ¥g2 64.¤e5† ¢c3
    65.¤g6 ¢c4 66.¤e7 ¥b7 67.¤f5 ¥g2
68.¤d6† ¢b3 69.¤xb5 ¢a4 70.¤d6
    1–0
    Kasparov prepared and used the Tarrasch
 Defence for a very short period. After two
This is the picturesque position every chess consecutive losses to Karpov, he never again
aficionado or future world championship used the Tarrasch against top opposition.
92 Grandmaster Opening Preparation
In 1987-88 he won three games with it against
less strong grandmasters, which is a reminder
that there is nothing wrong with the defence
itself.
I started to play the Tarrasch Defence in 1977
after the training camp with Vladimir Yurkov,
and it was my main defence with Black in
the semi-final of the USSR Championship
in 1983, which was my first big tournament
success. However, I did not use it in the 1984
USSR Championship. As in Kasparov’s case,
there was nothing wrong with the opening. I
had suffered a painful loss against Lev Psakhis
in 1984, and such an occurrence usually
triggers a change of opening for me. I never
collected or tracked my statistics with different
openings and I liked to change them a lot. One
of the main reasons was that I did not have a
very good team of coaches with whom I could
work out my opening repertoire.
The Tarrasch is a good opening per se, but
most of the time I was playing for qualification
spots. Usually only the top places in these
tournaments are counted as success. The Soviet
Union was by far the strongest chess country
in the world, but to succeed you needed to
qualify from a very strong field to advance
further.
Keeping the Tarrasch Defence in my
opening repertoire no longer made sense when
I was looking to become first among equals;
I looked for more aggressive openings. From
time to time I experimented with the King’s
Indian and the Grünfeld, though finally the
classical approach prevailed and I had many
successful years with this approach.
The Tarrasch Defence disappointed me
once, however, when in a must-win situation
in New Delhi in 2000, I could not beat rising
star Alexander Grischuk with the white pieces.
Jaan Ehlvest – Alexander Grischuk
New Delhi/Teheran (4.4) 2000
1.d4 d5 2.c4 e6 3.¤c3 c5 4.cxd5 exd5 5.¤f3
¤c6 6.g3 ¤f6 7.¥g2 ¥e7 8.0–0 0–0 9.¥g5
cxd4 10.¤xd4 h6 11.¥e3 ¦e8


  
   
   
    
    
 
  

I knew that Grischuk was most likely going
to play the Tarrasch and I prepared something.
The FIDE knockout system does not provide
much time between the mini-matches, so I just
went for a line which theoretically promised
something, but in practice this was not a very
clever choice on my part.
12.£b3
Like Karpov, I was looking for a forcing,
concrete solution to get some advantage. This
game was played at a shorter time control. I
had lost the first rapid game and now only a
win could keep me in the match. 12.£b3 is a
good move, but in this situation it is probably
not a good idea.
However, on the morning of the match I
did not know that I would need to win this
particular game. Ideally, a player should have in
his repertoire different lines to play in different
situations. One line which looks terrific overall
may not work at all in certain situations.
 Chapter 2 – Evolution of the Isolated Pawn 93

Still, I hoped that I could crack my young


opponent with this line.   
12...¤a5 13.£c2 ¥g4 14.¤f5 ¥b4 15.¥d4
   
¥xc3 16.¥xc3 ¦xe2 17.£d3 ¦e8 18.¤e3    
¥e6    
     
      
     
       
    
     27.¦d4!? f5 28.¦ed1?
    28.¦d2!? ¤g6 29.£b4 £f6 30.¦ed1 f4 and
Black has enough counterplay.
   28...¤c6 29.£xd8 ¦exd8
    Now there are no winning chances for
 White, and the match was over.
This position can be practically forced after 
12.£b3, and I was hoping to get at least a
slightly better endgame after winning back the    
d-pawn. The endgame of bishop against knight    
with pawns on both wings should give White
some winning chances. But now, facing this
  
position in my particular situation, I started   
to play creative chess because the endgame did
not seem so good when I had to win.
    
    
19.£b5!? b6 20.¦ad1 ¦c8 21.¦fe1!?   
After 21.¥xf6 £xf6 22.¤xd5 ¥xd5 23.¥xd5
White has some advantage, but with careful    
defence Black can hope to defend. 
30.¦4d2 d4 31.¤c2 d3 32.¤e3 ¤e5 33.b3
21...¥d7 22.£f1 ¥e6 23.£a6 ¤c6 24.£a4 ¤g4 34.¤f1 ¦c2 35.h3 ¤f6 36.¦xc2 dxc2
¤e7 25.¥xf6 gxf6 26.£h4 ¢g7 37.¦c1 f4 38.¦xc2 ¦d1 39.¢h2 ¤d5
Short of time (it was a rapid game with 30 40.gxf4 ¤xf4 41.¤e3 ¦a1 42.h4 ¤xg2
minutes plus 10-second increments), I was not 43.¢xg2 a5 44.¦b2 a4 45.¤c2 ¥d5† 46.f3
sure that I would have enough after 27.¥xd5! ¦d1 47.¤e3 ¦d3 48.¤xd5 a3 49.¦c2 ¦xd5
¤xd5 28.¦xd5 £xd5 29.¤xd5 ¥xd5. It is 50.¢g3 ¦d4 51.h5 ¢f6 52.¦c6† ¢g5
difficult to win this position, but objectively 53.¦xb6 ¦d2 54.¦b5† f5 55.f4† ¢xh5
this was the best opportunity I had. 56.¦xf5† ¢g6 57.¦a5 ¦xa2 58.¦a6† ¢f5
59.¦xh6 ¦b2 60.¦h5† ¢f6 61.¦a5 ¦xb3†
94 Grandmaster Opening Preparation

62.¢g4 ¦b6 63.¦xa3 ¦b5 64.¦a6† ¢f7 This looks like a strange move, but the idea
65.f5 ¦b1 is to gain a tempo and vacate the d1-square for
½–½ the rook as quickly as possible.

The positional approach should be best


against the Tarrasch. There are many early  
deviations where White tries to punish Black,  
but these lines are not too important, in part
because Black often avoids the straightforward     
Tarrasch move order. After 1.d4 d5 2.c4 e6    
3.¤c3 c5, White can try 4.cxd5 exd5 5.dxc5!?
which can be unpleasant for Black for practical
    
reasons. White needs only to prepare some    
forcing lines and Black might end up in a
slightly worse endgame with no counterplay,

for example: 5.dxc5 d4 6.¤a4 ¤c6 (6...b5!?   
7.e3 bxa4 8.exd4 with a complicated position) 
7.e3 ¥xc5 8.¤xc5 £a5† 9.¥d2 £xc5 10.exd4 9...¥b6 10.¤c3 ¤c6 11.¥g5 ¥e6 12.£a4?
£xd4 11.¥c3 £xd1† 12.¦xd1 ¤f6 White has 12.¦ad1 was more consistent, and I believe
a small but pleasant advantage in the endgame. I just could not remember the line correctly.
Playing with Black you can switch to the 12...h6 13.¥h4 £e7 14.£c1 (14.£a4 ¦fe8
Tarrasch much later – after 1.d4 ¤f6 2.c4 e6 15.¦xd5!? eventually led to a draw in Kasparov
3.g3 d5 4.¤f3 ¥e7 5.¥g2 0–0 6.0–0 c5, for – Behrhorst, Hamburg [simul] 1987) 14...¦fd8
instance. Preparing one surprise line against 15.e4 d4 16.¤a4 ¥g4 17.£f4 ¥xf3 18.¥xf6
the Tarrasch is not enough. £xf6 19.£xf6 gxf6 20.¥xf3 ¤e5 21.¥h5
When I had the Tarrasch in my opening d3 22.¤c3 ¥d4 23.¤d5 ½–½ Veingold –
repertoire as Black, I was worried about some Campos Gambuti, Marchena 1991.
surprise lines, and seeing one idea I tried it
myself as White. I believe the idea belongs to 12...h6 13.¥h4 ¥c5 14.¦ad1 g5 15.¤xg5
Estonian IM Alexander Veingold and was even
used once by Kasparov himself, though only 
in a simul.   
Jaan Ehlvest – Giorgi Giorgadze   
  
Minsk 1986
   
1.d4 d5 2.c4 e6 3.¤f3 c5 4.cxd5 exd5 5.g3    
¤f6 6.¥g2 ¥e7 7.0–0 0–0 8.dxc5 ¥xc5
9.£c2
    
The USSR Team Championship seemed  
to me a good opportunity to try some
extraordinary ideas. My teammates probably
  
did not like it, but I did not ask them. 
 Chapter 2 – Evolution of the Isolated Pawn
This sacrifice is more common in the open
games, such as the Italian Game, in which the
current World Champion obtained a huge
advantage against Sergei Karjakin at Wijk aan
Zee in 2017.
15...hxg5 16.¥xg5 ¥e7 17.¥xf6?!
White could force a draw with 17.¥h6
¦e8 18.£f4 ¤h5 19.£f3 ¤f6 20.£f4, which
would have been the logical end to the game.
17...¥xf6 18.¤xd5 ¥xb2 19.£b5 ¥g7
20.£xb7
White has three pawns for the bishop, but
Black has the better game because the pawns
are not dangerous and in the open position the
bishop is much stronger.
20...¦c8 21.e3
After 21.¤f4 £c7 22.£xc7 ¦xc7 23.¤xe6
fxe6 24.e3 White has good drawing chances,
but this position looked too boring to me and
instead I continued to play aggressive chess
without any basis, and quickly lost the game.
21...£a5 22.f4?! £a3 23.¢h1 ¥g4 24.¦de1
£xa2 25.f5 ¥e5

  
  
   
  
   
    
  
  

26.¦c1?
Here I had the very strong move 26.£b5,
but it was difficult to find in time trouble.
With a computer it looks easy, though.
95
26...¥e2 27.¦f2 ¦fd8 28.¢g1 £a6 29.£b1
¥d3 30.£b3 ¤d4 31.¦xc8 ¦xc8 32.¤e7†
¢f8 33.£b4 ¦c1† 34.¥f1 ¥d6 35.¤g6†
fxg6 36.£xd4 gxf5 37.£h8† ¢e7 38.£g7†
¢d8 39.£h8† ¢d7 40.£g7† ¢c6
0–1
There are various other lines and ideas on
the White side of the Tarrasch of course, but
the decline in the popularity of the Tarrasch
Defence is not connected with the quality of
the opening. In my opinion, people rely more
on computer evaluations nowadays, and it is
much easier to learn some forcing lines in the
Semi-Slav than to learn the Tarrasch.
The Tarrasch is not considered a classical
opening system. “Classical” in my interpretation
means an approach where the very first
principles of play commonly acknowledged by
chess authorities are followed. For instance, the
rule about the centre: one should defend the
centre from the very first move, and it is not
wise to postpone the fight for the centre for
very long. From this we can easily understand
which openings should be classified as classical
– 1.e4 e5 and 1.d4 d5 undoubtedly involve the
classical approach.
The Tarrasch Defence is not classical in the
sense that Black voluntarily creates an isolated
pawn. This is why at the very beginning the
Tarrasch Defence did not find many followers.
The ideas that developed during the evolution
of the Tarrasch Defence nevertheless became
classical. This has been somewhat hidden from
the public, so I have just tried to show that
this is the case. There are lot of applications of
these classical ideas.
Applications of classical ideas
Theoretically it is possible to avoid positions
with an isolated pawn. Unfortunately,
professional chess is like any other job –
sometimes you need to study things that
96 Grandmaster Opening Preparation
you do not like or enjoy. Nobody argues that
studying theoretical rook endgames a pawn
down is a waste of time. The argument, even
from a very strong player, that he or she never
loses a pawn does not convince a coach.
Everyone who wants to become a top player
needs to be acquainted with every opening.
A wide chess education raises his or her chess
intellect. Chess professionals do not talk about
the openings so much with strangers, but to
understand and know the opening problems is
valuable, not only for the tournament player,
but also for the commentator, and of course is
the bread and butter of the coach or second.
There are hundreds of opening lines and it
is difficult to grasp all of them. On the other
hand, in trying to understand isolated pawn
strategy from the defensive point of view, the
know-how presented in this book should be of
great benefit. How have I benefited from this
knowledge? Measuring the impact on my level
is not possible (at least today), but there is no
doubt that having access to these secrets when
I was only 15 years old benefited my chess
understanding.
The chances that a player will never have
to play with or against an isolated pawn is
effectively zero, so a few hours spent on this
material must be a good idea. In my practice,
the Lein approach helped me on one occasion
when drawing a game against Vladimir
Kramnik in a super-tournament in Zurich.
Vladimir Kramnik – Jaan Ehlvest
Horgen 1995
1.¤f3
In the second part of the 1990s Kramnik
was one of the best players in the world, and he
finally took the title from Kasparov in 2000.
His opening repertoire was at that time strictly
limited to 1.¤f3 followed by 2.c4. Later a series
of books was written by Alexander Khalifman
about Opening for White according to Kramnik.
Kramnik’s opening repertoire was a little too
timid in my opinion, but on the other hand it
promised him good results without taking too
much of a risk. He himself never made any big
positional mistakes, and when he was playing
with the white pieces there were basically only
two possible results. Every player knew before
the game that it was going to be a long game.
Most likely you would need to defend a very
slightly worse position until the very end;
psychologically, this is a difficult situation.
I, however, had a big plus, already being a
veteran compared with many other players.
I had the experience of playing many games
with Anatoly Karpov, who had the same style
as Kramnik.
1...e6 2.g3 b6 3.¥g2 ¥b7 4.0–0 ¤f6 5.c4
¥e7 6.¤c3 0–0
The classical Queen’s Indian. I do not think
I ever spent enough time or worked enough on
these positions, but I had a lot of experience,
especially against Karpov, the most fearsome
positional player at the end of 1980s. I was
ready to suffer a slightly worse position.

  

   
    
   
   
 
  

7.¦e1
Kramnik’s trademark.
Against Karpov, from a different move order,
I was defending another line: 1.d4 ¤f6 2.c4
e6 3.¤f3 b6 4.g3 ¥b7 5.¥g2 ¥e7 6.¤c3 ¤e4
 Chapter 2 – Evolution of the Isolated Pawn 97

 
     
  
   
        
      
        
     
    

7.¥d2 ¥f6 8.£c2 ¤xd2 9.£xd2 0–0 10.0–0    
d6 11.¦fd1 ¤d7 12.d5 e5 13.£c2 g6 14.e4 
¥g7 15.¥h3 h5 16.b4 £e7 17.£e2 ¦fd8 Black has many options here, and during
18.a3 ¥h6 19.¦a2 ¤f6 20.¦c2 a5 21.¦a2 the game I was not sure which was the best.
¦db8 22.¦da1 ¤h7 23.¥g2 ¥c8 24.¤b5 I made a move that was popular a long time
 ago, at the beginning of the 1980s, just trying
  to avoid following Kramnik’s well-known
path.
  
    10...£c8!? 11.¤e5
  11.¦c1 ¦d8 is Black’s plan with ...£c8. An
   earlier game of mine with White continued:
    12.¤e5 c5 13.dxc5 bxc5
  
      
  
24...¤g5 25.¤xg5 ¥xg5 26.c5 bxc5 27.bxc5    
¥a6 28.a4 dxc5 29.£c4 ¦b6 30.£c3 ¥xb5    
31.axb5 ¦xb5 32.¥f1 ¦b4 33.¦xa5 ¦xa5     
34.¦xa5 ¦b1 35.h4 ¦c1 36.£b3 ¥h6 37.¢g2
¥f8 38.¦a8 ¢g7 39.¥c4 £d6 40.¦a6 £d7
    
41.¥e2 ¥d6 42.¦a2 c4 43.£e3 ¦e1 44.£c3  
¦b1 45.¥xc4 ¦b8 46.¥e2 ¢h7 47.£e3 £e7    
½–½ Karpov – Ehlvest, Haninge 1990. 
14.e4 ¤xe4 15.¤xe4 dxe4 16.£b3 ¥d5
7...d5 17.£e3 f6 18.¤c4 f5 19.f3 ¥f6 20.¥e5
There are other moves, such as 7...c5 and ¥xe5 21.¤xe5 ¦e8 22.£f4 exf3 23.¥xf3
7...¤e4 8.¤xe4 ¥xe4 9.d3 ¥b7 10.e4 c5, but ¥xf3 24.¤xf3 £b7 25.£xf5 White had some
after the text move the rook on e1 is not very advantage in Ehlvest – Wojtkiewicz, Haninge
well positioned. 1990.

8.cxd5 exd5 9.d4 ¤a6 10.¥f4 11...c5!?

98 Grandmaster Opening Preparation
When in doubt in the opening, the best
strategy is to create a confrontation as quickly
as possible. There was nothing wrong with
11...¦d8, but I was not sure what to do next
– to play ...c5 or ...c6 anyway, or first to move
my queen to e6. During the actual game, you
do not have enough time to figure it out. Also,
I tended to get into time trouble, and the text
move was a good idea to clarify the situation
in the centre.
12.dxc5 ¥xc5!
At this point I already had Lein’s idea in
mind.
12...¤xc5? loses the d5-pawn without
compensation: 13.¤xd5 ¥xd5 14.¥xd5±
13.e3
13.¤xd5?! ¤xd5 14.¥xd5 ¥xf2† 15.¢xf2
£c5† is fine for Black.
13.¦c1!? £e6 14.¤d3 was possible, to prevent
the plan Black employed in the game.

 
 
   
   
    
    
  
   

13...¥b4!?
If you do not know this plan or idea, it
is very difficult to come up with it during a
game. The position is very complicated and
to give up the good bishop seems like giving
up hope. When a beginner gives up his bishop
it is a patzer move; when a grandmaster does
the same thing you need to figure out what
is going on. Kramnik certainly would not
underestimate his opponent, and although
at first it looks as though White is securing a
better position, he probably realized very soon
that Black’s position is solid.
It may be possible to come up with this plan
by using a computer, but computers do not
tell you the background of the idea. I was very
pleased with myself after the game that I had
been able to use one of the ideas I had known
for so many years, but had not previously had
the chance to try out.
14.¦c1 £e6 15.¤d3
15.¤f3 ¤e4 16.¤d4 £f6 is good for Black.

  
 
  
   
    
   
  
   

15...¥xc3!
Of course, this is the point of ...¥b4.
16.¦xc3 ¤e4 17.¦c1 f6
Aiming to keep the white bishop on the
h2-b8 diagonal.
There was nothing wrong with 17...¦ac8
18.¥e5 ¦xc1 19.£xc1 ¦c8 20.£d1 ¤b4,
but this is a computer line and the last move
is typical computer tactics. I was playing
the position without much calculation, just
following positional principles. The line offered
 Chapter 2 – Evolution of the Isolated Pawn 99

by the computer is equal, but it changes the 23.¤e2 ¤e5 24.¤d4 £d7 25.¥g3 ¦c8
position and with limited time it is difficult 26.£d2 a5 27.¦d1 ¥a6 would have given
for a human to calculate and evaluate all the Black counterplay on the light squares.
positions that may occur. Sticking to one plan
is a much more practical solution. 23...¤d7 24.¤b3 ¤de5 25.¤d4 £d7
26.¥g3 ¦c8 27.¢h2
18.g4 27.f4 leaves a hole on the e4-square:
18.f3 ¤ec5 is okay for Black. 27...¤c4 28.f5 ¤fe5 29.¤e6 £b5 Black has
good counterplay.
18...¦ac8 19.h3 ¦xc1 20.¤xc1 ¤ac5

   
    
       
       
       
     
      
      
    
 27...¤d6
21.f3 The active 27...¦c3! was possible, as in
21.¤e2!? ¥a6 22.¤d4 also offers White a the famous Petrosian – Spassky game, where
slight advantage. Spassky confused his opponent with an active
rook move, but I was not playing for a win – a
21...¤d6 22.b3 draw was my goal.
The computer’s evaluation is not really
relevant here. White has a certain advantage, 28.h4 ¦e8 29.£b3 ¢h8 30.b5 ¤dc4
but it is not enough to break Black’s defences
on the light squares.

  t L
22.¤e2!? ¤c4 23.¤d4 £d7 24.b3 ¤e5  
25.¥g3 ¦e8 26.£d2 would be similar to the
game.     
p  
22...¤f7 23.b4?
It is true that this move weakens the c4-
 +m P
square, and it received some negative criticism q  
from commentators. On the other hand, how
can you improve the position without pushing
  
your opponent’s pieces back?     

100 Grandmaster Opening Preparation
31.a4 £f7 32.h5 h6 33.¥f4 ¤d7 34.£c2
¤c5 35.¥f1 ¤e5 36.¢g3
36.a5 ¤e6 37.¥xe5 (37.¢g3 bxa5÷)
37...fxe5 38.¤xe6 £xe6 39.£c7 could be
tried, though 39...¥c8!? 40.£xa7 £f6 should
give Black enough counterplay.
36...¦c8 37.£f5 ¦e8

   
  
    

  
   
    
   

38.£c2
38.¥xe5! ¦xe5 39.£g6 was an interesting
try, missed by both players: 39...£xg6 40.hxg6
¤xa4 41.¦a1 ¤c5 42.¦xa7 ¦xe3 43.¥g2
and White has good compensation. Still it
looks as though after 43...¢g8 44.¤f5 ¦e8
45.¤d6 ¦a8 46.¦xb7 ¤xb7 47.¤xb7 ¦a1 it
is still a draw, but this is not really relevant to
a practical game. Using computers is fun, but
still too far from a real human game.
38...¦c8 39.£f5
We were reaching the time control and
Kramnik could not see how to strengthen his
position further. The position is still better for
White, which it is easy to see using a computer,
and indeed any GM can see it by looking at
the position. The question is whether it is
enough to win the game. Kramnik came to the
conclusion that it was not.
½–½
Mastering isolated pawn positions allows
a player to build up his opening repertoire
without being afraid of slightly worse
endgames with an isolated pawn. This,
however, needs additional knowledge of these
endgames. There is one line in the Queen’s
Gambit Declined where you may land directly
in an endgame from the opening.
Valery Salov – Andrei Sokolov
Moscow 1988
1.d4 ¤f6 2.¤f3 d5 3.c4 e6 4.¤c3 ¥e7
5.¥g5 0–0 6.e3 ¤bd7 7.£c2 c5 8.cxd5
¤xd5 9.¥xe7 £xe7 10.¤xd5 exd5 11.¥d3

 

    
   
    
  
 
   

Which move is better here: 11...h6 or 11...g6?
It depends! Both moves are fine, though the ...g6
move has some apparent advantages over ...h6.
The only drawback is that having the pawn on
g6 might not be so good in the endgame. Very
deep analysis is needed to figure out if this is
the case. If it turns out having the pawn on g6
is not so important in the endgame, then ...g6
is definitely better: it restricts the white bishop
and protects the f5-square.
11...g6
11...h6 12.dxc5 ¤xc5 13.0–0 ¥g4 14.¤d4
¦ac8 15.¦ac1 a6 16.¥h7† ¢h8 17.¥f5 is
 Chapter 2 – Evolution of the Isolated Pawn 101

one possible line where White can use the f5- 25...h5
square: 17...¥xf5 18.£xf5 White has a serious The idea of this move is to gain space and
advantage. defend against White’s plan of f2-f3, g2-g4
and h2-h4.
12.dxc5
12.0–0? c4 13.¥e2 b5 favours Black. 26.¥e2 ¥d7 27.h4 ¥a4 28.¦b1 ¢d6 29.¥d3
¢e5 30.¢e2 ¢d6 31.f3 ¤c5 32.¦c1 ¥d7
12...¤xc5 13.0–0 ¥g4 14.¤d4 ¦ac8 33.¢d2 ¤xd3
15.¦ac1 a6 This move was not really necessary, but now
I do not want to comment on all the we have the critical position on the board.
individual moves here, but on the endgame The whole theoretical line is put to the test
that arises. in this endgame. If White cannot win this
 ideal-looking good knight against bad bishop
   position, then the whole line is safe for Black.

  34.¦xc8 ¥xc8 35.¢xd3

   
       
      
      
    
        
   
16.¥e2
16.£d2!? has also been tried.
  
    
16...¤e4 17.£d3 ¥d7 18.¥f3 £b4 19.£b3
£xb3 20.¤xb3 ¦xc1 21.¦xc1 ¦c8 22.¦d1

The main problem of this type of endgame
¥e6 23.¤d4 ¢f8 24.¢f1 ¢e7 25.¢e1
from Black’s side is how to handle the kingside
 pawns. In this particular game, Black has
    already committed his pawns, but if the
h-pawn were still on h7 then White could play
  g2-g4, not allowing ...h5. We will look at such
  a situation below.
    35...¥d7 36.¢c3 b6
    Valery Salov, along with all Soviet players
    from his generation, was a very strong
endgame player. We can believe him or analyse
   the positions ourselves, but there is no way
    White can win this position without somehow


102 Grandmaster Opening Preparation

tricking Black, as Karpov did to Kasparov in a 33.g3

similar position. 33.¤f3 ¥f5† 34.¢d4 ¥e4 35.¤e1 g5 is also
fine for Black.
37.¤e2 ¢e5 38.¤g3 a5 39.f4† ¢d6 40.¢d4
¥f5 41.e4 33...g5 34.f4 gxh4 35.gxh4 ¥g4 36.¢c3
½–½ ¢c5 37.¤b3† ¢d6 38.¢d4 ¥d1 39.¤c1
¥f3 40.¤a2 ¥g2 41.¤c3 ¥f3
Andrey Shariyazdanov – Sergey Grigoriants 
Nizhnij Novgorod 1999     
     
  
     
     
       
       
         
        
   
     White cannot make progress.

