Nimzowitsch: Move by Move

by Steve Giddins

Contents

About this Publication
Bibliography
Series Foreword
Introduction

Chapter One: Nimzowitsch the Attacker
Chapter Two: Nimzowitsch the Defender
Chapter Three: Nimzowitsch the Blockader
Chapter Four: Nimzowitsch the Strategist
Chapter Five: Nimzowitsch in the Endgame

Back to Contents Page

Nimzowitsch: Move by Move

Aron Nimzowitsch is one of the most important figures in chess history. He was one of the world's strongest players
and contributed enormously to the development of chess both through his games and his writings, which influenced
many grandmasters who followed him. Nimzowitsch was a leader of the Hypermodern School, which formed
revolutionary ideas on chess strategy to challenge previously held beliefs and created many new opening systems.

In this book, Steve Giddins selects and studies his favourite games by Nimzowitsch and examines Nimzowitsch's
skills in the vital areas of attack, defence, strategy and endgame play. He demonstrates how we can all improve by
learning from Nimzowitsch's masterpieces.

Move by Move provides an ideal platform to study chess. By continually challenging the reader to answer probing
questions throughout the book, the Move by Move format greatly encourages the learning and practising of vital
skills just as much as the traditional assimilation of knowledge. Carefully selected questions and answers are
designed to keep you actively involved and allow you to monitor your progress as you learn. This is an excellent
way to improve your chess skills and knowledge.

· Learn from the games of a chess legend

· Important ideas absorbed by continued practice
· Utilizes an ideal approach to chess study

Steve Giddins is a FIDE Master and a former editor of British Chess Magazine . He spent a number of years of his
professional life based in Moscow, where he learnt Russian and acquired an extensive familiarity with Russian
chess literature and the training methods of the Russian/Soviet chess school. He's the author of several
outstanding books and is well known for his clarity and no-nonsense advice. He has also translated over 20 books,
for various publishers, and has contributed regularly to chess magazines and websites.

Publisher Information

Nimzowitsch: Move by Move

by Steve Giddins

First published in 2014 by Gloucester Publishers plc (formerly Everyman Publishers plc)
Northburgh House, 10 Northburgh Street, London EC1V 0AT.

Copyright © 2014 Steve Giddins

The right of Steve Giddins to be identified as the author of this work has been asserted in accordance with the
Copyrights, Designs and Patents Act 1988.

All rights reserved. No part of this work may be reproduced, or transmitted in any form or by any means, without
prior permission of the publisher.

British Library Cataloguing-in-Publication Data
A catalogue record for this book is available from the British Library.

All other sales enquiries should be directed to Everyman Chess, Northburgh House, 10 Northburgh Street, London
EC1V 0AT
tel: [phone]; fax: [phone]
email: [e-mail]
website: www.everymanchess.com

Everyman is the registered trade mark of Random House Inc. and is used in this work under licence from Random
House Inc.

EVERYMAN CHESS SERIES (formerly Cadogan Chess)
Chief Advisor: Byron Jacobs
Commissioning Editor: John Emms
Assistant Editor: Richard Palliser

Back to Contents Page

Back to Contents Page

Bibliography

Books by Nimzowitsch

Blockade (Chess Enterprises Inc. 1980)
My System (G. Bell & Sons 1948)
Chess Praxis (Dover 1962)

Books by other writers

Aron Nimzowitsch: A Reappraisal , Raymond Keene (G. Bell & Sons 1974)
Aron Nimzowitsch: On the Road to Chess Mastery 1886-1924, Per Skjoldager & Jørn Erik Nielsen (McFarland &
Co. 2012)
Aron Nimzowitsch 1928-1935, Rudolf Reinhardt (New in Chess 2013)
Moi Veliki Predshestvenniki, Tom 1 , Garry Kasparov (Ripol Classics 2003)
Learn from the Grandmasters, (ed.) Raymond Keene (Batsford 1975)
Play 1 b3!, Ilya Odessky (New in Chess 2008)

Back to Contents Page

Back to Contents Page

Series Foreword

The Move by Move format is designed to be interactive, and is based on questions asked by both teachers and
students. It aims - as much as possible - to replicate chess lessons. All the way through, readers will be challenged
to answer searching questions and to complete exercises, to test their skills in key aspects of the game. It's our firm
belief that practising your skills like this is an excellent way to study chess.

Many thanks go to all those who have been kind enough to offer inspiration, advice and assistance in the creation
of Move by Move . We're really excited by this series and hope that readers will share our enthusiasm.

John Emms
Everyman Chess

Back to Contents Page

Back to Contents Page

Introduction

Although he was neither world champion, nor even a direct world championship challenger, Aron Nimzowitsch was
one of the most important and influential players in chess history. He belongs to that select band of players (the
other main one, ironically enough, being his arch-enemy Tarrasch) who have influenced the development of chess
style as much (or more) by their writings than by their play. His books The Blockade, My System and Chess Praxis
have had an enormous influence on the game, and until the computer became the main source of chess experience
for the present generation of players, it was hard to find any strong player who had not read these books.

Given the wealth of literature on Nimzowitsch (see the

Bibliography

for the main sources used in this book), I see

no point in giving too much biographical information here. Suffice it to say that he was born on 7th November 1886
(new style) into a Jewish family in the city of Riga, which was then part of the Russian Tsarist Empire. Of course,
Nimzowitsch's name is indelibly associated with Denmark, but he did not settle in that country until after the First
World War. Bent Larsen, who was greatly influenced by Nimzowitsch, famously joked that while he doubted that
there really was a 'Soviet School of Chess' ("How can Tal and Petrosian belong to the same school?"), there is a
Danish school, "even if it was founded by a Jew from Riga!".

Although a strong master before the Great War, it was in the mid-1920s that Nimzowitsch really started to flourish
as a player, and was one of the top 4-5 players in the world during the late 1920s and early 1930s. However, like
some other possible challengers of the pre-1948 period, he was destined never to get a shot at the world
championship, although it is probably fair to say that his eccentric and somewhat unstable play would not have
given him much chance in a match against such titans as Capablanca or Alekhine. He died in Denmark on March
16th, 1935, at the sadly early age of 48.

As a player, Nimzowitsch is inextricably associated with the Hypermodern School, that group of highly gifted young
masters (including also Réti and Breyer), who emerged to prominence after 1918, with their revolutionary ideas
regarding chess strategy. These masters challenged the accepted idea that 1 e4 and 1 d4 were the only truly
correct opening moves, and more generally the theory that one should seek to occupy the centre with pawns in the
opening. They developed new opening systems, based on piece control of the centre from a distance, and
propagated their ideas in what became classic works of chess literature, such as Réti's Modern Ideas in Chess and
Nimzowitsch's My System and Chess Praxis.

As with most such movements (e.g. modern art, atonal music, etc), there was fierce resistance to the new ideas
from certain of the 'old guard', but gradually the effectiveness of the new approach became understood, and a
Hegelian synthesis took place, with the best of the new methods being incorporated into the play of leading stars,
such as Capablanca and Alekhine, whilst the more extreme and eccentric aspects were left to die a natural death.
The eventual result was a great enrichment of our understanding of the game.

In Nimzowitsch's case, it was ideas such as restraint, blockade and over-protection which were the most dominant,
although he also laid enormous stress in his writings on centralization. His contribution to modern opening theory
was huge, with a large number of systems that today are taken for granted, being originally based, to a greater or
lesser extent, on his ideas. Pride of place goes to the Nimzo-Indian Defence, which remains to this day possibly the
single most popular and respected defence to 1 d4 at GM level, but one can also throw in the Queen's Indian, the
Winawer variation of the French, various lines of the English and Réti Openings, as well as a host of other, less
popular variations, such as 1 e4 Nc6. Few great masters in history can have contributed more to the openings.

Given the wealth of material available, both by and about Nimzowitsch, I should say a word about what I see as the
aim of the present volume. What I have sought to do here is present a selection of Nimzowitsch's best and most
instructive games, annotated for the less experienced player, in the style of the Move by Move series. Famous
though Nimzowitsch's own books are, he was not always the easiest of commentators to follow, and tended to wrap
many of his ideas in rather flowery language. Furthermore, few of his most famous games have been re-evaluated
in the light of the computer's analyses, which so often throw a completely new light on games we know so well.
Kasparov's Great Predecessors series, for example, only includes three Nimzowitsch victories, two of those being
the immortal games against Johner and Sämisch. There are many other famous Nimzowitsch games, where the
computer throws up interesting discoveries, and I have presented many of these in this volume.

I should also comment on the sources which have influenced this book. Readers familiar with Ray Keene's book
Aron Nimzowitsch: A Reappraisal will easily see that this is the volume to which I owe the greatest debt. That book
just happened to be one of the first chess books I ever obtained, for the simple reason that it was one of the very
few serious chess books to be found in the Dorchester branch of the W.H. Smith bookshop chain, to which I
repaired as a chess-mad 13-year-old, holidaying in Dorset and clutching the accumulated pocket money I had
saved up for the purpose. It was a fortunate choice, because I found the book enormously inspiring, and after
reading and re-reading it, soon reached the point where I could remember vast swathes of it by heart. The passage
of four decades since that time has still not materially dimmed its influence.

I was also fortunate in the timing of the present book. A year or so before I started work, the Danish chess
enthusiasts Per Skjoldager and Jørn Erik Nielsen produced their monumental volume

Aron Nimzowitsch: On the

Road to Chess Mastery, 1886-1924, the fruit of many years' diligent research. This made available to me many
hitherto unseen games and annotations, both by Nimzowitsch himself and other contemporaries. Then, just weeks

hitherto unseen games and annotations, both by Nimzowitsch himself and other contemporaries. Then, just weeks
before I put pen to paper (or, more accurately, fingers to keyboard), New in Chess published German enthusiast
Rudolf Reinhardt's similar volume, Aron Nimzowitsch 1928-1935, which did much the same job for Nimzowitsch's
later years. Admittedly, there is a gap of four years between the periods covered by these two volumes, but they
were both enormously useful nonetheless, and the chess world owes their authors a profound debt of gratitude for
many years of painstaking research.

This has been a very enjoyable book to write, and I hope my readers will derive similar pleasure from reading it.
Nimzowitsch was a great player, whose best games are wonderfully instructive and also highly entertaining. The
reader will find powerful attacks, fearless defences, the deepest positional manoeuvres and filigree endgame
technique. Most amazing of all are some of the positions Nimzowitsch was able to bring about - I know of no other
master who could so often reduce opponents to zugzwang on a full board, or achieve such visually striking
positions, often even against opponents of the front rank. Enjoy!

Steve Giddins
Rochester, Kent,
May 2014

Back to Contents Page

Back to Contents Page

Chapter One

Nimzowitsch the Attacker

One does not think of Nimzowitsch as primarily an attacking player, in the way one does with, say, Alekhine or Tal.
However, like any great player, he was a superb tactician, and the natural outcome of superior play is often the
opportunity for a kingside attack. Steinitz emphasised that in such situations, the player with the advantage must
attack, or else risk losing his advantage, and so one would naturally expect to find many fine kingside attacks
amongst the games of any great master. Nimzowitsch was certainly no exception.

Levenfish-Nimzowitsch

sees a vigorous counterattack, soon after the opening, which culminates in the crushing of

a young but highly gifted master.

Nimzowitsch-Haakanson

is one of Nimzowitsch's most famous miniatures, which

shows the crushing effects of his favourite anti-French Defence strategy, and is crowned with a neat queen
sacrifice.

Nimzowitsch-Wolf

is a longer effort, arising from a favourite Nimzowitsch opening scheme.

Nimzowitsch-Olson

is a curiosity, notable especially for Nimzowitsch's opening plan, which was a forerunner of the

modern 'Botvinnik System' in the English Opening.

Nimzowitsch-Rubinstein (Berlin 1928)

is another fascinating

flank opening battle, whilst

Nimzowitsch-Ahues

shows how deeply Nimzowitsch could calculate. Finally,

Nimzowitsch-Flohr

completes this chapter, with the crushing of a player who within a year or two had become one

of the world's strongest players, and, moreover, one noted for his solidity.

Back to Contents Page

B10

Levenfish,G
Nimzowitsch,A

Russian Championship, Vilnius

1912

[Steve Giddins]

1.e4 c6 QUESTION: Hardly the most

attacking of opening lines is it?
ANSWER: Indeed! The Caro-Kann has a
reputation as a rather solid, defensive line, in
which Black is primarily playing for equality.
However, it is possible to interpret it in a more
dynamic, unbalancing fashion, which tended
to be how Nimzowitsch played it. Indeed, in
his book 'Chess Secrets I Learned from the
Masters', Edward Lasker quoted a
conversation he once had with Nimzowitsch
about this opening, in which the latter said of
the opening that: "The ambitious aim of 1...c6
is to prove that 1 e4 is premature". That is
perhaps overstating the case a little, but
reflects Nimzowitsch's thinking in the line.

2.c4 This is a relatively rare approach by

White, although it has had periods of
popularity, especially in the 1980s.

[ The main line is, of course, 2.d4 d5 3.

c3

dxe4 4.

xe4 , whereupon, rather than the

most popular and solid move

f5 ,

( Nimzowitsch's more ambitious

interpretation of the black position was

4...

f6 . )]

2...e6?!

[ QUESTION: That looks a bit strange. What

is wrong with 2...d5 ?
ANSWER: Nothing at all. The latter is the
best move, of course. ]

3.

f3

[ 3.d4 looks more natural. ]

3...d5 4.exd5 exd5 5.cxd5 cxd5 6.

b5+

c6 7.0-0

[ Keene suggests 7.

c3 in the hope of

preventing the ...Bd6 and ...Nge7 set-up
adopted by Black in the game, but then
White must reckon with the reply d4
. As was often the case with the older
masters, Levenfish develops his pieces
somewhat mechanically, but the latter was
something of which Nimzowitsch was
scarcely ever guilty, and he immediately
seizes the chance to break the symmetry. ]

7...

d6 8.d4 ge7 QUESTION: Why put the

knight here, rather than on f6? Surely the
latter is the most natural square for the knight?

[ ANSWER: With the bishop on d6, 8...

f6

would run into the rather awkward 9.

e1+

e6 10.g5 . If Black wanted to develop his
knight to f6, then he should have played
his bishop to e7.
QUESTION: OK, but in that case, why didn't
he do that last move?
ANSWER: The plan of ...Bd6 and ...Nge7
was chosen quite deliberately, as a means
of breaking the symmetry, as well as
avoiding the pin Bg5 after ...Nf6. Basically,
what is happening here is that Nimzowitsch
is trying to play for a win, by unbalancing a
symmetrical structure. This is something on
which he was a great expert, and indeed, in
his book 'Chess Praxis', he devotes an
entire chapter to the topic of 'The
Asymmetric Treatment of Symmetrical
Variations'. In the course of that chapter,
he states as a rule of strategy that, in such
positions, "Nf3 should be answered by ...
Ne7 and Ne2 by ...Nf6". As always with
Nimzowitsch, one wrinkles one's nose at
the somewhat categorical tone of 'should',
but the fact that such asymmetrical
development is frequently fully justified
cannot be disputed. ]

9.

g5 f6 10.h4 0-0 11.bd2 QUESTION:

It looks as though Levenfish is applying
Nimzowitsch's own precepts, answering ...Nc6
with Nd2?

[ ANSWER: Maybe, but actually, the move

just looks a bit passive. Presumably White
was worried about the pin after 11.

c3 g4

, but then 12.

e2 seems perfectly adequate.

At least then the white knight would be
actively placed on c3, attacking Black's
IQP. ]

11...

g4 12.xc6? QUESTION: Why give up

the bishop like this?
ANSWER: It is a classic case of a poor
positional decision, resulting from a tactical
miscalculation. Levenfish understood only too
well that, ceteris paribus, it makes no sense
to surrender the bishop-pair like this, but he
thought a tactical point justified the decision.

xc6! An unpleasant surprise for White, I
suspect.

[ Levenfish probably thought the text was

impossible because of the reply in the game,
and therefore that Black would be forced
into 12...bxc6 , after which 13.

c1

leaves White with some pressure against
the backward c-pawn. ]

13.

b3 This is the point of White's previous

move, and the only consistent follow-up. Two
black pawns are attacked.

b4 The only way to defend both pawns.
QUESTION: So is this what Levenfish had
missed?
ANSWER: No, I doubt it. He had seen this
and planned...

14.

e5 This was the real point of White's

play.
QUESTION: What is the idea?
ANSWER: Well, what are you going to play?
Your bishop on g4 is attacked.

[ QUESTION: OK, but so is your knight on d2.

Suppose after 14.

e5 I take it with xd2 ?

ANSWER: Then I continue 15.

xg4

and it is your move again. I am attacking the
pawn on b7.
QUESTION: OK, but can't I win material
with g5 ( I can see that 15...

xd4??

loses to 16.

d3 ) 16.g3 f5 followed by ...

f4?
ANSWER: Not so fast, Moriarty! Now I have

17.

xb7 c8 ( 17...a5? leads to mate

after 18.

h6+ h8 19.e5+ ) , and now

the shock move 18.

e5! which rescues my

bishop with tempo, as I am threatening mate
in one with 19 Nh6#. Black has nothing
better than

xe5 , granting White the

choice between 19.

xc8 ( or 19.xd5+

f7 20.h6+ g7 21.xf5+ and White
has three pawns for the piece ) 19...

axc8

20.

xe5 c2 and Black regains the pawn

with equality. ]

[ QUESTION: Quite impressive calculation by

both players!
ANSWER: Yes, although the computer
points that after 14.

e5 xd2 15.xg4

, Black also has

d7 16.h3 a5

followed by ...Bb6, when his pressure
against d4 gives him some advantage. Even
so, Nimzowitsch's choice in the game is
superior. ]

14...

xe5! 15.xb4 d3! Firmly seizing the

initiative.

16.

xb7 e2 17.fb1 The only square, but

pitifully passive.

c8!? Nimzowitsch relished this rather
picturesque example of the triumph of
centralization. The white forces find
themselves scattered to the four winds, whilst
Black's army dominates the central lines. Even
so, the text does offer a second pawn, which
the ruthlessly materialistic computer is loath
to do.

[ The more parsimonious 17...

f7

may have been a little better. ]

18.

f1?!

[ The silicon beast prefers to have some

more material for which to suffer, and
advocates 18.

xa7 . Then after a line such

as

f7 19.a3 b6 , Black's initiative

compensates for the two pawns, but
whether it does any more than that remains
unclear. ]

18...g5 Initiating a vigorous attack on the

enemy king, whose major pieces have so
cruelly abandoned it.
QUESTION: Isn't it dangerous to advance the
pawns in front of the black king like this?
ANSWER: Not really. White has no pieces
with which to bring any threats against the
king, so its exposure really does not matter in
this position.

19.

g3 f5 20.e5 f7 21.a6 f4 22.e1?

Faced with the threat of ...f3, Levenfish caves
in and jettisons an exchange, without really
succeeding in relieving his position.

[ The main line is 22.f3

c2 (threatening ...

Nxe5) 23.

a3 ( the computer's 23.e6

is a tougher defence, but then simply

d7

24.

xd7 xd7 leaves White tied hand and

foot: for example, 25.b4

xe5 26.dxe5 d4

and the passed d-pawn will soon prove
decisive ) 23...

xe5 24.dxe5 b6+ 25.h1

f2 . ]

[ Probably the best defence is the computer's

cold-blooded 22.b4 f3 23.

e6 when Black

again does best simply to remove the
queens and pursue his initiative in the
endgame:

e8 24.xe8+ xe8 25.gxf3

xf3 26.b3 (with the point xe5 27.e3!
) and although Black is still better, White
remains on the board. ]

22...

xe1 23.xe2 xg2 Perhaps Levenfish

had just overlooked this simple blow, and
instead thought he was picking up two pieces
for the rook. Now it is all over.

24.

d2 h4 25.f3 g6 26.h1 g4

27.

d2 d7 28.g1 c2 29.h3 g3

0-1

C02

Nimzowitsch,A
Haakanson,A

Kristianstad

1922

[Steve Giddins]

1.e4 e6 2.d4 d5 3.e5 c5 4.

g4

This was Nimzowitsch's last word in his
evolution of the Advance French for White.

[ QUESTION: 4.

g4 seems to allow the

pawn centre to be destroyed after cxd4 ?
ANSWER: It allows the d4-pawn to be
eliminated, yes, but Nimzowitsch would
have vigorously rejected any suggestion
that White's whole pawn centre is being
wiped out. ]

[ After starting out defending the d4-pawn

with the normal move 4.c3
(see Nimzowitsch-Salwe, for example),
Nimzowitsch had gradually developed his
ideas to the point where he decided that it
was the e5-pawn that matters. In a sense,
he had always understood this, of course, as
had every other master – that was why
they defended the d4-pawn with 4 c3, so
that the d4-pawn in turn would support its
colleague on e5. But what was new about
Nimzowitsch's understanding of the position
was the idea that one could allow the d4-
pawn to fall, providing one defended the e5-
pawn with pieces, and thus maintained it
that way. ]

[ QUESTION: But I still don't understand the

specific move 4.

g4 . What does that

contribute to the defence of the e5-pawn,
and besides, one is always told not to bring
out the queen in the opening?
ANSWER: Well, the latter is a bit of a
shibboleth. Strictly speaking, one should
not bring out the queen in the opening,
unless there is a good reason to do so,
which, of course, does not really help a
great deal as a rule of thumb.
Nimzowitsch's argument was that the queen
is good on g4, because she prevents Black
from developing his king's bishop, due to
the attack on g7, and also can come to g3,
to support the key e5-pawn.
QUESTION: So, you like 4 Qg4 then?
ANSWER: Actually, no. Theory has
established that Nimzowitsch's strategic idea,
although in itself an excellent one, simply
cannot be achieved with the early queen
move. Black has a number of good defences,
and the move is now completely out of
fashion. ]

4...cxd4 5.

f3 c6 6.d3 f5? I am not

very keen on this move.
QUESTION: Why?
ANSWER: Mainly because Black deprives
himself of the chance to attack the e5-pawn
with ...f6. As we have said, the e5-pawn is the
key to White's idea, and Black should not give
it such an easy life, in my view.
QUESTION: So what should Black play?

[ ANSWER: He has several decent moves,

but the simplest and probably strongest is

6...

c7! , immediately attacking the e5-pawn.

The main line then is 7.

f4 ( if 7.0-0

then Black can just take the pawn:

xe5

8.

xe5 xe5 9.f4 f6! 10.g3 h5

and although White has some lead in
development, it is doubtful if it is worth two
pawns; 7.

g3 is the other way to defend e5,

but then Black continues in consistent
fashion by undermining e5 with f6!
: for example, 8.exf6

xg3 9.f7+ xf7

10.hxg3

f6 and Black is again simply

better, with an extra pawn and the queens
off ) 7...

b4! 8.0-0 ( 8.xd4? xd3+

9.cxd3

b6 10.b3 b4+ 11.1d2? g5!

, winning a piece, is an old trap ) 8...

xd3

9.cxd3

c2 10.xd4 xb2 and, once

again, White has some compensation for
his material, but it is not clear whether it is
really sufficient. This simple line,
immediately challenging the e5-pawn, is
close to being a refutation of 4 Qg4, in my
opinion. ]

7.

g3 ge7 8.0-0 g6?! This just walks

into the advance of the white h-pawn.

[ 8...

b6 is probably a better move, planning

...Bd7 and long castling, but Black still has
some problems in developing his kingside.
The basic focus of Nimzowitsch's strategy is
clear – all the while the e5-pawn remains on
the board, it exerts a significant cramping
effect on the black position. ]

9.h4! Black had hoped to have time for ...Be7

and ...0-0, but this would now drop the g-
pawn.

c7 10.e1 We can see the difference
between this position and the one in the
variations starting 5...Qc7. Here, Black does
not have the resource ...f6, to eliminate the
cramping e5-pawn.

d7 11.a3

[ QUESTION: What is that for? I expected
11.h5 .

ANSWER: There is no objection to the
immediate 11 h5, but there is also no hurry
for that move either, as Black cannot
prevent it. Instead, Nimzowitsch
anticipates Black's plan of castling long
(where else can he castle?) by preparing b4.
Among other things, that will open up the
possibility of Bb2 and recapturing the
sacrificed pawn on d4, as well as launching
a queenside attack.
QUESTION: You sound very enthusiastic

about the white position?
ANSWER: I am. The computer may say the
position is equal, but I do not really believe
that. White's position has great potential,
whereas it is difficult for Black to do
anything at all active. ]

11...0-0-0 12.b4 a6?!

[ As usual, this move just facilitates the later

opening of lines in front of his king by b4-b5.
I prefer something such as 12...

g8

, preparing ...Be7, and if 13.h5

h8

, followed by ...g5 and ...Nf7. Black would
still be worse, I think, but it looks a better
try than the game. ]

13.h5

ge7 14.d2 Defending the b4-pawn,

in readiness for a4.

h6?! Again, rather pitiful-looking, but it is hard

to recommend a decent move for Black.

[ QUESTION: What about 14...

e8

going after the h5-pawn? Isn't that a little
awkward for White?
ANSWER: Against this, the computer shows
a vigorous continuation: 15.a4!

xh5

16.

g5 d7 17.h2! e8 18.b5

and White has a virtually winning attack, for
instance,

b8 19.c3! and the c-file will be

torn open for the white rooks. ]

[ The pawn sacrifice 14...g5 may be the best

chance, although after 15.

xg5 g8

16.

h2 Black's position is still deeply

unenviable.
QUESTION: It seems amazing that his
position should have become so bad so
quickly?
ANSWER: Well, he has simply failed to
oppose White's basic strategy, which was
to maintain the e5-pawn and cramp Black.
Once White succeeded in that, Black was
always in for a difficult life.
QUESTION: So it all comes down to the
move 6...f5, then?
ANSWER: That is certainly the root cause
of Black's problems, yes. ]

15.a4 g5 16.b5 White's game almost plays

itself.

f4 17.

g4 b8 18.c3! This opening of the

c-file is decisive. Black has no way to meet
the threats which are coming.

e8 19.cxd4 d8 The way the black army
is huddled on the back rank reminds one of a
boxer, desperately trying to cover up on the
ropes. The towel could be thrown into the ring
at any point from here.

20.

c1 Now we get one of those amazing

positions, which Nimzowitsch was so adept at

bringing about.

b6 21.a5 a7 22.b6 a8 QUESTION:
Wow! It is surely hard to believe that such a
position should arise between two players,
without cooperation on Black's part?
ANSWER: Yes. Nimzowitsch was incredibly
good at achieving such visually striking
positions; I cannot think of another master
who reached such 'strangulation' positions so
often. Admittedly, this was only an exhibition
game, against a much weaker player, but later
in this book, we will see Nimzowitsch reach
such positions even against front-rank
masters, such as Sämisch and Flohr.

23.

c7 f5 24.c3 e7 Now a lovely final

combination crowns White's positional
achievement.

25.

xd5! xd4 A desperado sacrifice of the

knight, in an attempt to save material, but a
nasty shock awaits Black at the end of the
variation.

26.

xd4 exd5 EXERCISE: Can you spot the

crisp finish?
ANSWER: 27.

xd7+!

[ The smothered mate after 27.

xd7+ xd7

28.

e6# seems the only truly logical finish

to a game which has already seen the black
queen 'smothered' in the corner.
A lovely game by Nimzowitsch. Admittedly,
his opponent was totally outclassed, but
even so, the complete manner in which
Nimzowitsch achieved all his positional aims,
and then crowned things with a queen
sacrifice finish, deservedly makes this one of
his best-loved games. ]

1-0

A06

Nimzowitsch,A
Wolf,H

Karlsbad

1923

[Steve Giddins]

1.

f3

[ With regards to the note to move 22, below,

P.Clarke-R.Keene, British Championship,
Eastbourne 1973, saw 1.d4

f6 2.c4 b6

3.

c3 b7 4.f3 e6 5.e3 b4 6.d2 0-0

7.

e2 d6 8.0-0 bd7 9.a3 xc3 10.xc3

e4 11.c1 f5 12.b3 f6 13.d2 g6

14.

xe4 xe4 15.f3 g5 16.f2 b7

17.

d2 e8 18.b4 f6 19.a4 h5 20.f4

h4 . After 21.xh5? xh5 Clarke did not
last very long, being slaughtered by the now

unopposed enemy light-squared bishop:

22.d5

h6 23.h3 g6 24.h1 c6 25.dxe6

c5 26.bxc5

xh3+ 27.g1 exe6 28.cxd6

xg2 0-1. ]

1...d5 2.b3

[ QUESTION: I thought the Réti Opening

usually involved either 2 c4 or 2.g3 ?
ANSWER: The pure Réti or King's Indian
Attack do indeed involve those moves, but
Nimzowitsch was fond of the text, which
usually amounts to a sort of Queen's
Indian Defence, with a move in hand.
QUESTION: So what is the difference from
Larsen's Opening 1 b3?
ANSWER: The main difference is that
Nimzowitsch's move order prevents Black
establishing his pawn on e5, such as
happens after 1 b3 e5 2 Bb2 and now either
2...Nc6 or 2...d6.
QUESTION: Is Nimzowitsch's approach
better?
ANSWER: Not necessarily. It is more a
matter of taste. Larsen and other followers
were perfectly happy to invite Black to
occupy the e5 point, so as then to use that
pawn as a target. Nimzowitsch's approach
is slightly less ambitious. One can argue in
its favour that the move 1...d5 slightly
weakens the central dark squares
(precisely because it is harder to fortify the
e5-square, in the absence of the move ...d6),
and so it makes sense to put the bishop on
the a1-h8 diagonal now. ]

2...

f6 3.b2 c5 4.e3 c6!? QUESTION:

Why the punctuation? Isn't it the most natural
thing in the world to put the knight on c6?

[ ANSWER: Yes, but it is also a tad

cooperative in this position. White is clearly
building his development around the control
of the e5-square, and inviting him to pin and
later capture the knight on c6 just allows him
to pursue that plan all the more. It would be
a little more difficult for him to do so after

4...e6 , since then 5.

b5+ d7 6.xd7+

bxd7 would leave Black with better control
of e5. ]

5.

b5 d7

[ QUESTION: This looks a bit passive. Isn't
5...

g4 better?

ANSWER: It is another approach, but it is
not clear whether it is better, and Black
would risk having his queenside pawns
doubled. Nimzowitsch did in fact face this
move once, and the resulting miniature A.
Nimzowitsch-H.Johner (Exhibition game,

Berne 1931) is worth examining:

6.

xc6+ bxc6 7.h3

A) QUESTION: What is wrong with 7...

h5

?
ANSWER: Probably nothing at all. I
suspect Johner was afraid of the radical
plan 8.g4

g6 9.e5 c7 10.h4

when it looks as though he will be forced
into a horrible weakening of his pawns
after h6 . ( However, this is not the case,
as he has the resource 10...

e4 11.f3?

xg4!! and Black is winning: for example,

12.

xg4 g3+ 13.f2 xf3 , etc. );

B) 7...

xf3 8.xf3 c7 9.d3 e6 10.c3

QUESTION: Isn't there a pawn en prise on
f6?

( ANSWER: Yes, but after 10.

xf6 gxf6

11.

xf6 g8 12.f3 g7 Black has

good compensation.

) 10...

d6 11.a4 e5 12.xe5 xe5

13.0-0

d7 14.e2 0-0 15.d2 f5

16.

a5 fb8? ( Black would have been

fine after 16...

d6 , but instead, he

initiated some ill-judged tactics:

) 17.

xc5!! QUESTION: Err, doesn't this

lose material?
ANSWER: Technically, yes, but
Nimzowitsch had seen further.

b5 18.xd7 xa1 This desperado is
the point of Black's idea, but he had
underestimated the danger from White's
queen and knight duo.

19.

c7! c3 ( And not 19...xa2?

20.

xc6 .

) 20.a4!!

b4 21.e5 Extremely simple,

but highly effective. White simply intends
Qf7+ and Qxe6, with all manner of
smothered mate-type threats against the
enemy king. With his queen and rook shut
out of play, Black cannot defend.

B1) 21...

h8 22.f4 ( 22.e7! )

22...

bb8 23.f3 c5 24.f7+ g8

25.

d6! f8 26.g3 g6 27.h4 b6!

( White wins too after both 27...

ad8

28.

b7; and 27...xc2 28.h5! d1+

29.

h2 xh5+ 30.h3 ) 28.e7 d8

29.

xe6+ g7 30.xf5+ h8 31.e7

c7 32.xc6 ac8 33.d4 and wins.;

B2) 21...h6 22.

f7+ h8 23.xe6

is not really any better: for instance,

b7

24.

g6+ h7 25.xf5 f6 26.xf6

gxf6 27.

f4 and White has three pawns

for the exchange, with a winning
endgame advantage.;

B3) 21...

f8 Alternatively: 22.xc6!

h8 23.d6!! 1-0
Annexing a full rook in broad daylight.
An excellent example of the fearsome
attacking strength of a queen and knight
together. ]

After that somewhat lengthy digression, we
return to the game and Wolf's 5...Bd7:

6.0-0 e6 7.d3

e7 8.bd2 0-0 9.xc6

QUESTION: Isn't this a bit odd, exchanging
without waiting for Black to play ...a6?
ANSWER: As we will see elsewhere in this
book, such unprovoked exchanges were a
Nimzowitsch speciality. He was not a great fan
of the bishop-pair, and generally preferred to
bring about positions where his knights would
be stronger than the enemy bishops. Here,
White's play is all about the e5-square, and so
he is happy to exchange off one of the enemy
minor pieces which defends that square.

xc6 10.e5 This position is very much a
tabiya for this whole variation, and now Black
must make a major decision about the threat
to capture on c6.

d7!? QUESTION: That is the last move I
would have thought of! Surely Black should
do something to stop the c-pawns being
doubled, shouldn't he?
ANSWER: An interesting question. One
would probably think so, yes, but this position
is much deeper than it appears.

[ For an exhaustive and masterly discussion

of the position (and, indeed, the entire
variation), I can strongly recommend Ilya
Odessky's magnificent book 'Play 1 b3',
which devotes a whole chapter to the
position after 10 Ne5. We do not have
sufficient space to go into matters in the
same level of detail here, but it is certainly
true that any of the moves 10...Be8, 10...
Rc8 and 10...

c7 are perfectly playable,

and preserve the integrity of the black
pawn structure. However, Odessky actually
says that 10...Nd7 is the move he least likes
to face as White, so we should certainly
not condemn it. ]

[ Nimzowitsch reached the tabiya position in

a number of games, and faced all of the
obvious black moves here. As an illustration,
let us look at the course of his game against
Rubinstein (Semmering 1926), probably the
strongest opponent against whom
Nimzowitsch played this position. Rubinstein,
a great lover of the bishop-pair, predictably
chose 10...

e8 .

QUESTION: That looks very logical. Why

not preserve the bishops?
ANSWER: Well, the move does indeed keep
the two bishops, but it also costs a tempo,
which White exploits to strengthen his
control of e5: 11.f4

d7 (Black needs to

challenge the knight quickly, else White will
follow up with Rf3-h3, and the kingside
threats will rapidly become extremely
dangerous) 12.

xd7 ( 12.g4 was a later

Nimzowitsch recommendation, but Odessky
suggests White has nothing concrete after

f6 13.df3 g6 ) 12...xd7 13.e4 f6

14.

f3 f7 15.a4 b6 16.ae1 a6 17.f5

dxe4 18.

xe4 e5 19.e3 b5 20.g3 h8

21.

f3 . Now Black would have been

absolutely fine after

A) Instead, Rubinstein blundered with
21...bxa4? and was hit with 22.

xe5! e8

( 22...fxe5 23.

xe5 f6 24.xf6!

loses at once ) 23.

g4 g8 and now the

quickest win was 24.

g6+ ,

( but Nimzowitsch settled for an extra

pawn with 24.

xf7+ xf7 25.xa4

and went on to win. );

B) 21...

d6 . ]

11.

df3 QUESTION: Now you are confusing

me! Why on earth didn't White just take on c6
after 10...Nd7?

[ ANSWER: One would certainly expect the

great exploiter of doubled pawns,
Nimzowitsch, to do exactly that, but in reality,
Odessky suggests that White has nothing
after 11.

xc6 bxc6 12.e4 f6! .

QUESTION: But why? Aren't the pawns
weak?
ANSWER: Usually, they would be, but as
Odessky explains, in this case the key issue
is the white knight on d2, which is misplaced.
In these structures, the knight should be on
c3, so that White can follow up with Nc3-a4
and c4, blockading and attacking the pawn
on c5 (see Nimzowitsch-Johner, quoted
above, for an example of this plan). With
the knight on d2, this plan is not available,
and White cannot really exert any
significant pressure on the pawn on c5.
Meanwhile, the simplifications have also
deprived him of any hopes of a kingside
attack, so he really doesn't have anything in
this position.
QUESTION: It seems strange that the
assessment of an entire structure should
be changed by the location of a single
piece.
ANSWER: Maybe, but that is often the case.

Chess is a very deep game, and what look
like insignificant differences can have a
major effect on a position. ]

[ 11.

g4 is also a logical try, but once again,

the omniscient Odessky shows the way:

xe5 12.xe5 g6! and White has no
advantage, ( although 12...

f6?! 13.f4

xe5 14.fxe5 gives White an edge. )]

11...

c8 12.e2 xe5?! QUESTION: Why

do you criticise this? Surely it is the logical
follow-up to 10...Nd7, isn't it?

[ ANSWER: Seemingly, yes, but in fact Black

has a deep and subtle idea here, which is
stronger. He should play 12...

b8!

(Odessky).
QUESTION: What? Are you sure he hadn't
been at the vodka?
ANSWER: 'Au contraire'! This move is a
very subtle positional idea, about which
Mark Dvoretsky has written a good deal,
using the term 'the superfluous piece'. The
point is that White has two knights, both of
which are pointing at the e5-square. One is
occupying that square, but the other does
not really have a decent outpost. White
would be quite happy to exchange off one of
his knights for its opposite number, and
then occupy e5 with his other knight. In
such cases, the defender frequently does
best to avoid the exchange, as with 12...Nb8,
and leave the two white knights squabbling
over the same outpost.
QUESTION: So how should the game
continue after that?
ANSWER: White has little real choice, since
Black intends ...Be8 and then ...f6, driving
the white knights back in disarray. White
has nothing better than 13.

xc6 xc6

14.c4 , but he has nothing at all after

f6

(Odessky). ]

13.

xe5 e8 14.g4 f5?! QUESTION:

What is wrong with this?

