Sunteți pe pagina 1din 223


Title page
Introduction: The Cadillac of Openings
The Development of the Najdorf Sicilian

Chapter 1: Va Banque: 6.Bg5

Chapter 2: The Classicist’s Preference: 6.Be2
Chapter 3: Add Some English: 6.Be3
Chapter 4: In Morphy’s Style: 6.Bc4
Chapter 5: White to Play and Win: 6.h3
Chapter 6: Systematic: 6.g3
Chapter 7: Healthy Aggression: 6.f4
Chapter 8: Action-Reaction: 6.a4
Chapter 9: Odds and Ends

Index of Complete Games

Swipe left for next chapter

Bryan Smith

The Najdorf in Black and White



© 2017 Bryan Smith

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

Publisher: Mongoose Press

1005 Boylston Street, Suite 324
Newton Highlands, MA 02461

ISBN: 978-1-936277-86-5

Distributed to the trade by National Book Network, 800-462-6420
For all other sales inquiries please contact the publisher.

Layout: Stanislav Makarov

Editor: Jorge Amador
Cover Design: Alexander Krivenda
Printed in the United States of America

First English edition


! a good move
? a weak move
!! an excellent move
?? a blunder
!? an interesing move
?! a dubious move
 only move
= equality
 unclear position
 with compensation for the sacrificed material
 White stands slightly better
 Black stands slightly better
 White has a serious advantage
 Black has a serious advantage
+- White has a decisive advantage
-+ Black has a decisive advantage
 with an attack
 with initiative
 with counterplay
 with the idea of
 better is
 worse is
N novelty
+ check
# mate
Introduction: The Cadillac of
With this book, I present a collection of games played in
the Najdorf Sicilian. The purpose of this book is not to
be exhaustive - that would require at least ten times the
content, and even then it would not encompass a
fraction of the analysis and relevant games played in the
Najdorf. This book also does not suggest a repertoire
for either White or Black - although players can glean
some ideas, since I have generally picked games played
in the lines that I favor. I think it is dishonest for a writer
to try to portray an opening in only a positive light:
ultimately, even the most objective writers of repertoire
books have to massage the facts and minimize the
problems of an opening - and every opening has them.
The purpose of this book, rather, is to show how to
play the Najdorf, with White or Black, through
archetypal games. I believe that by studying the games
in this book, one can develop a solid general sense of
the different types of games resulting from the Najdorf
as played in the twenty-first century. It is my hope that
readers will also gain some degree of enjoyment or
entertainment from the games, which have been
selected not only on their instructional merits, but also
for their aesthetic value.
The games are grouped according to variation, and while
it is impossible for me to cover every single option for
Black, I have covered every reasonable sixth move by
White - the 1nain branch point where the first player
determines which way the game will go. The games
selected have featured the defining Najdorf move ...e7-e5,
whenever applicable. Thus transpositions to the
Scheveningen or Dragon are generally avoided here. The
moves 6. Bg5 and 6.Bf4 generally do not allow Black's ...
e7- e5 advance, but this book deals with the unique and
identifiably "Najdorf" positions resulting from those
moves, as well.
Having a lifelong opening that one knows inside and out
like one's own house is a major advantage to a chessplayer.
It means that the player can always rely on reaching
positions that he understands in general terms and knows
something about. Perhaps more importantly, though, it
gives confidence. There is no worrying about what opening
to play, no wasting energy before a game trying to decide,
and no regrets: just the ever-deeper exploration of the
opening's secrets. A sufficiently rich opening will provide
immunity against the winds of theory - if one variation is
refuted, another can be found, so long as the opening is
built on proper principles.
I believe the Najdorf can be such an opening. Some
may imagine that it is a theoretical labyrinth, suitable
only for those with an incredible memory and a
willingness to play twenty or more moves of known
theory before beginning the game. It is true that there are
certain lines in the Najdorf where this is the norm - for
instance, the Poisoned Pawn Variation (6.�g5 e6 7.f4
�b6). However, the reader will see in this book that these
variations can be sidestepped, and that it is indeed
possible to play the Najdorf "by the light of nature," with
experience providing a guide. Most of the games I have
chosen feature ways of avoiding these quagmires.
Despite its sharpness, the Najdorf is an opening built on
solid positional principles. It is basically a positional
Without any more introductions, let's begin our
journey into the secrets of the Najdorf.

Bryan Smith
Philadelphia, November 2017
The Development of the Najdorf Sicilian

The Najdorf can trace its origins to the nineteenth-century German master Louis Paulsen. Paulsen was an innovator of
defense. In an era when 1.e4 e5 was the dominant opening and direct attacking play was the main method of winning,
Paulsen understood the concept of asymmetrical play and counterattack. His openings and positional play were often a
full century ahead of their time.
Paulsen invented much of the Sicilian Defense, including the Dragon and the strategic basis of the Paulsen Variation.
Most importantly for our subject, he was the originator of the ...e7-e5 advance in the Classical Sicilian:

This advance, often credited to Isaac Boleslavsky, was first played by Paulsen a number of times in the 1880s, first in
the Vienna Congress of 1882.
Here is his game against Isidor Gunsberg, which looks quite modern:

Isidor Gunsberg – Louis Paulsen

Frankfurt 1887

1.e4 c5 2.Nf3 Nc6 3.Nc3 Nf6 4.d4 cxd4 5.Nxd4 d6 6.Be2 e5 7.Nb3 Be7 8.Be3 0-0 9.g4 Be6 10.g5 Ne8 11.Rg1
Nc7 12.Nd5 Nb8 13.Qd2 Nxd5 14.exd5 Bf5 15.Bd3 Qd7 16.Nc1 Rc8 17.Nb3 a5
18.c3 a4 19.Nc1 a3 20.b3 Na6 21.Bxf5 Qxf5 22.Ne2 Nc5 23.Ng3 Qg6 24.h4 Ne4 25.h5 Nxd2 26.Kxd2 Qxg5
27.Bxg5 Bxg5+ 28.Kc2 h6 29.Ne4 Be7 30.Rg2 Kh7 31.Rag1 Rg8 32.b4 g6 33.c4 b6 34.Kb3 gxh5 35.Rxg8 Rxg8
36.Rxg8 Kxg8 37.Ka4 h4 38.Kb5 f5 39.Nd2 h3 40.Nf1 Bh4 41.Kxb6 Bxf2+ 42.Kc6 e4 43.c5 e3 44.Nh2 e2 45.Nf3
e1=Q 46.Nxe1 Bxe1 47.b5 dxc5 48.b6 h2 0-1

Nearly fifty years passed after this game, during which the ...e7-e5 advance was all but forgotten. The move does,
after all, have some positional liabilities. The d5 square is weakened, as is the pawn on d6. Black’s dark-squared bishop
might turn out bad, particularly if there is an exchange of light-squared bishops.

In the meantime, the moves we now call the Najdorf had been introduced, but without all the ideas of that opening –
in particular, the principal advance ...e7-e5. As early as 1926, what we now know as the Najdorf was used by another
opening innovator, Savielly Tartakower. However, the opening was reached after 2...a6, and later essentially became a

Frederick Yates – Savielly Tartakower

Budapest 1926

1.e4 c5 2.Nf3 a6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 d6 6.Bd3 Nc6 7.Nde2 g6 8.0-0 Bg7 9.Qe1 0-0
10.f4 Nb4 11.Kh1 Kh8 12.Be3 Ng4 13.Bg1 f5 14.Rd1 Nxd3 15.cxd3 e5 16.h3 Nf6 17.fxe5 dxe5 18.d4 Nxe4
19.dxe5 Qe8 20.Nxe4 fxe4 21.Rxf8+ Qxf8 22.Nc3 Bf5 23.Bd4 Re8 24.Nxe4 Bxe5 25.Bxe5+ Rxe5 26.Qc3 Qe7
27.Rd5 Bxe4 28.Rxe5 1-0

There were a few other uses of what we call the Najdorf in the late 1920s and early 1930s. One in particular stands
out. A forgotten 1928 game from a tournament in the U.S. might be the first use of the “real” Najdorf:

Lewis Isaacs – Abraham Kupchik

Bradley Beach 1928

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Be2 b5 7.Bf3 e5 8.Nb3 Bb7 9.0-0 Nbd7 10.Bg5 Be7 11.Re1 0-
Undoubtedly recognizable as a Najdorf position, nearly a decade before Miguel Najdorf played the opening.

12.Rc1 Nb6 13.Na5 Rb8 14.Nxb7 Rxb7 15.b3 Rc7 16.Qd3 Nbd7 17.Be3 Nc5 18.Qd1 Qa8 19.Bg5 Ncd7 20.Nb1
h6 21.Bd2 Rfc8 22.Ba5 Rc6 23.g3 Nc5 24.Nc3 Bd8 25.Bxd8 Rxd8 26.Nd5 Nxd5 27.Qxd5 Qc8 28.Red1 Ne6 29.Bg4
Rc5 30.Qd2 Rc3 31.Re1 Qc5 32.Re3 Rxe3 33.fxe3 Ng5 34.Qd3 d5 35.exd5 Rxd5 36.Qe2 Qc3 37.h4 Rd2 38.Qe1 Ne4
39.Bf5 Nf2 40.Bd3 Nxd3 41.cxd3 Qxd3 42.Rc8+ Kh7 43.Rc1 f5 44.a4 b4 45.g4 Re2 0-1

This one game aside, the early use of the Najdorf move order was combined not with ...e7-e5, but rather with a
transposition to the Dragon with an early ...g7-g6 or to a Scheveningen with ...e7-e6. Vladimir Alatortsev, for example,
used the Najdorf this way a number of times in Soviet tournaments in the late 1930s.

During World War II, Boleslavsky began experimenting with ...e7-e5 in the Classical Sicilian, just as Paulsen had
done half a century earlier. In one of the first post-war tournaments, Groningen 1946, he showed his idea to the world,
winning a famous game against Gösta Stoltz:

Gösta Stoltz – Isaac Boleslavsky

Groningen 1946

1.e4 c5 2.Nf3 Nc6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 d6 6.Be2 e5

Boleslavsky and some other players had begun to recognize that this advance had merits that compensated for its
downsides. Not only is the white knight pushed from the center, but Black gains space, a tempo for development, and a
solid central bastion, providing a higher level of security against direct attacks than do the Scheveningen or Dragon

7.Nf3 h6!

Just as Paulsen used to do, Boleslavsky prevents Bg5xf6, which would have allowed White to gain control over d5.

8.Bc4 Be7 9.Qe2 0-0 10.h3 Be6 11.0-0 Rc8 12.Bb3 Na5 13.Rd1 Qc7

This move seriously weakens White’s king position, without any benefit.

14...Nxb3 15.axb3 a6 16.Kh1 b5! 17.b4

If 17.Rxa6, then 17...b4 18.Nd5 Bxd5 19.exd5 Qxc2, and all of White’s pawns are weak.

17...Qc4 18.Qxc4 Rxc4 19.Re1 Bc8 20.g5 hxg5 21.Bxg5 Bb7

Far from the backward d6-pawn’s being weak (as a classicist would expect), it is White’s e4-pawn which proves to be
the real weakness, along with White’s queenside pawns in general.
22.Kh2 Rxb4 23.b3 Rc8 24.Re3 Rd4!

Black manages to extract the rook from the b4 square using this tactic.

25.Rg1 Kf8 26.Bxf6 Bxf6 27.Rg4 b4 28.Na4 Rxc2!

In return for the exchange, Black wins several pawns and gains a passed pawn. Combined with the bishop pair,
Black’s advantage is decisive.

29.Nxd4 exd4 30.Re1 Rxf2+ 31.Rg2 Rf3 32.Rc2 d3 33.Rc7 Be5+ 34.Kg1 d2 35.Rd1 Bd4+ 36.Kh2 Rf2+ 37.Kg3
Bxe4 38.Rc4 Rf3+ 39.Kh2 Be5+ 40.Kg1 d5 41.Rc8+ Ke7 42.Rxd2 Bf4 43.Rb2 Be3+ 44.Kh2 Rf1 45.Kg3 Rg1+
46.Kh2 Rg6!

This retreat seals White’s fate: mate soon follows beginning with ...Bg1 or ...Bf4.


By this point, Karel Opočenský had already begun to experiment with the Najdorf move order, combined with the
...e7-e5 advance:

Vlastimil Stulík – Karel Opočenský

Prague 1945

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Be2 e5

This is the first known use of the system combining the move order with 5...a6 and the immediate advance ...e7-e5.

7.Nb3 Be7 8.Be3 Be6 9.Nd5 Nxd5 10.exd5 Bf5 11.a4 Nd7 12.a5 Bg5 13.Bg4 Bxe3 14.Bxf5 Ba7 15.Qg4 g6
16.Bxd7+ Qxd7 17.Qxd7+ Kxd7 18.c4 Rhc8 19.Nd2 f5 20.b3 e4 21.f3 e3 22.Nb1 Rab8 23.Na3 b6 24.axb6 Bxb6
25.Nc2 Bc5 26.Rb1 Rb7 27.Ke2 a5 28.Kd3 Rcb8 29.b4 Bxb4 30.Nxe3 Bc5 31.Rxb7+ Rxb7 32.Nc2 Rb3+ 33.Kd2 a4
34.Ra1 a3 35.h3 Ke7 36.Ne1 Bb4+ 37.Kc2 Rb2+ 38.Kd3 Kf6 39.f4 Bxe1 40.Rxe1 a2 41.Ra1 Rxg2 42.Kc3 g5
43.fxg5+ Kxg5 0-1

It has been said that Opočenský is the real author of the Najdorf, and that he showed the moves to Miguel Najdorf
sometime during the late 1930s. But clearly the ideas had been kicking around for some time.

Najdorf himself first used the Najdorf variation in the 1939 Olympiad in Buenos Aires:

Christian Poulsen – Miguel Najdorf

Buenos Aires Olympiad 1939

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Bg5

An early use of what later became the sharpest and most direct counter to the Najdorf. This move renders 6...e5
inferior, since White is pinning the knight which controls d5.

6...e6 7.Qf3 Nbd7 8.0-0-0 Qc7 9.Be2 Be7 10.Rhe1 0-0 11.Qg3 b5 12.Bh6 Ne8 13.Bg5 Bxg5+ 14.Qxg5 b4 15.Na4
Bb7 16.Bd3 Nc5 17.Nxc5 dxc5 18.Nf3 c4 19.Bf1 c3 20.Qe5 cxb2+ 21.Kxb2 Qb6 22.Re3 Nf6 23.Rd6 Qc7 24.Rd4
Qc6 25.Rb3 a5 26.Nd2 Nd7 27.Qb5 Ne5 28.Re3 Rfd8 29.Rxd8+ Rxd8 30.Qxc6 Bxc6 31.Bd3 a4 32.a3 Nxd3+
33.Rxd3 Rxd3 34.cxd3 b3 35.Kc3 Kf8 36.Nc4 Ke7 37.Nb2 f5 38.Kd4 fxe4 39.dxe4 Kf6 40.Ke3 Bb5 41.g3 Kg5
42.f3 e5 43.Kf2 Kf6 44.h4 Ke6 45.Ke3 Kd6 46.Kd2 Kc5 47.Kc3 h6 48.f4 exf4 49.gxf4 Kd6 50.Kd4 Ke6 51.Ke3
Kf6 52.h5 Be8 53.Kf2 Ke6 54.Ke3 Kd6 55.Kd4 Ke6 56.Ke3 Bc6 57.Kd4 Kf6 58.Ke3 Bb5 59.Kf2 Be8 60.Ke3 Ke6
61.Kd3 Kd6 62.Kd4 Kc7 63.Kc5 Kb7 64.e5 Kc7 65.f5 Bxh5 66.Nxa4 g5 67.fxg6 Bxg6 68.Kd4 Kc6 69.Nb2 h5
70.a4 h4 71.Ke3 Bf5 72.Kf4 h3 73.Kg3 Kc5 0-1

After the war, Najdorf was one of a number of top players who started to use this opening. David Bronstein also
played a number of games with the Najdorf, while Opočenský continued to use “his” variation, and even Kupchik, who
as we saw might have been the first player to use the Najdorf move order together with the ...e7-e5 advance, continued
to use the line, which was now known to the world. Alexander Kotov also often used the Najdorf, although he always
played it in Scheveningen style with ...e7-e6.
As the top player who tended to play the opening in the “new” style (with the ...e7-e5 advance), Najdorf’s name
slowly came to be associated with the opening. But even as late as 1961, in the book covering the Bled tournament of
that year, the opening was referred to as the “Modern Paulsen Variation.”
By the mid-1950s, the Najdorf had gone from a rare and obscure opening to one seen frequently in tournaments all
over the world. Chess had changed enormously after World War II, in particular due to the experiments of the Soviet
players and their new dynamic way of playing. Naturally, the classical 6.Be2 started to become less automatic. With the
new understanding of the power of Black’s blocking move 6...e5 came moves designed to inhibit this – in particular,
The theory on the sharp variations with 6.Bg5 quickly began to build up. The concept of deep opening preparation
already existed, and these lines of the Najdorf became a battleground for it. For example, in the 1955 Gothenburg
Interzonal, a famous episode happened where three players from Argentina – Oscar Panno, Herman Pilnik, and Najdorf
himself – prepared a special line against 6.Bg5:

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Bg5 e6 7.f4 Be7 8.Qf3 h6 9.Bh4 g5!? 10.fxg5 Nfd7

On the same day, each of them reached the same position, facing the Soviet players Yefim Geller, Boris Spassky, and
Paul Keres. The Soviets all sacrificed a knight and then made a surprising bishop clearance:

11.Nxe6! fxe6 12.Qh5+ Kf8 13.Bb5!

This move clears f1 for White’s rook while preventing the black king from escaping via e8 in some situations. The
Soviets won all three games, and the variation was considered refuted. Later, however, Bobby Fischer introduced the
move 13...Rh7!, after which Black is surviving, but just barely.
The advance ...g7-g5 may have failed in this specific move order, but the concept survived and became a key idea in a
number of variations of the Najdorf. This thrust allows Black to conquer the e5 square, which may prove to be more
relevant than the exposure of the black king and some weaknesses on the kingside.
By the 1960s, the Najdorf had become a mainstream opening, used by many top players. Fischer, in particular, made
it the foundation of his opening repertoire – he used it in nearly every game against 1.e4. He also popularized the move
6.Bc4 against the Najdorf.
A major new development that came later was the rise of the English Attack featuring 6.Be3. Based on Rauzer’s ideas
for combating the Dragon, players began to see the strength of White’s simple setup: Be3, f2-f3, Qd2, 0-0-0, and a
kingside pawn storm. In the late 1990s you could see many games featuring the following:

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.f3 e6 7.Be3 b5 8.g4 Nfd7!? 9.Qd2 Nb6 10.0-0-0 N8d7 11.Qf2
Bb7 12.Bd3 Rc8:
More recent developments in the Najdorf include the rise of the 6.h3 variation, which was previously a very rare
sideline but has now become fairly common, and the acceptance of the idea of preventing White’s g2-g4 advance in a
number of lines by playing the seemingly weakening ...h7-h5. For example:



Another new trend has been the revival of the move 6...Nbd7 against 6.Bg5. Avoiding the massive theoretical
labyrinth that occurs in most variations after 6...e6, Black sets out on his own. Not only is this line less explored, but
also the flexibility of Black’s setup (he can play a later ...e7-e5; adopt a Dragon-style development with ...g7-g6; essay
an early ...b7-b5; or revert to the normal structure with ...e7-e6) means that it will likely remain rich for a long time.
Despite the opening’s great popularity and constant use at the top level for many decades, the Najdorf remains
mysterious and has its unexplored areas, with new ideas waiting to be born. Its attraction for the chess professional
today is easy to understand, since it is an opening where it is possible to play for a win with Black, while it is also
unquestionably sound. Although positionally and tactically very sharp, the Najdorf player still controls his own fate.
Chapter 1
Va Banque: 6.Bg5

Undoubtedly, White’s sharpest and most theoretical answer to the Najdorf is the rapid development with 6.Bg5:

If 5...a6 delays Black’s development by a tempo, White continues in the most aggressive fashion, seeking to take the
black position by storm. Black’s key Najdorf advance 6...e5 is also essentially prevented, since it would be met by
7.Bxf6 Qxf6 8.Nd5 Qd8 9.Nf5, with a large advantage for White.
This line saw great development in the early days of the Najdorf, and theory grew exponentially. White seemed to be
walking a thin road, and Black – a tightrope.
Besides 6...Nc6, which transposes to the Richter-Rauzer, Black has two options: the usual 6...e6 and the slightly
“crooked” 6...Nbd7. The latter delays Black’s kingside development, but keeps various options open – for example,
...e7-e5 might still be played, depending on the circumstances, while developing the bishop to g7 is still possible.
After 6...e6 7.f4, a large number of lines are possible. The main line since the 1950s has been 7...Be7 8.Qf3 Qc7 9.0-
0-0 Nbd7. Although it might be Black’s most “solid” choice, it has not really been the most popular for many years.
The move 7...b5!? was Lev Polugaevsky’s brainchild, and is therefore justifiably known as the Polugaevsky
Variation. This line leads to wild complications typical of the 6.Bg5 line, with Black enjoying major positional
advantages but having an unsafe king.
7...Qb6 is the iconic Poisoned Pawn Variation, where Black grabs the b-pawn and destroys White’s queenside at the
cost of several tempi. First explored in total innocence by David Bronstein and Bobby Fischer as Black and Mikhail Tal
and Paul Keres with White, this line is now a massive theoretical morass of wild – often computer-generated –
variations and forced draws.
In this book, I have chosen to focus on the interesting and somewhat less-explored (and less forcing) 7...Qc7. With
this move, Black prevents e4-e5 and plans to continue with the rapid ...b7-b5-b4 advance before completing his
development. Driving White’s knight from c3, Black can assume the initiative in the center and on the queenside.
Game 1 is a masterpiece which shows a beautiful queenside attack. In the notes we discuss two of White’s three
attempts to take the black position by storm: 8.Qf3 b5 9.f5!?, and 8.Qf3 b5 9.0-0-0 b4 10.Nd5.
Game 2 features a defensive “un-brilliancy,” covering White’s third attempt to refute 7...Qc7: 8.Qf3 b5 9.0-0-0 b4
10.e5!?. For most of this game, White is the one who does the attacking, yet the game is a miniature win for Black.
Fierce battles of attack and defense are why we love the Najdorf, and in a deeper sense the defense is as aggressive as
the attack.
Game 3 deals with the more sedate 8.Bxf6, which is a critical response to the 7...Qc7 system. White exchanges a
bishop for a knight, doubling the black pawns and leading to rich positions similar to the Richter-Rauzer. In this game,
White carries out a thematic Nc3-d5 sacrifice. While there is no immediate attack on the king, White is able to reduce
Black to almost complete immobility.
Game 4 covers the move 6...Nbd7, a sharp move used frequently in the 1950s and 1960s which has only recently
become very popular again. The play there is less well explored than after 6...e6 and, in some ways, even more unusual.
The game features a sharp attack by White that comes seemingly out of nowhere, after Black had apparently achieved
his strategic goals – reminding us that, although Najdorf players are often called upon to play with their king in the
center, this is not always an easy thing to do.

Game 1
Thomas Luther – Leonid Yudasin
Budapest 1989

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Bg5 e6 7.f4 Qc7

This little move is one of the sharpest answers to 6.Bg5. Black prevents e4-e5, planning the quick advance of the b-
pawn without getting involved in the theoretical quagmire of the Polugaevsky Variation (7...b5).


A critical move is 8.Bxf6, leading to positions similar to the Rauzer. The doubled f-pawns are compensated for by
Black’s two bishops and mass of central pawns. This will be seen in the game Ivanchuk – Vachier-Lagrave.

8...b5 9.0-0-0
The downside of playing this line for Black is that there are three va banque attempts for White. Although they all
look scary, theoretically Black stands well in each, and the results bear that out. Nevertheless, the resulting positions are
very dangerous and it would be hard to remember complex theory of lines that you rarely meet. You must trust in your
Najdorf spirit to find the right moves over the board, and know that your task is no more difficult than White’s.
There is 9.0-0-0 b4 10.e5!?, 9.0-0-0 b4 10.Nd5!?, and, on this move, 9.f5.
9.f5 b4!. (Black should not hesitate to play the principled move. In some games, Black has taken the “easy” route and
played 9...Nc6, but this is not as good.) Now White replies 10.Ncb5, sacrificing a piece for the initiative. In these lines,
we see a common scenario in the Najdorf – White is sacrificing material and attacking. Black has to use exceptional
judgment to know which attack is deadly and which is a paper tiger, when material can be returned to take over the
initiative, and when to simply take everything and consolidate. In this position:

Position after 10.Ncb5 (analysis)

Black captures 10...axb5, and White has:

a) 11.Bxb5+ Bd7 12.fxe6 Bxb5 13.Nxb5 Qc5 (Again, Black is up a piece for some pawns and the white pieces are
hanging in air – yet hovering near the black king. Don’t forget, however, that the white king is itself not so well
covered.) 14.Bxf6:
Position after 14.Bxf6 (analysis)

14...fxe6! (this is a very important Zwischenzug, and much better than 14...Qxb5 15.Bxg7 Bxg7 16.Qxf7+ Kd8
17.Qxg7 as in Hector – Rashkovsky, Espergærde 1992) 15.Nd4 (now both of White’s pieces are hanging) 15...gxf6
16.Nxe6 Qc4 17.Qxf6 Nd7 and now the game A.Smith – Hillarp Persson, Stockholm 2007, continued 18.Nc7+ Qxc7
19.Qxh8 Qc5 20.Qxh7 b3! (A typical blow in this variation. The white king will not find safety. Nevertheless,
20...Qe3+ 21.Kf1 0-0-0 is also strong. White has a nominal material advantage but his rooks will not see activity, while
Black’s threats are very immediate.) 21.cxb3 Qe3+ when Black played with the draw in hand, while White still had to
be very accurate (0-1, 32).
b) Also possible is 11.fxe6, when Black needs to complete his development immediately by 11...Be7, and now 12.e5
(or 12.Nf5 0-0! 13.e5 Bb7 14.Qg3 dxe5 15.Bxf6 Bxf6 16.e7, when Black need not allow the draw by 16...Re8 17.Nh6+
Kh8 18.Nxf7+ as in Vymazal – Červený, Czech Republic 2010, but can instead play on with 16...Nc6!? 17.exf8=Q+
Kxf8, with great compensation for the exchange) 12...dxe5 13.Bxf6 gxf6 14.Bxb5+ Kf8 15.Nf5 Bxe6 16.Nxe7:
Position after 16.Nxe7 (analysis)

16...Ra5! is a typical activation of the rook along the fifth rank. After 17.Qxf6 Rxb5, Black is bound to get two minor
pieces against a rook and a pawn, where it will not be easy to activate the white rooks. The passed e-pawn will be a
great force for Black.

9...b4 10.Nce2

White allows Black to carry out his plans. White can also try 10.e5, which is obviously a very critical move. White
tries to take Black’s position by storm. This will be covered in the game Psakhis–Tukmakov (page 20).
10.Nd5 is the third of White’s attempts at blowing Black off the board. White’s standard Sicilian sacrifice does not
necessarily lead to a fierce attack, but there is actually a positional idea involved – to render the f8-bishop useless.
Nevertheless, with accurate play Black stands well. The game Li Chao – Saravanan, Fujairah City (UAE) 2012,
continued 10...exd5 11.Bxf6 gxf6 12.exd5 Ra7!:
Position after 12...Ra7 (analysis)

(this seventh-rank defense is a critical resource) 13.Bd3 Qc5 14.Bf5 Kd8! (the king will find relative safety on the
queenside) 15.Kb1 Rc7 16.Rd2 Bxf5 17.Nxf5 Nd7 18.Re1 Kc8 (White has succeeded in “stalemating” the bishop on
f8. Nevertheless, it is an extra piece and it will not be easy for White to maintain his hold. In the game, Black scored a
major upset.) 19.c3 Nb6 20.cxb4 Qxb4 21.Re8+ Kb7 22.Rd4 Qc5 23.Qd1 Ka7 24.Re3 Qb5 25.a4 Qd7 26.g4 h5 27.h3
hxg4 28.hxg4 Qc8 29.Rb4 Rc4 30.Reb3 Rxb4 31.Rxb4 Qc5 32.Rd4 Rh2 33.Qe1 Qc2+ 0-1.

10...Nbd7 11.g4

If 11.Ng3 Bb7 12.Bd3, then logical is 12...h6, seeking to resolve the position of the g5-bishop. After 13.Bxf6 Nxf6
14.Nh5 Nxh5 15.Qxh5 g6 16.Qh3 0-0-0, Black had a good game in Trujillo Villegas – Domínguez Pérez, Santo
Domingo 2007.

11...Bb7 12.Ng3

Yudasin makes this typical breakthrough in the center immediately. It is also possible to play 12...Rc8, e.g. 13.Bd3
Qb6, chasing the knight from the center, followed by ...a5-a4 with a good game for Black.


After 13.e5 Ne4 14.Nxe4 dxe4 15.Qe2 h6 16.Bh4 g5 the game gets very messy. 17.fxg5 hxg5 18.Bxg5 Qxe5 19.h4
saw Black’s central pawns pitted against White’s h-pawn in Outerelo Ucha – Terán Álvarez, Spain 1993.
Another possibility is 13.Bxf6 Nxf6 (13...dxe4!?) 14.e5 Nd7, when Black has no problems in this French-like
structure. Black threatens the thematic break ...g7-g5.


Black drives the knight from the center.

14.Nb3 dxe4 15.Nxe4 Be7 16.Rhe1 Rc8

Black is threatening 17...Nxe4 18.Bxe4 Qxc2+. With pressure against c2 and on the long diagonal, and the prospect
of ...a5-a4, it appears that Black stands well. Nevertheless, White’s counterplay with f4-f5 renders the situation sharp.

17.Re2 0-0 18.Bxf6

White prepares f4-f5, since the immediate advance 18.f5 can be met by 18...Nxe4 19.Bxe7 Ne5, winning material.
If instead 18.Qh3, then 18...Bxe4! is the correct capturing sequence. After:
a) 19.Bxe4 h6, the f4-pawn will be captured with check. If 20.Bxh6, then 20...Nxe4 wins;
b) 19.Rxe4 Nxe4 20.Bxe4 (20.Bxe7 Qxf4+ 21.Kb1 Rfe8 22.Rf1 Qe5 23.Bxb4 Ndf6 and White doesn’t have enough
compensation) 20...f5 21.Bxe7 (21.gxf5 Bxg5 22.fxg5 exf5), 21...fxe4 threatens checkmate on c2.

18...Nxf6 19.f5

Black has no time to delay, but presses forward with the attack. Of course not 19...exf5 20.Nxf6+ Bxf6 21.Qxf5,
when the threat against h7 cannot be met.

20.Kb1 a4 21.Nbd2

If 21.Nd4, then 21...e5 22.Nb5 Nxe4! 23.Bxe4 Bxe4 24.Qxe4 Qb6 and the b5-knight is trapped (Yudasin).


Bringing the last piece into the attack, in a way that foreshadows the removal of the key defender on d3. Black has
several other strong options, such as 21...Nd5 with ideas of ...Nf4, when White is in trouble.

22.fxe6 fxe6 23.Qh3?

This loses to Yudasin’s fine combination. Instead White must play 23.Nxf6+ Bxf6 24.Ne4, but then Black has
24...Bxb2!! 25.Kxb2 (25.Qh3 g6 does not help) 25...Qe5+ 26.Kc1 b3 with a strong attack.

23...Rxd3!! 24.Qxd3 Ba6!!

The c2 square is the target!


25.c4 bxc3 26.Qxa6 c2+.

25...Qxc2+ 26.Ka1 Qxd1+ 27.Nb1

Black has regained his material and could calmly play 27...Qd7 now, with an extra pawn, although White has some
slight hopes for survival. Instead Black continues the attack and the game ends in sparkling fashion.


28.Nxf6+ gxf6 29.Qxe6+ Kh8 leaves White without checks, and after 30.Qe4 Rxb1+ 31.Qxb1 Qxe2 Black is up a


Black opens the long diagonal.


29.bxa3 would have been met by 29...Ne4!!, clearing f6 for the bishop. Then after 30.Qxe6+ Kf8 there is no good
answer to the threat of 31...Bf6+, for example 31.Rxe4 (31.g5 Bxg5 is no better) 31...Bf6+ 32.Re5 Rxb1+! 33.Nxb1
Qd4+ 34.Nc3 Qxc3+ 35.Kb1 Bxe5 and Black wins.

29...Kh8 30.bxa3

Black’s threat was (after 30.Qxe7, for example) 30...axb2+ 31.Kxb2 Qc2+ 32.Ka1 Qc3#.

This elegant retreat makes ...Bf6+ inevitable. White resigned.


Game 2
Lev Psakhis – Vladimir Tukmakov
USSR 1979

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Bg5 e6 7.f4 Qc7 8.Qf3 b5 9.0-0-0 b4 10.e5!?
This is one of White’s three va banque attempts against Black’s opening – and probably the most dangerous of the
three. And it is logical, in a way. While White has been developing purposefully, Black has moved the queen and
pushed the b-pawn. The problem is that it is not so easy to pierce the second player’s position, and the black pieces have
a tendency of being able to fight off the white army right from their home squares...

10...Bb7 11.Qh3

White plans to sacrifice on e6. The main alternative is 11.Ncb5, sacrificing in a different way. Other moves such as
11.Qg3 or 11.Qe3 do not appear threatening, and White should not have enough compensation for the pawn following
11...dxe5 12.fxe5 Nfd7.
11.Ncb5 axb5 12.Bxb5+ Nbd7! is important to know. In one of the first games in this line, Krum Georgiev –
Kasparov, Malta Olympiad 1980, the future world champion played 12...Nfd7, but after 13.Nxe6! fxe6 14.Qh3 Kf7
15.f5, White had a very strong attack for the two pieces and went on to win.
After 12...Nbd7, White has to tee up the sacrifice on e6 with 13.Qh3; after other queen moves, Black just plays
13...Ne4 with a winning position. But if 13...Ne4 now, then 14.Nxe6 wins, so in a game shortly after Georgiev–
Kasparov, the shocking 13...b3! was introduced (apparently first played in Kosten – Kuligowski, London 1981). This
...b4-b3 move, placing the pawn en prise four different ways, is a very important disruptive resource for Black in this

Position after 13...b3! (analysis)

14.Qxb3 (Fascinating play follows after 14.exf6 bxa2, but it very clearly favors Black. For example, 15.Nxe6 [Best is
15.Kd2, but after 15...0-0-0!, the initiative is firmly in Black’s hands. White will not be able to defend his king while
dealing with the a-pawn. Don’t forget that, sometimes, with the battle raging around him, the black king can still escape
by castling queenside.] 15...a1=Q+ 16.Kd2 Qca5+ and Black wins.) 14...Bd5 (winning time to defend e6) 15.c4 Ne4.
Black is up a piece for two pawns; however, White retains just enough pressure that Black is not clearly better. The
game Miton – Vachier-Lagrave, Dresden 2008, went down a somewhat forcing line after 16.Rhe1: 16...Nxg5 17.fxg5
Bxc4 (another possibility is 17...0-0-0 18.Ba6+ Bb7 19.Nb5 Nc5! with an unclear game, Docx – Shytaj, Novi Sad
2009) 18.exd6! Bxb3+ 19.dxc7 Bxd1 20.Rxd1 (Black is now up a whole rook, but White’s passed c-pawn and the pin
on the a4-e8 diagonal give him just enough play) 20...Bd6 21.Nxe6! Ke7 (21...Bxh2!?) 22.Nd4 Bf4+ (not 22...Bxc7
23.Nf5+) 23.Kb1 Nb6 24.Nc6+ Ke6 25.Nd8+ Ke7 (Black should take the draw, as 25...Kf5 26.Rf1! creates big
problems for him) 26.Nc6+ Ke6 27.Nd8+ with a draw in Miton – Vachier-Lagrave, Dresden Olympiad 2008.


An example of the kind of thing Black needs to avoid is 11...bxc3 12.exf6 cxb2+ 13.Kb1, Hunt – Bresciani, Bratto
2013, when White’s king is perfectly safe and all the white pieces will soon assail the black king: the sacrifice on e6 is
already threatened.


White plays the sacrifice now that Black is committed to ...dxe5. However, White can also play differently.
12.fxe5 is the main alternative. After 12...Qxe5 13.Bxf6, 13...gxf6! is considered best. (13...Qxf6 14.Ncb5! is quite
dangerous for Black – the victor in our main game came to grief in the game Chiburdanidze – Tukmakov, USSR 1980,
which continued 14...Bc5 15.Nxe6 axb5 16.Bxb5+ Nc6 17.Bxc6+ Bxc6 18.Nc7+ Kf8 19.Nxa8 and White won
shortly.) Now Black has to put up with some various hacking attempts like:
a) 14.Bb5+ axb5 15.Rhe1 Qf4+ (perhaps better is 15...Qg5+!?, when the queen controls the h5 square) 16.Kb1 Ra6!
(the only move; Black must stop the sacrifice on e6) 17.Qh5 bxc3 18.Nxe6 Rxe6 19.Rxe6+ Be7 20.Qxb5+ Bc6 21.Rxc6
Nxc6 22.Qxc6+ Kf8 23.Qxc3 Bd6, where Black is somewhat better and won after a hard struggle, Yudasin –
Tukmakov, Frunze 1981.
b) 14.Bc4 bxc3 15.Bxe6 fxe6 16.Nxe6 was Kineva – Galliamova, Samara 2003, although here Black should win
easily after 16...Be4.

12...axb5 13.Bxb5+

This is the best move. After other replies, White’s attack becomes very dangerous:
a) 13...Nbd7 14.Nxe6 and now this sacrifice is crushing: 14...Qb6 15.Bxf6 and White wins.
b) 13...Nfd7 14.Nxe6 Qb6 15.Bxd7+ Nxd7 16.Nxg7+! Bxg7 17.Qxd7+ Kf8 18.Rd6! Qe3+ 19.Kb1. White threatens
20.Qxb7 with a winning position, while if 19...Be4 then 20.Qd8+! followed by checkmate.
c) 13...Ke7. Clearly a move like this cannot lead to success. 14.fxe5 Qxe5 15.Qh4! threatens Nf5+ followed by
Qxb4+, leaving White with an overwhelming attack.
d) 13...Nc6. This is the obvious move, and it’s not so easy to crack. Nevertheless, it is not as good as Tukmakov’s
choice. After 14.Nxe6 fxe6 15.Qxe6+ Be7 16.fxe5, the position is very messy.

14.fxe5 Bxb5 15.exf6

15.Nxb5 Qxe5 leaves White down a piece with no attack.


This simple retreat is the refutation of White’s play. The main idea – the sacrifice on e6 – is stopped. Black is up a
piece for one pawn. The black pieces are all still at home, so you might expect that White could somehow pick up
another pawn or create some kind of attack “with practical chances.” The reality is that the open lines on the queenside
mean that counter-threats will come – even from pieces sitting on their original squares.
Instead, 15...Qe5 could be met by 16.Nf3 and, after 16...Qc7, White does not need to repeat and can play 17.fxg7
Bxg7 18.Rd8+ Qxd8 19.Bxd8 Kxd8 20.Qh5, when the active queen should defeat Black’s scattered pieces.


White throws everything at his opponent, since with quiet play the counterattack will be swift. For example:
a) After 16.Rhe1, the simplest is 16...gxf6!. Now 17.Bxf6 is met by 17...Qf4+. If the bishop retreats, Black can just
play ...Rxa2.
b) 16.Kb1 is met by 16...Qa5, when both a2 and g5 are hanging.
c) Meanwhile, the other violent try, 16.Nxe6, is refuted by 16...Bxe6 (not 16...fxe6 17.fxg7 followed by Qh5+)
17.Rhe1 gxf6!, when again 18.Bxf6 runs into 18...Qf4+. Otherwise, Black is just up two pieces while a2 will be
captured next.


The dangerous pawn should be eliminated. The black king will still be safe in the center, despite having White’s
pieces hovering over him.

17.Bxf6 Rg8 18.Qxh7 b3!

Once again, this typical disruptive thrust upsets White’s attacking hopes and forces him to tend to the defense of his


A last-ditch attempt.
On 19.axb3 Ra1+ 20.Kd2 Rxg2+ 21.Ke3 exf5, with moves like ...Bc5+ coming, the white king will be caught in the

19...Bxd6 20.Rxd6

20.Qxg8+ Bf8 21.axb3 Ra1+ 22.Kd2 Qf4+ 23.Ke2 Rxd1 24.Kxd1 Qxf6 and with three pieces against a rook and the
white king exposed, Black wins easily.

After this demure little move, White had to admit that the attack had failed. Black threatens both 21...Qxd6 and
21...bxa2. The knight on b8 and rook on a8 never moved, but played a great role in any case – the knight in the defense,
the rook in the counterattack.


Game 3
Vassily Ivanchuk – Maxime Vachier-Lagrave
Istanbul Olympiad 2012

1.Nf3 c5 2.e4 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Bg5 e6 7.f4 Qc7 8.Bxf6

This move is a critical answer to Black’s system. The resulting positions are similar to those arising from the Rauzer
Sicilian. Black faces some strategic difficulties, but generally stands solidly and the complex nature of the position
makes it ideal for playing for a win.