 42.b3 ¥g2 43.¤e2 ¥f3 44.¤g3 ¥g4 45.¢c3

25...a5 26.¢h2 ¢f8 27.¢g3 ¢e7 28.h4 ¢c5 46.¤f1 ¥f3 47.¤d2 ¥g2 48.b4†
In this game, as in the previous example, axb4† 49.axb4† ¢d6 50.¤b3 ¥h3 51.¤d4
Black will not allow the ideal set-up for White. ¥g4 52.¢d3 ¥h3 53.e4 ¥g2 54.¤f5† ¢e6
55.¤d4†
28...f6 29.¤d4 ¥d7 30.¢f3 ¢d6 31.¢e2 ½–½
b6 32.¢d3 h6
 In these games Black did not allow White to
  have the ideal set-up. Black was able to prevent
the white king’s march to d4. Only by playing
   his pawn to f4 could White achieve that, but
  then his pawn formation would be spoiled.
In the position below, Black’s task is not easy.
   Precise defence is required. The main question
  is where to keep the bishop to defend the
d-pawn and how to handle the pawns on the
   kingside.
 
 

 Chapter 2 – Evolution of the Isolated Pawn 103

 8...h5! This is the idea behind 6...¢e6, and

    
 Black should hold. 9.gxh5 ¥xh5 10.¤h1
   ¥d1 11.¤g3 ¥f3 12.¤f1 Despite the reduced
material, Black still has to defend the position
   with accuracy.
   
     4.¤e2 g5 5.hxg5 hxg5 6.¤c3 ¥g8 7.f4 ¥e6
8.b4 axb4 9.axb4 ¥f7 10.fxg5 fxg5 11.e4
   dxe4 12.¤xe4† ¢e7 13.¤xg5 ¥b3 14.¤e4
    ¢e6

     
     
1...f6!     
1...¥d7 2.¤f4 ¥c6 3.g4 h6 4.g5 hxg5
5.hxg5 ¥b7 6.e4! dxe4 7.fxe4 ¥c6 8.¤d5 b5
   
9.¤f4 b4 10.axb4 axb4 11.e5† and White has     
excellent winning chances.   
After 1...h5? 2.¤f4 Black is in zugzwang and    
is forced to weaken his position further with     
2...b5.
    
2.¤f4 ¥f7 3.g4 h6 
3...g5 is also possible: 4.hxg5 fxg5 5.¤h3 The extra pawn is not enough to win the
h6 6.¤f2 (6.f4 gxf4 7.exf4 ¢e6! 8.¤f2 h5! game.
transposes) 6...¢e6 7.f4 gxf4 8.exf4
 Many players from the Soviet Chess School,
such as Andrei Sokolov, Yuri Balashov and
     Yuri Razuvaev, played this Queen’s Gambit
    Declined line with success. Knowing the
    endgame that arises from this opening line is
    only one small part of the opening preparation,
    but a vital one.
     I have also used this opening line in my
     practice, but I did not like the very drawish
     line and opted for a more aggressive one.

104 Grandmaster Opening Preparation

Alexander Khalifman – Jaan Ehlvest Garry Kasparov – Lajos Portisch

Tallinn (rapid) 2012 Brussels 1986

1.d4 ¤f6 2.c4 e6 3.¤f3 d5 4.¤c3 ¥e7 1.d4 ¤f6 2.c4 e6 3.¤f3 d5 4.¤c3 ¥e7
5.¥g5 0–0 6.e3 ¤bd7 7.£c2 h6!? 5.¥g5 0–0 6.e3 ¤bd7 7.£c2 h6 8.cxd5!?
This is a very ambitious move. I have analysed this position a lot with and
 without computer engines. Back in the 1980s
it was considered too dangerous to accept the
  sacrifice.
  8.h4 does not attack anything and Black may
    just continue with 8...c5.
    
     
     
     
      
     
8.¥h4 c5 9.cxd5 ¤xd5 10.¥xe7 £xe7
11.¤xd5 exd5    
White no longer has the tempo move ¥d3,  
attacking the h7-pawn. The following moves
are forced.
  

12.dxc5 ¤xc5 13.¥e2 ¥g4 14.0–0 ¦ac8 8...exd5?
Black has an easy game compared with other 8...hxg5 9.dxe6 fxe6 10.¤xg5 ¤b6
games where White had an extra tempo. 11.h4 and now 11...¥b4! 12.h5. White has
compensation, but it is difficult to say, even
15.¦ac1 using computer engines, whether White has
A draw was agreed. an advantage.
½–½
9.¥f4 c5
Why then does Black not include this simple 9...c6 10.h3! offers White an edge.
move 7...h6 before playing ...c5? The following
game may provide an answer. 10.¥e2 b6 11.0–0 ¥b7 12.¦fd1 ¦c8 13.dxc5
bxc5 14.a4 £a5
 Chapter 2 – Evolution of the Isolated Pawn 105

 stick with your principles all the way. Giving up

   on your bluff in the opening against a very well-
  prepared opponent is not a good strategy.
Accepting the piece sacrifice looks scary, and
     probably it looked even more scary against
    Kasparov, as it would be naive to think that
    Kasparov did not know the ...h6 idea and was
not prepared for it. Still, the correct play was to
    accept the sacrifice. Portisch knew of course that
  by not accepting the sacrifice he would be much
worse when compared with similar lines in the
    Queen’s Gambit Declined.
 As with other Hungarian players, such as
Gyula Sax and Zoltan Ribli, Portisch was a good
15.¤h4
The early ...h6 has made the f5-square theoretician, but they were all players who had
vulnerable. Kasparov did not miss this much better results with the white pieces. Their
opportunity. opening preparation was always very concrete.
However, when the position did not suit their
15...¦fd8 16.¤f5 ¥f8 17.¤b5 ¤e8 style, usually in a situation where you needed
to make some big decision over the board, they
 played below their usual strength.
 
The most refreshing recent game where the
  isolated pawn concept was used was the
     following.
  Radoslaw Wojtaszek – Veselin Topalov
    Shamkir 2017
    
  1.d4 d5 2.c4 c6 3.¤f3 ¤f6 4.e3 ¥f5 5.¤c3
e6 6.¤h4 ¥g6 7.¤xg6 hxg6 8.¥d3
    
   
18.¥d6! ¤xd6 19.¤fxd6 ¦b8 20.¤xb7
¦xb7 21.¦xd5   
White won the pawn and eventually the
game.
 
...1–0    
Portisch made a typical mistake in the opening.
   
If you do not want to accept a slightly worse    
position against a very strong player (in this case,
the World Champion) and you are playing some
  
sideline or using some trick, then you need to   

106 Grandmaster Opening Preparation

Now Black introduces a very nice idea.


   
8...c5!
It is much easier to find this kind of idea if
 
you are familiar with isolated pawn positions.   
   
9.£b3 £d7 10.cxd5 exd5 11.dxc5 ¤c6!
An excellent novelty!     
  
11...¥xc5 12.£b5! The isolated pawn in
the endgame is a weakness in most cases.
  
12...£xb5 13.¤xb5 ¢d7 14.0–0 ¤c6 15.¦d1    
¢e7 16.¥d2 ¥b4 17.¤c3 ¦ac8 18.¦ac1 ¦hd8
19.a3 ¥d6

14.¤a4?
 14.¤e2 ¥b6 15.¥b5 0–0 16.0–0 ¤e4
    17.¦c2 ¦fe8 18.¤g3 and White can still hold
   the position together.
   14...¥d6 15.¤c5 ¥xc5 16.¦xc5
    If you are familiar with the material presented
     in this chapter, you will already know that
    here the two knights are much better than the
    bishops.
    
    
20.¤e2 (20.g3! would have given White a
long-term advantage, because the advance  
...d4 does not promise much for Black in   
the endgame as it only clears more space for
the white bishops. This is not the case in the
   
middlegame, when the advance ...d4 can be     
combined with a kingside attack.) 20...¤e5
½–½ Kramnik – Gelfand, Saint Vincent 2005.
  
  
12.¥d2 ¥xc5 13.¦c1 ¦d8    
Now we have a typical situation in which a
very strong player quickly loses the thread in 
a position with the isolated pawn, most likely 16...d4?
because he did not have much experience in Topalov, however, makes this standard move
these positions. The following attacking move too carelessly.
is a mistake, and instead some accurate defence
was called for. 16...a6 17.¦c1 d4 was the correct move order.
 Chapter 2 – Evolution of the Isolated Pawn 107

17.¥b5?

17.e4! £e7 and most likely both players    
missed 18.¦b5! ¤xe4? 19.¦xb7 and White is
winning.
   
  
17...0–0 18.¥xc6 bxc6    
Now Black is clearly on top. It is difficult
to understand what White could have missed     
– it is not hard to foresee that this position is    
hopeless for White.
 
19.f3 £e7 20.¦c2 ¤d5    
20...¦fe8 was even stronger. 
25.¢g3
21.¢f2 ¦b8 22.£a3 25.¢f1 was the only move, and after
 25...¤f4 26.£c3 £b6 27.£b3 £d4 28.£c4
    £d1† 29.¢f2 ¤d3† 30.¢g3 £xh1 31.£xd3
£e1† 32.¢h3 £e6† Black still has some
    winning chances.
   25...£f4† 26.¢f2 ¦b8 27.£c1 £d4†
    28.¢g3 ¤e3 29.¦c5 ¦b2 30.¦g1
     
       
     
      
     
22...¦xb2
Spectacular, but again not the best.     
   
22...dxe3† 23.¥xe3 £h4† 24.g3 £h3 25.¥c5
¦fe8 and White is helpless.
  
    
23.£xb2 dxe3† 24.¥xe3 £xe3† 
30...¦xa2
30...¦c2! 31.¦xc2 ¤f5† 32.¢h3 £h4#

31.h3 £d6† 32.f4 £d3 33.¢h2 £e4 34.¦g5

¦c2
0–1
108 Grandmaster Opening Preparation

Conclusion

The evolution of the isolated pawn strategy presented here using the Tarrasch Defence is not
complete of course. It only shows how deeply one can dig into an opening line using the history
of the line, identifying the relevant players and games. The chess detective is prone to mistakes
as well, and I apologise if I have missed some games or players who contributed no less value
than the players and games presented in this chapter. Having previously presented this material
in some private lecture sessions to my students, I am happy to now offer it to a wider audience. I
would also like to finish this chapter with a selection of exercises for the reader to test his or her
understanding of isolated pawn positions.

I believe that the work schedule with an opening line should be in the following order.
First, you should go through the games of today’s top players. From these games we can find
the critical positions, that is, those positions where one side, usually White, cannot find any
advantage, or the other side, Black, cannot find how to equalize.
Next, it is very useful to understand how these positions evolved from the past. Here some good
detective work is a must. The quality of this work and how deeply it goes depends on the level of
our student. For example, if our student at the beginner level wants to play some classical lines
after 1.d4 – let’s say the Nimzo-Indian Defence – we could recommend to him the book about
the 1953 Zurich Candidates tournament, written by David Bronstein.
On the other hand, if we have a student who is going to compete in the World Junior
Championship, he definitely needs to work as quickly as possible to be ready for the event. In
this case, we should be aware of the problem that just remembering the computer lines and the
latest games from the top players on the subject might yield only short-term success.
 Exercises
Ladva – Tomashevsky, Minsk 2017 Gralka – Rosicki, Jastrzebia Gora 2016
 
1     3
  
   
        
       
       
     
     
   
   
 
What is the best move? Should White play Should Black take the f3-knight, spoiling
13.b3 trying to develop his bishop, 13.¤g3 White’s pawn structure, or continue to develop
immediately attacking the centralized knight, his pieces with 12...¤c6?
or 13.¥d2?

Kastek – Schnepp, Bad Wiessee 2016 Reshef – Cruz, Barcelona 2016

 
2   
  4
  
    
       
       
      
    
     
     
 
Should Black take the c3-knight to continue Should Black take the f3-knight, or retreat
the positional battle, or take the f3-knight the bishop to h5 or to f5?
hoping for a tactical solution of the position?
110 Grandmaster Opening Preparation

Savchenko – Grachev, Sochi (blitz) 2016 Lekgau – Tumelano, Johannesburg 2016

 
5   
  7
  
     
       
       
       
    
    
     
 
Should Black take the f3-knight or retreat Should Black take the f3-knight, or retreat
the bishop to e6? the bishop to e6, or play 12...£c7?

Dzindzichashvili – Lein, Leningrad 1971 Stefansson – Seo, Uppsala 2016

 
6   
  8
   
     
       
       
       
     
   
         
 
Should Black take the f3-knight or retreat Black has a difficult position. Should he
the bishop to f5? keep the bishop pair and play 26...f4, or take
the f3-knight?
 Chapter 2 – Evolution of the Isolated Pawn 111

Grieme – Vachylya, Korbach 2016 Neiksans – Novik, Finland 2016

 
9    
  11
  
    
       
       
       
      
     
       
 
Black has good compensation for the pawn, The a-pawn is under threat. Should Black
but how should he continue? Should he attack play 16...¦b8 targeting the b3-pawn if White
the b-pawn with 18...£b6, at the same time exchanges on c6, or take with 16...¤xd4 and
controlling the d4-square, or take the f3-knight? then play 17...a6, or take the f3-knight?

Strating – Hummel, Hoogeveen 2016 Kurbedinov – Predke, Sochi (rapid) 2016

 
10  
  12
  
   
      
       
       
      
     
       
 
Should Black take the f3-knight or play Black can play 13...¦ab8 with the idea of
14...£d7? 14...£xc5, or he can take the f3-knight, giving
up his bishop but hoping for counterplay on
the kingside. Which plan is better?
 112 Grandmaster Opening Preparation

Li Yunshan – Zhao Xue, China 2016 Leroy – Erneste, Katowice 1984

 
13  
  15
  
   
       
       
       
     
    
      
 
Should Black take the f3-knight, or keep the Black has a nice position in the centre, but
tension and play 15...¥h5? how should she continue? Should she attack
the white bishop with 16...¤a4, or play
16...¦c8 or 16...¦e8?

Narayanan – Xu Yi, Bhubaneswar 2016 Gauglitz – Espig, Zittau 1989

 
14   
  16
  
   
        
       
     
      
   
      
 
Should Black exchange queens on d4 and hope In this typical position with an isolated pawn,
for quick counterplay on the c-file, or take the Black has three options. Should he play
f3-knight? 13...¤e4, or immediately take the f3-knight,
or play 13...¦e8?
 Solutions
1. Ottomar Ladva – Evgeny Tomashevsky 2. Thomas Kastek – Gunnar Schnepp
Minsk 2017 Bad Wiessee 2016

13.¥d2! 16...¥xf3?
This is the correct solution – White should 16...¥xc3! was the correct solution: 17.¥xc3
not be afraid to trade his bishop for the active 
knight on e4.   
Less accurate is: 13.b3 £f6 14.£d4 £xd4
  
15.¤fxd4 ¤c3 16.¤xc3 ¥xc3 17.¦b1 ¥xd4=    
Black reaches a drawn endgame.    
  
Also weaker is 13.¤g3 ¦c8! and Black has a  
strong initiative.
  
    
   
  17...¥f5! 18.¤d2 ¤xc3 19.£xc3 ¦e8 Black
has an excellent position, because of his full
    control over the central squares.
   
    17.£xf3 ¤d4 18.¦xd4 ¥xc3 19.¥xc3 ¤xc3
20.£d3
   White has an obvious advantage, because
  the black knight on c3 does not compensate
for the weakness of the d5-pawn.
   
 20...£b6
13...¥d6 
After 13...¤xd2 14.¤xd2 White has a
positional advantage, because in the position
  
with the isolated pawn, the knight is stronger   
than the bishop.     
14.¦ac1²    
White has completed his development and     
has a pleasant edge.
  
  
    

114 Grandmaster Opening Preparation

21.¦b4?! 3. Przemyslaw Gralka – Marcin Rosicki

This eases the pressure on the d5-pawn.
Correct was: 21.e3! ¦fd8 (21...£xb3?? 22.¦b4) Jastrzebia Gora 2016
 12...¥xf3?
 T + In the endgame the value of the bishop is
   more important.
    
    Better was just 12...¤c6! and Black has a
     satisfactory position.
P   13.gxf3 ¦d8 14.0–0–0 ¤c6
   
     
   
22.¥xd5 ¤xd5 23.¦xd5 ¦xd5 24.£xd5 ¦d8  
25.£c4 White emerges with an extra pawn.
   
   
    
  
    
 

15.¥b5?!
Stronger was 15.f4! d4 16.e4 and White has
the better game.
 Chapter 2 – Evolution of the Isolated Pawn 115

4. Omer Reshef – Cristhian Cruz 5. Boris Savchenko – Boris Grachev

Barcelona 2016 Sochi (blitz) 2016

10...¥xf3 13...¥xf3?
Black should avoid exchanging the bishop, Keeping the bishop with 13...¥e6! is the
although 10...¥h5?! would be a mistake: 11.g4 right choice here. After the move played, the
¥g6 12.¤e5 0–0 13.¤c3 White has a strong light squares become too weak.
initiative.
14.¥xf3 ¤d4 15.¥g5 ¤xf3† 16.£xf3 ¥d4
However, retreating with 10...¥f5! was
strongest.

  
11.¥xf3 0–0 12.¤c3   
Now White has the advantage, because
having more pieces on the board helps White     
and the bishop is stronger than the knight.    
     
     
    
       
    
     17.¦ad1 ¥xc3 18.bxc3
White has a clear advantage, because Black’s
   d-pawn is very weak.
   
  

12...£d7 13.¥g2 h6 14.¥e3²
116 Grandmaster Opening Preparation

6. Roman Dzindzichashvili – Anatoly Lein 19...£h6

19...f5!? 20.¥g2 (20.¥xf5? d3!–+) 20...f4
Leningrad 1971 21.£e4 £d6 22.gxf4 ¤c6 would offer Black
decent compensation.
16...¥xf3?!
16...¥f5! was stronger. In this position it is 20.£c7
better to keep the bishop, because the d-pawn 20.¢g2!? b6 21.¦d1 also gives White an
has advanced and the white bishop is going to edge.
be very strong on the h1-a8 diagonal.
20...¦fe8
16...¥h5 was also possible.

17.¥xf3 ¤e5 18.¥e4 ¦ad8   
  
        
      
        
        
      
        
   
     21.¦c5
White could play the solid 21.¢g2 and
 Black still needs to find counterplay. White has
19.£c2 the advantage.
After 19.£f4 ¦fe8 20.¦c2 White has the
advantage.
 Chapter 2 – Evolution of the Isolated Pawn 117

7. Mothupi Lekgau – Lesabe Tumelano 13.£xf3 ¤e5 14.£f5

14.£d1 d4 15.cxd4 ¤xd3 16.£xd3 £xd4
Johannesburg 2016 leads to an equal game.

12...¥xf3? 14...¤xd3
12...¥e6?! 13.¥f4 is also pleasant for White. 14...¦e8 was a better try.

The correct option is: 12...£c7! 13.¤f5 ¥xf5 15.£xd3

14.¥xf5 ¦fe8 15.¥c2 ¦ad8
 
     
     
        
       
    
       
    
     
    
Black is threatening ...d4 and has an excellent
game. 16.¤d4 ¤xd4 17.cxd4 ¥b6 18.¥e3 
£d6 Black has the advantage, because with a White has a small advantage.
fixed pawn structure the knight is always better
than the bishop.
118 Grandmaster Opening Preparation

8. Vignir Stefansson – Jung Min Seo 9. Jens Grieme – Andriy Vachylya

Uppsala 2016 Korbach 2016

26...¥xf3! 18...¥xf3!
This a strong move – in this position with This is much stronger than: 18...£b6?! 19.b3
isolated pawns, the bishop is weaker than the d4 20.¤xd4 ¤xd4 21.cxd4 £xd4 22.¤c3
knight. 
   
After 26...f4? 27.¤fd4 ¦cc8 28.h3 ¥d7 29.e6   
White has a winning advantage.
    
27.£xf3     
27.gxf3!? was an interesting move, improving    
the pawn structure and opening the g-file.    
 
27...f4 28.h3 ¦e6
    
 
    White has the advantage, because Black’s
    bishop is somehow irrelevant as White can
build up his position using only the dark
    squares.
    19.gxf3
     19.¦xf3? loses to 19...¤d4.
  
     
       
    
Black has counterplay.    
    
   
  
   

19...¤d4! 20.£d2 ¤f5 21.¦d3 ¦fe8
Even stronger is: 21...¤h4! 22.¢h1 £c6!
23.£g5 £b5 24.¦d1 (24.£d2 £a6 25.¤a3
¦xb2–+) 24...¤xf3 25.£e3 £xb2 Black
should win.
 Chapter 2 – Evolution of the Isolated Pawn 119

22.¤a3 ¦e6 10. Sybolt Strating – Joop Hummel

 Hoogeveen 2016
   
   14...¥xf3?
Correct is 14...£d7! 15.¦ad1 ¦ad8 and
    Black has a comfortable position.
   
      
    
      
        
     
Here too, Black has a strong attack.
   
  
    

15.exf3!
In this open position the two bishops are
very strong, and the doubled pawns are not
such an important factor.

15.¥xf3 ¤e5 16.¦ad1! was also strong.

15...¤e5
15...£b6 was relatively best, but after
16.¦xe8 ¦xe8 17.¦d1 d4 18.¤e2 ¦d8 19.f4
White has a big advantage, because it is
difficult to defend the d-pawn.

16.¤xd5
White has simply won a pawn.
120 Grandmaster Opening Preparation

11. Arturs Neiksans – Maxim Novik 12. Ramis Kurbedinov – Alexandr Predke
Finland 2016 Sochi (rapid) 2016

16...¥xf3! 13...¥xf3!
After 16...¤xd4 17.¤xd4 a6 18.¦ac1 White 13...¦ab8?! is weaker: 14.¤d4 £xc5 15.¦b5
has a small but pleasant advantage. £d6 16.¤xc6 £xc6 17.a4 White has the
better game.
If White meets 16...¦b8 with 17.¤xc6 bxc6
18.¦xa7 ¥xf3 19.¥xf3 ¦xb3 20.¦c1 ¦xb2 14.gxf3 ¤e5
21.¦xc6 h5! then even if Black loses his d5- Black has a good game.
pawn it will still be a theoretical draw in the 
rook endgame.
However, White can reply 17.¦fc1 and Black
  
has nothing better than to take on f3 anyway.  
17.¤xf3     
After 17.¥xf3 ¤xd4 18.exd4 a6 Black even    
has the better position, because in the long run
the d4-pawn is more difficult to defend than
    
the d5-pawn.   
    
     
  
    15.¥e2 ¦fc8 16.a4 ¦c6
16...£d7! was better, not allowing £f5.
   
     17.¢h1
White should have taken the opportunity to
   play 17.£f5! with good chances.
    17...£d7 18.f4? £h3
    
   
17...¦e4! 18.¦a4 ¦xa4
18...¦b4! 19.¦xb4 ¤xb4 20.¤d4 ¦c8 and  
Black is better.    
19.bxa4 d4 20.exd4 ¤xd4 21.¤e5 h5?    
After 21...b6! the endgame is equal.    
22.¥xb7 ¦b8 23.¥a6 ¦xb2 24.¥c4²    
White now had some pressure and eventually   
won.
 

 Chapter 2 – Evolution of the Isolated Pawn 121

Faced with 19.fxe5 ¤g4 20.¥xg4 £xf1#, 13. Li Yunshan – Zhao Xue
White resigned.
However, the resignation was premature, China 2016
because after 19.f3! ¤h5 20.fxe5 ¤g3†
15...¥xf3!
21.¢g1 ¦g6 22.¥d3 Black has only perpetual
After 15...¥h5 16.e3 White is able to defend
check.
the centre.
0–1
16.¥xf3
16.¤xf3? just loses a pawn to 16...¤xg3.
16.exf3? allows a neat tactic:

 
 
   
   
   
 
   
  

16...¤xg3! 17.fxg3 £f6 Black regains the piece
with a decisive advantage.

16...£f6!
Increasing the pressure on the d4-knight.

17.e3 ¤xd4 18.¥xd4 ¥xd4 19.exd4 ¦c3

20.¥g2 g6

  
 
   
   
   
  
   
  

Black has a big advantage.
122 Grandmaster Opening Preparation

14. Srinath Narayanan – Xu Yi 15. Christophe Leroy – Inguna Erneste

Bhubaneswar 2016 Katowice 1984

16...¥xf3! 16...¦e8!
This is stronger than: 16...£xd4 17.¤xd4 16...¦c8 17.¤b3 ¦e8! is also good.
¥xe2 18.¤xe2 ¦fc8
 However, 16...¤a4?! is not a good idea:
  17.¤d4 ¤ec5 18.b3 ¤xc3 19.£xc3 ¥f6 20.e3
¦c8 21.¥h3 ¦c7 22.¦ac1 White manages to
  keep equality.
    
    17.¤b3 ¥f8 18.¦fe1
    
      
 
     
    
19.¤d4! (19.f3 ¤d2 20.¦e1 ¦c2 gives Black    
sufficient counterplay) 19...a6 20.f3 White
has the advantage, because the knight on d4 is     
guarding the entry square on c2.   
17.¥xf3 £xd4 18.exd4 ¦fc8 19.¦e1 ¦c4
 
20.¦d3 ¦ac8     
 
   18...¤xb3
Keeping up the pressure with 18...¦c8 looks
  even stronger.
     19.axb3 ¤c5 20.£f4
    White will have some compensation for the
   pawn he is losing.
  
  
    

Black has reached an equal endgame.
 Chapter 2 – Evolution of the Isolated Pawn 123

16. Gernot Gauglitz – Lutz Espig 14...¥h5

14...¥xf3? 15.exf3! again gives White the
Zittau 1989 advantage.