[ ANSWER: Black had to do something

about the threat of 15 Nc6, but the text
concedes the e5-square forever, which is
something Black really should try to avoid.
Odessky suggests that there is nothing
seriously wrong with Black's position after

14...

f6 : for example, 15.f4 g6 16.f3

g7 17.h3 . Now Black has the surprising
tactical solution f6! .
QUESTION: Eh? Isn't e6 hanging with
check?
ANSWER: Indeed it is, and it gets better
after 18.

xe6+ f7!! .

QUESTION: Insane!
ANSWER: It certainly looks it, but the white
pieces turn out to be awkwardly placed.

19.-- ( Now White loses material after
19.

xf7? xf7 20.g4 f5 ,; and 19.f3

c6 20.g4 f5 .; As Odessky points out, his
only move is 19.

xh7! fxe5 20.xg7+

xg7 and "If he is lucky, White's activity
might be enough for a draw" (Odessky).
The computer agrees: 21.

xe5+ f8

22.

xg6 c6 23.h5 d7 24.h8+ e7

25.

h4+ with perpetual. )]

15.

e2 f6 16.c4 e7 17.f4 QUESTION:

So what is going on here?
ANSWER: White is certainly rather better,
thanks to his powerful outpost on e5, but it is
not clear exactly how large his advantage is
against accurate defence. His basic plan is to
advance with g4 and open the g-file.
QUESTION: But doesn't that expose his own
king?
ANSWER: To some extent, but he can tuck
his king away on h2, where it is fairly safe.
Meanwhile, opening the g-file allows the rooks
to work in combination with the bishop on b2,
in a concentric attack on g7.
QUESTION: And what is Black going to be
doing in the meantime?
ANSWER: Ah, well, funny you should ask that!

f7? The answer is...sitting on his hands!

[ Odessky goes so far as to describe this

move as "madness", and asks: "How could
one not play 17...b5 ?". This is clearly what
Black should do, after which he would have
a reasonable chance of counterplay. Instead,
Wolf simply shuffles his pieces around and
allows White a free build-up on the
kingside. ]

18.h3

fd8 19.h2 c7 20.f2 e8

21.

g1 dxc4 22.bxc4 xe5? QUESTION:

This looks insane!
ANSWER: It does, and this is a very
interesting moment. Ray Keene told me that,
when he first saw this game as a youngster,
he was dumbfounded by this exchange,
surrendering the entire dark squares and long
diagonal. He was even more dumbfounded
some years later, when the highly
experienced English master Peter Clarke
played the same exchange against him, in a
very similar position with colours reversed, as
we saw above.

[ Going back to Nimzowitsch-Wolf, it is very

striking to note that the computer does not
hugely condemn the exchange on e5,

thinking that Black is still equal after this
move. Similarly, the strong Hungarian
master Geza Maroczy, annotating the game
at the time, also does not condemn the
move, merely noting that "Into consideration
came 22...b5 ". Fritz even considers Black
to stand better after the latter move, which
seems a bit of an over-valuation of the
bishop-pair, but Black would certainly still
be very much in the game. ]

23.

xe5 cd7 24.g4 fxg4 25.xg4 g6

26.d4 QUESTION: It looks as though White

has obtained his dream position!
ANSWER: Very much so. I am truly amazed
that Fritz should still assess this position as
completely equal. White has huge pressure
against g7, total domination of the dark
squares, and now creates a passed pawn on
the d-file. These plusses should add up to a
decisive advantage, in my opinion.
QUESTION: But what about the opposite-
coloured bishops? Aren't they a drawish
factor?
ANSWER: It is true that pure opposite-
coloured bishop endings tend to have a high
draw factor, but where there are other pieces
on the board, this is usually much less
significant. Indeed, in a middlegame position
such as this, the presence of opposite-
coloured bishops can often strengthen the
attack, since the attacker effectively has an
extra piece, if he can direct play against the
squares controlled by his own bishop.

cxd4 27.exd4

f5 28.h5 g6 29.e2

f8 30.g5 f5 31.fg2 xg5 32.xg5

f8

[ Maroczy's suggestion of 32...

b4

is more active. It is true that White faces
some technical problems in making progress,
chiefly because his king is somewhat
exposed and Black has play on the light
squares. ]

33.

g4 f5 34.g2 h6?! This is an

inaccuracy, which makes the g7-square more
vulnerable.

[ A better choice would be 34...

g6 , ]

[ or even 34...g6 .

QUESTION: But surely the latter puts
another pawn on the same colour squares
as his bishop. That cannot be good, can it?
ANSWER: In general, it is true that one
should seek to avoid putting one's pawns on
the same colour squares as one's bishop,
but opposite-coloured bishop positions are
something of an exception. The 'fortress'

idea is the key element in such positions,
and the defender frequently does better to
place his pawns on the same colour squares
as his bishop, so as to keep them securely
defended. That is the problem with 34...h6
here – now a later ...g6 will always leave the
g-pawn weak. ]

35.

g3 h7 36.d5 This has been White's

plan ever since move 26. He creates a passed
pawn in the centre, with which he hopes to
distract his opponent. The principle of "two
weaknesses" once again applies – White
cannot win by simply attacking g7, but the
combined threats from the d-pawn and the
pressure on g7 should eventually overwhelm
the defence.

f7?

[ Wolf's abiding sin in this game has been

passivity, and this latest example proves the
decisive mistake. He absolutely had to
exchange on d5 with 36...exd5 37.cxd5
and then try to create some counterplay with
a move such as b5 . The extra open lines
on the queenside will always offer some
counter-chances. ]

37.d6 g6

[ QUESTION: Isn't 37...b6 more sensible, to

prevent White's next?
ANSWER: Yes, but it does not save the
game. I suspect Wolf avoided it because of
the combination 38.

xg7+ ( White actually

does better to continue 38.a4 : for example,

d7 39.a5 f7 40.axb6 axb6 , and only
now 41.

xg7+ ; the point being that after

xg7 42.xg7+ xg7 43.xg7 e5

44.

xe5 g6 he picks up a third pawn by

45.

d4 ) 38...xg7 39.xg7+ xg7

40.

xg7 , but then e5 ( if 40...xg7?

41.d7 ) 41.

xe5 still gives Black some

drawing chances in the pure opposite-
coloured bishop ending; White's two passed
pawns are rather too close together. ]

38.c5

c8 39.c3 c6 QUESTION: Is this

forced?
ANSWER: Unfortunately, yes, although the
resulting ending is lost.

[ Black cannot allow the breakthrough c5-c6:

for instance, 39...

d7 40.c6 bxc6 41.xc6

d8 42.c7 and Qc6 follows. ]

40.

xc6 bxc6 41.h4 g8

[ QUESTION: Why doesn't Black take the

open file with 41...

b7 and why didn't White

prevent this with 41 Rb3 himself?
ANSWER: It doesn't make any difference.
White replies 42.

b3 anyhow, and Black

cannot exchange rooks, because the d-
pawn promotes. ]

42.

b3 Now it is over. The white rook

penetrates, and Black has not a shred of
counterplay.

d7 43.b8+ f7 44.c8 e4 45.g3
The white king plans an incursion on the dark
squares.

h5 46.

f2 d5 47.e3 g2 48.d4 f3

49.

h8 e5+ Desperation.

[ In the event of further passive play, White

has various ways to win, but Keene gives
the attractive variation 49...

d5 50.e5

f3 51.g7! d5 ( 51...xg7 52.c7 )

52.

h6 and the mate threat on f8 decides. ]

50.

xe5 d5 51.f6 f3 52.g5 g7

53.f5!

g4

[ The point of White's last was to block the

h3-d7 diagonal, so that 53...gxf5 54.

c7

wins, as Black no longer has ...Bg4. ]

54.f6+

h7 55.xc6

1-0

B21

Nimzowitsch,A
Olson,A

Copenhagen

1924

[Steve Giddins]

1.f4

[ The easiest way to understand

Nimzowitsch's line of thinking at move 4
(see below) is to consider a typical Closed
Sicilian structure, such as the following: 1.e4

c5 2.

c3 c6 3.g3 g6 4.g2 g7 5.d3

d6 6.f4 e6 7.

f3 ge7 8.0-0 0-0

. One of Black's main ideas in such positions
is to put his knight on d4, and invite White
to take it. Let us assume (without pretending
that the next few moves are necessarily
terribly good) that this happens at once: 9.h3

d4 10.xd4 cxd4 11.e2 d7 .
Looking at this structure, we can see that, as
a result of the exchange, White has a
marked weakness on c2. Black has the
obvious plan of putting his rooks on the
open c-file, pressing against the backward
c2-pawn. For this reason, White usually
cannot afford to exchange off the d4-knight
in such positions, but must either live with it,
or go to elaborate lengths to remove it, by
moving his knight from c3 and then playing
c2-c3.
But now imagine the last position, but with

the white pawn on c4. That would be an
altogether different kettle of fish, as there
would be no weakness on c2 for Black to
aim at.
QUESTION: So that is the idea of
Nimzowitsch's set-up? Just to avoid a
possible weakness on c2 later on?
ANSWER: Effectively, yes. It is simple
prophylaxis against the enemy plan of ...Nd4,
which will now be met by just taking on d4,
without any worry over the recapture ...
cxd4 creating a target on c2.
QUESTION: OK, I see that, but what about
the hole on d4?
ANSWER: Nimzowitsch is hoping that this
will not matter too much. As we have
remarked elsewhere in this book, one can
sometimes afford to accept one weak
square, if there are sufficient chances to
cause trouble for the opponent on the other
63. ]

[ Nimzowitsch's strategic idea, although not

necessarily all that strong in this specific
position, is a deep and profound one, and is
the forerunner of the so-called Botvinnik
System in the English Opening: 1.c4 e5

2.

c3 c6 3.g3 g6 4.g2 g7 5.e4 . ]

1...c5 2.e4 Transposing into a position, which

became quite popular in the 1980s, thanks to
the efforts of various English players,
principally Mark Hebden.

c6

[ Eventually, it was established that the

gambit 2...d5 3.exd5

f6! is quite

promising for Black, as a result of which
the popularity of 2 f4 has declined
markedly. ]

3.d3!? A very modest approach by White, but

one which conceals a specific and quite
revolutionary idea for the time.

[ The usual move is 3.

f3 . ]

3...g6 4.c4!? QUESTION: This looks a

strange way to play. White moves only pawns,
and creates a hole on d4. What is he playing
at?
ANSWER: Nimzowitsch himself was very
proud of this idea, and awards his last move
two exclamation marks. Objectively, of course,
that is a wildly hyperbolic piece of punctuation,
but in some ways, the move 4 c4 does
deserve to be hailed as something quite
extraordinary. As usual with Nimzowitsch,
prophylactic thinking lies at its heart.
QUESTION: You mean he wants to stop Black
playing ...d5?

ANSWER: No, no, nothing so crude. As
Nimzowitsch himself points out, there is no
hope of doing that anyway, as Black can just
prepare the advance with ...e6. His idea is
much deeper than that. Basically, he wants to
play a kind of Closed Sicilian set-up, but with
a prophylactic eye on Black's counterplay.

g7 5.c3 b6?! Not an especially
impressive response.

[ Both 5...e6 followed by ...Nge7, ]
[ and 5...d6 followed by ...Nf6, are more

natural-looking alternatives, which would
give Black a perfectly satisfactory position. ]

6.

f3 b7 7.g4!? QUESTION: Goodness

me! That looks pretty extravagant.

[ ANSWER: Yes, but it makes some sense.

The natural English Opening/Closed Sicilian
approach would be 7.g3 , but later in the
middlegame, White often ends up advancing
g3-g4 anyway, as part of a kingside pawn-
storm. The text seeks to economize a tempo,
by playing the pawn to g4 at once.
Nimzowitsch was no doubt encouraged in
this by the fact that the enemy queen's
bishop has gone to b7, with the result that
a subsequent ...d6 will not attack the g4-
pawn. ]

7...e6 8.

g2 ge7?! 9.b5?! Another very

surprising move, fully in accordance with
Nimzowitsch's predilection for such 'bizarre'
moves.
QUESTION: He threatens 10 Nd6+?
ANSWER: Yes, but that is easily dealt with by
the reply in the game. Nimzowitsch explains
that the real purpose of Nb5 is to weaken b6.
QUESTION: What?
ANSWER: That is right! The idea is that Black
will sooner or later be unable to resist kicking
the knight away by ...a6, which will weaken the
b6-pawn.
QUESTION: But that can hardly matter,
surely?
ANSWER: I am inclined to agree that it looks
a bit fanciful, but Nimzowitsch was so pleased
with himself that he again gave the move two
exclamation marks!

[ However, I would suggest that the

straightforward 9.0-0 would be the choice of
most players. ]

9...d6 10.0-0 a6 Allowing himself to be

provoked. No doubt Nimzowitsch felt a glow of
self-justification at seeing this move, but I am
not convinced that there is any objective
reason for such a feeling. Black is doing fine.

11.

a3 Continuing the idea begun at move 9.

The knight heads towards c2, from where it
covers the d4-square, and also helps support
a possible break with b4 later on, trying to
expose the 'weakness' at b6.

0-0 QUESTION: Can't Black play more

energetically than this?

[ ANSWER: Yes, indeed, and possibly he

should. The computer likes the
counterattack 11...h5!? 12.g5 and now

c7

, intending long castling. After 13.

b1 0-0-0

14.

d2 d5! Black looks to be doing fine,

which just underlines the rather eccentric
nature of White's 9th move. ]

12.

e2 d7 13.e3 b4!? Nimzowitsch

writes: "Otherwise there follows Rad1 and d4,
with advantage to White", but this seems
rather an optimistic assessment.

[ I don't see any problems at all for Black

after either counterblow in the centre, with

13...f5 ]
[ or 13...d5 . ]

14.

c2! QUESTION: Doesn't this lose a

pawn?
ANSWER: Yes, it is a positional sacrifice,
which, Nimzowitsch says, "...is only possible
because b6 is undefended". Tactically, this is
true, but it seems a rather slim justification for
the rather fanciful manoeuvre Nb5-a3-c2.

xb2 15.ab1 c3 16.xb4 xb4

[ 16...cxb4 17.

xb6 is the point of

Nimzowitsch's play. ]

QUESTION: What does White have for his
pawn?
ANSWER: Nimzowitsch's idea is that, in order
to hang on to the extra pawn without losing b6
in return, Black has been forced to misplace
his dark-squared bishop, which in turn leaves
his kingside dark squares weakened.

17.

c1?! QUESTION: This looks rather slow!

ANSWER: It does, although it was all part of
Nimzowitsch's plan, and earns another
exclamation mark from him.

[ However, the direct 17.f5 is certainly a

more natural way to follow up the pawn
sacrifice, and leaves White with reasonable
compensation after exf5 18.gxf5 . ]

17...f6 18.

b2 e5? QUESTION: What is

wrong with this? It looks logical to shut out the
enemy bishop.
ANSWER: It is in principle, but tactically, the
blockade on e5 can be undermined.

[ I think Black should just get on with

something on the queenside, such as

18...b5 , which has the merit of removing the

weakness on b6. I assume Nimzowitsch

would have played something such as 19.g5

fxg5 20.

xg5 , trying to attack the dark

squares, but it is not particularly convincing
after, for example, h6 21.

h3 d5 . ]

19.g5! Hammering away at the dark squares.

c6?!

[ Nimzowitsch gives 19...fxg5 20.

xg5

(threatening Bh3)

c6 21.f5 . This certainly

gives White compensation, but is no worse
for Black than the game. ]

20.gxf6 Now the barricades on the long dark-

square diagonal a1-h8 start to crumble, and
the fact that the black bishop on b4 is out of
play really begins to be felt.

g4

[ 20...

xf6 21.fxe5 dxe5 22.xe5 xf1+

23.

xf1 xe5 24.f6 is winning for White

– a striking illustration of how quickly
Black's position can collapse, once the
long diagonal is opened. ]

21.fxe5 dxe5 22.

e3! An excellent move,

which unpins the knight and aims the white
queen at the weakened kingside dark squares.
White has conjured up a winning attack,
almost from nowhere.

h5

[ 22...

xf6 23.xe5 xf1+ 24.xf1 xe5

25.

xe5 e6 26.a1 is no better. Then,

after

e7 White has a choice between

27.

f2 ( and the more striking 27.c1!

when Black is defenceless against the
queen and bishop battery on the long
diagonal. )]

23.

g5 c8 24.f7+ g7 EXERCISE: Can

you spot a neat way to crown the attack?

25.

f4!! The prelude to a delightful finish.

h6 26.e6+! exf4 27.g7# A lovely final
mate.
QUESTION: Hmm...This was a strange game.
It looked as though Nimzowitsch's fancy ideas
were really not at all convincing, yet suddenly,
it all fell into place beautifully.
ANSWER: Yes, I find this quite a difficult
game to understand. In fact, the whole plan of
weakening b6 and then using that weakness
to lure the enemy bishop to b4, so as to mate
him on the kingside dark squares, looks so
far-fetched, that it makes one suspect it is the
sort of thing that was dreamt up after the
game, in a bid to pretend that what happened
rather randomly in the game was really all pre-
planned.
QUESTION: Is that what you think?
ANSWER: I am not sure. Some players,
notably Botvinnik, have been accused of this,

with some justification, I believe. In the
present case, though, the move 9 Nb5 is
actually rather hard to explain in any other way,
if one discounts Nimzowitsch's own
explanation, so I am inclined to give him the
benefit of the doubt. Whatever the truth about
the 9 Nb5 adventure, the game is a very
notable one, if only for the plan of e4-d3-c4,
which has proved to have an enduring legacy
in the English Opening to this very day.
1-0

A07

Nimzowitsch,A
Rubinstein,A

Berlin

1928

[Steve Giddins]

1.

f3 d5 2.b3 f5 Unlike Wolf, Rubinstein

seizes the opportunity to develop his bishop
outside the pawn chain, rather than shutting it
in with ...e6. However, as we will see, the
harassing of this bishop forms a major part of
White's subsequent plan.

3.

b2 e6 4.g3 h6 QUESTION: Isn't this a

waste of tempo in the opening?
ANSWER: In many positions, such a move of
the rook's pawn would be a tempo loss, but
here it is quite standard. Black wishes to
preserve his bishop from being exchanged by
a subsequent Nh4. The reason he is able to
get away with it is that White's plan, with the
double fianchetto and holding back of his
centre pawns, is quite slow and the position is
rather closed, as a result of which time is less
critical. In an open position, of the type
starting 1 e4 e5, with gambits and an early
opening of central lines, a tempo loss such as
...h6 would be much more dangerous, but
here it is perfectly feasible.

5.

g2 d7 6.0-0 gf6 7.d3 e7 8.e3

QUESTION: That looks like a slightly funny
move?
ANSWER: It is a tiny bit unusual, perhaps,
but perfectly sensible, nonetheless. White has
two main ideas in such positions: either to
advance c2-c4 or e2-e4. As we will see,
Nimzowitsch intends the latter, leaving his
pawn on c2, so that the c2-d3 pawn chain will
blunt the effect of the enemy bishop on f5.
QUESTION: OK, but if he intends e4, isn't his
last move just a tempo loss?

[ ANSWER: Not really. At the moment, e4 is

not possible without some preparation. The

most direct method is 8.

bd2 0-0 9.e1

followed by e4, but as we will see,
Nimzowitsch intends a slightly different,
slower method. He is not in a rush to commit
his knight to d2, and besides, there are the
same number of tempi involved in playing
Qe1, e4, and Qe2 as there are in e3, Qe2,
and e4, so the difference is to a large
extent just a matter of taste. ]

8...0-0 9.

e2 c6 QUESTION: This looks a bit

passive. Can't Black play more actively?

[ ANSWER: He could play 9...c5 , with the

idea of queenside play by a later ...b5 and
eventually ...c4. The text is more common in
such positions, as the central pawn triangle
c6-d5-e6 blunts the effect of White's bishop
on g2. Another small point is that, if Black
is going to play ...c5, then he would
probably prefer to have his queen's knight
on c6, where it is more active; somehow, c6
fits in more naturally with ...Nbd7, although
this is not a major factor. ]

10.

h1 QUESTION: That looks a funny

move!
ANSWER: It is a typical piece of
Nimzowitschian prophylaxis. He intends to
advance e4 and later f4, starting a kingside
pawn advance. In that case, his king may
prove vulnerable on the g1-a7 diagonal, so
tucking it into the corner is a useful precaution.
The other point is that he still refrains from
committing his queen's knight, which retains
the option of coming to c3, if circumstances
permit (although, I must admit, it is hard to
envisage a specific line where this is likely to
happen). If the full truth be told, I suspect
Nimzowitsch just liked playing such
'mysterious' moves as 10 Kh1, keeping his
opponent guessing as to his intentions.

a5 11.a4

[ QUESTION: Interesting. I am never sure

whether to play the text, or 11.a3
in such positions. Is there a clear rule of
thumb?
ANSWER: Not really. In many cases, it
amounts to a question of taste. Playing a4
has the drawback of ceding the b4-square to
a black minor piece, and may also allow
lines to be levered open with ...b5. On the
other hand, a3 allows the enemy a-pawn to
come to a4, with b4 answered by ...c5,
again trying to open lines. In this specific
position, I think Nimzowitsch's mind was
made up by the position of the black knight
on d7. If that knight were still on b8, within

easy reach of the b4-square via ...Na6-b4,
then I suspect he would have put his pawn
on a3, but with the black knight not able to
reach b4 so easily, it makes more sense to
play a4. ]

11...

c5 Black attacks the e4-square again,

still holding back White's intended e4
advance.

12.

d4 h7 13.f4 Thus, White gets his f-

pawn moving first, also incidentally securing
his knight's outpost on d4, by preventing
Black advancing ...e5. Now White intends to
follow up with Nd2 and e4.
QUESTION: The bishop on h7 does not look
very active!
ANSWER: Indeed, and that is a leitmotiv of
such positions. The solid white pawn structure
c2-d3-e4 keeps this bishop out of the game,
and White frequently plays to exploit the
inactivity of this piece. Admittedly, it does
defend the black kingside, but if the play
develops on the other side of the board, the
bishop can frequently be little more than a
spectator. For a classic example of this, I can
recommend the game Capablanca-Lilienthal,
Moscow 1936.

fd7 14.d2 c7

[ QUESTION: That is a bit of a surprise. I

thought Black would break with 14...e5
, as in similar Dutch structures.
ANSWER: That is certainly desirable
positionally, but here it fails tactically, as the
knight on e5 will be loose. Nimzowitsch
gives the variation 15.fxe5

xe5 16.f5

(threatening 17 d4, winning a piece)

xf5

17.

xf5 g6 18.af1 and White has a

clear advantage, with his bishop-pair and
open f-file. ]

15.e4 QUESTION: White seems to have

achieved his goal. Is he just better now?
ANSWER: It is actually not entirely clear.
Nimzowitsch certainly thought so, and
annotated the game accordingly, but the
computer is less convinced. I think White is a
little better, mainly because of the relatively
passive bishop on h7.

dxe4 16.

xe4 xe4 QUESTION: If you are

so worried about the passive bishop on h7,
shouldn't Black unload it, while he has the
chance?

[ ANSWER: He could, but the cure would be

worse than the disease. After 16...

xe4

17.dxe4 White has more space and the

bishop-pair. His position would be very
pleasant. ]

17.dxe4 e5 QUESTION: What is going on? It

looks as though Black's play is very logical.
Isn't White seeing his pawn structure
damaged?
ANSWER: This is a critical moment of the
game. White is, indeed, sustaining some
damage to his pawn structure, but he hopes
that his open lines will give him at least
sufficient dynamic compensation.

18.

f3! QUESTION: Why is this good?

[ ANSWER: The punctuation is

Nimzowitsch's own, and reflects a very
typical piece of thinking on his part. As so
often in his games, the key issue is one of
blockade. The strategical battle rages
around the e5-point – Black wants to
establish that square as a key central
outpost, whilst White tries to prevent this.
The text places the knight on a square from
where it attacks e5. By contrast, 18.

f5?!

xf5 19.exf5 f6 "with a solid strongpoint
in the centre at e5" (Nimzowitsch) is less
effective. ]

QUESTION: OK, I see the strategical point,
but doesn't 18 Nf3 leave the f4-pawn hanging?
ANSWER: It does, yes, but Nimzowitsch has
seen a tactical justification – a typical
example of tactics coming to the service of
strategy.

18...exf4 19.gxf4

fe8

[ Now we see Nimzowitsch's point:
19...

xf4?? drops a piece to 20.e5

, a simple enough tactical point, but the sort
of little tactic which is so often the key to
good strategic play. One can have all the
depth of strategic understanding in the world,
and fully appreciate that your knight wants to
go to f3 at move 18, so as to fight for the
e5-square, but if you don't see the tactical
refutation of 19...Qxf4, you cannot play the
move. It is no accident that the world's
greatest strategic players, such as
Nimzowitsch, Capablanca, Petrosian and
Karpov, all had fantastic tactical ability,
which was an essential part of making
their strategic conceptions work in practice. ]

20.e5 QUESTION: So, what is going on here?

White has hanging pawns on e5 and f4. I
thought one usually wanted to keep these
pawns adjacent, and not to have to advance
one of them?
ANSWER: Quite right, and the assessment of
this position depends crucially now on the f5-
square. If Black can maintain his hold on that
square, and keep his blockade of the hanging

pawns, he will be fine, and can even obtain
the advantage, but if White manages to break
the blockade and push his pawn to f5 in
favourable circumstances, the potential
energy of his position will be unleashed
decisively. Once again, it all comes down to a
question of 'blockade'.

c5 21.d4 QUESTION: Ah, so now I
understand that move – White takes control of
f5 and threatens f5.
ANSWER: Quite, so Black immediately reacts
by challenging the enemy knight.

e6 22.ad1

[ QUESTION: Why not 22.f5 ?

ANSWER: The problem is

xd4 23.xd4

f6 with an awkward pin. Remember, I said
White would be doing well if he could
achieve f5 "in favourable circumstances". ]

[ QUESTION: OK, I see that, but isn't
22.

xe6 possible? Surely after fxe6

, the e6-pawn is weak?
ANSWER: It is somewhat weak, but is fairly
easy to defend, and, more importantly, the
black blockade of f5 is strengthened, which
was not in Nimzowitsch's plans at all. ]

22...

xd4 23.xd4 f5 24.e4 QUESTION:

At least I can understand the last few moves.
The battle is all about that f5-square.
ANSWER: Indeed, and, hitherto, both sides
have played very accurately, but now
Rubinstein goes astray.

xe4+?!

[ Nimzowitsch himself does not criticize this

move, but 24...

c8 was better. Nimzowitsch

claimed this was met simply by

A) instead, 25.

g1 is probably stronger,

when a subsequent Qd3 will anyway force
the exchange on e4; Black's best,
according to the computer, is

e6!

26.

xf5 ( the subtle point of Black's last is

26.

d3 xe4+ 27.xe4 f5! , exploiting

the pin to establish the blockade on f5
again ) 26...

xf5 27.d3 xd3 28.xd3

ad8 and White still only has a small
edge;

B) 25.

d3 , apparently forcing the

exchange on e4 anyway , but the
computer points out the typically awkward
tactical line

g4 26.d2 d8! when the

pin on the d-file leaves White no time to
push his f-pawn. ]

25.

xe4 ad8 26.e6 White's strategy has

triumphed, and he has broken through on the
central light squares.

f8 27.e5? Having achieved his strategic

aims, Nimzowitsch almost spoils everything
with a tactical oversight.

[ As he himself pointed out in his notes, and

the computer confirms, he should have
preferred 27.f5 with a substantial
advantage: for instance,

e7 28.de1 . ]

27...

c8? The losing move.

[ Black should have played 27...

xd1

28.

xd1 ( 28.xc7? xf1+ 29.g2 xe6

is even worse, and leaves Black better )

28...

e7 , when White cannot defend the e6-

pawn, and has nothing better than 29.exf7+

xf7 with equal chances, ]

[ or the immediate 27...

e7 . ]

28.f5 fxe6

[ QUESTION: Can Black hang on with 28...f6

?
ANSWER: Unfortunately not. There follows

29.

xf6 gxf6 30.xd8 xd8 31.g4+ g7

32.

d1 e7 33.d7 f8 34.f7

, winning. ]

29.f6! This is decisive.

xd1

[ After 29...gxf6 , 30.

g1+ with mate in six,

is the most precise:

g7 31.xg7+ xg7

32.

g1+ , etc. ]

30.f7+

h8 31.xd1 d8 EXERCISE: Once

again we find Nimzowitsch finishing crisply.
How?

32.

g6!

[ A nice finish. Even being able to take a rook

with check does not save Black: 32.

g6

xd1+ 33.g2 d2+ 34.h3 and mate on
h6. ]

1-0

A11

Nimzowitsch,A
Ahues,C

Frankfurt

1930

[Steve Giddins]

1.

f3 d5 2.b3 e6 3.b2 f6 4.e3 bd7

5.c4 c6 Reaching a kind of Semi-Slav set-up,

but where White has not committed his d-
pawn to d4.

6.

c3 d6 7.c2 e7 8.d4!?

QUESTION: That looks a bit of a funny move.
ANSWER: It is typical of Nimzowitsch's
original, some might say convoluted, style.
Rather than the routine advance of the d-pawn,
he seeks fresh pastures, where he has more
chance of outplaying his weaker opponent.
QUESTION: So what is the concrete idea?

ANSWER: Once again, prophylaxis is at the
heart of the move. Black is preparing a
central pawn advance with ...e6-e5, whereas
now this would allow the unpleasant knight
jump into f5. On the other hand, chasing the
knight with 8...c5 would be met by 9 Nb5.
QUESTION: But the knight is not going to be
able to stay on d4 forever, is it?
ANSWER: Agreed, and it is far from clear that
White's last move is of any great objective
merit, but Nimzowitsch is hoping to extract
some sort of concession from Black, before
retreating his knight.

[ If White does not wish to transpose into

more normal Semi-Slav positions with 8.d4
, ]

[ another, rather more orthodox option, was
8.

e2 . White prepares an aggressive plan

with long castling, followed by a kingside
pawn storm with g4. This has been played
by a number of strong GMs in recent years,
including Mikhail Gurevich and ex-women's
world champion, Antoaneta Stefanova.
Readers interested in finding out more about
this can find an excellent chapter on the line
by Tony Kosten, in the Everyman book
'Dangerous Weapons: Flank Openings'. ]

8...a6 Clearly preparing a subsequent ...c5,

without allowing the knight into b5. This is a
perfectly reasonable response, the move ...a6
being one that frequently fits into Black's
opening scheme anyway.
QUESTION: What other options are there?

[ ANSWER: Black has plenty of reasonable

moves. Interestingly, the computer way is to
play 8...g6?! just covering the f5-square and
preparing ...e5. Objectively, this may be OK,
but from a human perspective, it is not
very attractive to open the long diagonal in
this way, given that Black will probably wish
to castle short, and the white queen's bishop
is sitting on b2. I am sure that this is the
sort of concession Nimzowitsch was hoping
to extract, when he played his 8th move. ]

[ QUESTION: What about the immediate
8...c5 advance? Does Black need to be

afraid of the knight jump to b5?
ANSWER: That is a very good question.

9.-- ( After 9.

db5?! b8 , the knight on b5

is doing nothing.; Nimzowitsch himself
acknowledged this in his notes, and
indicated that he would probably have
played 9.

f3 .

QUESTION: But then Black has gained a
tempo, by avoiding ...a6, hasn't he?

ANSWER: Indeed he has. This just
underlines the objective harmlessness of
White's 8th move, in my opinion. What
White has achieved, though, is to put paid
to Black's plan of ...e6-e5, which would now
lose the d-pawn, and this was the main aim
of Nimzowitsch's plan.
QUESTION: Can't Black play dxc4 10.bxc4
and then e5 ?
ANSWER: In that case, I think Nimzowitsch
would have continued with 11.e4!
, establishing a grip on d5. Note that his
queen's knight is pointing at that square,
whereas Black's queen's knight is on d7,
and so does not point at the equivalent
outpost on d4, in the white position. )]

9.

e2 0-0 10.0-0 c5 11.f3 QUESTION:

Hasn't White simply lost two tempi with his
knight?
ANSWER: Well, Nimzowitsch would argue,
I'm sure, that he has weakened the enemy
pawn centre a little. The point is that now the
d5-pawn can potentially become a target,
after cxd5 and d4, aiming to saddle Black with
either an IQP on d5, or, after ...b6 by Black,
with hanging pawns on d5 and c5.
QUESTION: And these pawns would be
weak?
ANSWER: Not necessarily! I would say that
Black's pieces are quite well-placed to
support his pawns, especially in the hanging
pawn structure. However, in general, playing
with or against such pawns tends to be a
matter of taste, and Nimzowitsch usually
preferred to play against them, hoping to
emphasise their weakness. Objectively, I
think Black would be fine in such a structure,
but I imagine Nimzowitsch would have been
happy with his position nonetheless.

b6?! Thus far, Ahues has reacted perfectly
soundly to Nimzowitsch's early provocation,
but this is a funny-looking move. The knight is
just misplaced on b6.

[ It looks much more natural to play 11...b6

, followed by ...Bb7, with equal chances. ]

12.e4! This central blow is strong, but had to

be very well calculated. White is ahead in
development, and the decentralized knight on
b6 is misplaced, so the strike in the centre is
very logical.

[ QUESTION: But can't Black keep things

closed after 12.e4 with d4 ?
ANSWER: This is the first point of White's
tactical calculations. He intends to meet that
move with 13.e5! dxc3 14.

xc3 xe5

15.

xe5 and White is better, with his two

bishops and dark-square control. ]

12...

xe4 13.xe4 dxe4 14.xe4 e5

QUESTION: Err, what is going on here? Can't
White take that pawn? Ah, no; the pin on the
e-file is a problem, isn't it?
ANSWER: Funny you should ask... This is the
critical moment of the game. Black realizes
that, if he can free his position with this move,
followed by ...f5, then he will be fine; if not,
he just stands worse. Hence, 14...e5 is the
move he wants to play, if he possibly can.
White, in his turn, had to consider the
consequences of this move very carefully, in
choosing his 12th move. Just as Black had to
play 14...e5 in order to avoid standing worse,
so White has to take on e5, if he is to avoid
equality, at best. So, it is all a question of
which player has seen further.

15.

xe5!

[ 15.

xe5? avoids the pin against the queen

on e4, but loses material prosaically to the
wider pin on the e-file:

xe5 16.xe5 xe5

17.

xe5 e8 18.f4 f6 , etc. ]

15...

e8

[ QUESTION: Is the alternative 15...f6

any better?
ANSWER: Against that move, Nimzowitsch
had seen 16.

d3 , defending his queen

(and thus threatening Nf3), as well as
threatening h7. The latter cannot be ignored;
indeed, fxe5 17.

xh7+ f7 18.f4!

yields a decisive attack: for instance, e4

19.

ae1 and the black king is caught in a

deadly crossfire.
I would also point out an important feature
of this position: although White is a piece
down in terms of the overall count, all of
his pieces, bar his king, are taking part in
the attack, whereas Black's queen's rook,
queen's bishop and queen's knight are all
out of the game. Thus, in terms of pieces
where it matters, at the centre of the action,
White actually has a material superiority.
This is crucial to successful attacking play,
and we will see this again later in the
game. ]

[ There is one other move worth considering,

namely 15...f5 . According to Kurt Richter,
as quoted by Reinhardt, this was in fact
Ahues' original intention. He thought that
after 16.

e3 f4 , he was winning the knight

on e5, since White can no longer support
the piece by f4, but now he realized that
White has the saving move 17.

c3

, with the final point

e8 18.f3 xe2?

19.

xg7# . ]

[ QUESTION: But what is the difference from

the variation after 15 Bxe5?
ANSWER: Now, after 15...

xe5 16.xe5

xe5 17.xe5 e8 White has the saving
move 18.

c7 .

QUESTION: Ha! So, ironically enough, if
Ahues had played 8...c5 at once, saving the
tempo on 8...a6, this position would have
been winning for him, since his b6-knight
would still be defended by the a7-pawn.
ANSWER: Indeed! White's plan in the game,
beginning with 12 e4, would have been
tactically unsound. This is a striking example
of the difference that such seemingly
insignificant moves as ...a6 can make.
QUESTION: What does it say about the
interaction of strategy and tactics?
ANSWER: A good point. What it shows is
how deep strategy must always be founded
on accurate tactical calculation. You can
have all the profound depth of strategical
understanding you like, but if you fail to spot
a little tactical nuance like 18 Bc7,
attacking the loose knight on b6, then
everything falls to pieces. ]

16.f4 f6 17.

h5! QUESTION: How is this

different from the note to 15...f6 above?

[ ANSWER: The subtle difference is that the

inclusion of the moves ...Re8 and f4 now
means that 17.

d3 no longer unpins the

knight: Black can just reply g6
and the knight on e5 is still pinned. Hence,
White must find another resource. ]

17...

f8

[ QUESTION: And I suppose 17...g6

loses to a sacrifice on g6?
ANSWER: Indeed. After 18.

xg6 hxg6

19.

xg6+ f8 ( not 19...g7? 20.xe8+ )

, Nimzowitsch gives 20.

g4

A) 20...

g7 is a tougher defence, but

even so, after 21.

xg7+ xg7 22.xf6

f7 ( 22...d8 23.d5+ picks up the
unfortunate loose knight on b6 once again;
clearly, Nimzowitsch understood the force
of Nunn's law LPDO: loose pieces drop
off ) 23.

xe8 xe8 and White has a rook

and no fewer than four (!) extra pawns,
against the black bishop and knight –
clearly a winning advantage;

B) 20...

xg4 21.xf6 , winning: d7

22.

h6+ g8 23.h8+ f7 24.g7+

e6 25.ae1+ and it is carnage. ]

[ Likewise, 17...fxe5 loses to 18.

xe8 xe8

19.fxe5

f8 , and although Black has two

pieces for the rook, he is still undeveloped
on the queenside, and White has two extra
pawns. A simple plan such as 20.

f4

, preparing to double rooks, leaves Black
with no real defence.
QUESTION: Are there any other moves? ]

[ ANSWER: The only other line worth

mentioning, as given by Nimzowitsch himself,
is 17...

d8 which loses nicely to 18.f7+

h8 19.g6+ hxg6 20.xg6 g4
(the only sensible move to stop Qh5 mate)

21.

ae1 and there is nothing to be done

against Re3-h3. ]

18.d3?! Nimzowitsch awarded himself an

exclamation mark for the text, but in fact, it is
not the best, as the next note shows.