White has a number of other possibilities:

a) 9.f5. This direct move is probably not the most dangerous. The famous last game of Anand and Kramnik’s 2008
World Championship match in Bonn (where Kramnik needed to win as Black) went 9...Qc5! (a typical move whereby
Black takes control over some dark squares; this is probably better than 9...Nc6) 10.Qd3 Nc6 11.Nb3 (After 11.0-0-0
Black trades queens: 11...Qxd4 12.Qxd4 Nxd4 13.Rxd4 Bd7 and this kind of queenless middlegame is slightly better
for Black, thanks to his control over the dark squares. The king is safe, the doubled pawns are not a problem, and the
open g- and c-files will be valuable.) 11...Qe5 12.0-0-0. And now, rather than the risky 12...exf5 13.Qe3!, when White
soon got his draw and the World Champion title from a position of strength, Black should have preferred 12...Bd7, with
a complex battle in prospect.
b) 9.Qh5?! is considered to be an error due to 9...Qc5, when the exchange of queens is forced. In general, in these
structures Black should welcome a queen trade. After 10.Qxc5 dxc5 11.Nb3 Bd6 12.g3 Ke7, Black has a good position.
c) 9.Be2 has ideas of Bh5 at some later point. Black should challenge the white knight with 9...Nc6, and now 10.Qd2
transposes to positions after 9.Qd2 Nc6, with White having lost the option of Bf1-c4. After 10.Nb3, in V.Shulman –
Wojtkiewicz, Riga 1982, Black adopted an interesting setup: 10...Bg7!? 11.0-0 b5 12.Kh1 0-0 13.a3 Ne7, when White’s
14.f5 was met by simply capturing the pawn: 14...exf5 15.exf5 Bxf5 16.Rxf5 (otherwise the bishop would play the role
of a pawn on g6) 16...Nxf5 17.Nd5 Qa7 18.Bd3 Ne3 19.Qh5 f5 20.Bxf5 Nxf5 21.Qxf5 Rfe8. Next the rook came to g6
and White was left without enough compensation.
d) 9.Qf3. After this, it definitely makes sense to develop the bishop to b7. This position could also be reached by the
immediate 8.Qf3 b5 9.Bxf6 gxf6. Now Smeets – Ivanchuk, Wijk aan Zee 2010, went 9...b5 10.a3 Bb7 11.Be2 h5 12.0-
0-0 Nd7 13.f5 e5 14.Nb3 Rc8 15.Kb1 Nb6, when the knight is placed well, controlling the d5 square, and the position
is dynamically equal.


No doubt the most aggressive move, but as we will see Black faces some strategic problems. One of these is that the
pawn on b5 provides a “latch” for White to potentially open the queenside by a2-a4. This means that the black king
might have trouble finding safety. The other is that with the bishop going to b7 Black might find that the e6 square is
weaker than usual. The more solid move is 9...Nc6 10.0-0-0 Bd7:

Position after 10...Bd7 (analysis)

And now White decides between putting the bishop on c4 or on e2:

a) 11.Kb1 Nxd4 12.Qxd4 Be7 13.Be2 h5 (a typical move, preventing Be2-h5) 14.Rhf1 Qc5 15.Qd2 b5 was the
logical approach that Magnus Carlsen used in his game against Predojević in Sarajevo 2006. The black king does end up
staying in the center, while Black works up an initiative on the queenside.
b) 11.Bc4. After this move, the most aggressive response is 11...Nxd4 (Black should avoid adventures such as
11...Na5 12.Bb3 Nc4 13.Qd3 Rc8 14.Rhe1 b5 15.Kb1, when he failed to make much of an impression on the queenside
and faced an eventual central breakthrough in Azarov – Malisauskas, Warsaw 2008) 12.Qxd4 Qc5 13.Qd3 (as usual,
White should avoid the queen exchange, while if 13.Qxf6 then 13...Rg8 and Black regains the pawn with an active
position) 13...b5 14.Bb3 b4 15.Ne2, leading to a very double-edged game in Gershkowich – Drori, Maccabiah Games

10.Bd3 Bb7 11.0-0!?

This was a novelty at the time, and surprisingly it has been played only a couple of times since then. Instead of the
expected 0-0-0, White sends the king to the kingside. But now Black faces a problem with the eventual a2-a4, while he
also has to watch for the push f4-f5.


A natural move, but it might be better to eventually develop the knight to d7. For one thing, Black will not be
vulnerable to the Nd5 sacrifice as happens in the game. Additionally, the knight can go to squares like b6 or c5. Black
could begin with 11...h5, running the h-pawn down the board to create some weaknesses in White’s kingside, e.g.
12.Kh1 h4 13.h3 Nd7 with a very complex position. Black faces some difficulties regarding where to put the king, as it
will not be safe on the queenside in view of a2-a4. And yet attempts by White to open the game can easily backfire.

12.Nb3 h5

Perhaps it is worth considering Wojtkiewicz’s setup as seen above: 12...Bg7!? 13.f5 0-0. However, after 14.a4 White
has some initiative; after 14...b4 15.Ne2 the knight can eventually come to f4, pressuring e6.

13.Kh1 h4 14.h3

White should stop Black’s pawn before it reaches h3, when the white king would always be vulnerable.


The black king is not safe on the queenside. In the subsequent game Ziatdinov – Shyam, Nagpur 2012, Black tried
14...0-0-0 but after 15.a4 b4 16.Nd1 Rg8 17.Qe2 Qb6 18.a5 Qa7 19.Ne3 White had a serious initiative and Black
lacked counterplay.


White is preparing a typical Sicilian sacrifice, and Vachier-Lagrave decides to call White’s bluff.

15...b4! 16.Nd5!?
16...exd5 17.exd5 Na7

In return for the knight, the e-file is opened and the opposing knight has been driven away. But will it be enough?

18.Nd4 Bxd5?!

Black continues to take everything. But ultimately he gets into a bind. Perhaps better was to move the king to relative
safety on f8 and soak up White’s pressure: 18...Kf8 19.Nf5 Re8 20.Qxb4 Bxd5 21.Qd4 Bb7, when White has nothing
better than to capture the ugly bishop on e7: 22.Nxe7 Rxe7 23.Qxf6 Rh5, with a deceptive position, typical of the 6.Bg5
line of the Najdorf. Black has weathered the storm. White has two pawns for the piece, but no real attacking prospects
left, while Black has a number of consolidating moves (the knight will soon re-enter the scene) and the makings of a
counterattack – against g2, for example.

19.Nf5 Nc6 20.Be4!

Simple positional play. The defender of the light squares is traded off, making White’s knight on f5 invulnerable.

20...Bxe4 21.Rxe4 Kf8

A better defense seems to be 21...Rd8 22.Rfe1 Rd7, when the black king can go to d8, although White retains a strong
grip. In the game, however, Black ends up in a worse kind of pin on the e-file.

22.Rfe1 Re8 23.Qe2 Rh7

Black defends g7 in advance. 23...d5 24.Qg4 Rh7 25.R4e2 would transpose to the game.


The g7 square is defended, but Ivanchuk has a nice idea planned. Despite Black’s extra piece, the position is in fact of
a relatively slow nature.


This move does not help Black. In fact, he was just hanging on after 24...Qd7, when if 25.g3 then 25...Nd8! and the
knight reaches e6 to defend g7. Now after 26.gxh4 Ne6 27.Rxe6 (27.Rg1 Ng7 defends and wins for Black) 27...fxe6
28.Rg1, Black is just able to save himself by 28...Qc6+ 29.Kh2 Qxc2+ 30.Kh1 Qc6+, with perpetual check.

25.R4e2 Qb6

Black’s position is very difficult. It is now too late for 25...Qd7 to be effective. After 26.g3 Nd4 (the only try; if
instead 26...Nd8, then 27.gxh4 Ne6 28.Rxe6 fxe6 29.Rg1 and there is no check on c6 as in the above variation!)
27.Nxd4 Qxg4 28.hxg4 hxg3+ 29.Kg2 Rh2+ 30.Kxg3 Rxe2 31.Rxe2, Black has an awful endgame. However,
objectively speaking this is the best chance.

Ivanchuk is after more than simply regaining the piece by 26.Nxe7 Nxe7 27.Rxe7 Rxe7 28.Qc8+ Kg7 29.Rxe7,
although that, too, gives White a big advantage.


The idea is 26...hxg3 27.Qxg3 followed by 28.Rg2, when Black cannot escape.

27.Nxe7 Nxe2 28.Qg8+ Kxe7 29.Rxe2+ Kd6 30.Rxe8

In the end, the black king turns out to be the more exposed one.

One nice variation after 30...Qf2 is 31.Qf8+ Kc6 32.Rc8+ Kb5 33.Rb8+ Kc6 34.Qc8+ Kd6 35.Qxa6+ Ke7 36.Qb7+
Ke6 37.Re8+ Kf5 38.Qc8+ Kg6 39.Rg8+ Kh6 40.Qf8+ Kh5 41.g4#:

31.Qf8+ Kc6 32.Rc8+ Kd7

Or 32...Kb5 33.Rb8.

33.Qe8+ Kd6 34.Rd8+ Kc5 35.Qe3+ Kc6 36.Rd6+ 1-0

Game 4
Bryan Smith – Hristos Banikas
Rethymnon 2009
Every chess book deserves to have at least one game by its author, does it not?

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Bg5 Nbd7!?

An old and new move – it was played frequently in the 1950s and again in the 2010s – and not so much in-between.
Black avoids the doubling of the f-pawns in a way that keeps some more options open. The e-pawn might go to e5, or
Black might revert to a Dragon development with ...g7-g6, depending on how White plays. This is a very attractive
move for Black because it sends the game onto less well-trodden paths, but it is also not without its dangers.


This move, I now think, is not the most dangerous. For a number of years I have considered 7.Bc4 to be the critical
choice. Black’s most frequent reply is 7...Qb6. It looks strange to threaten b2 when, unlike in the Poisoned Pawn
variation, White can meet the threat with the natural 8.Bb3. Yet Black’s move has its points. One of them is that White
will need to develop the queen on d2 rather than e2. The bishop on b3 might be a liability if White plays as in an
English Attack by f2-f3, g2-g4, et cetera. And setups with f2-f4, while critical, also run into the problem that ...h7-h6
can force the exchange of the bishop, since after Bg5-h4 the ...Nf6xe4 tactic often works. For 8...e6 9.Qd2 Be7 10.0-0-0
Nc5 11.Rhe1 see the supplemental game Cubas – Mareco, São Paulo 2015. Or, for 11.f3, see Ortiz Suárez – Li, Las
Vegas 2014.
Also critical is 8.0-0 or 8.Qd2, when the capture 8...Qxb2 results in a new and fascinating kind of Poisoned Pawn
variation. See the supplemental game Wei Yi – Xu, Ho Chi Minh City 2017.
7.Qe2 has been played frequently in this position, in recent years. For this move, see the fascinating game Kokarev –
Wang Hao, Sochi 2012, in the supplemental games.

7...Qc7 8.Qe2!?

I headed into a position that I knew and had played before after 6.Bg5 e6 7.f4 Nbd7 8.Qe2, although Black does not
have to agree to the transposition. Black’s idea is 8.Qf3 h6 and if 9.Bh4 g5!:
Position after 9...g5! (analysis)

with the standard Najdorf sacrifice to conquer the e5 square. 10.fxg5 hxg5 11.Bxg5 Qc5! is the idea. Now 12.Be3 (or
12.Nf5 e6 13.Bxf6 Nxf6 14.Nxd6+ Bxd6 15.Qxf6 Bg3+, when there could be no question about Black’s compensation
for the two pawns in Rieger – Fedoseev, Budva 2013) 12...Ne5 13.Qe2 Neg4 14.Bg1 Bh6 gives Black great
compensation for the pawn, Abdumalik – Tari, Doha 2014.


Fans of 6...Nbd7 have tended to avoid this move, since it transposes back to better-known variations. For example,
8...b5!? 9.Bxf6 Nxf6 10.e5 b4 11.Ncb5 axb5 12.exf6 gxf6 was the sharp course of the game Dzhangobegov –
Khismatullin, St. Petersburg 2012. And the more “solid” 8...e5 has been played many times, e.g. 9.Nf3 (or 9.Nf5 h6)
9...h6 10.Bxf6 Nxf6 11.g3 Be7 12.0-0-0 0-0, Savchenko – Korobov, Sochi 2013.

9.0-0-0 Be7

More common now is 9...b5, although it may end up transposing. It is generally accepted that White should play
10.a3, since the alternatives 10.g3 and 10.g4 can both be met by 10...b4 when it looks like White does not get enough
after the sacrifice 11.Nd5. After 10.a3, the game Shirov – Ivanchuk, Jūrmala 2012, continued 10...Be7 11.g4 Bb7
12.Bxf6 Bxf6 13.g5 Bxd4 14.Rxd4 e5 15.fxe5 dxe5 16.Rd2 0-0 17.Nd5 Bxd5 18.exd5 and White had a small plus due
to the passed d-pawn and better attacking chances with h2-h4-h5, et cetera.


White threatens e4-e5, and the old 10...Nb6 11.Qe1! gives White great chances – the knight is misplaced on b6, while
White will develop harmoniously with Qg3 and Bd3. The alternative is 10.g4.


Instead, Banikas played this move, intending a pawn sacrifice.

11.Bh4 g5

The point of the previous move. This is a standard sacrifice in the Najdorf, to gain control of the e5 square.

12.fxg5 hxg5 13.Bxg5 b5 14.a3 Bb7 15.Qe1 Ng4 16.Qg3 Nde5 17.h3?!

Maybe not the most accurate. I underestimated my opponent’s next move. You can say that 17.h3 loses “half a
17.Bxe7 Kxe7 (17...Qxe7 18.h3) 18.Nxe5 Nxe5 19.Qg5+ f6 20.Qe3 leads to a similar kind of position while
avoiding letting the black queen into c5. White might end up not playing h2-h3. Nevertheless, Black cannot complain
about the result of the opening.


Played instantly, as if to point out that my last move had been a mistake. I had miscalculated this when I played


I had been planning 18.hxg4, but then I saw that I had missed something: 18.hxg4 Rxh1 19.Bxb5+ (if 19.Bxe7 Kxe7,
and now 20.Bxb5 is no longer check) 19...axb5 20.Rxh1 Nxf3 21.Bxe7 Qe3+! (I overlooked this when playing 17.h3)
22.Kb1 Nd2+. This variation is the point of 17...Qc5.
Changing the move order with 18.Bxb5+ first leads to wild tactics, with many pieces hanging: 18...axb5 19.Bxe7
Kxe7 20.hxg4 Rxh1 21.Rxh1 b4 (a strong counterblow which scared me off of this variation, but I could still play it)
22.Nxe5 bxc3 23.Qh4+ Ke8 24.Qh8+ Ke7 25.Qh4+, with a draw.

18...Kxe7 19.Nxe5 Nxe5 20.Qg5+ f6 21.Qf4 Rag8?!

Black aims to prevent White’s development, but White has a harmonious answer which my opponent probably
missed. Otherwise, the rook would stay on the queenside. The immediate 21...Bc6 was better, planning ...a6-a5 and
...b5-b4. White has his resources, although I would certainly prefer the black pieces here.


The only way to develop!

22...Bc6 23.g4 a5

Black’s attack looks more threatening at this point, but now I came alive.

24.Rf2 Rg6 25.Ne2!

Improving the knight is crucial.

25...b4 26.axb4 axb4 27.Nd4 Ba4

Preventing Nb3. If 27...Ra8, then 28.Nb3.

28.Kb1 Ra8

Black will soon have powerful threats down the a-file. But here I was able to wrest control of the game with a
surprising series of sacrifices. In the Najdorf, the black king often has to take up residence in the center. But even a
seemingly comfortable home like in this game can often prove to be vulnerable. The question of the initiative here is a
delicate one, more so than in any other opening.

First offering a pawn, to divert the knight and rook. It has to be taken, otherwise h4-h5 will follow and f6 falls.

29...Nxg4 30.Rg2 Rag8

Black needs a rook on g6 to defend the kingside, so he could not play 30...Ne5, which would be met by 31.Rxg6
Nxg6 32.Qg4 Nf8 33.Qg7+ Ke8 34.Qxf6, with an extra pawn and a continuing attack (one threat is 35.Qxf8+).


A sacrifice out of the blue. It seems scarcely believable that this works, but the sacrifice is correct.
31...Kxe6 32.Rd5 Qc7

Black must defend d6; 32...Qe3 meets with 33.Qxd6+ Kf7 34.Qc7+ Kf8 35.Rd8+ Be8 36.Bb5.


A little move, but the point of the combination. The bishop actually is hanging, and White is preparing Bf1-c4, with a
deadly lineup of pieces.


Black decides to return the piece rather than facing the full force of White’s attack. The slightly open position of the
white king can give Black some hope of counterplay.
If instead 33...Bc6, there is 34.h5 R6g7 35.Bc4 Ke7 36.Rdd2! (the pin on the g-file is a big problem) 36...Bd7 37.h6
Rg6 38.Bxg8 Rxg8 39.h7 Rh8 40.Rxg4 Bxg4 41.e5!!. This amazing “sting” at the end of the combination wins by
opening up the diagonals so that the bishop can be taken with check. Needless to say, this was not foreseen during the
game – but the intuition is very clear that White’s initiative must lead somewhere.
White also stands much better on 33...Bd7: 34.h5 R6g7 35.h6 (another possibility is 35.Rgd2!? Ke7 36.Rxd6 Ke8
37.Bc4, when White’s attack is very dangerous) 35...Nxh6 36.Qxh6 Rxg2 37.Bxg2 Bc6 38.Qh3+ Kf7 39.Rd4 and
White will win the b4-pawn, while the black king remains exposed. White has excellent winning chances.


I kept my cool here. The originally intended 34.Rxe5+ was less clear: after 34...dxe5 35.Qf5+ Kf7 36.Bc4+ Qxc4
37.Rxg6 (37.bxc4 Rxg2 and Black would have too much for the queen) 37...Bxb3 38.Qxf6+ Ke8, White has probably
no more than a draw.

34...Rxg6 35.bxa4!

Just taking the piece back. It was hard to switch from the attitude of attacking without regard to material, to simply
regaining the piece. Now White is up a pawn and his king is not too open.


Very tricky was 35...b3!?, sacrificing another pawn in the hope of creating some counterplay. The players then must
navigate turbulent waters, but in the end White reaches clear skies: 36.cxb3 Qc3 37.Rb5 Rg1 38.Qf5+ Ke7 39.Rb7+
Kd8 40.Qf2!. The only move. Now if 40...Qd3+ (best is 40...Kc8 instead, but after 41.Qxg1 Kxb7 42.Qg2! White is up
two pawns and should manage to overcome the technical difficulties) 41.Ka2 Rxf1 42.Qb6+ Ke8 and White forces
checkmate after 43.Rb8+ Kf7 44.Qb7+ Kg6 45.Rg8+.


It turns out that this is a mistake, although this evaluation rests on some very complicated variations. White should
have been content with 36.Qf5+ Ke7 (if 36...Kf7, then only now 37.Rb5! and the black king’s position on f7 means that
the variations below do not work for Black) 37.Rxe5+ fxe5 38.Qxg6 Qe1+ 39.Ka2 Qxf1, when the queen ending gives
White excellent winning chances.


Black had plenty of time on the clock, while I was in time pressure. Nevertheless, the position is very complicated,
and Banikas was not able to discover the route to the draw, which I also failed to see at the time or even for years
afterward. Computer analysis, however, shows it.
36...Rg1 was what I calculated during the game. White is winning after 37.Qf5+ Ke7 38.Rb7+ Kd8 39.Qxf6+ Kc8
40.Rb5. Now Black has his “turn” to pursue the initiative, but it fizzles out and White gets the final word: 40...Qe1+
41.Ka2 Rxf1 42.Qh8+ Kc7 43.Qb8+ Kd7 44.Rb7+ Ke6 45.Qe8+ Kf6 46.Qe7+ Kg6 47.Qg5#.
However, Black had 36...Qe1+! 37.Ka2 Rg3!, and surprisingly White cannot win. For example, 38.Bc4+ Ke7!
39.Rb7+ Ke8! (39...Kd8 loses after 40.Qxf6+ Kc8 41.Rb8+! Kxb8 42.Qxd6+ Kb7 43.Qe7+! Kb6 44.a5+!, when the a-
pawn joins the attack decisively) 40.Bb5+ Kd8 41.Qxf6+ Kc8 and, after 42.Rb8+ Kxb8 43.Qxd6+ Kb7, there is no
more than a perpetual, since the white bishop is on b5 (rather than c4 as in the previous variation) and therefore the
black king is safe on a8 or b7.
37.Qf5+ Kc7 38.Qf2!?

I could have gone straight into the queen ending seen above, but decided that this was clearer.
On 38.Rxe5 dxe5 39.Qxg6 Qe1+ 40.Ka2 Qxf1, White should probably win but this was hard to evaluate beforehand.

38...Nd7 39.Bd3?!

Time pressure. 39.Qa7+ Kd8 40.Rb7 Qc6 41.Bb5 was an immediate win, e.g. 41...Rg1+ 42.Qxg1 Qxb7 43.Qg8+
Kc7 44.Bxd7 Kxd7 45.Qg7+ Kc8 46.Qxb7+ Kxb7 47.h5.

39...Nc5 40.e5!

This is good enough nevertheless. The knight must stay on c5 because of Qa7+, so White gets to open the board for
the final attack.

40...Rg4 41.exd6+ Kxd6 42.Rb6+ Kc7 43.Rxf6 Qd4

Black reaches the endgame, but it is hopeless.

43...b3 was more fighting, but obviously there will be a mate: 44.Qh2+ Kb7 45.Rf7+ Kc6 46.Qc7+ Kd5 47.Qd8+
Ke5 48.Rf5+ Ke6 49.Qd5+ Ke7 50.Rf7+ Ke8 51.Bb5+.

44.Qxd4 Rxd4 45.Bb5 Kb7 46.h5 Rh4 47.h6 Rh1+ 48.Kb2 Rh2 49.Rc6 Nd3+ 50.Kb3 Ne5 51.Re6 Ng4 52.Re7+
Kb6 53.Kxb4 1-0

54.a5# is unstoppable.

Supplemental Games

José Cubas – Sandro Mareco

São Paulo 2015
1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Bg5 Nbd7 7.Bc4 Qb6 8.Bb3 e6 9.Qd2 Be7 10.0-0-0 Nc5
11.Rhe1 h6 12.Bxf6 Bxf6 13.f4 Qc7 14.Kb1 0-0 15.g4 b5 16.g5 hxg5 17.fxg5 Be5 18.Nf3 Rb8 19.Nxe5 dxe5 20.Qe2
a5 21.Qh5 g6 22.Qh4 Rd8 23.Re3 Rxd1+ 24.Nxd1 Qd6 25.Kc1 a4 26.Rh3 Kf8 27.Qh8+ Ke7 28.Qf6+ Kd7 29.Qxf7+
Kc6 30.Qxg6 axb3 31.cxb3 Bb7 32.Nc3 b4 33.Nd5 Rd8 34.Rh6 Kb5 35.Qf7 Bxd5 0-1

Isán Ortiz Suárez – Ruifeng Li

Las Vegas 2014
1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Bg5 Nbd7 7.Bc4 Qb6 8.Bb3 e6 9.Qd2 Be7 10.0-0-0 Nc5 11.f3
Qc7 12.h4 Bd7 13.Kb1 b5 14.a3 0-0 15.g4 Rab8 16.h5 h6 17.Bxh6 gxh6 18.Qxh6 Nh7 19.g5 Bxg5 20.Rhg1 Kh8
21.Rxg5 Rg8 22.Rdg1 Rxg5 23.Rxg5 Rg8 24.Rxg8+ Kxg8 25.Nf5 exf5 26.Bxf7+ Kh8 27.Nd5 Qd8 28.Nf4 1-0

Dmitry Kokarev – Wang Hao

Russian Team Chp, Sochi 2012
1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Bg5 Nbd7 7.Qe2 b5 8.0-0-0 Bb7 9.Bxf6 gxf6 10.Qh5 Rc8
11.Nd5 Nc5 12.Be2 Nxe4 13.Rhe1 Rc5 14.Nb3 Rxd5 15.Rxd5 e5 16.Rd3 Nxf2 17.Rf3 Ne4 18.Bd3 Ng5 19.Rf2 Be7
20.Bf5 Qb6 21.Rd2 d5 22.Kb1 h6 23.h4 Ne4 24.Bxe4 dxe4 25.Rde2 Qe6 26.Nd2 f5 27.Rf1 Bc8 28.g4 Qg6 29.Qxg6
fxg6 30.h5 0-0 31.hxg6 fxg4 32.Rxf8+ Kxf8 33.Rf2+ Kg7 34.Rf7+ Kxg6 35.Rxe7 g3 36.Re8 Kf7 37.Rxc8 g2 0-1

Wei Yi – Yinglun Xu
Ho Chi Minh City 2017
1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Bg5 Nbd7 7.Bc4 Qb6 8.0-0 Qxb2 9.Nd5 Nxd5 10.Rb1 Qc3
11.Bxd5 Qc7 12.f4 e6 13.Re1 Nc5 14.f5 Be7 15.Bxe7 Qxe7 16.fxe6 fxe6 17.Nf5 Qc7 18.Bc6+ Qxc6 19.Nxd6+ Ke7
20.Qg4 Nd7 21.e5 1-0
Chapter 2
The Classicist’s Preference: 6.Be2

The simple development with 6.Be2 is perhaps not the most aggressive or frightening way to meet the Najdorf, but one
with poison nevertheless. Not straying into unreason and sidestepping sharp theoretical lines, White builds up his game
in a rational manner. Black is invited to carry out the “Boleslavsky” advance 6...e5, with White hoping to gain an
advantage through various subtle positional schemes.
In Game 5, we deal with the lines following 6.Be2 e5 7.Nb3 Be7 8.0-0 0-0: the early try 9.f4, the preparatory move
9.Kh1, Karpov’s dangerous positional system with 9.Be3, and 9.Re1 (in the main game). Black later induces a
profitable change in structure and takes the initiative on both win
In Game 6, we cover White’s early change in the structure via Nc3-d5, with Qd1-d3. The main game combines these
with queenside castling, while kingside castling is dealt with in the notes. Black’s inaccurate opening play allows White
to develop an unstoppable attack based on classical positional principles.
Game 7 features an alternative to the standard Najdorf push 6...e5 with 6...Nbd7!?, a tricky and flexible move that
retains the possibility of choosing between an improved Scheveningen by ...e7-e6 or an improved “Dragadorf” by ...g7-
g6. This wild battle sees neither king castling – a typical, high-tension Najdorf encounter, where the player with the
stronger nerves wins.

Game 5
Vasif Durarbeyli – Anton Korobov
FIDE World Cup, Tromsø 2013

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Be2


This is the classical answer, and the manifestation of Karel Opočenský’s original idea. Even if Najdorf players find
the alternatives attractive, they are advised to study and play these positions nevertheless – the ideas are transferable to
other Najdorf positions featuring ...e7-e5.


7.Nf3 is played very rarely nowadays. The best answer is 7...h6, preventing Bc1-g5.

7...Be7 8.0-0

Black is advised to pay attention to 8.Bg5, with the idea of exchanging on f6, giving up the two bishops to maintain
control over d5. Theoretically Black has little to worry about there, but should have some ideas on how to handle
White’s simple plan. See the supplemental game V.Onischuk – Vachier-Lagrave, Corsica 2016, for an example of one
of the world’s leading Najdorf authorities combining solidity with counterplay.


White has a wide choice of ways to approach this position. The play is governed more by ideas than by forcing
variations, with the players of the white pieces making many attempts over the years to extract some advantage using
various positional subtleties. Here is a general account of White’s various tries.
The direct 9.f4 was one of White’s earliest attempts: Black can respond at once with 9...b5, for example 10.a4 b4
11.Nd5 Nxd5 12.Qxd5 Qb6+ 13.Kh1 Bb7 14.a5 Qc7 15.Qd3 Nd7 16.Bd2 Nc5 17.Nxc5 dxc5 18.Qf3 Rad8 19.Be3 f5,
and Black stood well in Larsen – Romanishin, Riga 1979.
Black has a number of good responses to 9.Kh1, with the idea of preparing f2-f4 (or perhaps f2-f3) and meeting 9...b5
with 10.a4. Black can play 9...Nc6, when contrary to usual the knight does not stand so badly on c6. Black’s main idea
is to play in a way not dissimilar to the Kalashnikov, e.g. 10.f3 Be6 11.Nd5 a5! 12.c3 a4 13.Nxf6+ Bxf6 14.Nd2 Ne7,
when Black had a solid position in the center and soon carried out the ...d6-d5 break in Tseshkovsky – Svidler, Moscow
9.Be3 Be6 10.Qd2 (after the immediate 10.a4, Black can employ a typical idea – now that the b4 square has been
weakened, 10...Nc6 makes good sense) is a system Anatoly Karpov used to great effect. White intends a2-a4-a5
followed by a number of maneuvers to stamp out Black’s counterplay and slowly improve the first player’s position:
Nb3-c1-a2-b4-d5, Ra1-a4-b4, et cetera. Black needs to be accurate to find sufficient counterplay against this system.
After 10...Nbd7 11.a4:
Position after 11.a4 (analysis)

there are two decent options:

a) 11...Rc8 (11...Qc7 comes to the same thing after 12.a5 Rac8) 12.a5 Qc7 13.Rfd1 Rfe8, when White finds great
difficulty in doing anything against Black’s solid position; moves like ...h7-h6 and ...Qc7-c6 can follow, while Black
has the ever-present possibility of changing the structure with ...Nd7-c5: this was the equalizing idea that Karpov’s
opponents never tried.
b) 11...Nb6, bringing the knight here before White plays a4-a5. Now 12.a5 Nc4 13.Bxc4 Bxc4 14.Rfd1 had been
considered slightly worse for Black since Karpov – Quinteros, Lucerne 1982 (see the supplemental games). However, in
Lékó – Shirov, Dortmund 2002, Black carried out a strong pawn sacrifice which equalized: 14...Rc8 15.Nc1 d5!?
16.Bb6 Qe8 17.exd5 Bb4 18.d6 Qd7 19.Nd3 (White cannot keep the pawn, e.g. 19.Bc7 Nd5) 19...Qxd6 20.Nxb4
Qxb4, when the position was level and Lékó’s 21.Ne4 forced a drawn ending. White has therefore sought an edge in
15.f3, delaying the transfer of the knight. However, after 15...Rc6, Black again uses tactics to carry through the ...d6-d5
break: 16.Nc1 (for 16.Kh1, see the supplemental game Jakovenko – Carlsen, Nanjing 2009) 16...d5 17.exd5 Nxd5
18.Nxd5 Bxd5, when 19.Qxd5? is met by 19...Rd6.

9...Be6 10.Bf3

With this move, White places an emphasis on controlling the d5 square. In effect, it is similar to the 6.g3 e5 7.Nb3
line. If allowed, White might carry out the maneuver Nb3-d2-f1-e3, as in that line.
10.Bf1 is a system devised by Yefim Geller in the 1980s. White’s idea is to protect e4 and thus prepare Nc3-d5. After
Black captures on d5, White plays exd5, followed by a general advance of the queenside pawns. Black has several ways
of seeking counterplay, and the variation is not considered so dangerous nowadays. 10...Nbd7 11.Nd5 (A common
mistake is 11.a4, which looks natural, but after 11...Rc8 Black plans ...Qd8-c7, when the pressure down the c-file
prevents White from playing Nd5. And if 12.Nd5 at once, then after 12...Bxd5 13.exd5 Nb6!, the only way to save the
d-pawn is by the seriously weakening 14.g4 as in Rowson – Gallagher, Lloyds Bank Open 1994.) 11...Nxd5 12.exd5
Bf5 13.a4 (White should not play 13.c4 too soon; after 13.c4 a5!, Black’s play was very thematic in the game Amann –
Bosch, Austria 1999: 14.Be3 Bg5 15.Qd2 Bxe3 16.Qxe3 Bc2 and Black’s excellent dark-square control gave him a
good position) 13...Rc8 14.c3. Now Black has several good paths, for example 14...Re8 (14...Bg6 15.a5 Bg5 is also fine
for Black) 15.a5 e4 16.Bf4 Bf8 17.Qd4 g5 18.Bd2 Bg6 19.Qe3 h6, when Black was building up play on the kingside in
Ye Jiangchuan – Sakaev, Moscow 2004.

10...Nbd7 11.a4 Nb6

The knight heads for c4 before White can prevent this with a4-a5.
Black has had good results with 11...Qb8!?, intending the maneuver ...Bd8-a5(b6). For example, 12.Nd2 Bd8 13.Nf1
Ba5 14.Bd2 Qc7 15.Ne3 Rac8 was Vovk – Vachier-Lagrave, Doha 2014.


White has generally covered the c4 square with 12.Nd2. After 12...Rc8, Black has a reasonable game.


Often, in the Najdorf, Black faces the question of how to capture on d5. The capture with the f6-knight generally loses
a tempo (exd5 attacks the e6-bishop), while the capture with the bishop gives White a potentially dangerous unopposed
light-squared bishop. Here, there is even a third possibility, that of capturing with the knight from b6. In this case,
though, it is a fairly easy choice. Black should keep the light-squared bishop, while from b6 the other knight will be able
to stay on c4, not only frustrating White’s queenside play, but also with some active designs of its own...

13.exd5 Bd7!

This is the best retreat, drawing the a-pawn to a5 and enabling the bishop to go to b5 later.

14.a5 Nc4

The opening has gone Black’s way. The position of the bishop on f3 has not been justified, and White is unable to
advance the queenside pawns. Black, meanwhile, enjoys great play on both sides of the board.

This is a very unnatural move. White should exchange the c4-knight immediately with 15.Nd2, when 15...Nxa5 loses
to 16.b4. After 15...Rc8, though, Black has no problems.

15...Rc8 16.Be2 Bb5

The point of 13...Bd7.


Again, White needs to play 17.Nd2, meeting 17...Nxa5 with 18.Bxb5 axb5 19.Rb4.


Now Black takes over the initiative completely.


This move often enters the equation in such structures. Here, there is no question of a weakness being created on a6,
while Black will develop pressure against f2.

19.axb6 Qxb6 20.c3 Bh4 21.Qc2

White had to drive the bishop away by 21.g3 Bf6, although Black has induced some weaknesses in the white kingside
and has excellent prospects.


This move might seem slightly non-standard for this structure, since the black pawns lose their flexibility. However,
White can no longer shake off the pressure against f2, while the c1-bishop is seriously restricted. Furthermore, Black
develops an additional possibility.

22.Bd3 g6! 23.Rf1

White tries to defend f2, hoping for a moment of respite to play Nb3-d2. Black has a large advantage after 23.Bxg6
Bxf2+ 24.Qxf2 Qxf2+ 25.Kxf2 hxg6.

However, this powerful move shows that White’s position is tactically fragile.


After 24.fxe3 fxe3, Black threatens 25...e2+. Now if 25.Be2 Rxf1+ 26.Kxf1 Rf8+ 27.Kg1 Bxe2 28.Qxe2 Bf2+
29.Kh1 Qxb3, Black should win. Or 25.Kh1 Rxf1+ 26.Bxf1 Bxf1 27.Bxe3 Bxg2+! 28.Qxg2 Qxe3 with an extra pawn
and an overwhelming position.

24...fxe3 25.c4 e4!

The black pawns press forward. 25...exf2+ 26.Kh1 Kg7 27.Nd2 would give White chances to fight.


26.Bxe4 would be met by 26...Rxc4.

26...exf2+ 27.Kh1 e3 28.Qe4 Bd7

Due to the en prise knight on b3, Black saves both bishops.

29.Nd4 Bf6 30.Ne6 Rce8 31.Ra3 Bg5! 32.Qg4 Qxb2 33.Rxa6 Rf4!

Black forces the removal of the e2-bishop, and the pawns will break through.

34.Qxg5 Qxe2 35.Raa1

35...Qxf1+! 0-1

White resigned in view of 36.Rxf1 e2.

Game 6
Fidel Corrales Jiménez – Orlen Ruiz Sánchez
Panama City 2011

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Be2 e5 7.Nb3 Be7 8.Be3

When White delays castling, Black should be careful. Normally some aggression is on the agenda.

Thus, 8...0-0 would likely be met by 9.g4!?. Of course, it is entirely possible for Black to allow this and continue with
9...Be6 10.g5 Nfd7. The resulting position looks like a 6.Be3 e5 7.Nb3 line, with Black having lost some options and
the move Be2 substituted for f2-f3. In this case, Qd1-d2 followed by an early f2-f4 seems logical. Black normally
prefers to deny White this opportunity by beginning with 8...Be6.


The idea of the early Nc3-d5, combined with Qd1-d3, has been played fairly frequently in the last few years. Black
has several reliable ways to meet it, yet White’s play has a logic and ease to it that make it attractive.
It is equally possible to lead with 9.Qd3 Nbd7 (Vachier-Lagrave has used 9...Nc6!? 10.a3 d5 to draw games with
Anand and Nakamura) 10.Nd5. Or 9.0-0 0-0 (but Black also can avoid White’s next with 9...Nbd7) 10.Nd5.

9...Nbd7 10.Qd3 Bxd5

It is clearly safer for Black to delay this capture until White has castled, thus avoiding the attacking setup White used
in this game.
Thus more common is 10...0-0. Now White can reach the main position after 11.0-0, or try 11.c4.
11.c4 can be met by 11...b5!, when 12.cxb5 (for 12.Nd2 Nc5, see the supplemental game Anand – Vachier-Lagrave,
London 2015) 12...axb5 13.0-0 Bxd5 14.exd5 Nb6 15.Bxb6 Qxb6 16.Qxb5 Qa7 as in Svidler – Gelfand, Haifa 2000.
This appears to give Black enough compensation.
11.0-0 reaches the main tabiya, which can be reached by various move orders. Only now should Black play 11...Bxd5

Position after 12.exd5 (analysis)

Here there is a great variety of approaches. The position is highly interesting and positionally tense, with White’s two
bishops and potentially mobile queenside pitted against Black’s strong dark-square blockade and mobile kingside
pawns. Black now has:
12...Nc5!?. This move was introduced by Najdorf legend Walter Browne. Paradoxically, Black grants White a passed
d-pawn after a trade on c5. However, Black’s dark-squared bishop will then go to d6, and with ...e5-e4 Black will gain
space in the center and on the kingside. In the first game played in this line, White replied 13.Nxc5 and, after 13...dxc5
14.Rfd1, Browne carried out his plan clearly: 14...e4 15.Qd2 Bd6 16.a4 Qc7 17.g3 Rae8 18.a5 Nd7! 19.Bf4 Ne5 20.c4
f5 21.Rac1 h6 22.h4 Ng6 23.Bxd6 Qxd6:

Position after 23...Qxd6 (analysis)

The exchange of dark-squared bishops has left Black with a great positional advantage. King – Browne, Reykjavík
In such positions, it is possible for White to push f2-f4, but then Black builds up strength on the e-file, e.g. (after
12...Nc5 13.Nxc5 dxc5) 14.c4 e4 15.Qc3 Qc7 16.a3 Bd6 17.h3 Nd7 18.f4 exf3 19.Bxf3 Rae8 20.Rae1 Qd8 21.b4 b6
22.Bd1 Re5 23.Ba4 Nf6 24.Bc2 Rfe8, and in Paragua – Bu Xiangzhi, Beijing 2008 (rapid), Black had pressure on the e-
file while the white king did not feel comfortable.
However, White has found a way to put some pressure on Black with 14.Bf3!, making it difficult for Black to
advance the e-pawn. For example, in Nakamura – Topalov, Thessaloniki 2013: 14...Qc7 15.c4 g6 (or 15...Bd6 16.g3
Rae8 17.Qc2! e4 18.Bg2 when the pressure on the e-pawn hinders Black’s ...Nf6-d7/...f7-f5/...Nd7-e5 maneuver)
16.Rae1 h5 17.Bg5 Ne8 18.Bxe7 Qxe7 19.Qe3 Nd6 20.b3 Rfc8 21.Qxe5 Qxe5 22.Rxe5, and Black’s attempts to hold
the blockade on the dark squares proved unsuccessful.
It is likely, therefore, that Browne’s classical method, while a theme worth seeing, is not the most promising for
Black. The two other principal approaches are to “develop” with ...Nf8-e8 and ...f7-f5 (or ...g7-g6 and ...Ne8-g7-f5), or
to rearrange the kingside pieces with 12...Re8 followed by ...Be7-f8, when the advance ...e5-e4 becomes crucial. Black
can combine either of these ideas with queenside moves such as ...Ra8-c8 or ...a7-a5:
a) 12...Ne8 intends ...f7-f5, and looks like a promising plan for Black, e.g. 13.Bg4 g6 14.Bxd7 Qxd7 15.f4 f5 16.c4
Nf6 17.fxe5 dxe5 18.Nc5 Bxc5 19.Bxc5 Rfe8 20.b4 b5! and Black stood well in Padmini – Adhiban, Pune 2015.
b) 12...Re8 intends ...Be7-f8, with an eventual ...e5-e4, hoping to obtain play on the kingside:
Position after 12...Re8 (analysis)

White’s play was ineffective in the game Ruiz Aguilar – Bruzón Batista, Bogotá 2011: 13.Rfd1 Rc8 14.Rac1 g6 15.c4
b6 16.Nd2 a5 17.Nb1 Nc5 18.Bxc5 (normally, if White has to make this exchange then the opening has not turned out
well for him; in the resulting position with opposite-colored bishops, Black will have mobile pawns on the kingside and
can build slow pressure there, while White has no prospects anywhere) 18...bxc5 19.Nc3 Bf8 20.Qh3 e4! and Black
stood well.
White showed much greater subtlety in Hovhannisyan – Andriasian, Yerevan 2012: 13.c4 Bf8 14.Nd2! a5 15.Bd1!
(White improves each of his pieces; the knight should go to c3, the bishop to c2 or a4) 15...Nc5 16.Qe2 g6 17.Nb1!
Nh5? (17...Nfd7) 18.Bxc5 dxc5 19.Qd2 Ng7 20.Ba4 Re7 21.d6 Re6 22.d7 Nf5 23.Qd5 when White had a big
advantage – the passed pawn was tying down the black pieces.
c) Finally, it is very logical and solid to begin constructing the queenside blockade with 12...a5, although this has not
been played very much. For example, 13.c4 b6 14.Nd2 Nc5 15.Qc2 Nfd7 16.b3 Bg5 17.Bxg5 Qxg5 18.a3 Nf6 19.b4
was agreed drawn in Palac – Vera González, Lucerne 1997.

11.exd5 Nc5?!

Black utilizes Walter Browne’s original idea. However, the idea does not work well when White can castle queenside
and push the g-pawn.
It is better simply to play 11...0-0, and although this gives White the additional possibility of 12.g4, Black can look
for counterplay here, avoiding ...Nc5: 12...Nb6!? (12.0-0 transposes to the above lines) 13.c4 Na4! was an idea
suggested by Viktor Korchnoi and put into practice in some later games, such as Kaufman – De Firmian, Foxwoods
2002 and Mészarós – Wojtaszek, Turkey 2004. Black controls the c5 square with ...Nf6-d7 and eventual counterplay
can come by ...f7-f5 or ...f7-f6 after White plays g4-g5.

12.Nxc5 dxc5 13.0-0-0 Bd6 14.g4!