13...¦e8! However, 14...¥f5 is a stronger retreat, with

After 13...¤e4? 14.¤d4! ¤xd4 15.£xd4 the idea of meeting 15.¤d4 with 15...¥e4.
¥f6 16.¦xc8! ¥xc8 17.£xa7 White is better.
15.¤c5 ¤e4 16.¤d3
13...¥xf3? 14.exf3! would give White a big
advantage, because of the bishop pair and the 
weak pawn on d5.  
14.h3
 
    
    
     
    
     
     
  
Black should now go for the good and
  aggressive 16...g5! with a complicated game.
  

Chapter 3
Key Ideas and Positions

Deep study of an opening line, starting with the evolution of the line, takes a lot of time. However,
time is the most valuable commodity – there is no bank that can sell you wasted time back.
Blindly memorizing lines is a supreme task and works for a short period of time. We can see
this when following world championship matches. Both players usually stick stubbornly to their
opening preparation.
A typical match in which two very well-prepared players, Anand and Gelfand, were unable to
break the opening barricades of their opponent was in Moscow 2012. Boris Gelfand prepared
some opening lines especially for this match, specifically in the Grünfeld Defence. This opening
requires remembering a lot of forced lines and the margin of error is very thin. Gelfand was
already a veteran (he was over 40 years old at the time of the match), yet he still preferred concrete
opening preparation and worked hard to store all these lines in his memory.
After the match, however, he did not play these lines very often in tournament chess. The
reason is simple: one cannot, without working every day, keep in one’s memory opening lines that
require exact knowledge. However, this is the final grind of opening preparation. First you need
to know or find the relevant positions and lines to remember.

There is an easy approach where one can just follow the current trend of opening theory and
copy it. The drawback is that when following or borrowing an opening repertoire from a trusted
or prominent player, the hidden subtleties may be lost to the borrower. It is not so important if
the borrower is also a very strong player and can look through the positions he has not analysed
himself and quickly get acquainted with them. Top players are top players because they have this
kind of ability and can switch quickly from one opening to another depending on the current
state of affairs, or because they just like to follow the trend.
Still, even at the top level not all players are universal players; there is a mixture. There are and
were players who relied only on their own analysis and work. This approach started with Mikhail
Botvinnik and was followed by his student Garry Kasparov.
On the other side there were players who did not mind following the common trend, players
like Boris Spassky, Mikhail Tal, Anatoly Karpov and Viswanathan Anand.
126 Grandmaster Opening Preparation
Boris Gelfand belongs to the first of these
categories. That is why the clash in Moscow in
2012 was like the opposition of two different
schools of chess: the hard-working Gelfand
versus the talented Anand.
Opening preparation before and during
world championship matches is almost the
only thing the players and their seconds
work on. Psychology and fitness are also
an important part of their preparation, but
receive less attention. Gelfand in my view won
the opening preparation battle, but ultimately
he lost the match by a minimal margin.
There is a strong trend towards moving
from hard-working home preparation to
the universal attitude, and this is due to
computers. It is so easy to follow and check
other people’s games and come to a conclusion
using a computer, and it saves a lot of time
compared with the situation in the past. The
only drawback is that we chess players do
not have the same hard drive and abilities of
a computer. To keep the correct information
in our memory without understanding how it
works is not possible.
The most recent world championship
matches have shown clearly that you cannot
win the match only by being better prepared in
the openings. There are no longer any sudden
one-move improvements during these matches
like those we saw in the past.
Bobby Fischer against Boris Spassky, in their
historic battle in Reykjavik in 1972, tried to
surprise Spassky by changing openings when
playing with the black pieces in nearly every
game. Spassky belonged to the universal
type of player and did not have a recipe for
every possible opening up his sleeve, and
could not threaten Fischer’s risky strategy. I
do not categorize Fischer here, because he
probably belonged to both groups – he was
hard-working himself and also a universal
player.
In 1995, in the match between Kasparov and
Anand, the latter could not withstand the
opening pressure from Kasparov.
Garry Kasparov – Viswanathan Anand
New York (10) 1995
1.e4 e5 2.¤f3 ¤c6 3.¥b5 a6 4.¥a4 ¤f6
5.0–0 ¤xe4
Anand tried playing lines that require a good
memory while preparing these lines. To play
these dynamic lines against Kasparov was a
mistake. It worked for a while, until Kasparov’s
team’s hard work during the match uncorked
the genie.
6.d4 b5 7.¥b3 d5 8.dxe5 ¥e6 9.¤bd2
¤c5 10.c3 d4 11.¤g5 dxc3 12.¤xe6 fxe6
13.bxc3 £d3

  
   
 
   
    
  
  
  

In my opinion Anand was unwise to repeat
this dynamic line once more. In Game 6,
after 14.¤f3 Black had a comfortable game.
In this game, Garry quickly made a different
move which refuted the whole line. Today
with computer help it is easy to find the
correct move, but back then it was not so easy.
However, my point is that Anand was too
naive in playing such lines against Kasparov,
because Anand belongs to the universal type
 Chapter 3 – Key Ideas and Positions 127

and Kasparov was the researcher. Even without 18.¥xe6

a computer, the researcher sooner or later finds All these moves were more or less forced, and
the correct path. it was obvious that Garry was still following
his home preparation.
14.¥c2!!N
Who exactly was the author of this move 18...¦d8
is not clear. It has been pointed out that the Anand found the best move and made it
associated idea was suggested by Tal. It is not rather quickly.
about one move, the idea is to sacrifice a whole
rook. Other options are weaker. In nearly all lines
the ¥d7† check is winning. For instance, after
14...£xc3 18...c6 there follows: 19.¥d7† ¢f7 20.£f4†
This is the only move. It seems likely that ¢e7 21.¥a3† ¢xd7 22.£g4† ¤e6 23.¦xa1
Vishy had not considered White’s last move in ¥xa3 24.£f3 ¥e7 25.¦d1† ¢c7 26.£f7+–
his pregame preparation, because he thought
for four minutes on this obvious reply.

   
15.¤b3!    
This move, played instantly, came as a
surprise to Anand.   
   
15...¤xb3
Now it became clear to everybody that
   
Vishy was in unknown territory, as he spent     
45 minutes on his move.
  
16.¥xb3 ¤d4 17.£g4    
 
   19.¥h6
Kasparov made this move instantly as well.
   
   19...£c3
Instead of the text move, Black could
    have the same endgame with an extra tempo
    after 19...£b2 20.¥xg7 £e2 21.¥xh8 £xg4
22.¥xg4, but this is objectively lost as well.
   
   20.¥xg7 £d3
    There is no time for 20...¥xg7 because of
21.£h5†!.

17...£xa1 21.¥xh8
There is no time for 17...¤xb3, because of
18.£xe6† ¥e7 19.¥g5 winning.
128 Grandmaster Opening Preparation

 Today there is computer help to analyse

    dynamic positions and it is available to every
    player. It looks as if computer software is
levelling the competition with regards to
   opening preparation in the same way that the
    Colt revolver supposedly made all men equal.
    Still, there is something more than just
brutal computer calculation. Before the
    computer’s evaluation process and the player
   remembering the lines, there should be a
concept. What is the concept of your opening
    preparation?
 Kasparov was very aware of this and used
a very good concept against Anand. I have
21...£g6
In my opinion, Black had more practical already mentioned a few times that Anand and
chances after 21...¤e2† 22.¢h1 ¦d4! Kasparov belong to two different categories
23.£h5† £g6 24.£xe2 £xe6. of players. Kasparov was strongly influenced
by Mikhail Botvinnik, who put professional
22.¥f6 ¥e7 23.¥xe7 £xg4 analytical work first, whereas Anand was
The only move, as after 23...¢xe7 White has very talented, but somehow did not have
the simple idea of 24.£h4† ¢e8 25.¥g4!. Botvinnik’s kind of discipline.
Before their match, Kasparov had won
24.¥xg4 ¢xe7 25.¦c1! another theoretical battle against Anand earlier
the same year.

     Garry Kasparov – Viswanathan Anand
    Riga 1995
   
1.e4 e5 2.¤f3 ¤c6 3.¥c4 ¥c5
   
    
     
   
        
     
Limiting Black’s counterplay. White is   
winning, but Kasparov never had very precise    
technique, so he spent a lot of time on every
move in this winning endgame, to make sure
 
he did not spoil it. White won on move 38.  
...1–0

 Chapter 3 – Key Ideas and Positions 129

Today this is one of the most popular opening


positions in super-tournaments. White usually  
tries some positional plans here, and the most
successful player in these positions is World
  
Champion Magnus Carlsen, who plays it    
with both colours. The gambit line played     
by Kasparov against Anand is not so popular
anymore because of computer analysis. It is     
possible to find some forcing lines that promise     
nothing for White.
 
4.b4!?    
Anand had no experience in this particular
line prior to this game.

18...¤f7?
Stronger was 18...¥e6, although White
4...¥xb4 5.c3 ¥e7 6.d4 ¤a5 7.¥e2 exd4 has excellent compensation. Even strong
8.£xd4! ¤f6 9.e5 ¤c6 10.£h4 ¤d5 grandmasters err in positions that are not
11.£g3 g6 12.0–0 familiar to them. Anand is in general an
 attacking player and rarely ends up defending.
  19.cxd6 cxd6 20.£e3 ¤xh6 21.£xh6 ¥f8
 22.£e3† ¢f7 23.¤d5 ¥e6 24.¤f4 £e7
   
       
      
      
      
       
     
12...¤b6?
Anand panics and will lose the game in a
 
dozen more moves.    
Better was 12...0–0 13.¦d1 ¤b6 14.a4,
when White has a strong initiative.

25.¦e1!
1–0
13.c4 d6 14.¦d1 ¤d7 15.¥h6! ¤cxe5
16.¤xe5 ¤xe5 17.¤c3 This opening disaster may look like a singular
17.¥g7 ¥f6 18.¥xh8 ¥xh8 19.¤c3 is also accident, but it actually had some underlying
promising for White. causes. Kasparov was testing Anand’s
knowledge of a rare opening line; Anand was
17...f6 18.c5
130 Grandmaster Opening Preparation
not ready. This gave Kasparov good reason to
bluff again in the future against Anand.
I put Anand in the same category as Karpov,
though he was never a classical player like
him. Karpov started his chess career with
openings like the Ruy Lopez and only on very
rare occasions played the Sicilian Defence.
His main weakness in later years was opening
preparation, because he could not adjust
to constantly working on openings, which
gave Kasparov an edge in their matches.
Karpov relied more on his ability to grind his
opponents in positional chess, especially in
endgames.
Little by little, Kasparov started to prevail
in their matches, as well as in tournament
chess. The 1988 USSR Championship and
also the World Cup series in 1988-89 showed
that Karpov could perform at the same level
as Kasparov when playing against other top
grandmasters. Nevertheless, Kasparov was the
better player – he was World Champion at that
time – and also the better theoretician. Against
Karpov, however, Kasparov never bluffed in
the opening.
Anand had his weaknesses and Kasparov
exploited them quickly. The game in Riga was
just a small test before their match in the same
year. The concept of playing dynamic lines
worked for Anand at the beginning of the
match, but the analytical power of Kasparov’s
team changed the course of the battle. We saw
above the line in the Open Ruy Lopez that was
refuted.
If you want to become the best, you need to
be perfect in all areas of chess. To play dynamic
or forcing lines you need to remember them,
but also the margin of error is very small.
Anand did not like the positional approach,
like that of Karpov, and it was too late during
the match to change his opening plans.
Kasparov, on the other hand, found and
exploited openings that required deep
preparation. First he played the Scheveningen
system, but here Anand showed good
preparation and, after a few draws, won a
game. Now the situation in the match become
critical for Kasparov and he switched to his
other opening line. The Sicilian Dragon came
as a shock to Anand – Kasparov’s boldness paid
off. The main reason for his success was simply
that Anand was not ready for the Dragon.
The Dragon itself may or may not be correct,
but in world championship matches players
usually avoid taking such risks in the opening.
Only Fischer, in his match against Spassky
in 1972, had played some risky openings
with success. Neither Spassky nor Anand was
ready to refute their opponent’s risky opening
strategy.
The idea of taking a risk against a player who
was not constantly working on his openings
worked perfectly for Kasparov. In an ideal
world, players must know all the sidelines
and, even more so, the current key lines and
critical positions. They should also have some
of their own analysis in sidelines that may not
necessarily offer an advantage, but which can
be difficult for opponent to play when seeing
them for the first time.
Viswanathan Anand – Garry Kasparov
New York (11) 1995
1.e4 c5 2.¤f3 d6 3.d4 cxd4 4.¤xd4 ¤f6
5.¤c3 g6
Garry prepared the Dragon especially
for this match. He had not played it in any
tournament games.
6.¥e3
Vishy hesitated a little, but still went for
the most principled line. This is a situation
where you would want to have your own odd
 Chapter 3 – Key Ideas and Positions 131

line ready to counter-surprise your opponent. 16...£a5!

Apart from the dynamic 6.¥e3 there are other This move had first been played in Suetin –
more positional lines that are playable, but Szabo, Leningrad 1967.
Anand, as we know, is not really a positional
player. Still, in a situation in which you have However, the most common move here is
lost your previous game, some caution is called 16...b4, which was tested in Wolff – Ki.
for. Georgiev, Biel 1993, where White got the
 upper hand after 17.¥xg7 ¢xg7 18.¤d5
¤xd5 19.exd5 £a5 20.b3 ¦c5 21.g4. With
  Patrick Wolff being one of Anand’s seconds,
  most certainly this line did not go unnoticed
by Kasparov’s team.
   
     17.¥xg7 ¢xg7 18.¤f4 ¦fc8
    
        
    
     
   
6...¥g7 7.f3 0–0 8.£d2 ¤c6 9.¥c4   
Garry himself preferred 9.0–0–0. It is
obvious that Anand had some idea of how    
to play against the Dragon. Most likely he  
followed one of the games played by his
second, GM Patrick Wolff.
 

9...¥d7 10.0–0–0 ¤e5 11.¥b3 ¦c8 12.h4 The timid professional approach in openings
h5 13.¢b1 ¤c4 14.¥xc4 ¦xc4 15.¤de2 b5 is that when you are faced with a novelty, or you
16.¥h6 have not analysed or do not know the position
 in depth, it is better to bail out with a quick
draw. This approach was even used by Karpov
    at the end of the 1970s and the beginning
   of the 1980s, when he won nearly every
tournament. Karpov could afford this because
    it was a very rare occasion when his opponent
   could surprise him in the opening, and on the
   other hand he won a lot of tournaments by
such a big margin that one draw could not
    spoil the final result. With his next move it
 seems that Anand is aiming for a draw.

  19.¤cd5


132 Grandmaster Opening Preparation

With this move it is difficult to refute the After 28.¤xe7 ¦e8 29.¤d5 ¥xd5 30.b4
Dragon. As Kasparov pointed out, White axb4 31.axb4 ¦c4 32.¦xd5 ¦xb4† 33.¢c1 f5
needs to push g2-g4 at some point to fight for 34.¦xd6 fxe4 35.¢d2 it would be equal.
the advantage.
28...axb4 29.axb4 ¦c4
19...£xd2
Kasparov was happy with his opening play,

and after White’s 19th move he thought that    
Anand was going to split the point and so    
offered a draw.
  
  
      
      
      
      
   
    30.¤b6??
  Anand could not stop his hand making this
  move.

 He had to try 30.c3! instead, and after

20.¦xd2
Anand declined Kasparov’s offer – a very
strange decision. The first rule is that if you
start to think, you need to accept the offer.
There was only one move to make if he did not
want to accept the offer. Spending even four
minutes was a mistake.
20...¤xd5 21.¤xd5 ¢f8 22.¦e1 ¦b8
Black’s position is solid, but instead of this
last move, which is just a waste of time, it was
better to play 22...¥e6 immediately.
23.b3 ¦c5 24.¤f4 ¦bc8 25.¢b2 a5 26.a3
¢g7 27.¤d5 ¥e6 28.b4?
Anand, as if under hypnosis, continues to
play for the full point and is preparing the final
blunder.
30...¥xd5 31.¦xd5! ¦xc3 32.¦e2! he could
still save the game.
30...¦xb4† 31.¢a3 ¦xc2
0–1
After 32.¦xc2 ¦b3† 33.¢a2 ¦e3† White
would be two pawns down. Anand just had a
bad day at the office.
After this game the match was over. Anand
could not find any remedy against the Dragon
and even lost another game against it.
However, he made one good move: choosing
the right opening in the 14th game of the
match. This may look like a desperate attempt,
but it seems to me that it was actually a very
good idea, forcing Kasparov to solve the same
problem that he himself had set earlier when
choosing the Dragon.
 Chapter 3 – Key Ideas and Positions 133

Garry Kasparov – Viswanathan Anand 8.f4?!

In the opening the most difficult task is to
New York (14) 1995 make positional decisions. Tactical decisions
are much easier to make because you might be
1.e4 d5 able to calculate correctly to the very end. It is
This bold move was a counter-shocker and a not possible, however, to calculate a positional
very good one. If you can play the Dragon, I decision. You need to make a judgement
can play the Scandinavian. taking into account the overall situation. The
situation might force you to take some extra
2.exd5 £xd5 3.¤c3 £a5 4.d4 ¤f6 5.¤f3 c6 risks, or conversely you might play cautiously.
 The psychological factor is important here.
  Kasparov makes a risky decision. The set-up
with f2-f4 is aggressive, but positionally risky.
  If White cannot build up a concrete attack, the
    early pawn advance only weakens the position.

     For some reason Kasparov did not like the

     simple 8.¤xd7 ¥xd7 9.0–0, when White has a
small advantage without any hassle.
   
  Most likely Kasparov was also overconfident
  and was at least in his mind sensing that his
opponent was sweating. This approach nearly
 backfired.
Now Kasparov did what Anand did not
do when he was surprised by Kasparov. Back

then, Anand followed the main line in the   
Dragon and Kasparov showed him his better 
preparation. Here Kasparov will not play the
main line. Probably he was not sure that he   
could remember all the subtleties, so he made     
a different move to avoid Anand’s home
preparation. This is the situation when it is
    
good to have a sideline in your pocket.    
6.¤e5  
6.¥c4 is the main line.   
6...¥e6!? 7.¥d3 ¤bd7 
Anand shows that he is playing for equality. 8...g6 9.0–0 ¥g7 10.¢h1 ¥f5 11.¥c4 e6
Black now has an excellent blockade on the
More complicated but more risky lines appear f5-square.
after 7...g6!? 8.0–0 ¥g7, but Anand was not in
aggressive mood and his move is theoretically 12.¥e2
a good choice. 12.g4? fails to 12...¤xe5!.
134 Grandmaster Opening Preparation

12...h5 13.¥e3 ¦d8 14.¥g1 0–0 15.¥f3 26...bxa6 was better and Black still has some
¤d5 advantage.
Black has an excellent position.
 27.¤e5! £e6?
27...fxe5 28.fxe5 ¤e4 29.¥xd8 ¦xd8 30.g4
    hxg4 31.¥xg4 ¥xg4 32.£xg4 ¤f2† 33.¦xf2
  £xf2 34.£xg6† ¥g7 35.¦c7 £f1† results in
a draw.
  
    
         
     
   
      

16.¤xd5 exd5 17.¥f2   
If, instead of having the bishop on f2, White    
could move back his pawn to f2, the position
would be equal.
 

17...£c7 18.¦c1 f6 28.g4! hxg4 29.¤xg4 ¥g7
18...¥h6! 19.g3 ¤f6 was very strong, with a Too passive. Anand, as too often in this
huge positional advantage. match when on the defensive, chooses the
most timid move.
19.¤d3 ¦fe8 20.b3 ¤b6 21.a4 ¤c8 22.c4
£f7 23.a5 ¥f8 24.cxd5 cxd5 25.¥h4 ¤d6 29...¤e4!? 30.¦c7 b5 31.¤e3 (31.¦xa7? £b6
 and Black has an advantage) 31...£b6 32.£c1
would have given a complicated game.
  
   30.¦c7 ¤e4?!
Now this hyperactive move is a mistake.
    30...¦a8! was needed.
 
     31.¤e3! ¥h3?
31...¦c8 32.¦xc8 ¦xc8 33.¤xf5 gxf5
  34.¥g2 ¦c7 and White is better, but Black
    should defend this position.

  32.¦g1 g5

26.a6 b6?
 Chapter 3 – Key Ideas and Positions 135

 40...¦c2??
   The last move in time trouble and the final
     mistake.

   After 40...¦c3 41.¦xd5 ¦xb3 42.¦d8† ¢f7

    43.¦d7† ¢f8 44.d5 ¦a3 Black could save the
    game.

  41.¦xd5

     1–0

   Anand learned a very valuable lesson from

 this match, and in the future his opening
preparation in world championship matches
33.¥g4
33.fxg5!? was also good, but in mutual time was up to the task. Also, in my opinion he
trouble the text move is easier to play. benefited a lot from the fact that computers
came to the fore and it was possible to check all
33...¥xg4 34.£xg4 £xg4 35.¦xg4 the lines to perfection. This is very important if
35.¤xg4!? was also possible, as pointed out you play dynamic chess, which Anand did. The
by other commentators, but it is difficult to classical approach is not so computer-sensitive.
play sophisticated chess in time trouble. When the two best players meet in a world
championship match there is always some
35...¤d6 36.¥f2 ¤b5 37.¦b7 ¦e4 38.f5 luck involved. The key factor for success is still
¦xg4 hidden in the opening phase overall, because
38...¦xd4? 39.¦xd4 ¤xd4 40.¦xa7 ¤xb3 nowadays there is not much room between
41.¦b7 d4 (41...¤c5 42.¦xb6 d4 43.a7) 42.a7 where opening preparation is over and the real
and it’s over. game starts. The concept of opening choices
influences the rest of the game so much that
39.¤xg4 ¦c8 40.¦d7 we can be sure that the right opening is the
40.¢g2 was simple and strong, and I doubt key to success. The battle of two great minds is
that in a practical game Black could defend about the concept of opening choices.
this position.
 Opening concepts
   When people talk about chess they like to
    compare it to art, sport or science. This
imaginary chess can belong wherever the
    player takes it. It is arbitrary; usually it depends
  on the viewer’s perspective. David Bronstein,
    after failing to win the dramatic world
championship match with Mikhail Botvinnik
    in 1951, started to talk about the other values
     of chess, becoming more interested in the
artistic side than the sporting element.
   

136 Grandmaster Opening Preparation
When talking about opening concepts we can
fit chess players into all of these categories.
If you like to play beautiful positions with
nice ideas, but which are not supported by
critics or computers today, you tend to value
the artistic value of the game.
If you are looking in your opening
preparation only at the future value or gain in
Elo-points or tournament wins, then chess is
only a sport for you.
Opening preparation can be done like
scientific research. However, we cannot
demand that a schoolchild work like Botvinnik,
who definitely belonged to the scientific type.
In his training sessions he even tried to mimic
a real game situation and allowed his sparring
partners to smoke during their training games.
Today smoking is not allowed in the playing
venue and this kind of additional training
method is not needed.
We can decide which opening to play by chance
or because of any influence that moves us. This
might be friendly advice, or something noticed
by chance in a book of theory. Initially it may
not matter too much what opening we start
to play. Even so, experienced coaches know
what openings should be best for a beginner.
Yes, they have their own concept of the
opening.
The influence of a leading coach may strongly
influence all their students, and we might say
that some kind of chess school has been created.
Unfortunately, nobody has documented this
kind of occurrence and the name of the chess
school tends to be an emotional one rather
than a scientific one. “Scientific” means that
there is something that we can measure, and in
chess there are things that we can measure – for
instance, we can retrieve a lot of information
from a database.
When talking about players from Riga, say,
we can look at the statistics of the top players
from the Latvian capital: Mikhail Tal, Aivars
Gipslis, Alexei Shirov, Aleksander Wojtkiewicz
and Alexander Shabalov. There are others,
but even this sample is impressive. Are
these players representatives of a Riga Chess
School?
Certainly they are from an emotional
standpoint, but if we work with the data we
would probably end up finding something.
That something is not the moves, but the
concept. All of these players have something
to do with Tal. The legendary attacking player
had an influence on those who followed,
especially Shirov and Shabalov.
A chess school also needs a leading coach
– Grandmaster Vladimir Bagirov, from
Baku, relocated to Riga and started to coach
the young talents. He was a proponent of
Alekhine’s Defence, which Shabalov happily
picked up and played throughout his career.
Typical of his style of play is the following
game.
Juri Vetemaa – Alexander Shabalov
Haapsalu 1986
1.e4 ¤f6
This is not a very popular opening in super-
tournaments. In world championship matches
only Fischer has dared to use it, against Spassky
in 1972.
2.e5 ¤d5 3.d4 d6 4.c4 ¤b6 5.f4
This very active line was prepared by Estonian
national master and coach Juri Randviir for
leading Estonian women players in the 1970s
and 1980s. Unfortunately, Vetemaa also
participated in these training sessions and had
taken it up.
5...dxe5 6.fxe5 c5
Shabalov was never shy of complications and
chooses the riskiest line for Black.
 Chapter 3 – Key Ideas and Positions 137

 17...¤a4 18.¤d4 ¥xd4 19.£xd4 ¤6c5

  20.¥xc4
  
       
      
        
       
     
    
    
7.d5 e6 8.¤c3 exd5 9.cxd5 c4 10.a3
This is too modest.
  