[ The computer finds the strongest

continuation in 18.

d3! . Now fxe5

( 18...

c7 is probably the best defensive try,

but after 19.

ae1 White has an extra pawn,

the knight on e5 is still immune, and he is
basically winning ) loses much as in the
game: 19.fxe5

c7 ( 19...xe5?

loses at once to 20.

xe5 xe5 21.xf8+

xf8 22.d8+ ) 20.xf8+ xf8 21.f1

d8 22.f7+ h8 23.xd8+ xd8

24.

d5 e7 25.f7 , etc. ]

[ EXERCISE: After 18.d3 can you spot how

Black might have defended? ]

18...fxe5?

[ ANSWER: The only defence was 18...g6

, after which the line 19.

xg6 hxg6

20.

xg6 xe4 21.dxe4 leads to a position

in which Nimzowitsch claimed that "Black
stands very unfavourably", whereas the
computer thinks White has only a very small
plus. Keene probably has the assessment
correct: "[White] has a powerful central/K-
side pawn roller and superior mobilisation
for a piece. However, after

f7 22.ad1

c7 White has no forced win available and I
believe that Black's position is capable of
immense resistance". ]

19.fxe5

c7 20.xf8+ xf8 21.f1 d8

22.

f7+ h8 23.e6 A picturesque position.

QUESTION: Black still has his extra piece!
ANSWER: Yes, but a mere glance at the
position is enough to see that White has a
crushing advantage. The two extra pawns
White has are not really the point – just as we
pointed out in another variation above, what
matters is who has more pieces active on the
part of the board that matters. Black's
queenside rook, knight and light-squared

bishop are totally out of play, and even his
king's bishop only controls a couple of useful
squares at d8 and d6. Effectively, Black only
has his queen playing, whilst White has queen,
rook, two bishops, and the e-pawn (which
certainly must be counted as an attacking unit)
in the assault.

xe6

[ After 23...

e7 the computer confirms

Nimzowitsch's refutation 24.

xg7+ xg7

25.

g4+ and it is mate in five: h6 26.f5

, and only spite checks and desperado
sacrifices (e.g. ...Qh4) remain to delay the
inevitable. ]

24.

xe6 g5 25.f7 An excellent example

of accurate and deep calculation by
Nimzowitsch, who when playing his 12th move,
had to see most of the rest of the game. It
also illustrates the importance, when
conducting an attack, of looking not just at the
overall material situation, but at who can bring
more pieces to bear in the crucial area of the
board.
1-0

A18

Nimzowitsch,A
Flohr,S

Bled

1931

[Steve Giddins]

1.c4

f6 2.c3 e6 3.e4 The so-called

Mikenas Attack, one of White's more
aggressive approaches to the English
Opening. Rather than engaging in slow flank
manoeuvres, such as g3 and Bg2, seeking to
influence the centre from the flanks, White
instead occupies the centre with his pawns, in
classical fashion.

d6?! QUESTION: Is this really necessary

against the threat of the e4-e5 advance?
ANSWER: No, certainly not. It is a really
rather feeble and passive move, and not what
one would expect from a master of Flohr's
strength.

[ Black has two preferable responses, both

involving a central pawn thrust: either 3...c5
, allowing e5, but preventing the white d-
pawn advancing to d4 in support, ]

[ or 3...d5 , immediately challenging White's

existing central pawn bastions.
QUESTION: Which is better?
ANSWER: There is no clear consensus.
Both offer Black decent chances, although

the variations can be quite sharp in both
lines. For fuller coverage of both, I will take
the liberty of recommending my own book
'The English Opening: Move by Move' –
"available from fine bookstores everywhere",
as the late Christopher Hitchens was wont to
say. ]

4.d4 Now White just enjoys a strong pawn

centre, of the sort he dreams of in the
opening.

e5 Reaching an Old Indian structure, where

Black has simply lost a tempo, by taking two
moves instead of one, to get his pawn to e5.
The Old Indian is a solid opening, but rather
passive, and obviously, being a tempo down
only serves to reduce its attractiveness further.

5.

ge2

[ QUESTION: Can't White cause Black

problems here by 5.dxe5 dxe5 6.

xd8+

xd8 when Black loses castling rights?
ANSWER: No, this is definitely not to be
recommended. The resulting structure is
known to be nothing for White. Indeed, he
would have to be careful not to end up
somewhat worse, as he has a significant
'hole' on d4, to which Black will try to direct
his knights. Black, by contrast, can cover his
d5-square with the move ...c6. Black will
continue with ...c6, ...Kc7, and then aim for
moves such as ...Bc5, ...Nbd7-f8-e6, etc.
With the queens off the board, the loss of
castling rights is of minor significance only. ]

5...

e7 6.f3 QUESTION: Is this the Sämisch

Variation?
ANSWER: Against the King's Indian Defence,
with the black bishop on g7, it is named after
Sämisch. The idea here can be similar –
White intends to develop with Be3, Qd2, and
probably long castling, followed by a kingside
pawn advance with g4 and h4.
QUESTION: This looks very aggressive!
ANSWER: Yes, it is one of the most
aggressive ways to meet these Indian set-ups.
Against the King's Indian, much work has
gone into developing methods of counterplay
for Black, but here in the more passive Old
Indian, it is rather harder for him to create any
real activity.
QUESTION: It sounds as though you don't
think much of Black's position?
ANSWER: I don't, especially with his having
lost a tempo already.

c6 7.

e3 c7 QUESTION: How come Black

doesn't just castle?
ANSWER: Flohr was concerned about

'castling into it'. Committing his king to the
kingside early would just clarify things for
White, who would be encouraged to get on
with his g4 kingside pawn storm. Flohr tries to
keep his opponent guessing, by leaving his
king in the centre and retaining at least the
theoretical possibility of castling queenside,
although it is not clear how serious this really
is.

[ QUESTION: Couldn't Black free his position

by exchanging in the centre with 7...exd4

8.

xd4 and then trying to organize ...d6-d5?

ANSWER: That is certainly a perfectly
plausible plan, and, indeed, it is Ray
Keene's recommendation at this point.
However, as Keene himself acknowledges, it
is far from clear whether Black will succeed
in arranging ...d6-d5 under favourable
circumstances. If played at once, for
example, then after d5? ( a better try is

8...0-0 9.

e2 d5 , but even this is

positionally fairly revolting after 10.cxd5

cxd5 11.e5

e8 12.0-0 followed by f4 )

9.cxd5 cxd5 10.

b5+ he would be

practically busted already.
In general, in the Old Indian, Black usually
tries to hold the e5 strongpoint for as long
as possible, and only to exchange on d4
when he has a concrete follow-up, but
against accurate white play, that follow-up
frequently does not materialize. The
essentially passive nature of the opening is
the precise reason why the King's Indian
development of the bishop to g7 became
more popular in the early 20th century. ]

8.

d2 bd7 9.d5 QUESTION: Now Black

loses the chance of ...exd4 forever!
ANSWER: Yes, and White secures a space
advantage.

b6?! QUESTION: Why don't you like this
move?
ANSWER: b6 is just such a terrible square for
the knight. OK, it frees the way for the
queen's bishop to move, and it attacks the
white c4-pawn, but in the long run, the knight
just does not want to be on this square.
QUESTION: So what should Black do?
Castle?

[ ANSWER: Well, 9...0-0 really plays into

White's hands. With the centre closed and
no danger of counterplay by ...exd4 and ...
d5, White would be only too happy to
continue 10.g4 and launch a huge kingside
attack by moves such as Ng3-f5, g5, etc. ]

[ QUESTION: What about blocking things up

with 9...c5 so as to close the queenside?
ANSWER: That, too, would play into White's
hands. He would again continue 10.g4
and follow up with Ng3 and 0-0-0, secure in
the knowledge that it is much harder for
Black to create counterplay on the
queenside.
QUESTION: OK, you keep rejecting all my
ideas, so why don't you suggest something
better for Black? ]

[ ANSWER: I agree it is not an easy position

for him. As the Russians say in such
position, "Good advice is expensive for
Black". However, I think Black should play

9...h5!? .

QUESTION: Good grief, that looks potty!
Where is Black going to castle after that?
ANSWER: He probably isn't going to castle
at all. Now the centre is closed, the black
king can stay where it is for a fair while.
Meanwhile, 9...h5 radically stops White's
intended g4 advance. On a good day and
with a following wind, Black can follow up
with something like ...h4 and ...Nf8-g6-h5,
and try to get some sort of grip on the f4-
square.
QUESTION: It all looks a bit optimistic. I
thought the strategy books tell you not to
make pawn moves on the side where the
opponent intends to attack?
ANSWER: That is the classical advice, yes,
but especially in recent years, such
prophylactic pawn advances have become
respectable in various lines. For example, in
the popular Najdorf Sicilian variation, 1 e4
c5 2 Nf3 d6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 a6
6 Be3 e5 7 Nb3 Be6 8 Qd2 Be7 9 f3, the
seemingly outlandish idea of 9...h5!?,
radically stopping White's planned g4-g5
advance, has proved itself a viable approach
to the black position. ]

10.

g3 d7 11.b3 QUESTION: With what

idea?
ANSWER: Nimzowitsch just indulges in a little
of his beloved prophylaxis. This move stops
any black ideas of ...cxd5, followed by ...Nc4
or ...Na4, perhaps getting a semblance of
activity and/or exchanging off a piece or two,
which might relieve his cramp.

h5 QUESTION: Looks like what you were

suggesting on the previous move!
ANSWER: Indeed, and still not such a bad
idea, I think, although Keene is critical of the
move (but he does not suggest any
alternative). Black is in a fairly bad way, in

any event, but this makes more sense to me
than castling short and just waiting to take it
on the chin.

12.

d3 g6 13.0-0 h7?!

This is inconsistent.

[ I think Black should follow up with 13...h4
14.

ge2 h3 ( 14...h5 is another

possibility) ) 15.g4 and then

f8 intending

to put the king on g7, whilst leaving the rook
on h8, to cover the h6-square and defend
h3.
QUESTION: And then you think Black is
OK?
ANSWER: No, I would not go that far. White
is still better, but I think this would have
been a tougher defence. ]

14.a4

[ The immediate 14.f4 is also very strong,

but Nimzowitsch is not in a hurry, and first
sets about gaining some more space on the
queenside. ]

14...h4 15.

ge2 c5 16.f4! The key advance.

White has to open some lines, in order to get
at Black's king and his other weaknesses,
such as the f7-pawn.

exf4 QUESTION: Is this forced?

[ ANSWER: No. Black should perhaps have

preferred the computer's 16...h3 17.g3

f6

, but even so, his position is pretty grim after

18.fxe5 dxe5 19.

g5 . The text is an

attempt to create a dark-square blockade
with his 17th and 18th moves, but
Nimzowitsch's energetic reply refutes it. ]

17.

xf4 g5 18.f2 f6 Desperately trying to

shore up the barricades on the dark-squares.
Given a free tempo or three, Black could
transfer a knight to e5 (and, with another
dozen or so free moves, his other knight to f7),
and would then have decent chances of
holding his lines, but as it is, Nimzowitsch
demolishes the blockade with an energetic
thrust.

19.e5! To the modern eye, this pawn sacrifice

cries out to be played, and it is entirely in
accordance with Nimzowitsch's theories of
restraint and blockade. He himself wrote of
how such sacrifices bring to life the forces in
the rear – most obviously, in this case, the
bishop on d3 and the knight on c3 (the latter
gaining the square e4).

dxe5

[ QUESTION: Presumably, 19...fxe5

is even worse?
ANSWER: It certainly feels worse intuitively,
doesn't it? Opening the f-file must assist

White, as the concrete variations confirm:

20.

g6+ d8 21.f7 f8 22.xe7 xg6

( 22...

xe7 23.xg5# is a nice mate )

23.

g7 f4 and now one of various winning

lines, suggested by the computer, is 24.a5

c8 25.b5 b8 26.xf4 gxf4 27.xf4!

exf4 28.

xf4 with a picturesquely crushing

position. It is particularly hard to believe
that the pathetic huddle of black pieces in
the north-west corner of the board could
belong to no less a player than Salo Flohr. ]

20.

g6+ d8 21.e4 See the note to move

19 – the knight assumes the wonderful
square vacated by the e-pawn, attacking the
f6-pawn and also threatening to win a piece
with 22 d6. Black is quite lost.

c8

[ QUESTION: Isn't 21...

f8 a better

defence?
ANSWER: Not really. There follows 22.d6

xd6 23.xd6 xg6 and now, in addition
to 24.

f7+ , ( there is the even stronger

24.

xg5! fxg5 25.xg5+ e7 26.f7+

and curtains. )]

22.

xh7 xh7 23.xf6 Not so much

regaining the sacrificed pawn as removing the
lynch-pin of the black kingside pawn chain.
Now the g5-pawn is not long for this world,
and Black could objectively have resigned.

g7 24.xg5 d6 EXERCISE: Can you
spot how Nimzowitsch broke through
decisively?
ANSWER: 25.

xh4! The start of a final

combination.

f5 26.xf5! The point.

xf5 27.d6! Neatly exploiting the indirect pin
on the e7-bishop.

xd6 28.xd6+ xd6 29.h5+ e7

30.

f1 d3

[ Keene gives the nice variation 30...

g6

31.

f8+ e8 32.g7! d7 , and now a

mass liquidation by 33.

xe7 xe7 34.xe8

xe8 35.xe8 xe8 36.c3 leaves White
easily winning the ending. ]

31.

f8+ d7 32.xa8 xe2 33.xe7 xe7

34.

g3 And the smoke clears, to leave White

an exchange and pawn ahead, with an
elementary winning position. The rest needs
no comment.

d1 35.f5 d8 36.xa7 c6 37.a8

d7 38.e3 xb3 39.f2 e4 40.xd8+

xd8 41.a5 a2 42.g4
1-0

Back to Contents Page

Chapter Two

Nimzowitsch the Defender

The ability to defend well is an essential quality in any great master. Nimzowitsch's special contribution to this
subject was the concept of what he termed "defence on heroic lines". In Chess Praxis, he laments the failure of the
chess world to take on board as part of defensive technique his new positional stratagems: prophylaxis, restriction,
centralization, and over-protection. 'Heroic defence' tends to imply risk-taking, and Nimzowitsch contrasts his
approach with what he describes as "the cautious need for niggling correctness in each move, the shy avoidance of
untrodden paths and, particularly, the fear of the 'colossal' (of anything on a big scale)...".

Amongst more modern GMs, the two players whose approach to defence perhaps reminds one the most of
Nimzowitsch are Larsen and Korchnoi. Larsen was well-known as a disciple of Nimzowitsch, and his similarity in
this particular respect is most obvious in his willingness to experiment in the openings (c.f. the above reference to
"the shy avoidance of untrodden paths").

Korchnoi is not generally linked with Nimzowitsch, his name being more frequently associated with that of Emanuel
Lasker, but in fact, I think Nimzowitsch and Lasker were quite similar in their attitude to risk and defence. Both were
willing to take considerable risks, in order to unbalance positions and create greater winning chances, even if this
involved accepting an objectively difficult position, which entailed walking a tightrope. Korchnoi was also fond of
this approach, especially in his legendary fondness for accepting hot-looking gambit pawns.

This attitude to defence inevitably risks suffering the occasional defeat, when the player goes a bit too far and
over-balances from the tightrope, but it also tends to bring a larger number of wins, especially as Black, than a
safety-first approach. In this chapter, we will see four examples of Nimzowitsch's defensive play in action.

Shoosmith-Nimzowitsch

is really a very early example of a Hedgehog-style set-up by Black (albeit arising from an

Old Indian), in which Nimzowitsch adopts a flexible set-up within his first three ranks, and then pounces with a
classical central counterattack when White over-stretches.

Nimzowitsch-Alekhine

is a grand battle with one of the

game's immortals, in which Nimzowitsch is initially forced on the defensive by a typically dynamic Alekhine pawn
sacrifice, but eventually hits back with a winning counterattack.

Brinckmann-Nimzowitsch (Berlin 1927)

and

Spielmann-Nimzowitsch (New York 1927)

are both cases where

Nimzowitsch's provocative opening play invites a strong attack, which he then defends, using his famous principles
of over-protection and restraint. The latter, in particular, shows him overdoing things a bit, and falling into a very
dubious, probably objectively lost position, but he still emerges in one piece, thanks to seeing further than his
opponent at the crucial moment.

Back to Contents Page

A55

Shoosmith,H
Nimzowitsch,A

Ostend

1907

[Steve Giddins]

1.d4

f6 2.c4 d6 3.f3 bd7 4.c3 e5

5.e4

e7 The Old Indian, which was the

forerunner of the King's Indian Defence.
QUESTION: What is the difference?
ANSWER: The key difference is in the
position of the black king's bishop – in the
King's Indian, this is fianchettoed on g7,
whereas in the Old Indian, it goes to e7.
QUESTION: Which is better?
ANSWER: The Old Indian is very solid, but
also rather passive. The King's Indian offers
Black much more dynamic play, thanks both
to the tactical possibilities on the long
diagonal and the idea of organizing a kingside
pawn storm with ...f5, after the closure of the
centre.

6.

d3 A slightly unusual development.

[ White generally prefers 6.

e2 , which gives

him more control over the d4-square. ]

6...0-0 7.0-0 exd4?! This is a very important

moment. The move was criticised by some of
Nimzowitsch's contemporaries, notably
Teichmann in the tournament book.
QUESTION: Why?
ANSWER: Basically, because it entails what
Tarrasch and other classicists used to call
"the surrender of the centre". The same issue
arises in various kingside openings, such as
the Philidor. The classical view was that Black
should maintain his central strongpoint on e5,
but Nimzowitsch took a different view. In line
with hypermodern thinking, he argued that
what mattered was not so much occupation of
e5, as the control of that square. He often
liked to make the exchange on d4, freeing up
the e5-square and half-opening the e-file,
along which he then hoped to pressure the
white pawn at e4.
QUESTION: So which do you think is the
correct approach?

[ ANSWER: In general, the modern

understanding is that it all depends on the
position. If Black can exert sufficient
pressure on the centre after the exchange,
then the Nimzowitsch approach can be
perfectly acceptable, but Black has to be
careful not to give White too free a hand in
the centre. In the Old Indian specifically, the
early exchange on d4 is generally frowned

upon, and the usual modern interpretation
would be to hold the centre with a move
such as 7...c6 , intending ...Qc7, ...a6, and
...b5, for example. ]

8.

xd4 e8 9.b3 e5 10.c2 a6

QUESTION: What is the idea of this?
ANSWER: I think Nimzowitsch is lining up a
potential ...b5 break, 'nibbling' at the white
centre. For example, he can follow up with ...
c6, ...Bd7 and then ...b5, looking to eliminate
the c4-pawn.

11.

b2 d7 12.h3 QUESTION: And this?

ANSWER: Again, a typical move in such
positions. White intends to play f4 at some
point, gaining additional space and driving the
enemy knight from e5. The move h3 takes
away the g4-square from Black's pieces.

f8 13.f4 g6 14.f3 c6 QUESTION:
How would you assess this position? It looks
as though White is much better, with far more
space and good central control.
ANSWER: White is slightly better, but one
should not over-estimate his advantage. It is
typical for such positions. He does have more
space, yes, but he also has more weaknesses.
The move 13 f4, especially, is quite committal
– it weakens the e4-pawn significantly (White
can never reinforce this pawn with f2-f3), and
also the white king, which may be vulnerable
along the g1-a7 diagonal.
QUESTION: So are you saying 13 f4 was
bad?
ANSWER: No, but it is double-edged. This is
always the key battle in such positions. In
order to exploit the extra central space
granted him by Black's exchange ...e5xd4,
White has sooner or later to advance f4 and
try to squeeze Black, but the latter, in turn, is
waiting for the chance to exploit the
weaknesses which will be created in the white
position, when that happens. Nimzowitsch was
a great believer in the potential strength of
such cramped but solid positions, as Black
has here. He liked to lure the white player
forward, in order to over-stretch him and then
land a counterblow. The Hedgehog formation
is a typical modern set-up, which has the
same underlying philosophy.

15.

ae1

[ 15.

ad1 may be better, eyeing up the

potentially weak d6-pawn. Shoosmith's
move is typical of a player who does not
believe in the black position, and just
thinks White has a free hand to build up on
the kingside. ]

15...b5! See the note to move 10. Black

initiates some counterplay, and threatens to
give White a nasty weak pawn on c4, by
exchanging on that square.

16.

d3 For the reasons explained above, the

alternatives were not so attractive to White,
but this queen move looks a little clumsy.

[ On the other hand, exchanging on b5 is

also unattractive for White. After 16.cxb5?!

axb5 , Black has the open a-file, and the

elimination of the c4-pawn has weakened
White's central control (as Nimzowitsch
points out, the d6-pawn is more likely to be
able to advance to d5, now the c4-pawn is
gone). ]

16...

c7 17.h1 This precaution was

perhaps overdue, just getting the king off the
exposed g1-a7 diagonal, but it does not
eliminate the structural weaknesses in the
white monarch's residence – the g3-square is
weak, for example, and there may be
sacrifices on h3 at some future date.

ad8 18.b1 b4?! QUESTION: Why don't
you like this?

[ ANSWER: It is not necessarily a bad idea in

itself. Black intends to transfer his attack
from the c4-pawn to the b3-pawn, following
up with a knight transfer to c5, and then ...a5-
a4. But I think the flexible 18...

b7!

is even better. This retains the option of a
later ...b4, whilst taking an 'X-ray' glance
both at the bishop on b2 (in the event of ...
bxc4 opening the b-file), and the weakened
e4-pawn, after a later ...c5. White's
position would be rather uncomfortable after
this.
QUESTION: You seem to be saying White is
worse here?
ANSWER: Definitely.
QUESTION: But how come?
ANSWER: Basically, because he has lost
the strategical battle over the committal
pawn moves such as f4. As we said, such
moves can enlarge White's space
advantage and potentially give him a
crushing position in the right circumstances,
but if he fails to suppress Black's
counterplay and generate real threats
himself, then he will just create weaknesses,
on to which Black can fasten. That is what
has happened here. White's pieces may
superficially look well-placed, but he has no
real threats (the black king is solidly
defended, for example), whilst White has
weaknesses, or potential weaknesses, on c4,

e4, b2, g3, etc. The experienced player can
just sense that the white position is hanging
by a none-too-secure thread, and it would
not take much to snap it. ]

19.

d1 c8?! Black continues his intended

manoeuvre. The bishop comes to the long
diagonal, to pressure e4, and the knight from
f6 intends to transfer to d7 and c5. However,
it is slow, and gives White a chance to
regroup.

20.

f3 d7 21.f5 c5 22.g4?

[ The slowness of Black's plan over the last

few moves has allowed White to improve
his position, and now 22.

de3 would give

him rather the better of it. However, he
instead gets completely carried away, and
thinks he can just crush Black on the
kingside. This is typical of what often
happens in such positions, as the white
player overrates his space advantage and
underestimates the fragility of his own
position. Now a vigorous counterattack
teaches him the error of his ways. ]

22...

e6 White's last move weakened the f4-

pawn, which Black immediately attacks.

23.

g3 b7 The vis-à-vis of bishop and king

on the long diagonal starts to look alarming,
but Shoosmith seemed oblivious to the
danger.

24.h4? Another absurdly optimistic lunge,

which places White in a critical position.

d5! QUESTION: Countering in the centre!

ANSWER: Exactly, as all the good books tell
one to do, in response to a flank attack. Black
wrenches open the long light-square diagonal,
at the end of which the white monarch is
sitting, looking distinctly under-dressed.

25.e5? With his f4-pawn attacked again,

White's resort to the text is perhaps
understandable, but now the long diagonal
opens with decisive effect.

[ The computer suggests, as the only way to

stay on the board, the pawn sacrifice 25.c5!
, although

xc5 ( 25...gxf4 26.e5

is White's idea, desperately trying to keep
lines closed, although even here,

xc5

27.

xf4 xf4 28.xf4 f8 gives Black

decent chances ) 26.h5

e7 is still better

for Black. At least White would still be on
the board, though, which is not the case
after the text. ]

25...c5

[ 25...dxc4 26.bxc4 c5+ 27.

g1 d2

is even simpler, but Nimzowitsch's move is
good enough. ]

26.cxd5

xd5 27.g1

[ QUESTION: Isn't 27.

e4 possible here?

ANSWER: Alas, no. Nimzowitsch refutes it
with

xd1 ( 27...d2 may be stronger still )

28.

xd1 xe4+ etc. ]

27...

d2 It is interesting just to compare this

position with the previous diagram. One can
see at a glance how the white centre has
evaporated, and the enemy pieces are now
pouring through the breach, into the white
position. This is absolutely typical of what
happens when these large space advantages
go wrong – the counterattack undermines the
white position, and the potential energy of the
defender's forces is released.

28.

fe3

[ 28.

f2 is a slightly more tenacious defence,

but even so, after

ed8 Black is in total

control: for example, 29.

d6 ( this pawn

sacrifice is the best hope, since 29.

de3

allows the neat

gxf4 , when capturing

twice on f4 leaves the b2-bishop hanging:

30.

xf4 xf4 31.xf4 xb2 , etc )

29...

xf2 30.xf2 gxf4 and the white

position is falling apart. ]

28...

c6 With the horrendous threat of 29...

Qh1 mate. This again exposes most visually
how the weaknesses of the white position
have come back to haunt him.

29.

f3 xf3

0-1

B02

Nimzowitsch,A
Alekhine,A

Semmering

1926

[Steve Giddins]

1.e4

f6 2.c3 A relatively tame reply.

[ 2.e5 is the only way to fight for an

advantage. ]

2...d5 3.e5

fd7 4.f4 e6 5.f3 c5 6.g3

By holding back his d-pawn, Nimzowitsch
avoids any transposition into a French
Defence, and heads for a more original
position. However, objectively, Black should
have no problems.

c6 7.g2 e7 8.0-0 0-0 9.d3 b6!?
The idea of this move becomes clear after
Black's next.

[ Other natural and fully reasonable moves

include 9...

b8 ]

[ and 9...f6 , the latter being suggested by

Nimzowitsch himself. ]

10.

e2 d4 This is Black's idea. He frees the

d5-square for his knight, whilst also exploiting
the fact that after White's last move, his
queen's knight can no longer reach the e4-
square.

11.g4?! This type of kingside pawn advance

is typical of such closed Sicilian-type
structures, but here it rather weakens the
white kingside.

[ Possibly White should prefer 11.c4 dxc3
12.bxc3 , although his central structure is

then none too secure. All in all, White's
handling of the opening has not been
impressive, to say the least. ]

11...f6?! QUESTION: Why don't you like this

central break now?
ANSWER: It is less good than it would have
been on either of the two previous moves.

[ Instead, Nimzowitsch recommended
11...

e8 12.g3 f8 13.e4 d5

, after which chances are about equal, ]

[ but the most consistent move is 11...

d5

12.

g3 b5 when Black is doing well. ]

12.exf6 gxf6 QUESTION: That looks rather

weakening. Is it forced?

[ ANSWER: Had Black played the ...f6 break

earlier, he could have recaptured with the
bishop on f6, but now 12...

xf6 would be

well met by 13.

g3 e5 14.f5 , and White

follows up by planting his knight on e4 and
organizing the pawn advance g5. ]

13.

g3 d5 14.e2 d6 15.h4 ce7

[ This creates the positional threat of 15...--
16.

xd5 exd5 17.hf5 , with a total

blockade of the black kingside.
QUESTION: But isn't it dangerous to give up
the fianchettoed bishop?
ANSWER: Usually, one would think twice
about exchanging off the fianchettoed
bishop in front of one's king, but here, one
can see easily that the white king is not
really in any danger. The long diagonal a8-
h1 is blocked by the black pawn on d5, so
Black is unable to create any threats along
that line, with his unopposed light-squared
bishop. One must always take such
specifics into account in assessing
positions. ]

16.

d2

[ 16.

h5 was more aggressive, and was

Nimzowitsch's own recommendation. ]

16...

c7 17.f2 c4! The start of a typically

energetic Alekhine counterattack.

18.dxc4

e3! This is the point of Black's

play.

QUESTION: Is he sacrificing a pawn?
ANSWER: Not really, because White cannot
take twice on e3 anyway. Meanwhile, with the
rook on f1, and pawns on c2 and g4 all
hanging, White has little choice but to take
once, thus losing his valuable dark-squared
bishop, after which his central dark squares
will be vulnerable.

19.

xe3 dxe3 20.f3

[ Most certainly not 20.

xe3?? c5 . ]

20...

xc4 QUESTION: What is the

assessment?
ANSWER: Black has regained the pawn
temporarily sacrificed, and has excellent
counterchances, with his two bishops and the
weakened white centre and kingside. However,
Black also has weaknesses on the kingside,
so the position is very sharp and double-
edged.

21.

e4 c7 22.b3 d4

[ 22...

xc2?? 23.ac1 drops a piece. ]

23.c3

b6 24.h1 d5? QUESTION:

What is wrong with this?
ANSWER: As we will see in the next note,
there is a tactical problem.

[ 24...f5! was stronger. The text allows White

a very dangerous line. ]

[ EXERCISE: Can you spot the flaw in

Alekhine's play with 24...

d5 ? ]

25.f5? Returning the compliment.

[ ANSWER: Nimzowitsch claims White

"would have won" after 25.g5 and, after
some initial doubts, the computer confirms
that this is indeed extremely dangerous.

A) Nimzowitsch gives 25...fxg5? 26.

xg5

xf4? 27.h5 and White is certainly
winning,;

B) and 25...f5 26.

h5 fxe4 ( 26...xf4

is no better: 27.

f6+ xf6 28.gxf6 c7

29.

g1 is decisive ) 27.xe4 again with a

completely winning position for White.;

C) Other defences also fail to solve

Black's problems: for example, 25...

f7

26.c4 f5 27.cxd5 fxe4 28.

xe4;

D) and 25...

d7 26.gxf6 xf6 27.g3+

h8 28.g5 , with a clear advantage to
White in both cases.;

E) 25...-- ]

25...

f4 26.fd1? Nimzowitsch does not

comment on this move, but it is a second
successive mistake, after which White finds
himself on the verge of defeat.

[ He could maintain the balance by 26.

ae1

. ]

26...

h8?! QUESTION: Why is this bad? It

looks quite natural to tuck the king away off
the half-open g-file.
ANSWER: It is, but this is a very sharp and
concrete position, where such general
considerations have to take a back seat to
specific tactical variations.

[ Black had a much stronger move in 26...e2!
27.

d2 exf5 ( Nimzowitsch mentions that

Wolf's suggestion 27...

b5 followed by ...

Qe5 is also very strong ) 28.gxf5

f7

, and White's position is hanging by a
thread. ]

27.

f1 Taking control of the e2-square, in a

bid to contain the enemy e3-pawn's "lust to
expand".

exf5 28.gxf5

e5?!

[ Again, Nimzowitsch does not comment on

this, but the computer indicates that 28...

e8

was stronger.
QUESTION: Why?
ANSWER: Well, again it is all about
variations, but the main general
consideration is that the white knight on e4
is playing a very important role in blocking
off both the e-file and the long diagonal a8-
h1. 28...Re8 makes a start in attacking the
knight, and intends the follow-up 29...Qc6,
doing so again. White has a lot of problems:
for instance, 29.

c4 c6 30.g1

(defending the knight on e4 tactically, by the
threat of mate on g8)

d7 31.ad1

(White continues to use tactics to try to
maintain his knight on e4, but it does not
work) e2! 32.

d4 ( 32.xe2 xe4

33.

xd7 xe2 wins material ) , and now

after 32...

xe4! 33.xe4 xe4! 34.xe4

c6 35.xc6 bxc6 the ending is much
better for Black.
QUESTION: The black e-pawn has a
charmed life!
ANSWER: Indeed. Ever since move 18, it
has looked as though this pawn could not
possibly survive, but it continues to do so,
and to cause White enormous problems in
the process. ]

29.

e1 d7 30.xe3 The pawn finally dies,

but in the meantime, the black pieces take up
extremely dangerous-looking positions.

c6 QUESTION: This pin on the long
diagonal looks a killer!
ANSWER: It does, but the computer shows
that, in actual fact, White is just about holding
here. Even so, it is extremely perilous for
White, and the slightest misstep is likely to
prove fatal.

31.

ae1? This should lose.

[ It was essential to deny the g8-square to

Black's rook with 31.

c4 , when White is

still hanging on by his fingertips. ]

31...

d5!?

[ The text is strong, but better still was
31...

g8! : for example, 32.c4 g7 33.b4

(what else? White can hardly move a piece)

d8 , and a decisive invasion follows with ...
Rd2. ]

32.

d3 xc3?

[ Missing a clear chance again, with 32...

g8!

. The text starts a combination, but Alekhine
had either missed or underestimated
Nimzowitsch's counter-combination. ]

33.

g6+! hxg6 34.g4! QUESTION: This

looks ridiculously slow!
ANSWER: It is amazing, yes. Keene
comments "A piece to the bad, and with a
large section of his army tied up by an
embarrassing pin, White can still afford to play
a quiet move and win".
QUESTION: So is White actually winning?
ANSWER: No, it seems not. The computer
shows that the combination is only good
enough for a draw, but even that is a
remarkable result, given the position.

f7? The losing move.

[ Nimzowitsch himself showed the best

defence, and analyzed the best follow-up for
both sides, as follows: 34...

g8! 35.fxg6

g7 36.d7+ xd7 37.xd7+ xg6

38.

d3! h6 39.h3+ g7 40.g1+

xg1+ "and the win would still be far away".
In fact, after 41.

xg1 xe4 42.xe4 ac8

I don't see that White is winning at all; the
position is just drawn. ]

35.

h3+ g7

[ Or 35...

h7 36.fxg6 c7 ( not 36...xh3?

37.

xh3+ and mate in two ) 37.xh7+ xh7

38.gxh7

xe4 39.xe4 , and White is

winning. ]

36.

c4

[ 36.

xg6+ was also good enough: f8

37.

c4 d5 38.xd5 xd5 39.h8+ e7

40.

xa8 , etc. ]

36...

d5 37.fxg6?!

[ Still winning, but the above variation was

much simpler: 37.

xd5 xd5 38.xg6+

f8 39.h8+ , etc. ]

37...

xe4?!

[ 37...

c7 was rather more tenacious,

although after 38.

h5 f5! 39.xf5 e8

40.

h7+ g8 41.h5 White wins. ]

38.gxf7+

f8 39.xe4

[ Or 39.

g8+ e7 40.f8+ xf8 41.h7+

e8 42.xd5 . ]

39...

xe4+ 40.xe4 The complications finally

end, and White is clearly winning.

e7 41.f8+ xf8 42.d5 d6 43.xb7+

d8 44.d3 d4 45.e4 e8 46.xd4
Although the computer shows up a number of
mutual errors, as is almost always the case in
such difficult tactical positions, this was still
a grand battle between two great players.
1-0

B01

Brinckmann,A
Nimzowitsch,A

Berlin

1927

[Steve Giddins]

1.e4

c6 The Nimzowitsch Defence, which,

unlike his eponymous defence to 1 d4, has
never enjoyed a great deal of popularity at
GM level. Nevertheless, it remains an
underrated line, which can certainly be
employed with confidence at club and amateur
level.

2.d4 d5

[ On the relatively rare occasions when this

defence is seen at master level, notably in
the practice of the late Tony Miles, Black
usually follows up with 2...e5 , which leads
to completely different positions. However,
the text move was always Nimzowitsch's
own preferred interpretation. ]

3.exd5

[ 3.

c3 is regarded as the strongest move,

and the only one to pose Black real
problems. ]

3...

xd5 4.e3 e6

[ From the point of view of equalizing, the

simplest and objectively best move is
undoubtedly 4...e5 .
QUESTION: So why would Nimzowitsch
prefer the text, shutting in his queen's
bishop?
ANSWER: It is typical of his rather
provocative, problem-setting style. 4...e5
invites simplifications, which will certainly
leave Black with a safe and solid game, but
may also make it harder for him to play for a
win. The text, by contrast, avoids such a risk
of drawish simplifications, but at the cost of
an objectively less good position. ]

5.

e2?! QUESTION: Doesn't this drop a

pawn?

ANSWER: It certainly sacrifices one, yes.
Brinckmann sees that taking on g2 will
expose Black to quite a few dangers, and is
keen to invite such a position.

[ Objectively, 5.

f3 was the best move, and

gives White a small but safe edge. ]

5...

xg2 "We have faith in the strength of the

defence", wrote Nimzowitsch. This capture
grants White the initiative and some attacking
chances, as the black queen will be chased
around a bit, but it is clearly the only attempt
to punish White's last move. Objectively,
White probably does not have enough for his
pawn, although in practical terms, it is the sort
of sacrifice that would appeal to the attacking
player. Nimzowitsch relished such defensive
tasks, however.

6.

f3 g6 7.e2 b4!? QUESTION: That

looks pretty extravagant? Surely Black should
not be trying to attack, with hardly any pieces
developed?
ANSWER: Well, this is not actually an
attacking move, despite appearances to the
contrary. Nimzowitsch is faithful to his
theories of centralization, and intends to bring
the knight to d5, anticipating White's plan to
harass the enemy queen by Nf4.
QUESTION: But it still involves moving a
developed piece twice!
ANSWER: Yes, and as Euwe comments, the
move would no doubt have horrified Dr
Tarrasch and others of his ilk. Interestingly,
though, it is the computer's first choice in this
position.

8.

a3 d5 9.c4

[ Nimzowitsch indicates 9.

b5 as the best

continuation, with which the silicon concurs.

A) He then gives the variation 9...c6
10.

f4 xf4 11.c7+ d7 12.xa8

( the computer points out that 12.

xf4!

is virtually winning for White, in view of

b8 13.d5! ) 12...g2+ with an unclear
position, adding that "attempts by Dr
Lasker in conjunction with the author
resulted in a draw".
QUESTION: What a crazy variation!
ANSWER: Yes, and it looks a little
fanciful.;

B) Instead of 9...c6, the computer prefers
9...

d8 , which it assesses as equal.