This is not just a prelude to a kingside attack. Advancing the g-pawn also enables White to control the central light
squares, e4 in particular.


14...e4 15.Qd2 0-0 16.g5 occurred, by transposition, in the first game where White used this g2-g4 idea. In Khalifman
– Gelfand, FIDE World Chp playoff, Las Vegas 1999, after 16...Nd7 17.h4 Ne5 18.h5 Rc8 19.Rh4±, Black could not
defend the e-pawn and his queenside attack after 19...c4 20.Rxe4 c3 proved insufficient.

15.g5 Nd7 16.h4

White’s play is not very hard to understand.

16...c4 17.Qf5 c3 18.h5 Qa5

Black’s queenside demonstration looks superficially dangerous, but he is playing on only three files. White can
defend the queenside and the inevitable g5-g6 will come with much greater force.

19.Kb1 Qb4

19...cxb2 20.g6 leads nowhere for Black.


20.b3 was also sufficient. After 20...a5 21.g6, White’s play comes first.

20...cxb2 21.Bd2 Qa3 22.Rh3!

The rook is activated with gain of time.

22...Qa4 23.g6
23...hxg6 24.hxg6 Rxh3 25.gxf7+ Kd8 26.Bg4!

This is stronger than 26.Qxh3. Now on 26...Nf6 27.Bg5, White’s attack is unstoppable.

26...b4 27.Bxh3 Kc7 28.Rg1 Nb6

28...Bf8 29.Rg6 intends Rc6+, and Black will not last long.


White clears the seventh rank for the final assault.


29...Bxf8 loses to 30.Qxe5+ Kb7 31.Rxg7+ Bxg7 32.Qxg7+ Kb8 33.Bf4#.

30.Rxg7+ Kb8 31.Qh7 1-0

Game 7
Viktor Bologan - Ivan Cheparinov
European Club Cup, Kemer 2007

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Be3

The game we are considering began with this move order, and I will leave it unaltered. However, it soon transposes to
a position which can arise after 6.Be2 Nbd7!?:

Position after 6...Nbd7!? (analysis)

We will use this game to examine this interesting move. Black’s idea is flexibility. Depending on how White plays,
Black may play ...g7-g6 and enter a Dragadorf-style position, or play ...e7-e6 with an improved Scheveningen.
Now if 7.0-0 or 7.f4, then 7...g6 and Black reaches a good kind of Dragadorf, where White is committed to 0-0. For a
thematic example of play in such positions, see the supplemental game Ehlvest – Short, Parnu 1996.
The aggressive 7.g4 is also quite common. It will likely lead to the position in the game after 7...h6.
After 7.Be3, on the other hand, Black returns to Scheveningen play with 7...e6! and carries out a quick ...b7-b5.
(Here, 7...g6 8.Qd2 is playable, but this line is fairly dangerous for Black as White plays out a quick Be3-h6 and
maintains the possibility of castling on either side of the board.)
Now White can turn to aggression to try to punish Black’s provocative setup with 8.g4!?. (Instead 8.f4 b5! reaches an
improved line from the Scheveningen – White normally meets the quick ...b7-b5 by 0-0 and Be2-f3, not spending a
tempo on Bc1-e3. But here Be3 has already been played, and after 9.Bf3 Bb7 10.e5 Bxf3 11.Qxf3 dxe5 12.Nc6 Qc7
13.Nxe5 Rc8 Black had absolutely no problems in Adams – Rowson, England 2004. Or if 8.a4, then 8...b6 and the
bishop goes immediately to b7, as in Tal – Browne, Milan 1975.) Following 8...h6 9.f4 g5!?, we reach the position in
the game.

6...e6 7.Be2

More common, of course, would be 7.f3. But Bologan is trying to reach a 6.Be2 line with Black’s having committed
to ...e7-e6 already. On the other hand, White is committed to Be3.

7...Nbd7 8.g4 h6 9.f4 g5!

Every Najdorf player needs to be well acquainted with the ...g7-g5 motif. Very often, this move is justified whether it
is a pawn sacrifice or not. Black breaks up White’s pawn front and obtains control over the e5 square.
Veselin Topalov (for whom the victor in the main game worked as a second) played two games against Michael
Adams with 9...g6. His experiences there probably led to the development of the idea 9...g5.


The main alternative is 10.f5 Ne5 11.h3, when Black possesses the excellent e5 square for the knight, but that is not
the end of the story. White has some pressure on e6, d6, and f6, so the game is positionally complex. For example,
11...b5 12.a3 Qe7 13.fxe6 fxe6 14.Nf3 Bb7! (Black accepts the doubled e-pawns, but they help to control many central
squares. Trying to keep the use of the e5 square would leave Black under some pressure, e.g. 14...Nfd7 15.Nxe5 Nxe5
16.Qd2, planning 0-0-0 with pressure against d6, and h3-h4 on the horizon.) 15.Nxe5 dxe5 16.Bd3 Qc7 17.h4 Bc5,
when Black stood just fine in Ponomariov – Topalov, Thessaloniki 2013.

10...hxg5 11.Bxg5 b5

As so often, Black sacrifices the pawn on the wing and then continues with development nonchalantly. It is possible
to recover the pawn with 11...Qb6 12.Nb3 Rxh2, but after 13.Rf1 Bg7 14.Bf4 Rh3 15.g5 White regains the initiative
and Black finds it hard to complete his development.


Interesting is the direct 12.e5!?, which should be met by 12...dxe5 (and not 12...Nxe5? 13.Bf3 d5 14.Qe2, with a very
strong attack, e.g. 14...Bd6 15.Nxd5! exd5 16.Nc6 followed by Bxf6) 13.Nxe6! fxe6 14.Bf3 Rb8 15.Qd3 Kf7 16.0-0-0:

Position after 16.0-0-0 (analysis)

when White has a deceptive kind of compensation. Black is up a piece for only one pawn with White having no direct
attack. Nevertheless, it is hard to organize the black pieces and there is nothing for Black to latch onto in the white camp
to gain counterplay. Still, while there are practical difficulties, Black objectively stands well.


With this spectacular move, White begins to undermine Black’s otherwise rock-solid central position, and the
complications now begin.

13...Nxc6 14.e5

This is the point.


14...Nxe5 15.Bxa8 Be7 16.Qd4 and Black should not have enough compensation for the exchange.

15.Bxf6 Rh3!

Black activates the rook in a common way – the restricted advance on the h-file. Now it controls the third rank,
creating the threat of ...Qe3+. The alternative 15...Rxh2 would not have worked due to 16.Bxc6+.


This might be too risky. White has a couple of other possibilities in this highly complex position.
16.Bg5 covers e3, but after 16...d5 the threat of ...Nxe5 necessitates a sacrifice on d5: 17.Nxd5 exd5 18.Qxd5 Rxf3!
(Black must remove this important bishop) 19.Qxf3, and here Black can play the solid 19...Be6, when the two minor
pieces are far more active than the rook, giving Black good play despite White’s three extra pawns.
16.Kf1 is met by 16...Bb7 17.exd6 Ne5! 18.Bg2 Bxd6, when White is under strong pressure.
16.Rf1 might be best. Play could then go 16...Qe3+ (16...d5!?) 17.Qe2 Qxe2+ 18.Nxe2 d5, when Black regains the
h2-pawn with a sharp endgame.

16...Qe3+ 17.Be2 Ne5!

Black increases the energy of his pieces, rather than chasing after material gains at the cost of the initiative by
17...Rxh2? 18.Rxh2 Qg1+ 19.Bf1 Qxh2 20.Qf3 Bb7 21.0-0-0, when White has completed his development ideally and
is now winning.

Black clearly gets the upper hand after this. White had to first include 18.Rf1 Bb7 and only then play 19.Qc1, when
19...Nf3+ would win the exchange but allow White to eliminate the threats. 20.Rxf3 Qxc1+ (not so good is 20...Rxf3
21.Qxe3 Rxe3 22.0-0-0) 21.Rxc1 Bxf3 22.Rd1 Bc6 23.Ne4! and White, with two pawns for the exchange and a strong
passed d-pawn, is not worse.
Instead, after 18.Rf1 Bb7 19.Qc1, Black should prefer the simple 19...Qxc1+! 20.Rxc1 Bxd6 when White’s position
remains uncomfortable. The black pieces are more active and White will hardly be able to keep the extra pawn.

18...Nf3+ 19.Kd1


Not the strongest. The queen should stay more active on the fifth rank. After 19...Qc5!, Black can now meet 20.Ne4
with 20...Qd5+ 21.Bd3 e5, planning ...Bxg4, when he is in full control.
Or 19...Qc5 20.Qf4 e5! 21.Qe4 Qxd6+ 22.Bd3 Rb8 23.g5 b4 24.Ne2 Bb7, also with a strong initiative for Black.


This is possible here, allowing White to stem some of Black’s initiative.

With the queen on b6, 20.Qf4 is met simply by 20...Bxd6, since 21.Qe4 Bb7 leads to a winning position for Black.


Nevertheless, Black retains all the chances. This move opens the diagonal for the bishop from c8 and also cuts off
White’s f6-bishop, enabling ...Qd4+.


White must capture the troublesome knight, even at the cost of abandoning some light squares. Otherwise, there was
no easy way to meet the threat of 21...Qd4+ nor any way for White to unwind his position.

21...Rxf3 22.Qd2 Bxg4 23.Kc1

White is still without the rook on a1 and there are some problems with the light squares, but finally the king reaches
some kind of safety, while the passed d-pawn still provides counterplay.
It was possible to simplify the position by 23.d7+ Bxd7 24.Qd5, threatening 25.Qxa8+, 25.Nd6+, and 25.Qxe5+; but
after 24...Qd4+ 25.Qxd4 exd4 26.Re1 Be6, Black’s two bishops would have continued to make life uncomfortable for


In this highly complex situation, no doubt on the edge of time pressure, Bologan now falters.


White meets the threat to the knight with a counter-threat of 25.d7+, but he overlooks Black’s next move, after which
the white pieces are “hanging in air.” It was necessary to play 24.Re1 Bf5, when:
a) 25.Ng3 Rd3! is a powerful blow. Then 26.Rxe5+ Kd7 27.Qg5 (White must prevent 27...Bh6+) 27...Qxc2+!
28.Kxc2 Rc8+ 29.Rc5 (29.Kb1 Rd1#) 29...Rxc5+ 30.Bc3 Rxg3+ (30...Rdxc3+!? 31.Kd2 Rc2+ 32.Ke3 Be6 and the
white position is also practically very difficult to play) 31.Qxf5+ Rxf5 32.hxg3 Bxd6 and White has not quite reached
the safety of a draw yet – there are some problems defending the g-pawn.
b) 25.Ng5 also leads to problems for White after 25...Bh6 when the pin is very uncomfortable.
c) However, 25.d7+! is the route to a rough equality. After 25...Bxd7 26.Bxe5 0-0-0 27.Nd6+ Bxd6 28.Qxd6 Qxd6
29.Bxd6 Rf2, White’s extra pawn is not relevant, but he will succeed in quelling Black’s slight initiative, with a likely


Black not only threatens the knight (which must guard the bishop), but also attacks d1, so the game is over now.

25.d7+ runs into 25...Bxd7, maintaining the threats, when 26.Nd6+ Bxd6 27.Qxd6 Qxd6 28.Rxd6 Rf1+ wins.

25...Rxf6 26.d7+ Kd8 27.Qa5+ Qb6 28.Nb7+ Kc7

Here White resigned, since after 29.Qxb6+ Kxb6 30.d8=Q+ Rxd8, either capture on d8 is met by 31...Bh6+.


Supplemental Games

Vladimir Onischuk – Maxime Vachier-Lagrave

Bastia 2016
1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Be2 e5 7.Nb3 Be7 8.Bg5 Be6 9.Bxf6 Bxf6 10.Qd3 Be7 11.Nd5
Nd7 12.Rd1 Bxd5 13.Qxd5 Qc7 14.c3 Nf6 15.Qd3 0-0 16.Nd2 Rfd8 17.Bf3 b5 18.0-0 g6 19.Ra1 Nd7 20.Rfd1 Nb6
21.g3 d5 22.exd5 f5 23.h4 e4 24.Nxe4 fxe4 25.Bxe4 Nc4 26.Bxg6 Bc5 27.d6 Rxd6 28.Qf3 Rf8 29.Bf5 Kh8 30.b4 Rdf6
31.Rd5 Ne5 32.Qe2 Bxf2+ 33.Qxf2 Rxf5 34.Qc5 Qg7 0-1

Anatoly Karpov – Miguel Quinteros

Lucerne Olympiad 1982
1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Be2 e5 7.Nb3 Be7 8.0-0 0-0 9.Be3 Be6 10.Qd2 Nbd7 11.a4
Nb6 12.a5 Nc4 13.Bxc4 Bxc4 14.Rfd1 h6 15.Nc1 Rc8 16.Nd3 Be6 17.Bb6 Qe8 18.f3 Nd7 19.Be3 f5 20.exf5 Rxf5
21.Nb4 Nf6 22.Nbd5 Nxd5 23.Nxd5 Bf8 24.b3 Qf7 25.Nb6 Rc6 26.c4 Be7 27.Qe2 Bd8 28.Nd5 Bxd5 29.Rxd5 b5
30.axb6 Bxb6 31.Kh1 Bxe3 32.Qxe3 Qc7 33.Rad1 Rf6 34.h3 Qb6 35.Qd3 Qb4 36.Kh2 Qb8 37.Re1 Rf8 38.Re4 Qc7
39.Rg4 Rf6 40.b4 Qb6 41.c5 Qc7 42.Kh1 a5 43.Rc4 Qb8 44.cxd6 Rcxd6 45.bxa5 Qd8 46.Rcc5 1-0

Dmitry Jakovenko – Magnus Carlsen

Nanjing 2009
1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Be2 e5 7.Nb3 Be7 8.0-0 0-0 9.Be3 Be6 10.Qd2 Nbd7 11.a4
Nb6 12.a5 Nc4 13.Bxc4 Bxc4 14.Rfd1 Rc8 15.f3 Rc6 16.Kh1 Qc8 17.Rac1 Rd8 18.Nd5 Bxd5 19.exd5 Rc4 20.Qd3 e4
21.fxe4 Rxe4 22.c4 Re8 23.Bg1 Bf8 24.Nd4 g6 25.Rf1 Bh6 26.Qf3 Rf4 27.Qd3 Ng4 28.Nf3 Rfe4 29.Rc3 Ne3 30.Re1
Qg4 31.Re2 Qh5 32.Bxe3 Rxe3 33.Rxe3 Bxe3 34.Qe2 Qh6 35.c5 dxc5 36.d6 Re6 37.d7 Bg5 38.Qd1 Bd8 39.Rxc5 Qf8
40.Rd5 Qb4 41.b3 Re3 42.Nd2 Qc3 43.Nf3 Qb4 44.Nd2 Qf4 45.Nf3 Rc3 46.Qe2 Qe3 47.Qxe3 Rxe3 48.Rd4 Kf8
49.Rb4 Rd3 50.Rxb7 Rd1+ 51.Ng1 Bxa5 52.g4 Ke7 53.Kg2 Rxd7 54.Rxd7+ Kxd7 55.Kf3 Kd6 56.Ke4 Kc5 57.Kd3
Kd5 58.Nf3 Bd8 59.h3 h6 60.h4 h5 61.gxh5 gxh5 62.Ke3 Kc5 63.Kd3 Kb4 0-1

Viswanathan Anand – Maxime Vachier-Lagrave

London 2015
1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Be2 e5 7.Nb3 Be7 8.Be3 Be6 9.Nd5 Nbd7 10.Qd3 0-0 11.c4 b5
12.Nd2 Nc5 13.Bxc5 dxc5 14.b3 Bxd5 15.cxd5 Ne8 16.0-0 Nd6 17.a4 Bg5 18.Nf3 Bf4 19.axb5 f5 20.Nd2 Qg5
21.Rad1 axb5 22.exf5 Ra3 23.Ne4 c4 24.Qc2 Qxf5 25.Qb2 Rxb3 26.Qxb3 cxb3 27.Nxd6 Qg6 28.Nxb5 e4 29.d6 b2
30.Nd4 Qxd6 31.Bc4+ Kh8 32.Ne6 Bxh2+ 33.Kh1 Rxf2 34.Ng5 Bg3 0-1

Jaan Ehlvest – Nigel Short

Pärnu 1996
1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Be2 Nbd7 7.f4 g6 8.0-0 Bg7 9.Bf3 0-0 10.Kh1 e6 11.Nb3 Qc7
12.Be3 Nb6 13.Nd2 Rd8 14.a4 Rb8 15.Qe2 Bd7 16.a5 Nc8 17.Rfd1 Bc6 18.Nb3 Re8 19.Bd4 Nd7 20.Bxg7 Kxg7
21.Qd2 e5 22.f5 Nf6 23.Nc1 b5 24.axb6 Rxb6 25.Nb3 Ba8 26.Rf1 Ne7 27.Ra4 Rd8 28.Qf2 d5 29.Nxd5 Nexd5
30.exd5 Bxd5 31.Bxd5 Rxd5 32.fxg6 hxg6 33.Qe2 e4 34.Rc4 Qe5 35.Rc3 Rbd6 36.g3 a5 37.Ra1 Qg5 38.Rc4 Rd2
39.h4 Qxh4+ 40.gxh4 Rxe2 0-1
Chapter 3
Add Some English: 6.Be3

Together with the setup of f2-f3, Qd2, 0-0-0, and a kingside pawn storm, the move 6.Be3 first become popular in the
1970s thanks to the games of various English grandmasters – Nigel Short, John Nunn, and Tony Miles, among others.
Therefore the line is known as the “English Attack.” This is now one of the critical variations against the Najdorf.
Black can respond by changing to a Scheveningen structure with 6...e6, by attempting to disturb the bishop with
6...Ng4, or by carrying out the typical Najdorf advance with 6...e5. As elsewhere in the book, I have limited the games
to this last move.
In Games 8 and 9, we cover the variation with 7.Nb3 Be6 8.f3 h5!?, whereby Black prevents g2-g4 at some cost in
the weakening of the kingside.
Game 8 features White’s aggressive plan involving queenside castling. In this game we see a thematic way for Black
to exchange the dark-squared bishops, gaining a positional advantage and ultimately victory, with the help of a small
simplifying combination.
Game 9 covers White’s more positional treatment with kingside castling. Here we see an elegant breakthrough for
Black, gaining the initiative in a surprising fashion.
In Game 10, the older main line with ...Nbd7 and ...b7-b5 is covered in the course of a strange game featuring an
unusual arrangement of pieces on the a-file.
Game 11 deals with White’s alternative (and more positional) retreat 7.Nf3. Here we will see some ways that Black
can obtain a fighting position. In this game, simplification does not diminish Black’s initiative. Black’s positional
advantage persists even into the pawn ending.
Note that the move 6.f3 is also an English Attack, and from the point of view of the games presented here, makes no
difference after 6...e5 7.Nb3, when Bc1-e3 will follow shortly. The main point of 6.f3 is to avoid Black’s possibility of
6...Ng4, but it also costs White his opportunities for setups avoiding f2-f3 and allows Black the additional opportunity
Game 8
Allan Stig Rasmussen – Tiger Hillarp Persson
Danish Team Chp 2008

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Be3 e5

Black uses the thematic Najdorf move, rather than Kasparov’s favorite 6...e6, which steers the game into
Scheveningen waters; or 6...Ng4, aiming to displace the white bishop.


In Andriasian–Nepomniachtchi we will examine the more positional alternative, 7.Nf3.


If Black wants to use the ...h7-h5 system as in this game, he should normally delay ...Bf8-e7, since in many cases
(such as in the current game) the bishop does not go to e7 at all, or at least its development is delayed. 7...Be7 is played
frequently, though, and it can introduce a number of other systems for Black.


This move is the basis of the English Attack setup: White anchors e4 and g4 by playing f2-f3, then Qd2, 0-0-0, and
g2-g4. It is a dangerously simple system where Black can just as easily fall under direct attack as come under positional
pressure as a consequence of attempts at counterplay. However, White has major alternatives:
For 8.Qd2 (delaying f2-f3 and perhaps preferring f2-f4 depending on circumstances), see the supplemental game
For 8.h3 (which has some commonality with the 6.h3 system), see the supplemental game Anand–L’Ami.
And for 8.f4, see the supplemental game Pikula–Damljanović.

The widespread acceptance of this move, which spends a tempo and potentially creates weaknesses in order to
prevent g2-g4, represents a shift in the understanding of Najdorf structures that has occurred in the last fifteen years or
so. It can now be found in most variations where White aims for g2-g4 – primarily in this one as well as in the 6.h3 line
(see Chapter 5). This advance allows Black to keep the knight on f6, while in the majority of cases, if Black plays right,
the pawn on h5 won’t constitute a weakness of the black kingside. In some cases this advanced pawn can even provide a
benefit to Black. In particular the advance of the h-pawn to h4 or even h3 could become a factor, cramping White’s
kingside. There are also ideas of ...h5-h4 with ...Nh5-f4. And very importantly, White’s advance h2-h3 can be met by
...h5-h4, locking down on the g3 and f4 squares.


White immediately decides to change the structure. This move can be paired with castling kingside or (as in this
game) queenside. White can also delay Nd5 or do without it altogether, castling kingside or queenside.


One of the ideas of White’s immediate 9.Nd5 (as opposed to 9.Qd2 Nbd7 10.Nd5) is that, if 9...Nxd5 10.exd5 Bf5,
then 11.Bd3 Bxd3 12.Qxd3 and White has gained a tempo since the queen did not go to d2 first. Although this line is
playable, the loss of a tempo has tended to dissuade people from trying it as Black.


White has gained the two bishops, which is a long-term strategic advantage here. Black’s central pawns are on the
dark squares, so in the absence of the light-squared bishop there is a palpable weakness of the light squares.
Nevertheless, this variation has a pretty good reputation. Black’s central fortifications are very strong. The advance f3-
f4 tends to give Black play on squares such as e5 or the open e-file, or to allow a passed pawn by ...e5-e4. Meanwhile, it
is not easy for White to carry out a c4-c5 advance, since Black will blockade the queenside with ...a6-a5 and ...b7-b6.
Thus White’s possibilities of starting an attack are limited. Black, meanwhile, has a plan of exchanging the dark-squared
bishops, which would give him a positional advantage. Depending on which side White castles, Black can also begin
operations against the white king in various ways.
10...Nbd7 11.Qd2 Qc7

Black stays flexible with this move. The game Anand – Topalov, Wijk aan Zee 2008, went 11...g6 12.0-0-0 Nb6?!,
aiming to force the exchange of the dark-squared bishops. After 13.Qa5 Bh6 14.Bxh6 Rxh6:

Position after 14...Rxh6 (analysis)

Black had achieved his positional goal of trading off these bishops. White was left with a light-squared bishop blocked
by some pawns, while Black had the makings of a superior minor piece. Nevertheless, Anand showed that the resulting
position was promising for White. A slight looseness of the black kingside and of the dark squares in general could be
felt. While the exchange of bishops is a positional goal for Black, it is worth remembering that the dark-squared bishop
was keeping the black position solid. “Bad bishops defend good pawns,” as Mihai Şubă has said.
The game proceeded with 15.Kb1 Rc8 16.Qb4 Kf8 17.c4 Kg7 18.g3! (Anand continues his unhurried play. The
bishop will come to a good diagonal h3-c8, with the potential of eliminating a knight on d7. White is also planning to
play f3-f4 at some point, opening the position. If Black were able to play ...Nbd7, ...b7-b6, ....Nc5, ...a6-a5, and then
concentrate heavy pieces on the e-file, he would stand well. But the lack of coordination caused by ...Nb6 and ...Rxh6
has cost him.) 18...Rh8 19.Rc1 Qc7 20.Bh3 Rce8 21.Rhd1 Re7 22.a3! (Since Black cannot do much, Anand makes every
useful move. The white king now has a secure escape on a2.) 22...Rd8 23.Nd2 Nbd7 24.Qc3 a5 (White was planning to
meet 24...Nc5 with 25.b4, although perhaps it was worth it for Black to induce that move. After 25...Ncd7 the white
king will always be a little drafty, which will constrain his ability to open the game safely.) 25.Bxd7! (The bishop was
not achieving much, so White exchanges it. Black’s potentially superior minor piece is gone, and White is simply left
with a space advantage and the safer king.) 25...Nxd7 26.f4 Nf6 27.Rf1 b6 28.h3! Qd7 29.f5 Rf8 (29...gxf5 is met by
30.Qe3 Nh7 31.Nf3, followed by Nh4) 30.Qe3 e4 31.g4! hxg4 32.hxg4 Re5 (if 32...Nxg4, then 33.Qg5 Nf6 34.Rh1)
33.Rf4 Qd8 34.g5 and White had a winning attack.
Besides White’s plan of castling queenside, as used in our main game, White can also castle kingside after Nd5.
11...g6 12.Be2 followed by 0-0 is the game Caruana–Gelfand (found in the Supplemental Games). In that contest, White
used a unique plan, avoiding playing c2-c4 and instead playing g2-g4, attempting to gain space and restrict Black’s
kingside play. This brought victory, although Black has ways to reach a good position.

A frequently seen way of trying to prevent a queenside blockade. Thus ...b7-b6 will be met by Nc6.
The natural 12.c4 b6 13.Be2 g6 14.0-0 Bg7 15.Rac1 a5 allows Black to blockade the queenside: 16.Na1 0-0 17.Nc2
Nh7!? 18.Na3 Nc5 19.Nb5 Qe7 20.Rce1 f5, and Black had promising play in Nijboer – Navara, Wijk aan Zee 2011.


This development of the bishop is typical in the ...h7-h5 variation. In most cases the bishop will stand better on g7
than on e7, or it can even be exchanged by ...Bf8-h6.



A paradoxical idea that has been seen in a number of similar positions and should be part of a Najdorf aficionado’s
repertoire. The knight returns home, but from here it can support ...Bf8-h6 or it can come via e7 to f5. Also, the f-pawn
is now free to advance.


After 14.0-0, 14...Bh6 is pretty solid for Black.


The knight heads for f5 with gain of tempo.


It was worth considering complicating the game by 15.g4!?, which prevents Black’s plans of ...Nf5 and ...Bh6 but
gives Black a great structural advantage after 15...hxg4 16.fxg4. Now if 16...Nf6, White should get some play for the
sacrificed pawn. Black could play something else, such as 16...Rh3, but there will always be some uncertainty about the
black king’s home, which will keep the game sharp.
15...Nf5 16.Bf2 Bh6

Part of what makes Black’s play effective is the hanging position of the a5-knight. Were the knight on b3, for
instance, White could now play 17.Qd3, with g2-g4 to follow.

17.Qb4 Be3!

Now Black achieves the sought-after bishop trade in favorable circumstances.

18.Rhf1 Bxf2 19.Rxf2

Compared to Anand–Topalov (see above), here the exchange of dark-squared bishops has resulted in a pleasant
position for Black. Rather than an awkwardly placed knight on b6 and rook on h6, the black pieces are placed well. It
will not be easy for White ever to expel the knight from its strong post on c5. The knight on f5, meanwhile, aims at
squares like d4 and e3.


It was possible to exchange queens at once with 19...Qc5, and Black would have a better endgame after 20.Qxc5
Nxc5. Nevertheless, it was not bad to take the more complicated path, as Hillarp Persson did.


There is very little for White to do, so most likely White did not mind giving Black ...b7-b5 with tempo, since this
advance is after all double-edged, allowing White the c6 square for the knight.

20...b5 21.Na5 Qb6 22.g4

White tries to complicate matters, looking for chances in the loose position of the black king. It was better to instead
play 22.Nc4, when Black should agree to a repetition by 22...Qc7 23.Na5, after which he can play for a win by a move
like 23...Rc8, rather than 22...Qa7 23.g4! hxg4 24.fxg4 bxc4 25.gxf5 when the position opens up favorably for White,
whose king is completely safe while Black’s is not.
22...hxg4 23.fxg4 Ne3 24.Rd2 f5!

Ignoring the loosening of his king’s position, Black prepares to anchor the knight on e3 while creating two connected
passed pawns. These pawns, in addition to the strong knights and the weakness of h2, will constitute Black’s
overwhelming positional advantage. The question remains whether White will succeed in causing problems for the
exposed black king.

25.gxf5 gxf5 26.Bf3 f4 27.a3 Kd7

The black king heads for relative safety on the queenside and connects the rooks.

28.Rde2 Kc7 29.Re1 Nb7

The exchange of the white knight would be decisive. In this way Black manages to advance his play on the queenside,
achieving the ...a6-a5 push.

30.Nc6 a5 31.Qd2

A better chance was 31.Qc3, but then after 31...Nc5 Black would soon follow up with ...b5-b4.

Black now simplifies the game by force.

32.Qe2 Qe3!

The exchange of queens will leave Black with his positional advantage while removing all of White’s counterplay.

33.Qf1 Qxf3! 34.Rxf3 Nd2+ 35.Kc1 Nxf1 36.Rexf1 Rxh2

With connected passed pawns, Black is easily winning.

37.Rg1 Nc5 38.b4 axb4 39.Rg7+ Kb6 40.axb4 Rh1+ 41.Kd2

41.Kb2 loses to 41...Na4+.

41...Rh2+ 42.Kc1 Rh1+ 43.Kd2 Ne4+ 44.Kd3 Nf6

Now the d5-pawn falls as well.

45.Kd2 Nxd5 46.Na5 Rh2+ 47.Kc1 Rc8 48.c4 Rxc4+

This is not at all necessary, but the four connected passed pawns will handily defeat the rook.

49.Nxc4+ bxc4 50.Ra3 Nxb4 51.Rg6 Nd3+ 52.Kd1 Kc5 53.Ra5+ Kb4 54.Ra6 c3 0-1

Game 9
Pouya Idani – Yuriy Kuzubov
Dubai Open 2015

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Be3 e5 7.Nb3 Be6 8.f3 h5 9.Qd2

In Rasmussen–Hillarp Persson, we examined the immediate 9.Nd5, combined with castling queenside. In this game
we will examine the more positional plan of castling kingside, without an early Nc3-d5.

9...Nbd7 10.a4

White’s play could equally begin with 10.Be2, which would most likely just transpose.


A somewhat rare move, although it will likely transpose to the 10...Be7 line later. It is also possible to anticipate
White’s a4-a5 by playing 10...Nb6. White can, however, use the time Black spends to create a different kind of bind as
in Lastin – Kovchan, Dagomys 2008: 11.a5 Nc4 12.Bxc4 Bxc4 13.Na4 Nd7 (Or 13...Rb8 14.Nb6 Be6 15.0-0-0!? Be7
16.Qc3 Nd7 17.Nc5!, as in Shirov – J.Polgár, Prague 1999. The point is that 17...dxc5? is met by 18.Rxd7 Bxd7
19.Rd1, when White will recapture on d7 with a total bind in return for the exchange.) 14.Nc1 g6 15.b3 Be6 16.0-0.
White plans to play c2-c4 and Ne2-c3, and Black will find difficulty getting any counterplay.

11.Be2 Be7

An important resource in such positions is 11...Nc5. However, in De la Villa García – Van Wely, Pamplona 1998,
White gained an advantage after 12.Nc1 Qb6?! 13.Nd3 Nxd3+ 14.cxd3 Qb4 15.d4! exd4 16.Bxd4 Be7 17.a5!, with the
threat of 18.Ra4 Qb3 19.Bd1. Then 17...Bb3 18.Na4! Qxd2+ 19.Kxd2 let to a better endgame for White.

12.0-0 0-0 13.a5 Rac8


In the game Grischuk – Zhang Zhong, Shanghai 2001, White won in absolutely crushing style with the subtle move
14.Kh1: 14...Rfd8 15.Nc1 (there is a hidden trick after the natural 15.Rfd1: 15...d5! 16.exd5 Nxd5 17.Nxd5 Bxd5
18.Qxd5 Nf6, when the white queen is trapped and White therefore has to give back the piece with 19.Qc4, when he is
struggling to maintain equality in the endgame) 15...h4 16.Rd1 Nc5 17.Nd5:
Position after 17.Nd5 (analysis)

17...Bxd5? (a major positional mistake, as Black will sorely miss the light-squared bishop, while the white pawns will
advance unhindered; Black had to play 17...Nxd5 18.exd5 Bf5) 18.exd5 Nh5 19.b4 Nd7 20.c4 f5 21.c5 Ndf6 22.c6 Rf8
23.b5 axb5 24.Bxb5 Bd8 25.cxb7 Qxb7 26.Bc6 Qf7 27.a6, and Black’s desperate counterattack on the kingside failed.
However, Black was more successful in Lahno – Areshchenko, Kramatorsk 2003, with 14...Rfd8 15.Nc1 Re8 (having
caused the knight to go to c1, Black now maneuvers) 16.Rd1 Bf8 17.Bf1 Red8, when White had trouble playing the
intended Nc3-d5 jump due to the pressure on c2. The waiting game lasted for a few moves before White allowed Black
the ...d6-d5 break: 18.Qf2 Be7 19.Rd2 Qc6 20.Nd3 d5, with easy play for Black.


Black is making useful moves while waiting for White to change the structure with Nd5. Black has his own way to
change the structure by 14...Nc5, although at this point White would get slightly the better chances after 15.Nxc5 dxc5
16.Qe1 c4 17.Na4.


This is White’s only active plan. Naturally, it was possible to do nothing, although this would lead to no good for


It is important to understand why Black can trade the light-squared bishop here, whereas it was fatal in Grischuk–
Zhang Zhong above. Black’s follow-up in this game, 16...Nc5, will securely blockade the queenside and keep White’s
unopposed light-squared bishop restricted.


Black can often create the structure with White’s passed d-pawn in these positions. See, for example, Walter
Browne’s system in the 6.Be2 lines (Chapter 2). The d-pawn will be securely blockaded by a bishop on d6, which will
have its own potential due to the break ...e5-e4. Black then develops insidious pressure against the kingside. The
weakening brought about by f2-f3 becomes very relevant.

17.Nxc5 dxc5 18.c4 Bd6 19.b4 Qe7 20.b5 e4 21.f4?

After this, Black’s position becomes very dynamic. White has to first play 21.bxa6 bxa6 and only then 22.f4, after
which the position is approximately balanced.
21...axb5! 22.cxb5 c4!

This move liberates the dark-squared bishop and turns Black’s passed c-pawn into a danger for White. Capturing this
pawn, on the other hand, gives the black knight access to the g4 square.


It is hardly possible to ignore the pawn. After 23.a6 c3! 24.Qa2 (24.Rxc3 loses to 24...Rxc3 25.Qxc3 Nxd5, followed
by 26...Nxe3 and 27...Bc5) 24...bxa6 25.bxa6 Bc5 and Black removes White’s dark-squared bishop, bringing the white
king into danger and winning the d5-pawn. Meanwhile, the a6-pawn is not going anywhere.

23...Rxc4 24.Bxc4 Ng4 25.Bd4

On 25.Kh1 Black has 25...Bc5, forcing the exchange of this important bishop. Then after 26.Bxc5 Qxc5 27.Qa2 (the
only move) 27...Nf2+ 28.Kg1 Qd4! there is no sufficient defense to the threat of 29...Nd3+ 30.Kh1 Nb4, as the queen
cannot protect both bishop and rook.

25...e3 26.Qe2

This allows a powerful centralization, but after 26.Qc2 Bxf4 27.g3 Be5 28.Bxe5 Qxe5 Black has a winning attack,
for example 29.Re1 h4 and the white king’s home is torn open before the white pawns can become a threat.

26...Qe4! 27.Bb6 Bxf4 28.g3

28.h3 is met by 28...Bh2+ 29.Kh1 Bd6!, with the deadly threat of 30...Qe5.

28...Nf2 0-1

Game 10
Lenier Domínguez Pérez – Alexander Morozevich
Wijk aan Zee 2009
1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Be3 e5 7.Nb3 Be6 8.Qd2 Nbd7 9.f3 b5

This natural advance is one of Black’s main systems against the English Attack. The Najdorf player must, of course,
realize that this advance has its risks as well. It is far from clear that the queenside will be Black’s domain, and the
weaknesses created (primarily on c6) by this advance could become important. One of White’s main ways of exploiting
this is by Nc3-d5, followed (after Black captures on d5 and White recaptures) by Na5-c6.


Previously, the main response to 9...b5 had been 10.a4, attacking the queenside at once. Then 10...b4 11.Nd5 Bxd5
12.exd5 Nb6 13.Bxb6 Qxb6 14.a5 Qb7 was frequently seen. Black’s queenside pawns are broken up, a6 and b4 are
weak, and Black often needs to sacrifice one of them. At the same time, White will miss the dark-squared bishop. With
opposite-colored bishops, Black can often whip up an attack involving the ...e5-e4 advance.


Morozevich prioritizes arranging the queenside pieces before kingside development. The natural alternatives are:
a) 10...Be7:
a1) 11.g4!? b4 12.Nd5 Bxd5 13.exd5 Nb6 14.Na5 Nbxd5 15.Nc4!? – for the rest of this fantastic game, see Carlsen
– So, St. Louis 2015, in the Supplemental Games.
a2) The immediate 11.Nd5 might lead, after 11...Bxd5 12.exd5 Nb6 13.Bxb6 Qxb6 14.Na5 Rc8 15.Nc6 Nxd5
16.Nxe7 Nxe7 17.Qxd6 Qxd6 18.Rxd6, to an interesting endgame where White can try to claim a microscopic
advantage due to the queenside majority and the bishop against a knight. But basically Black should have sufficient play
in the resulting position, as long as he avoids exchanging rooks.
b) 10...Nb6 is very natural: it makes room for the other knight on d7, controls d5, and aims at knight jumps to c4 or
even a4. However, Black has experienced some problems after 11.Qf2 Nfd7 (11...Nc4 can be met by 12.Bxc4 bxc4
13.Na5!, which was Bologan’s idea back in 2005. The knight cannot be captured due to 13...Qxa5 14.Bb6 Qb4 15.a3,
trapping the queen. The knight might look unstable on a5, but White is following up with Bb6 and Nd5, or even just
Rd2 and Rhd1, as Anand demonstrated in one game.) 12.f4 Rc8 13.f5 Bc4 14.Bxc4 Rxc4 15.Qd2 Qc8 16.Na5 Rc7
17.g4, when White had an edge in Svidler – Gelfand, Moscow 2004.
11.g4 Nb6 12.g5 b4!?

A novelty by Morozevich and previously played by him against Péter Lékó. However, this was never repeated after
the current game. Instead, Black has played both knight moves:
On 12...Nfd7, the f3-f4 advance has typically been White’s plan, e.g. 13.Kb1 Be7 14.f4; while after 12...Nh5 White
is slowed down on the kingside but can look to exploit the queenside weaknesses with 13.Na5! Qc7 (in the event of
13...b4 14.Nd5 Nxd5 15.exd5 Qxa5 16.dxe6 fxe6 17.Kb1, White has good compensation for the pawn thanks to his
play on the weakened light squares: Bf1-h3 is coming, as well as f3-f4) 14.Kb1 Be7 15.Nd5 Nxd5 16.exd5 Bd7 17.c4
Bd8 18.Nc6 (another pawn sacrifice to conquer the light squares, although this one is only temporary) 18...bxc4
19.Bxc4 Bxc6 20.dxc6 Qxc6 21.Rc1, when White was regaining the pawn on either d6 or a6 with advantage in Lastin –
Kokarev, Novokuznetsk 2008.


With this desperado, White wants the knight to be taken on a4 rather than on c3. This move was far better than Lékó’s
13.Nb1. After 13...Nfd7 14.Bxa6 Nc4 15.Bxc4 Rxc4 16.Qd3 Qa8 17.N1d2 Rc6 18.Kb1 Be7, Black had enormous
compensation for the pawn and went on to win in Lékó – Morozevich, Yerevan (Rapid) 2008.


With this counter-desperado, Black’s kingside remains intact. After the immediate 13...Nxa4 there follows 14.gxf6
gxf6 (14...Qxf6? would abandon the queenside: 15.Bxa6 Rb8 16.Ba7! Ra8 17.Bb5+ Bd7 18.Bxd7+ Kxd7 19.Qxb4!
Rxa7 20.Na5, and White wins back the piece with a winning attack) 15.Qxb4 Qc7 16.Rd2, and in comparison to the
game, the black king will never be able to find a secure home on the kingside.

14.fxe4 Nxa4 15.Qxb4

It is much better to remove this pawn than the a6-pawn. After instead 15.Bxa6 Rb8, Black’s position is fine. He will
castle kingside and the cramping effect provided by the b4-pawn, in conjunction with the open a-file, will give him a
strong attack. In taking the b4-pawn, White plans to invade on the queenside – and in fact most of this invasion occurs
in spectacular manner along the a-file.


A useful interposition: by threatening checkmate, Black forces White to either tie down one piece or weaken the
king’s position.


16.Qxa4+ is possible, but leads only to an equal endgame: 16...Bd7 17.Qc4 (due to the threat of checkmate, White
has to give back the piece) 17...Qxc4 18.Bxc4 Rxc4. Black has the two bishops, but the passive position of the dark-
squared one and the weaknesses on d6 and a6 make this position about equal.
16.Bd3 is not very good, since after 16...Bd7 the bishop can no longer take on a6. Black is meanwhile threatening
...d6-d5, so the white queen has to retreat. Black will complete his kingside development and enjoy good play along the
open lines on the queenside.
16.c3, avoiding having the rook tied down, is worth considering, but it allows 16...Nxc3!?, with the point 17.Qxc3
(and not 17.bxc3 d5, with a winning attack for Black) 17...Qb7 18.Nc5 dxc5 19.Qxe5, with a double-edged position.


There is a tricky possibility in 16...Nxb2!?, with the idea of 17.Kxb2 d5, with complications. However, White can
insert 17.Bb6 Qc6 18.Na5 Qa4 before taking the knight with 19.Qxb2, when Black is down a piece with only some
slight practical chances. For example 19...d5 20.Qb3 and White should be able to keep control.


Relying on the fact that the queen trade favors him, Domínguez removes the queen from the f8-bishop’s diagonal.
17.Bxa6 is possible but, after 17...d5 18.Qb7 d4 19.Qxc7 Rxc7 20.Bf2 h6!, while White is up a pawn he will have
trouble holding the position together on both sides of the board.