Stronger is 10.d6 ¤c6 11.¤f3 ¥g4 12.¥f4 20...£b5!
and it is difficult to see how Black can escape I was told by Shabalov that his opponent was
White’s grip. really shocked by this move, and it took some
time for him to realize what was going on.
10...¥c5 11.¤f3 0–0 12.¥e2 ¥f5 13.¥g5
£d7 14.£d2 h6 15.¥f4 ¤a6 21.¦d2
 Mate arrives at b3 after either 21.¥xb5
¤b3# or 21.¤xb5 ¤b3#.
  
  21...¤xc3
    0–1
We can conclude that Randviir’s chess school
   lost the battle this time to the boys from Riga.
    A chess school itself is imaginary and does
    not have a physical building. The “school” is
   an abstract idea, and connected more with the
teacher. The Lvov Chess School is associated
    with its coach, Victor Kart. There are a lot of
 famous Ukrainian grandmasters from Lvov –
Black has a good game, but White would Alexander Beliavsky, Oleg Romanishin and
still be doing fine after 16.¦d1. others. However, I have not analysed the
essence of the Lvov Chess School and cannot
16.0–0–0? ¦ac8 17.h3 describe it in depth.
Too slow.
I was first influenced a lot by the Soviet Chess
The only chance was 17.¥xh6 gxh6 18.£xh6, School. My first coach, Tonu Truus, was not
seeking counterplay at any cost. just someone who loved chess and teaching
138 Grandmaster Opening Preparation

children, but he had graduated from the This was the specific line that was in some
Moscow Sports Institute as a certified coach. way the work of the Estonian Chess School.
Tonu Truus also brought valuable contacts Even Ortvin Sarapu, the only other Estonian
with other coaches from Moscow, from which apart from Keres to face Fischer over the board,
I was lucky to benefit during my career. played this line. Juri Randviir was again the one
The Soviet Chess School was first of all a responsible for its popularity among Estonian
conception or a system of how to build up players. The line itself also attracted other
your chess. The classical approach prevailed, players, among them Nigel Short, and even
but it had some restrictions on openings. The Kasparov played it once in his very early years.
Riga players did not really follow the classical
concept of opening choices, despite being

children of the Soviet Chess School.  
The imaginary Estonian Chess School had

one weakness: it had no theoreticians. In my   
juvenile years, opening preparation (or the     
lack of it) was my main problem. Tonu was not
able to build up my opening repertoire when   
I started to advance to the master level of play    
– by today’s standards a rating of about 2400.
Initially the Scandinavian Defence as Black,
 
and some strange English Opening without  
knowledge of any theory was fine. Soon I
changed my Scandinavian to the more solid

5...exd4 6.0–0 ¥e7 7.¦e1
French Defence, and started to play 1.e4 7.e5 ¤e4 is the major alternative.
instead of the boring English Opening.
There was one line in the Ruy Lopez that 7...0–0
was popular among Estonian players, and 7...b5! is the critical line, with play
Latvian players also used it a lot, especially continuing: 8.e5 (8.¥b3 d6 9.¥d5 ¤xd5
Edvins Kengis. This line served me very well 10.exd5 ¤e5 11.¤xd4 0–0! 12.f4? ¥g4 and
until the Under-17 World Championship in Black is better) 8...¤xe5 9.¦xe5 d6 10.¦e1
Belfort in 1979. I started the tournament very bxa4 11.¤xd4 ¥d7 12.£f3 0–0 13.¤c6 ¥xc6
well, winning the first five games – among my 14.£xc6
victims were Nigel Short and Joel Benjamin.
Then I slowed down and finished in fifth place. 
One of the reasons was my limited opening   
repertoire. In the following game the opening   
went well for me, but I still lost the game.   
Jaan Ehlvest – Alon Greenfeld
    
   
Belfort 1979     
 
1.e4 e5 2.¤f3 ¤c6 3.¥b5 a6 4.¥a4 ¤f6
5.d4
   

 Chapter 3 – Key Ideas and Positions
Back in 1979 it was considered that White
had a small advantage here, because of the
better pawn structure. Today we know that
Black has more than sufficient counterplay and
it is White who must be careful to keep the
balance after 14...d5!.
8.e5 ¤e8 9.c3 d6
9...dxc3 10.¤xc3 d6 11.exd6 ¤xd6÷ is a
common alternative.
9...d3 was played in Short – Tempone,
Belfort 1979, the game ending in a draw.
10.cxd4 ¥g4

 
 
  
    
  
   
  
  

11.¤c3!?
11.¥xc6! is the computer’s suggestion. I
was fond of bishops throughout my career
and rarely played against the two bishops
myself. This stubborn attitude probably cost
me some half points, but it was the result of
my emotional reaction in my early training
sessions on seeing how strong the bishop pair
could be in certain positions. Later, when
working with Karpov, I realized why he was
such a flexible and strong player. He liked to
play positions with knight and bishop.
11...b5 12.¥b3 dxe5 13.dxe5 £xd1 14.¥xd1
¤b4 15.¥g5 ¥d8 16.¥xd8?!
Losing the initiative.
139
16.¦e4! ¥xf3 17.¥xf3 ¥xg5 18.¦xb4 would
keep Black under pressure.
16...¦xd8 17.¦e4 ¥xf3 18.¥xf3 c5 19.a4
¤c2 20.¦c1 ¤d4=

  
  
   
   
  
   
   
    

21.¥d1?!
In this game my love of the bishops cost me
dearly.
21...¤c7 22.¤e2 ¤de6 23.¢f1 ¤d5
24.¤c3 ¤df4 25.axb5 axb5 26.¤xb5 ¦d2
27.g3?
27.¥b3 ¦xb2 28.¥c4 should allow White to
defend.
27...¤d3 28.¦c2 ¦xd1† 29.¢e2 ¦d8
30.¤d6 f5 31.¦e3 ¤xe5 32.¢xd1 ¦xd6†
33.¢e1 ¤d4 34.¦cc3 ¦d5 35.¢f1 ¢f7
36.¦a3 ¢g6 37.¦a6† ¢h5 38.¢g2 ¤c2
39.¦c3 ¦d2 40.¦xc5 ¤e3† 41.¢h3 ¤3g4
0–1
It was obvious that I needed to switch my
opening preparation to another chess school!
To some extent I had already done this,
because despite my opening repertoire being
limited, even this was built up with help
from Moscow. Vladimir Yurkov was not a
very extrovert person and he did not publish
140 Grandmaster Opening Preparation
articles in popular magazines. He was a master
himself, but found his calling in one-to-one
high-level coaching. I would credit him with
creating some kind of chess school of his own.
His students or cooperation partners, among
others, were grandmasters Yuri Razuvaev,
Yuri Balashov, Andrei Sokolov, Alexander
Morozevich and myself. I started to work
with him from time to time in 1977, and the
last time he helped me as my second was at
the Moscow World Cup in 2002.
Yurkov kept notebooks for nearly all
openings, but he had some favourites. He was
a representative of the Soviet Chess School,
and a solid opening repertoire was a must.
Yurkov’s most successful student was Andrei
Sokolov, who reached the Candidates final in
1987. With the black pieces Sokolov played
the Nimzo-Indian or Queen’s Indian against
1.d4, and the Sicilian Defence against 1.e4.
With White he opened with 1.e4, and his
attacking skills made life for the Black side
very difficult.
I tried to play several openings against
him, including various Sicilian lines and the
Pirc. I lost nearly all these games. Finally, on
one occasion I prepared the Ruy Lopez. This
time, in 1980 in a USSR junior qualification
tournament, I was successful with my opening
choice and beat him. However, Sokolov still
won the tournament and I shared second
place. I will say more about my battles with
Sokolov from page 237.
In their Candidates match, Karpov’s choice
against Sokolov was the ultra-solid Caro-
Kann Defence. This did not suit Sokolov,
and the theoretical battle was won by Karpov.
With the black pieces, Sokolov defended
isolated pawn positions against Karpov. The
opening positions that evolved from Queen’s
Indian lines were fine for Black, but still
Karpov outplayed Andrei. Like Bronstein,
when failing to beat Botvinnik in 1951,
Andrei never fully recovered from this defeat.
Andrei was a good friend of mine and we
had several training camps together. In
these camps Yurkov worked mainly with
Andrei, and I was not fully introduced to
their analytical kitchen. I had a one-to-one
training camp with Yurkov in 1977, at the
very beginning of my rise to being one of
the top junior players in the USSR. During
that camp I accumulated knowledge mainly
in the Tarrasch Defence and the Sicilian
Scheveningen. From the Tarrasch I learned
a lot about how to play isolated pawn
positions, as we have seen.
The Scheveningen is the cornerstone of the
Sicilian Defence. Understanding not only
the typical positions and plans, but also the
move orders, is essential to mastering the
Scheveningen, since its main position may
be reached from other lines in the Sicilian –
the Najdorf and the Taimanov for instance.
Black’s first critical decision to make is about
the Keres Attack. If we chose the Taimanov
Variation move order, then the Keres Attack is
avoided but 5.¤b5 is one of the critical lines.
You can avoid both problems if you play the
Scheveningen via the Najdorf. Unfortunately,
there are other lines you need to be aware of
when playing the ultra-sharp Najdorf.
I allowed the Keres Attack and I had good
results against it overall. The ideas and plans
looked very attractive to a young player.
Many years later I managed to use one of
these ideas against the strongest woman
player, Judit Polgar, who was also famous
for being an attacking player. Winning
against her with the black pieces in the
Keres Attack was a reward for studying these
lines.
 Chapter 3 – Key Ideas and Positions 141

Judit Polgar – Jaan Ehlvest in the Keres Attack, and this extra tempo has
made every move apart from 6...h6 disappear
Tallinn 2001 from practice.

1.e4 c5 2.¤f3 e6 7.g5

The move order is already important here. This is the old line.
It depends which sidelines bother you most.
After 2...d6 there is 3.¥b5† and after 2...¤c6 The critical line is 7.h4 ¤c6 8.¦g1 h5 9.gxh5
again 3.¥b5. ¤xh5 10.¥g5 ¤f6 and now 11.¦g3!.

3.d4

3.d3 is the most solid sideline, used in  
practice by none other than Bobby Fischer.   
  
3...cxd4 4.¤xd4 ¤f6 5.¤c3 d6     
    
      
    
  
    
     White has scored well from this position.
    7...hxg5 8.¥xg5 ¤c6 9.£d2
     9.h4!? is more precise, not giving Black the
  ...¦h3 idea, and keeping the queen on d1; in
  some lines the queen can move to e2. This
move was recommended by the late Latvian
 GM and theoretician Aivars Gipslis.
6.g4!
This game was played in the Keres Memorial 9...£b6 10.¤b3 ¤e5 11.¥e2 ¦h3!
tournament, so why not test his system? I was This is the idea of ...¤e5.
expecting this, which is the most principled 
move in this position.
 
6...h6
In the Najdorf line it is popular to play a
  
slow version of the Keres Attack starting with    
6.h3 followed by g2-g4. The difference is that     
in the Scheveningen the pawn is still on a7,
and in many lines Black never plays ...a6. Still,    
none of the other moves besides 6...h6 are   
really playable here. This is obvious, because
in the English Attack White needs to prepare
 
g2-g4 with f2-f3, a move which can be avoided    

142 Grandmaster Opening Preparation

12.¥e3 ¦xe3! 18.¤d2 ¤fd7 19.b4?! ¥e7 20.¦a3?! ¥f6

The computer does not like this idea and gives 21.¢d1 ¤g4µ 22.e5 ¤gxe5 23.b5 axb5
instead 12...£c7, when Black is fine. In 1977, 24.¥xb5 d5 25.£a4 ¥e7 26.¦a2 ¦a8
fortunately, there were no computers available 27.¤b3 ¤c4–+
to spoil one’s preparation. I was excited back
then about this idea and never even analysed

the other options. It is a practical sacrifice, and   
even today it looks playable.  
13.fxe3!   
Judit, as I assumed during the game, did not   
know the ...¦h3xe3 idea, but she played the
correct recapture.
  
   
After 13.£xe3 £xe3 14.fxe3 a6 15.a4 g6 Black
has fine compensation.
  
  
13...¥d7 
 28.¦e1 ¢f8 29.¥xc6 bxc6 30.¤d4 ¦a6
   31.£b3 ¤c5 32.£b1 £xh2 33.¤ce2 ¤xe3†
34.¢c1 ¤c4 35.¤f3 £h6† 36.¢d1 ¥d6
  37.¤ed4 ¤e4 38.£b3 ¦a8 39.a6 ¦b8 40.a7
    ¦xb3 41.cxb3 ¤b6 42.¦a6 ¤a8 43.¦xc6
¥c7 44.¤b5 £f4 45.¦xe4 £xf3† 46.¢c1
     £f1† 47.¢b2 £xb5
    0–1

    Throughout my career I played the Scheveningen

 
   

14.a4?!
This looks too aggressive, but Judit was
unable to change her style and start to
consolidate, hoping to realize her material
advantage in the long run.
14...a6 15.£d4 £c7 16.a5 ¥c6 17.¦f1 ¦c8
Black has a fantastic knight on e5 and White
has no clear follow-up. The positional sacrifice
has worked perfectly against a hyper-aggressive
opponent. The rest of the game was very one-
sided.
and also followed the trends. The move order is
a matter of taste and depends on the specific
opponent and the tournament situation. The
only drawback is that the Scheveningen looks
like a passive defence. This is true, but by
practising it you gain a lot of experience in how
to defend against a direct attack.
The Scheveningen was tested in the world
championship matches between Karpov and
Kasparov, starting from 1984. Later, in 1988,
when I was playing in the World Cup series
and Kasparov was one of my opponents, I tried
to find a way to crack the Scheveningen myself.
In Reykjavik I had the white pieces against
Kasparov and it was a crucial game for the
tournament standings. I hesitated a lot, and
 Chapter 3 – Key Ideas and Positions
finally after a few nights working with my
second without finding anything in the English
Attack (Kasparov had won against Nigel Short
in the 1988 Belfort tournament in this line), I
suddenly switched to the English Opening and
lost without a fight.
It was a terrible experience, and probably
most chess aficionados did not know at the
time what had really happened. I tried to refute
at home the opening that I played myself, but
only got confused – and disaster struck. The
next time I had the chance to test Kasparov’s
Sicilian was in 1989 in Sweden, when I
achieved a big opening advantage, but could
not win the game anyway.
The opening concepts I got from Yurkov
served me very well. My only mistake was that
I could not develop and build enough opening
concepts of my own during my career. I was a
player with a Soviet Chess School background
enhanced by the Yurkov school. My opening
preparation was enough to advance towards the
top, but at certain moments my preparation was
not good enough. There were several reasons,
one of which was motivation. I failed to qualify
from the Manila Interzonal for the Candidates
in 1990; the World Cup tournament series shut
down; and the only good news was that the
Soviet Union was on the verge of collapsing.
I got back into top-level chess again in 1995.
Suddenly urgent help was needed, and
again Vladimir Yurkov helped me. Before
the Novgorod super-tournament in 1995, I
received some of his advice for my opening
preparation. Due to the short time we had, he
just tutored me on some sidelines. It was not
that I had any disrespect for my opponents,
but in two games my choice against the French
Defence was 1.e4 e6 and now 2.£e2!?. The
sideline did not give me any real advantage, but
I managed to beat Rafael Vaganian and drew
against Artur Yusupov.
143
Jaan Ehlvest – Rafael Vaganian
Novgorod 1995
1.e4 e6 2.£e2 ¤f6
Vaganian, after Wolfgang Uhlmann and
Viktor Korchnoi, was one of the top specialists
in the French Defence. I beat him in the 1989
World Cup tournament in Skelleftea, Sweden,
in a sharp theoretical line, but as I have
explained, this 2.£e2 line was prepared for
this tournament. We did not consider Black’s
last move at all. My opponent played it very
quickly though, because in an earlier round
against Yusupov I had already used this rare
2.£e2 move.
The most common answer is 2...c5 3.g3 ¤c6
4.¥g2.

 

   
    
   
    

 

3.e5 ¤d5
In a previous game my opponent had
played: 3...¤g8 4.f4 b6 5.¤f3 ¥b7 6.g3 h5
7.d4 Vasiukov – Vaganian, USSR 1981.
4.¤f3 d6
4...c5!? 5.¤c3 ¤xc3 6.dxc3 d5 could be
tried, with an unclear game.
5.d4 ¤e7
After this move it does not matter if the
knight has travelled to e7 via d5 or g8.
144 Grandmaster Opening Preparation

6.h4 b6 7.¤c3 d5 20.¦b1 ¥xd4 21.b4 ¤c4 22.b5 ¥xf2†?

 22...¥xe5 23.bxc6 £xc6 24.¤f3 ¥d6 would
give a complicated game – Black has enough
  compensation.
   23.¦xf2 ¤6xe5 24.£g3 0–0–0 25.¥c3
    In this open position the extra bishop is
    much stronger than the three pawns.
     
       
   
      
  
8.£d1 ¥a6    
8...c5!? 9.¤e2 ¥a6 10.c3 ¤bc6 11.h5 looks
just slightly better for White.     
  
9.¥xa6 ¤xa6 10.¤e2 £d7 11.c3 c5 12.£d3
¤b8 13.h5 h6 14.¥d2 ¤bc6 15.¤f4 ¤a5    
16.0–0 ¤ec6 17.b3 cxd4 18.cxd4 ¥a3 
19.¤e1!? 25...¦hg8 26.¤c2 ¢b7 27.a4 £c7 28.¦e1
During the game I just overlooked ...¥b2, ¤d7 29.¤b4
but it worked out well for me. My opponent White’s attack is decisive.
realized that I had missed his move, but
overestimated his position. 29...¤c5 30.£f3 ¤e4
 30...¤xa4? loses to 31.¤bxd5 exd5 32.¤xd5
£c5 33.¤b4†.
   
      
     
      
      
    
      
       
     
19...¥b2
19...¤b4?! 20.£g3 is good for White, as 
20...0–0? loses to 21.¤fd3! ¤xd3 22.¥xh6.
 Chapter 3 – Key Ideas and Positions 145

31.¦xe4 a5 Jan Timman – Jaan Ehlvest

Black also loses after 31...dxe4 32.£xe4†
¢c8 33. ¦c2. Novgorod 1995

32.bxa6† ¢a7 33.¦xe6 fxe6 34.¤xe6 £d6 1.e4 e5 2.¤f3 ¤c6 3.¥b5 d6 4.d4 ¥d7
35.¤xd8 ¦xd8 36.£f4 d4 37.£xd6 ¦xd6 5.¤c3 exd4 6.¤xd4 g6
38.¥xd4 This was the specific line that was used by
1–0 Yurkov’s students.

I used this concept just to avoid mainstream
theory, but it worked and I shared second
 
place after Kasparov. 
  
Avoiding theoretical opening lines can be too
extreme in some cases. Constantly playing    
1.b3, or going for an early deviation in the    
Sicilian with 1.e4 c5 2. ¤f3 a6, like the
Hungarian GM Richard Rapport, does not
    
promise long-lasting success. Theoretical  
experiments are like bluffing in poker. If your
opponents are aware of it in advance, the
  
surprise effect no longer works and the results 
are disappointing. Successful players should, 7.h4?! ¥g7 8.¥g5 ¤f6 9.¥xc6 bxc6 10.£d2
however, keep sidelines in their arsenal, ready h6 11.¥f4 ¤g4 12.f3 ¤e5 13.b3 £b8=
to use at critical moments. 14.¥e3 a5 15.h5 c5 16.¤de2 g5 17.0–0 a4

Sidelines are boring to study and there may
never be any reward for this work. In 2001
  
during the FIDE World Championship in   
Moscow, Yurkov was my second again. When     
preparing against Daniel Campora, Yurkov
was shocked that I did not know much about    
the sidelines in the Ruy Lopez.   
In 1995, during the Novgorod tournament,
I had played a sideline with the black pieces
  
against Jan Timman. The line belonged to the 
Yurkov school, and many others, including
Balashov and Morozevich, had used this
   
line in their practice. As I mentioned above, 
preparation for this tournament was done in a 18.¦ad1 axb3 19.axb3 0–0 20.¤d5 £d8
very limited amount of time. 21.¤g3 ¦e8 22.¥f2 ¤c6 23.c3 ¥e6 24.£c2
¤e7 25.¤e3 ¦a3 26.¦b1 £a8 27.¤gf5
¥xf5 28.exf5 ¤d5 29.¤xd5 £xd5 30.c4
£a8 31.¦fe1 ¦xe1†
½–½
146 Grandmaster Opening Preparation

Campora was known for playing some passive a difficult task defending his centre. After
lines in the Ruy Lopez, and I did not have 7...exd4 8.¤xd4 ¥e7 9.h3 (or 9.f3 0–0
the slightest idea how to play. I was lacking 10.¥e3) 9...0–0 10.¥e3 White has a nice
knowledge about all these lines, most likely position. Because of this ¥xc6 move, we
because having been one of the strongest cannot play the whole line. There are similar
players for years, nobody had had the courage cases in many sidelines. You need to discover
to test my knowledge in some risky sidelines. or know the refutation.
With a few hours work we fixed some of the
holes in my opening preparation, and I won 4.0–0 d6 5.d4 ¥d7 6.¤c3 ¥e7
convincingly. Campora had not played this line before.
This approach cannot be recommended
today, when players such as Baadur Jobava

and others are willing to take the risk and   
test your opening preparation in these risky 
sidelines. A player must be aware of all the
subtleties of the main line. When confronting    
a sideline, it is a good idea to have something    
prepared beforehand and to counterattack
psychologically with an even rarer line.
   
   
Jaan Ehlvest – Daniel Campora  
Moscow (2.2) 2001   
1.e4 e5 2.¤f3 ¤c6 3.¥b5 ¤f6

7.b3!?
As mentioned above, 3...d6 4.d4 ¥d7 is a
There are other options as well, but I wanted
line I played myself, but nobody discovered
to play some less common move.
and played the correct move against me:
5.¥xc6!
7...0–0 8.¥b2 ¦e8 9.¦e1 ¥f8 10.h3 exd4
 11.¤xd4 d5?
  

     
     
       
      
     
 
   
5...¥xc6 6.¤c3 ¤f6 (6...exd4 7.£xd4 ¤f6  
8.¥g5 ¥e7 9.0–0–0 is a well-known position
where White stands better) 7.£d3! Black has
   

 Chapter 3 – Key Ideas and Positions
12.¥xc6 ¥xc6 13.e5 ¤d7 14.e6 ¤f6
15.¤xc6 bxc6 16.exf7† ¢xf7 17.£d3
¢g8 18.¤e2 ¤e4 19.¤g3 ¤c5 20.¦xe8
£xe8 21.£c3 ¦d8 22.¦e1 £g6 23.¥a3
d4 24.£c4† £f7 25.£xf7† ¢xf7 26.¥xc5
¥xc5 27.¢f1 ¦d5 28.¤e4 ¥b4 29.¦d1 ¢e6
30.¢e2 ¢f5 31.¢d3

    
   
   
  
   
 
 
   

31...h5 32.h4 ¥e7 33.f3 ¢f4 34.¦h1 ¦a5
35.a4 ¦d5 36.¦h3 ¦d7 37.¦g3 ¥xh4
38.¦h3 ¥e1 39.¦xh5 ¦d5 40.¦h8 ¢f5
41.¢c4 ¢g6 42.¦f8 ¦h5 43.¦f4 a5 44.¦g4†
¢f7 45.¢xd4 ¦d5† 46.¢e3 ¦d1 47.¤c5
¥c3 48.¦c4 ¥b4 49.¤e4 c5 50.¤xc5 ¦d5
51.¤e4 c5 52.¤f2 ¥a3 53.¤d3 g5 54.¦e4
¦d8 55.¢d2 ¦h8 56.¢c3 ¦h2 57.¦e2 ¦h4
58.¤e5† ¢g7 59.¤g4 ¦h1 60.¢c4 ¦c1
61.¤e5
1–0
When choosing an opening at grandmaster
level, you can just follow the main trend or work
for yourself, as discussed earlier. However, the
first question is which opening to play? There
is an ECO classification of openings, but they
are not in order of good or bad, risky or solid.
I like to classify openings first into two groups:
classical openings and others.
147
Classical openings are well known, and most
of the chess community has some kind of
agreement about which openings fall into this
category. The Ruy Lopez is the first one that
comes to mind when talking about classical
openings. It is said that mastering the Ruy
Lopez is like getting a higher education in
chess. Of course there are a lot of rare and risky
sidelines in the Ruy Lopez as well.
In the closed openings, the Nimzo-Indian
and the Queen’s Indian are considered to
be classical openings. In general, openings
that follow the classical approach to the
importance of the centre, trying to defend
it from the very first moves, are classical
openings. This is why the Queen’s Gambit
Declined is one of the oldest of them, used in
the world championship match between Jose
Capablanca and Alexander Alekhine back in
1927, and still one of the most popular choices
in the world championship matches between
Karpov and Kasparov during their numerous
encounters.
However, classical openings can be boring, and
players do not usually want to have only solid
openings in their repertoire. Just as Yurkov
influenced a lot of his students’ opening
choices, so did other famous coaches. Mark
Dvoretsky was one of the first to develop
files about certain middlegame positions.
He was not fond of opening preparation,
and in my opinion this has had a certain
influence on his students regarding opening
choices. When talking about his students,
I am referring mostly to Sergey Dolmatov,
Artur Yusupov and Alexander Chernin. They
had one common tendency – with the black
pieces their favourite openings were dynamic
lines with a lot of tactical lines and a quick
climax.
148 Grandmaster Opening Preparation

Fikret Sideif-Sade – Sergey Dolmatov 12.g3! is the correct move.