QUESTION: But surely Black must be in
trouble here?
ANSWER: It is not clear. It does look
rather shaky, but similar lines of the 4...
Qh4 Scotch have also proven surprisingly

resilient for Black. White has various
attractive-looking lines, but Black seems to
be holding the balance in them all: for
example, 10.

f4 ( or 10.c4 b4+

11.

bc3 xe3 12.fxe3 h6 ) 10...f5

11.

xd5 exd5 12.g1 c6 13.c3 f6

, and although the black king must provide
his opponent with long-term compensation
for the pawn, there is no sign of anything
more than that. All in all, this is typical of
the Nimzowitsch risk-taking, 'heroic
defence' approach. Steinitz used to play in
much the same way.;

C) 9...-- ]

9...

f6 10.g3 d8!? QUESTION: This

looks as though he is deliberately taking the
rise out of White! In the space of ten moves,
Black has moved his queen five times and his
knight three times, and, furthermore, the
queen has returned home again.
ANSWER: It is certainly an extraordinary way
to play. Nimzowitsch explains the queen
retreat by a desire to over-protect the knight
on d5. Mad as it may seem, Black has his
extra pawn, and has avoided any weakening
pawn moves. This is absolutely the way
Steinitz used to defend, having unshakeable
faith in the soundness of his position.
QUESTION: But surely the white lead in
development should be overwhelming?
ANSWER: It may look that way, but relatively
few lines are open. In particular, the central d-
and e-files are both closed, or at least, only
half-open. If Black can keep it that way, while
he catches up with his development, then he
will be OK. The question is whether White can
find a way to open the position, before Black
catches up in development. This is where
Nimzowitsch's "over-protection" of d5 comes
in – as he points out, Black's idea is to cover
that square, since White's main line-opening
idea in the centre is to play c4 and d5.

11.

d2 gf6 12.e5 c6 "Sustained over-

protection..." (Nimzowitsch). Note that each of
his last two moves has also strengthened the
black hold on d5.
QUESTION: But surely this is not serious?
ANSWER: On the contrary, Nimzowitsch was
entirely serious, I am sure. He had great faith
in these strategic conceptions, especially over-
protection and centralization.

13.0-0?! QUESTION: But surely that is

wrong? Why doesn't he castle long?
ANSWER: I agree, castling short looks
incorrect.

[ Nimzowitsch gives 13.0-0-0

as the alternative, and continues the
variation

b4 14.c3 e7 followed by

preparing long castling with ...Qc7(or ...Qa5),
...Bd7, ...Rf8 and ...0-0-0. The computer
basically agrees with this, and can find
nothing especially dangerous for White.
The obvious plan is 15.c4

xe3 16.xe3

, hoping to break with d5, but after 0-0
Black is very solid, despite the open g-file.
White has some compensation for the pawn,
but it is far from clear that it amounts to
anything more than equality. Note that
insisting on the break 17.d5? just gives
Black the advantage after exd5 18.cxd5

cxd5 – a triumph for Black's strategy of

over-protecting the d5-square. ]

13...

d6 14.h1 c7 15.c4 xe3 16.xe3

c5 QUESTION: This looks a bit loosening?

ANSWER: There is nothing wrong with the
move, as the game shows, but Nimzowitsch
also commented that it loosens the position a
little.

[ He suggested 16...

d7 as a solid

alternative, although then 17.

ae1

maintains some pressure. ]

[ If Black does not want to break in the centre

at once, then 16...0-0 is a perfectly good
move, although, as we will see, Nimzowitsch
seemed set against castling into the open g-
file. ]

17.

g5? Brinckmann allows himself to be

provoked into trying to punish Black for his
last move, but this fails.

[ Nimzowitsch correctly gave 17.

ae1

as White's best try, although after cxd4

( 17...0-0 is a good alternative ) 18.

xd4

, the move

A) is better than Nimzowitsch's 18...

d7

, which allows White to regain the pawn
with 19.

xd7 xd7 ( if 19...xd7

20.

d1 xg3 21.xd7+ xd7 22.xb7

b8 23.c6 ) 20.xg7 .;

B) 18...0-0! ]

17...cxd4 18.

ae1 When playing his last

move, Brinckmann must have overlooked or
underestimated the reply.

[ Now 18.

xg7 g8 19.xf6 xe5

is better for Black, but White's choice is
even worse. ]

18...

f8 QUESTION: He is determined not to

castle short!

[ ANSWER: Indeed. Nimzowitsch claims that
18...0-0 is wrong, because of the variation
19.

g1 h6 ( 19...e8! is simpler ) 20.f4

xe5 ( and now 20...h7 is good enough )

21.

xe5 d7? ( even here, 21...h8

is better for Black ) 22.

h5! , and Black is

actually getting mated: for instance,

xe5

23.

xh6 g6 24.g5 h8 25.xg6+

forces mate. However, Black has many
improvements in this variation, as we have
indicated. Thus, there is nothing
objectively wrong with castling short, but
even so, Nimzowitsch's move also gives
Black the advantage. ]

19.

g1 h6 20.f4 xe5 21.xe5

QUESTION: Why exchange queens?

[ ANSWER: It is capitulation to do so, I agree,

but now 21.

xe5 d7 forces the queens

off anyway. White is basically just lost. ]

21...

xe5 22.xe5 QUESTION: Two pawns

down!
ANSWER: Yes, White is lost. All that remains
is for Black to get his pieces out and
consolidate.

d7 23.b5 a6 24.b3 c5 25.a3 b8

26.b4

d7 27.c5 White thrashes about, in a

desperate bid to create some problems for his
opponent, but it is all insufficient.

e5 28.e1 xf3 29.xf3 d7 With the
threat of 30...Bc6.

30.

d3 d8 The pawn is indirectly defended

by the threat of ...Bc6+.

31.

g1 b5 32.d2 d5 33.e4 e7

34.

d6 g5+ 35.h1 c6+

0-1

C02

Spielmann,R
Nimzowitsch,A

New York

1927

[Steve Giddins]

1.e4

c6 2.f3 e6?! The text leads to a

kind of French structure, which was frequently
Nimzowitsch's aim when playing his
eponymous defence, but it must be admitted
that this is probably not a very good French.

[ Alternatives include 2...d6 (which was Tony

Miles' favourite), ]

[ as well as the move 2...e5 , transposing

back into regular Open Games, ]

[ whilst the creative Moscow IM Nicolai

Vlassov has extensively analysed and tested
the move 2...d5 , reaching a type of
Scandinavian Defence. ]

3.d4 d5 4.e5 QUESTION: How is this any

different from the positions reached via 2 d4

d5 3 Nc3 e6 4 e5? This can also be reached
by the move order 1...e6 2 d4 d5 3 Nc3 Nc6
which has enjoyed some popularity of late,
hasn't it?
ANSWER: The difference here is that White
has not committed his knight to c3. That
deprives Black of the chance of ...Bb4,
pinning the knight, and also allows White to
reinforce his centre with c2-c3.
QUESTION: So you don't like this variation
for Black?
ANSWER: Not especially. I am anyway a trifle
suspicious of these very slow French Defence
structures, where Black delays ...c5 and
instead manoeuvres (a more pejorative
commentator might say "grovels") around with
stuff like ...b6 and ...Qd7, etc. The version in
this game strikes me as even less attractive
than usual. Players such as Nimzowitsch and
Petrosian could handle the black side of such
structures very well, but few other players are
able to, and I really don't think they are
objectively to be recommended.

b6 QUESTION: So, what is the idea of this?

The long diagonal does not look an attractive
place for the bishop.
ANSWER: Well, as we will see, Nimzowitsch
intends to put the bishop on a6, in a bid to
exchange off his bad bishop. Such
manoeuvres are well known in these
structures.
QUESTION: It looks very slow!
ANSWER: Yes, that is the main objection.
Black's justification is that the position is
blocked, so time is of less consequence than
would be the case in an open position, but
even so, it is all asking a bit much, I think. On
the other hand, Black has few other sensible
plans, and is already committed to the type of
play which ensues.

5.c3

ce7?! QUESTION: Another weird-

looking move!
ANSWER: Yes. Nimzowitsch was clearly in
his element. We have already seen how much
he liked such fanciful and 'mysterious' moves.
It is all part of an elaborate scheme to
establish a blockade on the square f5.
QUESTION: I am not sure that is the first
square I would have thought of in this position.
ANSWER: Maybe not, but it is in total
accordance with Nimzowitsch's theories about
the pawn chain. The white pawn on e5 creates
a 'qualitative majority' for White on the
kingside. He wants to attack the base of the
black chain, on e6, so the obvious long-term

strategic advance for him to aim at is f4-f5.
Ergo, the f5-square is the key point, which
Black needs to try to control, and establish a
blockade there.
QUESTION: Deep stuff!
ANSWER: Yes, although thanks to
Nimzowitsch himself, 'every Russian
schoolboy' nowadays understands that f4-f5 is
a key idea in such positions. But, although his
basic strategic understanding of the position
cannot be faulted, even blockading on f5 is not
a panacea for Black, who is terribly cramped,
and has to create horrible dark-square
weaknesses in order to achieve the desired
light-square blockade.

6.

d3 a5 QUESTION: Yet another non-

developing move! But this one I guess I
understand – he is preparing ...Ba6, yes?
ANSWER: Indeed, which idea White takes
immediate steps to prevent.

7.

e2 f5 8.h4 QUESTION: What is the

idea?

[ ANSWER: White wants to drive away the

enemy knight with g2-g4, but the immediate

8.g4 allows the reply

h4 , hence the

preparatory text. However, the computer
points out that even this may be good for
White, thanks to a very nasty tactic: 9.

xh4

xh4 10.d2 , and now I suspect
Nimzowitsch (and Spielmann) assumed
Black is doing well after the counterblow h5
. However, the demon silicon beast points
out that White then has 11.

f3 , and now

xg4? ( 11...h3 is a better try, although
after 12.gxh5

xh5 13.g1 it is still not

very nice for Black ) fails to the queen-trap

12.h3

g2 13.h2 . ]

8...h5 Black stops 9 g4, but at the cost of

weakening g5.

9.

g5

[ 9.

g5 was also tempting. ]

9...g6? QUESTION: Isn't this part of his

blockade plan on f5?
ANSWER: Yes, but it is a really rather
horrible move. Alekhine himself wrote that
"Black's position could, perhaps, have
withstood the eccentricities committed so far",
but that this permanent weakening of f6 was
too much.

[ Nimzowitsch himself later agreed, and

recommended 9...

ge7 10.d2 c5

, although I still don't think much of Black's
game after simply 11.

df3 . ]

10.

d2 ge7

[ The attempted larceny of 10...

xh4?

is justly punished by 11.

b5+! e.g. d7

12.

xd7+ xd7 ( 12...xd7 13.xh4 )

13.

xf7 , winning. ]

11.

f1 QUESTION: That looks a bit surprising.

I assumed the knight would come to f3.
ANSWER: That would also have been good,
but Spielmann's move is very logical too. He
intends to challenge the black blockade on f5,
by playing f3 and g4, so his knight covers the
g3-square in advance.

c5 Black finally makes this thematic French

pawn break, exerting some pressure on the
white centre.

12.f3 c4?

[ Black's last chance to obtain a vaguely

playable position was 12...cxd4 13.g4 hxg4

14.fxg4 , and now the computer's suggestion
dxc3! ( 14...

h6 15.cxd4 is just positionally

great for White, who has refuted the
attempted kingside blockade plan,; whilst

14...

xh4? 15.f2 loses a piece for Black )

15.gxf5

xf5 16.bxc3 c7 .

QUESTION: So Black has sacrificed a
piece?
ANSWER: Yes, but he has two pawns, and
has broken up the white position a good
deal. The white king has no particularly safe
home.
QUESTION: So Black is OK here?
ANSWER: I am not sure about that. White is
still better after 17.

d2 , for example, but

Black has some play and some practical
chances. ]

13.

c2 Now the character of the game has

changed completely from the line given in the
previous note. The position remains closed,
but Black has no real counterplay at all.
QUESTION: What is the problem?
ANSWER: The move 12...c4 has lifted the
pressure from White's centre, which is now
very securely defended, and he has a free
hand on the kingside.
QUESTION: But Black can play ...b5-b4, can't
he, with a queenside counterattack?
ANSWER: Yes, but that is little more than a
pin-prick for White. Black cannot bring any
extra pressure to bear against c3, which is
anyway easy to defend. Meanwhile, Black
has terrible problems on the kingside.
QUESTION: So your verdict would be 'Don't
try this at home'?
ANSWER: Definitely. If pressed for a
judgement, I would say Black is strategically
lost.

b5 14.g4

g7

[ Not 14...hxg4? 15.fxg4

xh4 ( or

15...

xh4 16.f2 ) 16.xh4 xh4 17.f2

and wins ]

QUESTION: This looks like another typically
bizarre Nimzowitsch position!
ANSWER: Yes, one does not see too many
fianchettoed knights. But that is for a good
reason – this position is just dreadful for Black.
His only hope is that the fairly blocked nature
of the position will make it hard for White to
break through.

15.

g3!?

[ Logical enough, although the computer is

very keen on the plan of bringing the white
queen to f4: 15.

d2 c6 16.f4

, and after

e7 17.d2 Black is appallingly

tied up. Admittedly, White still has to break
through, though. ]

15...

c6 16.g2 It was still possible to revert

to the plan of bringing the queen to f4, but
Spielmann has another idea. He intends to
break in down the g-file, with the aid of a
sacrifice on f7.
QUESTION: Was that really necessary?
ANSWER: Maybe not, but a sacrifice is the
usual way to deal with these blocked
structures. Once White has all his pieces
mobilized, his massive space advantage
usually allows him to transfer pieces to the
critical sector of the board much more readily
than Black is able to do, suffering as he does
with such a cramped position, and lacking
lines of communication for his pieces. As the
later notes will show, Spielmann's plan was
perfectly correct and should have won.

e7 17.gxh5 gxh5

[ 17...

xh5? 18.xh5 xh5 allows an

immediate winning breakthrough with

19.

xf7 . ]

18.

g1 a7 19.xf7!? The critical moment

of the game has been reached.

[ The text is objectively good enough to win,

as the following note shows, but Alekhine's
suggestion of one more preparatory move
with 19.

e2 was even better.

QUESTION: Why is that?
ANSWER: The king move serves two
purposes. Firstly, it gets the king out of the
way of a check from h4 (we will see the
move ...Bxh4+ cropping up in various of the
subsequent variations), and it clears the
back rank, in anticipation of the queen's
rook being able to swing over to the
kingside.
QUESTION: But presumably Spielmann was

afraid of giving Black extra time to arrange
his defences. Can't he make profitable use
of the tempo gained?
ANSWER: Analysis suggests not. 20 Nxf7 is
now a devastating threat, so the obvious
thing to do is to cover f7, by means of

f8

. However, then 20.

h7! threatens Nf6+

and/or Bh6 and Nxf8. Black is helpless, as
the computer confirms:

d7 21.h6 c7

22.

f1 (the threat is 23 Nxf8, simply winning

a piece, and Black is so helplessly cramped
and tied up that he has no way to avoid the
loss of material)

f5 23.xf8 xf8 24.xf8

xf8 25.xf5 exf5 26.g5 and White has
an extra exchange and still crushing
pressure on the dark squares. ]

19...

xf7 20.xh5? This second piece

sacrifice is a step too far, however.
Spielmann had probably missed Black's 23rd
move.

[ Alekhine suggested various possible

alternatives for White, one of which is

20.

e4! , which the computer confirms is

the strongest, and just wins by force:

xh4+

(now we see why the preparatory 19 Ke2
would have been useful, so this move does
not come with check, but White is winning
anyway) 21.

e2 g8 ( or 21...e8

22.

d6+ e7 23.h3 and wins ) 22.d6

f8 ( if 22...c7 23.g6 with ideas of Bh6
or even Ne8; after

f8 24.h6 the attack is

crushing: the threats include Ne8 and Rg2
followed by Rag1 ) 23.

h3 e7 ( 23...d8

24.

xc8 ) 24.xg7+! xg7 25.xc8 c7

26.

xe6+ f7 27.xe7+ xe7 28.b6

and White has material equality (two pawns
for the exchange), and still a raging attack,
with Be3 and Rg1 coming. ]

[ Another Alekhine suggestion was 20.

e2

, after which Keene gives the variation

xh4+ 21.d1 g8? ( however, in this line,
Fritz points out that Black is hanging on after

21...

g8 , so 20 Ne2 is not so good ) 22.f4

f7 23.g6 e7 24.xh8 xh8 25.g6
, winning. ]

20...

xh4+ 21.e2 xh5 22.g6+ e7

23.

xh5 d7!!

[ No doubt a very unpleasant shock for

Spielmann, who had presumably only
reckoned on 23...

xh5 24.g7+ e8

25.

g6+ , although even here, things are

still fairly unclear after

d7 26.xh5 e7

. ]

24.

g7+ e7 25.f7 h2+ 26.d1 c7

[ 26...

f8 was also winning, but at most

points from now on, Black had other equally
good choices than the moves played. The
rest does not really require detailed
commentary. ]

27.

f4 xb2 28.h7 b6 29.g8 c7

30.

h8 d8 31.g6 g2 32.h1 xg6

Black gives back some of his extra material,
to eliminate one of the white attacking pieces.
This is not the only way to win, but is a typical
defensive technique, of course.

33.

xg6 b4 34.g7 c6 35.h8 a4+

36.

e1 c6 37.xc8? Allowing a decisive

counterattack, but it really does not matter
anymore. White could and should just resign.

h4+ 38.g3 xg7 39.xh4 c2 40.d8+

xd8 41.b8+

[ He runs out of checks at once after
41.

xd8+ b7 . ]

41...

b7 QUESTION: But surely Nimzowitsch

was just lucky here, wasn't he? He was
outplayed and had a lost position, only for his
opponent to miss a tactic and sacrifice
unsoundly.
ANSWER: That is all true, but this is in many
ways typical of the 'provocation' style, of
which Nimzowitsch was fond. His opening play
was deliberately very risky and provocative,
tempting his opponent to attack and to burn
his boats in the process. It is true that, this
time, Nimzowitsch went a bit far, and
objectively was losing, but that is always an
inherent risk with such an approach.
Other defenders, who have adopted a similar
approach, include the young Korchnoi, and he
too would lose a few games that way, and an
even greater number of post-mortems, where
his opponents would succeed in showing that
they should have won. However, 'should
haves' and 'could haves' do not yield even a
fraction of a point in the tournament table,
and overall, the Nimzowitschs and Korchnois
won an awful lot more games as Black than
they would have done with a correct,
equalizing approach.
0-1

Back to Contents Page

Chapter Three

Nimzowitsch the Blockader

'Blockade' is probably the idea most associated with Nimzowitsch. Of course, it is just one among many strategic
ideas, and so it could have been subsumed under the general strategy chapter, but in view of its importance in
Nimzowitsch's practice, and the many outstanding examples, I decided to give the subject a chapter of its own.

This chapter contains a couple of Nimzowitsch's most famous games.

Nimzowitsch-Salwe

was a classic early

example of his theories, specifically on the white side of the French Defence. Rather than maintain his d4-e5 pawn
chain, Nimzowitsch exchanges both pawns off, but succeeds in controlling the centre by a piece blockade on these
two squares.

Johner-Nimzowitsch

is his immortal blockade game, and is the first of several examples of his handling

of his eponymous defence to 1 d4. Black exchanges on c3, doubling the white pawns, and then establishes his
blockade, in this case with a famous queen manoeuvre d8-d7-f5-h7.

Johner-Nimzowitsch

and

Matisson-Nimzowitsch

are also both Nimzo-Indians, and see highly talented masters

reduced to utter helplessness by their failure to cope with Nimzowitsch's blockading strategy.

Matisson-Nimzowitsch

sees White resign inside 30 moves, in a position where he still has material equality and is not threatened with
mate, whilst in

Bogoljubow-Nimzowitsch

, a double world championship challenger is humiliated. Finally,

Mannheimer-Nimzowitsch

is another of those games where Nimzowitsch succeeds in establishing absolutely

monster blockading knights, in this case entrenched on c4 and e4. This game, like several others, also shows his
excellent use of the strategy of 'alternation', i.e. switching the attack from one weakness to another, on different
sides of the board.

Back to Contents Page

C02

Nimzowitsch,A
Salwe,G

Karlsbad

1911

[Steve Giddins]

1.e4 e6 2.d4 d5 3.e5 c5 4.c3 QUESTION:

I thought Nimzowitsch preferred to sacrifice
the d4-pawn in this Advance French?
ANSWER: He did tend to do so later in his
career. His treatment of the variation evolved
over the years, starting out with this orthodox
defence of d4, and then moving on to lines
where he sacrificed d4, but hung on to the e5-
pawn like grim death.

[ His final theoretical word was 4.

g4

, which we examined in Nimzowitsch-
Haakanson. ]

4...

c6 5.f3 b6 6.d3 d7?!

QUESTION: Why the punctuation? Isn't this a
perfectly normal developing move?
ANSWER: Yes, but the fact is that, as a direct
result of this one single game, the move
virtually disappeared from master practice and
has not returned to this day.
QUESTION: So is it really so bad?
ANSWER: Actually, I think not, but it does
give White an extra option, adopted by
Nimzowitsch in this game.

[ Objectively, Black is probably fine if he

defends correctly, as the later notes will
show, but Black has one less variation to
know if he sticks with 6...cxd4 7.cxd4

d7

. ]

7.dxc5!? This is the key idea.

QUESTION: It looks strange to give up the
centre like this. I thought the whole point of
the Advance variation was to keep the pawns
on d4 and e5, so as to cramp Black?
ANSWER: That was, indeed, the
understanding of the line, before this and
some similar Nimzowitsch games. The latter's
great contribution was to show that White can
still cramp Black's position, even by
exchanging off the d4- and e5-pawns, if he
can maintain a piece grip over the d4- and e5-
squares. It is all about restraining and
blockading the black e- and d-pawns, which
of course was meat and drink to Nimzowitsch.

xc5 8.0-0 f6? This, rather than move 6, is
Black's first real mistake.

[ He should have prevented White's next with
8...a5 , keeping the bishop on c5, where it

controls the d4-square. Black would then
have had no real problems. ]

9.b4! This is much more than simply a space-

gaining move. Rather, it is part of White's
overall strategy of fighting for the dark squares,
especially d4 and e5. Black's bishop is driven
from the a7-g1 diagonal, which weakens his
control of d4.

e7 QUESTION: Doesn't this block the
development of the king's knight?
ANSWER: Not really, as Black intends to
exchange on e5 and then bring the knight to f6.
He believes, by so doing, that he is freeing his
game, since he will have eliminated White's
central pawn duo on d4 and e5, but
Nimzowitsch's understanding of the position is
far deeper.

10.

f4

[ QUESTION: That is a bit of a surprise. I

would have expected the tempo-gaining

10.

e3 .

ANSWER: It is a good example of
Nimzowitsch's consistent and planned
approach to the position. Kicking the enemy
queen with tempo is superficially attractive,
but it does not fit in with White strategic plan,
which is control of d4 and e5. In that plan,
the bishop stands better on f4, over-
protecting e5, and also keeping the e-file
open, so that White can play such moves as
Re1 and/or Qe2, further reinforcing his
control of e5. ]

10...fxe5 11.

xe5 xe5 12.xe5 f6

[ Black would like to challenge the

blockading bishop with 12...

f6?

, but as Nimzowitsch himself points out, this
is tactically impossible because of 13.

h5+

g6 14.

xg6+ hxg6 15.xg6+ e7

16.

xf6+ xf6 17.g7+ , winning

material. ]

[ QUESTION: So how should we assess the

game position after 12...

f6 ?

ANSWER: It all depends critically on the
blockade. At the moment, Black's e-pawn
is backward and weak, and his bishop on d7
is shut in. If White can keep his grip on e5,
and thus prevent Black from freeing himself
with ...e6-e5, then White can look forward to
an advantage. However, this is not so
easily achieved, and White has his own
weaknesses, notably the backward c3-pawn,
which can become a source of counterplay
for Black. If Black succeeds in liquidating
the blockade on e5, and advancing his e-
pawn, then White's position could very
quickly deteriorate markedly, as Black would
have the open f-file, and pressure against f2

and c3. So everything depends critically on
what happens over the next few moves.
White must fight tooth and nail to maintain
his blockade on e5. ]

13.

d2 The knight heads to f3, to reinforce

the blockade.

[ Tarrasch claimed that 13.

c2? was better,

but Nimzowitsch showed that this fails to a
typically dynamic response: 0-0! (Tarrasch
failed to consider this pawn sacrifice)

14.

xf6 xf6 15.xh7+ h8 16.g6 e5

, when White has won a pawn, but at the
cost of giving up his blockade on e5 and
allowing the black forces great freedom.
Black would be clearly better here – an
excellent illustration of how the strategic
battle over the e5-blockade is pre-eminent in
this position, and is often more important
than a small material investment such as a
pawn. ]

13...0-0 14.

f3 d6

[ QUESTION: Isn't 14...

b5 a typical French

idea, seeking to exchange off the bad
French bishop?
ANSWER: It is, but it does not contribute to
the battle over the e5-square, and it also
weakens the black centre. Nimzowitsch
gives the line 15.

d4 a6 16.xb5 xb5

17.

g5! , winning a pawn: for instance, c6

18.

e1 and a pawn is lost. ]

[ QUESTION: OK. If we have to fight for e5,

maybe 14...

g4 is an idea?

ANSWER: It certainly is, and may well be
the best move. Keene then gives the line

15.

g3 f6 16.h3 h6 17.e5 f7

( instead, the computer's 17...

b5

is fine for Black; thus, 14...Ng4 seems
perfectly adequate for equality ) 18.

xf6

gxf6 19.c4! "and White is still in control". ]

15.

e2

[ QUESTION: Isn't 15.

d4 just as good?

ANSWER: The general view, since
Nimzowitsch's own day, has been no. This
is another variation, which illustrates how
the critical importance of the e5 blockade
informs the play of both sides. Nimzowitsch
gives a dynamic tactical way for Black to
break out:

c7 16.e2 ( however, in this

line, 16.

e1 , not considered by

Nimzowitsch or subsequent analysts, is the
computer's recommendation, and is
probably better, the main point being that
now

g4?! 17.h3 e5 fails to 18.hxg4 exd4

19.

b3! with a clear advantage; thus, it is

not entirely clear that 15 Bd4 is wrong,

although the text move is certainly simpler )

16...

g4! 17.h3 ( Keene, quoting Botterill,

also considers an interesting line starting

17.

xh7+ xh7 18.g5+ g6 19.xg4

, but the computer points out that after

xh2+! 20.h1 f4 21.xf4 xf4
Black is certainly not worse ) 17...e5!
and the blockade is broken (Nimzowitsch). ]

15...

ac8?!

[ This is the last chance to prevent White

regrouping with Bd4 and Ne5. Tarrasch
suggested Black should prefer 15...

xe5

16.

xe5 ac8 , but after 17.c4

Nimzowitsch claimed "White has an
enormous position". That seems slightly
hyperbolic, although White is definitely for
preference. ]

16.

d4 c7 17.e5 QUESTION: It looks as

though White has won the battle for e5?
ANSWER: Yes, indeed. With his last move,
White announces that he has succeeded in
annexing control of the key squares d4 and e5,
and has a firm grip on the centre. Note that
his bishop on d4 also performs the very
useful extra task of defending the backward
c3-pawn.
QUESTION: So now what happens?
ANSWER: Having secured the blockade,
stage two is to exploit it. First, White brings up
his last reserves to secure the blockade fully.

e8 18.ae1?! Natural, but missing a small
tactical point.

[ 18.f4 was better, avoiding the trick shown

in the next note. ]

18...

xe5? QUESTION: Wow, that looks a

shocker of a positional move!
ANSWER: Yes, it is a dreadful move.
Admittedly, Black's position is pretty grim, but
this really amounts to positional capitulation.
It is interesting to compare this with Wolf's
22...Bxe5 in Nimzowitsch-Wolf, which is very
similar – Nimzowitsch seemed to have a
knack for inducing his opponents to make
such hideous positional exchanges.

[ Unlike the Wolf game, here even the

computer frowns on Black's last. It suggests
as the best chance the move 18...

e4!

, after which White's advantage still might
not be so huge: for example, 19.

xe4 dxe4

20.f4 ( 20.

xe4 b5 is the point, revealing

the tactical weakness of White's 18th move )

20...exf3 21.

xf3 h5 22.e3 ( if

22.

xe6+ f7 ) 22...ce8 and although

White retains some advantage, Black is not
without his chances. ]

19.

xe5 c6 20.d4 QUESTION: Why

retreat from the cherished e5-square, after
making such efforts to get it?
ANSWER: Well, as Nimzowitsch and the rest
of the Hypermoderns always liked to
emphasise, the battle in such cases is always
over control of squares, not necessarily
occupation of them. But in fact, White's last
did cause another argument with Dr Tarrasch,
who recommended leaving the bishop on e5
and playing something like 20 f4. However,
Nimzowitsch regarded 20 Bd4 as an important
move, the aim of which was to force the black
bishop to commit itself.

d7 Now the bishop loses the possibility of
coming to g6 or h5.

21.

c2 QUESTION: That looks a slightly odd

square for the queen.
ANSWER: The point is to start the process of
converting White's blockade into a kingside
attack. The text threatens to win a pawn by
taking on f6 and h7, so Black must respond to
this threat. Moving the h- and g-pawns are
both unattractive, so Salwe defends indirectly.

f7 22.e3! Now we see the other point of
White's last – he cleared the e-file for this
rook transfer to the kingside.

b6 23.

g3 Now the h-pawn is again

threatened, since after 24 Bxf6, the recapture
with the g-pawn is prevented. Hence Black's
next...

h8 ...which, however, allows a tactic:

24.

xh7! e5!? QUESTION: Can't the bishop

be taken?

[ ANSWER: Alas, no. After 24...

xh7

there follows 25.

g6 with the irresistible

threats of Bxg7+ and Rh3, ( but 25.

h3?

f5 is less clear. )]

25.

g6 e7 QUESTION: So Black has finally

lifted the blockade, albeit at the cost of a
pawn.
ANSWER: Yes. We saw several variations
earlier, in which such a transaction was quite
favourable for Black, but here, of course,
things are very different. Black gets some
temporary counterplay, but his material deficit
and weak king are fatal disadvantages.

26.

e1 d6 27.e3 d4 28.g5 xc3

29.

xc3 dxc3 30.xc3 g8 31.a3

QUESTION: What a funny move! What is the
point?
ANSWER: He just defends his queenside
pawns and prepares to transfer the queen to
h4.

[ The immediate 31.

c4+ e6 32.h4

can be met by

xa2 , and White has

nothing immediately clear. ]

31...

f8 32.h4 The bishop heads to g3, to

attack the weak e5-pawn.

e8 33.f5 d4 Otherwise the e5-pawn will
drop, but now the simplifications give White a
winning endgame.

34.

xd4 exd4 35.xe7 xe7 36.d3 d6

37.

xf6 gxf6 38.f1 c6 39.h4 A classic

game, which played a very important role in
developing the modern understanding of
central control.
1-0

E41

Johner,P
Nimzowitsch,A

Dresden

1926

[Steve Giddins]

1.d4

f6 2.c4 e6 3.c3 b4 4.e3 0-0

[ This is, of course, a perfectly normal and

good move, but if Black wishes to reach the
structure in the game, without the move ...d7-
d5, then the usual modern move order
(named after Hübner, but actually developed
in the 1950s by Lajos Portisch) is to
dispense with castling and play immediately

4...c5 5.

d3 c6 6.f3 xc3+ 7.bxc3 d6

. ]

5.

d3 c5 6.f3 c6 7.0-0 xc3

[ 7...d5 is the more usual way to continue

nowadays, but this was not Nimzowitsch's
interpretation of his eponymous defence. He
is intending to double the white pawns and
play against them. ]

8.bxc3 d6 QUESTION: So what is Black's

plan?
ANSWER: His main immediate idea is to
challenge in the centre with ...e5, hoping to
induce White to close the position with d5.
This will leave the white bishops relatively
inactive, and free Black's hands for flank
manoeuvres with his knights.

9.

d2! This is a good move, as Nimzowitsch

himself acknowledges.
QUESTION: What is the point?
ANSWER: There are a couple of points. The
first is to free the f-pawn to advance.

[ The other with 9.

d2 is to anticipate

Black's intention of e5 10.d5

a5

( 10...

e7 would be the modern,

Hübneresque approach ) , planning to attack
the white c4 weakness with ...b6 and ...Ba6.

White will now meet this with 11.

b3

, forcing the exchange of the black knight. ]

[ The continuation 9.e4 e5 10.d5

e7

is also perfectly respectable (even the main
line), but invites Black to establish what he
is hoping for, namely a blocked pawn
centre. ]

9...b6 This is motivated by the idea of freeing

the b7-square, so that in the variation above,
Black could meet Nb3 by retreating the knight
to b7.
QUESTION: That looks a bit artificial.
ANSWER: Yes, it is.

[ It would be more natural, especially to

modern eyes, to play 9...e5 and then retreat
the knight to e7. ]

10.

b3?! Now this makes little sense and just

misplaces the white knight.

[ Nimzowitsch, endorsed by Larsen (who

annotated this game at length in the classic
1975 Batsford book 'Learn from the
Grandmasters') recommended 10.f4
, with the continuation e5 11.fxe5 dxe5

12.d5

a5 13.b3 b7 14.e4 e8

, which he assessed as equal. Black has a
nice outpost on d6 for his knights (although
they cannot both go there, of course), but
White has extra space and the possibility of
a4-a5. The computer strongly prefers White,
in fact, but I do not really believe that
assessment – computers tend to overrate
space and passed pawns, but the white d5-
pawn is going nowhere in this position. ]

10...e5 11.f4

[ QUESTION: Why not play 11.d5

as intended?
ANSWER: Now this would be well met by

e4! followed by ...Ne5 – a common theme in

these positions, when White has not himself
played e4 or taken control of the square. ]

11...e4 12.

e2 d7!? Probably the most

famous move in the game, and one of the
most famous in Nimzowitsch's entire
tournament practice.
QUESTION: What is the idea? It looks a bit
peculiar!
ANSWER: It is another of those deep and
rather convoluted strategic ideas, of which
Nimzowitsch was so fond, and all tied up with
the concepts of prophylaxis and the blockade.
His reasoning was as follows: Black has the
advantage on the queenside, thanks to the
doubled white c-pawns. Given a free hand, he
will play ...Ba6 and gradually lay siege to the
c4-pawn. However, before doing so, he must

attend to White's counterplay. On the kingside,
White has what Nimzowitsch called a
"qualitative majority", i.e. not an actual
numerical majority of pawns, but a superior
pawn position, thanks to the advanced f4-
pawn. Basically, White wants to advance on
the kingside with g4 and develop an attack on
that flank. Black's basic task, therefore, is to
counter this plan.

[ QUESTION: Aha, I see. So 12...

d7

just stops 13.g4 . ( But surely White can
prepare the advance with, say, 13.h3
, can't he?
ANSWER: Yes, but Nimzowitsch's plan goes
deeper than just stopping g4 for one move. )]

[ The best way to understand Nimzowitsch's

choice of 12...Qd7 is from his note on the
alternative, which was the straightforward

12...

e8 . Now he gives the line 13.g4 f5

14.d5

e7 15.g5 . As we see, the position

has become extremely blocked, even
petrified. Nimzowitsch makes it clear that
this was what he was trying to avoid; he
regarded his position as already being so
good, that he had to avoid anything too
drawish.
QUESTION: But is that really the case?
ANSWER: Probably not. Just as most
masters of his day (and especially before
him) probably underestimated the
drawbacks of such doubled pawn
complexes as White has here, so
Nimzowitsch in his turn tended to
overestimate their weakness, and assumed
that Black had the advantage, in many
positions where the modern player would
not consider him to be more than equal,
often not even that. This is probably such a
case, with Nimzowitsch already thinking that
his opponent will be desperate to block
things up and try to hang on for a draw, at
the first opportunity. ]

[ Indeed, Kasparov suggests that White has

no need to play to close the position after

12...

e8 , and should prefer 13.d2 f5

14.

e1 "followed by h3-g4 and Bh4, with

some initiative for White". ]

13.h3!? QUESTION: This looks logical.

ANSWER: It is, and is not necessarily bad in
itself, but White also had several alternatives.

[ Firstly, the Hungarian GM Szabo (who also

annotated this game in the same 'Learn
from the Grandmasters' as Larsen – such
was the enormous influence of this game on
a whole generation of players) suggested

the eminently sensible move 13.f5
, just radically putting paid to the idea of a
blockade by Black on the f5-square.
Interestingly, Nimzowitsch does not mention
the move. Szabo assumes this was because
of

e7 14.g4 h5 , but even so, after 15.h3

the position remains unclear. ]

[ A second option is 13.

d2 , but then we

see another point of Black's 12th move:

e7

14.

e1 a6 and there is the possibility of ...

Qa4 in the air. ]

[ This last possibility led Larsen to suggest

flicking in 13.a4 a5 and only then 14.

d2

. Kasparov continues this variation with

e7

15.

e1 a6! 16.d2 d5 17.f2 fc8

"with some initiative for Black". ]

All in all, White had a number of choices here,
but the key thing is that he has to understand
what is going on in the position, and
appreciate the importance of preventing a
blockade materializing on the kingside. Alas,
as soon becomes clear, Johner either did not
get the point, or miscalculated a tactic at the
crucial moment.

13...

e7 14.e1?

[ QUESTION: Well, that I don't understand!

Why not play 14.g4 , which is surely the
logical follow-up to his last move?
ANSWER: I agree totally. I guess the point
is that Johner thought g4 could not be
prevented, so he did not need to hurry. ]

[ Szabo also suggests that the preparatory
14.

h2 is possibly even better than the

immediate advance of the g-pawn. What is
clear is that Johner's move is just pointless,
as his queen has nothing to do on the
kingside. Worse than that, it also allows
Black's next. ]

14...h5! Of course! Nimzowitsch does not

need asking twice. He seizes the chance to
clamp down on White's intended g4 advance.

15.

d2 QUESTION: It seems strange that a

master player such as Johner could not see
that he was allowing 14...h5. His 13th move
shows that he understood the need to
advance g4, so why would he now forget
about it?
ANSWER: A good question. I think this is
another example of a poor strategical move
actually resulting not from a deficiency in
strategical understanding, but from a tactical
oversight. I think Johner understood that he
should not allow Black to advance 14...h5,
and believed that he was still preventing it.