It is indeed possible to exchange the queens. The endgame should be better for White, but it is not so simple. For
example, 17...Qxa5 18.Nxa5 Nc5 19.Bxc5 dxc5 (19...Rxc5 20.b4 Rc3 21.Bxa6). Now the exchanging combination
20.Rxd7 Kxd7 21.Bh3+ Kc7 22.Bxc8 Kxc8 23.Nc4 would be expected to give White an advantage due to the strong
knight against the bad bishop and Black’s fractured queenside. However, the advanced g5-pawn makes this all unclear
(were the pawn on g3 instead of g5, White could certainly claim a clear advantage). Black has the prospect of ...h7-h6 at
any moment, opening the h-file in an unpleasant way. White should consider instead 20.Rg1.


This capture looks risky, but it is strong. We will soon see whose pieces get more tangled on the queenside.


Pinning the bishop with 18...Ra8 is met by 19.Bb5! Qxe4 20.Bxd7+ Kxd7 21.Qb5+ Qc6 22.Qf1, avoiding the trade
of queens. The black king will have to stay in the center, giving White the better prospects.
18...Rc7 loses immediately to 19.Bb5, e.g. 19...Qxe4 20.Qxc7 Qxh1+ 21.Rd1.

The rook occupies the traditional Najdorf weakness in a non-traditional way, plugging the long diagonal and possibly
supporting an eventual Ba6-b5.


Black has to hurry up to finally finish developing the kingside.


There can be no delay. After one more move, Black will have castled into complete safety, and White will start to
regret his queenside adventures when his pieces get tangled up and pinned.


Creating an unusual lineup of pieces on the a-file. However, this move allows White to carry out his dream of
exchanging some pieces – in particular, the light-squared bishops. The only solution for Black is to go into an exchange-
down position, after a series of complications, with 20...Bxg5+! 21.Kb1 Bd8! 22.Qd2 (22.Rxe5+!? leads to a
completely wild position: after 22...dxe5 23.Qxe5+ Be6 24.Bxb8 Qxa6 25.Qxg7 Rf8 Black has a piece for three pawns.
While it will not be easy for Black to bring the rook into the game, it is also impossible for White either to create any
kind of attack or to advance the queenside pawns – the position is wholly unclear) 22...Rxb3! (Black has to give up the
exchange in any case, so in this way he weakens White’s queenside) 23.axb3 Qxa6 24.Be3 Bc7 (Black cannot save the
knight, since 24...Nb6 25.Rxd6 wins material) 25.bxa4 Qxa4. In the resulting position, despite White’s extra exchange
and passed pawn, Black has quite reasonable compensation. Black’s light-squared bishop is particularly strong, while
White’s e4-pawn is weak and the king is on shaky ground. White might be able to gain an advantage by giving back the
exchange in the right way, but it is complicated. In any case, this was the only chance.


Forcing an eventual exchange of the light-squared bishops, which will provide White a clear route to consolidation.

Morozevich continuously refuses to capture the g5-pawn with check, reserving the possibility for later, although it
never actually happens.
If 21...Bxg5+ 22.Kb1 Qb7 23.Qxa4 Rxa7 24.Na5 Qc7 25.Rhd1 Be7 (on 25...0-0 26.Bxd7 Qxd7 27.Qxd7 Rxd7
28.Nc4, White has a big advantage in the endgame – the white pawns will be faster and better supported) 26.c4, and
White has a huge advantage. Capture of the g5-pawn was possible for Black on any of the following moves, but it
would just lead to the same position as in this note.

22.Qxa4 Rxa7 23.Na5!

23...Qc7 24.Rhd1 Bxb5

Black wants to avoid the endgame at all costs, seeking practical chances with the queens on the board in an
objectively lost position. White has an extra pawn and will have a strong bind on the light squares in addition.

25.Qxb5+ Kf8 26.Kb1 g6

By now, there was no point in capturing the g5-pawn. After 26...Bxg5 27.Nc4, White just breaks through to d6.

27.Nc4 Rb7 28.Qa4 Qb8

After 28...Kg7 29.Nxd6 Bxd6 30.Rxd6, such a position is certainly not without practical chances for Black, since it
will not be a simple matter to advance White’s passed pawns on the queenside. Nevertheless, with good technique and
calm nerves White should be able to win.
29.b3 Ra7 30.Qc6 Rc7

Black’s hopes of giving “perpetual check” to the queen prove unfounded after White’s next move.

31.Rb5! Qa7

The endgame after 31...Rxc6 32.Rxb8+ Kg7 33.Rxh8 Kxh8 is clearly hopeless for Black.

32.Qd5 Qf2 33.Qd2 Qf3 34.Nxd6 Bxg5?!

This allows White to win immediately in a nice way. However, Black’s position was lost in any case.

Now if 35...Bxd2 36.Rb8+ followed by checkmate, or if 35...gxf5 36.Rb8+ Kg7 37.Qxg5#.

35...Rc8 36.Rb8!

A final nice blow to conclude the game. After 36...Rxb8 37.Qd6+ followed by 38.Qxb8+ leads to checkmate, so
Black resigned.


Game 11
Zaven Andriasian – Ian Nepomniachtchi
Aeroflot Open, Moscow 2010

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Be3 e5 7.Nf3

The retreat of the knight to f3 rather than b3 changes nothing in the structure (at least not right away), but the choice
of this square has a dramatic effect on the course of the game. In contrast to 7.Nb3, putting the knight on f3 leads to
much quieter, more positional play, where White tries to dominate the d5 square. And why is this? Whereas 7.Nb3
allows for White to play f2-f3 with queenside castling and a kingside pawn storm, after 7.Nf3 this is not possible. White
will almost certainly castle kingside. In the meantime, b3 is left free as a retreat square for the bishop from c4.
Consequently, rather than opposite-side castling and mutual attacks, you get a more positional struggle.


The main alternative is 7...Qc7, preventing the development of the bishop to c4. This, however, costs Black in time
and flexibility. White will normally change his plan to prepare Nc3-d5 and the recapture with the e-pawn followed by a
general pawn advance. For example, 7...Qc7 8.a4 (White prepares a4-a5, locking down the b6 square, while 8...b6
would be well met by 9.Nd5 Nxd5 10.Qxd5 Bb7 11.Qb3, and if 11...Nbd7 then 12.Bc4) 8...Be7 9.a5, followed by Bf1-
e2 and 0-0. Black is not without his chances in this line – one possibility is to prepare to advance the b-pawn,
recapturing after axb6 with ...Nxb6, when Black has a weak a-pawn but some dynamic play on the queenside thanks to
the open b- and c- files and the possibility of ...Nb6-c4. In some ways this line is more double-edged than the one in the
main game, since the structure is likely to become unbalanced when White plays Nd5 and exd5, rather than a battle over
d5 resulting in many piece exchanges.
7...Be6 looks like a logical way to both develop and prevent White’s bishop from taking the a2-g8 diagonal. The
problem with this, however, is 8.Ng5, when White is ready to capture the bishop and exert pressure on the light squares,
either by Bf1-c4 or g2-g3 and Bf1-h3.


The bishop takes over this nice diagonal, where it influences d5. Yet Black will gain some time attacking the bishop,
and perhaps eventually exchange it.


If Black wants to play an early ...Bc8-e6, it is very important that he do so after 8...0-0 9.0-0. On the immediate
8...Be6?! White should not let Black get away with his inaccuracy, but instead play 9.Bxe6! fxe6 10.Ng5 Qd7 11.Qf3,
intending Qh3. A famous game Ye Jiangchuan – Shirov, Yerevan 1996, continued 11...h6?! (Better is 11...d5, after
which many games have gone 12.exd5 exd5 13.0-0-0 d4 14.Nce4, when White has a pleasant positional advantage due
to his occupation of the central light squares. However, even stronger is a tactical idea: 12.0-0-0 d4 13.Bxd4! exd4
14.e5, when White will inevitably regain the piece with a better position.) 12.Qh3 Nc6, when capturing the pawn on e6
loses material, but 13.Na4! instead, threatening to enter the b6 square, puts Black in a desperate situation. Ye
Jiangchuan actually played 13.0-0-0? and Shirov was able to take over the game with 13...Rc8, and now 14.Na4 will be
met by 14...Nb4, while otherwise the white pieces on g5 and h3 will soon find themselves poorly placed.

9.0-0 Nc6

The unofficial main line is 9...Be6 10.Bb3 Nc6 11.Bg5 (White aims to remove a defender of d5) 11...Nd7 12.Bxe7
Qxe7 13.Nd5 Qd8, when Black is very solid, but technicians could get excited over White’s continuing control of d5.
There is also a large variety of other ways for Black to play, such as 9...Qc7 combined with the delay of ...Nb8-c6 (so
that the knight may later go to d7).

White completes his development and makes room for a rook on the d-file. More often seen is 10.Bg5, which might
lead to the main line of the previous note after 10...Be6 11.Bb3 Nd7 12.Bxe7 Qxe7 13.Nd5 Qd8. There is nothing
wrong with this position at all. But for those who like to take a less-beaten path, I can recommend to study the games of
Ian Nepomniachtchi, the victor in the main game. This Najdorf adherent frequently uses obscure and experimental
ideas. For example, see his game against Geetha Gopal, Moscow 2013: 10...Nd7!? (it looks unusual – and yet it is
nevertheless playable – to make this move before ...Bc8-e6) 11.Be3 Na5 12.Bb3 Nf6 13.Bg5 Nd7 14.Qd2 Bxg5
15.Nxg5 Nb6, where Black played slightly provocatively, sending both knights to the queenside to occupy c4.


The capture of the bishop will now come with an immediate threat: 11...Nxd5 12.Nxd5, threatening Be3-b6. The
obvious alternative is 11.Bb3, which is normally played in the 7.Nf3 variation. White allows the trade of the light-
squared bishop, but the recapture axb3 (or in some cases even cxb3, opening the c-file) will give White a stronger
queenside pawn structure. Control over c4 is particularly important. But the specifics of how Black has played – the
delay of ...Bc8-e6 in favor of the immediate hunting of the bishop – seem to give Black good prospects. Very thematic
was 11...h6! (preventing Be3-g5) 12.Rfd1 Nxb3 13.axb3 Qe8! 14.h3 Be6 15.Nd5 Nxd5 16.exd5 Bd7 17.Nd2 f5, and
without any difficulty at all, Black secured a promising game against a strong opponent in Smirin – Wolff, Biel 1993.
The two bishops and mobile kingside pawns give Black the better prospects.
11.Bd3!?, keeping the bishop, looks slightly passive but has some concrete ideas. After the natural 11...Be6, White
intends 12.Na4, gaining time for c2-c4, heading to a position similar to the Kalashnikov. For example, 12...Nd7 13.c4
Nc6 14.Rad1 Nc5 15.Nc3 was the course of A.Kovačević – D.Šarenac, Kragujevac 2013. White’s position is slightly
better, and moreover this type of position might not appeal to Najdorf players very much. Instead of 11...Be6, however,
it was possible to consider an interesting pawn sacrifice: 11...b5 12.b4 (this is White’s idea; otherwise his play makes
little sense) 12...Nc4! (12...Nc6 13.a4 gives White a strong queenside initiative. Instead, Black sacrifices a pawn to gain
the two bishops and expose White’s weaknesses.) 13.Bxc4 bxc4 14.Qxc4 Bb7, and Black had compensation for the
pawn in K.Le Quang – Van der Stricht, Charleroi 2014. The compensation is long-term and purely positional in nature –
Black’s strong and unopposed light-squared bishop keeps constant pressure against the pawn on e4, while the b2-b4
advance has seriously weakened White’s queenside.


Black does not capture the bishop yet. Instead, he leaves it on d5 where it is not the most effective piece to occupy the
outpost, and simply goes about completing his development.

12.Rfd1 Rc8 13.h3

This was not really necessary. In the game Erenburg – Ma.Tseitlin, Israeli Team Chp 2012, White played 13.a3
immediately, creating a retreat for the bishop. After 13...b5 14.Nd2 Be6 15.Ba2 Bxa2 16.Rxa2 h6 17.Nf1 Rxc3!
18.bxc3 Nxe4 19.Qd3 d5, Black had carried out the thematic Sicilian exchange sacrifice and enjoyed sufficient

13...b5 14.a3 h6 15.Nh2?!

White intends to use this knight in the fight for d5 by playing Ng4 and removing one of the black pieces covering that
square. However, this leads to no good. It was better to play as in Erenburg’s game above with 15.Nd2, although White
has no advantage.

15...Nc4 16.Bxc4 Rxc4

In the end, Black has removed White’s light-squared bishop, as normally happens in this line. However, this capture
usually occurs on b3, allowing axb3. In this game White has not been given the opportunity to strengthen the queenside
in this way, and Black can now freely put pressure on the c-file and against e4, while White can only defend passively.
A very successful outcome of the opening from Black’s point of view.

17.f3 Be6 18.Nf1 Qc7 19.Bf2


Without any further ado, Black opens lines on the queenside. Otherwise White would have the chance to play Nf1-e3.

20.axb4 Rxb4

The a-pawn is very weak, but so is White’s b2-pawn. When that pawn is removed, Black will have strong pressure
down the c-file and will be able to slowly turn his attention to the more populated kingside.


Not the move you would like to play, but after 21.Qxa6 (of course not 21.Rxa6 Bc4) 21...Rxb2, the c2-pawn is lost as
well. And after 22.Nb5 Qxc2 23.Rd2 Qb3 24.Nxd6, White maintains material equality, but his pieces continue to be
awkwardly tied up and out of play. After, for example, 24...Rxd2 25.Nxd2 Qc3 26.N2c4 Rd8, it is very hard for White
to defend.

21...Rfb8 22.Qxa6 Rxb2 23.Rxb2 Rxb2 24.Qd3

Despite the reduction of pawns on the queenside, Black keeps strong pressure. The pawn on c2 is far weaker than the
one on d6, and Black’s pieces are more active.


The threat of bringing the knight to f4 induces a serious weakness in g2-g3.

25.g3 Bc4

Black begins forcing play.

26.Qd2 Bg5 27.Qxd6 Qc8!

Black could play 27...Qxd6 28.Rxd6 Rxc2, when his pieces are more active and he keeps some pressure, but with all
the pawns on the kingside and the queens off the board, White has good prospects of defense. Instead, Nepomniachtchi
retains the queens and takes aim at the white king.

28.h4 Bxf1


This allows a nice breakthrough. One alternative, 29.hxg5, would have lost after 29...Qh3 30.Rxf1 Nxg3 31.Bxg3
Qxg3+ 32.Kh1 Rxc2, when there would be no defense.
Instead, there was a very difficult route to salvation: 29.Qxe5! would defend the kingside and leave two black pieces
hanging, while protecting the knight on c3. After 29...Qh3 (on 29...Bf6 30.Qxh5 Qxc3 31.Rxf1, White is fine) 30.Rxf1
Bf4!, 31.Qf5! is the only way to defend, but a sufficient one. (Not 31.gxf4 Ng3 32.Bxg3 Qxg3+ 33.Kh1 Rxc2, and
Black checkmates.) Then, after 31...Qxf5 32.exf5 Nxg3 33.Bxg3 Bxg3 34.Ne4 Bxh4 35.Rc1, the game should be


Black breaks through to this key point. This is not really a sacrifice, but rather a desperado, since the c3-knight is


30.hxg5 is met by 30...Qh3.

30...Be3+ 31.Bf2 Qxc3

If White loses the c2-pawn, then Black’s 4-3 majority on the kingside, with the queens on the board and the white
king exposed, will be enough to win. Therefore White tries to defend with his next move, which however allows a
winning pawn endgame for Black.


Black has to finely calculate the pawn endgame to make these trades.

33.cxd3 Rxf2! 34.Rxf2 Kh7

The material in the pawn ending will be equal, but Black’s advantage consists of the weakness of the h4-pawn and the
fact that the e5-pawn is holding back the white pawns.

The alternative was to try to delay the black king’s entry into the game with 35.h5. This also had to be carefully
calculated before Black went into the endgame. For instance, 35...g6 36.Kg2 Bxf2 37.Kxf2 gxh5 38.f4 f6! 39.Kg3
Kg6, and now after 40.Kf3 (or 40.fxe5 fxe5 41.Kh4 Kf6 42.Kxh5 Kg7 43.Kg4 Kg6 and Black wins due to the outside
passed pawn, while White cannot create his own passed pawn), 40...h4! is the only way: 41.Kg4 h5+! 42.Kxh4 exf4
43.d4 Kf7 and Black wins by bringing the king to c4, forcing d4-d5, then after ...Kc5 and ...f6-f5 undermining and
winning the white pawns. In the meantime, the white king cannot attack the black pawns and can only wait.

35...Bxf2 36.Kxf2 Kg6


Black’s win also comes down to the wire after 37.f4. The only route to victory then is 37...exf4 38.Kf3 Kh5 39.Kxf4
Kxh4 40.d4 g5+ 41.Kf5 (or 41.Ke3 g4 42.d5 Kg5 and Black catches up to the d-pawn and will later be able to
undermine it with the f-pawn) 41...g4 42.d5 g3 43.d6 g2 44.d7 g1=Q 45.d8=Q+ Qg5+ and Black forces the exchange of
queens, winning.

37...Kh5 38.Kh3 g5

Now Black creates an outside passed pawn which will divert the white king.

39.hxg5 Kxg5 40.Kg3 h5 41.f4+

Otherwise, after king moves the black king will enter f4 and the h-pawn’s advance will divert the white king,
allowing Black to capture the white pawns.

41...exf4+ 42.Kg2 h4 43.d4 Kg4 44.d5 h3+ 45.Kf2 f3

Since the black h- and f-pawns are self-supporting, Black could also win by bringing the king back to stop the d-pawn
and undermine it by ...f7-f5. However, Nepomniachtchi’s method is quicker and nicer, allowing the white pawn to
queen and then forcing the win of the new queen.
46.d6 h2 47.d7 h1=Q 48.d8=Q Qg2+ 49.Ke3 Qe2+ 50.Kd4 Qd2+ 51.Ke5!

A last try – but a world-class player will not casually capture the queen on d8, with stalemate.


This instead forces the trade of queens again, and Black’s f-pawn then queens, so White resigned.


Supplemental Games

Alexander Morozevich – Veselin Topalov

Monte Carlo (rapid) 2004
1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Be3 e5 7.Nb3 Be6 8.Qd2 b5 9.0-0-0 Nbd7 10.f4 Be7 11.Kb1
Rc8 12.f5 Bc4 13.g4 Nxg4 14.Rg1 Nxe3 15.Qxe3 g6 16.Bxc4 bxc4 17.Nd2 Nf6 18.Qh3 Qd7 19.Nf1 gxf5 20.Ne3
Nxe4 21.Ncd5 Nf2 22.Qh5 Nxd1 23.Rxd1 Rc5 24.Nxe7 Qxe7 25.Nxf5 e4 26.Ng7+ Kf8 27.Qh6 Qe5 28.Nf5+ Ke8
29.Nxd6+ Ke7 30.Nf5+ Ke8 31.Qg5 Qc7 32.Ng7+ 1-0

Viswanathan Anand – Erwin L’Ami

Wijk aan Zee 2013
1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Be3 e5 7.Nb3 Be6 8.h3 Be7 9.Qf3 0-0 10.0-0-0 Qc7 11.g4 Rc8
12.Kb1 a5 13.a3 a4 14.Nc1 b5 15.N1a2 Bxa2+ 16.Nxa2 Qxc2+ 17.Ka1 Nc6 18.Rc1 Qxe4 19.Qxe4 Nxe4 20.Bxb5
Nc5 21.Bxc5 dxc5 22.Rc4 Na5 23.Rxa4 Rcb8 24.Bc4 e4 25.Bd5 Bf6 26.Nc3 Ra6 27.Rc1 c4 28.Ka2 Nb3 29.Rc2 Nc1+
30.Kb1 Rxa4 31.Nxa4 Nd3 32.Bxc4 h5 33.Bxd3 exd3 34.Rd2 hxg4 35.hxg4 Re8 36.Rxd3 Re4 37.Nc3 Rxg4 38.b4 g5
39.Nd5 Be5 40.Kc2 Kf8 41.a4 Rc4+ 42.Kb3 Rd4 43.Rxd4 Bxd4 44.f3 f5 45.Kc4 Bf2 46.Kd3 g4 47.Ke2 Bd4 48.Ne3
g3 49.f4 Kf7 50.Kf3 Ke6 51.b5 Bb6 52.Nc4 Bc5 53.a5 Bd4 54.a6 1-0

Dejan Pikula – Branko Damljanović

Vrnjačka Banja 2012
1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Be3 e5 7.Nb3 Be6 8.f4 exf4 9.Bxf4 Nc6 10.Qe2 Be7 11.h3 Qc7
12.0-0-0 0-0 13.g4 Rfe8 14.g5 Nd7 15.h4 b5 16.Qd2 Nce5 17.h5 Bf8 18.Nd5 Bxd5 19.exd5 Nc4 20.Qh2 Ne3 21.Bxe3
Rxe3 22.Qf4 Rae8 23.Bd3 R3e7 24.g6 Ne5 25.Bf5 Kh8 26.gxh7 Qc4 27.Qg3 Qa4 28.Kb1 Nc4 29.Bd3 Ne5 30.Rh4 b4
31.Be2 Nd7 32.Bd3 Ne5 33.Bf1 Rc7 34.Re4 Rec8 35.Bh3 Re8 36.Nd4 Rc5 37.Nb3 Rc7 38.Nd4 Rc5 39.Qb3 Qxb3
40.Nxb3 Rc4 41.Nd4 Kxh7 42.Bf1 Rxd4 43.Rexd4 a5 44.Be2 Be7 45.Rf4 Bg5 46.Rf2 Kh6 47.a3 bxa3 48.bxa3 Be3
49.Rf5 Bc5 50.Ka2 f6 51.Bb5 Rb8 52.c4 Rb7 53.Rd2 Nf7 54.Re2 Ne5 55.Re4 Rc7 56.Kb3 Rc8 57.Re2 Rc7 58.Re4 Rc8
59.Re2 Rc7 ½-½

Fabiano Caruana – Boris Gelfand

Wijk aan Zee 2014
1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.f3 e5 7.Nb3 Be6 8.Be3 h5 9.Nd5 Bxd5 10.exd5 Nbd7 11.Qd2
g6 12.Be2 Bg7 13.0-0 0-0 14.Rac1 b6 15.h3 Re8 16.g4 hxg4 17.hxg4 Nh7 18.g5 f5 19.gxf6 Bxf6 20.Rf2 Bg5 21.Rg2
Bxe3+ 22.Qxe3 Ndf8 23.Bd3 Ra7 24.Rf1 Rf7 25.Qh6 Kh8 26.Nd2 Rf4 27.Rg4 b5 28.Ne4 Nd7 29.Rxg6 Rg8 30.Ng5

Magnus Carlsen – Wesley So

St. Louis 2015
1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Be3 e5 7.Nb3 Be6 8.f3 Nbd7 9.Qd2 b5 10.0-0-0 Be7 11.g4 b4
12.Nd5 Bxd5 13.exd5 Nb6 14.Na5 Nbxd5 15.Nc4 Nxe3 16.Nxe3 0-0 17.Bc4 Nd7 18.h4 a5 19.g5 Rc8 20.Bd5 Nb6
21.Kb1 Qc7 22.Rhf1 Nxd5 23.Nxd5 Qb7 24.f4 f5 25.Qe3 e4 26.h5 Rc5 27.h6 g6 28.Qb3 Rf7 29.a4 Bd8 30.Rd4 Kf8
31.Rfd1 Rc6 32.Ne3 Bb6 33.Nc4 Bxd4 34.Nxa5 Qb6 35.Nxc6 Bc5 36.Qd5 e3 37.a5 Qb5 38.Nd8 Ra7 39.Ne6+ Ke8
40.Nd4 Qxa5 41.Qg8+ Kd7 42.Qxh7+ Kc8 43.Qg8+ Kb7 44.c3 bxc3 45.Qb3+ Qb6 46.Qxb6+ Kxb6 47.bxc3 Bxd4
48.Rxd4 Kc6 49.Kc2 Ra2+ 50.Kd1 Rf2 51.Ke1 Kd7 52.Ra4 Ke6 53.Ra8 Rh2 54.c4 Kf7 55.Rb8 Ke6 56.Rg8 1-0
Chapter 4
In Morphy’s Style: 6.Bc4

The Najdorf didn’t exist when Paul Morphy was playing chess, but we have to surmise that if it had, he would have
played 6.Bc4. Indeed, this was the move usually chosen by his (possible) reincarnation, Bobby Fischer.
With 6.Bc4 White prepares to castle quickly while developing the bishop to its famous “Italian” diagonal, aiming at
f7. However, unlike in the open games, a bishop on c4 here can be stymied by ...e7-e6, which is how Black normally
replies immediately.
The outcome hinges on White’s attempts to reactivate the bishop, either by f4-f5 or, more radically, by a piece
sacrifice on e6, d5, or f5. White may also attempt to breach the center by e4-e5, making use of his lead in development,
or begin a kingside pawn storm by f2-f4, Qd1-f3, and g2-g4-g5.
Black’s counterchances lie in his pressure against e4 and the opportunity to gain the two bishops. Often White finds
himself fighting against the positional tide, as Black’s small but strong center and open c-file give him a long-term
In Game 12, we see a nice counterattacking game by the sadly departed Aleksander Wojtkiewicz. White’s system
with 9.f4 is not particularly common nowadays, but it is logical and aggressive – White seeks to carry out f4-f5 without
delay, and does not shrink from sacrificing the e-pawn. A wild position develops, with many pieces hanging on both
sides. Black wends his way through the complications, capturing just the right pieces while leaving the poisonous ones
alone. Finally, a counterattack comes.
In Game 13, we move on to Black’s alternative 7...Nbd7. Here White combines rapid development and queenside
castling with the f4-f5 advance – a very aggressive setup. When White begins some unjustified tactics, Black quickly
cuts him down with a swift counterattack, showing the hidden resources of the second player’s position.
Game 14 was a crazy draw from the 1993 World Championship match between Nigel Short and Garry Kasparov. In
this game, Short goes for the throat with the hyper-aggressive line 7...Nbd7 8.f4 Nc5 9.e5, and nearly wins a game for
the ages. However, time pressure causes him to miss a win, and the game ends in a draw.
Game 15 covers another sacrificial possibility for White, in the line 7...Nbd7 8.0-0 Nc5 9.f4, a variation made famous
by the spectacular game Topalov – Kasparov, Amsterdam 1996. Despite being known for two decades, this variation is
still unresolved. Black defends well and obtains a reasonable position; however, a slight stumble allows White to renew
the attack.

Game 12
Franciszek Borkowski – Aleksander Wojtkiewicz
Polish Championship, Słupsk 1989

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Bc4 e6

Black almost always blunts the white bishop in this way. One of the primary points of the 6.Bc4 line is to prevent the
typical Najdorf push ...e7-e5, which here would leave White well placed to dominate the d5 square.


This flexible anticipatory retreat is now by far the most common continuation, to the point that other moves such as
7.0-0 are considered offbeat. Nevertheless, these less-popular systems are worthy of interest.
7.a4 seeks to prevent ...b7-b5 and also provides a more distant retreat for the bishop on a2. On the other hand, this
move weakens the b4 square. It often becomes profitable for Black to hit back in the center with ...d6-d5. See the
supplemental game Geller – Belov, Sochi 2006.
7.0-0 is natural and might be followed up in various ways, including a later a2-a4. 7...b5 8.Bb3 transposes to the main
line. For those who prefer to avoid the 7...b5 line (perhaps preferring 7...Nbd7), then the logical move is 7...Be7. Black
prepares to castle and awaits White’s future play. The eventual retreat Bc4-b3 would allow Black to revert to ...Nbd7-
c5. Unlike after 7.Bb3 Be7, here 8.g4 is obviously not a good idea.


This is the most common (and probably also the safest) move. For the sharp 7...Nbd7, hunting down the bishop and
also pressuring e4, see Game 14.
7...Be7 is very natural, but actually fairly uncommon. The reason is that the absence of direct pressure on the white
center allows White to play 8.g4, which gains space and foreshadows a kingside attack. Indeed, achieving g4-g5 and 0-
0-0 is about the most aggressive setup White could have in these Sozin-type setups. Nevertheless, it is a provocative
way to play as Black, providing winning chances and perhaps you could say “more known risks” than the 7...Nbd7 line.
For example: 8...h6 (Black should certainly play this move) 9.Be3 b5 10.f3 Qa5!? 11.Qd2 Nfd7 12.Bxe6? (12.Nf5)
12...fxe6 13.Nxe6 Ne5 14.Nxg7+ Kf7–+, Dragnev – Van Kampen, Munich 2016.


The most aggressive setup is 8.Bg5 but, as so often against the Najdorf, simple rapid development and direct
aggression fail to yield dividends. Black does not appear to have big problems, although he should certainly have some
notion of what he is doing here. We will see this line in the game Ivanchuk–Karjakin.

8...Be7 9.f4!?

This is an old variation which is mostly abandoned today in favor of 9.Qf3. White’s idea, though, makes sense
strategically as he is trying to open up the diagonal for the Sozin bishop on b3 by f4-f5, or perhaps to open the center in
a different way with e4-e5. At the same time, e4 is seriously weakened.
9.Qf3 is by far the main line. Now Black has 9...Qc7 and 9...Qb6 with the idea of 10.Be3 Qb7. Both of Black’s
approaches are solid. For an example of 9...Qc7, see the supplemental game Blomqvist – Anand, Gibraltar 2016.
On the other hand, it is very important to be aware of a certain theme, known since the 1930s. After 9...Bb7?, White
will sacrifice a piece on e6: 10.Bxe6 fxe6 11.Nxe6 Qd7 12.Nxg7+. With three pawns for the piece and a persistent
attack, White stands much better. Sacrifices on e6 are abundant in the Najdorf, and the players must develop good
judgment about the resulting positions. Here the sacrifice is very promising since White has managed to capture g7,
obtaining a third pawn while Black is lacking in the counterplay which might exist if the white king were, for example,
castled queenside.


Black immediately pressures e4, although the capture of this pawn is still too dangerous. The alternative, 9...0-0, is
also okay. White can now play:
a) 10.f5 b4! (Black should chase the knight from c3, reducing White’s control over d5) 11.Na4 e5 12.Ne2 Bb7
13.Ng3 Nbd7! and Black has no problems.
b) 10.Be3 b4 (again, chasing the knight from its position of influence over e4 and d5 is a good idea) 11.Na4 Bb7
(however, Black should not be in a hurry to capture the e-pawn: after 11...Nxe4?! 12.f5 d5 13.fxe6 fxe6 14.Rxf8+ Bxf8
15.Qf3, White’s pieces are very active and give him a dangerous attack). Now 12.e5 dxe5 13.Nxe6!? (13.fxe5 Nd5 is
very good for Black) 13...fxe6 14.Bxe6+ Kh8 15.Qxd8 Rxd8 16.fxe5 Ng8 17.Nb6 Ra7 18.Rad1 Nc6 19.Rxd8 Nxd8
resulted in an unclear position in Kaidanov – De Firmian, Las Vegas 1994.
c) 10.e5 is a fairly dangerous attacking line, where White normally sacrifices a pawn. But with accurate play, Black
can neutralize it, for example 10...dxe5 11.fxe5 Nfd7 12.Be3 (12.Qh5 Nc6! 13.Nxc6 Qb6+ 14.Be3 Qxc6 15.Kh1 Bb7
and Black has good counterplay, Gheorghiu – Polugaevsky, Petrópolis 1973) 12...Nxe5 13.Qh5 Nbc6 14.Nxc6 Nxc6
15.Rf3 b4 (15...Qd6!?, Kulaots – Nisipeanu, Medellín 1996) 16.Rh3 h6 17.Rd1 Qa5 18.Nd5 exd5 19.Rg3 d4 20.Bd5
Bg5 21.Bxg5 Qxd5 22.Bf6 Qxh5 23.Rxg7+ Kh8 24.Rg6+ Kh7 25.Rg7+ ½-½ Short – Kasparov, Novgorod 1997.


In this line, White can also play 10.e5, although Black is well placed for a counterattack. 10...dxe5 11.fxe5 Bc5!
12.Be3 Nc6 13.exf6 Bxd4 14.fxg7 (14.Qe1 is another line which has been played in the past. A sharp answer is
14...gxf6!?, when Black looks for counterplay on the g-file.) 14...Bxe3+ 15.Kh1 Rg8 16.Qf3 Rxg7 17.Bxe6:

Position after 17.Bxe6 (analysis)

17...Qe7! has been played a number of times. White should not have enough compensation; for example, Inkiov –
Ribli, Plovdiv 1983, went 18.Bd5 (18.Nd5 is also met by 18...Nd4) 18...Nd4 19.Qh3 Bd2 20.Bxb7 Qxb7, when the
pressure against g2 made it impossible for White to create any attack to compensate for the missing piece.

10...b4 11.e5
A reverse of the typical situation in the Sicilian, where e4-e5 is met by ...b5-b4, fighting for d5 and e4. Alternatively,
White can play 11.Na4 Bxe4 12.c3 (12.f5 e5 and Black will successfully defend and keep the pawn), and now 12...a5!.
Black keeps the knight sidelined on a4 and White does not have enough compensation. It would be a serious mistake to
cede the initiative by 12...bxc3 13.Nxc3 Bb7 14.f5, when White’s attack is very strong.

11...bxc3 12.exf6 Bxf6 13.Ba4+

13.bxc3 has been played, although Black has the better pawn structure and doesn’t face any real attack. 13.f5!? is
perhaps the best attempt, trying to transpose to the next note after 13...e5 14.Ba4+ Nd7 15.Ne6!. Black, though, should
prefer 14...Ke7! 15.Ne2 cxb2 16.Rb1 Qc7 17.Rxb2 Nd7, when White has some compensation, but Black’s strong
center should be more important than the temporarily displaced king.

13...Nd7 14.f5

Black avoids 14...e5 15.Ne6! fxe6 16.fxe6. Instead, Wojtkiewicz removes the king into safety immediately and will
activate the pieces and gain a strong passed pawn on b2.

15.fxe6 Nc5 16.Nc6

White intends to sacrifice on f6. After the alternative 16.exf7+ Rxf7 17.Bb3 Nxb3 18.axb3 cxb2 19.Rb1 Qe7, White
will not be able to deal with the strong pressure exerted by the black bishops, while the b2-pawn is still alive.


The position looks very wild after 17.Bxc5 cxb2 18.Rb1 dxc5 19.Qd7, but with 19...Qb6! Black can count on the
pawn on b2. For instance, 20.e7 c4+ 21.Kh1 Rfe8 22.Rxf6 gxf6 23.Qg4+ Kh8 24.Qf4 Qc5! 25.Qxf6+ Kg8 and White
has no more checks, while his pieces are hanging in air.

17...cxb2 18.Rb1 Nxa4

Black of course will not take the rook. If 18...gxf6? 19.e7 Rfe8 20.Qg4+ Kh8 21.Bd4 Nd7 22.Nd8!, White removes
the d7-knight and wins.



Wojtkiewicz finds his way through the complications. Not 19...Rfe8? 20.Rxf7! Kxf7 (or 20...Qxc6 21.Rxg7+! Kxg7
22.Qg4+ Kf6 23.Rf1+, and Black must give up the queen to avoid getting checkmated) 21.Qd5+ Kg6 22.Rf1, followed
by checkmate.

20.exf8=Q+ Rxf8

The clouds begin to clear. Black is down the exchange, but the pawn on b2 will cost White material. Additionally,
mate is threatened on g2 while the f6-rook is also under attack.


Black wins after 21.Rf2 Nc3 22.Qe1 Nxb1 23.Qxb1 Qc3.

21...h5 22.Qg3 Nc3

Black threatens 23...Ne2+ as well as the rook on b1.

23.Rf2 h4

Avoiding any problems associated with 23...Nxb1 24.Bd4.

24.Qg4 Bc8 25.Qd4 Nxb1 26.Qxb2 Nc3

Black is up a piece and the game quickly comes to an end.

27.Bd4 Nd1 28.Qa1 Nxf2 29.Bxg7 Bb7 0-1

Game 13
Dmitry Chuprov – Peter Svidler
Smolensk 2000

1.e4 c5 2.Nf3 d6 3.d4 Nf6 4.Nc3 cxd4 5.Nxd4 a6 6.Bc4 e6 7.Bb3 Nbd7
A sharper and slightly less common approach compared to 7...b5. Black sends the knight to c5, where it attacks e4
and can remove the b3-bishop at the right moment.


For the main alternative 8.Bg5, see the supplemental game Abergel – Vachier-Lagrave.
It’s important to note that unlike after 7...Be7, 8.g4 does not work well here, since ...Nbd7-c5 puts pressure on
White’s center. After 8...Nc5 9.Qe2 (this position has occurred sometimes in the move order 8.Qe2 Nc5 9.g4?!) 9...b5
10.g5, and now Black has the typical counterblow 10...b4! (10...Nfd7 is also acceptable, but check out the spectacular
game Tate – Yudasin in the supplemental games). Now after 11.gxf6 bxc3 12.fxg7 Bxg7 13.bxc3 Qh4 Black stood very
well in V.Zaitsev – Semakin, Voronezh 1991.

8...Nc5 9.Qf3

A natural but rather older move. For the critical alternatives 9.0-0 and 9.e5, see below Amin–Areshchenko and Short–
Kasparov, respectively.


One of the principles of Black’s play in this line is that the pawn goes to b5 before ...Bf8-e7 is played. Thus Black can
play ...Bc8-b7 (or, as happened in this game, ...Bc8-d7) and create immediate pressure on e4 without having to spend a
tempo preventing Nd4-c6.


White creates immediate pressure on e6, since otherwise ...Bb7 comes and Black will be attacking e4. The push f4-f5
attempts to make the b3-bishop relevant. The question of this bishop tends to define the outcome in the Sozin.


Black keeps e6 defended while preparing ...Be7 (preventing Nd4-c6). Closing the center with 10...e5 gives White the
d5 square and opens the a2-g8 diagonal for the white bishop. Nevertheless, this move is typical for this variation.
However, after 10...e5 White has the strong 11.Nc6!, immediately bringing the knight into contact with d5 after
11...Qd7 12.Nb4, when White is following up with Bg5 and the occupation of d5. Meanwhile, Black should be careful
about playing 10...b4, since after 11.Na4 his position is seriously loosened, and meanwhile he cannot capture the e4-

11.Bg5 Be7 12.0-0-0 0-0

A very sharp position is reached, where Black should have adequate play. White now tries to force the issue –
illogically, it would seem, since Black has made no mistake.


It was better to preface this by 13.fxe6 fxe6, but then 14.e5 Nd5 15.Bxe7 Nxe7 16.Qe3 d5 locked down the bishop on
b3 and gave Black an excellent game in Doghri – Dao Thien Hai, Budapest 1996.
13.a3 was possible, but Black would nevertheless prepare to open lines with ...a6-a5 and ...b5-b4, with great attacking
Meanwhile, 13.Rhe1 could be met thematically by 13...b4 14.Nb1 e5 15.Bxf6 Bxf6 16.Ne2 Bc6, with a tremendous
game for Black.

13...dxe5 14.Nc6 Bxc6 15.Qxc6

Trading queens by 15.Rxd8 Bxf3 16.Rxf8+ Rxf8 17.gxf3 exf5 would leave Black up two pawns.

This is even stronger than 15...Rc8, which gives White the option to enter the endgame by 16.Rxd8 (16.Qf3 Qa5 leads
back to the games) 16...Rxc6 17.Rxf8+ Kxf8 18.fxe6, when White will be down a pawn, but with at least some fighting

16.fxe6 Rac8 17.exf7+ Kh8

So White has achieved the goal of f4-f5xe6. But that is it. And now Black will have free rein to attack on the
queenside. As we will see, White has an additional tactical problem in the exposed bishop on g5.

18.Qf3 Nxb3+ 19.axb3 b4!

This move is always hanging over White’s head in the Najdorf. Now not only does Black drive away the defending
knight, but he also clears the fifth rank.

There is no time to insert 20.Bxf6 since 20...bxc3 threatens checkmate on a1. Trying to block the e-pawn by 20.Ne4
also fails after 20...Nxe4 21.Bxe7 Qa1#.


It was not obvious when Black played 15...Qa5, but there were secret designs on the g5-bishop, which now become


Other queen moves would be met by different discoveries, i.e. 21.Qf4 Nh5 or 21.Qe3 Ng4.


And just like that, Black wins material.

22.Bxe7 Nxg3 23.Bxf8 Ne2+

This is the most accurate way.


24.Kb1 Rxf8 leaves Black up a queen for a rook.

24...Qg5+! 0-1

25.Kxe2 is now met by 25...Rxc2+, with checkmate in a few moves, while 25.Ke1 Qe3 also wins for Black, so White
resigned. Comically, the game Kozlitin – Pavlidis, Pardubice 2014, followed exactly this course (presumably not
intentionally, at least by White), with the exception that Black used the move order 15...Rc8 16.Qf3 Qa5 and White
only resigned after 25...Qe3.
Game 14
Nigel Short – Garry Kasparov
PCA World Chp (8), London 1993

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Bc4 e6 7.Bb3 Nbd7 8.f4 Nc5 9.e5

The fastest way to begin an attack. On the other hand, unlike after f4-f5 the b3-bishop still faces a defended pawn on
e6 – barring a piece sacrifice.


The sharpest move. A solid option is 9...Nfd7 10.exd6 Nf6!, when Black has lost two tempi, but on the other hand the
e-pawn has been exchanged. After 11.Be3 Bxd6 12.Qf3, White plans 0-0-0 and g2-g4. Black must be accurate, but
should not face too many problems, for example 12...0-0 13.0-0-0 Qc7 14.Kb1 Bd7 15.g4 Nxb3 16.axb3 Rfe8! (an
important idea: Black prepares to meet g4-g5 with ...Nf6-d5, when the rook will be aimed at e3 in the event of Nxd5
exd5) 17.Rhg1 Rac8 18.g5 Nd5 19.Nxd5 exd5, and Black had counterplay in the center and on the queenside in Nukin –
Kokarev, Tomsk 2012.

10.fxe5 Nfd7 11.Bf4 b5

In this way, Black places the bishop on its best diagonal (the long diagonal) before White can prevent it by Qd1-f3.
Such a position might look good for White on the surface – the e5-pawn confers some space advantage and White has
rapid development, plus the f-file is open and the white pieces are placed in threatening-looking positions. But such is
the poison of the Sicilian. Black too has his advantages, and they tend to be more long-lasting. The bishop which will
come to b7 will be very well placed. The advanced e5-pawn is not only a strength, but a weakness. And most
importantly, Black has a well placed knight on c5 and a substantial advantage in space on the queenside – the advance
...b5-b4 is constantly looming over White, and the b3-bishop, if not activated in some dramatic fashion, could turn out to
be a complete dud.