Rostov 1980 12...£b6 13.a4 0–0–0

1.d4 d5 2.c4 c6 3.¤f3 ¤f6 4.¤c3 e6 
This system became popular thanks to
Botvinnik, and is considered one of the most
   
active systems for Black today.  
5.¥g5 dxc4
  
Dolmatov is following the sharpest line.    
More solid and popular today is 5...h6.   
     
    
    
   
     14.£c1 b4 15.¤d1 c5 16.¤e3 cxd4
    17.¤xc4 £c5 18.¥f4 ¥xg2 19.¦g1 ¥d5
20.¤d6† ¥xd6 21.¥xd6 £xc1† 22.¦xc1†
    ¢b7 23.¦c7† ¢b6 24.¦g7 e5 25.¥c4 ¤xf6
  26.a5† ¢xa5 27.¥xd5 ¦xd6 28.¥xf7 ¦d7
  29.¥c4 ¦xc7 30.¦xc7 ¢b6 31.¦e7

 
6.e4 b5 7.e5 h6 8.¥h4 g5 9.¤xg5 hxg5     
10.¥xg5 ¤bd7 11.exf6 ¥b7 12.¥e2     
     
       
     
       
      
       
  
  31...¢c5 32.b3 ¦h5 33.¦xa7 ¤e4 34.¦a5†
   ¢c6 35.¦a6† ¢c7 36.¦e6 ¤c3 37.¢f1 e4
38.¦e7† ¢d6 39.¦e6† ¢c5 40.¢g2 d3
 41.¦e8 d2 42.¦c8† ¢b6 43.¦d8 d1=£
The aggressive opening choice pays off 0–1
quickly. White is not prepared and Black
quickly achieves a superior position.
 Chapter 3 – Key Ideas and Positions 149

The other opening system they used was the Again, just as in the previous Dolmatov
Sveshnikov Variation of the Sicilian Defence. game, White is not ready for this very concrete
opening line and is quickly lost.
Herbert Armando – Artur Yusupov
12...¤d4! 13.¤c7† £xc7 14.£xa8† ¢e7
Innsbruck 1977

1.e4 c5 2.¤f3 ¤c6 3.d4 cxd4 4.¤xd4 ¤f6    
5.¤c3 e5 6.¤db5 d6 7.¥g5 a6 8.¤a3 b5
  
    
    
       
       
     
      
     
  15.¦d1 ¤xc2† 16.¤xc2 ¥xc2 17.£d5
  ¥xd1 18.£xd1 ¥h6 19.£b1 ¦c8 20.¥e2
£a5† 21.b4 ¦c1† 22.¥d1 ¦xb1 23.bxa5
 ¦a1 24.a4 bxa4 25.0–0 d5 26.¥c2 ¦xf1†
9.¥xf6 27.¢xf1 a3 28.¥b3 d4 29.¢e2 e4 30.f3 d3†
This was the main line back then – the more 31.¢f2 e3† 32.¢e1 ¥g7
solid positional approach after 9.¤d5 ¥e7 0–1
10.¥xf6 ¥xf6 11.c4 was not yet known.
With White, Dvoretsky’s students preferred a
9...gxf6 10.¤d5 f5 11.exf5 ¥xf5 slow, positional approach. Yusupov preferred
 to play closed systems, like the Colle-Zukertort
   Opening and the Torre Attack.

  Artur Yusupov – Peter Scheeren

  Plovdiv 1983
 
  1.d4 ¤f6 2.¤f3 e6 3.e3
With this move Yusupov avoids long
  theoretical lines, and White can still chose
  different plans depending on where he develops
his other knight and whether he pushes the
  c-pawn or keeps it on c2. The position can still
 become very sharp, and the opening choice
12.£f3? is not about avoiding tactical play; instead it
150 Grandmaster Opening Preparation

avoids lines prepared long beforehand by the 18.£xh5! ¥xg5 19.¥xg6 f6 20.f4 £g7
opponent. 21.fxg5 ¤xg5 22.h4 ¤e4 23.¥xe4 dxe4
 24.¦f4
1–0
 
 In the next game, White pushed the c-pawn in
a similar structure and won in positional style.
   
     Artur Yusupov – Milan Drasko
     Sarajevo 1984
   
  1.d4 ¤f6 2.¤f3 e6 3.e3 b6 4.¥d3 ¥b7
5.0–0 d5 6.¤e5 ¥d6 7.f4 0–0 8.¤d2
 In these closed systems it can be difficult
 during the game to figure out where to put
your pieces. Back in 1984 there was no
3...c5 4.¥d3 d5 5.b3 ¤bd7 6.¥b2 b6 7.0–0
computer help, and even if you had the chance
¥b7 8.¤e5
to prepare against your formidable opponent,
8.c4 ¥e7 9.¤c3 0–0 10.cxd5 exd5 11.¤e5
you might still go wrong in the opening phase.
was the other option, but in this game White
keeps the pawn on c2. 
8...a6 9.¤d2 b5 10.¤xd7 £xd7 11.dxc5
  
¥xc5 12.£f3 ¥e7 13.£g3 0–0 14.¤f3  
¦ac8?!    
Better was 14...¥d6 15.¤e5 £e7 16.f4 ¥a3,
when White has a nice position but finds it    
difficult to create a kingside attack.     
15.¤g5! g6 16.£h3 h5 17.¦ad1 ¤h7?
   
  
     
 
8...¤e4 9.c4 ¤d7 10.cxd5 exd5 11.¤xe4
  dxe4 12.¥c4
  This position is already quite unpleasant for
Black.
    
  12...¤f6 13.£b3 £e8 14.¥d2 c6 15.a4 ¤d5

  16.a5 ¦b8 17.a6 ¥a8 18.¦fc1 b5 19.¥xd5

cxd5 20.¥b4 ¥xb4 21.£xb4
  

 Chapter 3 – Key Ideas and Positions 151

 This system suited Yusupov’s solid style. Black

  did not know what was going to happen –
   White could execute various plans: a kingside
attack or just solid positional play in the centre.
    This kind of opening strategy was later used
   by another Soviet Chess School representative,
    Vladimir Kramnik.

     Vladimir Kramnik – Anish Giri

    Stavanger 2017
    
 1.¤f3
Kramnik’s favourite first move, which has
White has a long-term positional advantage
served him for decades.
and convincingly won the game.
1...d5 2.d4 ¤f6 3.e3
21...f6 22.¤g4 £d7 23.¤f2 ¦fc8 24.¤d1
Giri is one of the hardest-working players
¦xc1 25.¦xc1 ¦b6 26.¦a1 ¢f7 27.¤c3 ¥c6
around today, and likes to memorize all the
28.h3 f5 29.¦a5 ¢e6 30.£f8 b4
lines in his opening repertoire. In this game,
 Kramnik is determined to avoid a theoretical
     duel. This solid move does not pretend to gain
an opening advantage, but is just an invitation
   to a long, balanced middlegame.
  
    
     
        
       
         
    
31.¤xd5 ¥xd5 32.£g8† £f7 33.£xf7†
¢xf7 34.¦xd5 ¢f6 35.¦a5 ¦c6 36.¦b5
 
¦xa6 37.¦xb4 h5 38.g4 hxg4 39.hxg4 fxg4 
40.¢f2 ¦a1 41.¦b5 ¦b1 42.¦e5 ¦xb2†
43.¢g3 a5 44.¦xe4 a4 45.d5 ¦b5 46.¦xa4

3...e6 4.¥d3 c5 5.0–0
¦xd5 47.¢xg4 g6 48.e4 ¦d1 49.e5† ¢f7
White could avoid Black’s next move with
50.¦a7† ¢e6 51.¢g5 ¦g1† 52.¢h6 ¢d5
5.b3, but in my opinion Kramnik was actually
53.¦a6 ¦g4 54.¦f6 ¢c5 55.¢g7 ¢d5 56.e6
hoping that Giri would lose his head and make
1–0
the next move.
152 Grandmaster Opening Preparation

 10...¤c6! was the correct move, when Black

  is actually doing fine. Now, I may seem to be
  contradicting my criticism of Giri’s 5th move.
My point is that the move itself was not bad,
    but if you are a player who relies only on lines
    stored in your memory, it is not a good idea
to change this attitude. There are very few
     players in chess history who were truly capable
   of solving new opening problems during the
  actual game, and Giri is not one of them. In
the past, Viktor Korchnoi was one such player
  who could refute his opponent’s ideas over the
 board.
5...c4??
11.a3 a5 12.¥b5† ¢f8 13.¤c5 £b6
Giri makes a practical mistake. Probably he
14.¤xb7 £xb7 15.£e2 g6?
did not give Kramnik any credit for his idea
at this moment. There is a Russian proverb 
that might be translated as, “Let well alone.”
Giri must know some Russian, as his mother
   
is Russian. There was no reason to play ...c4,  
because all normal lines here are fine for Black.   
Psychology? It would be interesting to know
what happened in his head at this point, but   
my assumption is that sometimes everybody     
gets too excited and makes childish moves.
   
6.¥e2 b5 7.b3 ¥b7 8.¤c3 b4 9.¤a4 c3  
10.¤e5    
 
   16.e4! ¤xe4 17.¥h6† ¢e7 18.f3 ¤d2
  19.¦fe1 ¢d8 20.¥f4
1–0
    An opening concept of genius from
    Kramnik, and I am sure that he was influenced
by Yusupov.
   
    Yusupov did not use this concept all the time
 of course. Against stronger opponents he tried
to use the principal mainstream lines. I lost
   against him in 1988 in the Candidates match
 in Saint John, and in our short match he
aimed to play the so-called normal lines. And
10...¥d6?
with Black he did not play the Sveshnikov, but
 Chapter 3 – Key Ideas and Positions
the Petroff Defence, which had been in his
repertoire alongside the Sveshnikov from the
very beginning of his chess career. Flexibility is
a must if you face very strong opposition.
Memorizing opening lines makes more sense
nowadays, as all the memorized lines can be
checked with a computer and you can be sure
you are not memorizing lines that may have
some holes. Yusupov aimed to play lines that
needed to be memorized with the black pieces,
but with White he wanted to play positions
that require the player to make his own
decisions from an early stage of the game.
Chernin adopted the same approach in general,
though with White his main opening choice
was the English Opening. However, Dolmatov
was more principled with the white pieces and
tried to follow the main lines after 1.e4.
Looking into the data we can see that Yurkov’s
best student, Andrei Sokolov, had very good
results with White, beating his opponents with
aggressive play after 1.e4. Dvoretsky’s students
had more even results with both colours.
Playing only tactical lines with Black was not
acceptable to the Soviet Chess School. Soviet
players had better chess education in general
and their coaches did not like openings where
the result was dependent on some tactical
mayhem.
In 1980 I was playing in a Soviet Union
junior qualification tournament in Sochi.
The main contenders for the only spot in the
World Junior Championship were myself,
Andrei Sokolov and Valery Salov. The only
difference between us was that Salov, being
a little younger, was going to play in the
Under-17 World Championship without
having to qualify, and the Sochi tournament
was presented to him as a training opportunity.
Moscow head coach Anatoly Bykhovsky
was sceptical of Salov’s opening choices and
forbade him to play the Sveshnikov. As I did
153
not like playing against the Sveshnikov at all
with White and I knew the head coach’s advice
to Salov, I was furious when Salov still played
the Sveshnikov against me. I lost the game,
and only by winning the last five games was
I able to share second place. Sokolov won
the tournament. Bykhovsky was not Salov’s
personal coach and Valery made his own
choices, so there was no punishment for
disobeying the boss.
Nevertheless, I think a player should listen to
his coach. It might sometimes be difficult to
accept what your coach is telling you, because
the computer evaluation might show a different
story; I think this is a main area of uncertainty
in coaching. You may follow the well-known
approach of one of the chess schools described
above, or create your own approach. The recipe
for success in opening preparation might still
be unclear. One thing is clear, however: you
need to find the right coach. The coach cannot
teach you, Yuri Balashov once said. We, the
coaches, can only help.
The critical or key position in the opening
The Soviet Chess School had numerous
training camps for the elite, which included
members of the Soviet team. In 1980 I was
in one of the junior team camps at the Pirita
facilities in Tallinn, Estonia, back then in the
Soviet Union. The facilities were built for the
1980 Moscow Olympic Games because the
Olympic regatta was held in Tallinn.
The camp included the usual suspects:
Dvoretsky himself and his best students –
Artur Yusupov and Sergey Dolmatov – future
World Champion Garry Kasparov and others,
including yours truly. Dvoretsky was fond of
his middlegame positions and tried to find
victims who would play out the positions.
Kasparov was giving time odds in blitz games.
There was no specific opening preparation.
154 Grandmaster Opening Preparation

That was done separately, and the training


camp, where we were all together, was not the   
place to do it. Nevertheless, one day Kasparov
showed me a position that appeared to be the
 
critical position of the Petrosian Variation of    
the Queen’s Indian.     
Wlodzimierz Schmidt – Adam Kuligowski   
 
Warsaw 1980
  
1.d4 ¤f6 2.c4 e6 3.¤f3 b6 4.a3 ¥b7 5.¤c3    
d5 6.cxd5 ¤xd5 7.e3 ¥e7 8.¥b5† c6 9.¥d3
¤xc3 10.bxc3 c5 11.e4 ¤c6 12.¥e3 cxd4

This was the position that attracted
13.cxd4 ¦c8 14.0–0 0–0
Kasparov’s attention. At that time I did
 not know in which game it had occurred,
   or whether we were just following some of
Kasparov’s own analysis. This game was played
  just a few months prior to our training camp,
   which was held at the end of the summer of
1980.
    
    19.¤d2?!
   19.d5! was Kasparov’s idea, which we
analysed: 19...exd5 20.exd5 ¥xd5 21.¤e5!
    
     
   
15.£e2     
Kasparov played 15.¦a2?! in a game in the    
last round of the World Junior Championship.    
Garry had already secured first place, but he
was not satisfied with his play in this game, and
  
the rook move is not the best. 15...¥f6 16.¥b1   
g6 17.£d3 £d7 18.¦d2 ¦fd8 19.£e2 ¤a5    
20.¤e5 ¥xe5 21.dxe5 £c7 22.¦xd8† ¦xd8 
23.¥g5 ¦d4 24.£e3 £xe5 25.f4 £d6 26.¥e7 This was the point. Our conclusion was that
£d7 27.¥f6 ¦d1 28.¥c2 ½–½ Kasparov – White has a dangerous initiative and has full
Karolyi, Dortmund 1980. compensation for the pawn. This position,
however, has never happened in an actual
15...¤a5 16.¦fe1 ¦c3 17.a4 ¤b3 18.¦ad1 game to my knowledge.
£a8
19...¦fc8 20.¤xb3 ¦xb3 21.f3 ¥b4
 Chapter 3 – Key Ideas and Positions 155

22.¦b1 ¦xb1 23.¦xb1 ¥d6 24.£e1 £b8


25.g3   
½–½  
Kasparov used the Petrosian Variation with     
great success. In the next game he managed to   
use the idea of pushing d4-d5!
    
Garry Kasparov – Miguel Najdorf    
Bugojno 1982   
    
1.d4 ¤f6 2.c4 e6 3.¤f3 b6 4.a3 ¥b7 5.¤c3
d5 6.cxd5 ¤xd5 7.e3 ¥e7 8.¥b5† c6 9.¥d3

20...¥f6
¤xc3 10.bxc3 c5 11.0–0 ¤c6 12.e4 0–0
This looks like the most solid move, but it is
13.¥e3 cxd4 14.cxd4 ¦c8 15.£e2 ¤a5
already not the best.
16.¦fe1 £d6
The experienced Argentinian grandmaster
20...¤c4! The middlegame is all about piece
somehow avoided the position that was
power. You need to improve your pieces and
analysed at the Pirita training camp, but a
limit the power of your opponent’s pieces. In
similar idea is executed anyway.
the opening phase of the game we need first
 to concentrate our attention on the centre.
   The knight on a5 is misplaced and it is very
logical to move it back into the game. White
  should still be able to save the game, but Black
    is absolutely fine.
     21.£g4 ¦ce8?
    Black loses the battle in just a few moves.
   22.¥d2 £xa1 23.¦xa1 ¥xa1
   
       
  
17.d5!?
17.¦ad1 £xa3 18.d5 was also possible.     
17...exd5 18.e5 £e6 19.¤d4 £xe5 20.¤f5
  
White is just concentrating on piece power    
and has sacrificed the whole centre. The    
computer does not approve of this, but in a
practical game Najdorf was unable to solve the    
tactics.     

156 Grandmaster Opening Preparation
24.¤xg7! ¥xg7
24...¥c8 25.¤e6† ¢h8 26.£f5+–
25.¥h6
1–0
Kasparov was very strong in open, dynamic
positions. The only player who could from
time to time stop him was the ninth World
Champion, Tigran Petrosian. In his own
Queen’s Indian system, but playing with the
black pieces, he beat Kasparov in Moscow in
1981. The game is annotated by Kasparov
himself in his book The Test of Time. Kasparov
gained an advantage from the opening, but
Petrosian managed to close the position,
and in a long battle Kasparov lost from a
promising position.
Botvinnik once recommended how to play
against Kasparov. His advice was to choose
closed systems against him. This is important
advice – we are not playing only against the
black or white pieces, but also against the
person moving those pieces. An opening
position that seems critical to one player may
not be critical at all for another.
Having a glimpse of Kasparov’s preparation
made a powerful impression. He showed
me just one position, but how many critical
positions did he have up his sleeve? Years later
I got to see the so-called “stolen analysis” file,
which contained hundreds of opening lines.
This was Kasparov’s team’s work from around
1990-95. The files are still circulating on
the internet if one wants to search for them.
Most of the ideas or lines never occurred in
tournament practice. Theoretical preparation
has a hidden part, like an iceberg.
Once, in a training camp in 1995 with
Karpov, he forced me to look at some
positions in the Ruy Lopez. He had played
these complicated positions against Kasparov
in their world championship matches, but
he was not satisfied with the overall result.
He asked my opinion about a few positions,
which I suspect were critical for the line. We
worked for a while, but he never used or
played these lines anymore. He switched to
the Caro-Kann, and although he was already
a veteran player he had some excellent results
in top competitions, such as beating Anand in
their mini-match in 1998.
What is the right percentage of time for you
to spend on opening preparation, studying
and analysing with a computer and seconds,
for use in tournament practice?
Good opening preparation needs to
be cooked over long hours by the player
himself. You need to find or figure out the
critical positions yourself. You need to have
an intelligent chess-related thinking process.
The opening concept and the lines may come
recommended, and some already prepared
files should be studied, but you need to
understand what is going on yourself. This
approach guarantees the stability of your
results.
The opening phase is still just one part of the
game; it does not solve the chess game. This
is why older, experienced players like to stick
to the lines that were their bread and butter at
the peak of their game. To understand the key
positions and plans from a new modern line
takes time; unfortunately we do not have an
unlimited resource of time.
Critical positions in the Scheveningen
I have a lot of experience with the
Scheveningen. It was the first line that I was
taught professionally by Vladimir Yurkov. The
line itself is very important for understanding
the Sicilian Defence.
I am not talking here about those lines that
are essentially just tactics. Of course, even
 Chapter 3 – Key Ideas and Positions
these tactical lines have their own motifs and
tricks. The difference is that in these lines
the position changes quickly and the climax
is reached in a few moves. Today computers
offer the best assistance for solving the critical
positions that arise from this type of line – the
Najdorf with 6.¥g5, the Dragon with ¥c4
followed by queenside castling, and so on.
To master the Scheveningen you must
master the concepts – the plans and ideas
are important. To find the critical positions
or move orders we need to understand the
essence of the Scheveningen.

 
 
   
    
   
    
 
 

The classical approach teaches us that the
centre is the subject of the game from the
very first move. In the diagram position the
black pawns on d6 and e6 are defending
the centre, but it looks like a passive kind of
defence. White has a free hand to develop his
pieces in several ways and to build different
set-ups.
Black is in passive mode right now
concerning the centre, but there is a plan
to attack the centre. The positional plan is
simple and straightforward. Black wants first
to push the white knight away from c3 and
after that to play ...e6-e5. In this case the
white knight is no longer controlling the
vital d5-square. In the Najdorf Black plays
157
...e7-e5 right away in many lines, weakening
the d5-square. In the Scheveningen Black is
more ambitious: he wants to push ...e5 only
when the circumstances are right.
To drive the knight from c3 Black needs
to play ...b5-b4. White is perfectly aware
of this plan and will try to be ready for it.
The Scheveningen is passive in the sense of
allowing White to build an attack because
White has secured the centre, and the plan
with ...b5-b4 takes time. In some cases,
after ...b4 White has a typical sacrifice on
d5.
Anatoly Karpov – Josif Dorfman
Moscow 1976
1.e4 c5 2.¤f3 d6 3.d4 cxd4 4.¤xd4 ¤f6
5.¤c3 e6 6.g4

 
 
   
    
  
    
  
 

6...¥e7
As this game and other later games showed,
Black cannot get a satisfactory game with this
passive move.
6...h6! is the correct line, as Black needs to
temporarily stop White’s advance.
7.g5 ¤fd7 8.h4 ¤c6 9.¥e3 a6 10.£e2 £c7
11.0–0–0 b5 12.¤xc6 £xc6 13.¥d4 b4
158 Grandmaster Opening Preparation

 Sozin Attack

 
  White has several plans or lines in the
Scheveningen. Besides the Keres Attack, there
  is the Sozin Attack. We should do the same
     research as we did for the isolated pawn in
    the Tarrasch Defence. Historically, the Sozin
Attack, as the name suggests, was created by
     the Russian master Veniamin Sozin, and one
  of the first games was played in 1931.

  Veniamin Sozin – Alexander Ilyin-Zhenevsky

 Moscow 1931
14.¤d5! exd5 15.¥xg7 ¦g8 16.exd5 £c7
17.¥f6 ¤e5 18.¥xe5 dxe5 19.f4 ¥f5 20.¥h3
1.e4 c5 2.¤f3 ¤c6 3.d4 cxd4 4.¤xd4 ¤f6
¥xh3 21.¦xh3 ¦c8 22.fxe5
5.¤c3 d6 6.¥c4 e6 7.0–0 ¥e7 8.¥e3 a6 9.f4?
 The nuances of the system had not been
  worked out yet, and in this early game in the
Sozin Attack the author of the line makes a
   slight mistake.
    
     
      
     
      
      
     
22...£c4 23.¦dd3 £f4† 24.¢b1 ¦c4 25.d6
¦e4 26.¦he3 ¦xe3 27.¦xe3 £xh4 28.£f3
 
£xg5 29.¦e1 £g2 30.£f5 ¦g6 31.¦f1   
£d5 32.dxe7 ¢xe7 33.£f4 a5 34.£h4†
¢e8 35.£xh7 £f3 36.£h8† ¢e7 37.£h4†

9...£c7
¢e8 38.£c4 £b7 39.b3 ¦e6 40.¦g1 ¦xe5
9...d5! takes advantage of White’s inaccurate
41.¦g8† ¢e7 42.£h4† ¢d7 43.£f6 ¦e7
move order. Black was not sufficiently familiar
44.£f5† ¢d6 45.£xa5 ¦e5 46.£d8† ¢e6
with the isolated pawn concept and avoided it.
47.¢b2 f6 48.¦f8 £g7 49.£c8† ¢d5
This is why general knowledge of the openings,
50.£c4†
even those that are not in your repertoire, is
1–0
important.

10.¥b3 ¤a5
 Chapter 3 – Key Ideas and Positions 159

10...0–0 11.£f3 b5 is another option. This is too early. Better was 11...b5!.

  12.axb3 0–0 13.g4 ¦b8 14.g5 ¤d7 15.£h5
g6 16.£h6 ¦e8!
   Suddenly we have a most typical
  Scheveningen position, all about attack and
    the defence of the kingside.
    
    
 
    
  
This is one of the critical positions if Black     
is not fond of the alternative line that occurred
in the Short – Kasparov game given on    
page 164. I was familiar with this position    
thanks to Yurkov back in 1978.
   
11.£f3    
 
  17.¦f3 ¥f8 18.£h4 ¥g7 19.¦h3 h5 20.gxh6
¥f6 21.h7† ¢h8 22.£f2 ¥g7 23.¢h1?
  23.¤f3!? b5 24.¤g5 ¦e7 25.¦d1 gives a
   complicated game.

     23...¤f6 24.¤de2
    White is pulling back his pieces for no
   reason, and Black seizes the initiative.

  

      
  
It is surprising that this position is critical  
even today, the point being that Black can     
reach it by a different move order – specifically
after: 1.e4 c5 2.¤f3 d6 3.d4 cxd4 4.¤xd4 ¤f6    
5.¤c3 e6 6.¥c4 ¤c6 7.¥e3 a6 8.¥b3 £c7   
9.f4 ¤a5 10.£f3 ¥e7 11.0–0?! However, in
this move order 11.g4! is much stronger, and
  
White can hope for an advantage.    
11...¤xb3?!

160 Grandmaster Opening Preparation

24...b5!

The typical plan is finally executed with   
devastating effect.  
25.¦g1 b4 26.¦g5 bxc3 27.¤xc3 ¥b7   
28.£g2 £c6    
0–1
   
Let us consider this typical Sozin Attack   
position.
  
    
  
  13...a5! 14.e5 dxe5 15.fxe5 ¤d7 16.¤xb5
   ¤c5 17.¥xc5 ¥xc5† 18.¢h1 £g5
Black has more than enough compensation
     for the pawn. Fischer managed to draw in 45
   moves.