[ I think he intended 15.

h4 , when it looks

as though the h-pawn cannot be defended.
However, he had overlooked that after

f5

16.

g5 , Black has the trick h7! 17.xh5

g3 winning the exchange.
QUESTION: That is very interesting. A bad
positional or strategical decision can result
just from a tactical oversight?
ANSWER: Yes, and that is why, if you are
trying to improve, it is vital to analyze your
games carefully and record your thought
processes during the game as honestly as
possible. In this example, a trainer, looking
at the bare game score, could easily
conclude that the white player's weakness
was in his strategical understanding, and
not appreciating the need to prevent the
blockade, whereas in fact, if my speculations
are correct, Johner's problem was with his
calculation of variations. ]

15...

f5! Continuing with the main kingside

blockading idea, envisaged at move 12.

16.

h2 Johner prepares Rg1, to support g2-

g4, but runs into the final point of the black
plan.

h7! QUESTION: So this is the big idea?
ANSWER: Yes, and it is one of the most
famous manoeuvres in Nimzowitsch's
tournament practice. From h7, the queen
prevents White getting in the desired g4
advance, because of the pin resulting from ...
hxg4. That in turn prevents White getting his
pawn to f5, so Black achieves his desired
blockade on the light squares on the kingside.
QUESTION: So White's kingside counterplay
is stopped?
ANSWER: Yes, and meanwhile, his
weaknesses on the queenside remain. Larsen
goes so far as to suggest that White may
already be lost, and he is probably correct.

17.a4

f5

[ Larsen's 17...a5! is probably more

accurate.
QUESTION: So why didn't Nimzowitsch play
this? After all, he does it on the very next
move.
ANSWER: It is another case of the wrong
strategic move resulting from a tactical
miscalculation. Nimzowitsch thought he
was threatening something on the kingside,
and that ...a5 could therefore wait, but it is
not clear he was correct. ]

18.g3 Nimzowitsch's opponent takes him at

his word, and assumes he needs to attend to
the threat of ...Ng4+, but it is probably not the
case.

[ Larsen, endorsed by Kasparov,

recommends 18.a5 , with the point that after

g4+ ( Kasparov points out that this is not
forced and that Black can just play 18...

d7

, but then "White's position is better than in
the game" ) 19.

xg4 ( 19.hxg4? hxg4+

20.

g1 g3 was Nimzowitsch's idea )

19...hxg4 20.axb6 "was a better chance

than [White's position] deserved" (Larsen).
The line continues gxh3 21.gxh3

h4

22.

g3 "and it is not so easy to prove a

clear win for Black" (Larsen). ]

18...a5! Returning to the correct path.

QUESTION: But surely this move leaves a
weak backward pawn on b6? Why are you so
keen to play this?
ANSWER: Because the b6-pawn is much
easier to defend than White's a4-pawn. This
has been shown many times, the most
famous probably being Game 5 of the
Spassky-Fischer match in 1972, which arose
from this same Hübner Variation of the Nimzo-
Indian.

19.

g1 h6 20.f1 d7 21.c1 ac8

With the kingside counterplay under control,
Black has to decide how to progress.
QUESTION: It is not clear how he can really
do much on the queenside. With the position
blocked, White can surely defend the a4- and
c4-pawns adequately, can't he?
ANSWER: He can, and in fact, as we see,
Black's plan is eventually to take over the
attack on the kingside, with the ...g7-g5 break.
But first, he puts paid to any possible
counterplay in the centre, by inducing White
to close the position with d4-d5.
QUESTION: This is all getting a bit confusing.
I thought the whole of Black's play hitherto
had been based around getting the advantage
on the queenside, and just making sure there
was no counterplay on the king's wing. Yet
now it seems the opposite is the case – Black
blocks the queenside and centre, preparatory
to attacking on the kingside?
ANSWER: It may seem a bit inconsistent, but
this is what frequently happens in chess. One
weakness is rarely enough to win, so creating
a second weakness and switching the
pressure between the two is standard
procedure. We will see this too in his game
against Mannheimer.
Nimzowitsch explains the present example in
typical (for him) language: "Every restraint
action is the logical prelude to an attack and
[...] every immobile complex tends to be a

weakness and therefore must sooner or later
become an object of attack". That is what has
happened here with the white kingside – what
started off as a primarily defensive restraint
and blockade action by Black forms the basis
of a later attack on the blockaded sector.

22.d5 A bit cooperative, but Black could force

it anyway, by playing ...Be6, threatening to
exchange on d4 and win the c4-pawn. Now
Black's hands are freed to build up his
kingside attack.

h8 23.d2 g8 24.g2 g5 25.f1 g7
All very logical and, for once with Nimzowitsch,
easy to understand. Black systematically
builds up the pressure down the g-file, in
readiness for the decisive breakthrough.

26.

a2 f5 27.h1 cg8 28.d1 gxf4

29.exf4

c8! Preparing to transfer the bishop

to a6. The finish requires careful calculation
and tactical control, as White starts to obtain
some counterchances, by attacking the b6-
pawn, for example.

30.

b3 a6 31.e2

[ 31.

xb6? loses at once, since after xc4

, Black plans to take on f1 and g3, whilst

32.

ag2? allows mate in two by g4+ . ]

[ QUESTION: Can White try 31.

d2

, with the idea of bringing the bishop to e1,
so as to defend the g3-pawn?
ANSWER: Nimzowitsch mentions that, and
demonstrates a fine combination in
refutation:

g6! 32.e1 g4+! 33.hxg4

( or 33.

g2 xc4! 34.xc4 ge3+ )

33...hxg4+ 34.

g2 xc4! 35.xc4

, and now the quiet move e3!! and there is
no defence to ...Qh3 mate, except 36.

xe3

whereupon

xe3+ wins the queen – the

point of Black's 34th move. ]

31...

h4! 32.e3

[ Nimzowitsch shows another delightful

variation, after the expected 32.

d2!

, trying to create counterplay by attacking
the e4-pawn:

c8 33.xe4 f5 34.f2 .

EXERCISE: Can you spot what it is?
ANSWER: The lovely finish is

xh3+!

35.

xh3 g4# . ]

32...

c8 33.c2 xh3! The sacrifices come

in every variation.

34.

xe4

[ 34.

xh3 allows a forced mate with f5+

35.

h2 ( or 35.g4 xg4 and it is mate in

six ) 35...

g4+ 36.h3 f2+ 37.xh4

h3# . ]

34...

f5 35.xf5 xf5 36.e2 h4

Finally opening the h-file decisively.

37.

gg2 hxg3+ 38.g1 h3 39.e3 h4

40.

f1 e8!

[ A very nice finish. The subtle 40...

e8

creates the irresistible threat of 41.--

xg2

42.

xg2 h1+ 43.e2 xg2+ . Now we

see the point of 40...Re8!.
Nimzowitsch's immortal blockade game, and
still highly instructive to this day. ]

0-1

E21

Matisson,H
Nimzowitsch,A

Karlsbad

1929

[Steve Giddins]

1.d4

f6 2.c4 e6 3.c3 b4 4.f3

[ A classic modern example of a similar

sacrifice to that which we'll see in the notes
to White's 7th, below, is the game Portisch-
Timman, Wijk aan Zee 1978: 4.e3 c5

5.

d3 c6 6.f3 xc3+ 7.bxc3 d6 8.0-0

e5 9.

d2 . After cxd4?! 10.cxd4 exd4

11.exd4

xd4 12.e1+ e6 13.a3 0-0

14.

b3 d7 15.e3 Portisch went on to

win in crushing style. People argue over
the greater understanding allegedly
possessed by modern masters, but I am
sure you would not have got Nimzowitsch to
take that pawn on d4 in a month of
Sundays. ]

4...

xc3+ QUESTION: That looks a bit

extreme. Why doesn't he wait for White to
provoke the exchange with a3?
ANSWER: Such was Nimzowitsch's love of
positions in which his opponent had doubled
pawns that he would usually seize the first
opportunity of making this exchange, without
worrying about whether he was losing a tempo
by not waiting for a3. This approach is less
common nowadays, although it is preserved in
the so-called 'Hübner System', as discussed in
the previous game. In fact, Nimzowitsch's
fondness for the exchange of bishop for knight
on c3 even led him several times to adopt the
exchange after 4 Qc2 Bxc3+, but that really is
going a bit far.

5.bxc3 d6 6.

c2 e7 With this move,

Nimzowitsch prepares the central advance ...
e6-e5.

[ QUESTION: I am a bit surprised. I thought

he would play 6...c5 , to fix the doubled
pawns?
ANSWER: Well, as we will see, he does end

up doing that, both in this game and in
others. But as a general rule, he would
prefer to keep the c-pawn back and play ...
e5 if he could. The idea is to induce the
reply d4-d5, when Black would have a
potential knight outpost on c5, as well as
the possibility of breaking open the c-file by
...c6, to attack White's doubled pawns.
The following game against Bogoljubow is
an example of this preferred strategy in
action. ]

7.

a3?! With this move, White prevents the

intended 7...e5, because of the pin on the d6-
pawn. Nimzowitsch therefore switches plans
and does now fix the doubled white c-pawns.
QUESTION: But isn't that a concession? After
all you were saying he preferred to avoid
playing ...c5.
ANSWER: Yes, but here the advance ...c5
has the added benefit of killing the activity of
the white bishop on a3, so I am sure
Nimzowitsch was quite happy with his position.

[ Nimzowitsch recommended instead 7.e4
e5 8.

d2! c5 ( or 8...c6 whereupon

Nimzowitsch gives 9.

b2 intending f3, Be2

and 0-0 ) 9.

d3 c6 and now his

recommendation is 10.

b3! , ( rather than

blocking the position with 10.d5 .

) ANSWER: It sacrifices one, yes.

Nimzowitsch gives 10...exd4 11.cxd4

b4

12.

e2 xd3+ 13.xd3 xe4+ 14.xe4+

xe4 15.f3 f6 16.f4 "without an
advantage for Black". It is interesting that
he should have appreciated how White is
often better off in such structures
sacrificing material to open the position,
rather than keeping material equality at the
cost of blocking things up. This is
something which modern masters
understand very well. ]

7...c5 8.g3?! QUESTION: Why don't you like

this?
ANSWER: Because it leaves the c4-pawn too
vulnerable.

[ It would be better to play 8.e3 and put the

bishop on e2 or d3, from where it keeps a
weather eye on the vulnerable pawn. ]

8...b6 9.

g2 b7 10.0-0 0-0 11.h4

QUESTION: Is this good? It looks a bit
strange to exchange off White's fianchettoed
bishop.
ANSWER: The exchange of bishops itself is
fair enough, since, as Nimzowitsch himself
pointed out in his notes in 'Kagans Neueste
Schachnachrichten' at the time (once again, I

am indebted to Reinhardt's book for these
notes), "the terrain along the g2-b7 diagonal
needs clarification". However, the choice of
h4 was probably not good, as the knight has
no prospects there.

[ Nimzowitsch recommended instead 11.

d2

, and it is worth giving the whole of his
variation, because it again shows how he
was aware of the need for White to be
willing to sacrifice a pawn, in order to
undouble his pawns and seize the initiative:

xg2 12.xg2 c6 13.e4! (Nimzowitsch
stresses that it is only this pawn sacrifice
that justifies the move 11 Nd2) cxd4

14.cxd4

xd4 15.d3 e5 16.f4 c7

17.

b3 "with an attack". On this occasion,

though, the computer is unimpressed, and
considers Black to be much better after

fe8

. Indeed, it is hard to see a satisfactory
solution to White's problems with his weak
pawns on e4 and c4.
QUESTION: So are you saying Nimzowitsch
had it wrong?
ANSWER: Not really. Although this
particular variation may not work, this does
not invalidate the basic correctness of
Nimzowitsch's understanding of the position
– playing passively is hopeless in such a
structure, and White needs to find some way
to activate his game, even if it costs material.
In this particular instance, though, White's
earlier inaccuracies mean that it may already
be difficult to find a totally satisfactory
continuation for him. ]

11...

xg2 QUESTION: Doesn't this bring the

white knight back into play?
ANSWER: It does, and for that reason,
Nimzowitsch explained that he was reluctant
to make the exchange on g2.

[ However, after 11...

c6 he did not like the

line 12.e4 cxd4 13.cxd4

xd4 14.d3 e5

15.f4 when White again seeks some

initiative at the cost of a pawn. Once again,
the computer is not wholly convinced, but we
can by now understand the way
Nimzowitsch's mind was working in this
game – he wanted to leave White saddled
with his weak c-pawns and try to exploit
these, without allowing any counterplay. ]

12.

xg2? A strange choice.

[ White should, of course, have recaptured
12.

xg2 , thus bringing the knight nearer to

the e3-square, from which it can defend the
c4-pawn. Nimzowitsch then gives another
characteristic note:

c6 13.e4 ("always this

same point")

a5 ( "but not 13...cxd4

14.cxd4

xd4 15.d3 e5 16.f4 " ) 14.e3

"and White has completely consolidated". It
is true that Black has no real advantage in
the final position of this line, but once again,
Nimzowitsch overestimated White's
counterplay after the acceptance of the
pawn sacrifice. Black should play the line
ending 16 f4, when the weakness of e4 and
c4 again gives him a clear plus. ]

12...

b7+ The queen is on her way to attack

the c4-pawn.

13.

g1

[ 13.f3? drops a piece to g5 14.

d2 h6 , ]

[ but once again, it is better to re-centralize

the knight with 13.

f3 . ]

13...

a6 QUESTION: It appears Black is

getting what he has been after all along?
ANSWER: Indeed. From now on, it is really all
one-way traffic. White's many inaccuracies
have landed him in a totally prospectless
position, in which he has no real counterplay
at all, while Black just piles up on the
weakness at c4.

14.

b3 c6 15.fd1 a5 16.b5 xb5

17.cxb5 QUESTION: Oh, wait a minute!

White seems to have undoubled his weak
pawns! Why did Black allow that?
ANSWER: Nimzowitsch realized that, in actual
fact, the undoubling of the c-pawns does not
relieve White's position one bit. As he put it in
his own notes, "The weakness of the c4-
square itself, rather than the pawn sitting on
that square, suffices for the win." A key factor
is that the white bishop will be driven to the
back rank, thereby disrupting the
communication between the rooks, which will
prove fatal.

c4 18.c1 a6 Here is the rub. Because of
his undefended rook on a1, White has to
exchange pawns on a6, after which his a-
pawn soon drops off.

19.bxa6

xa6 20.dxc5 bxc5 21.g2

The knight belatedly hurries back from its
junket to the edge of the board, but in its
absence, its house has crumbled to dust.

d5 22.d3 fa8 23.e4 e5

[ The final position is a picture. White

resigned, although at this moment he has
material equality. Sadly, this state is about
to end, as he is losing a2, c3 and probably
e4 as well. Nimzowitsch gives the variation

23...

e5 24.d1 xc3 25.f1 (the only

safe square for the rook)

xa2 26.xa2

f3+! (bottling White up totally) 27.h1

xa2 and the e4-pawn is next on the
menu.
A remarkable crush of a player who was no
fool, by any means. Sometimes
Nimzowitsch's strategic ideas were so far
ahead of their time that even the best
masters of his day could be made to look
ridiculous, as here. ]

0-1

E21

Bogoljubow,E
Nimzowitsch,A

Karlsbad

1929

[Steve Giddins]

1.d4

f6 2.c4 e6 3.c3 b4 4.f3

xc3+ We have already discussed this move
in the context of the previous game.
Nimzowitsch loses no time in inflicting
doubled pawns on his opponent.

5.bxc3 b6 6.g3?! Also as in the Mattison

game, and not ideal, for the reasons
explained there.

[ Keene's suggestion of 6.

g5 followed by

e3, Qc2 and Bd3, is much more dynamic. ]

6...

b7 7.g2 0-0 8.0-0 e8!? QUESTION:

Well, there is a funny move! I suppose this is
what people mean by Nimzowitsch's
"mysterious rook moves"?
ANSWER: Absolutely. Typically, Nimzowitsch
himself awards the move a rather hyperbolic
two exclamation marks.
QUESTION: So what is the point?

[ ANSWER: Nimzowitsch describes it as a

case of "prophylaxis and centralization". The
idea with 8...

e8 is to anticipate 9.d2

xg2 10.xg2 , after which there would
follow e5 and now 11.e4? is not possible
because of exd4 12.cxd4

xe4 . ]

[ QUESTION: Wouldn't 8...d6 be a sensible

move?
ANSWER: The trouble with this is that White
has the option of 9.d5 , with the idea exd5

10.

h4 . Black can try to hang on to his

extra pawn by c6 , but 11.cxd5

xd5

12.c4 gives White excellent compensation,

thanks to his two strong bishops and the
open diagonals towards the black king. ]

[ For this reason, I prefer 8...

e4 , which is a

typical way of avoiding the d4-d5 trick. ]

9.

e1 d6 10.c2?! This threatens 11 e4,

and he hopes to profit from luring the black
bishop to e4, but it does not work.

[ Instead, 10.

d2 now seems the logical

choice, since after

xg2 11.xg2 e5 12.e4

the e4-pawn is adequately defended after
the exchange on d4. Nimzowitsch continues
the variation with

c6 (trying to induce

White to close the position with 13 d5?,
which would cripple his pawns) 13.

b2

"and Black has no clear method of
attacking White's pawn front". However, as
Keene emphasizes, even this is not terribly
attractive for White, since his bishop is
very passive. ]

[ QUESTION: So what should White play?

Surely he can't be worse already?
ANSWER: Probably not, although, as we
have already seen, these doubled-pawn
structures can turn against White very
quickly. I am inclined to think he should
accept that he already has no chance of
any advantage, and should adopt the
simplifying line 10.

g5 h6 11.xf6 xf6

12.

d2 xg2 13.xg2 e5 14.e3

, when the position is just rather equal,
notwithstanding the doubled pawns.
QUESTION: Why is it equal?
ANSWER: White has offloaded his
potentially bad bishop(s), and avoided fixing
the pawn structure with d5. Black does not
have any obvious knight outposts and the
pawn structure remains fluid – a break with
c4-c5 is always in the air, for example. White
is OK here. ]

10...

e4! 11.b3 c6 12.f1 QUESTION:

What is White up to now? Setting the pieces
up ready for the next game?
ANSWER: This is part of White's regrouping
plan, starting with 10 Qc2. He intends to drive
away the bishop by Nd2, and then advance in
the centre, and he wishes to avoid the
exchange of bishops on g2. However, a
timely central counter by Black prevents this
plan coming to fruition.

e5! This break interrupts White's dream of

Nd2 and e4. Now the pressure against d4
means that plan is impossible, and he is
forced into a decision about his d4-pawn
(which is threatened by 13...Bxf3, followed by
14...exd4).

[ QUESTION: But why doesn't White's last

move invite 12...

xf3 , doubling the white

pawns?
ANSWER: This exchange on f3 is rarely
good in these structures. After 13.exf3
, it is true that the f-pawns are doubled, but
the e-file is half-open for White's rook, and

Black is also weak on the light squares,
especially the long diagonal. White has
two increasingly active bishops, and will
follow up with f4, put his bishop on g2, and
even later aim for the break f5. White would
be better. ]

13.dxe5?!

[ 13.d5?

a5 is even worse, as Black can

use the c5-square for his other knight (via ...
Nd7-c5) and the ossified central pawns do
the white bishops no favours either. ]

[ Ideally, White would prefer to defend his d4-

pawn, but this not easy. 13.

e3? runs into

g4 , ]

[ so he should probably prefer 13.

d1

, although it is hardly attractive, and shows
up the total bankruptcy of the idea started
with 10 Qc2. ]

13...

xe5 14.xe5 xe5 QUESTION: So,

the central pawn position has been clarified.
What is the upshot?
ANSWER: It is not good news for White. His
doubled c-pawns are now isolated as well,
and there is a potentially tempting knight
outpost on c5 for Black.
QUESTION: But White does have two
bishops.
ANSWER: That is true, but they are not easy
to use actively. In particular, the dark-squared
bishop rather bites on granite, in the shape of
the rock-solid black pawn formation c7-d6,
and the white pawn on c3. Overall, this is a
depressing position for White, and the
subsequent play suggests that Bogoljubow
was indeed quite depressed with his game.

15.

f4 e8 16.f3 b7 17.ad1 Now 18 c5

is a threat, but naturally, Nimzowitsch is not
going to fall for that.

d7 18.e4 f6 QUESTION: Is this really so
bad for White?
He looks to have a solid centre and two
bishops.
ANSWER: Yes, but as we have said, the
bishops are not really very active, and the
'solid' centre is not actually going anywhere.
As we quoted Nimzowitsch saying in the
game against Johner, every blockaded
structure sooner or later becomes a target,
and in this case, that is true of White's centre
– Black will eventually attack the e4-pawn by
the break ...f5 (what Nimzowitsch called 'a
rolling-up operation'). The real key to
assessing the position is to ask yourself what
White can actually do with his position. The
answer is nothing. He has no pawn breaks,

and no weaknesses in the black position to
attack. Black, on the other hand, has the
weak enemy queenside pawns to target and a
potential pawn break with ...f5.

19.

g2 e5 20.d2 e7 21.ed1 c6

QUESTION: What is the point of this?
ANSWER: Black just takes prophylactic
measures against any potential c4-c5 break,
exploiting the undefended bishop on b7.
QUESTION: But that was not actually
threatened, was it?
ANSWER: No, but that is the difference
between 'defence' and 'prophylaxis': the
former is about meeting specific concrete
threats, whereas the latter is anticipating such
threats and side-stepping them, before they
even come about. If you want a less
pretentious way of thinking about it than is
conveyed by the word 'prophylaxis', then think
of Nunn's Law, LPDO – on b7, the bishop is a
'loose piece', that might be tactically
vulnerable after a c5-break, so Nimzowitsch
just 'tightens' it up. Of course, such
prophylactic/tidying up measures are only
possible because of the lack of white
counterplay.

22.

f2 ae8 23.f1? h6!?

A very interesting moment.

[ White's last move allows a tactic:
23...

xf3+ 24.xf3 g5 and the 'loose rook'

on f3 means that the bishop is pinned. The
white e-pawn would then soon drop off.
QUESTION: So did Nimzowitsch miss this?
ANSWER: No, I don't believe he did for a
moment. I think he just preferred to avoid
even the slight murkiness of this variation,
which does after all weaken the black king
position. Objectively, Black should be
winning here, as the computer confirms,
but Nimzowitsch understood that his position
was strategically winning anyway, and he
preferred to take things slowly and
methodically, rather than engaging in what
Keene describes as 'such brigandage'.
QUESTION: Do you think a top modern GM
would play the same way?
ANSWER: I very much doubt it. Even a
highly correct positional player, such as
Kramnik, would just take on f3, I suspect.
Partly this is because contemporary players
are much more concrete in their approach
than players of past generations. Thanks
to spending so much time working with
computers, modern players do not
generalize by thinking such things as:

"This weakens the black king a bit, it might
be dangerous, so I'll avoid it"; rather, they
will just calculate carefully and decide
whether the danger is real or not, and
proceed accordingly. Past generations
would be much more likely to take the
Nimzowitsch approach. It is one of the main
ways in which the computer has changed
chess. ]

Nimzowitsch's 23rd move is just another
prophylactic, improving move. He makes luft
for his king, and intends now to triple on the e-
file and then carry out his 'rolling-up operation'
with ...f7-f5.

24.

e2 h8 25.a3 e6 26.c1 f5!

[ 26...

xc4?! 27.xc4 xc4 28.xh6

was what White was hoping for, although
even here, f5 is still much better for Black.
However, Nimzowitsch is not going to be
distracted by such nonsense, as we have
already seen – he just carries out his long-
prepared 'rolling-up' of the white central
structure. ]

27.exf5

xf5 28.d2 f7 Going back to the

attack on the c4-pawn – yet again,
Nimzowitsch employs the 'two weaknesses'
approach, alternately pressing first on c4,
then on the kingside, then on c4 again, etc.
Eventually, such switching tactics will
overwhelm the defences.

29.

d4? Capitulation.

[ The "loathsome 29.

xe5 " (Keene) was the

only way to stay on the board, as the
computer confirms, but after

xe5

, White will still not succeed for long in
holding together his bag of weaknesses (it
seems too generous to call it a "position").
A sample variation is 30.

df1 e3 31.c2

h5 32.d2 c5 33.g2 g5!
and the threat of ...g4 decides. ]

29...

g6! Now there is no time to prevent

further catastrophic structural damage,
resulting from the exchange on f4, since
White must defend his other bishop on e2.

30.

d3 xf4 31.xf4 xf4 32.gxf4 f8

The rest is a simple mopping-up job. The
white pawn structure reminds one of
Shakespeare's Richard III's laments of his
bodily deformities:
"I that am curtail'd of this fair proportion,
Cheated of feature by dissembling nature,
Deformed, unfinish'd, sent before my time
into this breathing world, scarce half made up,
And that so lamely and unfashionably,
That dogs do bark at me, as I halt by them."

33.f5

d7 34.dd2 xf5 The first fruits.

35.

fe2 xe2 36.xe2 e8 37.f2 e5

38.

d5 g5 39.xe5 dxe5 40.c5 bxc5

Frankly, White could and should have
resigned, but Bogoljubow may well have been
rather shell-shocked to have been beaten so
one-sidedly. As Keene points out, just 34 days
(!) after the date on which this game was
played, Bogoljubow sat down to start his
world championship match against Alekhine.

41.

a6 e4 42.a4 g7 43.a5 exf3 44.xf3

f6 45.e3 e5 46.c4 g4 47.a6 h5

48.

c4 h4 49.a6 d1 50.b7 g4

0-1

C01

Mannheimer,N
Nimzowitsch,A

Frankfurt

1930

[Steve Giddins]

1.e4 e6 2.d4 d5 3.

c3 b4 4.exd5

QUESTION: Ugh. How boring!
ANSWER: Actually, although the Exchange
variation has a reputation for being boring,
there is plenty of scope for the black player to
play for a win. Both Alekhine and Nimzowitsch
won some instructive games against it,
demonstrating how Black can unbalance the
position if he wishes.
QUESTION: How?
ANSWER: Basically, by exploiting
Grandmaster Suba's profound insight:
"Black's knowledge is always greater by one
move". In other words, the white player, by
dint of having the first move, always has to
commit himself one move before his opponent,
and the latter can fashion his response so as
to take advantage of this slight extra
commitment. As mentioned in the notes to
Levenfish-Nimzowitsch, Nimzowitsch was
especially interested in this issue, and
devoted one chapter of 'Chess Praxis' to the
topic of the "asymmetric treatment of
symmetrical variations".

exd5 5.

d3 e7 Here we see

Nimzowitsch's asymmetrical treatment in
action. As quoted in our opening game, his
rule of thumb was to meet Nf3 with ...Ne7, and
Ne2 with ...Nf6. The fact that White has to
commit first is key to this approach.

6.

f3 bc6 7.h3 f5 This exchange of

light-squared bishops sows the first seeds of
what proves to be the harvest which defines

the game. White already has a very slight
weakening of his light squares, thanks to the
fixed position of the respective d-pawns, and
the exchange of bishops removes one of their
principal defenders.

8.

xf5 xf5 9.0-0?! QUESTION: Why is

this wrong?
ANSWER: The problem is that it allows Black
to implement his plan of doubling the enemy
pawns and weakening the light squares in his
camp.

[ Nimzowitsch gives as an alternative the line
9.

d3 xc3+ 10.xc3 e7+ 11.e3

xe3 12.xe3 xe3+ 13.fxe3 0-0-0
, and claims that White is worse, which must
be true, although it is only one weakness,
and relatively easy to defend. Black's plus
would not be very great. ]

9...

xc3 Nimzowitsch rarely needed a second

invitation to inflict such doubled pawn
complexes on his opponent. However, as he
himself pointed out, White is not yet
significantly worse, since he has counterplay
on the dark squares.

10.bxc3 0-0 11.

d3 d6 12.g5

QUESTION: Isn't this just a crude one-move
threat?
ANSWER: Not really. Nimzowitsch
emphasizes that White needs to create some
activity for his dark-squared bishop, as, if he
does not, then the black knights will settle into
some juicy outposts on c4 and e4, and White
will just be worse. It is similar to what we saw
in Matisson-Nimzowitsch – White must make
something of his plusses, in terms of his
bishop(s) especially, to offset his structural
weaknesses. The text forces Black to weaken
his dark squares, which the bishop on c1
hopes later to exploit.

g6 13.

f4 f6 14.d2 QUESTION: What

about just removing the knight on d6, before it
can reach c4?

[ ANSWER: Nimzowitsch mentions this, and

says he intended 14.

xd6 cxd6! .

QUESTION: But then both sides have a
doubled pawn. Why should Black's be any
less weak than White's?
ANSWER: The difference is that Black's is
not on an open file. There is no effective
way to attack Black's doubled d-pawns,
whereas Black can easily direct his heavy
pieces against the white doubleton on the c-
file, as well as aiming his remaining knight
at the c4-outpost.
QUESTION: So, rather like George Orwell's

animals, some doubled pawns are more
equal than others?
ANSWER: Well, a weakness is only a
weakness if it can be attacked. White's
doubleton can, Black's can't. It is as simple
as that. ]

14...h6?! The start of a typical piece of

Nimzowitschian prophylaxis, but in view of the
white improvement at move 17, it may be a
little too clever.

[ After the obvious 14...

fe8 , Nimzowitsch

says he was afraid of 15.f4 "with an attack",
but the fearless computer is not fazed by
this, and believes Black is better after simply

a5 16.f5 gxf5 . We saw in the Mattison
game that Nimzowitsch tended to take a
very 'pure' view of these doubled pawn
positions, in that he liked to win without
allowing any kind of murky counterplay,
even where that counterplay involved what
may objectively be an unsound sacrifice by
the opponent. This may be another example
– the fact is that one cannot avoid such
murky positions, if the opponent resolutely
refuses to be pushed around. ]

15.

f3 h7 16.h2! It looks as though

White's plan is working. He has provoked a
significant weakening of the dark squares on
the enemy kingside, and now threatens Ng4.
What is Black to do?

h8!? QUESTION: Good grief. I wouldn't
have thought of that!
ANSWER: It is in many ways a typical
Nimzowitsch move. Although he always
argued bitterly with Tarrasch about the
appearance of a move being unimportant (c.f.
his famous quote that "the beauty of a chess
move lies not in its appearance, but in the
thought behind it"), one can easily sense from
his annotations that he actually loved playing
such paradoxical and visually surprising
moves.
QUESTION: So what is the point?

[ ANSWER: 16...

h8 is just prophylaxis

against the threat of 17.

g4 , which would

now be met by h5 . ]

17.

e3?!

[ Nimzowitsch was obviously very proud of

his 16th move, and awarded it an
exclamation mark in his notes, but it is not
clear if it was really so great. The computer
points out the surprising reply 17.

f4!

when Black appears to have some problems,
due to his uncoordinated pieces. For
example,

ac8 ( defending the c7-pawn,

and so planning ...Nc4; 17...

f6?

does not even force a repetition, because
White has 18.

xh6! ) 18.g4 h5 19.e3

and now the white knight is ideally placed,
attacking the d5-pawn and covering c4.
Black has serious problems after

e7

20.

fb1 b6 21.e5 f6 22.xd6 cxd6

23.a4 , with a strong initiative. ]

17...

g7 Now Black manages to neutralize his

opponent's initiative, as he hoped, and the
white weaknesses gradually come to the fore.

18.

f3

[ 18.

g4 h5 leads to nothing for White. ]

18...

e4 19.c1 f5 20.d3 a5

QUESTION: It is starting to look ominous for
White. Those black knights are ready to settle
into the two big central outposts!
ANSWER: Yes, and this leads White to a
highly committal decision.

21.f4? QUESTION: Hmm – doesn't do much

for the bishop on c1, does it?
ANSWER: Indeed. White's idea is to secure
the e5-square for his knight, and for that
purpose, he is prepared to trade away the e4-
square to Black. The trouble is, as the late
Tony Miles once remarked of a similar
situation, the possession of one good square
on e5 does not really make up for White's
problems on the other 63. Besides, e5 is still
not even that secure, since, as Nimzowitsch
pointed out, it can be undermined by a later ...
g5.

[ Having said all that, it is extremely hard to

suggest anything really constructive for
White. He can mark time with something like

21.

b1 b6 22.f3 , for example, which at

least does not create any new weaknesses,
but his position remains extremely
unpleasant, and Black has a free hand to
decide how to strengthen his game. ]

21...

d7 22.f3 c6?! Rather pointless,

since, as the next note shows, c3 is not really
attacked.

[ More accurate was 22...

a4 23.e5 f6

, and Black is fully in control. ]

23.

e5 e6

[ 23...

xc3?? 24.xc3 xc3 25.d2

wins for White. ]

24.

b1 b6 Black's inaccuracy has slightly

discoordinated his pieces, but this is really
only temporary, since White cannot change
the basic assessment of the position. Black
just needs a bit more time to reorganize.

25.

h2 c4 26.e3 QUESTION: What

about exchanging on c4? At least it gets rid of

the huge enemy knight.

[ ANSWER: Yes, but it also gets rid of

White's only well-placed piece. That in turn
emphasizes the downside of the move 21
f4? – in a sense, the latter move commits
White to never exchanging his knight off, if
he can avoid it, because that would deprive
him of the only benefit of 21 f4 (namely, the
outpost on e5 for his knight), and leave all
the drawbacks, such as the bad bishop and
weakness on the e-file. After 26.

xc4 dxc4

White's position is terrible. ]

26...g5 QUESTION: Isn't this a bit risky?

ANSWER: It shouldn't be, providing Black is
careful. He has such control of the position,
and the white pieces are so inactive, that
there should be no danger from the
weakening of the black king position. Even so,
Black need not have hurried with this break,
which cannot in any event be prevented, so
he could first have brought his queen's rook
into the centre and regrouped further.

27.g3

f6 28.be1 g8 QUESTION: So, it

seems that Black is preparing to break
through on the g-file?
ANSWER: That is certainly his plan, yes.
QUESTION: So does that mean he has
forgotten about the weaknesses on the white
queenside?
ANSWER: Funny you should ask that...

29.

c1 b5! Here is the answer to your last

question! By no means has Nimzowitsch
forgotten about the queenside. His last move
introduces the threat of ...Qa6, winning the a2-
pawn.
QUESTION: OK, but isn't this just
inconsistent? I thought one was supposed to
form a plan and stick to it. Here, he first
attacks the queenside, then the kingside, and
now the queenside again.
ANSWER: Ah, well, now you see the
difference between a weak player and a
master – when the weak player does such a
thing, it is 'inconsistent dithering', but when
the master does it, it is 'a classic two-
weaknesses strategy'. Seriously, though, the
books tend to overdo the 'form your plan and
stick to it' routine. In reality, a single
weakness is rarely enough to lose a game,
and switching the attack from one weakness
to another, exploiting the military concept of
the 'second front', is usually a key element in
winning a game. The point here is that Black
cannot break in down the g-file, all the while
White is in a position to meet ...gxf4 by

recapturing on f4 with the bishop, defending
g3. On the other hand, once the bishop goes
to c1, it blocks its rooks from reaching a1, so
the a-pawn is now vulnerable.

30.

f3? This hastens the end, by allowing

Black to open the g-file in another way.

[ Nimzowitsch recommended waiting tactics

with 30.

e2 , after which, rather than going

after the a-pawn straightaway, he would
have continued his policy of threatening
both flanks, by regrouping his rooks with

ff8 , followed by ...Rg7 and ...Rfg8. The
white a-pawn is going nowhere anyway. ]

30...g4 31.hxg4

xg4! This is the difference.

Now the g-file is opened and Black can pile up
on g3.

32.

g1 fg6 33.f3 g8 34.e2

The knight has finally had to abandon any
hopes of getting back to its beloved e5 outpost,
further underlining the bankruptcy of his 21st
move.
QUESTION: OK, but Black seems to have
brought the maximum pressure possible
against g3, and White is still hanging on. How
do we finally break through?
ANSWER: The camel's back is strained to the
maximum – we just need one little straw...

h5! And here it is! The h-pawn supplies the

final touch. There is no defence to ...h4,
bringing the white barricades on the g-file
crashing down.

35.

g2 h4 36.h1 h6 37.h3 g6!

This is more than just a waiting move.

[ 37...

g6 creates the threat of 38.-- hxg3

39.

xh6+ xh6 40.xg3 h4 and White

is in zugzwang. After 41.a3 a6 , he has no
move to avoid the loss of material. A
typical Nimzowitsch finish. White avoids this
by bringing his bishop round within reach of
f2, but then Black makes one final switch to
the queenside. ]

38.

e3 a6! QUESTION: Ha! So the a-pawn

does finally drop off after all?
ANSWER: Yes. The manner in which
Nimzowitsch switched threats between
kingside and queenside is absolutely classic,
and makes this game immensely instructive.

39.

f2 xa2 40.e1 a5 Now the passed

pawn decides.

41.

f1 b1 42.g1 a4 43.e2 a3 44.f1

a2

0-1

Chapter Four

Nimzowitsch the Strategist

As might be expected, this is the longest chapter of the book, and features a series of wonderful games, embodying
various aspects of strategy.

Nimzowitsch-Taubenhaus

is an interesting take on the perennial problem of the IQP, in which Nimzowitsch shows

the possibility of exploiting the possession of the pawn to attack not the enemy king, but his queenside - a
neglected aspect of such positions.

Nimzowitsch-Spielmann (Stockholm 1920)

is another of his anti-French

Defence games. This time, in contrast to the earlier games against Haakanson and Salwe, he succeeds in
maintaining his pawn centre on d4 and e5, and shows how this space advantage can be a long-term strategic asset.

In an interesting interview with Ray Keene, conducted exclusively for the latter's book on Nimzowitsch, Bent Larsen
identified play on squares of one colour as a quintessentially Nimzowitschian device, and

Consultants-Nimzowitsch

is a typical example. As early as move 6, Nimzowitsch makes a seemingly senseless exchange of his queen's
bishop for an undeveloped white knight on b1, which proves to be the basis of a plan of controlling the central
square d5, and which in turn leads to a long-term plan of a kingside pawn storm.

Nimzowitsch-Brinckmann (Kolding 1923)

is a classic exploitation of the queenside pawn majority, whilst

Sämisch-Nimzowitsch

is perhaps Nimzowitsch's single most famous game, the 'immortal zugzwang' game.

Then

Yates-Nimzowitsch

sees an instructive exploitation of an open file, and

Nimzowitsch-Rubinstein (Dresden

1926)

is another famous one, involving a classic knight retreat to the corner, although, as the notes show, it is

nothing like as convincing as most previous annotators have suggested.