A logical, aggressive development. The queen is posted actively, aiming at both e6 and g7, without being exposed to
attack by ...g7-g6 (as it would be after 12.Qh5). Nevertheless, there are downsides to 12.Qg4. Two other queen moves
have been tried:
a) 12.Qe2 Bb7 13.0-0-0 Qb6 14.Kb1 Be7 15.h4 0-0-0 16.a3, when Black had some difficulty finding play, as can be
seen by his next move: 16...Nb8 17.Be3 Nc6 18.Ba2 Nxd4 19.Bxd4 Qc6 20.Rhf1 Rhf8 21.g3, and White had succeeded
in containing Black’s queenside activity, clearing the long diagonal, and pressuring f7, which gave him an edge in
Topalov – Anand, Wijk aan Zee 1996.
However, Black has had better results after 13...Qa5. For example, Reefat – Wojtkiewicz, Dhaka 1999, went 14.a3
(interesting is 14.Rhf1, when Black should probably choose the safer 14...Nxb3+ 15.Nxb3 Qc7, as in Motwani – Pigott,
Plymouth 1989, since 14...b4 15.Nd5!, as was played three times by Czech IM Pavel Vávra, is fairly dangerous)
14...Nxb3+ 15.Nxb3 Qc7 16.Rhf1 Be7 17.Nd4 Nf8!? (a typical maneuver in these positions after e4-e5; the knight
strengthens the kingside) 18.Ne4 Ng6, when Black had a reasonable game.
b) The longest possible move, 12.Qh5, has only been played a few times. The question is whether it helps or hurts
White to induce ...g7-g6. Most likely, this helps Black. The game Sión Castro – Nedobora, Mondariz 1996, went 12...g6
13.Qe2 Bb7 14.0-0-0 b4 15.Na4 Nxb3+ 16.axb3 Qc7, when Black stood well.


This surprising move points out one problem with 12.Qg4. It would seem that the advance ...h5-h4, although played
with gain of time, only serves to weaken the black position. However, the panorama here is not normal, and Black will
likely never castle kingside in any case, so the push to h4 actually greatly increases Black’s dynamicism. Down the line,
a disruptive ...h4-h3 further advance starts to become possible. Nevertheless, there are specific questions requiring
concrete answers. Another possibility is 12...Nf6!?, attacking the knight on d4, which was mentioned by Latvian IM
Antanas Zapolskis in Chess Informant 57. However, this idea is better carried out with the inclusion of ...h7-h5-h4 – see
the next note.

13.Qg3 h4 14.Qg4 g5!?

It’s a fascinating position already. It looks crazy to play as Kasparov has, leaving his king in the center, pushing
pawns, and not developing. But he does this to gain positional advantages, backed up by concrete variations.
However, perhaps a better possibility is 14...Nf6!, which was played by “The World” against Peter Heine Nielsen in a
correspondence match via the Internet. After 15.exf6 Qxd4 16.fxg7 Qxg7 17.Qe2 Be7:

Position after 17...Be7 (analysis)

in this “typical” Najdorf position, Black stands better. The black pieces are very active, with pressure against g2 coming
as well as play on the c-file. Don’t forget that the e6-pawn is passed, and that nearly every endgame will be a positional
disaster for White. Nevertheless, the black king lacks a secure home, which makes the game at least somewhat unclear,
and ultimately the game Nielsen – World, Internet 2000, ended in a draw.


Short sacrifices the bishop, which was the only reasonable choice. This line is not for the faint of heart.
On 15.Bxg5 Nxe5 White cannot guard g5 and d4 while saving the queen. So he has to go into the ending with
16.Bxd8. In all these endings, Black will have a massive advantage due to his center pawns, control of the h1-a8
diagonal (remember the weakness on g2) and especially the terrible bishop on b3.
The ending after 15.Qxg5 Qxg5 16.Bxg5 Bb7 17.Rg1 Bg7 is also terrible for White. Black will take back the pawn
on e5 with a huge advantage.


After this, Black faces a cascade of sacrifices. Kasparov’s recommendation was 15...Rh6, in order to cold-bloodedly
prevent sacrifices on e6, which was White’s main idea. Now 16.Bxg5 is met by 16...Rg6, forcing a favorable endgame.
White’s best is 16.Be3!, which was analyzed by Nielsen. Black faces a lot of problems here, although he might be able
to reach rough equality:
a) 16...Nxe5 17.Nc6! is crushing.
b) 16...b4 17.Ne4 Nxe5 18.Nc6! Ncd3+ 19.Rxd3! Nxd3+ 20.Kb1 Qc7 21.Qxg5 Qxc6 22.cxd3 (Nielsen), when
White will at least regain the material while keeping a strong attack.
c) 16...Nxb3+ is premature. After 17.axb3 Bb7 18.Nf3 Rg6 19.Ne4 Black is in danger.
d) 16...Bb7 seems to be best: 17.Nf3 and now White will bring the knight to e4, for instance 17...Rg6 18.Bxc5 Bxc5
19.Ne4. But then after 19...Qc7 Black will castle queenside and have navigated most of the dangers, with a reasonable
Black can also accept the piece sacrifice by 15...gxf4!?, which is playable but takes some courage and enormous
calculation. 16.Nxe6 Nxe6 17.Bxe6 Qe7! (17...fxe6 18.Qg6+ Ke7 19.Rd6 is a winning attack) 18.Nd5 (or 18.Bxd7+!?
Bxd7 19.Qf3 Ra7 20.Rhe1, with a very dangerous attack) 18...Nxe5! and the hanging pieces on both sides fizzle out to
an endgame: 19.Nxe7 Nxg4 20.Bxc8 Ne3 21.Bb7 Nxd1 22.Bxa8 Nxb2!, when White will be fighting for a draw down
a pawn.


The knight leaps into a trap, similar to the famous game Kholmov – Keres, USSR Chp 1959.

16...Nxb3+ 17.axb3 Qc5 18.Ne4!

No retreat, of course!

18...Qxc6 19.Bxg5

Black is up a piece, but his king is caught in the center and d6 is a gaping hole. He has a tempo to do something, but
there is not much else other than trying to catch up in development.

19...Bb7 20.Rd6!
White wants a knight on d6, and is willing to sacrifice more to get it.


Black must take the exchange, since if 20...Qxe4 then 21.Rxe6+ is decisive; while if 20...Nxe5 then White wins very
nicely: 21.Nf6+ Ke7 22.Rhd1! and if Black takes the queen, then 23.Rd7+ is mate next.

21.Nxd6+ Kf8 22.Rf1

White is down a rook, but around the black king he has a “material advantage” around the enemy king, and that is
what matters. Black cannot get his pieces over so easily because of the pawn on e5 and knight on d6, which cut his
position in half.

22...Nxe5 23.Qxe6 Qd5


This is very tempting, and certainly when you are down a rook you want to land some kind of blow. But it turns out
that it lets Black escape; instead, building the net more slowly was the right way. According to Short (and others),
24.Qf6 would win: 24...Rh7 (24...Rh5 25.Bh6+ Rxh6 26.Qxh6+ Kg8 27.Nf5) 25.Rf5! (the knight cannot be saved and
the attack continues) 25...Qxg2 26.Qxe5! (threatening 27.Be7+ and 28.Rg5+) 26...Qh1+ 27.Kd2 Qg2+ 28.Kc3 Kg8
29.Bf6 (the h7-rook is useless and White threatens 30.Rg5+, while the white king will escape from the checks)
29...Qh3+ 30.Kb4 a5+ 31.Ka3 and after this wild play, White is left in total control. 32.Rg5+ is unstoppable.


Trying to leave the white pieces hanging would lead to a spectacular mate: 24...Kg8 25.Rg7+! Kxg7 26.Nf5+ Kf8
27.Qe7+ Kg8 28.Qg7#.

25.Be7+ Kg7 26.Qf6+ Kh7 27.Nxf7

White is down two rooks for a knight, but look at the difference in activity! Despite the inaccuracy on move 27, Short
still has all the chances. Kasparov has to find the only moves just to survive.


There were threats of 28.Qh6+ and 29.Qg6# or 28.Ng5+, and this is the only defense.

28.Ng5+ Kg8 29.Qe6+ Kg7 30.Qf6+ Kg8 31.Qe6+

Short repeats moves to get closer to the time control, but he has no interest in a draw.

31...Kg7 32.Bf6+ Kh6

Every one of Black’s moves is forced, which makes it easier for him to play in the time scramble.

33.Nf7+ Kh7 34.Ng5+

Taking the rook would lead to an opposite-colored-bishops position where White is up a pawn but probably cannot
win: 34.Nxh8 Rxh8 35.Qd7+ Kg6 36.Bxh8 Qg5+ 37.Kb1 Qxg2 and White can chase the black king around but not
checkmate him. And it is unlikely he will be able to use his extra pawn.


Short’s choice does not win, but definitely gives Black the most problems. After 35.Qe7 Rag8 36.Nf7+ Kg6
37.Nxh8+ Rxh8 38.Bxh8 Qg5+ 39.Qxg5+ Kxg5 40.g3 hxg3 41.hxg3 Kg4 42.Be5 Bd5 43.Kd2 Kf3, the endgame is
drawn (Kasparov).


Not 35...Kxg5 36.Qe5+ Kg6 (36...Kg4 37.h3#) 37.Qf6+ Kh7 38.Qg7#.

36.Nf7+ Kh7 37.Qe7 Qxg2?

After hardly making a move of his own since the opening, Kasparov finally has a decision to make, and he makes the
wrong one. This gives Short another chance to win. Black needs to play 37...Kg8!!, leaving the pieces hanging in air.
Then there follows 38.Qxb7 Rf8!:
Position after 38...Rf8! (analysis)

with a funny position! After 39.Ne5 Rf1+ 40.Kd2 Qd6+ 41.Ke2 Qd1+ 42.Ke3 Kxh8 (chasing the king more does
nothing) 43.Qc8+ Kg7 44.Qc7+ Kf6 45.Qf7+ (the simplest draw) 45...Kxe5 46.Qe7+ Kd5 47.Qd7+ Kc5 48.Qc7+
Kd5 49.Qd7+, the checks continue along the seventh rank.


Facing time pressure and a practically irrational position, it would not have been easy for White to find the correct
solution. But 38.Bd4! would have interfered with checks on the f-file and won for White. 38...Qh1+ (or 38...Qf1+
39.Kd2 and there is no check on f2) 39.Kd2 Qxh2+ 40.Kc3! (the king escapes this way) 40...Qg3+ 41.Kb4 (the checks
come to an end and White has a decisive attack) 41...Bd5 42.Nd6+ Kg8 43.Qf6.
Regardless of the outcome of the match as a whole, this game would have become a classic had Short picked 38.Bd4.

38...Qf1+ 39.Kd2 Qf2+

Were the bishop on d4, this would be impossible.


40.Kc3 Qe1+.

40...Qf3+ 41.Kd2 Qf2+

The turbulent battle ends in a perpetual.


Game 15
Bassem Amin – Alexander Areshchenko
Antalya 2013

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Bc4 e6 7.Bb3 Nbd7 8.0-0 Nc5 9.f4

With this very direct move, White aims to play a quick e4-e5 or f4-f5, while daring Black to take the e-pawn.


While in many Sozin lines Black should decline the offered e-pawn, this is one case where taking the pawn is possible
– and indeed, promising for Black. However, there is a safer option which is often preferred, in 9...Be7. After 10.e5 (a
classic Najdorf specialist, Semyon Dvoirys, showed a thematic way to meet 10.f5: 10...Bd7!? 11.Qf3 e5! 12.Nde2 Bc6
13.Ng3 h5! [an idea of Bobby Fischer’s, in similar positions] 14.Bd5 h4 15.Nge2 Bxd5 16.exd5 b5, with good play for
Black in Hendricks – Dvoirys, Dieren 2000) 10...dxe5 11.fxe5 and now Black has two options:
The simplest way to roughly equalize is 11...Nxb3 12.axb3 Bc5 13.Be3 Nd5 14.Bf2 Nxc3 15.bxc3 Qc7 as in the
game Arakhamia-Grant – Ftáčnik, Sydney 1991.
Or 11...Nfd7:
a) 12.Qh5 can be met by 12...Nf6!, forcing a repetition by 13.Qd1, since 13.exf6 Qxd4+ 14.Kh1 gxf6 leaves White
without real compensation for the pawn. Then both players have their options to continue the game – Black by 13...Nd5,
or White by choosing a move other than 14.Qh5 in the event of (13.Qd1) 13...Nfd7.
b) 12.Bf4 is very natural, but Black has this aesthetic and byzantine maneuver, typical of such positions: 12...Nf8!.
The knight heads to g6, strengthening the kingside and clearing the d-file. The game P.Carlsson – Cheparinov, Albena
2012, went 13.Be3 Ng6 14.Qh5 Qc7 15.Nf3 Bd7 16.Rae1 h6 17.Bf2 Nf4 18.Qg4 g5, and Black was taking over the
c) 12.Qe2!? has been played rarely, but it might be the most reasonable continuation. After 12...0-0 13.Bf4 Qc7 we
have an unexplored position. White can attempt to build up an attack by Rf1-f3, et cetera, while Black must watch out
for Nc3-d5 or Nd4-f5 sacrifices. On the other hand, the b3-bishop is difficult to activate.

10.Nxe4 Nxe4 11.f5 e5 12.Qh5 d5!

This variation had been played a few times by weaker players, but its first high-level outing was the famous game
Topalov – Kasparov, Amsterdam 1996. In that game, the world champion preferred to keep the center solid and made
the awkward move 12...Qe7?!. After great complications, Topalov won in spectacular fashion: 13.Qf3 Nc5 14.Nc6!
Qc7 15.Bd5 a5?! 16.Bg5 Ra6? 17.Nd8!! f6 18.Nf7 Rg8 19.Be3 g6 20.Ng5! Rg7 21.fxg6, when Black’s position was in
ruins. Just two days later Topalov repeated the variation against Nigel Short, who played the improvement 12...d5.

13.Re1 Bc5

Black develops and pins the knight, which is very natural. But in recent years, an alternative has given Black success
– 13...Qc7!?, after which most games have gone 14.Ne6 (the idea is 14.Bxd5 Nf6, when White is struggling) 14...Bxe6
15.fxe6 Qb6+ 16.Be3 Qxe6, and White does not have enough compensation for his pawns. For example, 17.Bd4 g6
18.Qxe5 Qxe5 19.Bxe5 Bc5+ 20.Kh1 0-0 and Black exploited the extra pawn in Moosavian – Miroshnichenko, Al-Ain
13...exd4, on the other hand, loses by force: 14.Rxe4+ Be7 15.f6! dxe4 16.fxg7 Rg8 17.Qxf7+.

14.Rxe4 0-0!?

Black leaves both pieces hanging. This was touted as an improvement on the earlier move 14...Bxd4+ where, despite
Black’s victory, decent attacking chances for White were found. For example, 15.Be3 (also possible is the exchange
sacrifice 15.Rxd4!? Qb6 [and not 15...exd4 16.Bg5 followed by Re1+, with a crushing attack] 16.c3 exd4 17.Qe2+ Kf8
18.Qe5, when Black’s defense is not easy) 15...0-0 16.Rxd4 exd4 17.Bxd4 f6 was Topalov – Short, Amsterdam 1996.
Now 18.Qf3 would give White reasonable compensation after capturing the pawn on d5 and playing c2-c4.


The rook takes aim at the sensitive g7 point. Instead 15.Rh4? Bxf5! would exploit the pinned knight and overworked

15...Bxd4+ 16.Kh1 e4 17.Bg5

This was a novelty at the time. In earlier games, White played 17.Bh6, but after 17...Kh8 18.Rd1 Bxb2 19.Rxd5 Qb6
White’s attack was not breaking through, while Black could create counterplay with ...Qf2 coming. Megaranto –
Konguvel, Hyderabad 2005.


Black must block the f6 square. Instead 17...Qb6 18.Rf1 and f5-f6 is coming eventually. Also, 17...f6 18.Bh6 Rf7
19.Rd1 is crushing.

White protects the f-pawn, preparing Rh4. There was another possibility in the strange move 18.Qh6!?. However, this
fizzles out into an endgame that White probably cannot hope to win: 18...Kh8 (the only move) 19.Bxf6 Qxf6 20.Qxf6
gxf6 21.Bxd5 Bxf5 22.Rf4 Bg6 23.Bxe4 and White has a nominal advantage, but winning chances are very slim.

18...Bxg5 19.Rxg5 Qf6 20.Bxd5

White has regained the pawn, while the d5-bishop stands well. On the other hand, White has not succeeded in creating
real threats to the black king, so the rook is awkwardly placed on g5. The passed e-pawn could become a powerful force
for Black.

This move, however, is too abrupt and ends up losing the pawn. After 20...Re8! Black’s game would be in order. He
would follow up with ...Re5, ...Bc8-d7, and ...Rae8, only pushing the e-pawn later.


It turns out to be difficult to defend the pawn.


This was Black’s idea – to develop with gain of tempo and drive the bishop from its active square d5. If instead
21...Qxb2 then White plays 22.f6, winning.
Protecting the pawn was not easy – 21...Re8 22.Qe2 Qe5 23.c4 and White is preparing f5-f6, while the e-pawn can be
attacked again by Rff3. Meanwhile, 23...Bxf5 loses to 24.Rg5.

22.Be4 Rae8 23.Rxe3 Bc4 24.Rfe1 Qxb2?

This allows White to renew the kingside attack. Taking the other pawn by 24...Bxa2 obviously leaves the bishop in a
bad state after 25.b3. However, Black was better off playing 24...Re5 and doubling the rooks on the e-file, when he has
decent fighting chances despite the pawn minus.

White has been seeking this breakthrough since the opening, and finally he achieves it. Checkmate on h7 is

25...g6 26.Bxg6! fxg6

26...hxg6 also leaves the e8-rook inadequately defended after 27.Qh6 Qxf6 28.Rxe8, and here White’s win is much
easier than in the game.

27.Qh6 Qxf6 28.Rxe8 Bxa2 29.h3?!

White naturally wants to clear up the back-rank problem, but it was more accurate to play 29.Rxf8+ Qxf8 30.Qg5!,
which prevents the bishop from anchoring itself on the long diagonal. White ought to win without too much trouble.


White still has good winning chances, but after the bishop comes to c6 things will not be as easy. The bishop is
solidly anchored, it defends the queenside pawns, and the counter-threats against g2 constrain White’s actions.

30.Rxf8+ Qxf8 31.Qg5 Bc6 32.Rd1 Kg7 33.Rd8 Qf5 34.Qxf5

The weakness of the f-pawn after the recapture allows White to go into this endgame with good winning chances.

35.Kg1 f4?!

In principle, the trade of the f-pawn would lead to a drawn position, but due to specific tactics the black king ends up
stuck on the h-file. A better chance was 35...Kf6, with some drawing prospects.

36.Rd4 f3 37.Rg4+! Kh8

Black cannot go toward the center with 37...Kf6 because of 38.gxf3, when White wins a pawn. And the king is no
better off on h6 than h8 after 37...Kh6 38.gxf3 Bxf3 39.Rg3.

38.gxf3 Bxf3 39.Rg3 Bc6 40.Kf2

Now with the king trapped on the h-file, the white king comes over. Black cannot hold a fortress.


Had Black tried to keep the queenside pawns where they were, the white king would have come to b6, followed by an
exchange sacrifice on b7 or c6 and the promotion of the c-pawn.

41.Ke3 b5 42.Kd4 Be8 43.Kc5 b4

The exchange sacrifice could also have been carried out after 43...Bg6 44.Rxg6 hxg6 45.Kxb5, when the c-pawn

44.h4 h6

Rather than 45.Kb6? a4 46.Kc5 b3 47.cxb3 axb3 48.Rxb3, with a drawn game. The bishop stays on the h7-b1
diagonal, and White cannot breach the fortress. With his last move, White prepares a maneuver to drive the bishop from
e8 and invade on b5 with the king, not allowing the exchange of queenside pawns.


There are no good waiting moves.

46.Re2 Bh5

46...Ba4 47.Kb6 b3 48.c4! wins for White.

47.Re1 Bg6

Black cannot control b5, so he attacks c2.


But this interposition means that Black cannot maintain his delicate fortress.


48...Kh8 49.Re2 Bh5 50.Rd2 does not help, because if 50...Be8 then 51.Rd8 wins.

49.Re2 Kg7

Now if 49...Bh5, then 50.Rg2+ Kf7 51.Kb5 wins.


A last trap was 50.Kb5?? b3. But now Black cannot prevent White from taking the a- and b-pawns, gaining a passed
pawn on the queenside.

50...Kf6 51.Kb5 Be4 52.Rf2+ Kg6 53.Kxa5 Kh5 54.Kxb4 Bh7 55.Rh2 Kg4 56.c4 Kg3 57.Rh1 Be4 58.Rd1 1-0

Supplemental Games

Jakov Geller – Vladimir Belov

Sochi 2006
1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Bc4 e6 7.a4 Be7 8.0-0 0-0 9.Be3 Nc6 10.Kh1 Re8 11.Ba2 d5
12.exd5 exd5 13.f3 Ba3 14.bxa3 Rxe3 15.Qd2 Re7 16.Rfd1 Be6 17.Nce2 Rc8 18.c4 Nxd4 19.Nxd4 Kh8 20.cxd5 Bxd5
21.Nf5 Rd7 22.Bxd5 Rxd5 23.Qe1 Qd7 24.Ne3 Re8 25.Nxd5 Rxe1+ 26.Rxe1 Qxd5 27.Rad1 Qa5 28.Re7 h6 29.Rxf7
b6 30.Rb7 Qxa4 31.Rd8+ Kh7 32.h3 b5 33.Ra8 Qxa3 34.Raa7 Nh5 0-1

Erik Blomqvist – Viswanathan Anand

Gibraltar 2016
1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Bc4 e6 7.Bb3 b5 8.0-0 Be7 9.Qf3 Qc7 10.Qg3 0-0 11.Bh6 Ne8
12.Rad1 Bd7 13.f4 Nc6 14.Nce2 Kh8 15.Bg5 Nf6 16.Qh4 Rae8 17.f5 exf5 18.Nxf5 Bxf5 19.exf5 Bd8 20.Ng3 Ne5
21.Qd4 Qb7 22.Kh1 Neg4 23.Bxf6 Nxf6 24.c3 Bb6 25.Qh4 Bc5 26.Rf4 Be3 27.Rf3 h6 28.Rxd6 Bg5 29.Qb4 Re1+
30.Rf1 Rxf1+ 31.Nxf1 Re8 32.Bc4 Re1 33.Bd3 Ne4 34.f6 Bxf6 0-1

Thal Abergel – Maxime Vachier-Lagrave

Pau 2008
1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Bc4 e6 7.Bb3 Nbd7 8.Bg5 Qa5 9.Qd2 Be7 10.0-0-0 Nc5
11.Rhe1 0-0 12.Kb1 Qc7 13.f4 b5 14.e5 dxe5 15.fxe5 Ne8 16.Ne4 Nxb3 17.cxb3 Bb7 18.Rc1 Bxg5 19.Qxg5 Bxe4+
20.Rxe4 Qd7 21.a3 Rc8 22.Rxc8 Qxc8 23.Nf3 h6 24.Qh4 Nc7 25.Rg4 Kh8 26.Qg3 Qd7 27.Rxg7 Qd3+ 28.Ka2 Nd5
29.Rg4 Nc3+ 30.bxc3 Qc2+ 31.Ka1 Qxc3+ 32.Ka2 Qc2+ 33.Ka1 Qc1+ 34.Ka2 Rc8 0-1
Emory Tate – Leonid Yudasin
Chicago Open 1997
1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Bc4 e6 7.Bb3 Nbd7 8.Qe2 Nc5 9.g4 b5 10.g5 Nfd7 11.Bd5 Bb7
12.Bxb7 Nxb7 13.a4 bxa4 14.Rxa4 Nbc5 15.Ra3 Qb6 16.0-0 Be7 17.Kh1 0-0 18.b4 Na4 19.Nf5 exf5 20.Nd5 Qd8
21.exf5 Re8 22.Qh5 Nab6 23.Rh3 Nf8 24.f6 Nxd5 25.fxg7 Kxg7 26.Bb2+ Kg8 27.g6 Bf6 28.gxf7+ Kh8 29.Rg1 Re1
30.Rxe1 Bxb2 31.Re8 Nf6 32.Rxd8 Rxd8 33.Qh6 Ne4 34.Qh4 Nf6 35.Rg3 N8d7 36.Qh6 1-0
Chapter 5
White to Play and Win: 6.h3

The move 6.h3 was first suggested and played by the American master Weaver Adams. Adams was an unusual
character who passionately believed that White should win by force after the move 1.e4. He published a book and many
articles to prove his belief. His “refutation” to the Najdorf was the little move 6.h3.
The logic behind this move is deep. Unlike with the sharp moves 6.Bg5, 6.Bc4, or 6.Be3, White doesn’t attempt to
punish Black directly for the slight delay in development inherent in 5...a6. Instead, he plans to strike the positional
blow g2-g4, allowing an “improved” fianchetto – and possibly an extra tempo compared to the 6.g3 variation. In doing
so, he challenges Black to punish him for this achievement – something which Black might not be prepared to do. The
g2-g4 move is not merely a precursor to a kingside offensive, but most importantly it directs White’s efforts towards
controlling the strategically crucial d5 square. Weaver Adams might have had strange ideas, but in this case he was onto
The move 6.h3 slumbered for decades, despite three uses of it by Bobby Fischer in the 1960s. It was regarded as a
fairly rare sideline until the twenty-first century. However, in 2008 it suddenly woke up, and now it is frequently seen at
the top level. The reason for this can be traced back to one day – covered in Game 16.
For players of the white pieces, 6.h3 looks a bit like a holy grail of playing against the Najdorf – a positionally sound
line where play is governed by general principles, and yet White can count on frequently gaining a vicious attack. His
strategic plans are clear, and can have great force. Meanwhile, there are relatively few long, forcing, and theoretical
lines compared to other variations of the Najdorf.
Game 16 deals with the main development which brought 6.h3 back to life: the piece sacrifice 6...e6 7.g4 b5 8.Bg2
Bb7 9.0-0 b4 10.Nd5!. In the notes to this game, we also discuss Black’s other setups with 6...e6.
Game 17 features the typical Najdorf move 6...e5, with what has become arguably the main line of the 6.h3 Najdorf:
7.Nde2 h5!?. Black carries out a highly unusual positional pawn sacrifice, winning a game which opens new vistas in
strategic understanding. The notes to this game cover Black’s various other alternatives: 6...g6, 6...Nc6, and 6...b5, as
well as alternatives to 7...h5.
Game 16
Hikaru Nakamura – Nikolai Ninov
French Team Chp, Évry 2008

1.e4 c5 2.Nf3 d6 3.d4 Nf6 4.Nc3 cxd4 5.Nxd4 a6 6.h3 e6

Black opts for a switch to a Scheveningen-style position. This has its logic after 6.h3, since White cannot really
adopts a classical setup (Be2/0-0/f4), while the English attack setup with Be3/Qd2/0-0-0 will often cost a tempo when
White moves the h-pawn a second time. Therefore, in most cases White continues with an “extended fianchetto” (g2-g4,
Bf1-g2, and 0-0) and a slow-growing attack on the kingside, which has its own peculiarities. In Game 17 we will see the
Najdorf-style 6...e5 as well as consider some other possibilities for Black.

7.g4 b5

A very logical reaction: Black meets White’s fianchetto by opposing the bishop, while the eventual ...b5-b4 poses a
threat to the white e-pawn. Nevertheless, in view of the sacrifice in this game (which had been introduced by Sergey
Karjakin a few months earlier), this line has been mostly neglected recently. Instead, Black’s play has gone in three
different ways:
With the immediate 7...d5, Black carries out this desirable break before White is ready to prevent it. This move also
logically responds to wing moves with a counter in the center. The game, however, opens up very early and White is
able to gain some slight pressure in the center and along the long diagonal, often leading to a “two results” type of
position. For example, 8.exd5 (other options are 8.Bg2 and the tricky 8.Nde2!?) 8...Nxd5 9.Nde2 (an alternative here is
9.Bd2) 9...Bb4 10.Bg2 0-0 11.0-0 Nxc3 12.Nxc3 Qc7 (the exchange on c3 is considered to give White a better
endgame, with the two bishops in this open position being more relevant than the compromised pawn structure – in
particular, White will often bring the bishop to b6, controlling the d-file) 13.Qd4 Bd6 14.Be3 Bd7 15.Rad1 Bh2+
16.Kh1 Be5 17.Qb6 Bc6 18.Qxc7 Bxc7 19.f4, and White had a very pleasant endgame edge in Świercz – Firouzja,
Doha 2015. Black has trouble completing his development and the c5 and b6 squares are tender.
7...Be7 (the immediate 7...Nfd7 is also played here) 8.Bg2 Nfd7 has been fashionable for a number of years. Black
anticipates the advance g4-g5 while delaying castling, keeping open the options for queenside development, and
occasionally planning to use the dark-squared bishop. For example, 9.f4 Bh4+ would be disruptive. White often
switches to a plan of queenside castling by 9.Be3 or even the immediate 9.g5 (9...Bxg5? 10.Nxe6). In my view,
however, a promising plan was seen in Baklan – Esen, Turkey 2014: 9.0-0 Nc6 10.Nce2! Qc7 11.c3. The late
Belarusian GM Viktor Kupreichik, one of the pioneers of 6.h3, used to use this plan as well in similar positions. Not
often can White move the knight from c3 in this way, since Black can normally reply with ...d6-d5. But with the early
...Nfd7 this becomes possible. White thus makes it difficult for Black to capture on d4, lessens the effectiveness of ...b7-
b5, and blunts Black’s play on the c-file, while preparing a kingside attack with f2-f4, g4-g5, et cetera.
7...h6 is seen very frequently, and it can also be reached from the Scheveningen. One of Black’s ideas runs as follows:
8.Bg2 Be7 9.Be3 Nc6 10.f4 Nd7 11.0-0 0-0, and now if 12.Nce2 then 12...Nxd4 13.Nxd4 e5 14.Nf5 exf4 15.Bxf4
Ne5, when Black has been shown to have good counterplay, for example 16.Qd4 Bxf5 17.exf5 Rc8 18.c3 Nc4 as
occurred in Kulaots – Artemiev, Moscow 2017. For a more promising idea, starting with 12.Nf3, see the supplemental
game Bromberger – Grandelius, Munich 2017.

8.Bg2 Bb7 9.0-0!

This move was the real innovation, although the justification lies in White’s following move. In previous games
White had played either 9.Qe2, 9.a3, or 9.g5, all of which slowed down his play a little or were prematurely committal.
Because of 7...b5 in particular, the move 6.h3 was regarded as little more than a sideline for decades. But on the same
day in 2008 in the Amber Rapid tournament, both Magnus Carlsen and Sergey Karjakin played 9.0-0, ushering in a new
era for 6.h3.


Black calls White’s bluff. Simply completing development is not possible here, which shows a dark side of the early
...b7-b5 push: 9...Be7 10.e5 Bxg2 (or 10...dxe5 11.Bxb7 Ra7 12.Bc6+ Kf8 13.Ndxb5 axb5 14.Qxd8+ Bxd8 15.Bxb5,
with a winning position for White) 11.exf6 Bxf1 12.fxe7 Qxe7 13.Qxf1 and White’s two minor pieces were stronger
than Black’s inactive rook and pawn in Fedorchuk – Berbatov, Navalmoral 2009.
Black can avoid the challenge with 9...Qc7, protecting the bishop, but this implies some delay in Black’s play. White
should continue with 10.Re1 Be7 11.a4!, striking at the queenside. Then after 11...bxa4 (or 11...b4 12.Na2 [it is not
necessary to speculate with 12.Nd5 here] 12...d5 13.exd5 Nxd5 14.Nxe6! fxe6 15.Rxe6, and White’s attack is very
strong – he will win back the material with interest) 12.Rxa4 0-0 13.Rb4! Nfd7, the game Robson – Andrews,
Washington 2008, went 14.Nf5 and White gained a large advantage, but even stronger is 14.Rxb7! Qxb7 15.e5 and
Black is losing after either 15...d5 16.Nxe6! fxe6 17.Nxd5!, or 15...Qa7 16.Be3, when there is no good defense to
White’s threats.
Boris Gelfand, against Carlsen on the day in Nice when 9.0-0 was played simultaneously by both of the recent
opponents for the world championship, played 9...h6 here. However, after 10.Re1 e5 11.Nf5 g6 12.Ne3, White had the
advantage and went on to win.

Nc3-d5 sacrifices against the Sicilian have been known for a very long time, but not until 2008 was it recognized that
White could play this here. Certainly not 10.g5 bxc3 11.gxf6 gxf6 12.bxc3 Rg8, as in Matsenko – Efanov, Chelyabinsk


There is no serious alternative to accepting the sacrifice.


White’s compensation is of a general nature, but can be backed up by concrete variations. The e-file has been opened
and White now has access to the c6 and f5 squares for the knight. The advanced position of the b4-pawn even factors
into White’s compensation – in some cases, White captures this pawn without releasing his grip. While there is not
normally a direct checkmate, Black’s pieces end up in a tangle, allowing White to methodically increase the pressure.

11...Be7 12.g5

White drives the knight back. It is most accurate to play this move before the black king goes to e7, so that the knight
will not be able to retreat to e8, where it defends g7.
The first game with 10.Nd5, Karjakin – Van Wely, Nice 2008, featured the immediate 12.Nc6 Qc7 13.Nxe7 Kxe7
14.g5 Nfd7, which transposed to Nakamura–Ninov. However, that move order gave Black the extra option of 14...Ne8,
defending g7 and allowing the other knight to be developed to d7. While White has compensation there as well, there is
no need to give Black this option, so the immediate 12.g5 is more accurate.


Of course 12...Nxd5 walks into a pin, and after 13.Nf5 White regains the material with a large advantage.
It is hard to think that Black could equalize with the demure 12...0-0. After 13.gxf6 Bxf6 14.a3 a5 15.axb4 axb4
16.Rxa8 Bxa8 17.Re1, White was somewhat better in I.Popov – Lu Shanglei, China 2015.

This is the right way to continue. Black cannot afford to capture on c6, opening the long diagonal, so White will be
able to capture on e7, displacing the black king and removing the defender of the dark squares.

13...Qc7 14.Nxe7 Kxe7


It is important to realize that this variation is by no means exhausted, and while practically speaking it is not easy for
Black to play, it is not winning for White. Some players have taken on the black side with “open eyes” – perhaps to set
themselves a limited task: you just defend, and if your opponent doesn’t show anything, you might win! White’s choice
on this move is therefore very important. This centralizing queen move attacks g7, and can be followed up by the
attacking and developing moves Bc1-f4 and Rfe1.
However, there is another option, which Karjakin used in the original game in this line. 15.Re1+ forces Black to
immediately choose which way to retreat. In the Karjakin–Van Wely game, Black played 15...Kf8, and after 16.Qe2
was already in serious trouble. The game continued 16...Qd8 (if 16...g6 then 17.Be3, and the bishop comes to the long
diagonal with great force) 17.Bf4 Ne5 18.Bxe5 dxe5 19.Qxe5 and White had enormous compensation for the piece,
with already two pawns, domination of the center, and the black pieces frozen on both sides of the board. The game
concluded with little more than a flick of the wrist: 19...h6 20.Qf4 a5 21.g6 f6 22.Re6 Bc8 23.Re3 Qb6 24.Rae1 Bd7
25.d6 Ra7 26.Qc4 and Black resigned. It is better to meet 15.Re1+ with 15...Kd8. However, Black’s position is certainly
not easy to play and White should have enough compensation.

15...Kf8 16.Bf4 a5

Black must develop his pieces and clear the back rank. This is the only way to do that, which also has the benefit of
protecting the b4-pawn. If instead 16...Nc5, then 17.a3! bxa3 18.b4 Ncd7 19.Rxa3 and the rook comes to c3 with great

17.Rfe1 Nb6

Black intends to finish his development with ...N8d7, keeping both knights active. But the delay in defending the
back rank allows White’s next move. The alternative is 17...Na6. Black’s position, however, remains passive and he is
playing without the h8-rook. The game Dembo – Vlček, Rethymnon 2009, continued 18.Re2 Qc5 (18...Qb6 may be
better, leaving c5 free for a knight, but after 19.Qd3!, planning Qg3, White continues to enjoy good compensation)
19.Qe4 Ne5 20.Bxe5 dxe5 21.Qxe5, and White had typical compensation: two pawns for the piece and a dominating
central position, while the black pieces were not working together.


A spectacular-looking move, but really just a trade of rooks. On the one hand Black is relieved of the burden of the
stuck rook on h8, but on the other hand White wins a pawn or two and the black king is driven out of relative safety.

18...Kxe8 19.Qxg7 Bxd5

Removing this pawn takes a huge burden off Black. His pieces can finally breathe, and his king runs for the
queenside. In the meantime, however, the white queen will pick off some pawns. Note that 19...Rf8 would lose to

20.Qxh8+ Kd7 21.Re1

White sets his sights on the black king, at least for the moment. This, however, allows Black counterplay. The
alternative is to immediately start taking some pawns with the queen by 21.Bxd5 Nxd5 22.Qxh7, when 22...Nxf4
23.Qxf7+ Kc8 24.Qxf4 leaves White with three pawns for the piece. However, the position is not very clear. The black
knight will come to e5 and the white king won’t necessarily be so safe.

21...Nc6 22.Qxh7 Nd4

Black has escaped from a dangerous (and probably unexpected) opening line in reasonable shape. He has good
counterplay in this murky position. As is frequently the case, though, even a successful solution to the problem of a
dangerous and surprising opening line costs time and energy, leaving a player depleted of both. Black soon goes wrong.

23.Be3 Qxc2 24.Qh5 Bxg2 25.Bxd4 Qf5?

This move is the culprit, leaving White with the stronger minor piece and safer king. Black could have played
25...Bd5 26.Bxb6 Qxb2, or the safer 26...Qf5, with equal chances in either case. White’s king position is too insecure to
play for a win.

Probably Black did not realize this was possible, and expected to win after 26.Bxb6? Bf3 27.Qh4 Rh8!. However, it
was indeed possible, since now 26...Qd5+ is met by 27.Qf3, in which case 27...Qxd4 28.Qb7+ Kd8 29.Qe7+ Kc8
30.Rc1+ Nc4 31.Qe8+ Kb7 32.Qxf7+ is winning.

26...Nd5 27.Qf3! Qxg5+ 28.Kh1

The end result is that material is equal, but the bishop is much stronger than the knight, which is pinned, and the black
king is very unsafe. Black now cannot safely defend the f-pawn.


If 28...f6, then 29.Rg1 Qh6 30.Qf5+ Kc7 31.Be3! Nxe3 32.Rc1+ with a winning attack, e.g. 32...Kb6 33.Qd7!.

29.Qxf7+ Kc6 30.Qf3

The queen provides perfect cover for the king from f3. The extra pawn is not easy to exploit on its own, but Black has
too many other problems.


A better chance is 30...Qf4, hoping that the black pieces will be active in the endgame and his wandering king will be
transformed into a strength. But after 31.Qd3!, Black continues to struggle.

31.Be3 Qe5


One has to assume that there should be some immediate win in view of Black’s loose and awkward pieces. Indeed,
32.Bd2! Qd4 33.Bh6! would have ended the game immediately, due to the twin threats of Rd1 and Qf7+.

32...Kc6 33.Re1
Presumably had Black now repeated moves with the nonsensical 33...Kb7, 34.Bd2 would have followed anyway.

33...Rh8 34.Bd2 Qh5

Black could do worse than escape into this endgame. On the other hand, White’s outside passed pawn and bishop
against knight give him great winning chances.

35.Qxh5 Rxh5 36.Kg2 Kc5 37.Re4

The rook is placed very well on the fourth rank, not only covering squares like h4, but also keeping an eye on Black’s
queenside counterplay.

37...Rf5 38.h4 a4 39.Kg3


This drops another pawn, but in any case White was planning f2-f3 and Kg4, followed by the further advance of the
h-pawn, while Black had no counterplay.

40.Bxb4+ Kd5 41.Re2 Ne4+ 42.Kg2 Rf8

There was nothing Black could achieve with his momentary burst of activity.

43.f3 Rg8+ 44.Kh2 Ng3 45.Rg2 Nf1+ 46.Kg1 Ng3 47.Be1 Ne2+ 48.Kf1 Rxg2

The rook trade is finally forced and the game comes to an end.

49.Kxg2 Nc1 50.h5 Ke6 51.Bd2 Nd3 52.b4 axb3 53.axb3 d5 54.b4 d4 55.b5 Kd5 56.f4 1-0

Game 17
Fabiano Caruana – Maxime Vachier-Lagrave
Wijk aan Zee 2015
1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.h3 e5

It has been said that 6.h3 is aimed – in a discreet way – against the typical Najdorf move, 6...e5. Nevertheless, the
latter is fully playable. Let us now look briefly at Black’s other options:
Entering the Dragon with 6...g6 is quite logical. By this move Black declares that h2-h3 is of less value than ...a7-a6
in the Dragon – and he is not without some justification. This move will be popular among Dragon “refugees,” of whom
there are quite a few among Najdorf practitioners. Nevertheless, White has several schemes to aim for an advantage. It
is possible to combine g2-g4 and Bf1-g2 with Be3/Qd2/0-0-0, since it is not easy for Black to develop an attack on the
queenside. White can also castle kingside and prepare the occupation of d5. See the supplemental “just blitz” game
Carlsen – Ponomariov, Moscow 2008.
6...Nc6 7.g4 Qb6!? is a rare system, where Black drives the knight back from d4 immediately. I had never seen it
before when Reynaldo Vera played it against me in a tournament in Montréal. White can castle kingside with a slow-
growing attack: 8.Nb3 e6 9.Bg2 Qc7 10.0-0 Be7 11.f4 0-0 12.Be3 b5 13.g5, B.Smith – Vera González, Montréal 2012,
with reasonable prospects. But it is more promising to castle queenside, e.g. 8.Nb3 e6 9.g5 Nd7 10.Bf4!? (or 10.h4)
10...Qc7 11.Qd2 Be7 12.0-0-0 Nde5 13.Bg3 b5 14.f4 Nc4 15.Bxc4 bxc4 16.Nd4 Bd7 17.Nf3 planning f4-f5 with the
initiative, as in the game Sethuraman – Kelires, Gibraltar 2016.
Finally, 6...b5 was played in the famous Fischer – Najdorf encounter at Varna 1962, where Fischer responded with
7.Nd5!?. While that move is interesting, not all is clear, and in any case there is no real independent value to 6...b5 after
White just plays 7.g4. If then 7...b4?! 8.Nd5 Nxe4? 9.Bg2 Nc5 10.Nxb4, Black is too far behind in development;
therefore Black should play 7...Bb7 8.Bg2 e6, which transposes to Game 16, Nakamura–Ninov.