     ...½–½

  In the Sozin Attack, the sacrifices on d5, e6 and

  

White has two different plans depending
on where he chooses to castle: kingside or
queenside. Another factor is which pawn
White is going to advance: the f- or g-pawn.
Black has the same general plan mentioned
earlier: the advance of the b-pawn.
Robert Fischer – Boris Spassky
Reykjavik (4) 1972
1.e4 c5 2.¤f3 d6 3.d4 cxd4 4.¤xd4 ¤f6
5.¤c3 ¤c6 6.¥c4 e6 7.¥b3 ¥e7 8.¥e3 0–0
9.0–0 a6 10.f4 ¤xd4 11.¥xd4 b5! 12.a3
¥b7 13.£d3
f5 are constantly in the air. The arrangement of
the bishop on the a2-g8 diagonal along with
the powerful knight on d4 is very dangerous.
This is why Black needs to trade at least one of
these pieces. Nearly all Black’s plans of defence
are built on this theme.
There are different aspects of chess that affect
the evaluation of a position. One is the human
factor; a player might not like the position or
have bad feelings about it. The other is the
objective evaluation: the position might not be
playable. This also has different degrees; some
lines are good only up to a certain level.
The Scheveningen itself looks passive and in
many lines the king is under attack. White has
a lot of different attacking and development
schemes. Some of them are critical for
Black; the Sozin is not, because Black has
different move orders to reach the standard
Scheveningen position.
 Chapter 3 – Key Ideas and Positions 161

1.e4 c5 2.¤f3 d6 3.d4 cxd4 4.¤xd4 ¤f6 7.f4? d5! is why White needs to move his
5.¤c3 ¤c6 6.¥c4 bishop first. The classical rule – one must take
 care of the centre – applies here as well.

  7...¤a6!?
  7...¤bd7? 8.¥xe6! is a typical Sozin Attack
sacrifice: 8...fxe6 9.¤xe6 £a5 10.¤xg7†
    ¢f7 11.¤f5 White has more than enough
     compensation.

   8.f4 ¤c5 9.£f3

     
   
    
    
Against the Classical Variation, the Sozin     
Attack makes more sense, because the black
knight is already on c6.    
  
In the Scheveningen, Black can avoid this
specific Sozin position and keep the knight on  
b8 for a while.    
1.e4 c5 2.¤f3 d6 3.d4 cxd4 4.¤xd4 ¤f6

With a complicated game. This line is
5.¤c3 e6 somewhat out of fashion, the main reason
 being that in this move order White can play
  the Keres Attack. Of course, for a young player
who plays the Sozin wherever possible, this
  position might be one of the critical ones.
    The most aggressive approach by White is the
     plan with queenside castling. Here the move
    order is very important. This is why if you
pick some position and consider it as critical
     for a certain line you may miss some subtleties
  about how to actually reach the position.
  In every opening line there are moments
where either White or Black may choose the
 direction. White can choose to castle kingside
The Sozin Attack is not considered a or queenside. In the latter case, White again has
dangerous option here because of the following: a choice: playing the classical plan with £e2,
or the more modern plan with f2-f4 followed
6.¥c4 ¥e7 7.¥b3 by £f3. Let’s look first at the plan with £e2.
162 Grandmaster Opening Preparation

Andrei Sokolov – Valery Salov 8.£e2?

8.¥b3! is the correct move, when White has
Nikolaev 1983 the good option of switching to the f2-f4 plan.

1.e4 c5 2.¤f3 ¤c6 3.d4 cxd4 4.¤xd4 ¤f6 8...£c7 9.0–0–0 ¥e7?
5.¤c3 d6 6.¥c4 e6 7.¥e3 9...¤a5! is the refutation of White’s plan with
 £e2, because after 10.¥b3 b5 Black is already
  threatening the e4-pawn. The difference is that
instead of ...¥e7 Black has already started the
  offensive with ...b5, and White has not yet had
   time to advance the g-pawn.

     10.¥b3 0–0 11.¦hg1

   11.g4 ¤xd4! and now White is forced to
recapture with the rook: 12.¦xd4 (12...¥xd4?
     e5) 12...b5 13.g5 ¤d7
  
    
  
7...a6?   
It may seem too judgmental to brand this    
move a mistake, but it is.    
   
One interesting option here is 7...£c7!?. The
idea is to avoid the £e2 plan, because without
 
having spent a tempo on ...¥e7, Black can    
afford to play an early ...¤a5. 
This is a critical position for this line.
After 7...¥e7 then 8.£e2! is best, but not
8.¥b3 0–0 9.f4, as you will see in Ehlvest – 11...¤d7
Milos below. 11...b5 12.g4 b4 13.¤xc6 £xc6 14.¤d5!
 exd5 15.g5 ¤xe4 16.¥xd5 £a4

   
    
     
        
     
    
      
     
   

 Chapter 3 – Key Ideas and Positions 163

White has several options from here. This The correct defence is 15...b4! and according
position was critical some time ago, when to the computer evaluation Black is doing fine.
computer analysis was not available.
16.g6!! hxg6 17.¦xg6 ¤e5
12.g4 ¤c5 13.¤f5!

   
    
    
   
       
       
     
     
    
 18.¦xg7†! ¢xg7 19.¦g1† ¤g6 20.exf5 ¦h8
13...b5 21.¥d4† ¥f6 22.fxg6 fxg6 23.£g4 ¦h6
13...exf5 14.gxf5 gives White an extremely 24.¥xf6† ¢h7
dangerous attack, for example: 14...¤e5 24...¢xf6 25.£d4† ¢e7 26.£g7† ¢d8
15.¤d5 £d8 16.f4 ¤ed7 17.£g2 ¤xb3† 27.£xh6+–
18.axb3 g6 19.¥d4+– Lanc – Boensch,
Rostock 1984.

   
14.¥d5!?   
Prior to this game, the plan with ¤f5 and
¥d5 was not known, and Sokolov won this   
game in style. The Sozin Attack is not for the   
faint-hearted.
   
In purely chess terms, however, the better     
move is 14.¤xe7† ¤xe7 15.¢b1 ¥b7 16.f3
b4 17.¤a4 ¤xa4 18.¥xa4 ¥c6 19.¥xc6 ¤xc6
  
20.h4 and White has some advantage, because     
the bishop is stronger than the knight in this 
position. 25.¦e1 ¥xd5 26.¤xd5 £c8 27.¦e7† ¢g8
28.¦g7† ¢f8 29.¦g8† ¢xg8 30.¤e7†
14...¥b7 15.g5! 1–0
Only forward.
The next game is an example of the difference
15...exf5 if the black pawn is still on a7.
164 Grandmaster Opening Preparation

Jaan Ehlvest – Gilberto Milos White has an edge in this typical position.

Bali 2000 A similar position can arise from another move

order: 1.e4 c5 2.¤f3 d6 3.d4 cxd4 4.¤xd4
1.e4 c5 2.¤f3 d6 3.d4 cxd4 4.¤xd4 ¤f6
¤f6 5.¤c3 a6 6.¥c4 e6 7.¥b3 ¤c6 8.¥e3
5.¤c3 ¤c6 6.¥c4 e6 7.¥e3 ¥e7 8.¥b3 0–0
¥e7 9.f4 0–0 10.0–0 ¤xd4 11.¥xd4 b5 12.e5
9.f4?
dxe5 13.fxe5 ¤d7 14.¤e4 ¥b7 15.¤d6 ¥xd6
 16.exd6 £g5 17.£e2
  
    
   
  
        
        
       
  
      

 Only the positioning of Black’s queenside
9...¤xd4 10.¥xd4 b5! pawns is different. White had the better game
Black has saved a tempo be keeping his in Short – Kasparov, London (14) 1993.
pawn on a7 and achieves good counterplay by
pushing the b-pawn without ...a6. 13...¦b8
11.e5 dxe5 12.fxe5 ¤d7 13.£f3?! 
White should prefer 13.0–0! b4 14.¤e4 and
now:
  
a) 14...£a5!? 15.£h5 ¥b7 16.¤g5 ¥xg5  
17.£xg5 ¥d5 with an approximately equal    
game.
b) Not so good is: 14...¥b7 15.¤d6 ¥xd6    
16.exd6 £g5 17.£e2     
   
    

       
     
     14.¥xa7?!
14.¤e4 ¥b7 15.0–0–0 a5 and the misplaced
    bishop on b3 only creates problems for White.

    14...¥h4†

 Chapter 3 – Key Ideas and Positions 165

Black has a big advantage and won the game


in 36 moves.  
...0–1   
Somehow I managed to get tricked by the   
Brazilian grandmaster and lost without a fight,    
even though I had already gained quite a lot of
experience with the plan of f2-f4 followed by     
£f3 back in 1988.    
Jaan Ehlvest – Ilya Smirin   
  
Moscow 1988

12...¤d5?
1.e4 c5 2.¤f3 d6 3.d4 cxd4 4.¤xd4 ¤f6
12...b4!? 13.exf6 bxc3 14.fxg7 ¥xg7 15.bxc3
5.¤c3 ¤c6 6.¥c4 e6 7.¥e3 a6 8.¥b3!
£h4†! 16.¥f2 £e4† 17.£e2 ¥b7 18.£xe4
As mentioned above, 8.£e2? £c7 9.0–0–0
¥xe4 19.0–0–0²
¤a5! arrives at a critical position which is a
problem for White.
13.£f3 ¤xe3 14.£xa8 £d7
 
   
   
    
        
        
       
    
      
 
8...¤a5
15.¤cxb5! ¤xg2† 16.£xg2 ¥b7 17.£f1
After 8...¥e7 White even has two choices,
axb5 18.0–0–0 ¥d5 19.¤xb5 £b7 20.¦g1
9.£e2!? and 9.f4, a situation which is better
g6 21.¦d2 ¥e7 22.¤d6† ¥xd6 23.exd6
for Black to avoid. We should prefer to limit
0–0 24.¦g4 £c6 25.£f6 £xd6 26.c4 £c7
our opponent’s choices.
27.¦gd4 £c5 28.b4 £c7 29.b3 ¦a8 30.¦4d3
1–0
9.f4 b5
9...£c7!? is the main alternative.

10.e5 dxe5 11.fxe5 ¤xb3 12.axb3

166 Grandmaster Opening Preparation

Jaan Ehlvest – Konstantin Lerner 

  
Tallinn 1986
  
1.e4 c5 2.¤f3 d6 3.d4 cxd4 4.¤xd4 ¤f6   
5.¤c3 ¤c6 6.¥c4 e6 7.¥e3 a6 8.£e2 £c7
9.0–0–0 ¤a5
    
   
    
   
     
     
16.b3 ¤a5 17.¢b2 ¥b7 18.¤a2 ¦b6
   19.¦hf1 £c5?!
     Much stronger was the simple 19...0–0.

 20.f5 ¤g4 21.¥xe7 ¢xe7

   
     
10.¥d3
10.¥b3 b5 and White can hardly find any
 
aggressive idea here.   
10...b5 11.¥g5
   
The strange 11.¤b3 has never occurred in   
practice, but it is playable. White has not made  
any big mistake yet, so the position remains
sharp, and with deep preparation against an 
unprepared opponent it might be a good idea.   
The point is that today it is so difficult to find
an advantage in the opening that any small

22.¦de1?
idea for one game is welcome.
22.fxe6 fxe6 23.¤c3 ¦b4 24.¦a1 and White
is doing fine.
11...¥e7 12.a3 ¦b8
12...¥b7!? 13.f4 ¦c8 14.¥xf6 gxf6 15.£h5
22...¤e5?
£d7 16.f5 e5 17.¤de2 £c7 and Black was
22...¦b8! was strong and the white king is
better in Ehlvest – Schults, USSR 1986.
under attack.
13.f4 b4 14.axb4 ¦xb4 15.¤f3 ¤c4
23.¤xe5 £xe5† 24.¢a3 ¤c6 25.£f2 ¤d4
Black already has the advantage here. White
26.fxe6 fxe6 27.¤c3 ¢d8 28.¤a4 ¦c6 29.c4
has lost too much time and has not created any
¦c7 30.¦d1 ¥c6 31.¥b1 ¤xb3
threats yet.
 Chapter 3 – Key Ideas and Positions 167

 Bobby Fischer among others, but its popularity

     today is for some reason unjustifiably reduced.
    I myself had good results with the Sozin.

  Thanks to the Sozin, I once gained an opening

     advantage against none other than Kasparov
  himself. Probably this game was the reason why
Nigel Short, in their match in London 1993,
    tried the Sozin. He failed, but it was nothing
    to do with the opening line – he, like many
others, just could not keep pace with Kasparov
  where opening preparation was concerned.
 Jaan Ehlvest – Garry Kasparov
32.¢xb3?
32.£b6! ¤a5 33.¥a2 ¢c8 34.£xa6† ¢b8
Skelleftea 1989
35.¦b1† ¤b7 36.¦b4 and according to the
computer evaluation the position is equal. All
1.e4 c5 2.¤f3 d6 3.d4 cxd4 4.¤xd4 ¤f6
this is irrelevant of course; I just blundered in
5.¤c3 a6 6.¥c4 e6
time trouble.
Here the pawn is already on a6, which I have
criticized in some situations, but the plus for
32...¥xa4†
Black is that the knight is still on b8.
0–1

To sum up, in the classical Sozin Attack when
the knight is on c6, Black can play the normal
 
opening moves and then figure out what the  
critical position is. Actually it is more likely   
that it will be White who needs to determine
the critical position, because Black is on the
    
defensive side and White needs to prove that   
his attack is good enough.
Black has one major choice: either to
    
postpone castling or to do it as quickly as  
possible. Black needs to be aware of the move
orders and not play ...a6 too quickly. The move-
  
order problem is the first thing to be decided. 
Our goal as a Scheveningen player is to reach 7.¥b3
the classical position, but it can be reached via The usual waiting move, because White
the Najdorf or Taimanov, or with move orders wants to keep his options open – specifically
such as 1.e4 c5 2.¤f3 e6 3.d4 cxd4 4.¤xd4 castling queenside.
¤c6 5.¤c3 d6. In the Najdorf the pawn is
already on a6. This is why Black avoids playing 7.f4 is too early because of 7...d5!.
...¤c6 against the Sozin Attack in the Najdorf
after 6.¥c4. The Sozin attracted players such as 7...¤bd7
168 Grandmaster Opening Preparation

7...b5 is one of the main lines. 14.¥c4!

In the endgame the Sozin bishop is relatively
7...¥e7 is also possible, but White can try the useless. There is no kingside attack on the
aggressive 8.g4 against it. agenda anymore.

8.¥g5 h6 9.¥h4 £a5!? 10.0–0 £h5 14...¥e7 15.¥f2 ¦c8 16.¥f1 0–0 17.¤b3
This was the idea behind ...£a5. Because of ¦c6 18.a4 ¤hf6 19.¦d2 ¦fc8?
this, in a later game against Mikhail Tal I opted 19...¤e5! gives an equal game: 20.¤d4 ¦cc8
for 10.f3 instead of castling. 21.¤b3 ¦c6=
 20.¤d4 ¦6c7 21.a5! b5 22.¥xb5! axb5
  23.¤dxb5 ¦c4
  
     
     
       
       
    
      
   
11.£xh5 ¤xh5 12.f3 b6!
Very deep understanding. Usually the
    
Sicilian endgames are fine for Black, but one 
needs to be careful not to create weaknesses. This move was followed by a draw offer.
With the text move Black tries to make the White has a big advantage: the pawns are
queenside pawn structure safe. more valuable than the piece in this position.
However, frustrated by my tournament
13.¦fd1 ¥b7 position, I accepted the offer.
 ½–½

   Keres Attack

  One can just decide that the famous Keres
   Attack is too dangerous, and abandon
    altogether any idea of playing the Scheveningen
    with a move order that allows the Keres Attack.
Chess is not a poker game, but sometimes
   you can bluff. As we saw earlier, even Kasparov
  was capable of bluffing when he chose the
Dragon against Anand.
    When I played the Sicilian, I was expecting

 Chapter 3 – Key Ideas and Positions 169

a full-scale complicated game, and the stronger than your opponent, or in a must-win
Keres Attack was welcome. Practice and situation, for example.
understanding of the key positions helped me
to achieve good results. Let’s look at it step by step.
When playing lines that are not accepted
by mainstream theory, some risk is involved.

Some work of your own is a must. In chess  
history the Polugaevsky line in the Najdorf is  
the best example. Lev Polugaevsky created the
line and wrote a wonderful book about this    
subject.     
   
      
     
    
    
    In the diagram position there are many
different moves, but 6...h6 is the only sensible
     move. Why?
  We should consider the English Attack,
  which evolved around 1988, and whose typical
starting position is the following diagram:
 
However, today the line is less common in  
practice, mostly due to the computer’s help in
analysing forced lines.
 
  
Defending against the Keres Attack helped     
me to develop my chess knowledge and    
character. I started to play it immediately I    
became acquainted with the Scheveningen
at the age of 15. The only major setback was
 
my game against Alexander Grischuk in 2000  
in the Delhi FIDE World Championship. 
I lost this rapid game due to a blunder in a We can easily see that compared to this
drawish rook endgame. Despite my good position, in the Keres Attack White has not
results, the Keres Attack is powerful and Black played f2-f3, but has managed to push g2-g4
faces certain problems. There are some critical right away. The extra tempo is too valuable in
key positions that make Black’s life difficult. the Sicilian Defence, so we cannot really play
However, as I have already mentioned, going anything other than 6...h6.
into these positions must be a calculated risk After 6.g4 h6 White has several options:
and other factors might prevail. You may be the old straightforward 7.g5, the more
170 Grandmaster Opening Preparation

sophisticated 7.h4, and moves such as 7.h3 or


even 7.f3.  
  
Various move orders may reach the same
position where Black needs to decide which   
plan to choose. There is a plan with queenside     
castling and another more aggressive plan with
no castling at all.
   
    
Anatoly Karpov – Ulf Andersson   
Skara 1980  
1.e4 c5 2.¤f3 d6 3.d4 cxd4 4.¤xd4 ¤f6

9.£d2
5.¤c3 e6 6.g4 h6
9.h4 is perhaps a more precise move order,
 not allowing the ...¤e5 idea.
  9...£b6 10.¤b3 a6
   I used only 10...¤e5!? in my practice, trying
    to get counterplay as soon as possible. See
     Polgar – Ehlvest, Tallinn 2001, on page 141.

   11.0–0–0
     
    
    
  
7.g5     
7.h3 a6 8.¥g2 £c7 9.¥e3 b5! is one possible
idea.    
   
7.¥g2 ¤c6 8.h3 ¥e7 9.¥e3 0–0 10.£e2
¤xd4 11.¥xd4 e5 12.¥e3 ¥e6 13.0–0–0 ¦c8
  
14.¤d5 ¤xd5 15.exd5 ¥d7 16.¥e4 ¥g5 gave  
Black a good game in Grischuk – Ehlvest, New
Delhi/Tehran (4.3) 2000.

Here Black is at a crossroads – he needs
to decide which plan to execute: queenside
7...hxg5 8.¥xg5 ¤c6
castling or keeping the king in the centre. I
never played the former – it was too passive for
my taste – but Ulf was famous for his defensive
skills and he was not afraid to defend a slightly
worse position.
 Chapter 3 – Key Ideas and Positions 171

11...£c7 12.f4 ¥d7 13.¥e2 0–0–0 14.h4 Why not 9...¦xh5? It is not important in
¥e7 15.h5 ¢b8 16.¢b1 ¥c8 17.¥f3 ¤g8! most cases, but by taking with the knight
18.¦dg1 ¦h7 19.£f2 ¥xg5 20.¦xg5 ¤ge7 Black has the option of kingside castling in the
21.¦d1 g6 22.hxg6 ¤xg6 23.¦h5 ¦g7 future.
24.¦g1 ¦gg8 25.¦hg5 ¤ge7 26.¥h5 ¦xg5
27.¦xg5 f5 28.exf5 ¤xf5 29.¥g4 ¤ce7 10.¥g5 ¤f6
 
     
      
     
        
       
        
     
      
 
30.a3 ¦f8 31.¥e2 ¤c6 32.¥d3 ¤fe7 33.£e3 11.£d2
£d8 34.¤d2 £c7 35.¦h5 £d8 36.¤e2 ¤d5 This is the most popular option, but it is
37.£g3 £a5 38.f5 e5 39.¤c4 £c7 40.¦h6 difficult to understand why.
¤f6 41.¤c3 £e7 42.¤e3 ¤d4 43.¦g6
£d8 44.¦g5 ¤h7 45.¦g7 ¤f6 46.¦g5 ¤h7 The correct move is 11.¦g3!.
47.¦h5 ¤f6 48.¦h6 ¤g8 49.¦h5
½–½ 11...£b6 12.¤b3
In this position ...¤e5 does not make any
Ulf defended this plan with queenside castling sense, because there is no ...¦h3 anymore.
twice against Karpov, and in their third
encounter Karpov gave up the 7.g5 move and 12...a6 13.0–0–0 ¥d7
played 7.¦g1 instead, but still could not win
the game.

This was a psychological success for Black,   
and soon other plans evolved. Specifically,  
White started to play 7.h4 instead of 7.g5.
 
Zhu Chen – Jaan Ehlvest     
Gibraltar 2007    
   
1.e4 c5 2.¤f3 e6 3.d4 cxd4 4.¤xd4 ¤f6
5.¤c3 d6 6.g4 h6 7.h4 ¤c6 8.¦g1 h5   
9.gxh5 ¤xh5   

172 Grandmaster Opening Preparation

Here my opponents tried several moves, but Andrey Vovk – Alvaro Valdes Escobar
the conclusion – White is better, but Black has
enough counter-chances – never changed. Shenzhen 2011

14.¦g3 £c7 15.¥g2?! ¤h5 16.¦h3 ¦c8 1.e4 c5 2.¤f3 d6 3.d4 cxd4 4.¤xd4 ¤f6
17.¢b1 b5 18.a3 b4 19.axb4 ¤xb4 20.¥f1 5.¤c3 ¤c6 6.¥g5 e6 7.£d2 a6 8.0–0–0 h6
e5 21.¦h2 ¦b8 22.¥h3 ¥c6 23.¤d5 ¥xd5 9.¤xc6 bxc6 10.¥f4 d5 11.£e3 ¥e7 12.¥e2
24.exd5 ¥e7 25.£c3 £b6 26.¥e3 £b5 0–0 13.h4
27.¥f1 £a4 28.¥a7 ¦a8 29.£c7 
  
      
     
       
      
        
    
      
  
 13...¦e8 14.g4 ¤h7 15.£g3 ¥b7 16.g5 ¦c8
29...0–0 30.£xe7 ¦fc8 31.f3 £a2† 32.¢c1 17.¦dg1 d4 18.g6 fxg6 19.£xg6 ¥f6 20.e5
£xb3 33.¦dd2 £a2 34.¢d1 £b1† 35.¢e2 dxc3 21.¥d3 ¤f8 22.£xh6 cxb2† 23.¢b1
¤g3† 36.¢f2 ¤xf1 37.¦h1 ¦xc2 38.¦xc2 ¢f7 24.£h5†
¤d3† 39.¢g2 ¤f4† 1–0
0–1
And now let’s go back to the Keres Attack.
In this game a typical Sicilian battle occurred,
and I do not want to focus on the mistakes Ismael Teran Alvarez – Jaan Ehlvest
made by White. Today everybody can look
Gibraltar 2007
at the game using a computer and can point
out that theoretically White has the better
1.e4 c5 2.¤f3 e6 3.d4 cxd4 4.¤xd4 ¤f6
game. This is true, but the same goes for the
5.¤c3 d6 6.g4 h6 7.h4 ¤c6 8.¦g1 h5
Anand – Kasparov game we saw on page 130.
9.gxh5 ¤xh5 10.¥g5 ¤f6 11.¦g3!
My opponent definitely had not studied this
This is the main problem in the Keres Attack.
position for a week.
This waiting move creates a lot of problems for
Black.
However, there is more bad news for Black
in the Keres Attack. Let’s first examine the
following game.
 Chapter 3 – Key Ideas and Positions 173

 12...bxc6 13.£f3 d5 14.0–0–0 £a5 15.¢b1

  ¦b8 16.b3 ¦b4 17.e5 ¤h5 18.¦g1 £c7
   19.£e3

   
       
       
       
     
       
    
11...¥e7!?   
I was trying to trick my less experienced
opponent, and I succeeded.
 

Black cannot support the c6-knight with the 19...f6 20.exf6 gxf6 21.¥h6 ¦xh4 22.¥d3
bishop: 11...¥d7?! 12.¤db5! £b8 13.¦d3!± ¢d8 23.¤e2 ¥a3 24.¦h1 ¦xh1 25.¦xh1
The d6-pawn is difficult to defend, which is ¤g7 26.¦g1 ¤e8 27.¥f4 e5 28.¥g3 £b6
the other point behind ¦g3!. 29.£d2 ¥d7 30.¦d1 ¤d6 31.c4 d4 32.£c2
¢c8 33.f4 £c5 34.fxe5 fxe5 35.¦f1 £a5
11...a6 is the main move, but after 12.¤xc6 36.c5 ¤b5 37.¥xe5 ¦e8 38.¥xd4 ¤xd4
bxc6 13.£f3! I could not find any good plan, 0–1
despite analysing it at home for some time.
 In the above games I was still successful,
despite the fact that the players with the white
  pieces had some theoretical advantage from
    the opening.
 