D40

Nimzowitsch,A
Taubenhaus,J

St Petersburg

1913

[Steve Giddins]

1.d4 d5 2.

f3 f6 3.c4 e6 4.e3

A very modest move, but Nimzowitsch was
interested in the kind of symmetrical Tarrasch
position that arises in this line. This is a game
which is interesting for its middlegame, rather
than its opening.

c5 5.

d3 c6 6.0-0 dxc4 7.xc4 cxd4

8.exd4

e7 9.c3 0-0 So, we have a typical

IQP structure. Nimzowitsch was very
interested in this topic, devoting an entire
chapter of 'My System' to the subject, and
describing it as "one of the cardinal problems
of the whole theory of positional play".
QUESTION: Which side was he on?
ANSWER: In 'My System', he gives a
relatively objective presentation of the subject,
and advises the student to play both sides, so
as to get a feel for the strengths and
weaknesses of each side's position. But it is
true that, as his career progressed, he did not
so often reach the white side (i.e. the side
with the IQP) of such positions, generally
preferring to play against the IQP. Naturally,
this was another facet of his rivalry with his
great foe Tarrasch, who had unbounded faith
in the IQP.

10.

e3 QUESTION: That looks a rather tame

approach by modern standards, doesn't it?
ANSWER: Exactly right. The modern
understanding is that White should play more
dynamically. One would expect White's bishop
to go to g5 in such positions, and his rooks to
come to e1 and d1. In general, White attempts
to exploit the dynamic strength of the IQP, to
attack the black king. However, in 'My System',
Nimzowitsch spoke of this subject, and
commented that: "It seems to me that there is
no objective motive for plunging on a
desperate attack. At first, the utmost solidity is
called for." To this end, he advised a set-up
with the bishop on e3, defending the IQP, and
the rooks on c1 and d1.
QUESTION: So was he abandoning the idea
of a kingside attack?
ANSWER: No, but he felt it should be delayed
until Black weakens his king position, which in
practice, Nimzowitsch felt would usually
happen. For example, the f6-knight will
usually move to d5, occupying the blockade

square, in which case White is more justified
in trying to attack the king.
QUESTION: Do you agree with him?
ANSWER: Well, it is fair to say that modern
players do not generally share his view, and
tend to regard the direct attacking plan as
more justified, in most cases. Of course,
everything depends on the specifics of the
position, but in general, the modern approach
is more dynamic than Nimzowitsch's
preferred treatment.

b6 11.

e2 b7 12.fd1 Following the

scheme outlined in 'My System'.

b4 13.e5 QUESTION: But now he starts
to attack!
ANSWER: Well, not perhaps attack, but
certainly he makes a forward move. This is a
good example of what he meant by squares
often becoming available without one needing
to make any attempt to conquer them. The
outpost at e5 is one of White's principal
assets in IQP structures, but he frequently
needs to do no more than wait for it to fall into
his hands – the black queen's knight wants to
reach d5, and so e5 becomes available to
White without his having to do anything to get
it.

c8 14.ac1 bd5?! This looks extremely
natural, but runs into some problems, thanks
to Nimzowitsch's surprising reply.

[ The alternative 14...

fd5 was probably

better, although it is rather inconsistent
with Black's 12th move, which clearly
intended to put the queen's knight on d5. ]

15.

b5! QUESTION: This looks strange!

What is the idea?
ANSWER: Nimzowitsch explains this move by
a desire to avoid the standard black plan of
exchanging knights on c3 and then, after the
recapture bxc3, attacking the backward c-
pawn. However, to my mind it is much more a
case of tactical opportunism. Nimzowitsch
has simply seen that the attack on the
undefended black a-pawn is not as easy to
meet as one might think.

a6!?

[ Certainly the most natural response, but

after some thought, the computer prefers
the retrograde 15...

a8 . This is probably

objectively true, but it is a hard move for a
human to play. ]

16.

a7! QUESTION: Wow! This looks pretty

risky. Is the knight coming out alive?
ANSWER: This is what I mean by it being
tactical opportunism. Nimzowitsch has seen

that his knight is OK. The attack on the a6-
pawn means that Black's next is forced, and
the white knights gain access to the c6-
square.

a8 17.ac6 d6 18.xe7+ xe7
QUESTION: So, White has won the bishop-
pair. Presumably, this gives him the
advantage?
ANSWER: Well, White is somewhat better in
this position, yes, but it does not follow from
this that the bishop-pair is always a plus in
such IQP structures. These positions rather
suit knights, as a matter of fact, because the
pawn structure creates various decent
outposts for knights, such as e5 for White, and
d5 for Black.
QUESTION: So why is White better here?
ANSWER: It is really because of the
weakness of Black's queenside pawns. The
attack on the a6-pawn ties down his rook, and,
as we will see, advancing the pawn to b5
creates a new weakness at c5.

19.

d3?! QUESTION: Not preserving his

bishop-pair?

[ ANSWER: No. Nimzowitsch gave his move

an exclamation mark, but the computer
actually does want to keep the two bishops
here, and recommends 19.

g5 . That does

seem quite nice for White, but Nimzowitsch
had a different idea in mind. ]

19...

xe3 20.fxe3! This was Nimzowitsch's

plan.
QUESTION: He wants to get rid of his IQP?
ANSWER: Yes. He comments that: "The
centre position is now very strong". He
certainly thought the position better for White,
although that is not entirely clear.

b5?! QUESTION: You don't like this move?

ANSWER: No, not at all. It weakens the c5-
square and puts another pawn on the same
colour square as Black's bishop. Admittedly, it
disposes of the attack on the a6-pawn, but
the price seems rather high to me.
QUESTION: So what should Black do?

[ ANSWER: I think his best move is 20...

d5

. Nimzowitsch did consider this, and gave
the reply 21.

e4 .

A) Now he claims the tactical justification

is that if 21...

c3 22.xc3

( but Nimzowitsch seems to have

overlooked the simple pawn gain by the
desperado 22.

xh7+! ) 22...xe4 23.dc1

"and White has the advantage".;

B) Instead, Black should answer 21...

ac8

(taking advantage of the fact that White's

last move relieved the pressure against
a6), when he seems completely equal.
This all suggests that Nimzowitsch's plan
of allowing the exchange on e3 was not
the strongest, and he would have done
better to play 19 Bg5.;

C) 21...-- ]

21.

c5 fc8 22.dc1 g6

[ QUESTION: How would White recapture

after 22...

xc5 ?

ANSWER: That is a good question.
Nimzowitsch does not indicate what he
would have played. 23.dxc5 gives White a
passed pawn ( 23.

xc5 is probably the

move to play, when White retains a small
advantage after, say,

c8 24.b4 )

, but after 23...

d5 24.b3 e4 it is not

really going anywhere, and I am not sure
White has anything. ]

23.a3

e8?! The knight heads to d6, but now

Black starts to find his difficulties mounting.

24.b4

d6 25.f2 QUESTION: What is the

point of this?
ANSWER: The queen heads to f4, whilst also
exerting pressure against f7, provoking
Black's next move.

f5

[ The active 25...

g5 would be strongly met

by 26.

d7 , with threats to jump to f6 or

b6. ]

26.

f4 e8

[ Black eliminates the potentially 'loose'

knight on d6. It looks tempting at first to
seek some counterplay with 26...a5
, but then after simply 27.

xc8+ xc8

28.

xc8+ xc8 29.bxa5 xa3 30.h4!

xa5 31.f6 Black is in trouble. ]

27.

e2! QUESTION: Why is this good?

ANSWER: This is a nice idea, which intends
to exchange bishops. The enemy bishop on
b7 is a key piece in defending the rook at c8,
and once it goes, White conquers the c-file.

d6 28.f3 xc5? This is capitulation.

[ After 28...

xf3 29.gxf3! (taking the e4-

square from Black's knight), White has a
large advantage, but the text is even
worse. ]

29.dxc5! Much the stronger recapture,

opening the d-file for White.

e8

[ Nimzowitsch writes that Black's original

intention had been 29...

e4 , which would

fail to 30.c6 e.g. g5 31.cxb7!

f8 32.c8

( 32.

c6! is also winning ) 32...xb7

33.

xf8+ xf8 34.xe4 and wins. ]

30.

d1 f6

[ Now White penetrates decisively down the

d-file, since 30...

d8? is impossible:

31.

xd8 xd8 32.xb7 , ]

[ or 30...

xf3 31.d7 f6 32.gxf3 d8

33.c6 , which is decisive. ]

31.c6

c8 32.c7 a7 33.d8+ g7

34.

xc8 xc7 35.xg6! A nice example of

what is a relatively uncommon aspect of IQP
positions, namely the use of the c-file and the
c5-outpost, rather than the usual kingside
attack.
1-0

C02

Nimzowitsch,A
Spielmann,R

5th matchgame, Stockholm

1920

[Steve Giddins]

1.e4 e6 2.d4 d5 3.e5 c5 4.

f3 c6 5.c3

b6 6.e2 cxd4 This exchange on d4
became de rigueur in such positions, after
Nimzowitsch's game against Salwe. Black
fixes the white d4-pawn as a potential target
and avoids the exchange on c5.

7.cxd4

h6 8.c3

[ 8.

xh6?! is known to be bad after xb2

, when the queen emerges safely: for
example, 9.

e3 xa1 10.c2 b4+

11.

f1 d7 and the black rook comes

quickly to c8. ]

8...

f5 9.a4 The only way to defend the d4-

pawn satisfactorily.
QUESTION: It looks as though White is
having to resort to some contortions to do so,
though. Isn't this a sign of trouble?
ANSWER: Well, it all depends on whether he
is able in fact to consolidate his pawn centre
without serious consequences. If so, then he
will be fine. As it happens, he manages to do
this, although Black has enough play for
equality. This whole line is nowadays one of
the main lines of the Advance French.

a5+ 10.d2 b4 11.c3 d7

[ 11...b5 is the main move these days, after

which 12.a3

xc3+ 13.xc3 b4 14.axb4

xb4 15.b5 d7 16.xc6 xc6 17.0-0

a5 is regarded as equal. ]

12.a3

xc3+ 13.xc3 h5 QUESTION: Is

this good?
ANSWER: It is a double-edged decision.
Black wants to maintain his knight on its
excellent outpost, from where it attacks the

white pawn on d4. However, 13...h5 weakens
Black's kingside notably, especially the g5-
square, and I do not particularly like it. I would
prefer castling short.

14.0-0

c8 QUESTION: How do you assess

this position?
ANSWER: I would say that White is a little
better. He has completed his development
and maintained his pawn centre intact.
QUESTION: But is the pawn centre a serious
advantage?
ANSWER: It confers greater space on White,
and certainly represents a plus, although it is
true that the d4-pawn remains a potential
weakness. Indeed, Nimzowitsch emphasises
that this pawn is the key problem of the
position. Black hopes later to undermine its
defence by the advance ...g5-g4, driving away
the knight from f3, which White needs to
prevent.
QUESTION: So it all comes down to
prophylaxis again?
ANSWER: Yup!

15.

d2

[ QUESTION: Why didn't he chase the queen

with 15.b4 ?
ANSWER: Nimzowitsch explains that he
avoided this, because he did not wish to
drive the enemy queen to d8, from where
she supports the ...g5-advance.
QUESTION: But surely ...g5 is not a serious
threat is it? What about the black king –
where would it go after such an advance?
ANSWER: Well, the point is that with the
centre blocked, the black king is probably
fairly safe where it is. Such flank advances
with ...g5 are a common feature of many
French variations nowadays, and the
blocked pawn centre lies at the basis of
their viability. ]

15...

d8 Nimzowitsch explains that Black is

now threatening 16...g5, with the tactical point
that d4 falls if White captures on g5. White's
next move is directed against this.

16.h3 QUESTION: Er, how does that stop

16...g5?

[ ANSWER: The point of 16.h3 is to meet g5

with 17.g4 , when Nimzowitsch gives the
variation hxg4 18.hxg4

h4 19.xh4

( 19.

xg5? still allows xd4! which is

unclear at best ) 19...

xh4 , and now,

instead of Nimzowitsch's 20.

g2

planning Rh1, ( the computer points out the
much stronger 20.

b5 , when the knight

penetrates to d6, with a clear advantage. )]

16...

a5 17.ad1 QUESTION: That seems

odd. I thought White would put his rook on
the open c-file.
ANSWER: Well, tactical considerations, of
course, prevent that at the moment (there is a
fork on b3), but more fundamentally,
Nimzowitsch is following his beloved strategy
of 'over-protection', by adding further
coverage of the key d4-pawn.
QUESTION: But isn't the rook passive on d1?
ANSWER: To modern eyes, it is. But
Nimzowitsch had enormous faith in the value
of such pieces, claiming that they derive
strength from being in contact with
"strategically important squares", such as d4
here. It often sounds a bit overdone, but the
essential point is of some validity - the rook
will not sit there defending d4 forever, but just
takes part in a temporary consolidating
manoeuvre, after which White plans gradually
to drive back the active enemy pieces and
then go over to active operations himself,
utilizing the strength of his superior central
control. That, at least, is the theory. The
reality is that the position is actually just equal.

b6 18.fe1

[ Nimzowitsch justifies this by explaining that

it prepares Bd3, which if played at once, with

18.

d3 , would be met by xd4 19.xd4

xd4 20.b5 xe5 (hence the preparatory
18 Rfe1). But this is in fact nonsense, since
the black knight hangs on a5, so he is losing
a whole piece here. ]

[ QUESTION: So should he have played
18.

d3 at once?

ANSWER: Probably not, since after the best
reply

c4 19.xc4 xc4 , White has lost a

tempo over the game. So I think
Nimzowitsch's 18th move was probably best,
although his reasoning was faulty. ]

18...

c4 19.xc4 xc4 20.e2

QUESTION: More over-protection!
ANSWER: Yes, all according to plan. Now the
next step is to challenge the c-file, the rook on
d1 having been relieved of the need to defend
d4.
QUESTION: But is White better?
ANSWER: No, not at all. Things are still just
equal, although it is clear from his notes that
Nimzowitsch felt he stood better.

a4

[ 20...h4 is another option here, fixing the

kingside and ruling out a possible g4-
advance or the move Ng3, exchanging off
Black's strong knight. ]

21.

c1 b3 22.xc4 xc4 23.g3 e7

[ 23...

xg3 24.fxg3 was perhaps better,

with no real problems. ]

24.h4

g6?! The knight is not very effective

here.

[ 24...

c6 looks like a better way to preserve

the balance, maintaining pressure against
the d4-pawn. ]

25.

f1 xf1 26.xf1 e7 27.c1 0-0

28.b4

f5 QUESTION: How is the position

now? White appears to have made some
progress.
ANSWER: Yes, over the last few moves,
Black has committed a couple of slight
inaccuracies and White now has control of the
c-file. Even so, Black is quite solid, so the
advantage should not be exaggerated, but
White definitely has the initiative.

29.

c5 a6 30.c3 e2 31.c2! xd4?!

This wins a pawn, but the black knight loses a
lot of time, and White's endgame advantage
increases.

[ Nimzowitsch gives the alternative variation
31...

a6 32.a4 f6 33.a5 ( 33.a5

is probably stronger ) 33...

b6 34.c5 d8

( and here 34...

xc5 is probably essential,

although Black is suffering after 35.bxc5 )

35.

xa7 fxe5 36.xe5 , and now xh4?

loses outright to 37.

xf8+! . ]

[ However, the best move is probably
31...

xc2 . Nimzowitsch dismisses this

endgame as bad, because of the invasion
on c7, but after 32.

xc2 e8 33.c7 e7

34.

c8+ h7 Black is still holding, although

it is certainly very passive and unpleasant. ]

32.

xe2 xe2+ 33.f1 f4 34.c7 b5

35.g3

[ Grabbing the pawn back with 35.

xa7?

would cede the open file, when Black is
fine. ]

35...

d3 36.e2 b2 37.xa7 QUESTION:

But now he gives up the c-file anyway?
ANSWER: Yes, but after first activating his
king and driving the enemy knight offside.
Nimzowitsch claims to have calculated the
consequences of his last move very carefully.

c8 38.d4 c4 39.xb5 d4 QUESTION:
So what is going on here? It looks rather
double-edged.
ANSWER: White has won a pawn, and has
two connected passed pawns on the
queenside, but Black has counterplay, thanks
to his combined unit of rook, knight and
passed d-pawn. However, the big difference
between the two sides is that the white king is

playing, whilst the black king is not, and this is
enough to swing the balance to a decisive
advantage for White.

40.

c7 d3+ Exchanging rooks into a knight

ending would be hopeless, so Black has to
play this, but now his rook is driven offside.

41.

e3 g4 42.c1 g5 The best hope for

counterplay.

43.

d6!

[ 43.

b1? c4+ 44.xd3 xe5+ 45.c3

gxh4 46.gxh4

xh4 is far from clear. ]

43...gxh4 44.gxh4

xh4 45.b1 h3+

46.

d2 a4 47.b5 b6 48.b4

Preparing to remove the blockading knight.

f3 49.c4! d7

[ Nimzowitsch indicates one line he had

calculated, as follows: 49...

xc4+ 50.xc4

xf2+ 51.xd3 f3+ 52.c2 xa3

53.

c8+ g7 54.b6 a2+ 55.c3 a3+

56.

c4 a4+ 57.b5 a1 58.b7

, and wins. ]

50.b6

c5 51.b7 xf2+ 52.e3 e2+

53.

d4 xb7 54.xd3! This zwischenzug is

the key to White idea. With the elimination of
the d-pawn, Black's last hopes disappear.

g2 55.xb7 h4 56.e3 g5 57.d4 h3

58.

b2 h5 59.h2 The temporarily passive

rook is not important, as the white a-pawn will
advance decisively.

f6 60.

c4 f7 61.a4

[ Megabase gives several more moves, but

Nimzowitsch's notes say the game was
adjourned here and Spielmann resigned
without resuming. Play might continue 61.a4

g6 62.a5 f5 63.a6 h7 64.c5
, and White wins easily.
QUESTION: Nimzowitsch was apparently
very proud of this game, wasn't he?
ANSWER: Yes. He wrote that he considered
it "one of my finest performances". I am
less convinced, in that I don't think White
had anything much until relatively late in
the game, but Nimzowitsch was always
more concerned with showing the apparent
success of his ideas. In the latter sense, he
felt that his theory of maintaining the d4-e5
pawn chain, and this proving advantageous
in the long term, had been vindicated. ]

1-0

B00

Consultants
Nimzowitsch,A

Uppsala (consultation game)

1921

[Steve Giddins]

1.e4

c6 2.d4 d5 3.e5

[ 3.

c3 is the only serious try for an

advantage here. ]

3...f6

[ 3...

f5 is also perfectly sound, and

probably slightly safer, but Nimzowitsch
always preferred the more pugnacious text. ]

4.

b5 f5 5.f3 d7 6.c4 xb1?!

QUESTION: That is a surprising move! What
is the idea?
ANSWER: Nimzowitsch awarded himself no
fewer than two exclamation marks for this
move. It is typical of the paradoxical moves of
which he was so fond. The idea is to fight for
control of the square d5, on which Black has
hopes of establishing a knight one day. To
this end, he removes the white knight that
could fight for that square.
QUESTION: But he is losing a tempo and
giving up the bishop-pair!
ANSWER: That is true. As to the latter, we
have already seen in this book that
Nimzowitsch was generally not too attached to
the bishop-pair anyway, and tended to prefer
knights over bishops. The tempo issue is
perhaps more of a consideration, but the
central pawn structure is already semi-closed,
and so rapid development is likely to be less
of a consideration than would be the case in
an open position. Nimzowitsch is more
concerned with the permanent structural
features of the position, specifically the
weakness of the white d4-pawn.

7.

xb1 0-0-0 8.cxd5

[ QUESTION: If Black is playing to occupy

the d5-square, doesn't that suggest that
White should play 8.c5 to prevent this?
ANSWER: That is a good question.
Nimzowitsch says Black would continue g5
"followed by ...Bg7". The implication is that
he would have been fully satisfied with
Black's position, but I am less convinced.
Both 9.0-0 ( and 9.b4 look better for White
to me,; although 9.

a4?! is met by f5! . )]

8...

xd5 9.xc6 xc6 10.0-0 e6 11.e3?!

[ White develops the bishop in defence of d4,

which is logical enough in itself, but the
more ambitious 11.

f4! would have caused

Black some problems, thanks to the threat of

Rc1 and exf6, going after the c7-pawn.
Black probably has to play

d7 ( 11...g5

12.

g3 does not help, and only weakens

Black further ) 12.

c1 a6 ( he would

rather put the queen on d5, but 12...

d5

13.

a4! is too strong ) , but then 13.exf6

gxf6 14.

c2 leaves Black with a miserable

position. ]

11...

e7 12.e2 d5 QUESTION: So Black

has succeeded in completing his plan?
ANSWER: Yes, and he is now certainly equal,
although no more than that. To be fair,
Nimzowitsch himself acknowledged this, which
makes his earlier undue optimism slightly
harder to understand.

13.

fc1 d7 14.c4 b8 The start of a

defensive regrouping, which shows that
Nimzowitsch appreciated the need for care
here, although, as usual, he dressed it up in
the rather fancy clothes of "over-protecting
strategically important points".

15.

d2 c8 16.e1 e7 17.d3 hd8

QUESTION: Black is solid perhaps, but isn't
he also rather passive?
ANSWER: To some extent, yes. This is what I
mean by Nimzowitsch dressing his defensive
moves up as something grander. He
comments, à propos the rooks' positions on c8
and d8, that: "They are rewarded by the fact
that those who protect strategically important
points in every respect stand well. This
means that strategic points give lustre to their
protectors, just to express it with some feeling".
That sounds to me like a very fancy way of
saying that the rooks are performing a
valuable defensive role. So they are, but that
does not alter the fact that they are
defensively and passively placed, at least for
the time being.

18.

c2 f5 With this move, Black reveals his

intentions of advancing on the kingside, but
White has no weaknesses on that flank, and
the black pawns are not currently supported
by their own pieces, so there is no special
reason to think the attack will be terribly
effective.

19.

c1?!

[ 19.b4 looks rather more purposeful than

just banging his head against the brick wall
that is c7. Mind you, even advancing the a-
and b-pawns does not guarantee an
effective attack, since Black's king position
is absolutely rock solid. ]

19...g5 20.

c5? QUESTION: You don't like

this?

ANSWER: No. Nimzowitsch does not
comment at all on this move, but it seems to
me to be obviously wrong. White just invites a
position where he has a bad bishop against
Black's powerful centralized knight, and from
here onwards, he is steadily outplayed. After
some sort of non-committal waiting move, it
would have been very hard for either side to
make any real progress.

xc5 21.xc5 g8

[ 21...f4! looks more energetic. ]

22.

e2 h5!?

[ Once again, the straightforward 22...f4
23.

d2 g4 looks simpler and stronger.

Nimzowitsch was attracted by the trap
mentioned in the next note, but there seems
no need for such devices, when Black has
such an obvious and good plan available. ]

23.

d2

[ 23.

xh5? g4 was Nimzowitsch's idea. ]

23...h4 QUESTION: Things are looking pretty

bad for White though, aren't they?
ANSWER: Certainly. He has no counterplay
and his bishop is little more than a large pawn,
whilst Black has a huge knight on d5 and a
mobile kingside pawn phalanx. It is all one-
way traffic from now on, much as one would
expect with a great master playing three
unknown amateurs.

24.a4 g4 There was no objective reason not

to take the a-pawn, but Nimzowitsch ignores it
and sticks with his kingside plans.

25.a5 a6 26.b4 c6 Slamming the door on

White's belated attempts to open some lines
on the queenside.

27.

b1 f7 28.b3 f4 White has no

effective counterplay, and can only wait for
the axe to descend on the kingside.

29.

e4 f3 30.c1

[ 30.g3 hxg3 31.fxg3 is the other option, but

it is hopeless in the long run. Black has
several ways to continue, one such being

f2+ 32.

f1 h8 , winning easily. ]

30...fxg2 31.

xg2 cf8 32.f1 g3

The attack plays itself.

33.hxg3 hxg3 34.f4?! QUESTION: Is there

nothing better?

[ ANSWER: Well, 34.

xg3? xg3+ 35.xg3

g8+ mates quickly, ]

[ but 34.

f3 was probably a slightly more

tenacious defence. Then gxf2+ 35.

xf2

h5 36.xf8+ xf8+ 37.e1 leaves White
still on the board, although Black is
ultimately winning after

h8 . ]

34...

e7 Heading to f5, to defend the g3-

pawn.

35.

e1

[ 35.

xg3 f5 36.xg8 xg8+ 37.f2

g3 is also hopeless. ]

35...

f5 36.h1 g4 37.xg3 g6 38.e1

xg3 Nimzowitsch was clearly getting a bit
carried away when he annotated this game,
because he gives a long and flowery prose
comment to this move, describing it "one of
my most beautiful memories".

[ Of course, it prosaically wins a couple of

pawns, but the alternative 38...

xf4

was perhaps even stronger. ]

39.

xg3 fxf4 40.hh3 xd4 41.f2

xg3+ 42.xg3 e4+ 43.h2 xe5

44.

g2 d5+

0-1

D40

Nimzowitsch,A
Brinckmann,A

2nd matchgame, Kolding

1923

[Steve Giddins]

1.d4 d5 2.

f3 c5 3.c4 e6 4.e3 f6 5.c3

[ QUESTION: With regards to the note to

move 8, below, presumably Nimzowitsch
liked playing against hanging pawns?
ANSWER: In general, yes, although he
himself acknowledged that he came badly
unstuck against Tarrasch at St Petersburg
1914, losing a famous brilliancy, which
illustrates the dynamism which the hanging
pawns can contain. It is worth a short
digression to see the game in full. After the
same initial four moves, it continued: 5.

d3

c6 6.0-0 d6 7.b3 0-0 8.b2 b6

9.

bd2 b7 10.c1 e7 11.cxd5 exd5

12.

h4 g6 13.hf3 ad8 14.dxc5 bxc5

(Nimzowitsch had clearly learnt his lesson
by the time of the game against Brinckmann;
here, both the bishop on d3 and the knight
on d2 are misplaced) 15.

b5 e4 16.xc6

xc6 17.c2 xd2 18.xd2 d4!
. The classic advance in such structures,
unleashing a decisive attack on the
undefended white king.
Play concluded: 19.exd4

xh2+! 20.xh2

h4+ 21.g1 xg2! 22.f3 ( or 22.xg2

g4+ 23.h2 d5 , and the mate threat
forces 24.

xc5 , after which Black wins by

h5+ 25.xh5 xh5+ 26.g2 g5+

27.

h2 xd2 , etc ) 22...fe8 23.e4

h1+ 24.f2 xf1 25.d5 ( 25.xf1 h2+

wins the queen ) 25...f5 26.

c3 g2+

27.

e3 xe4+! 28.fxe4 f4+ ( 28...g3+!

mates in three ) 29.

xf4 f8+ 30.e5

h2+ 31.e6 e8+ 32.d7 b5# 0-1. ]

5...

c6 6.e2

[ QUESTION: That looks rather passive! Why

not 6.

d3 ?

ANSWER: The latter is certainly the more
natural-looking move, to modern eyes.
Nimzowitsch's explanation for his more
modest choice is that e2 is a better square
for the bishop in certain of the pawn
formations that could result from the
unresolved central tension. This is
especially true if White isolates the black d5-
pawn.
QUESTION: But what if he doesn't?
ANSWER: Well, there is the rub. Usually,
White would first exchange on d5, to clarify
the structure, and then select the
appropriate square for his bishop. ]

6...

d6 7.0-0 0-0 8.b3 cxd4

[ Black could maintain the symmetry in the

pawn structure with 8...b6 , and after 9.

b2

b7 10.cxd5 exd5 11.dxc5 bxc5
, Nimzowitsch comments that "the hanging
pawns are not to everybody's taste". This is
true, and it is also true that this is another
structure where the position of the white
bishop on e2 is more justified than d3, since
it keeps the d-file open and exerts more
pressure on the d5-pawn. ]

9.exd4

e4 10.b2 xc3 11.xc3 e7?

This allows White to establish a powerful
queenside pawn majority, which theme
dominates the rest of the game.

[ The natural move, recommended by

Nimzowitsch himself, was 11...b6
, when Black is basically fine. ]

12.c5

c7 13.b4 QUESTION: So, is White

better here?
ANSWER: Definitely, thanks to his strong
queenside majority.
QUESTION: But elsewhere in this book, you
have described the advantage of the
queenside pawn majority as a shibboleth and
a myth!
ANSWER: What I said was that it was a
shibboleth and a myth that a queenside
majority is of itself an advantage, regardless
of whether it is adequately supported by its
pieces. Like everything else in chess, it
depends on the specifics of the position.
Where the majority is mobile and advanced,
and supported by its pieces, as here, then it

can be a very powerful weapon.

g6 Black has a majority of pawns on the
kingside, and so he needs to try to use these
as the basis of his counterplay. This means
preparing the advance ...e5.

[ The text does that, but another way to do

the same thing was 13...f6 . ]

14.

e1! QUESTION: Why the exclamation

mark?
ANSWER: Nimzowitsch was quite proud of
this, which he described as "a fine
tournament move". It serves two functions: it
points at the e5-square, helping to restrain
the black pawn advance, and it prepares to
drop the bishop on e2 back to f1, if the enemy
knight jumps into f4.

d7 15.b5 e7 16.d2! Another move with
which Nimzowitsch was quite pleased.

[ 16.

d2 again is prophylaxis against the

advance ...e6-e5, the point being that now

e5 17.dxe5

xc5? runs into 18.b4

, winning the exchange. ]

16...

fc8 17.a4 h8? Rather a pointless

move, suggesting that Brinckmann simply
could not think of anything constructive.

[ QUESTION: Isn't 17...e5 possible, now the

rook has moved away?
ANSWER: Actually, no, it still fails tactically:

18.dxe5

xc5 19.b4 , and the d5-pawn

drops off. ( Strangely, in his notes to the
game, Nimzowitsch gives instead 19.

d4

, claiming the superior game thanks to the
blockade. That is true, but winning the pawn
is better, of course. )]

[ QUESTION: So what should Black do?

ANSWER: It is true that Black's position is
deeply unenviable, but a move such as

17...

f6 would be slightly better than the text.

Black can perhaps cause a few problems
with ideas such as ...Nf4 and a later tactic
with ...Nxg2 or ...Nh3+. At least this would
give White something to think about,
whereas the text is really just a 'pass'
move. ]

[ The immediate 17...f6 is another

possibility. ]

18.a5 White happily takes the free tempo to

press on with his queenside advance.

f6 19.a6 Forcing a very advanced passed

pawn.

b6 20.c6

e8 QUESTION: This looks like a

crushing position for White.
ANSWER: It does, but things are not quite as
simple as they appear. The passed pawn,
although a powerful asset and a long-term

threat, is not going anywhere for the
foreseeable future, so White needs to come
up with a plan to make further progress. It is
not immediately clear how to do so, since
there are no other obvious weaknesses in the
black camp.

21.

f1 f7 22.h4 QUESTION: Is he

planning to attack the king?
ANSWER: Not really, although that may be a
more medium-term aim. The immediate point
is to prevent Black advancing ...e5, due to the
threat of h5, driving away the supporting
knight.
QUESTION: Prophylaxis above all else!
ANSWER: Yes, this was always a major part
of Nimzowitsch's strategy, but perfectly
correctly, in this case. White needs to control
his opponent's counterplay, whilst he seeks a
way to increase the pressure.

d6 23.g3 c7 24.h3 e8 25.e3 f8

[ QUESTION: Isn't the move 25...e5

possible now?
ANSWER: No, it is still prevented, this time
by another tactical trick. Nimzowitsch gives
the variation 26.h5

f8 27.dxe5 fxe5

28.

ae1 d4 ( 28...e4 is met by the crushing

29.h6! , showing that White's 22nd move did

have long-term attacking ideas behind it, as
well as the purely prophylactic ) 29.

xd4!

exd4 30.

xd4 , and White wins, in view of

xh5 31.d7! . ]

26.

ae1 e7 QUESTION: White seems to

have found a second weakness in the enemy
camp, namely the e6-pawn?
ANSWER: Correct, although, as we will see,
this pawn can be defended adequately. But it
is one more problem for Black - he must
defend e6 and keep an eye on the passed c6-
pawn.

27.

b4! The next stage of the plan is to

exchange bishops, which will weaken the
dark squares in Black's camp. In the process
of exchanging bishops, White also wants to
seize control of the a3-f8 diagonal.

ae8 28.c3! xb4 This gives White
control of the diagonal, but it would in any
case be forced after Qa3.

29.

xb4 g8

[ QUESTION: What about 29...e5

now? Don't tell me it is still impossible!
ANSWER: I'm afraid it is. Nimzowitsch had
prepared the variation 30.dxe5 fxe5

31.

xe5 xe5 32.xe5 , and if xe5 then

33.

xf8+ . ]

30.

f5 g6

[ Here, Black finally did have the chance to

play 30...e5 , but he turned it down.
Nimzowitsch says he should have played it,
"probably with equality".
QUESTION: Equality? Surely not?
ANSWER: It is a bizarre assessment. In fact,
as the computer confirms, White has a near-
decisive advantage after 31.dxe5 fxe5

32.

g5 . It is extremely hard for Black to

move anything. The exchange of knight for
the bishop on f7 will leave the d5-pawn
extremely exposed: for example, h6

( alternatively, keeping the bishop with
32...

h5 still exposes the d5-pawn: 33.b3

d6 34.e4 and the pawn drops off )

33.

xf7 xf7 34.c2! with Bb3 coming. ]

31.

xg6 xg6 32.h5 f8 33.h4 f7

[ Yet again, the move 33...e5 is impossible,

this time because of 34.

f5 e6 35.h6! .

QUESTION: It is amazing how White has so
often managed to ensure that Black cannot
get in this much-desired advance.
ANSWER: That is the essence of great play,
especially prophylactic play. Note how it is
the use of little tactical ideas that has
enabled White to achieve this. We have
seen many more examples in this book,
where good positional and strategical play
is founded on tactics. The ability to
calculate little variations, and spot and use
tactical motifs is absolutely fundamental to
good positional play. ]

34.

g2 g6

[ 34...e5 35.

f5 e6 36.h6 is still

devastating: for instance, exd4 37.

xe6

xe6 ( or 37...xe6 38.d6+ ) 38.hxg7
, and wins. ]

35.hxg6+ hxg6 36.f4 Now Black was

threatening 36...e5, so Nimzowitsch slams the
door on the idea once and for all.

d8 37.f3 Black is hopelessly tied up, and
White just has to find a way to break through.
The open h-file is now to be one of the keys.

c7 38.h1 g8 39.ee1 h7 If Black just
waits, then he must reckon with Nh2-g4,
increasing the pressure even further.

40.

xh7 xh7 41.h1 f8 Black now

intends ...Re7-h7, to challenge the h-file.
Nimzowitsch tries a little feint, to disrupt this.

42.

h6! e7

[ 42...

g7 43.h2 would gain a tempo,

since the king now blocks the intended ...
Re7-h7. ]

43.

a3 g7? The text loses to a nice

finishing combination.

QUESTION: Does Black have any better
defence?

[ ANSWER: If Black had just waited,

Nimzowitsch outlines a remarkable winning
plan, consisting of preparing a knight
sacrifice on b6 (!), but this looks very
fanciful and unnecessary. Instead, after first
pushing his pawn to g5, to secure the
outpost e5 (as Nimzowitsch indicates),
White can win by just going after the g6-
pawn: for example, 43...

d8 44.c1 c7

45.g4

d6 46.g5 f5 47.e5 g7

( 47...

g7 does not defend the pawn, in view

of 48.

xg6 xg6 49.xg6 xg6 50.c7 )

48.

h1 , and there is no defence to Rh8 and

Qh6 mate. If

b4 49.h8 d2+ 50.g3

e3+ 51.f3 . ]

[ 43...

h7 is the other option, but then

44.

xh7 xh7 45.e3 f7 46.g4

, and once again, the pawn comes to g5,
securing control of e5, and then White wins
quickly. For example,

d6 47.g5 f5

48.

e5 d8 49.c7 c8 50.d6 e8

51.

e5 and wins. ]

44.

h8+! xh8 45.xf8+ h7

[ 45...

g8 46.xf6+ is effectively the

same. ]

46.

xf6 Now the knight has giant squares on

g5 and e5, and the white queen and knight
combination is actually stronger than Black's
queen and rook.

e7 47.g5+ h8 48.e5 c7 49.xe6

e7 50.h3+

[ Black resigned in view of 50.

h3+ g8

51.

c8+ f8 52.xf8+ xf8 53.e6+

f7 54.xg7 xg7 55.c7 . Megabase
actually gives these last moves as occurring
in the game, but Skjoldager and Nielsen
(following Nimzowitsch himself in 'Die
Blockade') show the game as ending at
move 50, which I take as correct. ]

1-0

E06

Sämisch,F
Nimzowitsch,A

Copenhagen

1923

[Steve Giddins]

1.d4

f6 2.c4 e6 3.f3 b6 The Queen's

Indian Defence is the natural counterpart to
the Nimzo-Indian, continuing the fight for the
e4-square.

4.g3

b7

[ QUESTION: Isn't 4...

a6 more popular

nowadays?
ANSWER: Yes, and it too was originally a
Nimzowitsch idea. For example, his game
against Sultan Khan from Liege 1930,
continued 5.b3

b4+ 6.d2 xd2+

( 6...

e7 is the main line nowadays )

7.

bxd2 b7 8.g2 c5 9.dxc5 bxc5

10.0-0 0-0 and Black stood well, eventually

winning after 60 moves. ]

5.

g2 e7 6.c3 Not surprisingly for a

game played almost 100 years ago, in the
earliest days of the development of this
opening, the move orders adopted by both
players are not entirely accurate, compared
with contemporary understanding.

[ The usual way to reach the game position is
6.0-0 0-0 7.

c3 . ]

6...0-0!? Not regarded as 100% accurate by

theory nowadays.

[ The best reply - certainly the safest, anyway

- is 6...

e4! . ]

7.0-0 And White, in his turn, settles for routine

development, instead of trying to profit from
the move order.