This is by far the most common move, although in recent years 7.Nb3 has started to appear, having some similarities
with the line 6.Be3 e5 7.Nb3 Be6 8.h3 (see Chapter 3). Black’s next move is what led some players to start considering
7.Nb3, since if White does not achieve g2-g4 and Ng3 the knight might look quite bad on e2.

This has emerged as the main and most solid line against 6.h3 in recent years. Three or four decades ago, most players
would not consider playing ...h7-h5, creating a huge weakness on g5 as well as leaving the h-pawn exposed, potentially
making castling difficult. Indeed, this move was played in only one amateur-level game before the year 2000. Not until
2008 was it played with any frequency at the grandmaster level.
But the modern Najdorf player has a different view of things. The advance ...h7-h5 helps to contain the knight on e2,
which in some cases can not only be left without prospects, but can even get in the way. But what of the weakening
inherent in ...h7-h5? The twenty-first century player has begun to realize that the weaknesses created in this pawn move
might be compensated for by Black’s gain in kingside space. The cramping advance ...h5-h4 comes into sight, while in
the event of g2-g3, the same advance could serve to open lines or weaken White on the dark squares. It is far from clear
that ...h7-h5 is more weakening than h2-h3. This same development can be seen in the 6.Be3 Najdorf (see Chapter 3).
The player who specializes in 6.h3 can expect to see many other routine developing moves here, such as 7...Be7 or
7...Nc6, which allow White to play 8.g4 followed by Bf1-g2, Ne2-g3, 0-0, and eventually f2-f4 or Ng3-f5. The position
is basically a tempo up on the 6.g3 variation, where White often continues with a later h2-h3 and g3-g4. White’s results
in those lines are very good.
Besides 7...h5, however, there are two legitimate moves connected with clear ideas:
a) One is 7...b5, meeting 8.g4 with 8...b4 9.Nd5 Nxd5 10.exd5 h5!, as in the game Caruana – So, Dortmund 2015.
Now 11.gxh5 (If instead 11.g5, then 11...h4! is the best answer, continuing to restrain the white knight and separating
the g5-pawn from its support. Black has had good results here.) 11...Rxh5 12.a3 bxa3 13.Rxa3 Nd7 was the
continuation, with a sharp position where neither king will be particularly safe. Black went on to win a messy game.
White has some other ways to seek an advantage after 7...b5. One is to switch plans with 8.Ng3, developing with
moves like Bf1-d3, Bc1-g5, and 0-0, attempting to use the f5 square or to play f2-f4. This is fairly unexplored so far;
Black’s best results have come from an eventual ...g7-g6, once again making the g3-knight a bad piece. Another way is
to capture with the queen after 8.g4 b4 9.Nd5 Nxd5 10.Qxd5 Ra7 11.Be3 Be6 12.Qd3, where White keeps the same
structure, although Black has reasonable chances here.
b) 7...Be6 8.g4 d5 is a logical way of carrying out the central counterblow, similar to 6...e6 7.g4 d5. However, in this
case the position of Black’s pawn on e5 weakens his light squares a little bit. The move was under a cloud since a 1964
Fischer simultaneous exhibition game, which went 9.exd5 Nxd5 10.Bg2 Nxc3?! 11.Qxd8+ Kxd8 12.Nxc3 Nc6
13.Be3, when White had a pleasant advantage due to the queenside pressure and went on to win easily. However,
around 2010-2012 this line was being tried by strong players in connection with 10...Bb4, when Black goes into the
same endgame but doubles the white pawns. Recently, though, this line has again dropped off the radar as White has
been getting the better position, e.g. 11.0-0 Bxc3 12.Nxc3 Nxc3 13.Qxd8+ Kxd8 14.bxc3 Nc6 15.Be3 Kc7 16.Rfb1!:
Position after 16.Rfb1! (analysis)

planning to install the bishop on b6, and meeting 16...b5?! with 17.a4 Bc4 18.a5, with a clear advantage as in the game
B.Vučković – Kostromin, Novi Sad 2016. The bishop is coming to b6, ensuring White of control over the d-file, while
the black king is in great danger – this far outweighs the doubled pawns.


White heads into a 6.g3 Najdorf, with the inclusion of the moves h2-h3 and ...h7-h5. One would assume this would
favor White, as in many cases he plays h2-h3 anyway; but as discussed earlier this is not so clear. The whole approach
with ...h7-h5 in this line is really terra nova in positional chess.
The main alternative is 8.Bg5, logically seeking to exploit the weakening of the g5 square. White will exchange the
f6-knight and occupy d5 with the c3-knight, making room for the other knight to come to c3. Nevertheless, the
exchange of the dark-squared bishop is a large concession, giving Black counterplay on the dark squares. 8...Be6 9.Bxf6
Qxf6 10.Nd5 Qd8:
Position after 10...Qd8 (analysis)

and now the natural 11.Nec3 has been played many times. White dominates the d5 square and can even sometimes
control and invade on b6 with a2-a4-a5, but then tends to reach an impasse while Black’s play on the kingside gathers.
Murky, complex positions with chances for both sides result, for example 11...Nd7 12.Bc4 g6 13.a4 Bh6 14.a5 Rc8
15.Ba2 0-0 16.0-0 Kg7 17.b4 Rc6 18.Qd3 Qg5 was Anand – Nakamura, London 2012.
Lately, the plan of castling queenside beginning with 11.Qd3 has been tried many times, with some success. In some
games Black has gone ...Nb8-d7 or played ...Bf8-h6 (after 0-0-0 by White), which has resulted in problems. A
reasonable setup, promising decent play, is 11...Nc6 12.0-0-0 g6 13.Kb1 Bg7 14.h4 (14.f4 was played in the first outing
of the 11.Qd3 line, but Black equalized easily after 14...Bxd5 15.Qxd5 exf4 16.Nxf4 Qf6 17.Nd3 0-0 in I.Šarić –
Wojtaszek, Istanbul Olympiad 2012) 14...Rc8 15.g3 b5, when Black didn’t have any big problems in Oparin – Gelfand,
Moscow 2015.


Black’s alternative is to develop with 8...Be7, with a view to playing an early ...b7-b5. Of course, these lines can be
compared to the 6.g3 variation. In many cases the inclusion of h2-h3 and ...h7-h5 will help White; in a few, they may
help Black (in particular where an advance with ...h5-h4 becomes strong).
8...Be7 9.Bg2 b5 10.Nd5 is White’s normal response, making a place for the other knight on c3 and not waiting for
Black to play ...Nbd7. Now 10...Nbd7 has been typical, with Black agreeing to the exchange of the dark-squared bishop
in order to gain more control over d5. (10...Nxd5 11.Qxd5 Ra7 12.Be3 Be6 13.Qd2 has been played in some games, but
definitely offers White the better chances.) White’s most ambitious path is 11.Nxe7 Qxe7 12.Bg5 Bb7 13.Nc3 Qe6
14.Nd5 Rc8 15.c3 Nxd5 16.exd5 Qg6 17.h4 f6 18.Be3 f5 19.Bg5, when White was slightly better in Shirov – Sjugirov,
Sochi 2013.

9.Bg2 Nbd7

In the 6.g3 Najdorf, Black would normally be happy to get 9...b5 in, without White’s being able to occupy d5 with a
piece as in the last variation. This, however, is a variation where the inclusion of h2-h3 and ...h7-h5 seems to work in
White’s favor by comparison. That said, 9...b5 is playable: 10.0-0 Nbd7 11.Be3 Nb6 (11...Be7 could be met by 12.f4,
which tends to be an uncomfortable kind of position for Black; White plays b2-b3, a2-a3, Qd1-d2, and some other
useful moves, while Black has a hard time doing anything; finally, at the right time, White plays f4-f5) 12.b3 Be7
13.Qd2 Qc7 14.a4 bxa4 (14...b4 would be met by 15.a5!, e.g. 15...Nc4 16.bxc4 bxc3 17.Nxc3 Bxc4 18.Rfb1, planning
Na4-b6) 15.Nxa4 Nxa4 16.Rxa4 0-0 17.Rfa1, and White had an edge in Giri – Wojtaszek, Biel 2014.


White does not have to prevent ...b7-b5, since 10.0-0 could transpose into the previous note. However, a2-a4 comes
into White’s plans at some point, and it is useful to reduce Black’s options.
It might look like g5 is a good square for the white bishop, and in some cases it could exert some pressure there, but
an instructive downside to the bishop’s position was shown in the game Strugnell – I.Šarić, Novi Sad 2016: 10.Bg5 Be7
11.0-0 Rc8 12.a4 Nf8!? 13.b3 Qd7 14.Bxf6 (if 14.Kh2, then 14...Ng4+; however, it is worth considering 14.h4)
14...Bxf6 15.Nd5 Bd8 16.Kh2 h4!:

Position after 16...h4! (analysis)

A typical example of how Black can benefit from the inclusion of ...h7-h5/h2-h3. After 17.g4 Bxd5 18.exd5 Nh7
19.Qd3 0-0 20.f4?! exf4 21.Nxf4 Bf6, Black was better as the bishop will be very strong on e5.

10...Be7 11.0-0 Rc8 12.Be3

In view of Vachier-Lagrave’s surprising idea, this move has come to be regarded as inaccurate. Things are not
absolutely clear, but indeed there is no need to give Black the option of the pawn sacrifice, and therefore 12.b3 is
preferred by up-to-date players. This move, covering c4, is inevitable, and by playing it first White avoids 12...Nb6,
which is now met by the immediate 13.a5. Black is in danger of falling into a completely passive position if he fails to
find the right plan; White has many improvements in his position to make, i.e. Bc1-e3, Qd1-d2, Rad1, a4-a5, and
eventually f2-f4.
Black’s best chance to reach a reasonable game looks like 12...Nc5 13.Be3 (13.Bg5 is also possible, as in Radjabov –
Sevian, Baku World Cup 2015) 13...0-0 14.a5 Bd7!?, Kovalenko – Gajewski, Warsaw 2016. The bishop prepares to
come to c6 and in some cases the knight can retreat to e6; meanwhile White is tempted to weaken his position with b3-
b4. White might be able to keep some advantage here, but at least his game does not play itself.


This must have been a surprising move at the time, since after the obvious 13.b3, what will the knight achieve on b6?
White intends next to play a4-a5, simply gaining two tempi.

13.b3 d5!

This is the point of Black’s play. If Black carries out this freeing break “for free,” he will stand better. Thus White is
forced to win a pawn.


Naturally, 14.exd5 Nbxd5 15.Nxd5 Nxd5 is more than adequate, while 14.a5 d4 is obviously comfortable for Black.

14...Qxb6 15.Nxd5

While capturing the pawn, White also frees the c-pawn to come to c4. If instead 15.exd5 Bf5, Black has an excellent
position. White will not easily be able to deal with the pressure down the c-file (...Be7-b4 might be coming later), and
the pawn on d5 is not so easy to defend. Black is likely to regain the pawn with advantage.

15...Nxd5 16.exd5 Bd7

So this is the result of a somewhat surprising positional pawn sacrifice. White has an extra pawn – and not just any
extra pawn, but a passed pawn that can be solidly defended by c2-c4. In theory, White’s queenside pawn structure could
produce two connected passed pawns at some point.
What does Black have in return for this? Mainly Black’s compensation for the pawn rests in his great control over the
dark squares and White’s inability to activate the knight. White’s entire queenside and central pawn structure are locked
on the light squares, and White has had to exchange the dark-squared bishop, which means that Black can solidly
blockade the white pawns. Meanwhile, the white king is not very safe – the nudge h2-h3 has loosened the kingside, and
with ...f7-f5 and ....h5-h4 Black can wrench open lines there. We shall soon see how this worked out in practice.

17.c4 Qd6!

The d6 square is not for the bishop but for the queen, who – at least in this instance – proves to be a good blockader of
a passed pawn. From d6 the queen also looks to the kingside, while the dark-squared bishop retains possibilities to
become active either on the long diagonal or on the b8-h2 diagonal.


A natural move, hoping to regain some control over the dark squares and not allowing Black to play ...a6-a5.
However, it does not address the situation on the kingside. It was better to play 18.f4, striking back in the center. In the
game J.Tan – Troff, Internet (rapid) 2017, Black played 18...e4!?, intending ...f7-f5 and ...h5-h4. Other moves came into
consideration as well, such as 18...h4. Now after 19.f5?! Qe5, Black had a good game with prospects of ...Be7-d6 and
...h5-h4, or just capturing the pawn on f5. Instead, 19.Nd4 f5 20.h4 would lead to a tough struggle. It is hard to see how
White can use his extra pawn. While not having a direct attack, Black retains constant pressure on the dark squares, after
20...Bf6 for example.

18...f5 19.Qd3

It is hard to find any kind of plan for White – “MVL” has created an unique positional conundrum!
If White plays the natural 19.Nc3, then Black just continues with his plan of 19...h4, when the knight achieves little
on c3 and would be better off back on e2, defending g3 and f4. Or if 19.h4 then 19...g5!, and Black will develop a fierce
attack on the kingside, with White having no counterplay.
19...h4 20.g4?!

White wants to gain the use of the e4 square for the knight, despite the risks. However, this move makes it possible
for Black to open lines, and most importantly it weakens White on the b8-h2 diagonal, which proves his undoing. It was
better to sit tight somehow – although that is easier said than done.


20...fxg4 is possible, although White will firmly blockade e4 and hide his king behind the black pawns, on h1.


White is hoping to follow up with gxf5 followed by Nc3-e4 or Bg2-e4, blockading the center. However, he might
have underestimated Black’s next move and the following bishop maneuver. It was necessary to play 21.gxf5 Bxf5
22.Qe3, although Black has more than enough compensation after 22...Qg6!, preparing ...Be7-c5 and controlling the e4

21...e4! 22.Qe3


A fine maneuver, which has been in the works for some time. Black simply intends ...Bc7 followed by the invasion of
the queen on h2, after which the attack will hardly be possible to repulse. This immediately moves White to desperation.


White sacrifices a piece for two pawns, after which he will have three pawns for the piece. However, Black’s dark-
square blockade will persist, and his two bishops will be very strong. Nevertheless, White had nothing better. For
example, 23.gxf5 Bc7 24.Rfd1 Qh2+ 25.Kf1 Bf4 gives Black a winning attack – 26.Qd4 (or 26.Qxe4 Bxf5 followed by
27...Bxh3) 26...Bxf5. Next Black will play ...Bxh3, with an overwhelming attack; ...Bf4-g3 and ...e4-e3 would come
next, and White still hasn’t got a shred of counterplay. 23.f4 exf3 24.Rxf3 Bc7 also gives Black a crushing attack.

23...fxe4 24.Nxe4 Qf4

Black does not want to allow f2-f4, which would let White fight back on the dark squares.

25.Qxf4 Rxf4 26.f3 Be7 27.Kf2

If 27.d6, then 27...Rxe4! 28.fxe4 Bxd6, when the bishops will overpower the rook and two pawns. White cannot
defend both the a- and e-pawns, and soon the white king will face problems on the diagonals as well. The white rooks
can achieve nothing.

27...Rcf8 28.Ke3 Be8!

Black intends ...Bg6, threatening to capture on e4.


White has finally made this crucial step, but now it frees the light-squared bishop to cause damage, and the white king
will be caught in the crossfire of the black bishops.

29...Bb5 30.b4

A desperate move – White is prepared to go down a whole rook, but with some strong pawns. If instead 30.Rf2 then
30...Rxe4+! is like the game – of course White cannot play 31.fxe4 in view of 31...Bxc5+.


30...Bxf1 is winning for Black, naturally; but MVL’s choice is a more effective and aesthetic way to finish the game.


31.fxe4 Bxf1 would leave White down too much – a couple of passed pawns would not compensate for two extra


The point: White cannot escape the bishops’ clutches.


32.Kd4 is met by an elegant checkmate: 32...Bf6#.

32.f4 Bxc5+ 33.Kf3 Re3+ 34.Kg2 Bxf1+ 35.Rxf1 Rg3+ 36.Kh2 Bd6 also leaves White without hope – ...g7-g5 is
coming next.

32...g5+ 33.Kf5 Kf7!

The threat is 34...Bf6, when it would not be possible to stop checkmate with ...Bd3, ...Bd7, or ...Re5. Although this
takes two moves to enact, White is stuck and cannot prevent it.

34.Rfe1 Bd3+ 35.Re4 Bf6!

Checkmate next is forced, so White resigned.


Supplemental Games

Stefan Bromberger – Nils Grandelius

Munich 2017
1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.h3 e6 7.g4 h6 8.Bg2 Be7 9.Be3 Nc6 10.f4 Nd7 11.0-0 0-0
12.Nf3 b5 13.g5 hxg5 14.Nxg5 Bxg5 15.fxg5 Nde5 16.Qh5 Bb7 17.Rad1 Rc8 18.a3 Nc4 19.Rf4 f5 20.g6 Rf6 21.Qh7+
Kf8 22.Qh8+ Ke7 23.Qxg7+ Ke8 24.Rh4 Rf8 25.Rh8 Rxh8 26.Qf7# 1-0

Magnus Carlsen – Ruslan Ponomariov

Moscow (Blitz) 2008
1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.h3 g6 7.g4 Bg7 8.Bg2 0-0 9.Be3 Nc6 10.0-0 Bd7 11.Nde2 b5
12.f4 b4 13.Nd5 Nxd5 14.exd5 Na5 15.b3 Bxa1 16.Qxa1 Rc8 17.Nd4 Nb7 18.f5 f6 19.fxg6 hxg6 20.Qe1 a5 21.Qh4
Rf7 22.Qh6 Rg7 23.g5 Nc5 24.Nc6 Bxc6 25.dxc6 f5 26.Bd4 Qf8 27.Re1 e6 28.Rxe6 Nxe6 29.Bd5 Re8 30.c7 Kf7
31.Bxg7 Qg8 32.Bd4 Ke7 33.Bb6 Kd7 34.Qh4 Qg7 35.Qc4 Qc3 36.Bxe6+ Rxe6 37.c8=Q+ 1-0
Chapter 6
Systematic: 6.g3

6.g3 is a move for thoughtful players, who prefer to build their attack slowly according to comfortable schemes. The
bishop on g2 controls the key d5 square, greatly strengthening White’s center. The main downside of this move is the
loss of control of some queenside squares – c4 in particular.
In this variation, White often continues later with h2-h3 and g3-g4. Undoubtedly, it is preferable to achieve g2-g4 in
one move. Nevertheless, the player of 6.g3 prefers to avoid the complications and uncertainty of Black’s sharp answers
to 6.h3. Instead, White’s goal is to reduce Black’s counterplay and take control of the position before launching an
Game 18 covers 6...e5 7.Nb3, a retreat which has recently supplanted 7.Nde2 as the main line. The game features an
instructive positional exchange sacrifice.
Game 19 deals with the older main line, 7.Nde2, which was previously almost always played. From an innocuous
position, Magus Carlsen manages to push his opponent to the edge, finally forcing an error.

Game 18
Bartosz Soćko - Anton Korobov
Karlsruhe 2016

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.g3 e5

As is the case after every one of White’s sixth-move options, Black can head for a Scheveningen structure by 6...e6.
But in this book we are concerned with the Najdorf structure after 6...e5.

This alternative to 7.Nde2 was not common in the past, but in the last few decades it has become the main move.
From b3 the knight controls some important squares like a5, while it can reach the crucial d5 square by the maneuver
Nb3-d2-f1-e3. After 7.Nde2, on the other hand, the c3 square must be cleared by an early Nc3-d5 in order for the king’s
knight to aim at that square.


As is often the case, the early 7...b5 is fairly risky. White responds with the normal 8.a4, forcing a loosening of
Black’s queenside. Then after 8...b4 9.Nd5 Bb7 10.Nxf6+ Qxf6 11.Bg2 Nc6 12.a5 Be7 13.0-0 0-0 White perhaps had
some advantage, although Black’s position was playable in Smirin – Shirov, Poikovsky 2015.
7...Nbd7 is the main alternative, although Black uses this setup in our current game as well, only after the inclusion of
...Bf8-e7 and ...0-0. In this line, it is important to understand that ...Nbd7 constitutes a preparation for ...b7-b5, since
Nc3-d5 (either immediately or after the inclusion of a2-a4 and ...b5-b4) is met by ...Nxd5 Qxd5 Nb6, defending the
rook on a8 and driving the queen away in the most convenient way. Thus, Black is not required to play an awkward
move like ...Ra8-a7. At the same time, the development of the knight on d7 makes White’s prophylactic a2-a4 more
desirable, since the knight can no longer be deployed to c6 and then to b4. In view of this, White should play 8.a4, since
8.Bg2 b5 would be a great step for Black. Now 8...b6 prevents the cramping a4-a5 and prepares to develop the bishop to
b7. Then 9.Bg2 Bb7 10.0-0 Be7 11.Re1 could reach the position after the note 11.Re1 in our main game if Black played
11...0-0. However, in the game Short – Kasparov, St. Louis (rapid) 2015, Black used a surprising idea: 11...Rc8 12.Nd2
h5!? 13.Nf1 h4 14.Ne3 g6 15.Re2 hxg3 16.hxg3 and now Kasparov made the standard Sicilian sacrifice 16...Rxc3
17.bxc3 Nxe4, with good compensation.


It is not necessary to prevent an early ...b7-b5 yet, and if 8.a4 then Black gains the additional and quite beneficial
possibility of developing the knight to c6 and then b4. This is an important theme to remember in the Najdorf: the
knight most often goes to d7, linking with its comrade in the battle for d5 or heading to c5 to attack e4, but if White
plays a premature a2-a4 then ...Nb8-c6 becomes good. Specifically, at this moment 8...0-0 9.Bg2 Bg4! 10.Qd3 Nc6,
with the idea of ...Nb4, is trouble for White.

8...0-0 9.0-0
Now is the latest moment when Black must choose between the two setups of ...Nbd7/...b7-b6/...Bb7 or


For the alternative setup with 9...Be6, see the supplemental game Naiditsch–Narayanan.
9...Bg4 is a tempting move, because clearly both 10.f3 and 10.Bf3 harm White’s position, with the bishop going back
to e6 anyway. The problem is that Black has no good way to take advantage of the queen’s position after 10.Qd3, for
example 10...Qc7 11.Be3 Nbd7 12.a4 Nb6 13.Nd2 Rac8 14.h3 Be6, Fressinet – Milman, Philadelphia 2006, when
White has continued his usual play and gained a tempo with h2-h3, while the queen’s placing on d3 is not detrimental.


Once ...Nbd7 has been played, it is time to prevent ...b7-b5.

10...b6 11.Be3

11.Re1 begins White’s main plan of transferring the knight to e3, in order to dominate d5. The value of the
...Nbd7/...b7-b6 setup depends on whether Black can find reliable counterplay in this position. 11...Bb7 12.Nd2 Rc8
13.Nf1 and now:
13...Rc5?! (this slightly strange move has been played a few times, with some players not finding the strategic
refutation) 14.Be3! (this blocks the knight’s path, but White will switch plans) 14...Rc7 15.g4!. Now White switches to a
plan more typical of the 7.Nde2 variation: the knight heads for g3. Black has lost a lot of time, and with Ng3 and
potentially g4-g5 and Nf5 White will dominate the center and kingside, Smirin – Kulaots, Moscow 2004.
Most games have gone 13...Nc5 14.Bg5 Ne6 15.Bxf6 Bxf6. For example, 16.Ne3 g6 17.Ncd5 Bg7 18.c3 Kh8
19.Qb3 was Popović – Polugaevsky, Sarajevo 1987. White has a stable plus, although Black went on to win.

11...Bb7 12.h3 Rc8 13.Qd3

The early Bc1-e3 has prevented White’s main plan (the transfer of the knight), and he has some trouble finding a plan

13...Qc7 14.Rfc1 Qc4!

Black offers a queen trade, which would benefit him. Without queens on the board, Black will create strong pressure
on the c-file and against the e4-pawn. White therefore avoids the queen trade, but this allows Black to move the queen
out of the way and to a more active post.

15.Qd1 Qe6

This is not the most normal location for the queen, but she stands well here. Trapping her is impossible, while the
queen looks at both sides of the board and stays out of the way of the c-file.

16.Qd3 Qc4

Black prefers to repeat moves.

17.Qd1 Qe6 18.Qd3

White is satisfied with a draw against a higher-rated opponent. Objectively he is justified, since the position after the
opening promises nothing special.


Black plays for a win. As András Adorján said, White wins more games because in general, too many people are
satisfied with a draw with the black pieces! Here Black cannot complain about the opening – he has full equality in a
complex position. And now he has taken the “psychological initiative.”


The endgame after 19.Nd5 Nxd5 20.exd5 Qg6 21.Qxg6 hxg6 is very pleasant for Black. There is pressure on the
queenside and potentially against d5, while Black also has some mobile pawns on the kingside. With 19.g4, White is
hoping to play Nc3-d5 and after the trade of knights meet ...Qe6-g6 with Bg2-e4.


Therefore the queen once again goes to the c4 square.

20.Qd1 Nc5?!

This shuts in the queen and is somewhat risky. Safer is 20...h6.

21.a5 b5


We will see how dramatically significant is White’s exchange of this bishop. The dark squares on the kingside will
become very weak. One has to assume White was pursuing some kind of tactical chimera, which did not work out.
Much better was 22.Nxc5, not only keeping the defender of the dark squares, but also exchanging a very poor piece.
Now after 22...dxc5 23.Nd5 Bxd5 24.exd5 Rd7 the black queen’s awkward position allows White to defend d5. 25.b3
Qc3 26.Bd2 Qd4 27.c3 Qd3 28.c4, and White is somewhat better.

22...dxc5 23.g5

23.Nd5 Nxd5 24.exd5 (24.Nd2 is met by 24...Nc3!, when Black saves the queen with a large advantage – this is most
likely what White missed when he played 22.Bxc5) 24...Rd7 is very bad for White. The black queen has the f4 square
and there is a lot of pressure against the d5-pawn.

23...Ne8 24.h4 Qe6 25.Nd5


This purely positional exchange sacrifice might seem surprising at first – Black simply refuses to move the rook on
c7. Yet the following play will show that Black’s judgment was correct. After instead 25...Rc8 or 25...Rd7, White would
have 26.c4, and Qf3 followed by Bh3 would be coming. Instead, Black just gets ahead with weakening White’s
kingside, saying that the knight on d5 is worth more than either of Black’s rooks.

26.gxf6 Rxf6!

Black even gives White a choice of rooks. This is much better than 26...Bxf6, which would be met by 27.Nxc7 Nxc7


White removes the king’s rook, which would otherwise become a problem after, for example, 27.c4 Rg6.

27...Nxf6 28.Nd2

This allows ...c5-c4, when the bishop becomes a monster on c5. A better chance was 28.c4 when, after 28...Nxe4
29.Qh5, at least Black would not be in full possession of the initiative.


With ...Be7-c5, ...Rc7-f7, and eventually ...Nh5-f4 coming, and no counterplay for White, Black has more than full
compensation for the exchange.

29.Qe2 Bc5 30.Rf1 g6

Now ...Nh5-f4 is prepared.

31.Nf3 h6
Black is not in a hurry to capture e4, which would give White counterplay on the long diagonal and e-file. Instead,
Black just prevents Nf3-g5 and goes ahead with his plans.

32.Nh2 Rf7 33.Rad1 Kg7 34.Kh1


Black continues with the plan to enter f4. This allows Qg4, but it is not such a big problem.


A better chance was 35.Qg4, when it would be nice to be able to play 35...Rxf2 since 36.Qxe6 is met by 36...Ng3+
37.Kg1 Rxf1#, and 36.Rxf2 Bxf2 37.Qxe6 is met by 37...Ng3#. However, White has 36.Rxf2 Bxf2 37.Rd7+ Kh8
38.Bf3, when the tables are turned – the advantage of the exchange now counts. Black can instead play 35...Qe7, when
the invasion squares are guarded and ...Nh5-f4 and ...h6-h5 are coming, to hopefully expel the white queen.
Nevertheless, at least White retains some chances here.

35...Nf4 36.Qd2 h5

Black prevents Bf3-g4. The h4-pawn is now a target.

37.Rg1 Nh3 38.Rg2 Kh7 39.Rf1 Qe7 40.Rg3 Qxh4 41.Kg2 Bc8 42.b4 cxb3 43.cxb3 Be6

White retains a nominal material advantage, being up the exchange for a pawn. But his pieces are completely
disorganized. The rook on g3 is awkwardly stuck, as are the knight on h2 and the bishop on f3. Black only needs to pick
off the various weak points in the white position.

44.Qc2 Bd4 45.Qc6

If the white queen stayed in contact with f2, by 45.b4 for example, then 45...g5 followed by ...g5-g4 would be an
entirely adequate way to finish things.

The breakthrough on the kingside is decisive.

46.Rxf2 Nf4+ 47.Kh1 Qxg3 48.Rf1 g5 49.Rg1

49.Qxa6 g4 50.Rg1 Qh3 is winning for Black.

49...Qh4 50.Qe8 g4 51.Bxg4 hxg4 52.Rxg4 Qe1+ 0-1

Game 19
Magnus Carlsen – Alexander Grischuk
Stavanger 2015
In this game, we will see Carlsen’s unique ability to keep up relentless pressure, even in simplified positions.
Grischuk was on the verge of holding the draw, but erred on the famous fortieth move.

1.Nf3 c5 2.e4 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.g3 e5 7.Nde2

This knight retreat looks more harmonious than 7.Nb3, and hence was preferred for many years. From e2 the knight
links up with its partner on c3, and is within the “circuit” of the d5 square. Additionally, the plan of h2-h3, g3-g4, and
Ng3 is possible. Nevertheless, in recent years 7.Nb3 has become popular.


Black’s alternatives are the immediate 7...b5 and 7...Nbd7, which serves as preparation for ...b7-b5, since after a
hypothetical 1.Nd5 Nxd5 2.Qxd5 Black will have 2...Nb6 (or, in some cases even 2...Qc7!, meeting 3.Qxa8 with
a) 7...b5 may look like the most aggressive, but in some ways it is the most solid! The reason is that by achieving this
advance, Black induces White to play an immediate 8.Nd5 to take advantage of it. This usually leads to various
exchanges on d5 and f6. The resulting position is not so exciting, and White might have a slight edge, but it is fully
playable for Black. For example:
a1) The preliminary 8.a4, which is often played in these situations, is not good here. After 8...b4 9.Nd5 Nbd7, the
knight on e2 cannot come to c3, while the exchange of pawns on e4 and b4 only helps Black.
a2) Some players ignore Black’s early ...b7-b5, but this should not trouble Black too much. For instance, 8.Bg2 Nbd7
9.0-0 Be7 (Black purposefully delays ...Bc8-b7 – an important concept to be aware of) 10.h3 0-0 11.g4 b4 12.Nd5
Nxd5 13.exd5 a5 with the idea of ...Bc8-a6, when Black has a very active game.
a3) 8.Nd5 Nbd7 9.Nec3 Bb7 10.Bg2 Nxd5 (10...Be7 transposes to our main game) 11.Nxd5 Nb6 12.Bg5 Qxg5
13.Nxb6 Rb8 14.Nd5 Qd8, when White had kept his hold on d5, but Black was very solid in this position, which soon
started to look like a Sveshnikov Sicilian, Alekseev – Karjakin, Dagomys 2008.
b) 7...Nbd7 and White must, as usual, decide whether to play a2-a4 or to allow Black to go ...b7-b5.
b1) If after 8.Bg2 b5 9.0-0 or 9.h3 leads to similar play as above. Black delays ...Bc8-b7, planning ...Ba6 instead after
a later ...b5-b4 and the exchange on d5; whereas 9.Nd5, meanwhile, is not a problem for Black, since ...Nbd7 has
already been played. Now 9...Nxd5 10.Qxd5 Nb6 is very comfortable for Black. For example, 11.Qd3 Be7 12.b3 0-0
13.Be3 Be6 14.0-0 Qc7 15.f4 f6! was Fedorchuk – Volokitin, Lvov 2015. Black is slightly better.
b2) 8.a4 seems to be the most critical. After 8...b6 White can either occupy d5 before Black plays ...Bc8-b7, or
continue with development:
b2a) 9.Bg2 Bb7 10.h3 Be7 11.g4 Nc5 12.Ng3 0-0 13.Be3 g6. Now an instructive moment occurred in Simon –
Polugaevsky, Buenos Aires Olympiad 1978: 14.Qd2?! allowed Black to carry out the advance 14...b5! (better was 14.0-
0, when after 14...Ne6 Black is fine). Then after the further error 15.g5?! Nfd7 16.a5 b4 17.Nd5 Bxd5 18.exd5 Rb8!,
with the idea of ...Rb5, simply capturing the a5-pawn, Black stood better. White’s kingside is also weakened and
vulnerable to the advance of Black’s f-pawn.
b2b) 9.Nd5 is an idea that Alexander Ivanov has used with success. In particular, he recognized that despite the move
g2-g3, the bishop need not necessarily be fianchettoed. For example, 9...Nxd5 10.Qxd5 Rb8 11.Nc3 Bb7?! (Despite the
obvious nature of this move, it is probably premature. In Ivanov – Najer, Moscow 2004, Black achieved an acceptable
position after 11...Be7 12.Bg2 [Here 12.Bc4 is not as good. After 12...0-0 13.Be3 Nc5 Black threatens ...Bc8-e6, the
exchange Bxc5 should not promise much, and on 14.Qd2, 14...b5! is good for Black.] 12...0-0 13.0-0 Bb7 14.Qd1 Rc8
15.Qe2 Nf6.) 12.Qd1 h6:

Position after 12...h6 (analysis)

13.Bc4! Nf6 14.Qd3 Ra8 15.f3 and White had a good grip on d5 and a pleasant advantage in Ivanov – Stolerman,
Marlboro 2002.

8.Bg2 b5

The inclusion of ...Be7 and Bg2, before all the above play, changes a few things. In general, I tend to think that trying
to unravel these things should not be part of chess! It should be about ideas, plans, combinations, structures – everything
but twisting our minds trying to find out if it is better to play ...b5 or ...Nbd7 before or after ...Be7. But the people would
really like to know...
If Black reverts to playing 8...Nbd7 (after 7...Be7 8.Bg2), then Ivanov’s Bc4 plan (see above) is avoided by this move
order, while at the same time the early ...Be7 means that a white knight entering d5 or f5 might be able to capture the
bishop. Black’s queenside play is delayed by one move in some cases. But there is not a single logical reason why
moving a bishop from its home square to e7, making room for castling, could be an error or even an inaccuracy.

In most games, White leaps into this square. Otherwise, after e.g. 9.h3 Nbd7 10.g4 b4 11.Nd5 Nxd5 12.exd5 a5,
Black gets good play, as we have seen before.


Another possibility is 9...Nxd5 10.Qxd5 Ra7 (Black exchanges the knight before it can be supported by Nec3, but has
to make this slightly awkward rook move). Now 11.Be3 Be6 12.Qd2 Rb7 13.Nc3 Nd7 14.Nd5 was reasonably solid for
Black, but White had a slight advantage in Fressinet – Gharamian, Cap d’Agde 2015.


Black is not concerned about White’s capturing on e7. The exchange of this knight, which moved several times, for
the passive bishop on e7, would give White the two bishops but allow Black to win the battle for d5 and e4.


10...Nb6 is another, probably better, possibility. 11.a4 now leads nowhere because of 11...Nfxd5 12.Nxd5 Nxd5
13.Qxd5 Be6. White should play 11.Nxe7 Qxe7 12.Bg5 h6 (12...Bb7!?) 13.Bxf6 Qxf6 14.a4 bxa4 15.Nxa4 Nxa4
16.Rxa4 with a small advantage in Giri – Wojtaszek, Qatar 2015.


White would have little chance for advantage by only using the disputed square d5. Weakening Black’s queenside is
the only way forward.


Black begins exchanging pieces on d5.

In Lutikov – Magerramov, Moscow 1983, Black simply gave up a pawn by 11...0-0 12.axb5 axb5 13.Rxa8 Bxa8
14.Nxb5 Bxd5 15.exd5 Qa5+ 16.Nc3, and went on to win, although he did not have enough compensation at this point.

12.Nxd5 Nf6
Black challenges the d5-knight at once. In the game Adams–Navara, played a month later at Biel, Black preferred
12...0-0 13.axb5 axb5 14.Rxa8 Bxa8 15.0-0, and White won in Michael Adams’s trademark creeping positional style.
For the rest of the game, see Adams–Navara in the supplemental games.


Since Grischuk basically had an okay position after this, more promising was 13.Nxe7, when Black suffered greatly
but eventually held the draw after 13...Qxe7 14.axb5 axb5 15.Rxa8+ Bxa8 16.Bg5 0-0 17.0-0 h6 18.Bxf6 Qxf6 19.Qd3
Bc6 20.Rd1 Rd8 21.Qc3 Rc8 22.Qb4 Rd8 in Malakhov – Kuzubov, Chartres 2017.


Black does not give White another chance. Now a white pawn will be on d5 and the b5-pawn is easily held.

14.exd5 axb5 15.Rxa8 Qxa8 16.Qd3

16...0-0 17.0-0

Of course, 17.Qxb5 is too risky. Black plays 17...Qa1 and if 18.0-0 then 18...Ba6 and Black wins the exchange.


The b7-bishop is restricted by the d5-pawn, so Black transfers it to d7, where it solidly defends b5.

18.Be3 Bd7 19.Rc1 h6

Black is completely fine, but the position is roughly equal, and the game goes on. Carlsen is very dangerous in such

20.Qd1 Qb7 21.Ra1 Ra8 22.h4

This space gain, and making a convenient square for the king on h2, was also used in Adams–Navara. A later ...Be7-
g5 is also prevented.

22...Ra6 23.b3 Bd8

The transfer of this bishop to challenge its white counterpart is entirely natural. In the ensuing play, the dangers are
extremely slight, but start to build slowly. The only threat to Black is White’s possibility of c2-c4 and c4-c5, creating a
passed pawn. In view of this, it was possible to play 23...Rxa1 24.Qxa1 b4, crippling White’s majority. Naturally, the
b4-pawn is very weak and can be attacked by the queen and dark-squared bishop, but c2 is also weak. For example,
after 25.Qa5 Bf5, Black is comfortably holding.

24.c4 bxc4 25.bxc4 Bb6 26.c5!

The only chance to play for a win. White sacrifices a pawn, creating a passed pawn which will tie down some black

26...Bxc5 27.Bxc5 dxc5 28.d6 Qb6

28...Rxa1 29.Qxa1 e4 is met by 30.Qe5, when Black still faces problems.

29.Rxa6 Qxa6

The d-pawn will not get through, so White hopes to create problems with an attack on f7. Grischuk was, as usual, in
serious time pressure.


More accurate is the immediate 30...Be8, defending f7 solidly but passively and challenging White to create
something. The queen stays on a6 for the moment, preventing White from leaving the d6-pawn. Objectively White has

31.Qb3 Be8 32.Qc3

White wins back the pawn and retains his passed pawn. Yet the lack of material means that Black still has good
chances to hold.

For now, Grischuk defends well in time pressure. It is better to give up the c-pawn rather than allow the white queen
to come to e5. If White could play Qxe5 and block the c-pawn by Bd5-c4, Black would be in danger of running out of


The queen exchange gives White no winning chances, naturally, while 33.Qxe5 c3 and the c-pawn becomes too
dangerous for White to do anything.

33...Bd7 34.Qb3 Qe8

Still holding. Black has found a way to securely block the d-pawn and defend f7. White will need new threats to
create problems.

35.Qf3 Kf8 36.h5 Kg8

Black now starts to wait, challenging White to make a breakthrough. It would be hard, under time pressure, to play
36...e4, when the pawn might become weak.

37.Qe4 Bc6 38.Bd5 Bd7

The exchange of bishops would make the passed pawn too strong. After 38...Bxd5 39.Qxd5 Black would have to give
up the e-pawn rather than allow the white pawn to d7 (with Qd6 to follow). The resulting position is won for White.

39.Kg2 Kh8 40.f4

Black has refused to budge, so on the fortieth move, with Grischuk’s flag (metaphorically) hanging, Carlsen takes his


And Grischuk cracks under the pressure. It was necessary to play 40...f6 41.fxe5 fxe5, when Black still faces a tough
defense, but with good chances to hold. The white king is rather open now, so it will be hard to put pressure on Black
without allowing counterplay.

41.Qxe8+ Bxe8 42.Bxf7!

Grischuk might have missed this move in time pressure, or misevaluated the resulting endgame.

42...Bc6+ 43.Kf2 fxg3+ 44.Kxg3 Bd7

If 44...g6 then White will capture 45.hxg6, and not with the bishop, after which Black could afford to give up the
bishop for White’s d-pawn.

Black resigned, as White will win the bishop for the d-pawn, and the resulting position will be winning for White
despite the h-pawn’s being of the wrong color. The presence of Black’s g7-pawn will allow White to win by using
Zugzwang to force the advance of that pawn. For example: 45...Kg8 46.Kf4 Kf8 47.Ke5 Kg8 (Black can only wait –
there is no preventing the journey of the white king to c7) 48.Kd5 Kf8 49.Kc5 Kg8 50.Kb6 Kf8 51.Kc7 Bg4 52.d7
Bxd7 53.Kxd7 (without the g7-pawn, the h6-pawn, or both, the position would be drawn) 53...Kg8 54.Ke7 Kh8
55.Ke8 Kg8 56.Bf7+ Kh8 57.Kf8 Kh7 58.Bg6+ Kh8:

59.Bb1 (finally Zugzwang is achieved) 59...g5 60.hxg6 (Barring the h6-pawn, this position would be stalemate. But with
it...) 60...h5 61.g7#.