     Positional ideas
   
In the Sicilian Defence there are some
    positional concepts as well, but most of the
   time concrete evaluation prevails.
    In closed systems, the plan or set-up should
 be decided before analysing the concrete lines.
This position resembles the above one from One feature of computer engines is that they
the Rauzer. Black has the centre, but here the are not so good in closed positions, so you
kingside is already open, which makes Black’s need to guide the program a lot.
life difficult.
One solid opening is the Slav Defence. I played
12.¤xc6 it myself for some time, but quit when I lost
12.£d2 was stronger. to Yusupov in our Candidates match in Saint
174 Grandmaster Opening Preparation

John in 1988. The Slav was used by Petrosian
against Spassky in their world championship
match in 1969. The book of the match was
like a bible for my first coach. In one of their
encounters, the Slav was annotated deeply
by both seconds – Isaac Boleslavsky and
Igor Bondarevsky. Bondarevsky offered an
interesting positional plan.
Instead, either 13...exd5 14.¤e1 ¥xe1! or
13...¥xd3! 14.£xd3 cxd5= should be fine for
Black
13...fxg6!
13...hxg6 14.¥d2 ¥e7 15.h4! ¥xh4 16.g3
¥e7 17.¢g2 gives White a dangerous attack
on the h-file.

14.¥d2 ¥e7
Boris Spassky – Tigran Petrosian

Moscow (7) 1969
 
1.d4 d5 2.c4 c6 3.¤f3 ¤f6 4.¤c3 dxc4 5.a4  
¥f5 6.e3 e6 7.¥xc4 ¥b4 8.0–0 ¤bd7 9.£e2
0–0
 
9...¤e4 10.¥d3! is good for White.    
10.e4 ¥g6 11.¥d3
   
A very well-known theoretical position.    
   
      
 
Spassky did not find any plan here, and just
  tried to make good moves.
     15.£e4
   15.a5 a6 16.¤e4 h6 17.¤e1!
   
    
      

    
11...¦e8
11...¥h5 is the most common move.    
    
Also possible is 11...h6, which I played myself   
against Lerner in 1986.    

12.e5 ¤d5 13.¥xg6 This was the set-up recommended by
13.¤xd5 cxd5?! 14.¥xg6 hxg6 15.¥d2 ¥e7 Bondarevsky.
16.¦fc1 ¤b8?! 17.£b5! would allow White an
edge. 15...¤f8 16.a5 ¦c8 17.¦fc1 a6 18.£g4 h6
 Chapter 3 – Key Ideas and Positions 175

19.¤e4 £d7! 20.h4 ¦ed8 21.¦c4 £e8 not matter much: the plan White wants to
22.¦ac1 execute is still the same.
22.b3! was the last attempt, with the idea of
¥c1-a3. 12.e5 ¤d5 13.¤e4 ¥e7 14.¤g3 ¥g6
15.¥xg6 fxg6 16.a5! a6 17.¥d2 h6 18.¤e4
22...¦c7 23.¤g3 £f7 24.¤e1 ¦cd7 25.¤f3 g5 19.¤e1 ¦f7 20.¤d3 ¤f8
¦c7 26.£e4 ¦cd7 27.£g4 White has achieved the Bondarevsky set-up.
½–½ 
I used this set-up as White in several games,   
and it was also picked up from me by Lembit   
Oll – a player who knew nearly everything.
 
In the next game I used the correct set-up but,    
instead of winning a good game, I messed up
and only won the game because of a terrible
   
blunder from my opponent. I still managed    
to drink champagne, not to celebrate this
particular game, but because I won the Reggio
  
Emilia super-tournament ahead of Karpov and    
Ivanchuk, and this game proved important. 
21.¦fd1
Jaan Ehlvest – Margeir Petursson Protecting the d-pawn, just in case. I did not
have any computer assistance back then, and
Reggio Emilia 1989
like Spassky I could not find the most precise
follow-up.
1.d4 d5 2.c4 c6 3.¤f3 ¤f6 4.¤c3 dxc4 5.a4
¥f5 6.e3 e6 7.¥xc4 ¥b4 8.0–0 ¤bd7 9.£e2 21...¦b8 22.£g4 ¢h7 23.¤dc5 ¤c7
¥g6 10.e4 0–0 11.¥d3 ¥h5 The computer prefers White here, but as I have
 explained elsewhere, the computer evaluation in
   closed positions can be misleading.

 
       
     
    
       
     
        
    
In closed positions the different moves do    

176 Grandmaster Opening Preparation

24.¦a3 £d5 25.¦h3


This plan with ¦a3-h3 was too ambitious.     
Black’s position is too solid and the direct
attack is not going to succeed.
   
  
25...¢g8 26.b4 ¤g6 27.g3?    
After good opening play I spoiled my
position in a few moves. Still, in this game luck   
was on my side.     
27...¥xc5 28.bxc5 ¤b5 29.¥e3 ¦f5
   
     
    
43...¦xc5?? 44.¦xd4 ¦xa5 45.¦d7 ¦b5
    46.h5
 1–0

  My compatriot Lembit Oll produced a much

   neater game, although we never actually
analysed this specific idea together.
    I should also confess that I used this plan
     once more, winning against Barua at the
    Yerevan Chess Olympiad in 1986. However,
at one point in the game I was totally lost, but
 my opponent blundered in time trouble, just
30.£h5? like Petursson.
Another mistake in mutual time trouble. What can I say? Opening preparation is
important, but like they say, against a strong
30...¤xe5 opponent before you count your chickens
30...£xe4 31.£xg6 £g4 32.£h5 £xh5 you need to win all three battles: opening,
33.¦xh5 ¦f3 34.¢g2 ¦bf8 35.g4 and White middlegame and endgame.
should just save the game.
Lembit Oll – Aleksander Wojtkiewicz
31.dxe5 £xe4 32.£g6 ¦xe5 33.¦xh6 £xg6
34.¦xg6 ¢f7 35.¦xg5 ¦d5 New York 1994
Now it is an equal endgame.
1.d4 d5 2.¤f3 c6 3.c4 ¤f6 4.¤c3 dxc4 5.a4
36.¦a1 ¦bd8 37.¢g2 ¦xg5 38.¥xg5 ¦d3 ¥f5 6.e3 e6 7.¥xc4 ¥b4 8.0–0 ¤bd7 9.£e2
39.¦a4 e5 40.¦e4 ¦d5 41.¥e3 ¢f6 42.h4 ¥g6 10.e4 0–0 11.¥d3 ¥h5 12.e5 ¤d5
¤d4 43.g4 13.¤e4 ¥e7 14.¤g3 ¥g6 15.¥xg6 fxg6
16.¤e4 h6 17.a5 a6 18.¤e1 £e8 19.¤d3
g5
 Chapter 3 – Key Ideas and Positions 177

 
     
   t+  
  P
     +  
        
        
     
       
 
20.b4!? 25.¦a3 ¦f6 26.¦g3 £f7 27.¥d2 £e8 28.£e4
The idea is just to play ¤dc5 next move. Black £h5 29.¦b3 £h4 30.f3 £xe4 31.fxe4 ¦xf1†
was too careless with 18...£e8, overestimating 32.¢xf1 ¤f6
his position. The correct plan was to defend
the b7- and e6-pawns as quickly as possible,

so that after ¤dc5 Black is not forced to trade    
on c5.   
20...£g6  
20...¤xb4 21.¤xb4 ¥xb4 22.¦b1 ¥xa5     
23.¦xb7 gives White a decisive attack.
   
21.¤dc5 ¤xc5 22.bxc5    
Black’s position would be fine if there were
no rooks on the board, but with them it is just    
positionally lost.    
 
   33.d5!
An attractive final breakthrough.
   
 33...cxd5 34.e5 ¤e4 35.¢e2 ¤xc5 36.¦c3
    ¤e4 37.¦c7 ¦d8 38.¥e3
1–0
   
     There are some closed positions where both the
critical plan and the move order are important.
   To find such a plan is not easy, and deep
    understanding of the position is required. You
cannot find these ideas by accident: you need
 to work a lot with the data and then perhaps
22...¦ad8 23.¤d6! ¥xd6 24.exd6 ¦d7 some spark will light up.
178 Grandmaster Opening Preparation

Baadur Jobava – Magnus Carlsen prepared to sacrifice a pawn. This is the idea
behind 7.¤d2.
Internet (rapid) 2017
A possible plan is 7...¤c6 8.e3 ¤e7, but still
1.d4 ¤f6 2.c4 e6 3.¤f3 b6 4.¤c3 ¥b4 White is somewhat better in this position.
5.¥g5 h6 6.¥h4 ¥b7
 The ...£e7 move looks strange at first sight, but
   I believe Carlsen’s idea was to improve on the
immediate 7...¥xc3 (which has been played in
  several previous encounters, including some
    games of Nisipeanu which are well worth
studying). Black is playing flexibly, making
     a generally useful move while keeping the
    option of converting to the ...¥xc3 positions
    under improved circumstances.

  8.£c2!?

  After 8.f3 then 8...d5 is possible, as we will
see in the next game.

In this well-known theoretical position, 8.a3 ¥xc3 9.bxc3 g5 10.¥g3 d6 11.f3 ¤bd7
Georgian players found a new idea. 12.e4 0–0–0

7.¤d2!?
7.e3 g5 8.¥g3 ¤e4 9.£c2 ¥xc3† 10.bxc3 d6
   
11.¥d3 f5 12.d5 ¤a6! (a modern improvement  
over the older 12...¤d7) 13.¤d4 ¤dc5    
     
     
       
       
    
   
    Here it is obvious that ...£e7 fits with Black’s
  plan.
    8...c5 9.a3 ¥xc3
 9...¥a5!? is also possible.
This is a critical position, but practice has
shown that Black has good counterplay. 10.£xc3
10.bxc3 d6 11.e4 g5 12.¥g3 ¤c6 13.¤b3
7...£e7!?
leads to a complicated game. Black can even
The direct attack with 7...c5 is not good
consider 13...¤xe4!?.
because of 8.d5! g5 9.¥g3 and White is
 Chapter 3 – Key Ideas and Positions 179

10...cxd4 11.£xd4 ¤c6 I am sure that the World Champion was aware
11...d6 12.e4 e5! 13.£d3 ¤bd7 14.f3 0–0 of the next game. In this earlier game the more
15.¥e2 ¤c5 16.£c3 ¦ac8 is fine for Black. principled move 8.f3! was played.

12.£c3 ¦c8 13.e3 0–0 14.¦d1 d5 15.¥xf6 Alexander Moiseenko – Ruslan Ponomariov
gxf6
Despite the doubled pawns Black is Sochi 2006
better here; the lead in development is more
important. 1.d4 e6 2.c4 ¤f6 3.¤c3 ¥b4 4.¤f3 b6
 5.¥g5 ¥b7 6.¤d2!? h6 7.¥h4 £e7!? 8.f3!
It seems that this is the correct way for
   White.
   
     
     
       
         
       
     
  
16.cxd5 ¤e5 17.£d4 ¦fd8 18.¥e2 ¦xd5
19.£h4 £d8 20.£xh6 ¤d3† 21.¥xd3 ¦xd3
 
22.e4 ¥xe4 23.0–0 ¥g6 24.h4 
 8...d5
This is the logical follow-up, but another
   plan with 8...c5 9.a3 ¥xc3 10.bxc3 d5 also
    deserves serious attention. It looks counter-
intuitive to open the position when White
   has the bishop pair, but Black has the better
     development.

     9.e3 0–0 10.¥e2 e5?!

    10...c5! 11.a3 ¥xc3 12.bxc3 ¤c6 13.0–0
    ¤a5 The position is complicated. The little
move ...£e7 creates a lot of lines which take
   time to analyse – one week at least!
 11.0–0 exd4 12.exd4 ¦e8 13.¦f2 g5
24...¦c5 25.b4 ¦cd5 26.h5 ¦xh5 27.£f4
¦hd5 28.¤e4 ¥xe4 29.¦xd3 ¥xd3 30.¦c1 13...£e6 14.¥d3 ¤bd7 15.cxd5 ¤xd5
¢g7 31.¦c7 16.¤xd5 ¥xd5 17.¦e2 £c6 18.¤e4 may be
0–1 slightly better for White.
180 Grandmaster Opening Preparation

14.¥g3 ¤c6 15.¤b3 37.¢f2 ¢f5 38.¢g3 ¢e4 39.¦f1 ¤xa3

15.a3!? might keep some advantage for 40.¦f6 ¥d5 41.¦xh6 ¤b5 42.¥b8 ¤xc3
White, for example: 15...¥d6 16.¥xd6 cxd6 43.¥xa7 b5 44.h4 ¤e2† 45.¢g4 ¢d3
17.cxd5 ¤xd4 18.¤de4! (18.¤c4 ¤xe2† 46.h5 ¥e4 47.¥c5 f5† 48.¢g5 f4 49.¦e6
19.¦xe2 £d7 20.¦xe8† ¦xe8 21.h4÷) f3 50.¦xe4 ¢xe4 51.h6 ¤f4 52.h7 ¤e6†
18...¤xe2† 19.¦xe2 ¤xe4 20.fxe4± 53.¢g4 ¤xc5 54.h8=£ f2 55.£h1† ¢e3
56.£f3†
15...dxc4? 1–0
15...¦ad8!= would keep the position balanced.
When starting to look at an opening line, the
16.¥xc4 ¦ad8 17.£d3 ¢g7 18.f4± first step is to study the games of prominent
 players. A theoretical book can be useful to
provide an overview of the structure of the
    particular line; visually seeing the tree in a
   book helps you to keep track of your work.
In the past the Encyclopedia of Chess Openings
    had this kind of tree structure, though with
     comments using only symbols. An opening
    textbook will have a lot of comments, but the
conclusions can be misleading. A player needs
   to learn the opening himself; the book should
   serve only as a reference.
During my long career I cannot name even
     one opening book that I really studied. This
 does not mean that they are completely useless,
18...¤g4 19.¦e2 £d7 20.d5 ¦xe2 21.£xe2 but they are just additional learning material.
¦e8 22.£d2 ¤a5 23.¥b5 ¤xb3 24.¥xd7 If a young player does not understand what he
¤xd2 25.¥xe8 gxf4 26.¥e1 f3 27.¥c6 ¥a6 or she is reading, it only hurts their studies.
28.gxf3 ¤xf3† 29.¢g2 ¤fe5 30.¥g3 ¤e3† The coach should always supervise opening
31.¢h1 ¤3c4 32.a3 ¥xc3 33.bxc3 ¤xc6 study. From my experience I would say that
34.dxc6 ¥b5 35.¥xc7 ¥xc6† 36.¢g1 ¢g6 even at a level of around 2000-2100, a player
is not capable of teasing out the essence of the
 opening line just by reading a book. Working
     alone with the data might also not lead to
    the right conclusion. The helper (the coach),
the book and the data should be combined.
   Finally, there is the computer evaluation,
     which may spoil an idea or plan that is already
approved and recognized.
    The next game has it all.
    
    
    

 Chapter 3 – Key Ideas and Positions 181

Lajos Portisch – Anatoly Karpov 13...¥xc3!?

This move needs deep explanation. At some
Bugojno 1978 level we always prefer the bishop over the
knight, but with this move Black gives away
1.d4 ¤f6 2.c4 e6 3.¤c3 ¥b4 4.e3 0–0 the bishop pair and strengthens White’s centre.
5.¥d3 c5 6.¤f3 d5 7.0–0 dxc4 8.¥xc4 cxd4 However, Black’s position is solid and he hopes
9.exd4 b6 to put pressure on the white centre. Today we
 can see many games with this theme.
  14.bxc3 £c7 15.c4
   15.¤d2! is the correct plan.
    
       

       
        
       
      
   
Positions with an isolated pawn are one of    
the most difficult types to master, but it is 
essential if the player hopes to advance to the White wants to use the e4-square and, after
next level. f2-f3 and ¤e4, retake with the f-pawn on e4.
In this case Black’s counterplay with ...e5 will
10.¥g5 ¥b7 11.¦e1 ¤bd7 12.¦c1 ¦c8 not be so effective.
13.¥d3
 15...¦fe8
15...¤g4? is just a shot in the air: 16.h3!
   ¥xf3 17.£xf3 £h2† 18.¢f1 and White is
 winning.

    16.£e2 h6
     16...£c6 17.h3 ¦cd8 18.¦cd1 ¤f8 19.£e5
¤6d7 20.£g3± Gligoric – Portisch, Lugano
     1968.
  
   17.¥d2
17.¥h4 £f4 18.£e3 £g4 19.h3 £h5 and
    Black has good counterplay.

182 Grandmaster Opening Preparation

 although the computer likes White’s position)

  20...e4 21.£h3 Again, the computer prefers
  White’s position.

    19...exd4 20.¦xe8† ¤xe8 21.¥f4 £c6

     22.¥f5 ¦d8 23.h3 ¤c5 24.¦d1 £f6
Black has comfortably dealt with White’s
    attacking attempts and is a clear pawn up.
   
    
        
     
The following idea was praised back in 1978
as a brilliant plan executed by Karpov.    
   
17...¥xf3!
Giving up the other bishop.
   
  
18.£xf3 e5
This was Karpov’s positional idea.
   
 
25.¥b1 £e6 26.¢h2 ¢f8 27.¥e5 £xc4
  28.£f4 ¤e6 29.£e4 £d5 30.£e2 ¤d6
   31.a4 ¤c4 32.¥g3 ¤c5 33.¥a2 d3 34.£e1
£d4 35.f3 ¤e3 36.¦d2 ¦e8 37.£c1 ¤xa4
     38.¢h1 ¤c5 39.¥f2 £e5 40.¥b1 ¢g8
     0–1

    This game is difficult to grasp, even for a strong

  
  
    

19.£g3?
Portisch probably chose not to play 19.d5!
because it is counter-intuitive – closing the
position with two bishops just does not look
like the correct plan. Still, this was the best
choice, and White has a good game: 19...¤c5
20.¥f1! (20.¥c2 e4 gives Black some initiative,
grandmaster. What is the final truth – should
I play like Karpov or not? What is the final
verdict on these critical positions?
We can simply answer that “it depends”
and ignore these questions. The lesson here
was about the pattern or idea executed by
Karpov. The game of chess is fortunately so
complicated that there are no ready answers for
every situation. Knowing the opening patterns
just helps to improve one’s game.
In the next example, Karpov demonstrated
how to play against White’s hanging pawns.
 Chapter 3 – Key Ideas and Positions 183

Mark Taimanov – Anatoly Karpov 15.bxc3 ¦c7 16.¦ac1 £c8 17.£a4

Moscow 1973 

  
1.d4 ¤f6 2.c4 e6 3.¤c3 ¥b4 4.e3 c5 5.¥d3
0–0 6.¤f3 d5 7.0–0 dxc4 8.¥xc4 cxd4
 
9.exd4 b6 10.£e2 ¥b7 11.¦d1 ¤bd7    
     
      
    
      
        
    
    17...¦c4!
A very strong positional pawn sacrifice.
 
    After 17...£b7 18.c4 ¦fc8 19.¥f4 ¦c6 20.h3
White is doing fine; the position is about equal.

12.¥d2 18.£xa7 £c6 19.£a3 ¦c8 20.h3 h6 21.¦b1
A very timid move, typical of Taimanov. ¦a4 22.£b3 ¤d5 23.¦dc1 ¦c4 24.¦b2
f6 25.¦e1 ¢f7 26.£d1 ¤f8 27.¦b3 ¤g6
12...¦c8 13.¥a6 ¥xa6 14.£xa6 28.£b1 ¦a8
 
      
     
    
        
        
     
     
       
 
14...¥xc3! 29.¦e4 ¦ca4 30.¦b2 ¤f8 31.£d3 ¦c4
A strong move, seemingly improving White’s 32.¦e1 ¦a3 33.£b1 ¤g6 34.¦c1 ¤xc3
pawn structure, but instead of an isolated pawn 35.£d3 ¤e2† 36.£xe2 ¦xc1† 37.¥xc1
White will now have hanging pawns with no £xc1† 38.¢h2 ¦xf3 39.gxf3 ¤h4
prospects. 0–1
184 Grandmaster Opening Preparation

Studying these classical games teaches a lot,


even to very strong players. I had the chance   
to use these motifs against none other than
Vassily Ivanchuk in our match in the FIDE

World Championship in 2000. I had the    
fortune to win the decisive rapid game and     
thus the mini-match.
   
Vassily Ivanchuk – Jaan Ehlvest   
New Delhi/Teheran (2.4) 2000   
   
1.c4 c6 2.e4 d5 3.exd5 ¤f6 4.¤c3 cxd5 5.d4
e6 6.¤f3 ¥b4 7.cxd5 ¤xd5 8.£c2 £c7!?

14.¤h4?
9.¥d2 ¤d7 10.¥d3
My opponent tries too hard. This move is
 not good and only Vassily’s bad nerves can
  explain such an antipositional plan.

 14...0–0 15.f4 ¥d5 16.¤f3

    White pulls back, admitting that his plan
    does not work.

     16.c4 was the correct move, and after 16...¦ac8

   17.¦ac1 ¥a8 the game is complicated. White
should not be tempted by 18.f5? e5 when
  Black is better, as the a2-h7 diagonal is closed
    by White’s own pawn.

 16...¥c4 17.¤e5 ¥xd3 18.£xd3 ¦ac8

10...¥xc3!? 19.¦f3 ¤d5 20.£b5 ¦fd8 21.¦af1
10...¤f4?! 11.¥xf4 £xf4 12.0–0 is
unpleasant for Black, for example: 12...a6?! 
13.¦fe1 ¥xc3 14.bxc3 g6 15.¦e4 White had a   
serious initiative in Spacek – Dejkalo, Prague
1989.
 
   
11.bxc3 ¤5f6 12.a4?!
12.0–0 b6 13.¦fe1 ¥b7 14.¤e5 ¤xe5
  
15.¦xe5 would lead to a normal game. When    
in doubt there is always the simple advice: try    
to play in the centre.
   
12...b6 13.0–0 ¥b7    

 Chapter 3 – Key Ideas and Positions 185

21...f5! To find and analyse the critical or key positions

Blocking the f-file. Black has an is not simply routine work, but is more like a
overwhelming advantage. Compared with the creative task that helps a player to advance.
Taimanov – Karpov game above, Black has not The idea that the game starts after the
even sacrificed a pawn. opening is a big misconception. In that view
we are at first in some kind of memory contest,
22.¥e1?! ¤xe5! 23.fxe5 £c4 followed by a practical exercise, and then in
Black is completely dominating on the light the final phase of the game we again test our
squares. knowledge in theoretical endgame positions.
 So-called Fischer Random chess (or
Chess960) ostensibly cuts off the first part
   of the game, but in my opinion this is
    unnecessary: we should not cut off the logical
process of the game. Nobody argues that we
    should remove some part of the game in other
  sports, such as the serve in tennis.
   Lastly, we start our lives when we are born
and live till we die. If in chess we want to see a
    model of life, we need to keep it as it is.
    Working on openings, despite there always
    being some computer-generated final answer,
 is essential if you are to master chess. The
24.¥h4?! £xb5 25.axb5 ¦d7 26.¦c1 ¦c4 concepts, the critical lines and the positions
27.¥e1 ¤c7 28.h3?! that you have in your mind make you a player.
28.¦b1 would not be enough to save the Chess is not a game where we just compare our
game: 28...¦d5 29.¦f2 ¤xb5 30.¦b3 ¤c7 knowledge, but we also compare our creativity
31.¦a2 ¦b5 32.¦xb5 ¤xb5 Black is winning. in a competitive environment.

28...¦dxd4!
White will soon lose a second pawn, for
example: 29.cxd4 ¦xc1 30.¢f1 ¤xb5–+
0–1
Chapter 4
Computer Preparation versus
Human Common Sense
What route should we follow in opening preparation – the computer evaluation or a practical
human decision?
There is always the final truth, which can be measured strictly by computer evaluation. However,
the number it gives is only relevant as a scientific measure of the position, and might be good if
we were preparing to play a match against some advanced aliens who might visit us in the future.

For decades the only measure was simply practice – this was the criterion of truth.
Practical players follow the trend and use modern opening lines; researchers try to find
something of their own. It becomes more and more difficult to find something new; the practical
approach prevails today. In the past you had to work or analyse yourself: you could not trust
practice blindly. There were no computers around to confirm what was right or wrong. The old
saying – trust but verify – applied.

Today one can reduce the amount of work one has to do to find the correct way because there
is the absolute truth of the computer evaluation. This attitude is very common among the new
generation of chess players. I experienced this while working with young players. Once, when
I tried to explain a certain position to a student in the training camp, he just ran to another
room to check the correct computer move. He was right of course: the computer move was
the strongest move. Then I realized that the coach has a very difficult duty to explain to even
advanced and clever students the difference between a computer evaluation and a practical human
decision.