[ Indeed, 7.

c2! is the most challenging

move for Black to face.
QUESTION: Why is this?
ANSWER: Well, as mentioned in the first
note to this game, the Queen's Indian is first
and foremost about the battle for the e4-
square, especially in these old-fashioned
main lines with the black bishop on b7. If
White can get in the move e4 then, ceteris
paribus, he will usually have some
advantage, thanks to his central control and
extra space, and so the fight develops
around his attempts to do this, and Black's
attempts to prevent it. The move 7 Qc2
threatens that move, and Black can no
longer respond 7...Ne4.
QUESTION: So is he just worse?
ANSWER: It is not quite as crystal-clear as
that, but certainly, he has to solve a few
problems. The move

A) so Black usually enters sharp lines

starting with 7...c5 8.d5! exd5 9.

g5

, with unclear complications. This variation
was tested several times in the 1974
Candidates Final match between Korchnoi
and Karpov, which showed that Black has
to tread carefully to avoid a disaster. The
famous miniature in Game 21 of that
match shows what can befall him if he
does not:

c6 10.xd5 g6 11.d2

xd5?! 12.xd5 b8?? 13.xh7!
and Black was completely busted.

e8

( 13...

xh7 14.h6+ g8 15.xg6+

h8 16.h5+ g8 17.e4 f5 18.d5+
forces mate in three ) 14.

h6 e5

15.

g5 xg5 16.xg5 xg5 17.xg5

xd5 18.0-0 xc4 19.f4 1-0, V.Korchnoi-
A.Karpov, Moscow 1974.
QUESTION: Hmmm, this line certainly
does look pretty dangerous for Black!
ANSWER: Well, it is not as terrible for
Black as Karpov made it look here, but it
definitely poses some problems. Anyway,
back to the game.;

B) 7...d5 , similar to the game, is not

entirely comfortable, thanks to the
imprisonment of the queen's bishop, ]

7...d5?! A slightly old-fashioned treatment,

which is rarely seen these days.
QUESTION: What is the objection to 7...d5
then?
ANSWER: The basic problem is that, having
gone to some lengths to place his queen's
bishop on a nice open long diagonal, Black
now closes that diagonal, and potentially
allows White to set up a pin along it as well. It
is not the end of the world, and, indeed,
Botvinnik had considerable success with the
black position back in the 1930s, but practice
has shown that, with accurate play, White can
secure a certain positional advantage.

[ The main line here is 7...

e4 , ]

[ whilst Sergey Tiviakov's favourite 7...

a6

is also perfectly satisfactory for Black. ]

8.

e5 c6?! Another sub-optimal choice.

[ 8...

a6 is relatively best, but White still has

some pressure against Black's central
pawns after 9.cxd5 exd5 10.

f4 c5

11.

c1 . ]

[ QUESTION: The move 8...

bd7

looks natural to me. ]

[ ANSWER: It is natural, but 8...

bd7?!

also runs into problems, mainly because of
the weakness of the c6-square. White
replies 9.

a4! , which threatens either to

come into c6 with the knight, or (more
immediately) to take on d7 as many times
as necessary, and win the d5-pawn.
QUESTION: OK, but can't Black just take on
e5?
ANSWER: Then after

xe5 10.dxe5

he still has a problem.

A) 10...

d7? just drops a pawn after

11.cxd5 , thanks to the tactical trick

xe5

12.d6! , winning a piece.

QUESTION: Nasty!
ANSWER: Another case of LPDO.;

B) 10...

e4 is slightly better, but still not

nice for Black after 11.cxd5

xc3

12.bxc3

xd5 ( 12...exd5 13.c4

continues to exploit the pin on the long
diagonal ) 13.

d1 e8 14.g4

, and Black still has problems, with Bh6
being an unpleasant threat.;

C) 10...-- ]

9.cxd5?! This exchange gives away the

opening advantage.

[ Black has more problems after the central

thrust 9.e4! , when he has no reliable way
to avoid at least some disadvantage: for
example, dxc4 ( 9...

bd7?! is worse,

thanks to the tactical blow 10.

xc6! xc6

11.exd5 exd5 12.cxd5

b7 13.d6!

- those dreaded four letters again: LPDO -

xg2 14.dxe7 xe7 15.xg2 , with a
clear extra pawn ) 10.

xc4 a6 11.b3

, with a plus.
QUESTION: Wait a minute! Can't Black win
a pawn here with b5 12.

e3 b4 13.e2

xe2 14.xe2 xd4 ?
ANSWER: Well spotted! Alas, though, White
has nice compensation after 15.

b2

. He has the two bishops, better
development, more space, and the extra
black pawn is a sickly backward one on c6. ]

9...cxd5 10.

f4 a6! QUESTION: What is the

assessment?
ANSWER: Black is now doing fine. The
exchange on d5 has relieved him of all his
problems.
QUESTION: But his bishop on b7 looks pretty
bad, still.
ANSWER: True, but Black's structure is rock-
solid and, actually, White's bishop on g2 is
not that much more active than its counterpart
on b7. It looks better, because its pawns are
not fixed on the same coloured squares, but
in reality, both bishops are biting on the
granite of the d5-pawn. It is similar to what
often occurs in the Stonewall Dutch.
QUESTION: Can't White attack in the centre
with e4?
ANSWER: He can, but that will open up
Black's bishop every bit as much as White's,
and will also leave White with a weak isolated
pawn on d4.
QUESTION: So why the exclamation mark for
Black's last move?
ANSWER: This was Nimzowitsch's own
punctuation. His plan is to advance ...b5 and

use the c4-square as an outpost for his knight.
However, the position is no more than equal,
at best, for the moment - it is only White's
later inaccuracies which lead him into trouble.

11.

c1 b5 12.b3

[ Kasparov suggests 12.a3 , when

c6?

( 12...

bd7 is met by 13.d3

"with a complicated, unclear position"; White
will follow up with b4 and Nc5 ) is tactically
impossible because of 13.

xd5! . ]

12...

c6?!

[ In view of the tactical possibility pointed out

by Kasparov in the next note, it would
probably be more accurate to play

12...

bd7 , with unclear chances. ]

13.

xc6?!

[ 13.

xd5 is stronger, forcing the

continuation

xd4 14.xe7+ xe7

15.

e3 .

QUESTION: What is so great about this?
ANSWER: Kasparov continues the variation
for some moves, beginning with

xg2

( Fritz prefers the immediate 15...

f5

, but even here, after 16.

c5 xc5

17.

xc5 xg2 18.xg2 g5! 19.c1! h6

20.g4 White has some initiative ) 16.

xg2

b7+ 17.f3 f5 18.f2 when White does,
indeed, have some advantage - Black's
knight will be driven out of the centre by e4,
after which White has definite pressure. In
any event, it seems clear that 13 Nxd5 is
stronger than the game continuation, after
which White starts gradually to drift into
difficulties. ]

13...

xc6 14.h3 d7 15.h2 QUESTION:

What is White doing? He seems to have run
out of ideas.
ANSWER: That is precisely what has
happened, I think. It is not easy to see a
constructive plan for White, and Sämisch just
shuffles around indecisively, almost as if he is
waiting for Black to do something and thus
spur him into a response. But Nimzowitsch
uses the time to start taking a grip on the
game.

h5 16.d2 f5 QUESTION: So we have a
kind of Stonewall set-up?
ANSWER: Yes, and in a relatively favourable
form for Black. It is worth thinking again about
the two light-squared bishops. Formally,
White's still looks the more active, since his
central pawns are not fixed on light squares.
Yet, if one looks more deeply at the position,
it becomes clear that the bishop on g2 is not
really doing much at all, just staring at the

brick wall on d5. Black, on the other hand, has
a plan to activate his own light-squared bishop,
by means of ...b4 and ...Bb5.

17.

d1? White has already lost a lot of time,

and stands very badly, but Kasparov suggests
that this is the final, decisive mistake.

[ He says White could still fight with 17.

b1

a5 18.

c2 .

QUESTION: But this looks terrible for White!
ANSWER: It does, indeed. Kasparov
continues his variation with f4 19.g4

d6

20.

f3 f6 21.fc1 fc8 , but the resulting

position is dreadful for White. Even so, it is
better than what happened in the game. ]

17...b4 18.

b1 b5 Now we see just how

'bad' the Stonewall bishop really is.

19.

g1 d6 20.e4 The text is a last,

desperate attempt to change the course of
events, by trying to exploit the tactical
possibilities against the 'loose' knight on h5.
Unfortunately, it loses by force, as
Nimzowitsch had seen, but White does not
really have anything better.

[ He can hold up the kingside threats by

means of 20.e3 , but after

f6 Black just

switches his attention to the queenside
and doubles on the c-file. White's position is
so cramped that his pieces are almost
literally tripping over each other, and he is
just lost. ]

20...fxe4! The prelude to one of the most

famous finishes in chess.

21.

xh5 xf2 Black already has two pawns

for the piece, and the white pieces still have
almost no moves.

22.

g5 QUESTION: What is the point of

that?
ANSWER: White wants to defend his g-pawn,
so he can unpin with Kh1. It still fails, but
what other constructive moves does he have?

af8 Now 23...R8f3 is a threat, hence White's
next.

23.

h1 8f5 24.e3 d3

Now he threatens 25...Re2.

25.

ce1 h6!! One of the most well-known

final positions in chess history.

[ After 25...h6 White has no safe move,

apart from 26.b3 (met by any waiting
move), ( or 26.a3 (met by a5
): for example,; 26.

c1 xb1 ,; 26.h2

5f3 ,; or 26.g4 5f3 .
QUESTION: Zugzwang on a full board, isn't
it?
ANSWER: Indeed, and a remarkable
achievement in practical play (although, as

we have already seen in Mannheimer-
Nimzowitsch, this was not the only occasion
when Nimzowitsch managed this in a
tournament game). )]

0-1

C01

Yates,F
Nimzowitsch,A

Semmering

1926

[Steve Giddins]

1.e4 e6 2.d4 d5 3.

c3 b4 4.exd5

Another Exchange variation; indeed, in those
days, it was the most popular response to the
then rather new Winawer variation. However,
as we have already seen in Mannheimer-
Nimzowitsch, Nimzowitsch made something of
a speciality of playing such positions for a win
as Black.

exd5 5.

d3 e7 6.ge2 QUESTION: We

haven't seen this before?
ANSWER: No, Yates is being rather clever, if
such spineless handling of the white pieces
can be regarded as clever at all. As we have
already discussed, Nimzowitsch's favourite
device in such variations was to place his
king's knight asymmetrically to its opposite
number - thus, if White played Nf3, he could
play ...Ne7, and vice versa. Here, Yates has
waited until Black has committed his king's
knight first, and then copies his deployment,
thus avoiding such asymmetric treatment.
QUESTION: So, was Yates playing for a
draw?
ANSWER: Funnily enough, I am not sure he
was. Quite apart from the fact that Yates had
a reputation as rather a fearless and brave
player, who was not usually intimidated by
reputations (witness his two tournament
victories over Alekhine, both as Black), the
1920s was a period where openings played
much less significance than they do nowadays.
Such a tame and symmetrical development
looks to us like a sign of a player seeking just
to make a draw, but in those days, many
masters were happy just to develop their
pieces in fairly routine fashion in the opening,
and then start the real battle in earnest in the
middlegame. The opening was not seen as
integral to the later middlegame plan, in the
way it subsequently became.

0-0 7.0-0

g4 Nimzowitsch, on the other

hand, was one player who never just

developed his pieces routinely. Despite being
thwarted of his favourite asymmetrical
development of his king's knight, he still
avoids a totally symmetrical piece
development. Rather than bringing the bishop
to f5, to exchange it off, he chooses a more
ambitious deployment for it, hoping to provoke
the weakening pawn move f3. Note too, that
he has avoided the ...Nbc6 development
which is typical of this variation; as we will see,
he has a more ambitious plan in mind.

8.f3

h5 9.f4 g6 10.ce2 Rather than

taking the bishop at once, Yates preserves
his pawn structure from the doubled pawns
resulting from the exchange on c3. We have
already seen in Mannheimer-Nimzowitsch how
adept Nimzowitsch was at exploiting such a
formation.
QUESTION: But now White can gain the
advantage of the bishop-pair. Isn't that better
for him?

[ ANSWER: He would probably be a little

better after 10.

xg6 hxg6 , but the position

would still be anything but symmetrical.
Nimzowitsch was willing to accept this, as
the price for unbalancing the position.
Furthermore, as we have already seen in a
number of previous games, he was in
general not a huge fan of the bishop-pair
anyway. He tended to prefer knights to
bishops, not just in blocked positions, but
even in positions where most masters would
prefer the bishops. As Keene pointed out in
his book, Nimzowitsch rather gave the game
away in 'My System', where he justified
including a chapter on the two bishops, by
saying "the reader naturally has the right
to expect that I should enlighten him, as far
as I can, on the dangers in which a pair of
enemy bishops may involve him". Note the
assumption that it is the opponent who will
have the bishops, not we ourselves. ]

10...

d6 11.e1?! QUESTION: That looks a

funny move!
ANSWER: Yes, it does. Yates clearly intended
to transfer his queen to the kingside, to attack
there, but this proves rather optimistic.
Nimzowitsch emphasized that the move
ignores the centre, which soon becomes the
theme of the game.

[ Interestingly, and characteristically in view

of the previous note, he recommended

11.

xg6 , ]

[ but I am sure most modern masters would

take the bishop-pair with 11.

xg6 hxg6

and then, say, 12.

e1 with perhaps a slight

advantage (although it is not very much). ]

11...c5! QUESTION: That too is a bit of a

surprise. After all, now Black has an isolated
pawn on d5.
ANSWER: Yes, and in fact, Nimzowitsch was
in generally fairly distrustful of the IQP, in
contrast to his great rival Tarrasch, who
almost always preferred to play with the IQP,
rather than against it.
QUESTION: So why does Nimzowitsch take
on the IQP here?
ANSWER: Well, as IQP structures go, this is
a pretty decent one for Black. White's pieces
are not well placed to bring any pressure on
the d5-pawn (his bishop on d3 is blocking the
d-file, and his queen is misplaced).
Furthermore, the fact that the pawn is on f3 is
a disadvantage, as it weakens the e3-square
and the g1-a7 diagonal. Indeed, this was all
part of Black's plan, from the moment he
played 7...Bg4. He intended to provoke the
move f3, and then attack the centre with ...c5.
That is also why he left his queen's knight at
home.

12.dxc5

[ QUESTION: Couldn't White refuse to 'play

ball' and just defend d4 with 12.c3 ?
ANSWER: Yes, but after

bc6 and ...Qb6,

Black develops pressure against d4 anyway.
It would be hard to avoid the exchange on
c5 sooner or later. ]

12...

xc5+ 13.h1 bc6 QUESTION: So I

guess you would say Black is doing well here?
ANSWER: Yes, I like his game. The
theoretical weakness of the d5-pawn is really
just that, i.e. theoretical. Black has free piece
play and good central control. Nimzowitsch
probably thought he already stood a little
better.

14.

d2 e8 15.xg6 hxg6 16.f4?

This really is rather a horrible move.
QUESTION: Why?
ANSWER: Because it weakens White's
central position even further, especially the
e4-square, and shuts in his dark-squared
bishop.
QUESTION: So what was Yates' idea?
ANSWER: I think he wanted to attack on the
kingside, just as with his earlier move 11 Qe1,
but as Nimzowitsch pointed out, White's
central position is just not strong enough to
justify this.

[ He would have done better to try something

like 16.

g5 f6 17.f4 , although after g5

18.

g3 d4 it is already a little uncomfortable

for White. ]

16...

f5 The weakness of e3 is now really

noticeable.

17.c3 d4! 18.c4

[ This turns Black's IQP into a strong passed

pawn, but the alternative was even worse:

18.cxd4?!

cxd4 , and White has serious

trouble down the d- and e-files: for example,

19.

d1 xe2 20.xe2 h4 (threatening

21...Ng3+) 21.

f3 ad8 , with enormous

pressure. ]

18...

b6 QUESTION: It looks grim for White?

ANSWER: Yes, he has been totally outplayed.
Black dominates the centre. For the time
being, the passed d-pawn is securely
blockaded, but as we will see, this will not
remain the case forever.

19.

f3 b4 QUESTION: Isn't there a pawn

en prise on b2?

[ ANSWER: Not really. After 19...

xb2?

20.

b1 White threatens to regain the pawn

on b7, whilst

xa2 21.a1 b2 22.b1

would force a draw by repetition. This is a
particularly clear-cut case where the tactics
do not favour Black, but we have already
seen in previous games that, once he
secured a safe positional edge, Nimzowitsch
was always reluctant to give his opponent
any tactical chances at all, even where
those chances were objectively favourable
to himself.
QUESTION: OK, I see why he didn't take on
b2, but why the text move? The d2-bishop is
a fairly bad piece, obstructed by the f4-pawn,
so it looks illogical to exchange it off.
ANSWER: It is all about the weakness on e3.
It is true that the bishop on d2 is not very
active, but it is the key defender of e3; once
the bishop is gone, Black's knight or rook
can occupy that monstrous outpost, like a
bone in White's throat. ]

20.a3

xd2 21.xd2 a5 QUESTION: Is that

really necessary?
ANSWER: Not exactly necessary, but a very
useful move to play. There is no hurry to
occupy e3, as White cannot prevent it anyway,
so Nimzowitsch takes time to secure another
small plus, by holding back the white b-pawn
and stopping a later b4 advance. As we will
see, he intends to use the weakness on b3
later on, to exert pressure against the bishop
on d3, which is the key blockader of the
passed d-pawn. With Nimzowitsch, no matter
what was going on in the position, thoughts of

blockade were never far away.

22.

g1 e3 QUESTION: Slightly surprising

again. I thought the knight would come into
e3.
ANSWER: The knight would also be strong
there, of course, but, as I said, Nimzowitsch is
thinking in terms of blockade. The e3-square
is a very nice outpost in its own right, but
Nimzowitsch is thinking of how it can be made
most effective use of. His answer is: as a
staging-post, to bring pressure against the
blockading bishop on d3. That is now his
main target. First, the rook attacks it, and later
the black queen will attack it again, from b3.

23.

f2 ae8 24.d1 b3 The pressure

against d3 intensifies.

25.

d2? After this new inaccuracy, Black is

able to force the exchange of the blockading
bishop.

[ QUESTION: How about 25.

c2

to exchange or drive off the queen?
ANSWER: That was in fact the best defence.
Black would continue

xc2 26.xc2

, and then probably

e1 . He retains the

advantage, but White is still resisting
stubbornly. ]

25...

d6 26.c5 The only way to avoid loss of

the c4-pawn.

c4 27.xc4 xc4 28.c2 d5
Mission accomplished. Now the blockader has
been eliminated, and White must face new
threats, involving the potential advance of the
passed d-pawn.

29.

c1 QUESTION: What is that for?

ANSWER: Basically, it is about the only legal
move that does not drop something. White is
tied hand and foot, and is struggling for
moves.
QUESTION: Almost zugzwang?
ANSWER: Yes, a kind of 'zugzwang-lite'. It is
remarkable how often Nimzowitsch managed
to reduce even strong masters to such straits.

e4 30.f5?! This just loses a pawn, but
White was desperate.

[ Black was threatening 30.--

xf3 31.xf3

( if 31.

xf3 xf4 ) 31...f5 , and he

maintains his grip, with such threats as ...
Re4 or ...Re3, followed by the advance of
the d-pawn. His advantage should be
decisive. ( In 'Chess Praxis', Nimzowitsch
claims Black's threat is, instead, here

31...

xf3 32.xf3 e2 "and the seizure of

the seventh rank should be decisive", but in
fact, this is not so clear. White can struggle
on with, say, 33.

g1 xb2 34.d1 c2

35.

xd4 xc5 36.xc6 bxc6 37.d8+

h7 38.a8 , and the rook ending offers
some hopes of salvation. )]

30...

xf3 31.xf3

[ 31.

xf3 gxf5 is equally hopeless. ]

31...

xf5 Nothing has really changed, except

that Black has an extra pawn into the bargain.

32.b4 axb4 33.axb4

xb4 QUESTION:

Giving up his passed pawn?
ANSWER: 'Trading it in' would be more
accurate. Black will be left with an extra
outside passed pawn, as well as a powerful
initiative against the white king.

34.

xd4 d3 35.c2 c8 The white c-

pawn is dropping off.

36.

c3 xc5 37.h4 b6 The rest is pretty

clear. Black has two extra pawns, and only
needs routine care to convert his advantage.

38.

g5 f8 39.f3 d7 40.c4

White has a slight temporary initiative, but
Black soon defuses it.

e6 41.d3 c8 42.b5 c1+ 43.h2

f4+ 44.g3 f2+ 45.h1 f1+ 46.h2

c5
0-1

A34

Nimzowitsch,A
Rubinstein,A

Dresden

1926

[Steve Giddins]

1.c4 c5 2.

f3 f6 3.c3 d5 4.cxd5

xd5 5.e4 At the time, this was
Nimzowitsch's own novelty. White sets his
pawn centre in motion, at the risk of a
backward d-pawn.

b4

[ The critical response, although

Nimzowitsch actually recommended

5...

xc3 6.bxc3 ( 6.dxc3 xd1+ 7.xd1

looks colourless, but Ulf Andersson has
shown that this position is also not without
some poison ) 6...g6 , transposing into a
type of Grünfeld. ]

6.

c4 e6 QUESTION: Why not check on d3?

[ ANSWER: Nimzowitsch dismisses this on

the basis of 6...

d3+ 7.e2 f4+ 8.f1

"with the threat of d4", but this is in fact
nowadays the main line. After

e6

, the critical variation is the extravagant-
looking positional gambit 9.b4!? cxb4

10.

e2 , a kind of English Opening version

of the Evans Gambit, in which White hopes

to capitalize on his central pawns after a
subsequent d4, ( although 10.

d5

is also possible. )]

7.0-0

8c6 8.d3 QUESTION: So what is

going on here? White seems to have a
backward pawn and a hole on d4.
ANSWER: Both of those statements are true,
but despite this, he has a decent position.
This is why the present game is so important,
as it is in many ways the forerunner of many
modern Sicilian systems involving the move ...
e5 by Black. What we have here is basically a
kind of reversed Sicilian formation, except
that the black e-pawn is on e6, rather than e5.
As in the Open Sicilian, the exchange of c-
pawn for the enemy d-pawn has left White
with a central pawn majority. The advance e4
gives him extra central control over certain
squares, and he argues that the weakness of
d4 is something he can 'play around'. We
have seen Nimzowitsch use a similar strategy
in other openings, such as in Nimzowitsch-
Olson. The other point here, again similar to
some Sicilian variations, especially the
Sveshnikov variation, is that the black knight
on b4 is in danger of losing time, and being
driven back to the poor square a6.

d4 Anticipating the threat of a3, Black frees
c6 for the other knight to retreat to, but now
White solves the problem of his backward d3-
pawn.

9.

xd4 cxd4 Now the d3-pawn is shielded

down the closed d-file.
QUESTION: Why not recapture with the
queen, and keep the backward d-pawn
exposed?

[ ANSWER: Sadly, Black cannot really afford

to do this, as he falls too far behind in
development. White can continue 9...

xd4

10.

b5 d8 11.f4 a6 12.g4

with a strong initiative – Black has real
trouble developing his pieces satisfactorily. ]

10.

e2 Now White plans to advance f4,

setting his kingside initiative in operation. He
stands somewhat better, although at least
here, Black manages to complete his
development.

a6 11.

g3 d6?!

[ 11...

e7 was safer, in view of the next

note. ]

12.f4

[ Nimzowitsch himself acknowledged that
12.

g4! would have been stronger here,

after which Black has some problems, due
to his vulnerable kingside. For example,

0-0? ( 12...g6 is better, but also unpleasant

after 13.

h6; Nimzowitsch gives 12...f6

as relatively best, but Black is still worse
after 13.f4 ) 13.

g5 (showing up the

downside of Black's 11th move)

e7

14.

h6 f6 15.e5! ( even better than

15.

xg7 e5 16.xf6+ xg4 17.xd8

fxd8 , although this is just a solid extra
pawn ) 15...

xe5 16.xg7 xg7 17.h5

, and wins. ]

12...0-0 13.

f3

[ Nimzowitsch now rejects the direct attack

with 13.e5 as inadequate after

c7

, but the computer shows that White still has
a promising position after 14.

d2 d5

15.

xd5 xd5 16.b4 e8 17.e4 . ]

13...

h8 QUESTION: What is that for?

ANSWER: Black is preparing the advance ...f7-
f5, to put paid to any possible f4-f5 threat.

[ However, this is rather weakening, so
13...

c6 looks safer. ]

14.

d2 f5 15.ae1 c6 QUESTION: How

do you assess the position?
ANSWER: As Nimzowitsch himself points out
in 'My System', Black has disposed of the
immediate threats to his king, but at the cost
of creating a new weakness on the e-file. This
now becomes White's main target. He still
stands better.

16.

e2 c7!? Nimzowitsch was critical of

this, on the grounds that Black thereby gives
up the possibility of a later ...Qf6.

[ However, it is not clear that his suggestion

of 16...

d7 is any better. ]

17.exf5 exf5 18.

h1!? This is the move for

which this game is famous.
QUESTION: What is the idea? It looks very
retrograde.
ANSWER: 'Reculer pour mieux sauter', as
they say in my local! The knight embarks on a
journey to g5, via h1-f2-h3. It is obvious that
on g5, the knight will cooperate extremely well
with the bishop on c4. However, it is a slow
manoeuvre, and although over the years, 18
Nh1 has accumulated more exclamation
marks than many chess players have had hot
showers, it is not clear exactly how strong the
move is objectively.

d7 19.f2 ae8 20.fe1 xe2 21.xe2

d8?!

[ 21...

e8 is natural, but after 22.xe8+

xe8 23.d5 Black is losing material
because of the threat on g8, which forces

e7 , dropping a pawn after 24.xd4 . ]

[ However, the computer suggests that White

has nothing after 21...

b4! 22.xb4

( 22.

c1? b5 23.b3 a5 is better for

Black ) 22...

xb4 . Indeed, a possible

variation now is 23.

h3 b5 24.g5 h6

( 24...bxc4? leads to a quick mate after
25.

h5 h6 26.g6 hxg5 27.e7 ) 25.f7+

h7 26.g5+ , with a draw by repetition.
This in itself is enough to show that the
hyperbole heaped on Nimzowitsch's 18th
move is misplaced. ]

22.

h3 c6!?

[ 22...h6 does not now stop 23.

g5

anyway, but this is still Black's best
continuation; now

b5 24.b3 g6

and, although the black kingside has been
significantly weakened, it is not clear that
White has so much after 25.

e6 xe6

26.

xe6 h7 . ]

23.

h5 g6 24.h4 Now the threats start to

become serious, however.

g7 25.f2! QUESTION: What is that for? It
looks inconsistent with the attempt to attack
the black king.
ANSWER: It is a typical example of the 'two
weaknesses' technique, which we have seen
many times in this book. Nimzowitsch realizes
that he cannot force an immediate kingside
breakthrough, so he indulges in a manoeuvre,
designed to disrupt the coordination of the
black forces. The d4-pawn is attacked and
must be defended.

c5

[ After 25...

b6 , Nimzowitsch gives 26.b4

, intending Bc3, ( but the computer points
out that 26.

e1! is stronger still, with

threats of Ba5 and/or Re7. )]

26.b4

b6? Although Nimzowitsch does not

comment on this move, it is a losing mistake.

[ 26...

e7 is more tenacious, giving up the d-

pawn to retain control of e7, although after

27.

xd4+ f6 28.c5 b6 29.f2

, White is a good pawn up. ]

[ EXERCISE: After 26...

b6 White's queen

can move again, but to where? ]

27.

h4? Missing his chance.

[ ANSWER: Nimzowitsch himself points out

that 27.

e1! is better, as the computer

confirms: for instance,

e4 28.f2

, and White simply wins a pawn. ]

27...

e8 28.e5 Again, the recipient of

numerous exclamation marks down the years,
but the computer shows that it is nothing for
White.

f7?

[ 28...

xe5? 29.fxe5 xe5 30.h6+ f6

31.

g5# is the main point, ]

[ whilst Nimzowitsch gives 28...h6 29.g4

"with a very strong attack"; however, the
computer is unimpressed after

d7 . ]

[ But perhaps the simplest defence of all was

the immediate 28...

d7 , after which 29.g4

h6 transposes to the previous note. It

appears that White does not have anything
here. ]

29.

xf7 xf7

[ If 29...

xe5 30.g5! , ( which is stronger

than Nimzowitsch's 30.fxe5

xf7 31.g5

g8 32.e6 when d8! 33.xd4+ f6
is holding for Black: 34.

d6 d8! 35.xd8

xd8 , and White cannot hold the e6-
pawn. )]

30.

g5 g8 31.xe8 xe8 32.e1!

QUESTION: Yet again the queen switches its
line!
ANSWER: Yes, it is an attractive feature of
the game, the way that the white queen
switches so effectively between squares on
the short diagonal e1-h4. This last switch is
decisive, as Black cannot prevent a winning
penetration to e5 or e7. As we have seen
elsewhere in this book, the combination of
queen and knight can be a devastating
attacking partnership.

c6

[ 32...

f8 can be met by the same idea as in

the game: 33.b5! . ]

33.

e7+ h8 34.b5

[ This elegant final blow has also been much

praised, but the soulless silicon beast points
out that the prosaic 34.

e6 is simpler and

even more deadly - there is no defence to
Qf6+: for example, h5 35.

f6+ h7 36.g4!

and the threat is -- 37.

f8+ h6 38.g5#

. ]

34...

g7

[ Giving up the piece is the equivalent of

resignation, but there is nothing better: for
instance, 34...

xb5 35.f7+ g7 36.e5+

h8 ( 36...h6 37.g4! mates quickly )

37.

f6+ g7 38.xb6 , ]

[ or 34...axb5 35.

e6 . ]

35.

xg7+ xg7 36.bxc6 bxc6 37.f3

Black should resign, and the rest requires no
comment.

c5 38.

e5 c7 39.c4 f7 40.g3 d8

41.

a5 e7 42.c7 e6 43.b6 h6 44.h4

g5 45.h5 g4 46.

e5 One of Nimzowitsch's

most famous games.
QUESTION: But it was not entirely convincing
in places, was it?

ANSWER: That is true. There were a few
missed opportunities, and also, his famous
manoeuvre starting 18 Nh1 was not
objectively even better for White. But the
originality of Nimzowitsch's concepts is what
makes the game important, especially the
opening idea with 5 e4, accepting the
backward d-pawn.
1-0

A61

Nimzowitsch,A
Marshall,F

New York

1927

[Steve Giddins]

1.c4

f6 2.d4 e6 3.f3 c5 The Modern

Benoni was an extremely rare opening in the
1920s. Indeed, it did not really exist as a
recognized opening at all, being dismissed as
'Irregular'. Nimzowitsch himself, no stranger
to experiments in the opening, described it as
an "extravagance" on Marshall's part.

4.d5 d6 5.

c3 exd5 6.cxd5 g6

QUESTION: Marshall seems to know what he
is doing, unknown opening or not!
ANSWER: Indeed. In fact, this game is
extremely impressive for the modernity with
which both players handle the opening and
early middlegame. Usually, when one sees
such modern openings used by masters of the
1920s or earlier, one finds oneself
involuntarily wincing at the rather ham-fisted
way they interpret them, but the present
game could quite easily be passed off as
having been played by two contemporary
GMs.

7.

d2 bd7 8.c4?! The only real blemish

on Nimzowitsch's handling of the opening.

[ It would be more accurate to preface this

with 8.e4

g7 and only now 9.c4

, so as to be able to meet

b6 with 10.e3

, reaching the game position. ]

8...

b6 9.e4 g7?! Returning the

compliment.

[ Marshall should have seized the chance to

capture on c4: 9...

xc4! 10.xc4 g7

, with a comfortable position. ]

10.

e3! QUESTION: So why is this good?

After all, e3 does not look like a great square
for the knight - it obstructs the queen's bishop,
and the e-file.
ANSWER: Yes, but that is a purely temporary
thing. White intends a4-a5, driving away the

enemy knight, whereupon the white knight will
return to c4. Black has less space, so White
is happy to avoid exchanges. Indeed, one can
see that Black is likely to face a bit of a
traffic jam in his position, as the queen's
bishop has no square to go to, except d7,
which is the square the queen's knight will
want to retreat to.

0-0 11.

d3 h5 This was widely criticised,

but as Keene suggests, it is probably not so
bad.

[ Keene himself quotes one of his own

games, where Black tried 11...

e8 12.0-0

c4 13.

c2 d7 , but came to grief. After

14.

d2 c8 15.h1 c5 16.f3 c8 17.a4

a6 18.

e2 White regrouped his queen's

bishop to d4 and queen to c3, with an
enduring advantage, and won in some style
(R.Keene-C.Pritchett, British Championship,
Brighton 1972). ]

12.0-0

e5 13.a4 f4

[ I am rather shocked that my version of Fritz

actually wants to play the repulsive-looking
move 13...a5? here.
QUESTION: Well, it does keep the knight
from being driven back!
ANSWER: Yes, but at what cost? The
permanent weakening of b5 looks
absolutely horrible. I suppose it is true that
it is only one square, but even so, most
humans would rather cut off their hand than
play such a move. ]

14.a5

d7 15.c4 QUESTION: Why doesn't

White preserve his bishop?

[ ANSWER: That is an interesting question,

and one that shows where the computer
really can change assessments. Keene
rejects the bishop retreat, in view of the
variation 15.

c2 h4 16.g3 h3 17.gxf4

xf4 18.g4 e5! "with complications not
unfavourable to Black". However, 40 years
later and silicon-armoured, we can say that
this is not the case: after the further moves

19.

f6+ g7 20.xf4 Black is quite lost.

So it seems that there was no real objection
to retaining the bishop with 15 Bc2.
QUESTION: So is that what Nimzowitsch
should have done?
ANSWER: Well, not necessarily, in fact. The
truth is that in these Modern Benoni
structures, knights tend to be very useful
pieces, and it is not entirely clear to me
that the white bishop on d3 is better than
Black's rather active knight on f4. So,
although there seems no actual tactical

objection to 15 Bc2, I am not sure that
White needs to avoid the exchange of
bishop for knight anyway. ]

15...

xd3 16.xd3 f5!? Another point where

the computer has a few things to say about
the received wisdom on this position.

[ Keene identifies Black's 16th as Marshall's

real mistake in this game, and instead
advocates 16...

d4 as best. He admits that

White is still better after a line such as

17.

f4 ( 17.xd6?? e5 is the tactical

point of Black's 16th ) 17...

e5 18.xe5

dxe5 19.

b5 , but says the advantage

would be "of by no means decisive
proportions". ]

[ However, once again the silicon beast

points out something extremely inconvenient
for Black after 16...

d4 : instead of 17.f4

, ( White has the much stronger 17.

h6!

, with the nasty point that after

e8

White really can take on d6: 18.

xd6! e5

19.

b5 and the knight is indirectly protected,

thanks to the attack on the rook on e8. )]

[ This pretty much refutes 16...Bd4, as far as

I can see, and means that Marshall's 16...f5
is actually Black's best chance.
QUESTION: So what are we saying about
this position? Is Black OK?
ANSWER: I think what we are saying is
'Don't start from this position!'. Things
have already gone wrong for Black, possibly
as early as move 9.
QUESTION: But even if that is right, it is
hard to believe that he can be virtually lost
already!
ANSWER: Well, that is actually in the nature
of the Modern Benoni, I'm afraid. It is a
very risky opening, where Black is
constantly walking a tightrope. When it
comes off, the results can be very attractive,
but, as any tightrope walker can tell you,
one tiny slip is all that it needs to plunge to
the ground, with potentially lethal
consequences. ]

17.exf5

xf5 18.f4? A critical moment.

[ Nimzowitsch himself admitted that this was

a surprising move, "since 18.

e4

gave a good game without any effort". Even
so, he clearly still thought his move to be
objectively good, but the computer
suggests otherwise. 18 Ne4 would, indeed,
have retained the advantage, whereas after
the text, things could have changed
markedly. ]

18...

d4+ 19.e3 xc3 20.xc3 f6

QUESTION: Isn't the d5-pawn hanging?

[ ANSWER: Taking on d5 at once is not so

good, in view of 20...

xd5 21.f5! ( Keene's

21.

b3 f6 22.f5 transposes to the game,

which as we shall see, was probably not that
good for White ) 21...gxf5 22.

f4

, and White has a very strong attack.
QUESTION: Can't Black defend with

f6

trying to get queens off?
ANSWER: That is probably the best try, but
even the exchange of queens does not
really relieve his position. His log-jam of
undeveloped pieces on the queenside, plus
the awkwardly-placed rook on d5, still
causes him terrible problems after 23.

g3+

g7 24.xg7+ xg7 25.e3 , and Black
is busted. ]

21.

b3 xd5 QUESTION: What about the

other capture?

[ ANSWER: This would, indeed, have been

fine for Black. Nimzowitsch gave 21...

xd5

22.

ae1 with the idea of Bd2 and doubling

rooks on the open e-file. He claimed "Black
would then have been quite helpless", but
the ruthless silicon circuits of the computer
are having none of it. After b5! 23.

xb5

d7 24.b3 b8 25.d1 f6
Black seems to be doing fine (and if 26.g4?

xe3 ). ]

22.f5 gxf5? This is the decisive mistake.

[ Nimzowitsch does not even consider the

other capture on f5, but Keene analyses

22...

xf5 23.g5 d3 24.xb7 as "highly

unpleasant for Black". However, once again,
the computer refutes this with the incredible

e4! 25.a6 d7!! .
QUESTION: What? Black is giving up a
whole piece?
ANSWER: Yes. This variation is actually a
rather nice example of the power of
opposite-coloured bishops in the attack.
After either capture on f6, the counterattack
against g2 ensures Black enough play to
draw:

A) or 26.

xf6 g4 with the same finish,

( whilst even 26...

xg2 leads to the same

drawn result: 27.

xg2 g4+ 28.h1

e4+ , etc );

B) 26.

xf6 g4 27.f2 d1+ 28.xd1

xd1+ 29.f1 e2 30.f2 d1+
, and a draw. ]

23.

g5 Now, however, even the computer

agrees that White really does have a winning
attack. The immediate threat is 24 Ne3.

d4 Allowing a deadly discovered check, but

there is nothing better.