Supplemental Games:
Arkadij Naiditsch – Sunilduth Lyna Narayanan
Isle of Man 2016
1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.g3 e5 7.Nb3 Be7 8.Bg2 0-0 9.0-0 Be6 10.a4 Qc7 11.Re1 Nbd7
12.Nd2 Qc5 13.Nf1 Bd8 14.h3 Ba5 15.Be3 Qc7 16.Bd2 Rac8 17.Ne3 Rfe8 18.g4 h6 19.Nf5 Nf8 20.Qf3 N6h7
21.Red1 Bb6 22.a5 Ba7 23.Be3 Bxe3 24.Qxe3 Red8 25.Qb6 Qxb6 26.axb6 Rd7 27.Rd2 Rc6 28.Na4 Bc4 29.Rad1 Bb5
30.Nc3 g6 31.Rxd6 Rxc3 32.bxc3 gxf5 33.exf5 Nf6 34.c4 Bxc4 35.Bxb7 Rxb7 36.Rxf6 Kg7 37.Rfd6 Bb5 38.Rc1 Bc4
39.Rb1 Bb5 40.Rc1 Bc4 41.Rb1 Bb5 42.h4 Nd7 43.c4 Bxc4 44.Rxd7 Rxd7 45.b7 Rxb7 46.Rxb7 Be2 47.Rb3 e4 48.Rg3
Kf6 49.g5+ hxg5 50.Rxg5 a5 51.Kh2 a4 52.Rg8 Kxf5 53.Kg3 f6 54.Ra8 Bd1 55.Ra5+ Kg6 56.Kf4 Bc2 57.Ra6 Kg7
58.Rc6 Bb1 59.Rc1 Bd3 60.Rc3 Bb1 61.Kg4 Kg6 62.h5+ Kh6 63.Rc6 Kg7 64.Rc7+ Kh8 65.Ra7 Bc2 66.h6 e3 67.fxe3
Bg6 68.Rxa4 Kh7 69.Rf4 1-0

Michael Adams – David Navara

Biel 2015
1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.g3 e5 7.Nde2 Be7 8.Bg2 b5 9.Nd5 Nbd7 10.Nec3 Bb7 11.a4
Nxd5 12.Nxd5 0-0 13.axb5 axb5 14.Rxa8 Bxa8 15.0-0 Nc5 16.b4 Ne6 17.h4 Bxd5 18.Qxd5 Nd4 19.c3 Ne2+ 20.Kh2
Nxc1 21.Rxc1 Qb6 22.Kg1 Rb8 23.Ra1 g6 24.Ra8 Rxa8 25.Qxa8+ Kg7 26.Bf1 Qc7 27.Qe8 Bf6 28.Bxb5 Qxc3
29.Bc6 Qf3 30.Qd7 g5 31.hxg5 Bxg5 32.Qxd6 Qc3 33.Kg2 Qb2 34.Qc5 Qd4 35.Qxd4 exd4 36.f4 Be7 37.b5 Bc5
38.e5 d3 39.Be4 d2 40.Bc2 f6 41.exf6+ Kxf6 42.Kf3 h5 43.Ke2 Bd6 44.Ke3 Bc7 45.Bd1 h4 46.gxh4 Kf5 47.Bc2+
Kg4 48.Ke4 Kxh4 49.Bd1 Bd8 50.Kd5 Kg3 51.f5 Kf4 52.Ke6 Ke4 53.Bc2+ Kd4 54.Kd7 Bb6 55.f6 Kc3 56.Bd1 1-0
Chapter 7
Healthy Aggression: 6.f4

The move 6.f4 walks the line between the direct aggression of moves like 6.Bg5 or 6.Bc4, and the classical positional
methods of 6.Be2. After 6...e5, White tries to combine threats on the kingside – often via the opening of the f-file or the
use of the f5 square – with attempts to exploit traditionally weak central squares like d5 or the weak pawn on d6. In this
variation, White’s light-squared bishop most often goes to d3 and the knight retreats to f3, behind the f-pawn.
The downside of White’s play is that the king will never be as safe on the kingside as it would be with three pawns
covering it. Chaos favors the side with the safer king, and chaos is native to the Najdorf. White’s position can become
overextended, with Black often gaining a base of operations (for a knight) on e5.
Game 20 covers the traditional Najdorf move 6...e5. In this game, we see White carry through his thematic attacking
ideas, including an exchange sacrifice on f6. In the notes to that game, we can see some safer ways for Black to reach a
playable position.
Game 21 is a fascinating correspondence game featuring 6...Qc7, Black’s main alternative to 6...e5 and one favored
by many Najdorf devotees. Complications and tactics proliferate in the adventurous game.

Game 20
Dimitri Reinderman – Loek Van Wely
Dutch Chp, Rotterdam 1999

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.f4 e5

The true Najdorf response, gaining a foothold in the center, although Black has many other moves. Besides the
Scheveningen-style 6...e6, there is 6...Qc7 (see Game 21) as well as 6...Nbd7.

7.Nf3 Nbd7

The other move played here is 7...Qc7. These moves can easily transpose, but there are some significant differences.
For one thing, 7...Qc7 would prevent the radical method White used in the game Volokitin–Harikrishna (see below) –
although this is definitely not something Black should worry about. More importantly, 7...Qc7 prevents White from
playing Bf1-c4 rather than Bd3. On the other hand, Black loses some options by committing the queen to c7 early on. In
particular, the options on move ten of 10...Nc5 11.Kh1 exf4 12.Bxf4 Bg4, and 10...exf4, meeting 11.Bxf4 with
11...Qb6+, are both unavailable since they do not involve the early ...Qc7 move.


White normally prevents ...b7-b5 at some early point. In the game Volokitin – Harikrishna, Cap d’Agde 2006, White
won with the wild 8.g4?!. However, this thrust borders on the unsound. After 8...Nxg4 9.Bc4 Nb6 10.Bb3 Be7
11.Ng5?! 0-0 12.f5 Nf6 13.a4 h6 14.h4 d5! 15.a5 d4 16.axb6 dxc3 17.Qf3 Bb4?! (17...Qxb6 is just winning for Black:
after 18.bxc3 Qc6!, for example, Black threatens 19...hxg5 20.hxg5 Qxe4+; White has no attack on the kingside) and
the further 18.bxc3 Qd4?? (A pretty move, but a gigantic blunder. Black forks his own pieces! Instead 18...Bc5 was still
much better for Black.), White won a piece with 19.Bd2! and went on to win the game.

8...Be7 9.Bd3

White would like to place the bishop on the more active a2-g8 diagonal with 9.Bc4 and also fight to use the d5-
square. However, there are some problems with the “loose” position of the bishop – primarily 9...Qa5, when Black
threatens 10...Nxe4, while 10.0-0 is not possible due to 10...Qc5+. Nevertheless, this is a crucial line. White plays
10.Qe2 b5 11.Ba2 bxa4 (11...b4 12.Nd5 is good for White) 12.0-0 0-0 13.Kh1. To me, this line looks pleasant for
White. For example, 13...Bb7 14.Nh4 Rae8 15.Nf5 Qb4 16.fxe5 Nxe5 17.Bd2 Bd8 18.Nd5 Qxe4 19.Nxf6+ Bxf6
20.Qxe4 Bxe4 21.Nxd6 Bxc2 22.Nxe8 Rxe8 was the course of Dolmatov – Ftáčnik, Moscow 1985. Black did not have
quite enough for the exchange.
After 9...Qa5 10.Qe2, Black has also played the more solid 10...0-0 11.0-0 exf4 12.Bxf4 Ne5, and here Black held
the draw after 13.Nd5 (more promising seems to be 13.Bb3 Bg4 14.Kh1 Nxf3?! [Black should wait to make this
capture, for example with 14...Rac8] 15.gxf3 Be6 16.Rg1, when White had good attacking chances in Smirin – Ehlvest,
Khanty-Mansiysk 2009) 13...Nxd5 14.Bxd5 Be6 15.Bxe6 Nxf3+ 16.Qxf3 fxe6 17.Qb3 Qc5+ 18.Kh1 Qc8 in Lékó –
Kasparov, Sarajevo 1999.

9...0-0 10.0-0 Qc7

Black has a very solid line here which has no doubt dissuaded many from playing 6.f4: 10...Nc5 11.Kh1 exf4
12.Bxf4 Bg4, with the idea of strengthening the kingside with ...Bh5-g6. The play is around equal. For example, 13.Qd2
Bh5 14.Nd4 Bg6 15.Nf5 Re8 16.Rae1 Bxf5 17.exf5 d5 18.Be5 d4 19.Bxf6 Bxf6 20.Rxe8+ Qxe8 21.Nd5 Qc6
22.Nxf6+ Qxf6 was Bilguun – Sandipan, Chengdu (China) 2017. The position is equal.
Black can also begin with 10...exf4, when 11.Kh1 Nc5 transposes, although White might try the pawn sacrifice
11.Bxf4!? Qb6+ 12.Kh1 Qxb2 13.Qe1 Qb6 14.Nd5 Nxd5 15.exd5 Qd8, when Black’s position is not particularly easy
to play, although most likely White does not have enough for a pawn.


This move normally is played at some point. The king will inevitably have to be removed from the a7-g1 diagonal, so
White does it immediately, while staying flexible otherwise. This position can easily arise after 7...Qc7.


This allows the dangerous exchange sacrifice which happened in the game. There are various ways for Black to
challenge White to carry out his thematic attacking ideas on the kingside, rather than releasing the tension in the center.
For instance, the 11...Nc5 12.Qe1 Re8 13.fxe5 dxe5 14.Qg3 Bd8!? of Solleveld – Giri, Netherlands 2012, looked
strange and provocative, but White made no headway on the kingside after 15.Nh4 (15.Bg5!?) 15...Nh5 16.Qf3 Nf4
17.Bxf4 exf4 18.Nf5 Bxf5 19.exf5 Re3, when Black stood well.
Alternatively, after 14.Bg5 Be6 15.Nh4 Nh5! 16.Bxe7 Qxe7 17.Nf5 Qg5 in Vignesh – Zhou Jianchao, Ho Chi Minh
City 2017, Black was controlling the area in front of his king while the knight’s impending entry into f4 was annoying
for White. The ...Nh5 move is an important idea, aiming at f4 and allowing some strengthening of the black kingside.
12.Bxf4 Ne5

The logical follow-up to Black’s previous move.


The 6.f4 line is not really just a blunt attacking weapon, and often has some similarities to the positional 6.Be2
variation. Here White is aiming at the d5 square. The threat is 14.Nxe5 dxe5 15.Bxf6 Bxf6 16.Nd5. At the same time,
White prepares to open the black kingside via an exchange sacrifice.


Black therefore has to defend the d5 square.

14.Nxe5 dxe5 15.Bxf6 Bxf6


This thematic exchange sacrifice breaks open Black’s kingside. White will not be able to checkmate Black directly,
but Black will suffer from a number of weak pawns and a lack of counterplay, and White will create passed pawns in
the center as well.

16...gxf6 17.Qh5?!

This move is very natural, but in view of the next note White should seriously consider 17.Nd5, inducing Black to
capture the knight since 17...Qd8 18.Qf3 allows White to take the f6-pawn quickly, and the black rook won’t get to the
g-file as in the next note. After 17...Bxd5 18.exd5, play would continue as in the game, while avoiding the next note. If
now 18...Kh8, then 19.Qh5 with a big advantage.

This move, planning to defend the h7-pawn by the uncomfortable ...Rh8, does not seem right. It seems that Black can
play 17...Kh8, planning to bring the rook to the g-file. Now if 18.Nd5 Qd8!, when Black has ideas of ...f6-f5, for
example in response to 19.Rf1. Then on 19.Qh6 Rg8 20.Nxf6 Rg7, the black king is well defended, while the white
pieces are awkwardly placed. White does not have compensation for the exchange here.

18.Rf1 Rh8 19.Nd5

This exchange sacrifice had been played several times before. The first was the game Glek – Malisauskas, Tallinn
1986, which continued with 19.Qh4 Qe7 and only now 20.Nd5 Bxd5 21.exd5, which was also good for White. The
continuation was 21...Rad8 22.c4 a5 23.Be4 b6 24.b4! axb4 25.a5! Rd6 (25...bxa5 26.c5 and the connected passed
pawns, in conjunction with the attack on the black king, make the defense practically impossible) 26.axb6 Rxb6
27.Qg3+ Kf8 28.Qe3 Rb8 29.c5 h5 30.d6 Qe6 31.Qd3 b3 32.c6, and White soon won.

19...Bxd5 20.exd5


It is not easy for Black to arrange a defense. With the rook tied down to h8 and the black rooks in general lacking
useful open files, in addition to White’s passed d-pawn and the weakened position of the black king, White must have
enough compensation. For example, if 20...Rad8 21.c4 a5 22.Qg4+ Kf8 23.Rxf6 Rg8 24.Qf3 Rg7 25.g3 White has not
managed to break through yet, but Black stands entirely passively. Nevertheless, this was preferable to what Black

21.c4 a5

The normal way of impairing White’s 4-2 majority on the queenside.

22.Qg4+ Kf8 23.Qh4

In comparison to the previous variation, when White takes on f6 he will now gain a tempo, in addition to the one
Black lost by playing ...Qd6 in the first place. This means that defending f6 is obligatory.

23...Kg7 24.c5 achieves this key advance: 24...Qd8 25.d6, and White has a huge advantage.


Not just eventually winning back the material, but also opening crucial lines.


If 24...Qxc5, then 25.Bxa6 bxa6 26.Qxf6 and White wins immediately.


25.Bxa6 Rg8! followed by ...bxa6 is not as clear.


A better chance was offered by 25...Qd8, when White must be very accurate not to let Black get away. Best seems to
be 26.Bxa6 bxa6 27.Qh6+ Ke8 28.Qg7! Rf8 29.h3, when White will take f6 and Black remains very tied up. For
example, 29...Qd5 30.Rxf6 Qxc5 31.Qxh7 and the black king has no safety, nor can Black use the f8-rook. Meanwhile
the white king will sit in peace on h2, the queen will come to f5, and the h-pawn will advance.

26.Qh5! Ke7

26...Qc7 27.Qh6+ Ke8 28.Bxa6 bxa6 29.Qxf6 is hopeless for Black.

27.Qxf7+ Kd8 28.Bxa6 bxa6 29.b3?!

There were more accurate moves here. For example, 29.h3 wins easily, since White just captures on f6 with the rook
and commences the mating attack.
29...Kc8 30.Rxf6 Qxc5 31.Qe6+ Kc7 32.Rf7+ Kb8 33.Qxa6

Black could play on with 33.Qxa6 Qc1+ 34.Rf1 Qc5, although down a pawn with the king open, the black position is
pretty grim, so Black resigned.


Game 21
Ger van Perlo – Grigory Sanakoev
“Serbia 30” Correspondence, 1983-85

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.f4 Qc7!?

The standard Najdorf move is 6...e5, but many players prefer to play something more crooked, rather than going into
the solid (but sometimes tricky) positions that result from 7.Nf3.


This is White’s normal setup in the 6.f4 line: the bishop goes to d3 and the knight to f3, and White prepares a kingside
attack by Qe1-h4 and f4-f5. The advantage of Black’s 6...Qc7 is flexibility. If White reacts in a different way, Black can
use a variety of setups, keeping his options open for ...g7-g6, ...e7-e5, ...b7-b5, and both developments of the b8-knight.
For example:
7.a4 prevents ...b7-b5, but now Black usually switches to a Dragon setup with 7...g6. See the supplemental game
7.Qf3 is an interesting move. White attempts to reach a type of position, after 7...e6 8.Be3, with 0-0-0, Bf1-d3, and
g2-g4-g5 to follow, where...Qd8-c7 is premature and might hinder an exchange sacrifice on c3. Black usually also
meets this by switching to a type of Dragon with 7...g6, for example 8.Be3 (8.f5!?) 8...b5 9.Bd3 Nbd7 10.0-0 Bb7 11.a3
Bg7 12.Kh1 0-0, with a harmonious setup for Black in Stefánsson – Kasparov, Reykjavík 1995.
7.Nf3, threatening an early e4-e5, is often met by 7...Nbd7 8.Bd3 g6, heading into the typical formation. However,
here Black can also use his flexibility to avoid falling in with White’s plans. 7...e6 is a good move, switching to the
Scheveningen structure after White’s Nd4-f3. White’s attacking setup is less effective after this than after the fianchetto,
where f4-f5 becomes more dangerous. Following 8.Bd3 Nbd7 9.Qe2 Nc5 10.0-0 Be7 11.Kh1 0-0 12.e5 Nd5!, Black
stood well in Atabayev – Grischuk, Baku 2015: since the isolated queen’s pawn position after 13.Nxd5 exd5 is fine for
Black in view of White’s weakening f2-f4.
7.Be2 abandons White’s usual setup with Bd3 and Nf3 and instead tries to reach a Scheveningen. However, Black
can play in a number of good ways, such as 7...e5 or 7...b5 8.Bf3 e5, when White’s pieces are not as harmoniously
placed as when the bishop goes to d3 and the knight retreats to f3.


More often, Black has adopted a development with 7...g6 and 8...Bg7, when the knight retreats to f3. White then plays
0-0 and Qe1-h4, normally taking a tempo to prevent ...b7-b5 with a2-a4. These positions are certainly playable for
Black, but I like Sanakoev’s rarer, sharper, but I believe sound approach. The ...b7-b5 advance is achieved before White
can prevent it, and Black puts immediate pressure on e4, which interferes with White’s play.


White continues his development in a normal way. The alternatives are:

8.a4 is a standard response to an early ...b7-b5, but here it does not mix so well with the f2-f4 plan. After 8...b4 9.Na2
e5 10.Nf3 Nc6, Black had gained space and confused the white pieces in Levy – Quinteros, Cienfuegos 1972.
8.Qe2 threatens 9.e5, but commits the queen to e2, when in many situations it goes to e1 and then h4. Black can reply
8...e5, when White has lost options and Black has achieved ...b7-b5, in contrast with the main 6...e5 lines. Black can
also play 8...Nbd7, heading for the ...g6-g7 and ...Bf8-g7 setup, also having achieved...b7-b5 which White normally
8.Nf3 would likely end up transposing to the game after 8...Nbd7 9.0-0, but of course after this move Black can also
consider switching to a Scheveningen structure with 8...e6.


Black again stays flexible and holds up an eventual threat of e4-e5.


This move is standard in the 6.f4 system. White threatens to push e4-e5 and therefore induces the ...e7-e5 advance,
allowing White to continue with his normal attacking schemes on the kingside. However, Black’s reply creates
immediate complications. Naturally, White has other choices.
9.Kh1 is a logical and flexible move, taking the king off the diagonal which – one assumes – he would inevitably
need to leave anyway. After 9...Bb7 10.Qe1 g6 11.Nf3 e5 White has achieved his objective of inducing ...e7-e5, while
the queen is on e1 and can go to h4. Nevertheless, Black has achieved ...b7-b5, whereas that move would normally be
prevented by a2-a4, and generally enjoys enough play. For highly instructive play by a classic maestro of the Najdorf,
see the supplemental game Adorján–Browne.
9.a3 stops the early ...b5-b4, but this kind of passive move is hardly the kind to terrify Black. Now 9...Bb7 10.Kh1 g6
11.Qe1 Bg7 12.Nf3 e5 13.fxe5 dxe5 14.Bd2 Nc5 was the course of Dussol – A.Čolović, France 2007. Black has a good


This disruptive thrust allows Black to attack e4 at once and leads to very complicated play within just a few moves.

10.Nd5!?, sacrificing a pawn, has never been played but is worth consideration. After 10...Nxd5 11.exd5 Black can
take a pawn with 11...Qc5+ (Black could also play 11...Bb7, when 12.a3 is also unclear) 12.Kh1 Qxd5, when after
13.Qe2 White has reasonable compensation, with Bd3-c4, Nf3-g5, and f4-f5 coming.

10...Bb7 11.Ng3


Black judges that his king won’t be safe on the kingside in any case, since White has two knights and some other
pieces in the vicinity. Therefore, he resolves to leave the king in the center and to open lines on the kingside, so that the
black king won’t be the only one in danger.
Earlier, Zoltán Ribli had twice faced this position and preferred the calmer 11...e6, which is not without its merits. For
example, 12.Qe2 Be7 13.Bd2 0-0 (in an earlier game against Victor Ciocâltea in the same tournament, he preferred the
sharper 13...h5 14.Ng5 h4 15.Nh1 d5?! 16.e5 Nh5 17.Qg4, when White had the initiative in Ciocâltea – Ribli,
Bucharest 1971) 14.Kh1 a5, with a double-edged position in, Ghinda – Ribli, Bucharest 1971.


In the only previous game, Stean – Rajković, Hastings 1972, White played 12.Nh4, which was a very awkward way
to prevent ...h5-h4. After 12...e6 13.f5 e5, Black had a good game.

12...h4 13.Nf5

White naturally prefers active play to the passive 13.Nh1.


The impudent pawn ensures that the white king will never be safe. Either the white king’s defenses will be destroyed
or the long diagonal will be seriously weakened.

White begins the attack. After instead 14.g3 g6 15.N5d4 e5 16.Nb3 a5, Black has a great game. The far-flung pawn
on h3 will be a problem for White for the rest of the game.

14...dxe5 15.fxe5


Black cannot play 15...hxg2 due to 16.Nd6+!, e.g. 16...Kd8 17.Nxf7+ Kc8 18.Qxg2, and Black loses material.
Therefore, he has to give up this important bishop. A series of forcing complications begins.

16.Qxf3 Qa7+

Black is able to protect the rook with gain of tempo, and now captures the e5-pawn.

17.Be3 Nxe5 18.Qg3 Qb7!

Black’s play depended on this. The e5-knight is immune due to checkmate on g2, while Black threatens to play
19...Qxg2+ eliminating the queens and entering the endgame with an extra pawn.


White preserves the queens and his main chance for attacking the black king. For this, however, he must give up the
important light-squared bishop, while his own king is also open. After the alternative 19.Be2 Qxg2+ 20.Qxg2 hxg2
21.Kxg2, White has some play thanks to his two bishops and Black’s undeveloped state, but bear in mind that Black is
not only up a pawn but has two connected passed pawns, so all the chances lie with Black.


Black gladly removes this dangerous bishop.


After the storm has passed, Black is left with a winning endgame, if he can reach it. White’s wrecked pawn structure
is tantamount to an extra pawn for Black. However, the active knight on f5 and White’s open lines mean that the
position still hangs by a thread.


This is much better than 20...g6 21.Nd6+ exd6 22.Bd4, according to Sanakoev, when the black king is again in


Simplification in this way allows Black to consolidate his advantage. White needed to seek chances with a knight
sacrifice. 21.Rae1 Rd8! leaves White with no way forward, while the knight will soon be hanging in earnest. But
21.Rac1!? is very complicated – White is invading on c7. On 21...exf5 22.Rc7 Qb5 23.Rfc1 the position is quite tricky.
Black must continue to defend e5, while White is planning Be3-c5. It appears that Black is better if he finds 23...f4!!,
when after 24.Bxf4 Be7 or 24.Qxf4 Rh5! Black succeeds in defending all the key squares and activating his pieces.

21...Bxd6 22.Qxd6 Rd8 23.Qg3 Rxd3 24.Rac1

Black has won a pawn and the white pawn structure is broken up, with the white king exposed and the e3-bishop
pinned. Yet the game is undecided nevertheless. The black king cannot easily find refuge, since 24...0-0 is met by
25.Rxf6. Meanwhile, White plans Rc1-c7 and Rac1. Sanakoev therefore resolves on a clever move.


Black uses this unique rook lift, threatening to pin the queen with ...Rg6 to achieve the exchange of some more pieces,
even though it ends up costing a pawn. Black could have also played 24...Nh5, when White must agree to the exchange
of queens by 25.Qc7, since on 25.Qe5 0-0 Black succeeds in bringing the king to safety, with a large advantage.
However, the endgame after 25...Qxc7 26.Rxc7 Nf6 27.Bc5 is not very clear, as the white pieces are active.
25.Rc7 Qa8 26.Rfc1

26.Bxh6 is not good due to 26...Rxg3+ 27.hxg3 Qd8 (Sanakoev), when 28.Bf4 Qd4+, and the centralization of the
queen is decisive.

26...Rg6 27.Rc8+ Rd8

Black can force the queens off by 27...Qxc8 28.Rxc8+ Kd7 29.Rc7+ Kd8 30.Qxg6 fxg6, but after 31.Bb6 Ke8
32.Rxg7 things are not clear. Black should stand better, but various black pawns are also weak, giving White serious
hopes for a draw.


White would like to destroy the black pawns and take over the c-file for the endgame. After 28.Rxa8 Rxg3+ 29.hxg3
Rxa8, now 30.Rc7 is met by 30...Nd5, while otherwise Black has excellent winning chances in the endgame.


Now the queens will remain on the board. Not 28...fxg6? 29.Rxa8 Rxa8, when 30.Rc7 is possible in view of 30...Nd5
31.Rxg7. With the active rook on the seventh and various weak black pawns, White has nothing to worry about.

29.Rxc8+ Qxc8 30.Qxg7 Ke7

Another quiet harbor has been reached. Given White’s doubled and isolated h-pawns, Black is essentially up a pawn.
If the queens are traded, then Black should win, having connected and passed pawns. However, with the queens on the
board, White has counterplay on the dark squares. The black king will still not find any safe place.


White pins the knight and creates threats of Bc5+ or Qc5+. The obvious 31.Bg5 would not be very effective after
31...Qc5+ 32.Kf1 Qf5+, when Black will take the b2-pawn after some more checks, defending the knight on f6, and
should win.
31...Qd8 32.Qc5+?

This is not the strongest. White has to centralize his pieces with 32.Bc5+ Ke8 33.Qe5, when White’s well-placed
pieces make it hard for Black to play for a win.

32...Qd6 33.Qc4?!

White need not retreat, since the capture of the queen would result in the loss of the b4-pawn. Therefore it is possible
to play 33.h4. Then, after 33...Nd5 34.Qxd6+ Kxd6 35.h5!, the endgame is reached that Black avoids by his 36th move
in the game. The white pawn has taken one more step, which should allow the first player to save the game.

33...Kd7 34.Bc5 Qd1+

The exchange of queens is not practically forced, since after a move by the white king, 35...Nd5 will give Black a
direct attack with the queen and knight combined.

35.Qf1 Qxf1+ 36.Kxf1 b3!

A key subtlety. Black gives away this pawn, but White’s resulting doubled b-pawns are not really better than one
pawn. This is much better than 36...Nd5 37.h4, when the pawn will reach h5 and divert the black knight enough to save
the game.

37.axb3 Kc6 38.Bd4

After other bishop moves, the black king will nevertheless reach the strong and invulnerable square e4.

38...Nh5 39.Kf2 Kd5 40.Bc3

A better chance is to prevent the black king from reaching e4 with 40.Ke3, although after 40...e5 41.Bb6 f5 42.h4
Nf6, with the idea of ...f5-f4+ followed by ...Ke6-f5, Black should reach a similar position to the game eventually.

40...Ke4 41.h4 e5 42.b4 f5 43.Be1 Kf4 44.h3 Ke4 45.Bc3 f4 46.b3

White is nominally up a pawn, but this is irrelevant as the doubled h- and b-pawns are each not worth more than one
pawn. The question is whether Black can advance his passed pawns without allowing the white h-pawn to advance,
which would tie down the knight.
Here the game was adjudicated, as sometimes happened in correspondence chess in those days. Below are some of the
complex and study-like variations submitted by Sanakoev to show a win, along with my occasional notes.

Position after 46...Nf6 (analysis)

Black plans ...Kf5 followed by the advance of the e-pawn, supported by ...Nd5 if necessary. First, we should consider
whether White can play actively with the king:
47.Ke2 Kf5 48.Kd3 Nd5 49.Be1 e4+ 50.Kd4, and here Black wins by 50...e3 51.h5 f3 52.Kd3 (or 52.h6 f2 53.Bxf2
exf2 54.h7 f1=Q 55.h8=Q Qa1+) 52...Kf4 53.h6 f2 54.Bxf2 exf2 55.Ke2 Kg3 56.h7 Nf4+ 57.Kf1 Ng6.
If 47.Bd2, then 47...Kd3 and the bishop has no good squares.
White therefore has to wait. This can be done in several ways. For instance:
a) 47.Ba1 Kf5 48.Bb2 e4 49.Bd4 (or 49.Bc1 Nd5 50.h5 Nxb4 51.h6 Nd3+ 52.Kg2 Kg6) 49...Nd7 50.Ke2 Ne5
51.h5 Nf3 52.Bc5 Kg5 53.Bd6, and now Sanakoev gives 53...Ng1+, which does not win after 54.Kf1 Nxh3 55.h6
Kxh6 56.Kg2. However, 53...Nd4+ 54.Kf1 f3 is stronger and Black wins.
b) 47.Bb2 Kf5 48.Ke2 e4, and now:
b1) 49.Bd4 Nd7 50.Kf2 Ne5 51.Bc5 Nd3+ 52.Kg2 (52.Ke2 f3+ 53.Ke3 Nf4) 52...Ne1+ 53.Kf2 Nc2 54.h5 (54.Ke2
e3 55.Kd3 Ne1+ 56.Kd4 e2) 54...e3+ (54...Kg5 is better than Sanakoev’s move. Then after 55.Bd6 f3, Black will
capture the h-pawn and should win.) 55.Kf3 Ne1+ 56.Ke2 Ng2 57.h6 Ke4 58.Bxe3 (58.Kf1! holds the draw, e.g. if
58...f3 59.Bxe3 Nxe3+ 60.Kf2 Nd1+ 61.Kf1 and Black must keep checking with the knight; or if 58...Kf3 then
59.Kg1! anticipates ...e3-e2+. The black knight will have to retreat to deal with the h-pawn, and it will not be able to
support the advance of the black pawns.) 58...fxe3 59.h7 Nf4+.
b2) 49.Bc1 Nh5 50.Kf2 Ng3 (this move is given by Sanakoev; however, it seems to me that this knight maneuver
leads nowhere, and Black should instead prefer 50...Ke5 51.Bb2+ [...Kd4 must be prevented] 51...Kd5 52.Ba1 f3!,
planning the invasion of the knight by ...Nh5-f4 – for example, 53.Ke3 Nf4 and Black is winning) 51.Bb2 (51.Bd2
Nh1+ 52.Kg2 e3 is given by Sanakoev, but after 53.Bc1 there is no win evident. White intends to capture the knight
after 53...Ke4, while after 53...Nf2 there is 54.Kf3 followed by sacrificing the bishop on e3) 51...e3+ 52.Kf3 Nf1 53.h5
Nh2+ 54.Kg2 f3+ 55.Kxh2 e2 56.Bc3 f2.
The adjudicators therefore accepted Sanakoev’s evaluation of the endgame (winning for Black) as correct, and gave
him the point.


Supplemental Games

Ilya Smirin – Anton Korobov

Poikovsky 2016
1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.f4 Qc7 7.a4 g6 8.Be2 Bg7 9.0-0 0-0 10.Kh1 e6 11.f5 Nc6
12.fxg6 hxg6 13.Nxc6 bxc6 14.Bf4 e5 15.Bg5 Nh7 16.Be3 Be6 17.a5 f5 18.Qd3 Nf6 19.Bb6 Qe7 20.Qg3 Kf7 21.exf5
gxf5 22.Bd3 e4 23.Be2 Rh8 24.Rad1 d5 25.Na4 Nd7 26.Bc7 Rag8 27.Bd6 Qf6 28.Bf4 Bf8 29.Qe3 Bh6 30.Bxa6 Qh4
31.h3 Bxf4 32.Rxf4 Qg5 33.Bf1 Nf6 34.Kg1 Rxh3 35.Qxh3 Qxf4 36.a6 Rg3 37.Qh2 Qe3+ 38.Kh1 Rg7 39.Ra1 Rh7
40.Qxh7+ Nxh7 41.Nb6 Nf6 42.Ra3 Qe1 0-1

András Adorján – Walter Browne

Wijk aan Zee 1972
1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.f4 Qc7 7.Bd3 Nbd7 8.0-0 b5 9.Kh1 Bb7 10.Qe1 g6 11.Nf3 e5
12.a3 Bg7 13.Qh4 0-0 14.fxe5 dxe5 15.Bh6 Nh5 16.Bxg7 Kxg7 17.Ng5 Ndf6 18.Rf3 h6 19.Nh3 Rae8 20.Raf1 Qe7
21.Qf2 Ng4 22.Qg1 f5 23.exf5 e4 24.Re1 exf3 25.Rxe7+ Rxe7 26.Qd4+ Nhf6 27.gxf3 Bxf3+ 28.Kg1 Re1+ 29.Bf1
Rfe8 30.Qd2 Ra1 31.Nf2 Ne3 32.h3 Rxf1+ 33.Kh2 Rh1+ 0-1
Chapter 8
Action-Reaction: 6.a4

The ...b7-b5 advance is not the only point of the Najdorf, nor is it usually played as early as move 7. Nevertheless,
White prevents it with 6.a4. To some, this could seem as an overreaction – a validation of the whole opening. Surely the
inclusion of ...a7-a6 and a2-a4 should benefit Black in virtually every Sicilian? Dangerous attacking setups involving
queenside castling – the Yugoslav Attack against the Dragon, the English Attack against the Najdorf/Scheveningen, and
the Velimirović and Richter-Rauzer Attacks against the Classical Sicilian – are ruled out.
To the proponents of 6.a4, however, White is simply following the same flexibility principle that underlies the
Najdorf. White gives up the chance to castle queenside, but retains the ability to react to Black’s next move in various
ways. For example, in response to 6...e5, White can play 7.Nf3 and later Bf1-g5 in one move, possibly gaining a tempo
to 6.Be3 e5 7.Nf3. At the same time, White points out the one flaw in the move 5...a6 – the weakness of the b6
square, which can be fixed by a4-a5.
Game 22 covers the response 6...e5, in which Black nevertheless sticks with the iconic Najdorf move. Here we see
that there can be a dark side to White’s attempt to save a tempo and play Bg5 in one move. In the game, Black elegantly
demonstrates Wilhelm Steinitz’s principle that the king is a valuable soldier that can look after himself, when the king
itself turns out to be a crucial piece in the fight for central squares.

Game 22
Alexander Panchenko – Alvis Vitolinsh
USSR Chp Semifinal, Dniepropetrovsk 1980

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.a4 e5

Also against 6.a4 Black can play in Najdorf style. Of course, there are other options:
6...e6 transposes directly into the Scheveningen, with White committed to playing a2-a4.
6...g6 is a Dragon with a2-a4 and ...a7-a6 included. This obviously makes plans involving queenside castling unlikely
for White. However, the early ...a7-a6 means that White can head for a variation where that move is not normally
played, and seek to exploit the weakness of b6. For example, 7.Be2 Bg7 8.Be3 0-0 9.0-0 Nc6 10.Qd2 is a line where
Black normally avoids ...a7-a6. See the supplemental game Adams–Gelfand.
6...Nc6 is often considered the critical response to 6.a4. Black develops in a logical way and seeks to find a flaw in
any of White’s different ways of developing. Now:
7.Bc4 leads, after 7...e6, to a variation of the Sozin with an early a2-a4 that is considered innocuous.
7.a5!? was a surprising pawn sacrifice introduced by Gata Kamsky in his 2011 match with Veselin Topalov.
And 7.Be2 can be met by 7...e5, when White loses the option of an early Bf1-c4. This is the most common line.
White now chooses between 8.Nf3, 8.Nb3, and 8.Nxc6 bxc6 9.f4.


This line has been unfavorably compared to 6.Be3 e5 7.Nf3, since White can hope to play Bg5 in one move and thus
gain a tempo (which has been used for a2-a4). However, this can often be shown to be an illusion. In particular, in the
6.Be3 e5 7.Nf3 line, the light-squared bishop often goes to b3, where it can be captured by a black knight on a5 or by
the bishop from e6. In this case, White would have to recapture with the c-pawn, harming the pawn structure.


Black chooses to prevent Bc4 in the same fashion as in the 6.Be3 e5 7.Nf3 Qc7 line (see Chapter 3). The alternative
is 7...Be7. For this, see the supplemental game Kamsky–Nakamura, which shows clearly the point in the previous note –
the advance a2-a4 does not necessarily make White’s queenside stronger.


White makes the move in one turn. However, the position of the bishop on g5 results in completely different play than
in the equivalent 6.Be3 variation, as Black will maroon the bishop on the kingside. 8.Be3 leads directly into the
variation 6.Be3 e5 7.Nf3 Qc7 8.a4.

Black must support the knight and prevent the destruction of his pawn structure by Bxf6.


With this move, White prepares Bf1-c4 or Nd2-c4-e3.

9...h6 10.Bh4 g5

Black’s thrust weakens the kingside, but chases the bishop away and enables ...Nd7-c5 on the next move without
allowing Bxf6.
Normal play by 10...Be7 gives White better chances to dominate d5. For example, 11.Nc4 b6 12.Ne3 Bb7 13.Bc4
was Tal – Marjanović, Subotica 1987. Black cannot capture the e4-pawn safely.

11.Bg3 Nc5

The resulting position is very double-edged. Black has significant weaknesses on the kingside, which can be poked by
h2-h4 or occupied by a later Nd2-f1-e3, Qd1-f3, et cetera. At the same time, his pieces are very active and he has a
good chance to carry out the freeing break ...d6-d5, when White might find his pieces misplaced.


This move might not be the best try for an advantage. There are numerous other moves:
12.a5?! looks too casual. After 12...Be6 13.h4 (this move also does not mix well with White’s previous move)
13...Bg7 14.Ra3?! (creative, but White’s last three moves make for an incoherent impression) 14...Rd8 15.Qe2 d5 16.f3?
(This is just resignation. Good or bad, White had to capture on d5: 16.exd5 Nxd5 17.Nxd5 Bxd5, and although White’s
position is uncomfortable, at least it is not all one-way traffic.) 16...d4 17.Nd1 Nh5 18.Qf2 Bf6, White was suffering in
S. Hansen – Nakamura, Malmö/Copenhagen 2005.
12.Qf3 Be7 13.Bc4 Ne6 (13...Be6 14.a5 g4 15.Qe2 h5 16.f3 Rc8 17.Bxe6 fxe6 18.fxg4 hxg4 19.0-0 occurred in Tal –
Eolian, Yerevan 1982. The black king is a little exposed, but he has a very strong central position. In the end, the former
world champion was the victim of a big upset.) 14.h4 Nf4 15.Bxf4 gxf4 16.Nd5 Nxd5 17.Bxd5 Be6 (17...Qxc2!?)
18.Bxe6 fxe6 19.Qb3 Kf7, with a good position for Black in Short – Chandler, Næstved 1985. As is often the case in
the Najdorf, Black has to play with his king in the center. However, he has a great pawn center and useful open g- and c-
files, and the white knight will not easily find activity.
12.h4 is best met by 12...Bg7, when it is not clear how the opening of the h-file will help White. Instead, 12...g4?!
13.h5! allows White to once again fight for d5 with Bg3-h4. Ermenkov – Akopian, Dubai 2000, ended in a quick draw
after 13...Be6 14.a5 Bg7 15.Bh4 Nxh5 16.Nc4, but White has a very promising position.


This equalizes. If instead White is allowed to consolidate his hold on the central squares, then he has a good chance
for an advantage, e.g. 12...Be6 13.Qe2 Rc8 14.0-0 Be7 15.b3 Ncd7 16.a5 h5 17.h4, when Black was struggling to find
useful activity in Jansa – Heinemann, Hamburg 1992.

13.Ncxe4 Nxe4 14.Bxf7+!?

It is risky to give Black the center and two bishops, even if the black king himself will have to lead his army into
battle! The alternative is the drawish 14.Nxe4 Qxc4 15.Nxd6+ Bxd6 16.Qxd6 Be6 17.Qxe5 0-0 and White is a pawn
up, but must give it back in order to castle: 18.b3 Qxc2 19.0-0 Qf5 (19...Bxb3 was possible, but the black king is very
exposed, so Black instead sought a queen trade and the draw) 20.Qxf5 and the game was agreed drawn in Moroz –
Kovchan, Alushta 2003.

14...Kxf7 15.Nxe4 Bf5

15...Be7 is safer. For example, 16.0-0 Be6 (16...Qc6, fighting for the light squares, looks more flexible) 17.c4! with
dynamic equality in the game Fishbein – Vigorito, US Chess League (Internet) 2013.


White cannot content himself with restraining the black center. The two bishops and pressure on the light squares are
a major, long-term advantage, and White needs to cause some discomfort to the black king.
Better is 16.Qf3!, when Black should play 16...Kg6 17.h4 h5! (the only move, in view of the threat of 18.h5+)
18.Nxg5 Be7 and in this strange position, it might be the black king who is now safer! The game is very sharp.


Black fights for the central light squares and pressures White along the long diagonal.

17.0-0 h5!?


Too passive; White had to show the spark of life with 18.f4!! gxf4 19.Bxf4 exf4 20.Rxf4, with a dangerous attack
against the exposed black king. Now, best seems to be 20...Kg6! (better than giving the tempting check on c5, after
which White will later be able to play Qf3, or Ne4 with tempo) 21.Nd5 Qxc2 22.Qf3, with a wild position. Black is not
yet consolidating.

18...h4 19.Bf2 h3!

White will suffer from the weakness on g2 for the rest of the game.

20.g3 g4 21.Be3

White is planning Nc3-d5 followed by fxg4. In spite of Black’s positional advantage, it is not obvious how to prevent
this and keep the grip on the long diagonal.

The black king steps up, not only leaving the dangerous f-file, but crucially covering the d5 square. The king is
perfectly safe on e6 – Steinitz would be proud.


White finds himself with no moves, needing to keep f3 defended and g2 shielded. With his last move, he plans to
develop the rook to a4.
On 22.Qe2 gxf3 23.Rxf3 (the endgame after 23.Qxf3 Qxf3 24.Rxf3 Bxc2 is hopeless for White) 23...Bg4, Black wins
the exchange without giving White any real counterplay.


There was no hurry for 22...gxf3, when White was probably planning 23.Ra4, eventually recovering the pawn by
covering the g4 and e4 squares.

23.Ra4 Raf8

Black focuses on f3, which will eventually not be able to hold.

24.Rb4 Bd8!

A last preparation: b6 is defended. Black is now threatening to capture on f3.