There are many factors that we need to consider when talking about a perfect or correct opening
line in terms of the computer evaluation. When we are annotating a game, we always notice and
point out any mistakes using computer help. All commentators today have this powerful tool and
the differences in their own chess strength do not count for much. Grandmasters do not have the
same magic aura that they had in the past, when only the great masters could reveal the hidden
secrets of the game to chess aficionados.
188 Grandmaster Opening Preparation
It might seem that everybody is capable of
analysing and preparing opening lines using
computer help, but this is not true – and I am
not making the argument here to defend the
profession of coach.
Years ago we did not have computers, yet in
1988 Lev Polugaevsky was convinced that
in correspondence chess he could play at the
same level as the World Champion, Kasparov.
The latter correctly pointed out, however, that
it is not enough to have nearly unlimited time
to analyse the position – one must first also
choose the correct opening or line to play.
There are vast numbers of opening lines.
Which line to choose?
The player is the most important factor,
not the opening line. The level or rating
of the opposition is also a very important
consideration. Some openings are good up to a
certain level. We might have a very emotional
bond with them, but we cannot fall in love
for ever. The King’s Gambit should be in the
young player’s repertoire, but then he needs
to move on and study and play other classical
lines: most importantly the Ruy Lopez.
Opening lines illustrate some kind of order
in chess. The young player must play attacking
chess. The first move 1.e4 must be the main
opening move. When the player advances
and their understanding of positional chess
develops, he or she can start the game with
1.d4 or 1.¤f3. Older players tend to play more
closed positions. Some players try to play both.
Which move is best – 1.e4 or 1.d4 – is not
relevant, but the character of the game after
each move is completely different. This is why
once you move from 1.e4 to 1.d4 it is not so
easy to switch back. Vladimir Kramnik has
done it with some success, but other players
have regretted that they moved from open
lines to closed lines.
Computer lines become more relevant when
young players approach grandmaster strength.
They tend to check the computer evaluation
all the time, and neglect the other lines.
Fortunately, chess is not just about black and
white decisions. At a certain level, a player may
be stronger in attacking positions, but lacking
in defensive skills. The computer’s truth is not
relevant until the player reaches the very top of
the game. Even then, there are some positions
which, despite the computer evaluation, are
easier to play for humans.
In the following game I had a position after the
opening which was evaluated as “...and White
has compensation for the sacrificed piece.” The
computer evaluation is not relevant, because
the compensation is positional, not dynamic
or tactical.
Jaan Ehlvest – Raj Tischbierek
Leningrad 1984
1.e4 c5 2.¤f3 d6 3.d4 cxd4 4.¤xd4 ¤f6
5.¤c3 ¤c6 6.¥c4

 
 
   
    
  
    
 
  

The Sozin Attack, which we discussed in the
previous chapter, is one of the most direct ways
to attack the black king. It has some positional
basis as well: developing the bishop to the
 Chapter 4 – Computer Preparation versus Human Common Sense 189

a2-f7 diagonal, controlling the light squares


and, in some positions after f4-f5, forcing  
Black to move his e-pawn or to take on
f5, after which White gets control over the
 
d5-square. Black must try to trade White’s    
light-squared bishop or the d4-knight. If they   
can work together, a lot of tactical motifs are
in the air.     
   
6...e6 7.¥e3 a6 8.£e2 £c7 9.0–0–0 ¥e7
10.¥b3 ¤a5 11.g4 b5 12.g5 ¤xb3† 13.axb3
  
¤d7     
 
  20.h4!
Who is the author of this plan? I got this
  idea from Yurkov, who showed me this in one
   of our training camps. His student Andrei
Sokolov won a game with it (see below), but
    the very first game (according to the database)
    was played by Florin Gheorghiu.

    20...¦c8 21.f4!

   21.¥f4? was played in the first game to
   reach this position, suggesting that h2-h4 was
just an accidental move made by Gheorghiu
 without any idea behind it: 21...¢f8 22.gxf6
White is not forced to sacrifice a piece here, ¤xf6 23.¦f5 ¦c5 24.¥h6† ¢e8 25.¦xc5 dxc5
but can just play 14.h4. However, practice 26.¥g5 ¦g6 27.f4 ¤g8 28.f5 ¦d6 29.£e4
shows that there is no advantage for White. ¢f8–+ Black went on to win in Gheorghiu –
Hamann, Vrnjacka Banja 1967.
14.¤f5! exf5 15.¤d5 £d8 16.exf5 ¥b7
17.f6 gxf6 18.¦he1 ¥xd5

The only move.  
19.¦xd5 ¦g8
 
In this position, a slow plan was introduced.    
Even today we can check the computer   
evaluation and see that all of the forcing lines
are about equal – sooner or later White needs     
to force a draw. But the computer does not    
indicate the positional plan here. When we
talk about a positional plan or move, we are
  
usually talking about the pawns, the pawn     
structure, etc. 
190 Grandmaster Opening Preparation

White’s plan should just be to play a positional


game a piece down! The compensation lies in    
the lack of coordination in Black’s camp. The  
position is about equal, but it is easier to play    
with White.  
21...¢f8 22.¢b1 ¦g6     
Sokolov’s game continued: 22...¤b8?   
    
      
   
    White has more than enough compensation
for the piece.
  
     23...fxg5 24.hxg5 ¢g8 25.¦d5 ¦e6 26.£d3
    £c7 27.¦e2 £c6 28.¦h2 ¤f8 29.¥d2 £b6
   30.¦g2 b4 31.f5 ¦e5
    
   
23.£h5 (23.£d3! would already give White a
big advantage) 23...¦g7 24.¥d2 ¦c5? 25.¦xc5   
dxc5 26.gxf6 ¥xf6 27.£xc5† ¢g8 (27...¥e7    
28.¦xe7! £xe7 29.£c8† £e8 30.¥b4† ¢g8
31.£xe8#)   
     
      
     
   
       
     
32.g6 hxg6 33.fxg6 £c6 34.gxf7† ¢xf7
    35.¦f2† ¢e8 36.¦d4 ¤e6 37.¦xb4 ¢d7
    38.¥c3 ¦h5 39.¦e2 d5 40.¦g4 £b5 41.£f3
    d4 42.¦ge4 ¦c6 43.¥xd4 £f5
 ½–½
28.£c7! h5 29.£xd8† ¥xd8 30.¦e8† ¢h7
31.¦xd8+– White had regained the piece In the next game the computer evaluation
and easily won the ending in A. Sokolov – might have helped White, but back then
Korzubov, Bukhara 1981. nobody knew the exact situation.

23.¦f5?
The correct way to keep the pressure on was:
23.h5! ¦g8 24.£d3 ¦g7 25.¥d4
 Chapter 4 – Computer Preparation versus Human Common Sense 191

Oleg Romanishin – Pertti Poutiainen 17.£g3 £xg3† 18.¤xg3 ¥xh1 19.¤xh1

dxe5 20.fxe5 ¦fd8 21.¤f2! (21.¦d1? ¦xd1†
Tallinn 1977 22.¢xd1 ¦d8† 23.¢e2 ¦d4 24.¤f2 ¦h4
25.h3 ¦h5) 21...¦d4 22.¢f1! White can still
1.e4 c5 2.¤f3 ¤c6 3.¤c3 d6 4.d4 cxd4 hope for some advantage.
5.¤xd4 ¤f6 6.¥c4 e6 7.¥b3 ¥e7 8.¥e3 17.¦g1 g6
0–0 9.f4 a6 10.£f3 ¤xd4 11.¥xd4 b5 
    
   
    
      
   
      
      
       
  
18.¦d1
    18.¦g3!? might be a try for the advantage.
 18...¥xe4 19.£xe4 £xh4† 20.¢e2÷
This was one of the tabiya positions in the A draw eventually resulted in Short –
Sozin Attack. It also featured in the 1993 Kasparov, London (12) 1993.
London match between Short and Kasparov.
I have played this position with both Black 15...¥b7 16.¤e4 ¥e7?!
and White. The position is so complicated The main problem for White is: 16...¥xe4
that once I misplayed it with White, despite 17.£xe4 d5
being familiar with it as early as 1979, when I 
won a game with Black in the Under-17 World    
Championship in Belfort.   
12.¥xf6   
12.e5 dxe5 13.¥xe5 ¦a7! 14.0–0 ¦d7   
15.¦ad1 £b6† 16.¢h1 ¥b7 was fine for Black    
in Sorensen – Sher, Hastings 1989-90.    
  
12...¥xf6 13.e5 ¥h4† 14.g3 ¦b8
This defensive trick was very well known.
  

15.0–0–0 I have used this position with my students
Short played: a lot. The question is: can you find the
15.gxh4 ¥b7 16.¤e4 dxe5 correct plan for White? It was rather strange
Kasparov had prepared this line. He that I missed the correct plan myself against
probably did not like that after 16...£xh4† Mednis in 1998, but the game was played in
192 Grandmaster Opening Preparation

Las Vegas and either due to the casino


atmosphere or something else, I could not    
remember anything. I knew that White has
a way to put pressure on Black, but could
  
not figure it out and the game ended in a    
quick draw after 18.£f3 ¥e7 19.f5 a5 20.c3    
a4 21.¥c2 b4 22.f6 gxf6 23.¥xh7† ¢xh7
24.£h5† ¢g7 25.£g4† ½–½ Ehlvest –    
Mednis, Las Vegas 1998.    
The correct move is 18.£d3! ¥e7 19.h4!
and now Black has only one move to keep the
  
balance: 19...£b6! Instead 19...£a5 20.f5 b4   
21.£f3 ¢h8 22.f6± gave White the advantage
in Morozevich – Mitenkov, Moscow 1991,

20.exd6
and 19...a5 20.c3 b4 21.¥c2 g6 22.h5 bxc3 It is difficult to understand why Romanishin
23.hxg6 hxg6 24.¦h6± is also good for White. did not play 20.e6. Most likely he thought
 that Black has some drawing chances after
    20...fxe6 21.£xe6† ¢h8 22.¦he1 ¥f6
23.£xd6 £xd6 24.¦xd6 a5, but you cannot
  expect to win immediately in the opening.
   Reaching an endgame with an extra pawn is
good value from the opening.
   
    20...¥xd6 21.¦hf1 a5 22.¦d3 a4 23.¥d5
   ¢h8 24.¦c3
24.¦xf7 ¦xf7 25.¥xf7 £xf7 26.¦xd6 £xa2
   27.£e7 was still equal.
   24...£a7 25.¦c6 f5 26.£d3 £e7
 
17.f5
17.exd6 ¥xd6 18.¦he1? (18.f5! ¥xe4     
19.£xe4 £g5† 20.¢b1 exf5 21.£c6 ¥e5
22.£xa6 f4 23.gxf4 ¥xf4 24.a3 and White
   
has better chances) 18...a5 19.¥xe6? fxe6    
20.£g4 ¥d5 and White did not have sufficient  
compensation in Hmadi – De Firmian, Tunis
1985.    
   
17...exf5
Black could also consider 17...¥xe4 18.£xe4
  
d5.    
18.£xf5 ¥xe4 19.£xe4 £c7

 Chapter 4 – Computer Preparation versus Human Common Sense 193

Black is now better. The opposite-coloured to learn from it as well. The next game shows
bishops are quite venomous, because Black will how an interesting idea was born in a different
effectively have an extra piece when attacking. Najdorf line.

27.¥g2 ¥e5 28.¢b1 g6 29.¦e1 £g7 30.£a3 Alvis Vitolinsh – Viktor Gavrikov
b4 31.£xa4 ¥xb2 32.¦ce6 ¥a1 33.£b3
¥c3 34.¦d1 ¦fd8 35.¦xd8† ¦xd8 36.£d5 USSR 1977
£c7 37.£f3 £d7 38.£d3 £c8 39.¦d6 ¦e8
40.¥f3 ¥g7 41.£b5 h5 42.¥d1 ¦d8 43.¦d3 1.e4 c5 2.¤f3 d6 3.d4 cxd4 4.¤xd4 ¤f6
¦xd3 44.£xd3 £c5 45.¢c1 £g1 46.£d2 5.¤c3 a6 6.¥g5
¢h7 47.£e2 In recent years this move has become popular
again. Years ago there was one principal line
 that was a headache for White.
    
    6...e6 7.f4 £b6!
The Poisoned Pawn was also played by
    Bobby Fischer in his match against Spassky in
   Reykjavik in 1972.

     8.£d2 £xb2 9.¦b1 £a3 10.f5 ¤c6 11.¤xc6

     bxc6 12.fxe6 fxe6 13.e5 dxe5 14.¥xf6 gxf6

  15.¤e4 ¥e7 16.¥e2 h5 17.¦b3 £a4

    
  
47...£a7 48.£c4 £e3† 49.¢b1 £g1     
50.£e2 £d4 51.¢c1 £b2† 52.¢d2 £c3†  
53.¢c1 ¥h6† 54.¢b1 £a3 55.£e7† ¥g7
56.c3 £xc3 57.£e2 £a1† 58.¢c2 £b2†    
0–1   
Everybody looks at the computer evaluation
   
in sharp lines. The Najdorf Variation of the 
Sicilian Defence is probably one of the most
complicated openings, especially the 6.¥g5
   
line. One of the best opening books ever 
written, Grandmaster Preparation by Lev This was one of the critical positions where
Polugaevsky, is about the line 6.¥g5 e6 7.f4 White got stuck. It seems that Black has
b5!?. This line is no longer considered playable everything defended. Back in the 1970s there
and we have computers to thank for this. The were no computers, so how did human players
question is: should we neglect the wonderful find the plan here? The brilliant Latvian master
work done by Polugaevsky? Of course not! who introduced this idea could not explain it
The line might be refuted, but the intelligent in words. He had some mental problems and
chess researcher can and must find something later he committed suicide by jumping off a
194 Grandmaster Opening Preparation
bridge in Riga. Still, his chess idea was that
of a genius, and later, in one of the Soviet
training camps where I was present, Alexander
Beliavsky explained it.
18.¤xf6†! ¥xf6 19.c4!
The evaluation of the position might be 0.00,
which is the common way for computers to
make it crystal clear that the position is a draw.
I was never a fan of playing such positions,
where you need to work a lot to remember
the correct path, and at the end of the day you
simply have a half point.

 
    
 
   
  
   
 
   

19...¥h4†
A good move; the idea is to provoke the
pawn to g3, limiting White’s play on the third
rank.
20.g3 ¥e7 21.0–0 ¦a7 22.¦b8 ¦c7 23.£d3
¥c5† 24.¢h1 ¢e7 25.£g6 ¢d6 26.£f6 ¦e8
27.¥xh5 ¦ce7 28.¦d1† ¥d4 29.¦xd4† exd4
30.£xd4† ¢c7 31.£b6† ¢d7 32.£d4†
½–½
Computers are just calculating machines and
they do not explain what is going on. In the
Soviet Union, chess players had the privilege of
having regular training camps. Most talented
youngsters received some tutoring from the
elite Soviet players and coaches.
Alexander Beliavsky – Peter Szekely
Frunze 1979
1.e4 c5 2.¤f3 d6 3.d4 cxd4 4.¤xd4 ¤f6
5.¤c3 a6 6.¥g5 e6 7.f4 £b6 8.£d2 £xb2
9.¦b1 £a3 10.f5 ¤c6 11.fxe6 fxe6 12.¤xc6
bxc6 13.e5 dxe5 14.¥xf6 gxf6 15.¤e4 ¥e7
16.¥e2 h5 17.¦b3 £a4 18.¤xf6†

 
    
 
   
   
   

   

Four-time Soviet champion Alexander
Beliavsky prepared some very specific material
for his lecture about the sacrifice in the
opening. Everybody knows some very general
ideas about sacrifices, such as how many
tempos you should gain when sacrificing a
pawn. In this specific example the situation is
not trivial at all. White is not winning a tempo
here to speed up his attack. Why then give up
the knight?
Apart from the tempo-winning motif there
are others – one of them is to open routes into
Black’s camp. The knight is sacrificed not for
immediate gain, but for long-time positional
gain. Black’s ugly pawns do not defend the
files and diagonals very well. White is not
even gaining a tempo. It is a purely positional
sacrifice, which most likely only humans can
discover.
Even so, the computer is a big help today,
when you can quickly check if the idea works
 Chapter 4 – Computer Preparation versus Human Common Sense 195

at all. However, the evaluation – even if it was 28.c5†! Again the same principle: open as
something in the region of –0.5 to –0.9 – may many diagonals and files as possible. 28...¢xc5
not be enough for a final verdict. 29.£e7† ¢d5 30.¥e4† ¢c4 31.¦c1† ¥c3
Mastering the discovery of these kinds of 32.¦xc3† ¢xc3 33.£c5† ¢d2 34.¦b2†
position and introducing them into practice is Followed by mate.
the dream of every researcher.
25...¥d4
18...¥xf6 19.c4 ¥h4† 20.g3 ¥e7 21.0–0 25...£a5 was the only move, but again very
¥d7? difficult to find.
This move is too passive and cannot be
recommended. 26.£xh5† ¢e7 27.£h4† ¢d6 28.c5†
Again the same motif.
22.¦b7 ¦d8 23.¥d3
As a result of the sacrifice of the knight for

the f-pawn, there is nothing left to defend the    
kingside, especially the g6-square.   
  
        
      
     
      
      
    
    28...¢xc5 29.£e7† ¢d5 30.¦xd7† ¦xd7
    31.£xd7† ¢c5 32.¦c1† ¢b6 33.¥e4 ¥c3
34.£d3 ¥d4 35.¦xc6† £xc6 36.¥xc6 ¢xc6
 37.£xa6† ¢d7 38.£b7† ¢d6 39.a4 ¦d8
23...¥c5† 24.¢g2 ¦g8 25.£e2 40.a5 ¦d7 41.£b8† ¢d5 42.£b5† ¢d6
25.£h6 was stronger and White is winning 43.¢f3 ¦c7 44.a6 ¢e7 45.h4 ¢f6 46.£d3
after: 25...¢e7 26.£f6† ¢d6 27.¦e1 ¥d4 1–0

    There are openings that have a fixed or nearly
fixed pawn structure out of the opening, and
   the set-up of the pieces and the plans are well
  known. These lines are sometimes considered
    a little passive, but easy to play. The computers
   do not tell you the set-up, just the moves.
    One approach is to build up an aesthetically
pleasant-looking position and then look to
   take action after that; the computers can help
     here.

196 Grandmaster Opening Preparation

In the following game Paul Keres built up
a nice position, but he then failed to take
any concrete action and finally lost from a
promising position.
Paul Keres – Tigran Petrosian
Bled/Zagreb/Belgrade 1959
1.e4 c5 2.¤f3 ¤c6 3.d4 cxd4 4.¤xd4 g6
5.c4 ¥g7 6.¥e3 ¤f6 7.¤c3 ¤g4 8.£xg4
¤xd4 9.£d1 ¤e6 10.£d2 d6 11.¥e2 ¥d7
12.0–0 0–0 13.¦ac1 ¥c6 14.¦fd1 ¤c5 15.f3
a5 16.b3 £b6 17.¤b5!

  
 
16.¤b5± ¥xb5 17.cxb5 ¢h8 18.¢h1 f5
19.exf5 gxf5 20.¦fe1 ¦ad8 21.¦c1 £c7
22.¥f1 b6 23.¥d3?! d5 24.¥g5 ¥f6 25.¥h6
¥e5 26.g3 f4 27.¥xf8 ¦xf8 28.£e2 ¥f6
29.gxf4 £xf4 30.¥b1 d4 31.¦g1 £h4 32.¦g4
£h5 33.¦cg1 e5 34.f4 1–0 Ehlvest – Kveinys,
Liepaja (rapid) 2016.
However, building up the position is one
thing; you need to have a follow-up. This is
where computers are very strong.
17...¦fc8 18.¥f1 £d8
Now we have a typical situation where the
computer evaluation is very good for White,
but some concrete plan of action is a must.
Although Keres made a series of good moves,
he eventually lost the thread.

   19.£f2

    The computer’s top choice is: 19.g3!

    
    
    
       
   
I like this plan and used it myself against   
Kveinys in a similar situation: 1.c4 c5 2.¤f3    
g6 3.d4 cxd4 4.¤xd4 ¤c6 5.e4 ¤f6 6.¤c3   
d6 7.f3 ¥g7 8.¥e3 0–0 9.£d2 ¤xd4 10.¥xd4 
¥d7 11.¥e2 ¥c6 12.0–0 ¤d7 13.¥e3 a5 I like this computer move. It is not a plan
14.b3 ¤c5 15.¦ad1!? £b6 (15...e6 16.¤b5±) of course, because there is no such definition
 in the program’s algorithm. We, however, can
   look for some plan or set-up after this move.
  There is no immediate tactical gain, and it
   actually weakens the pawn formation. It might
look as though the black bishop on c6 will have
     good prospects on the h1-a8 diagonal after this
   move, especially if Black is able to strike with
   ...f5 in the future.
  From White’s side, the space advantage very
   often allows him to make such a weakening

 Chapter 4 – Computer Preparation versus Human Common Sense
move, because although the space advantage
is pleasant, we need to increase our advantage
somehow. This means that we need to create
a pawn weakness or carry out a direct attack
against the kingside. Here Black has very good
defensive barricades on both wings. As practice
has shown, the plan of pushing the pawns on
the queenside is very difficult to accomplish.
White needs to play all over the board and for
this reason the last move fits this plan. Black’s
main idea is to build a blockade on the dark
squares, and to undermine this we need to
have the option of pushing our pawns on the
kingside.
Play may continue: 19...b6 (19...£f8
20.¥h3 ¦d8 21.¤c7! ¦a7 22.¥xc5 is already
winning) 20.¥h3 ¦cb8 21.¤d4 ¥b7 22.¦b1
And now, in the changed situation, the black
pawn is on b6, which favours White. The
black bishop no longer has protection on c6,
and after a2-a3 and b3-b4 the pawn on b6 will
be attacked by the white bishop on e3. This is
why White wants to keep the bishops on the
board in these positions.
19...£e8 20.¤c3 b6 21.¦c2 £f8 22.£d2
¥d7 23.¤d5 ¦ab8 24.¥g5 ¦e8

  
 
   
   
  
  
 
  

25.¦e1
Keres still has the advantage, but he does not
have a concrete plan.
197
This is how you can use a computer engine
during your opening preparation. First you
need to figure out the plan, and then ask the
computer, step by step, how to carry it out
with the most precise move order.
Some positions, such as isolated pawn
positions or some closed positions, are tricky
to analyse, because the computer evaluation
might not show the real situation on the board
and even a seemingly very good evaluation is
not enough to draw a final conclusion about
the position. This does not mean that the
player and his coach should analyse every
position to the very end; there is just not
enough time for this. They just need to check
with the computer engine if their assumptions
about the plan or set-up are correct or not.
25...¦b7 26.£f2 ¥c6 27.£h4
Keres has just been playing around with
his pieces, and Black begins to take over the
initiative.
27...f6 28.¥e3 e6 29.¤c3 ¦d7 30.¥d4

  
  
 
    
  
  
 
   

30...f5!
Suddenly Black has the better game.
31.exf5 gxf5 32.¦d2 ¥xd4† 33.¦xd4 ¦g7
34.¢h1³ ¦g6 35.¦d2 ¦d8 36.¦ed1 ¦d7
37.£f2 £d8
198 Grandmaster Opening Preparation

 
       
     
     
       
      
      
     
   
 
I consider the computer evaluation of 42.a3
equality here to be wrong. The position is not A very human and good move.
equal from a human point of view. It is very
difficult for White to play. The computer move 42.¤b5 is also strong, but
difficult to make, because who wants to trade
38.£e3?! the dominating knight?
The best move, 38.¦b1!, is not easy to find,
given that White had already abandoned this 42...£a8 43.¢g1?
plan earlier in the game. Keres is playing this game in too timid a
fashion.
38...e5! 39.f4
39.¥d3! would make it difficult for Black 43...h5 44.¦b1 h4 45.¦bb2 ¦g4 46.¦f2
to break through. It is true that computers are £d8 47.b4
very strong at finding this kind of move, but
you cannot build your openings on the hope

of finding computer moves during the actual    
game.    
39...e4?     
In time trouble Petrosian makes a serious    
positional mistake.
 
After 39...£h4! 40.fxe5 ¦h6 41.£g1 dxe5     
Black would have a big advantage.
   
40.¤e2 ¦dg7 41.¤d4 ¥d7    

47...¦g3 48.hxg3 hxg3 49.¦fd2 £h4
50.¥e2?
 Chapter 4 – Computer Preparation versus Human Common Sense
50.¤f3! exf3 51.gxf3 ¦h7 52.¥g2 ¦e7
53.£d4 ¦e1† 54.¥f1 ¤e6 55.£d5 ¦xf1†
56.¢xf1 £h1† 57.¢e2 g2 58.¦d1 g1=£
59.¦xg1† £xg1 60.¦d2 £h2† 61.¢f1 £xf4
62.¢f2 and White can still hold on.
50...¦h7 51.¢f1?
White had to try 51.¥h5! ¦xh5 52.¢f1.

   
  
    
   
  
    
  
   

51...£xf4†!
0–1
There are situations where the computer
evaluation is wrong. In closed positions,
computer power, at least my computer power,
is not enough to give a realistic evaluation.
The practical point is that a human may see
the correct evaluation, but in today’s world
the player will most likely abandon these
positions too quickly after seeing the computer
evaluation. This is where the human factor is
important. You should not teach or coach
only using a computer. There is still a need
for a good coach or an adviser who can tell
the difference between human and computer
evaluation.
199
Harika Dronavalli – David Anton Guijarro
Sharjah 2017
1.e4 e5 2.¤f3 ¤c6 3.¥b5 a6 4.¥a4 ¤f6
5.0–0 ¥e7 6.¦e1 b5 7.¥b3 d6 8.c3 0–0 9.d4

 
  
  
   
   
  
  
  

This was my pet line when I was a young
master. The Ruy Lopez is a complex opening,
and learning the main lines takes years.
However, young players should follow the
principled lines, because they can learn so
much about chess by playing the classical
established lines; thus 1.¤f3 and 2.g3 is a
no-no for a young master.
I started to play the main lines of the Ruy
Lopez later in my career, but I always followed
the development of the 9.d4 line.
9...¥g4 10.¥e3 exd4 11.cxd4 d5
This counterattack was very well known back
in the 1970s when I started playing the line,
and some theoretical overviews were published
in Soviet chess journals.
12.e5 ¤e4 13.h3
13.¤bd2 was the idea that was published in
a Soviet journal and I was eager to analyse it:
13...¤xd2 14.£xd2 ¥xf3 15.gxf3 ¥b4 16.£c2
200 Grandmaster Opening Preparation

 
     
     
     
     
         
    
     
         
 
Black cannot afford to take the rook: 16.£c3!!
16...¥xe1? (16...¤a5 was, however, considered This startling move, found by a computer,
a safe option) 17.£xc6 White has a big changes the evaluation. The idea works only
ad