White has no good defense, e.g. 25.Bb6 gxf3, and if 26.Rxf3 then 26...Bxc2, while 26.Rf2 Be7 leaves White a pawn
down with limited swindling chances.

25...gxf3 26.Rxf3

White prefers the quicker exit. The endgame after 26.Qxf3 Qxf3 27.Rxf3 Bxc2 is hopeless. After, for example,
28.Rxf8 Rxf8 29.Rxb7 Bd3 30.Bf2 Bxa5, not only is White down a pawn, but his king is also caught in a permanent
trap, leaving all of his pieces stuck.


This little tactical blow decides the game.

27.Qxd3 Qxf3 28.Qc4+ Kd7

White can stop checkmate only by going into the hopeless endgame with 29.Qg4+. Therefore, he resigned.


Supplemental Games

Michael Adams – Boris Gelfand

Dortmund 2006
1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.a4 g6 7.Be2 Bg7 8.0-0 0-0 9.Be3 Nc6 10.Qd2 Nxd4 11.Bxd4
Be6 12.a5 Rc8 13.Bd3 Bc4 14.Rfd1 Qc7 15.h3 Rfe8 16.Bxc4 Qxc4 17.Ra4 Qc6 18.Rb4 Qc7 19.Rb6 Nd7 20.Nd5 Qb8
21.Rb4 Ne5 22.Ra4 Rc4 23.Rxc4 Nxc4 24.Qc3 Bxd4 25.Rxd4 Ne5 26.Rb4 Nc6 27.Rb6 Qc8 28.b4 Qd7 29.b5 axb5
30.a6 b4 31.Qa1 bxa6 32.Qxa6 Nd4 33.Rxb4 Ne6 34.Rb7 Qc8 35.Qb5 Kf8 36.Qb2 Ng7 37.Nxe7 Qc4 38.Qf6 Qxe4
39.Nf5 Qe1+ 40.Kh2 Qe5+ 41.Qxe5 dxe5 42.Nd6 Rd8 43.Nxf7 Rc8 44.Ng5 h5 45.Nh7+ Kg8 46.Nf6+ Kf8 47.Nh7+
Kg8 48.Nf6+ Kf8 49.Nd7+ Kg8 50.Rb2 Kf7 51.Nxe5+ Kf6 52.Nf3 Ne6 53.Rb6 Ke7 54.Ne5 g5 55.c4 h4 56.Rb7+
Kf6 57.Ng4+ Kg6 58.Rb6 Kf7 59.Rb7+ Kg6 60.Ne3 Ra8 61.Rb6 Kf7 62.g3 Ra2 63.Kg2 hxg3 64.Kxg3 Nc5 65.Kf3
Nd3 66.Ng4 Ke7 67.Rb3 Nc5 68.Re3+ Kf7 69.Ne5+ Kf6 70.Ng4+ Kg6 71.Re5 Ra3+ 72.Ne3 Nd3 73.Rd5 Rc3 74.Rd4
Ra3 75.h4 gxh4 76.Rxh4 Ra2 77.Rh2 Kf6 78.Ke4 Nc5+ 79.Kd5 Nb3 80.f3 Ra5+ 81.Ke4 Nc5+ 82.Kf4 Ne6+ 83.Kg3
Ra3 84.Re2 Kg5 85.Nc2 Ra6 86.Nb4 Rb6 87.Nd5 Rd6 88.Re5+ Kh6 89.Nf4 Nc7 90.c5 Rc6 91.Kg4 Kg7 92.Kf5 Kf7
93.Ke4 Ra6 94.Rf5+ Ke8 95.Ke5 Rc6 96.Nd3 Ke7 97.Rh5 Re6+ 98.Kd4 Rf6 99.f4 Ra6 100.Rh7+ Kd8 101.f5 Ne8
102.Nb4 Ra1 103.Nc6+ Kc8 104.Rh8 Kd7 105.Ne5+ Ke7 106.Ng6+ Kd8 107.Kd5 Re1 108.c6 Rd1+ 109.Ke5 Kc7
110.Rxe8 Re1+ 111.Kd5 Rxe8 112.f6 Rd8+ 113.Kc5 Rd1 114.f7 Rc1+ 115.Kd5 Rd1+ 116.Ke6 Rf1 117.Ne7 1-0

Gata Kamsky – Hikaru Nakamura

U.S. Chp, Saint Louis 2012
1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.a4 e5 7.Nf3 Be7 8.Bc4 0-0 9.0-0 Be6 10.Bb3 Nc6 11.Bg5 Na5
12.Bxf6 Bxf6 13.Bd5 Rc8 14.Nd2 Qc7 15.Re1 Bg5 16.Nf1 Qb6 17.Rb1 Nc4 18.Qe2 Bh6 19.h4 Qb4 20.g3 Rc7
21.Kg2 Nb6 22.Bb3 Bxb3 23.cxb3 Qxb3 24.a5 Na4 25.Nh2 g6 26.Ng4 Bg7 27.Nd5 Rc2 28.Qe3 Nc5 29.h5 Qxe3
30.Ngxe3 Rd2 31.Nc4 Rd4 32.Nxd6 Rd8 33.b4 Nd3 34.Nxb7 Nxe1+ 35.Rxe1 Ra8 36.f3 Bf8 37.Rc1 Bxb4 38.Rc7
gxh5 39.Kh3 Kg7 40.Kh4 Ra7 41.Kxh5 Rxd5 42.exd5 Bxa5 43.Re7 Bb6 44.d6 a5 45.Kg5 a4 46.Kf5 a3 47.Nd8 a2
48.Ne6+ Kh6 49.Ng5 a1=Q 50.Nxf7+ Kg7 0-1
Chapter 9
Odds and Ends

In this chapter, for the sake of completeness, we cover White’s various obscure and often eccentric options on move 6.
Besides White’s main ways of meeting the Najdorf, there are a number of playable and semi-playable moves which are
not without interest. We will cover each one briefly. In rough order of soundness, they are:
a) 6.Bd3
b) 6.Qf3
c) 6.Qd3
d) 6.Nb3
e) 6.a3
f) 6.Rg1
g) 6.h4
Many of these moves have been played in recent years by the world’s top players, often with success.
While these moves are not considered theoretically critical, they all should be respected by those meeting them. Let us
not forget that there was once a time when 5...a6 was an obscure, eccentric move that might have provoked laughter!

a) 6.Bd3
A move that is not at all innocuous. The bishop is blocked by the e-pawn, but at the same time White’s center is
One of White’s ideas is to meet 6...e5 with 7.Nde2. White later plays f2-f4 and Ne2-g3, where the knight is aiming
towards the f5 square. The game begins to look like an improved version of the 6.f4 variation.
Another point is that after the natural 6...Nc6, White will play 7.Nxc6 bxc6 8.0-0, followed by b2-b3, Bc1-b2, f2-f4,
and so on. Black’s central pawns are strengthened, but this structure lacks a certain flexibility.
In my opinion, Black does best to enter an improved Dragon with 6...g6. In the Dragon, the white bishop rarely goes
to d3. This represents Black’s best way of exploiting the downsides of 6.Bd3.

Game 23
Josef Gheng – Arkadij Naiditsch
Neckar Open 2004

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Bd3 g6 7.0-0 Bg7 8.Kh1

One of the ways in which White has attempted to make sense of his position is with 8.Re1, which transposes directly
to a variation of the Classical Dragon sometimes used by Karpov – the bishop drops back to f1 (normally from e2).
However, this variation is not particularly dangerous for Black. For example, 8...0-0 9.Bf1 Nc6 10.Nb3 Bg4! is a good
way to exploit White’s delay in developing the dark-squared bishop. White has to agree to the weakening f2-f3 or else
place the queen either in front of that bishop (11.Qd2) or on an exposed square (11.Qd3).

8...0-0 9.f4 Nbd7

Black played thematically in the game Medina – B.Smith, Ohio 2005: 9...b5 10.a3 Bb7 11.Be3 Nbd7 12.Qf3 Rc8
13.Rae1 Rc7 14.Rc1 Qa8 15.Nde2 d5!, when Black opened the center in effective fashion.

10.Be3 e5 11.Nde2 b5
A position of the “Dragadorf” type arises, but White’s minor pieces on d3 and e2 are poorly placed. The knight will
have to go to g3, where it will be controlled by Black’s g-pawn, while the bishop on d3 just gets in the way.

12.Ng3 Bb7 13.fxe5 Nxe5 14.Qd2 b4 15.Na4 Bc6 16.b3

Decentralizing the white pieces in order to grab a pawn by 16.Nb6 Rb8 17.Qxb4? looks, and is, terrible. Black wins
after 17...Nfg4 18.Bg1 Nxh2! – a standard tactic which has been seen in similar positions. Then 19.Bxh2 is met by
19...Rxb6, when Black regains the pawn with a winning advantage (...Nf6-g4 and ...Qd8-h4 are coming), while
19.Kxh2 is met by 19...Qh4#.


As so often happens, this explosive advance gives Black the initiative, when White cannot answer with e4-e5.

17.Nb6 dxe4! 18.Nxa8 exd3 19.Nb6 Nfg4!

White has won the exchange, but his far-flung forces cannot stop the focused energy of the black army.

20.Bg1 Qh4

Black’s main threat is 21...Nxh2 22.Bxh2 Ng4. In fact, White has no good way to defend against this.


21.Qf4 loses to 21...Bh6, when the queen is lost, e.g. 22.Qxb4 Nf2+, or 22.Qd4 Be3.

21...Bh6 22.Raf1 Bxf4 23.Rxf4 Qh6 24.cxd3 Nxd3 25.Nf5 Qxf4 26.Ne7+ Kg7 27.Qxf4 Nxf4 28.Nxc6

White has survived the storm – surprisingly – but the endgame is hopeless.

28...Re8 29.g3 Nd3 30.Nd5 Re2 31.Ndxb4 Nxb4 32.Nxb4 a5 33.Nc6 Rxa2 34.h3 Nf2+ 35.Bxf2 Rxf2 36.Nxa5
Rb2 37.h4 h5 38.Kg1 Kf6 39.Kf1 Kf5 0-1

b) 6.Qf3
The queen moves 6.Qf3 and 6.Qd3 (which may sometimes transpose to each other) demand the utmost respect. In
recent years, for example, 6.Qf3 has enjoyed great success. The move is not without logic at all, and such Qd1-f3
moves are becoming more common in some other variations of the Sicilian (such as the Taimanov).
One of the points is that 6...e5 is less desirable in view of 7.Nf5. White can meet the change to a Scheveningen
(6...e6) by 7.g4, racing the g-pawn up the board, with good attacking prospects.
Once again, changing to the Dragon with 6...g6 is perhaps the best option. However, White has a setup that is not
without some poison. This is a good variation for players of the white pieces who like to attack but also want to play
their “own” moves.

Game 24
Judit Polgár – Dariusz Świercz
Istanbul Olympiad 2012

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Qf3 g6

6.Qf3 is aimed against 6...e5, but it is not clear that the latter is not playable. For example, 6...e5 7.Nf5 Be6 8.Bg5
Nbd7 9.0-0-0 Qc7 10.g4 Rc8 was seen in Pichot – Adly, Doha 2016. White’s setup looks aggressive, but where does he
go from here? Meanwhile, Black is planning counterplay on the queenside with ...Qd8-a5.
6...e6 is likely to be met by 7.g4, which looks like a fairly promising line for White – although 7...Nc6 is another
possibility for Black to look into. Instead, the originator of 6.Qf3, American master Andrew Karklins, liked to continue
with 7.b3. His record against grandmasters with this line was not very good, but he did have one major scalp: 7.b3 Qb6
8.Nde2 Qc7 9.Bb2 b5 10.a3 Bb7 11.g4 d5 12.exd5 Nxd5 13.Bg2 Nd7 14.0-0 Bd6 15.Qh3 Nxc3 16.Nxc3 Be5 17.Bxb7
Qxb7 18.Rad1 0-0 19.Qe3 Bb8 20.Ne4 Ne5 21.Bxe5 Bxe5 22.Nc5 Qc7 23.f4 Bf6 24.Rd7 Qb6 25.Rfd1 Rfd8 26.b4 a5
27.Qf3 axb4 28.axb4 Kf8 29.Kg2 Rdc8 30.R1d6 Qb8 31.Qd3 1-0 was Karklins – Svidler, World Open 1995.


This is White’s usual setup here. White prepares Bc1-e3 and 0-0-0.

7...Bg7 8.Be3 Bd7

Black prepares ...Nb8-c6 without allowing Nxc6 and e4-e5. This doesn’t seem to be necessary, though. Instead, 8...0-
0 9.0-0-0 Nc6 is playable, since 10.Nxc6 bxc6 11.e5 Nd5 12.Nxd5 cxd5 13.Qxd5 Be6 followed by ...Bxe5 is a dream
Dragon for Black: the open files on the queenside and open files for the bishops will lead to a winning attack. Instead,
10.Kb1 will possibly end up transposing to the game.


White goes to remove the knight on f6 with this early jump. Natural is 9.0-0-0 Nc6, and now either 10.g4, 10.Nd5 (as
in the game), or the pawn sacrifice 10.e5!?, which occurred in Pichot – Banikas, Doha 2016. The game was unclear after
10...dxe5 11.Nxc6 bxc6 12.Bc4 0-0 13.Bc5 Re8. This pawn sacrifice can be avoided by Black, using the move order in
the last note.

9...0-0 10.0-0-0 Nc6 11.Kb1 Ne5

It would be logical to continue to develop with 11...Rc8, rather than making this committal move.

12.Nxf6+ Bxf6 13.Qg3 Qa5 14.Nb3 Qc7 15.f4 Nc6

Black is inconsistent. It is better to play 15...Nc4 16.Bxc4 Qxc4 17.Bd4 Rac8, when Black has reasonable chances.
Instead he agrees to the loss of two tempi.

16.h4! Rfc8 17.c3

White’s attack is very dangerous, since Black lacks defensive pieces on the kingside and there is no easy way for
Black to get any counterplay.

17...h5 18.Be2 Bg7 19.Qf2 a5 20.Bb6!

Before continuing the kingside attack, White chases the queen into a trap. Black will suffer from the out-of-play
queen for the rest of the game.

20...Qb8 21.a3!
Another prophylactic move. White is not in a hurry, but instead prevents the a-pawn from reaching a3.

21...a4 22.Nd4 Nxd4 23.Bxd4 e5 24.fxe5 dxe5 25.Bb6 Be6 26.g4!

Black’s queenside is fatally blocked, meaning his pieces cannot come to the king’s aid, and this direct attack is

26...hxg4 27.h5 gxh5 28.Rxh5 Ra6

Desperation, but Black was facing decisive threats of Rh5-g5 followed by Qf2-f6 or simply Bxg4 and a doubling or
tripling of pieces on the g-file. Offering slightly more chances was 28...Rc6, in which case White should not take the
queen by 29.Rd8+ but rather just play 29.Be3, when the attack on the kingside will continue with Qf2-h4, Rdh1,
Bg5/h6, et cetera. Black should not be able to defend.

29.Bxa6 bxa6 30.Rg5 Qb7 31.Qf6 Qxe4+ 32.Ka1

32...Qh7 is met by 33.Qd8+!.


c) 6.Qd3
This is another unusual-looking queen move, which has become popular in recent years. Its biggest proponent has
been Spanish grandmaster Francisco Vallejo Pons. The position of the queen on d3 rather than f3 means that White can
avoid the h2-h3 move against Black’s Dragon setup.

Game 25
Pentala Harikrishna – Maxime Vachier-Lagrave
Biel 2014

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Qd3 Nbd7

This flexible move is among Black’s good options here. White’s most aggressive development now leads to an
obscure sideline of the 6.Bg5 variation.
White develops rapidly after 6...g6, although naturally that move is playable. For example, 7.Bg5 Bg7 8.0-0-0. Now
Black should play 8...0-0 9.f4 Nbd7, with a sharp position, since the immediate 8...Nc6 leads to problems, as two games
by Vallejo from the world rapid and blitz championships showed: 9.Nxc6 bxc6 10.e5 Ng4 11.exd6 Nxf2 12.Qf3 Nxd1
13.Bc4 Bf6 14.Bxf6 exf6 15.Re1+ Be6 16.Bxe6 0-0 17.Rxd1 fxe6 18.Qxc6 f5 19.d7 Qe7 20.Na4, and the powerful
passed pawn fatally tied up the black pieces in Vallejo Pons – Fedoseev, Dubai 2014 (rapid); and 9.Nxc6 bxc6 10.e5
Bf5 11.exf6 Bxd3 12.fxg7 Rg8 13.Bxd3 Rxg7 14.Rhe1 and White’s three minor pieces were definitely stronger than
Black’s queen in Vallejo Pons – Karjakin, Dubai 2014 (blitz).


This position could also be reached by 6.Bg5 Nbd7 and now the very unusual 7.Qd3. White has often preferred
7.Be2. Black can now choose a Dragon setup where White does not have the option of such aggressive play as in the
last note. For example, 7...g6 8.Bg5 Bg7 9.0-0-0 h6!. It is good to play this move immediately, making the bishop
retreat, rather than later, when White has been known to play h2-h4 and sacrifice the bishop. Oparin – Nepomniachtchi,
Zurich 2017, continued 10.Bh4 0-0 11.Kb1 Qc7 12.Qe3 b5 and Black had good counterplay.


Black goes into a classical kind of setup against the Bg5 variations, but with the white queen on d3. It is also possible
to play in a more crooked way: Ibarra Jerez – Świercz, Baku 2016, went 7...h6 8.Bh4 g6 9.0-0-0 e5 10.Nb3 Be7, with a
typical structural change in 6.Bg5 Nbd7 variations. The pawn on g6 defends f5 and does not really weaken the black

8.0-0-0 Be7 9.f4 Qc7

Normally the queen would be on f3 in this position, and it certainly cannot be claimed that she stands better on d3.

10.Be2 b5 11.Bf3 b4 12.e5

White begins complications, since retreating the knight would have allowed Black to achieve his opening goals.

12...dxe5 13.Bxa8 exd4 14.Ne2 h6 15.Bh4 0-0 16.Kb1

White should prefer 16.Qxd4, although Black has reasonable compensation for the exchange after 16...e5 17.fxe5
Nxe5, planning ...Bc8-f5 and ...Rc8.



This blocks the bishop’s retreat, leading to some problems. 17.Bf3? would be met by 17...e4 18.Bxe4 Nc5, but
17.Qd2 looks better.

17...exf4 18.Nxf4?!

White prefers to sacrifice a piece over playing 18.Bxf6, which trades off an important bishop. Objectively speaking, it
is probably not quite sufficient.

18...Ne5 19.Qe2 g5 20.Bg3 gxf4 21.Bxf4 Bd6 22.Rxd4 Re8 23.Rhd1?

23.Bxh6 is much stronger.

23...Bg4 24.Qxa6 Bxd1 25.Rxd1


Black is up a piece for two pawns, although White has practical chances in the long run, since it will not be easy to
consolidate the black position. However, Black misses the chance to take the initiative decisively on this move with
25...Rc8!, for example 26.Qxd6 Qxc2+ 27.Ka1 Nc4! 28.Qd4 Rxa8, when Black is winning.

26.Bxd6 Qxd6 27.Qxd6 Nxd6 28.Bf3

Perhaps Black should win this position, but it is not so easy. In the game White managed to scrape to a draw.



29.b3 Ne5 30.Rd4 Nxf3 31.gxf3 Rb8 32.Kb2 Kf8 33.c3 bxc3+ 34.Kxc3 Ke7 35.b4 Ke6 36.a4 Nd5+ 37.Kb3 Rc8
38.Rc4 Rxc4 39.Kxc4 Kd6 40.a5 Ne3+ ½-½

d) 6.Nb3
Surprisingly, this retreat, voluntarily removing the knight from the center, was played many times in the past – even
as early as 1946, when the Najdorf still lacked a name. Yefim Geller once played it. In recent years, it has been used by
a number of top players.
Naturally, the move is aimed at 6...e5, which can now be met with 7.Bg5, and the light-squared bishop will probably
find a better place than e2.
Instead, Black should choose between “improved” Dragon or Scheveningen setups, according to preferences.

Game 26
Viswanathan Anand – Maxime Vachier-Lagrave
Bastia (rapid) 2016

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Nb3 Nc6

Entirely logical. The white knight went backwards, the black knight – forwards. How could this benefit White? Of
course, 6...g6 and 6...e6 are also very playable.

7.Be3 g6

Black chooses a Dragon setup. The main alternative is 7...e6 when usual is 8.g4. Black can aim to occupy the c4
square, e.g. 8...b5 9.Bg2 Ne5 10.g5 Nfd7 11.Nd2 Nb6 12.f4 Nec4 13.Nxc4 Nxc4 14.Bc1 with equality in Naroditsky –
Robson, U.S. Chp, Saint Louis 2017.


White can instead play 8.Be2 and head for a Classical Dragon, where the voluntary retreat Nd4-b3 is typical.
Nevertheless, that is not a particularly threatening line, and Black’s ...a7-a6 and ...b7-b5 is considered a good plan in this
version. Most players of 6.Nb3 have instead preferred to go into lesser-known positions with the text move.

8...Nxd5 9.exd5 Ne5 10.Bd4

In view of Black’s reply, it might be more accurate to play 10.Be2. However, the type of position arising after
10...Bg7 11.0-0 0-0 is not too bad for Black. Driving the knight from e5 will require the loosening f2-f4 sally.


Black takes the fianchetto one step further. The bishop trade is avoided, while the bishop is not poorly placed on the
h6-c1 diagonal.

11.Be2 0-0 12.0-0 b5

The move weakens the c6 square, but White is not in good position to take advantage of it, as the bishop is rather
stuck on d4.

13.a4 b4 14.c3 Bb7!

14...bxc3 15.Bxc3 would be good for White, since he gains access to the weak squares on the queenside. Instead,
Black gives White a mass of pawns on the queenside but conquers the center for himself.

15.cxb4 Bxd5 16.Na5 e6!

The bishop is anchored on the excellent d5 square, and the queen is freed to come to the kingside. In such a lopsided
position, the side possessing the center has a big advantage, provided that the position is not very simplified.

17.Ra3 Rb8

The attack on the b4-pawn is awkward for White to meet. His next move renders 17.Ra3 useless.

18.Bc3 Qh4

19.Be1 Bf4 20.Rh3

If 20.g3 then 20...Qh3, and mate is threatened on g2.

20...Qg5 21.Rg3!?

White is willing to give up the exchange to eliminate Black’s strong bishop. He hopes that the kingside attack will be
ended and the queenside pawns will come into their own.
Instead, 21.f3 would lock the rook out of the game on h3 in ugly fashion. Then after 21...Rfc8 the black rook will
penetrate to c1.


Black is not in a hurry to capture the exchange.

22.Bxa6 Bxg3 23.fxg3

If 23.hxg3 then Black can return the exchange, similar to the game, with 23...Rxb4!, and if 24.Bxb4 then 24...Qe4
25.f3 Qxb4 and White is losing a piece – not only is the knight on a5 attacked, but ...Qb6+ is also threatened.

23...Qe4 24.Rf2 Rxb4!

Black returns the exchange to eliminate the dark-squared bishop and leave himself in total control. Otherwise, White
could begin to advance the queenside pawns and the situation could become unclear, as Black would lack clear targets.


25.Bxb4 Qxb4 immediately wins for Black, since the white pieces are stuck on the queenside. If 26.Qd2 then
26...Qxd2 27.Rxd2 Ra8, winning a piece.

25...Rd4 26.Qc1

26.Rxe4 Rxd1 27.Re2 Ra8 wins a piece for Black.

Black is in total control here, but this is the clearest finish.


27.Kh1 Qg4 and the knight cannot be captured, with ...Qh5 is coming. Black is not only up material but has a
winning attack.

27...Qxf3 28.Qc3 Qh1+ 29.Kf2 e5!

There is no good defense to the threat of 30...Qg2+ 31.Ke3 Qf3#, so White resigned.


e) 6.a3

White mimics his opponent in a teasing way. Can a2-a3 be as valuable as ...a7-a6? It is hard to believe. But, as with
many of these obscure moves, White challenges his opponent to find the variation where his move is the least valuable.
In some cases, a2-a3 could be useful to tuck the bishop away after Bc4, or to slow a ...b5-b4 push.
Both the Dragon (with 6...g6) and the Scheveningen (with 6...e6) are good for Black here. White can get away with
playing 6.a3, but as they say, “it is just a game.”

Game 27
Magnus Carlsen – Radosław Wojtaszek
Wijk aan Zee 2017

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.a3

The “hottest theory,” played by the world champion in a top-level tournament. Magnus Carlsen likes to outplay his
opponents over the board, rather than by preparation.

It is entirely possible to play this standard Najdorf move, although Black should be aware of White’s reply. Both
6...e6 and 6...g6 are fine – White’s a2-a3 is not likely to be a very threatening way to meet the Dragon or Scheveningen.


In comparison to 6.Be2 e5 7.Nf5?!, which is almost never played, White benefits from the a2-a3 move and the fact
that the queen is free to come to f3 or g4. 7.Nf3 is another possibility, although after 7...Qc7, a2-a3 is almost useless.


7...Nxe4 is too risky. The position of the black king is questionable after 8.Nxg7+ Bxg7 9.Nxe4 d5 10.Bg5! f6
11.Qh5+ Kf8 12.Be3, when White saves both of his pieces since 12...dxe4 now runs into 13.Bc5+ Kg8 14.Bc4+.

8.Bg5 d4 9.Bxf6 Qxf6

9...gxf6!? is interesting. Unlike in the game, the knight cannot go to d5 due to 10...Bxf5. Therefore, White can use
6.a3 in a surprising way: 10.Na2. Then, after 10...Bxf5 11.exf5 Nc6, an interesting and double-edged game results.

10.Nd5 Qd8 11.Qg4 Bxf5

Black has to trade the bishop because 11...g6 is met by 12.Qg3!, for instance 12...Nc6 13.Nxd4! exd4 14.Nc7+.

12.Qxf5 Bd6 13.h4!?

White gains some space on the kingside and prepares to develop the rook via h3.

13...Nc6 14.Bc4 b5 15.Bb3 Ne7 16.Qg4 0-0 17.Rh3 Nxd5 18.Bxd5

From a wild and unusual opening, Carlsen’s type of position has emerged: a tiny advantage, with a long fight ahead.
The white bishop is somewhat more active than its counterpart.

18...Ra7 19.Rg3 Qf6 20.a4

White seeks to create objects of attack on both sides of the board.

20...Bb4+ 21.Kf1 bxa4 22.Rxa4 a5 23.Ra1 Rc7 24.Bb3 Ra8 25.Kg1

White has secured his king, and although there is still nothing clear, Black has to be a little more careful, as there is
some pressure against f7.

25...Bf8 26.Qh5 g6?!

It is hard to just wait, but that is what Black should do.

27.Qg4 Ra6 28.h5 Qf4?!

This only drives the white queen to the queenside, where she is an even bigger problem for Black.


White keeps the queen, which can now enter the other side of the board.

29...Qf6 30.Qb5 Qc6?!

Black panics and drops a pawn, although his position was already getting very difficult.

31.Qxe5! Re7

Probably Black intended 31...Bd6, but only now saw that this would be met by 32.hxg6! Bxe5 33.gxf7+ Kf8

32.Qf4 a4

32...Qxe4 would be met by 33.Bxf7+, keeping the extra pawn. On 33...Kg7 34.Qxe4 Rxe4 35.hxg6 hxg6 36.Bc4
followed by Bd3, White has excellent chances to win the endgame despite the bishops of opposite colors, due to the
presence of rooks and Black’s numerous weak pawns.
33.Bd5 Qc7

33...Qxc2 would meet with 34.Rc1 Qxb2 35.Rc8, with the threat of Qf4-h6, and White is winning.

34.Qd2 Qb6 35.Ra2

This move looks a little awkward, but White keeps his queenside pawns defended, and will soon take the initiative on
the kingside again.

35...Rc7 36.Rf3 Qb4 37.Qe2 Rb6 38.hxg6 hxg6 39.g3 Kg7 40.Kg2 Rd7 41.Qd1!

A clever idea. White plans to remove the cramping a4-pawn.

41...Rf6 42.Rxf6 Kxf6 43.c3! dxc3 44.Rxa4

Now 44...Qxb2 gives White a winning attack after 45.Qd4+, e.g. 45...Kg5 46.e5! (threatening 47.Qh4#) 46...Kh6
47.Qh4+ Kg7 48.Qf6+ Kh7 49.Rh4+ Bh6 50.Rxh6+ Kxh6 51.Qh8+ Kg5 52.Qh4+ Kf5 53.Qf4#. Meanwhile, after
other queen moves White can simply capture on c3 with two extra pawns and a continuing attack. Therefore, Black


f) 6.Rg1

This move was popular around the end of the ’90s, and I have played it a few times. Vassily Ivanchuk even once beat
Garry Kasparov with it. That said, it is not a very good move.
As with 6.h3, White is preparing g2-g4. However, he does it in a riskier manner. Kingside castling is ruled out, and
the pawn on g4 will not be protected so well. On the plus side, White can often play h2-h4 in one move, while from g1
the rook could support a further advance of the g-pawn. The rook can also sometimes come to g3 with some effect.

Game 28
Alfredo Giaccio – Anton Kovaliov
Pinamar 2006

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Rg1 e5

Most other moves are playable here as well:

6...g6 leads to a Dragon where White will try to prove that Rh1-g1 is more useful than ...a7-a6. One would expect that
White is trying to open the h-file, and that therefore the rook should be on h1. However, there are some cases in the
Yugoslav Attack where the rook doesn’t stand badly on g1, while ...a7-a6 is often not very useful.
6...b5 is a good move. After 7.g4 Bb7 8.Bg2 Nfd7, Black has had good results. In the days when I was playing 6.Rg1,
I intended the pawn sacrifice 8.g5 Nxe4 9.Nxe4 Bxe4 10.Qg4 Bb7 11.Bg2, although it has to be admitted that White
should not have enough compensation here.
However, 6...Nc6 7.g4 Nxd4 8.Qxd4 is not the best, as it only helps White to develop. Now if 8...Bxg4? (or 8...Nxg4
9.Nd5, giving White great compensation for the pawn), then 9.Rxg4 Nxg4 10.Qa4+ Qd7 11.Bb5 axb5 12.Qxa8+ Qd8
13.Qxb7, with a big advantage. This is one of the tricks in the 6.Rg1 variation.
Of course 6...e6 is playable. Black should be aware, however, that the logical 7.g4 d5, contrary to what one might
assume, is not really better than the equivalent 6.h3 e6 7.g4 d5 variation, because with the rook on g1 White can capture
twice on d5 and attack the queen with Bf1-g2: 8.exd5 Nxd5 9.Nxd5 Qxd5 10.Bg2.


Now White’s ideal scenario would be to achieve g4-g5, Qd1-f3, Bf1-h3 (trading off Black’s light-squared bishop)
and 0-0-0. Black should hit back quickly to render Rg1 useless.


This approach is very effective against 6.Rg1. Normally the early ...b7-b5 has the downside that the queenside can be
weakened by a2-a4. However, in this variation White has given up the chance to castle kingside, so this would leave his
king without a safe home.

8.g4 Bb7 9.Qf3

White is better off defending e4 with the queen rather than the bishop, which would otherwise be doing nothing on

9...b4 10.Nd5 Nxd5 11.exd5 Nd7

Black can now build up pressure on the c-file, against d5, and with ...a5-a4. Black is already slightly better.

12.Be3 Rc8 13.Qe4 Be7 14.0-0-0 Qc7 15.g5 a5 16.Bb5 0-0 17.Kb1 f5 18.gxf6 Nxf6 19.Qg2 Rf7 20.Rd2 Bf8

Black has comfortably defended the one target on g7, and White is walking on eggshells on the rest of the board,
trying to defend d5, c2, and a4.


The breakthrough comes already. This move diverts the bishop from covering c4 and opens the a-file.

22.Bxa4 Nxd5! 23.Qg4

The point is that 23.Rxd5 is met by 23...Qc4, when the pinned rook cannot be defended: if 24.Rcd1, then 24...Qxc2+
25.Ka1 Qxd1+ and Black wins the exchange.

23...Ra8 24.Bb5 Kh8

The upshot is that Black has broken through the center and opened the a-file as well.

25.Qc4 Nxe3 26.Qxc7 Rxc7 27.fxe3 Rac8

The endgame is clearly better for Black. The b4-pawn holds back three white pawns, in a manner that is typical for
the Najdorf. White is tied down to the defense of c2, and the black bishops will also become a strong force, while the
white knight is ineffective.

28.Bd3 g6 29.Rf2 Bh6 30.Re1 Bg5 31.Rd2 Kg7 32.Bf1 d5 33.Rg2 Bh4 34.Rc1 Rf7 35.Rg4 Be7 36.Na5 Ba8 37.Ba6
Rc7 38.Bf1 h5 39.Rg1 Bh4

40...Bf2, winning the e-pawn, is now an unstoppable threat.

40.e4 Bf2 41.Rh1 dxe4 42.Bg2 Rf8 43.b3 e3 44.Bxa8 Rxa8 45.Nc4 e2 46.Rhe1 Bxe1 47.Rxe1 Rca7 48.a4 bxa3
49.Ka2 Re7 50.Rxe2 g5 51.Nd6 Rf8 52.c4 e4 53.Rxe4 Rxe4 54.Nxe4 g4 55.c5 h4 56.Kxa3 g3 57.hxg3 h3 58.g4 h2
59.Ng3 Rf3 60.Nh1 Rf1 61.Ng3 Rg1 0-1

g) 6.h4
With this move, we have truly reached the avant garde. “Nothing is prohibited!” and “Everything is possible!” is the
battle cry. This move is really the chess equivalent of a neck tattoo.
While nobody could honestly claim that 6.h4 is the best move, it is not possible to directly refute it, and Black faces a
psychological liability – he must try to choose the variation that best renders h2-h4 useless or harmful. If he fails, and
after ten or so moves a position arises where h2-h4 makes sense, the resulting depression could be fatal.
The move 6.h4 is truly a twenty-first century move, since the first recorded instance of its use was a 2004 Internet
blitz game, where White likely meant to play 6.h3.
In 2017, however, it has caught on with strong players who want something different. Ian Nepomniachtchi, Yu
Yangyi, Vladimir Fedoseev, Daniel Naroditsky, and Grigoriy Oparin are strong grandmasters who have used it in the
last year. All in all, I would not be surprised if 6.h4 turns out to have the highest average rating of practitioners among
all of White’s sixth moves. And surprisingly, White’s results have not been too bad. White strikes a psychological blow,
mocks the opponent, loosens his own mental restraints – and the resulting slightly inferior position makes him play
more resourcefully and keenly, as Emanuel Lasker has said.

Game 29
Milan Pacher – Hovhannes Gabuzyan
Mashhad 2017

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.h4 b5

It is very logical to play this immediately, preparing the development of the bishop on b7 and a counterattack against
e4. After 6.h4 White is not ideally situated to exploit the downsides of an early ...b7-b5.
Other moves are, of course, possible:
6...e5 has been played most often. After 7.Nb3 Be7 8.Bg5 Be6 9.Bxf6 Bxf6, 10.g3 (Fedoseev – Sadzikowski,
Karlsruhe 2017) is one example of White’s making use of the h2-h4 advance. An eventual Bf1-h3 is planned.
Alternatively, 6...e5 7.Nb3 Be6 8.f4 g6 9.Be2 Nbd7 10.g4 h5 11.g5 Ng4 12.Rf1 exf4 13.Bxf4 Nde5 14.Qd2 Qc7 15.0-
0-0 was the course of Oparin – Svidler, Zürich 2017. White’s position was quite promising, although he went on to lose.
6...e6 could be met by 7.h5, while 7.g4 d5 is another obscure possibility, waiting to be tried.
6...g6?! is the one move that makes no sense at all!


It is hard to make h2-h4 useful after 6...b5. The early attack on e4 means that White most likely won’t be able to reach
a normal position with opposite-side castling. Therefore, with this move White tries to claim that ...b7-b5 was premature
and can be exploited despite White’s useless and weakening h2-h4 lunge.

7...b4 8.Nd5 Bb7 9.Bc4

On 9.Nxf6+ gxf6, White would wish that the pawn were back on h2.

9...Nbd7 10.Bg5 e6 11.Nxf6+ Nxf6 12.Qe2 h6?

Black probably missed White’s move 14. Better is 12...Be7, when it is hard to see the point of White’s position.
Where will his king find safety? Meanwhile, the sacrifice 13.Bxe6 does not look sufficient.

13.Bxf6 Qxf6

13...gxf6 would leave Black with serious problems after 14.Qh5.

14.Bb5+! Ke7

14...Kd8 15.Bc6 is also better for White.

15.Nc6+ Bxc6 16.Bxc6 Qxb2 17.0-0

Black’s main problem now is that he cannot easily develop his kingside.


17...Ra7 is met by 18.Qe3 Rc7 19.Qb6.


It is simpler to just capture the a-pawn immediately.

18...Rc8 19.Qa7+ Kf6 20.Qxa6 Be7 21.a5 g5

With the centralizing 21...Qd4, Black would be fighting on. The queen would go to c5.

22.Qc4! Rc7 23.a6 Ra7 24.Bb7

Instead, the black rook ends up entombed.

24...Qe5 25.Qxb4 d5 26.Qb6 Bd6 27.f4 gxf4 28.exd5 f3 29.Rxf3+ Ke7 30.Ra4 Rg8

If first 30...Qh2+, then 31.Kf1 Rg8 32.Rf2 and Black runs out of checks, while the rook on a7 is under attack and
White is also breaking through with dxe6.

31.Raf4 f5 32.Qxa7 1-0

Index of Complete Games

Isidor Gunsberg – Louis Paulsen

Frankfurt 1887
Frederick Yates – Savielly Tartakower
Budapest 1926
Lewis Isaacs – Abraham Kupchik
Bradley Beach 1928
Gösta Stoltz – Isaac Boleslavsky
Groningen 1946
Vlastimil Stulík – Karel Opočenský
Prague 1945
Christian Poulsen – Miguel Najdorf
Buenos Aires Olympiad 1939
Thomas Luther – Leonid Yudasin
Budapest 1989
Lev Psakhis – Vladimir Tukmakov
USSR 1979
Vassily Ivanchuk – Maxime Vachier-Lagrave
Istanbul Olympiad 2012
Bryan Smith – Hristos Banikas
Rethymnon 2009
José Cubas – Sandro Mareco
São Paulo 2015
Isán Ortiz Suárez – Ruifeng Li
Las Vegas 2014
Dmitry Kokarev – Wang Hao
Russian Team Chp, Sochi 2012
Wei Yi – Yinglun Xu
Ho Chi Minh City 2017
Vasif Durarbeyli – Anton Korobov
FIDE World Cup, Tromsø 2013
Fidel Corrales Jiménez – Orlen Ruiz Sánchez
Panama City 2011
Viktor Bologan - Ivan Cheparinov
European Club Cup, Kemer 2007
Vladimir Onischuk – Maxime Vachier-Lagrave
Bastia 2016
Anatoly Karpov – Miguel Quinteros
Lucerne Olympiad 1982
Dmitry Jakovenko – Magnus Carlsen
Nanjing 2009
Viswanathan Anand – Maxime Vachier-Lagrave
London 2015
Jaan Ehlvest – Nigel Short
Pärnu 1996
Allan Stig Rasmussen – Tiger Hillarp Persson
Danish Team Chp 2008
Pouya Idani – Yuriy Kuzubov
Dubai Open 2015
Lenier Domínguez Pérez – Alexander Morozevich
Wijk aan Zee 2009
Zaven Andriasian – Ian Nepomniachtchi
Aeroflot Open, Moscow 2010
Alexander Morozevich – Veselin Topalov
Monte Carlo (rapid) 2004
Viswanathan Anand – Erwin L’Ami
Wijk aan Zee 2013
Dejan Pikula – Branko Damljanović
Vrnjačka Banja 2012
Fabiano Caruana – Boris Gelfand
Wijk aan Zee 2014
Magnus Carlsen – Wesley So
St. Louis 2015
Franciszek Borkowski – Aleksander Wojtkiewicz
Polish Championship, Słupsk 1989
Dmitry Chuprov – Peter Svidler
Smolensk 2000
Nigel Short – Garry Kasparov
PCA World Chp (8), London 1993
Bassem Amin – Alexander Areshchenko
Antalya 2013
Jakov Geller – Vladimir Belov
Sochi 2006
Erik Blomqvist – Viswanathan Anand
Gibraltar 2016
Thal Abergel – Maxime Vachier-Lagrave
Pau 2008
Emory Tate – Leonid Yudasin
Chicago Open 1997
Hikaru Nakamura – Nikolai Ninov
French Team Chp, Évry 2008
Fabiano Caruana – Maxime Vachier-Lagrave
Wijk aan Zee 2015
Stefan Bromberger – Nils Grandelius
Munich 2017
Magnus Carlsen – Ruslan Ponomariov
Moscow (Blitz) 2008
Bartosz Soćko - Anton Korobov
Karlsruhe 2016
Magnus Carlsen – Alexander Grischuk
Stavanger 2015
Arkadij Naiditsch – Sunilduth Lyna Narayanan
Isle of Man 2016
Michael Adams – David Navara
Biel 2015
Dimitri Reinderman – Loek Van Wely
Dutch Chp, Rotterdam 1999
Ger van Perlo – Grigory Sanakoev
“Serbia 30” Correspondence, 1983-85
Ilya Smirin – Anton Korobov
Poikovsky 2016
András Adorján – Walter Browne
Wijk aan Zee 1972
Alexander Panchenko – Alvis Vitolinsh
USSR Chp Semifinal, Dniepropetrovsk 1980
Michael Adams – Boris Gelfand
Dortmund 2006
Gata Kamsky – Hikaru Nakamura
U.S. Chp, Saint Louis 2012
Josef Gheng – Arkadij Naiditsch
Neckar Open 2004
Judit Polgár – Dariusz Świercz
Istanbul Olympiad 2012
Pentala Harikrishna – Maxime Vachier-Lagrave
Biel 2014
Viswanathan Anand – Maxime Vachier-Lagrave
Bastia (rapid) 2016
Magnus Carlsen – Radosław Wojtaszek
Wijk aan Zee 2017
Alfredo Giaccio – Anton Kovaliov
Pinamar 2006
Milan Pacher – Hovhannes Gabuzyan
Mashhad 2017