[ 23...

e6 24.xb7 c8 25.ae1 is lethal, ]

[ whilst 23...

g7 24.e3 e5 25.g4!

is a nice exploitation of the pin motif: fxg4

26.

xf6+ xf6 27.xf6 xf6 28.g8

and Black has no moves. ]

[ 23...

f8 seems to lose the least quantity of

material, but even here, after 24.

xf6 xf6

25.

e3 e6 26.xd5 d4+ (this queen

check is the point, reducing Black's net
losses to 'just' an exchange) 27.

e3 xd5

28.

xd4 cxd4 29.xf5 , the ending should

be winning easily enough. ]

24.

b6+ c4 25.c3 axb6 26.xd4 g7

27.

ae1 Black has lost an exchange, but,

even worse, he remains pinned and
hopelessly tied up, as the white rooks hover
over his king's head like buzzards circling a
corpse.

bxa5 Allowing an attractive final blow.

EXERCISE: How did Nimzowitsch finish off
the American no.1?
ANSWER: 28.

e8! Forcing mate.

xe8 29.xf6+ g8 30.h6

[ 30.

h6 f7 31.d8+ mates.

QUESTION: A nice final attack, but it looks
as though it was all unsound, as, indeed,
was much of the previous commentary on
the game!
ANSWER: Well, as Nigel Short once said,
"In such positions, the computer makes
fools of us all". It is inevitable that, when
one subjects such a sharp game to analysis
with a powerful modern engine, not only the
moves played, but even much of the
published analysis from the pre-computer
era will be seen to contain a surprising
number of mistakes. In exposing some of
them here, I certainly mean no disrespect
to either of the players, or to the annotators.
This was a splendid game, despite the
mistakes. ]

1-0

A36

Nimzowitsch,A
Asztalos,L

Bled

1931

[Steve Giddins]

1.c4

f6 2.c3 c5 3.g3 c6 4.g2 e6

5.

h3?! QUESTION: Rather a strange-

looking move!

[ ANSWER: Yes. Instead, the normal 5.

f3

d5 would lead to a Tarrasch QGD, or a

Pseudo-Tarrasch, depending on how Black
recaptures on d5. Nimzowitsch explained
the text move by a desire to avoid a
Tarrasch, which at that time was
considered satisfactory for Black. ]

5...

e7

[ 5...d5 is the most natural response, when

Nimzowitsch gives the variation 6.cxd5 exd5

7.

f4 d4 8.e4 xe4 9.xe4 g5!?

"with unclear play". It is all rather
unconvincing for White, in my opinion. ]

6.d3 d6 Settling for a quiet, English-style

position.

[ Black is very solid and has no objective

problems after this, but taking the centre
with 6...d5 would have been more
ambitious. ]

7.0-0 0-0 8.

f4 a6 9.b3 c7 10.e3 b8

11.a4 b6 12.d4 QUESTION: So what is the

assessment?
ANSWER: It is basically pretty equal. White
has a little more space in the centre, but
Black is totally solid and has a harmonious
development for his pieces.

b4 13.a3

[ QUESTION: I was expecting 13.

b2 .

ANSWER: That is a natural enough move,
but White wants to remove the enemy knight
from its active position, as we will see. ]

13...

b7 14.d5 e5 15.d3 Completing the

mini-plan begun at move 13.

xd3 16.xd3 be8 17.ae1 QUESTION:
Is White better here? He seems to have quite
a large space advantage.
ANSWER: He is certainly a little better, thanks
to the space, but I am not sure that it is an
enormous amount. Black is very solid still, and
if he can regroup his pieces sensibly, he
should be OK.

d7?! I am not hugely keen on this move.

[ It makes more sense to re-position the

queen's bishop with 17...

c8 . ]

18.e4

f6 QUESTION: What is Black up to?

ANSWER: He plans to double rooks on the e-
file.
QUESTION: But why? The file is closed!
ANSWER: It is actually a form of prophylaxis,
and Nimzowitsch himself applauded the idea.
Black sees that the main white break in this
position is f2-f4, and so he hopes that he can
exploit the e-file after this, by taking on f4.
However, the white e-pawn is defended very
securely, not least by the bishop at g2, so this
whole plan looks a little fanciful to me.

19.

c1 Nimzowitsch's regrouping, by contrast,

looks to be eminently logical. This move re-
routes the bishop to a sensible diagonal,
supporting the f4 thrust.

e7 20.d2 fe8 21.a1 QUESTION: But
what is this about? I thought White was
playing on the kingside?
ANSWER: This is another of Nimzowitsch's
famous 'mysterious rook moves'. Actually, it is
rather a nice little strategical and
psychological trap. The threat is a4-a5, by
means of which he hopes to induce Asztalos
to close the queenside with ...a5 himself.

a5? And he falls for it! This move rules out

forever a possible white break with a5, but it
also gives away the b5-square and abandons
any chance of counterplay on the queenside.
QUESTION: You obviously don't approve! But
we have already seen plenty of cases in this
book, where Nimzowitsch wins by alternating
the attack from flank to flank. Surely, by
closing the queenside, Black prevents that
and thus makes his defensive task easier?
ANSWER: That is a very interesting question.
What you have described is always a dilemma
for the defender, and it is often tempting to
shut one side of the board in this way, to
avoid having the threat of a 'second front'
breakthrough hanging over one's head. But
practice shows that the chance of counterplay
is usually more important.
Perhaps it is best summed up by Bent Larsen,
who has an extensive and hugely
enlightening discussion of this issue, in his
notes to Johner-Nimzowitsch, from 'Learn
from the Grandmasters'. Writing about a
similar situation in one of his own games,
Larsen says: "The point is, chess is not an
easy game, where you can avoid problems" -
Black must accept the unpleasantness of the
a4-a5 threat hanging over him, in order
himself to have a vestige of counterplay. As it
is, he puts all of his hopes on defence on the
kingside, which in practice rarely succeeds.

22.

ae1

[ The immediate 22.

b5 is tempting, but

Nimzowitsch sees no reason to rush with
this threat, and instead reverts to his
kingside build-up. ]

22...

d8 Black continues to shuffle around

rather aimlessly.

[ I think he would do better to play 22...g6

followed by ...Bg7, ...Bc8 and eventually try
to prepare ...f5. Nimzowitsch himself refers
to this plan in his notes, and soon takes

steps to prevent it. ]

23.h4

f8 24.f3 Nimzowitsch explains this

move precisely as aimed at the plan of ...g6
and ...Bg7.

a6 QUESTION: Why go to a6?
ANSWER: Black wants to have the option of
taking off the white knight, should it ever
come to b5, but again, I really don't like it. It
rapidly becomes clear that Asztalos has no
real idea what to do, and is just waiting for his
opponent to break through, when the battle
will become more concrete and tactical. The
trouble is, Nimzowitsch prepares so
meticulously for that moment, that, when it
finally comes, it is too late for Black to fashion
a defence.

25.

h3 a7 26.g2 e7 27.h1 a8

28.

d1 The knight has done its job on c3 and

now heads towards the kingside.

a7 QUESTION: This looks pretty ridiculous!
ANSWER: Black has run totally out of ideas,
so he just marks time, until White gives him
something concrete to think about. In the
meantime, though, Nimzowitsch is in no hurry.
In contrast to his opponent's pointless
shuffling, he has a carefully-thought out plan
in mind.

29.

e3 a8 30.eg1 a7 31.f1!

And here it is!
QUESTION: What is he doing?
ANSWER: Nimzowitsch is planning to
evacuate his king to the queenside, safely out
of harm's way, before breaking through on the
other wing.
QUESTION: But it seems very slow!
ANSWER: Maybe, but the fact that his
opponent has opted to sacrifice all semblance
of purposeful activity means that time is of no
consequence at all. Black can undertake
nothing, so White can afford half a dozen
moves to shift his king out of the way. There
is one other point to which I would draw the
reader's attention: the reason White's king is
so safe on the queenside, with no other
pieces there, is a direct result of the move
21...a5?. Imagine that Black still had his pawn
on a6, and had, say, put his bishop on d7.
Then White would have had to think twice
about moving his knight from c3, and sticking
his king over on the queenside, since it would
have been walking into counterplay by means
of ...b5. As it is though, the price of Black's
sealing up the queenside is that the white
king will now feel perfectly safe on that flank.

b8 32.e1 d8 33.d1 e7 34.c2

c7 White has completed his king evacuation,
and Black continues to shuffle around
aimlessly. All the preparations are now in
place, and White can proceed with the f4
break.

35.

c3 The bishop takes aim at e5.

ea8 36.e2 d8 37.f4! At last! Black has
kindly brought both his rooks to the closed a-
file (!), so this thrust gains in strength. The e5-
pawn is attacked, and exchanging on f4,
opening the g-file, is clearly too awful to
contemplate, so Asztalos shores up the pawn
as best he can.

f6 Horribly weakening the light squares, but

there is little else.

38.

e6+ xe6

[ If Black declined the bishop by 38...

h8

, then Nimzowitsch intended 39.f5
, followed by g4-g5, with a crushing
position. ]

39.dxe6 Now the d5-square beckons the white

knight, so Black immediately covers it.

b7 40.d3 e8 41.f5 Defending the e6-
pawn, and setting up the planned
breakthrough on the g-file with g4-g5. Black's
position is quite hopeless.

c6 42.g4 h8 43.g3 b8 44.hg1

ab7 45.d2! QUESTION: Yet more
preparation?
ANSWER: The move has two points, one
positional, one tactical. The positional point is
that, since the move 41 f5, the bishop now
bites on granite on e5, so it makes sense to
redeploy it to a square from which is attacks
g5.

d8 The tactical point of White's last is to
anticipate Black's potential threat of a
sacrifice on b5, trying belatedly to get some
counterplay.

[ Now after 45...b5 46.axb5

xb5 47.cxb5

xb5 , the b3-pawn is defended along the
third rank, so White can just play 48.

c4

, and Black's counterplay is stymied ( a4

49.bxa4 and the knight defends b2). ]

46.

d1 QUESTION: Now what? So many

preparatory moves! Why doesn't he just get
on with the g5 break?
ANSWER: It is the same theme as the whole
game - there is no hurry. Nimzowitsch pays
attention to every possible shred of enemy
counterplay, and takes prophylactic measures.
The knight now heads back to c3, so as to kill
off once and for all time any possible break
with ...b6-b5. Black cannot stop the
breakthrough on the g-file anyway.

bb8 47.c3 dc8 48.e2 b7 49.g5
Now everything is ready for the final push.

d8 50.g6 f8

[ QUESTION: Why not 50...h6 , trying to

keep things closed?
ANSWER: Then we see how good
Nimzowitsch's preparations have been.
There follows 51.

xh6! (now we see why

the bishop came back to d2) gxh6 52.g7+

g8 53.g6 followed by Qh5 (and we see
why the queen retreated to e2 at move 48),
with decisive threats. ]

51.gxh7

xh7 52.g6 White has numerous

ways to win: for example, pushing the pawn to
h6 (as mentioned by Nimzowitsch), or Qg4,
followed by a rook sacrifice on f6 (the
computer's suggestion). A superb example of
manoeuvring and prophylaxis.
1-0

Back to Contents Page

Chapter Five

Nimzowitsch in the Endgame

The endgame is one of the touchstones of chess mastery, and it is practically inconceivable that one could have a
great master of the game who was not also a great endgame player. Nimzowitsch was certainly no exception to this
rule, and played some wonderful endgames. This chapter presents five examples, each showing different aspects
of endgame play.

Nimzowitsch-Pritzel

is more of a queenless middlegame than a pure endgame, since the queens depart from the

board in the opening. It shows an interesting exploitation of the open file.

Nimzowitsch-Tarrasch

is another where

simplification takes place very early, and Nimzowitsch goes on to outplay one of the game's greatest masters (and
also his own most bitter personal rival) in highly instructive style. Then

Von Gottschall-Nimzowitsch

is one to show

anyone who thinks that opposite-coloured bishops mean an automatic draw. Nimzowitsch gradually outplays his
opponent from a position in which many players would have shaken hands.

Nimzowitsch-Reti

is in some ways my favourite of all the endgames in this book. After mishandling an overwhelming

advantage as White, Nimzowitsch finds himself swindled into what looks as first sight to be a drawn ending with
rook and bishop versus rook, but he then brilliantly exploits the fact that Black has a couple of extra pawns, to show
how this actually rebounds to his disadvantage. The most impressive thing of all about this ending is that the key
move, declining to remove Black's pawns from the board, was the sealed move, so Nimzowitsch had to see the
main point of the endgame at the board, rather than finding it in home analysis.

Finally,

Nimzowitsch-Voellmy

is a perfect illustration of a Minority Attack, where Nimzowitsch succeeds in gradually

weakening Black's queenside pawns.

One last point I should make concerns the selection of games for this chapter. The more erudite amongst my
readers may be surprised to see that Nimzowitsch's two most famous endgames, each of which could reasonably
be considered to be a sine qua non of any treatment of Nimzowitsch's endgame prowess, are both missing from this
book. I refer to his knight and pawn ending against Emanuel Lasker, from Zurich 1934, and his immortal good
knight v bad bishop ending against Henneberger from Winterthur 1931. The explanation for the omission is that
both of these endings were analysed deeply in my earlier Everyman book, The Greatest Endgames Ever Played. I
did briefly toy with the idea of reproducing them here anyway, but eventually decided that this would run the risk of
short-changing those most loyal amongst my readers, who have purchased both books, something I preferred not
to do.

Back to Contents Page

B07

Nimzowitsch,A
Pritzel,A

Copenhagen

1922

[Steve Giddins]

1.d4 g6 2.e4

g7 The Pirc/Modern Defence

was an extremely rare bird in those days, and
on the few occasions when it did appear,
tended to be labelled as 'Irregular', if not
downright 'Incorrect'. Interestingly, the present
game shows a quite modern treatment by both
players.

3.

c3 d6 4.e3 f6 5.e2

[ 5.

d2 would be the contemporary

approach, intending an aggressive build-up
with f3, 0-0-0, and g4. Nimzowitsch prefers
a slightly more careful set-up, but one which
is not totally without venom. ]

5...0-0 6.

d2 e5

[ 6...c6 would be an alternative treatment,

but there is nothing wrong with Black's
move. ]

7.dxe5 Opting for a technical handling of the

variation. White hopes to profit from the open
d-file.

dxe5 8.0-0-0

xd2+?! QUESTION: Is the

queen exchange forced?

[ ANSWER: No. 8...

e7 is probably stronger.

The text plays into White's hands a little,
but Black probably thought simplification
would enhance his chances of a draw. ]

9.

xd2 QUESTION: So how would you

assess this position? It looks very equal and
drawish.
ANSWER: Actually, it is more dangerous for
Black than first appears. He is slightly behind
in development, and must therefore be very
careful not to lose control permanently of the
open d-file. In fact, this is the theme of the
entire game, and Nimzowitsch uses it in 'My
System', as an archetypal demonstration of
the exploitation of an open file. Black's other
slight problem is the passivity of his king's
bishop, which is locked in behind the e5-pawn,
and does not contribute to the battle on the
queenside at all. By contrast, White's bishops
are both pointing at the enemy queenside, and
as the game proceeds, we will see this factor
in action.

c6? QUESTION: Why is this wrong? It looks

very natural to guard the squares b5 and d5,
against the enemy knight.
ANSWER: Yes, it does, and ...c6 is a typical
move in such structures, for precisely the

reasons you outline. However, the move also
weakens the d6-square severely, and Black
must be very careful about doing this, if he is
not in a position to protect that square, and
the d-file more generally. In the present case,
he is not able to do so, and soon regrets his
last move.

[ Nimzowitsch himself indicates 9...

c6

as the best move, which is probably correct,
He continues the variation with 10.h3
(planning Nf3, without allowing ...Ng4)

A) an interesting pawn sacrifice, to shake

off the chains; Nimzowitsch actually
prefers 10...

e6 11.f3 h6 12.hd1

( but the computer's 12.

b5 is stronger

here, and leaves Black in some trouble,
due to his vulnerable e5-pawn ) 12...a6
and claims Black is holding the balance,
thanks to his pressure on e4 and the plan
to oppose rooks on the d-file;

B) 10...

d4!? 11.f3 ( 11.xd4? exd4

12.

xd4 g4! is good for Black )

11...

xe2+ 12.xe2 , and White has only

a small advantage. I actually think this is
the best line for Black, and the approach
of using the dynamic resources of his
position, with ...Nc6-d4, etc, is usually
the correct one in such structures. Instead,
Pritzel's play is too passive, and he
gradually falls into a vice-like grip. ]

10.a4! QUESTION: What is the idea? Is he

afraid of ...b7-b5?
ANSWER: That is part of the idea, yes, but
there is a much deeper, more positive purpose
than this simple prevention of the ...b5
advance. It is the start of what Nimzowitsch
termed "a siege" of the black queenside.
White plans to push the pawn to a5 and then
use the c5-square for his pieces. Another
possible idea is to put a rook on d6 and then
undermine the black queenside with a5-a6.
To modern eyes, a move like 10 a4 seems
almost a reflex in this structure, after Black
has played ...c6, but of course, this was not
the case in Nimzowitsch's day.

g4 Black induces the exchange on g4, thus
securing the two bishops, but they are not
especially active in this position, so this is not
big deal. On the other hand, it is not easy to
suggest a good alternative for Black.

11.

xg4 xg4 12.ge2 d7?!

[ Nimzowitsch suggests 12...

a6

, planning ...Rfe8 and ...Bf8, to protect the
d6-square. Black's position would indeed be
pretty solid then, although White remains

better. ]

13.

hd1 Now the attack on the d7-knight does

not allow time for such a set-up by Black, and
the knight is driven to a bad square.

b6 14.b3 QUESTION: Taking c4 away
from the knight?
ANSWER: Exactly, and so preparing to drive
the poor beast back with a5. The pawn on b3
effectively 'dominates' the knight, and is a set-
up worth remembering.

f6 QUESTION: What is this for?
ANSWER: Black wants to challenge the rooks
on the d-file, but he will not get the chance, as
his minor pieces are now driven back in
disarray. Once again, though, there is no
obviously superior alternative. Exchanging on
e2 would leave the d7-square available for
the white rooks, after a5.
QUESTION: Black's position seems to have
become very difficult, within just a few moves,
and without his making any obvious mistake?
ANSWER: Indeed. This is what can happen in
these symmetrical structures, with a single
open file. One or two small inaccuracies can
result in White seizing control of the open file,
and suddenly his grip becomes really
paralyzing. Many less experienced players
tend to look at the outwardly simplified and
symmetrical position, and just assume it is
drawish, but this is frequently a serious
misjudgement.

15.f3

e6 16.a5 c8 17.a4!?

Now we see the effects of the plan begun by
10 a4. The black queenside is indeed
'besieged'. The bishop at e3 rakes the dark
squares and attacks a7, and the a5-pawn
makes it hard for Black to cover c5 by playing
...b6. The white knight now threatens to
invade on c5. However, it turns out that a
tactic allows Black actually to play ...b6 after
all.

b6!

[ Black exploits a tactical trick to get in his

desired advance; if White takes on b6, then
after 17...b6 18.axb6 axb6 19.

xb6 xb6

20.

xb6 , there is g5 , winning the

exchange.
QUESTION: So White has messed up?
ANSWER: Not necessarily. I am not sure if
Nimzowitsch had really intended to allow
this trick – probably not, I suspect – but as
it happens, Black's getting in ...b6 is not
the end of his troubles. After White's next,
b6 really is hanging, and Black's queenside
is still a target. ]

18.

d3! This reply avoids the pin which occurs

in the previous note, and so really threatens
to take on b6.

[ QUESTION: But why does it deserve an

exclamation mark? The same effect could
have been achieved by a king move, such
as 18.

b1 .

ANSWER: That is correct, but Nimzowitsch's
move embodies an additional motive. As
well as avoiding the pin on the c1-h6
diagonal, he also now has ideas of bringing
his rook to c3, to attack the c6-pawn that
way. ]

18...bxa5? Black loses his nerve and

irreparably damages his queenside pawns.

[ Of course, he had to defend his position

with 18...

b8 , and then see what White

intended to do next.
QUESTION: And how bad is the position
after that?
ANSWER: It is still very uncomfortable, but
there is no obvious way to increase the
pressure. White would probably continue

19.

ec3 ( 19.c3 c5 20.cd3 , hoping to

follow up with Nec3-b5, is met by c4 )
, but after 19...

e8 Black is holding. His

minor pieces cover the entry squares on the
d-file at d6, d7, and d8, and his queenside
pawns are all defended. White would need
another idea to strengthen his position. One
possibility is 20.axb6 axb6 21.

a2

, with the idea of bringing the knight to b4,
hoping to provoke ...c5 and weaken the
light squares. But then Black can defend
with

e7 , and there is no immediate

danger. All in all, 18...Rb8 would enable
Black to put up stiff resistance. ]

19.

c3

[ 19.

c5 looks even stronger, but White

already has a pleasant choice of good
moves. ]

19...

e7?!

[ The computer points out the more

tenacious alternative 19...

b6! , when

20.

c5 fe8 is still better for White, but

there is nothing decisive. ]

20.

c5 Now a5 drops off and Black is left with

two very weak queenside pawns.

fb8 21.ec3 a6 Getting the a-pawn away
from potential attack after Rxa5, but now the
weakness of the b6-square proves fatal.

22.

xa5 g7 23.b6 a7 24.ca4

The weaknesses of the black queenside are
now really being exploited to the full, as the
white knights jump all over his position. There

is no defence.

ab7 25.xa6 c8 26.xc8 xc8 27.c5

bc7 28.d6 d8? 29.xe6!
A nice illustration of how to seize and exploit
an open file in such structures. It is a pity
Black's serious inaccuracy at move 18 made
things so much easier for White, as it would
have been very interesting to see how
Nimzowitsch set about breaking down the
black defences after the correct 18...Rb8.
1-0

D41

Nimzowitsch,A
Tarrasch,S

Breslau

1925

[Steve Giddins]

1.

f3 f6 2.c4 c5 3.c3 d5

[ One might have expected Tarrasch to

prefer 3...e6 , angling for a transposition
into his eponymous defence to the Queen's
Gambit. ]

4.cxd5

xd5 5.d4 This strike in the centre is

one of three reasonable moves here.

[ The quiet 5.g3 leads to typical English

positions, ]

[ whilst the pugnacious 5.e4 is the most

popular move nowadays, and results in
some sharp and heavily-analysed lines. We
saw this line in Nimzowitsch-Rubinstein. ]

5...cxd4

[ Black has a major alternative here in
5...

xc3 6.bxc3 , and now e6 leads to a

Semi-Tarrasch, ( and 6...g6 to a Grünfeld
(or pseudo-Grünfeld type position, if White
avoids 7.e4 ). )]

6.

xd4 e6 QUESTION: This is surprising.

Why give White the chance to inflict an
isolated pawn on Black?
ANSWER: As we will see, Black is not afraid
of the IQP here.

[ However, he could also have played
6...

xc3 7.xc3 ( the exchange of queens

with 7.

xd8+?! xd8 8.bxc3 is not very

attractive for White, as any inconvenience
caused by Black's loss of castling rights will
be minimal, in the absence of queens,
whereas the weak white queenside pawns
are a more permanent drawback ) 7...

c6

8.e4 . However, modern-day theory regards

this as better for White (mainly because
Black has trouble developing his kingside),
although it requires very accurate play to

prove this. ]

7.e3 Rather a modest choice.

[ 7.e4 is more ambitious, and would

probably transpose into the previous note
after

xc3 8.xc3 c6 . ]

[ QUESTION: So why not give him the IQP?

ANSWER: After 7.

xd5 exd5 White will

have to lose a tempo with his queen after
a subsequent ...Nc6, and he really does not
have anything here. ]

7...

c6 8.b5 d7 9.xc6 QUESTION:

Nimzowitsch giving up his bishops again! He
really was rather anti-clerical, wasn't he?
ANSWER: In general, he was, although this
instance is rather different from some of the
others we have seen in this book. Here, he is
not really giving up the bishop-pair, since he
intends to follow up with Ne5, and take the
enemy bishop in turn. He wants to maintain
his queen on its central square, and has also
decided that the resulting simplifications are
not as harmless for Black as they at first
appear.

xc6 10.e5 xc3 11.xc6 xd4

[ Nimzowitsch does not mention it in his

notes in 'My System', but the computer
points out the interesting possibility

11...

g5!? here. After the practically forced

line 12.

e5 ( 12.xc3? xg2 wins for

Black ) 12...

d6! 13.bxc3 ( 13.xd6 d8

wins, since if the queen moves, there is
mate on d1 ) 13...

xe5 , Black is

comfortably off. ]

12.

xd4 d5 13.d2 QUESTION: So what

is going on here? The position looks totally
dead equal.
ANSWER: Nimzowitsch comments that,
although this may be the appearance, the
position is "full of poison". Such symmetrical,
queenless positions tend to look very equal at
a superficial glance, but development is
frequently a key factor. If White manages to
seize control of the open c-file, then he can
often turn out to have a paralyzing grip on the
game. Because of this, such simplified
positions often require extremely accurate
handling.

c5 Nimzowitsch actually criticizes this, but it
is not clear that he is correct.

[ He gives as preferable the variation
13...

e7 14.e4 b6 ( 14...f6 15.f3 0-0

looks a better try, when things are pretty
equal ) 15.

c1 0-0 16.e2 f6 17.e3

fc8 18.b3 xd4 19.xd4 , and claims
Black is fine, but this seems like one case

where his myopia towards the strength of
bishops against knights affected his
judgement. Bobby Fischer spent much of his
career winning such positions, with the
advantage of rook and bishop versus rook
and knight on an open board - just ask
Mark Taimanov, among others. This position
looks pretty unpleasant for Black. ]

14.

b3 b4

[ Nimzowitsch recommended 14...

b6

as "decidedly better", and it is probably
true that the move is superior to the game
continuation, though perhaps not by that
much.
QUESTION: What is wrong with the text?
Surely simplification helps Black?
ANSWER: Not necessarily. The problem
with exchanging bishops is that the dark
squares in Black's camp are weakened,
especially squares such as c5 and c7. On
b6, the bishop patrols these squares
effectively. Even so, the text is not really a
mistake. ]

15.

c1 d8 16.xb4 xb4 17.e2 e7

QUESTION: Isn't the a-pawn hanging?

[ ANSWER: Not really, After 17...

xa2?

18.

c7 ( 18.c4 b5 is less effective )

, White has the advantage: for example,

18...

d7? ( 18...0-0 is better, but still pretty

unpleasant after 19.

a1 b4 20.xb7 )

19.

c4! , when the knight unexpectedly finds

itself cut off in enemy territory. The point is
that now b5 is met by the check 20.

c8+

, forcing

d8 21.xd8+ xd8 22.a1 b4

23.

xa7 f8 24.d4 and Black is just

busted - his b-pawn is lost and his pieces
all hopelessly passive, compared with their
white counterparts. ]

18.

c4 a6?! QUESTION: Why retreat to the

edge of the board?
ANSWER: Good question. I think Tarrasch
was worried about the white knight jumping to
c5, and also wanted to protect the c7-square
against penetration by the enemy rooks.
However, it is obviously unattractive to put the
knight on the edge like this, and, indeed, the
knight never really recovers from this offside
position, as we will see.

[ Nimzowitsch claimed White was still better

after 18...

c6 19.hc1 ( on the other hand,

Nimzowitsch dismissed the immediate

19.

c5 because of a5 followed by ...b6,

but even so, after 20.

c3 b6 21.d3 c8

22.

hc1 xc3 23.xc3 the control of the

open c-file gives White some advantage )

, but it does not seem to amount to a great
deal after 19...

d7 20.c5 c7 21.xb7

xb7 22.xc6 xb2+ , etc, which is
nothing. ]

19.

hc1 QUESTION: So White has

conquered the open c-file?
ANSWER: Yes, and that is certainly an
achievement.
QUESTION: How much of an advantage does
he have?
ANSWER: He is somewhat better, but it is far
from decisive.

d7 20.f4! QUESTION: What is the idea of
this?
ANSWER: It is another piece of
Nimzowitschian prophylaxis. He sees that
Black is going to double rooks on the other
open file, the d-file, so he takes steps to
neutralize the counterplay along that file.
QUESTION: How does 20 f4 contribute to
that?
ANSWER: The idea is to plug up the d-file
with Nd4; 20 f4 stops Black from attacking the
knight with ...e5.

hd8 21.d4 f6 QUESTION: But now Black
is ready for ...e5 anyway, isn't he?
ANSWER: Maybe, but at a significant cost; ...
e5 will now entail taking on an isolated e-
pawn after the exchange on e5.
QUESTION: OK, but White will also have an
isolated pawn on e3.
ANSWER: Yes, but his pawn is easier to
defend than Black's. The e5-pawn will be
further from its home base, and Nimzowitsch
intends to bring his king to e4, attacking it. It
is the old story – a weakness is only that if it
can be attacked.

22.a4!? QUESTION: Now what is he doing?

ANSWER: This is another very instructive
moment. White continues his policy of
restricting his opponent and gaining space.
He intends to drive the knight back further by
b4-b5, which will also deprive the invasion
square c7 of a defender.

[ QUESTION: OK, I understand the idea, but

why not just play 22.b4 at once?
ANSWER: He should have done that, in fact.
Nimzowitsch mentions 22 b4!, and explains
that he feared the defence b5?
(hence White's preparatory 22 a4), but this
seems to be simply a tactical miscalculation:
after 23.

c6+ e8 24.xd8 bxc4 25.c6

White wins the pawn on c4 and still has a
firm grip on the game. ]

22...e5!? QUESTION: Don't you like this?

ANSWER: Not really, but this is the classic
defender's dilemma. If he does nothing, he will
be criticised for passivity, whilst if he strikes
out, as here, in an attempt to get some
freedom at the cost of a weakness, he gets
criticized for wantonly weakening his own
position!

[ The alternative was passive waiting, by

means of 22...

f7 23.b4 b6 . Now 24.b5

is met by

c5 .

QUESTION: So hasn't White's plan of b4-b5
been effectively refuted?
ANSWER: Certainly not. In order to prevent
the immediate b5 advance, Black has been
forced to weaken the c6-square, which is
now a juicy target for the white knight. White
can play a nice preparatory move such as
Kf3, just getting out of the way of a rook
check on d2, and then Black's position is
really nasty. He has absolutely nothing
active worth undertaking, and must
constantly worry about Nc6, or Rc7, or Rc8,
etc. White can also start gaining space on
the kingside, by pushing his g- and h-pawns.
Objectively, the computer may say Black is
not losing and can defend with best play, but
in practice, it is a desperately unpleasant
and difficult task to do so over the board. ]

23.fxe5 fxe5 24.

f3 So now, for the first

time in the game, White has a clear weakness
to attack.

e6 25.b4 b6 26.1c2! QUESTION: What
is the point of this?
ANSWER: Nimzowitsch was quite proud of
this move, and rightly so, although to a
modern technician, it is quite natural – Ulf
Andersson would play the move in his sleep, I
am sure. The idea is simply prophylactic again
– White just over-protects the d2-square, thus
preventing a rook check, should his knight
move.
QUESTION: It does not look terribly fearsome!
What is all the fuss about?
ANSWER: The point is that such little
consolidating, prophylactic moves are what
these technical positions are all about. It may
look insignificant, but such attention to detail,
and careful preparation, before undertaking
anything active, makes all the difference
between winning such positions and not doing
so.
QUESTION: OK, but what is White actually
going to do to make real progress?
ANSWER: Well, Nimzowitsch may well not
have had any definite idea in mind. Much will

depend on how Black reacts. Nimzowitsch
himself admitted that his last move was
partially a waiting move, and psychologically
very unpleasant for the defender - it is often
much easier to defend against direct threats
than to find a decent move in such quiet
positions, where there are no obvious threats.
Having said all that, the move does contains
one possible idea, which Nimzowitsch
mentions, namely, Ng5-e4. That would cover
the c5-square, and thus reintroduce the threat
of b5, driving the knight back. In the meantime,
Black is invited to find a move, which does
not worsen his position further.

h6 QUESTION: Ha! He found one!

ANSWER: Oh yes?

27.h4! White immediately reacts to Black's

last move. Far from being a totally neutral
waiting move, 26...h6 weakens the g6-square.
White pounces on that, and intends to fix it
with h5 next move.
QUESTION: But what relevance is the g6-
square? Surely there is no real threat to
penetrate there?
ANSWER: Not yet, no, and maybe there will
never be any such threat. But there may be.
The point is that it is one more small
advantage, to add to White's collection. This
is how such positions are played - the
stronger side seizes every small advantage
that he can. One never knows which will prove
to be the straw that eventually breaks the
camel's back.

d6 28.h5 d5 29.g4 Already, we see the
fruits of the last couple of moves - the black g-
pawn is now fixed backward, and is a
weakness. Admittedly, it is not on an open file,
so the weakness is less easy to attack than it
might otherwise be, but a weakness it remains,
and Nimzowitsch uses the fourth rank to
manoeuvre his rook over to attack it.

5d7 30.c6+! A highly unpleasant check,
which leaves Black with a difficult decision.

d6

[ Of the two king moves, 30...

f5?

walks into a mating net at once - after

31.

cg6 there is no effective defence to 32

e4 mate, since e4 allows 32.

f4# . ]

[ 30...

d5 is stronger, but also runs the king

into trouble. Nimzowitsch gives

A) however, the computer finds a still more

effective riposte: 31.

c1! , and if e6

then 32.

g6+! f5 ( or 32...d5

33.

d1+ e4 34.d2+ and Black must

surrender the exchange to avoid mate )

33.

c4 followed by g4 mate;

B) 31.

cg6 e4 32.d2 , which is still

better for White. ]

31.

g6+!

[ QUESTION: Surely White can win a pawn

with 31.

xd6+ xd6 32.xg7 , can't he?

ANSWER: Yes, but this might allow a
smidgeon of counterplay, such as

xb4

33.

xa7 g8 , etc. Instead, Nimzowitsch

finds the most precise way of winning
material. ]

31...

e7

[ Forced, since 31...

d5? loses at once:

32.

cxd6+ xd6 33.e4+ c6 34.b5+

, etc. ]

32.

xg7+ f8 33.xd6 xd6 34.xa7

xb4 35.xe5 Thus, White has won two
pawns, and the black king is more passive
than in the variation in the note to move 31.
Even here, though, some care is needed to
overcome Black's last line of resistance.

e6 36.g6+ g8 37.e7+ f8 38.f5
The knight is ideally placed on f5, attacking
the h6-pawn and cooperating with his own
rook.

d5 39.g4 f4+ 40.f3 d3 QUESTION:
Surely Black should just resign here, shouldn't
he?
ANSWER: Possibly, yes, but even so, it is
worth observing the patient, methodical way
in which Nimzowitsch converts his advantage.

41.

a8+! Another nice little piece of precision.

[ Black's defensive idea was 41.

h7 e5+

42.

f4 f7 , defending the h-pawn and

leaving White's rook boxed in.
Nimzowitsch's little subtlety avoids this. ]

41...

f7 42.h8 Now the black king occupies

the key f7-square, so the h6-pawn is doomed.

c5 43.h7+ Once again, the utmost
accuracy. Rather than take on h6 at once,
Nimzowitsch first drives the enemy king one
square further back. It does not actually
change the result of the game, in that taking
on h6 immediately would still win, but why not
make things as easy as possible? Seize every
small advantage!

g8 44.xh6 xh6 45.xh6+ f8 46.f5

xa4 Black at last has some tiny glimmer of
counterplay, in the shape of a passed b-pawn,
but it is easily dealt with.

47.h6

g8 48.g5 h7 49.g4 c5

50.

h5 e6 51.g6+ g8 52.h7+ h8

53.

h6

1-0

C01

Von Gottschall,H
Nimzowitsch,A

Hannover

1926

[Steve Giddins]

1.e4 e6 2.d4 d5 3.

d3!? Hardly the most

terrifying response to the French, but a
perfectly playable move in itself.

c5

[ 3...dxe4 4.

xe4 f6 is another very

sound reply. ]

4.c3 dxe4 5.

xe4 f6 6.f3 bd7 7.e2

e7 8.0-0 0-0 9.e3 cxd4 10.cxd4!?
QUESTION: Is it wise to take on an IQP?
ANSWER: Objectively, maybe it is reasonable,
but it is certainly not the most dynamic of IQP
structures.

[ The knight recapture 10.

xd4 would avoid

a long-term weakness. ]

10...

b6 11.bc3 d7 QUESTION: That

looks a bit odd!
ANSWER: Yes, one would not usually
contemplate putting the queen on d7 in an
IQP structure. But this case is different,
because the white king's knight is on e2,
rather than the usual f3, so there is no threat
of Ne5. Black intends to bring his rook to d8.

12.

c1 d8 13.b3 fd5 14.xd5 xd5

15.

xd5? QUESTION: What is wrong with

this?
ANSWER: It is a classic case of a player
playing for a draw, by hoovering off the pieces.
The trouble is, in a structure with an IQP,
mass exchanges are usually the last thing the
player with the IQP needs. He should play to
the strengths of the structure, which are the
extra central space and chance for an initiative.
In an endgame, the pawn is liable just to be a
weakness.

[ As it happens, White would stand well after
15.

c3 , fighting for the d5-square.

QUESTION: Wasn't he afraid of the
exchange on e3?
ANSWER: That may well have been what
von Gottschall was afraid of, but he needn't
have been. That exchange would strengthen
the white centre and open the f-file. He
would be fine. We saw a similar structural
transformation in Nimzowitsch-
Taubenhaus. ]

15...

xd5 16.xd5 xd5 17.c3

[ White cannot exploit the open c-file, since
17.

c7 is met simply by d6 . ]

17...

a5 18.fd1 b4 QUESTION: What is

this for? Surely Black would prefer to keep his
bishop-pair?
ANSWER: Possibly, but White was
threatening to eliminate his only weakness
with d4-d5, so Nimzowitsch has to surrender
one of the bishops, in order to prevent this.

19.a3

xc3 20.xc3 d7 21.c5

Still in hoover mode. White sees an active
enemy piece and seizes the chance to
eliminate it, also offloading his IQP in the
process.

xc5 22.dxc5 c6 QUESTION: So, what is
the assessment? It looks like a dead draw!
ANSWER: In his notes, Nimzowitsch gleefully
quoted that opinion, in order to knock it down:
"No, there is still a great deal in the position,
and the play has still to begin".
QUESTION: But, equal pawns, and opposite-
coloured bishops?
ANSWER: True, but the presence of the
rooks reduces the drawing tendencies of the
opposite bishops. In addition, the pawn
structure is asymmetrical: White has a
queenside majority, Black a kingside one.
QUESTION: Yes, but isn't the queenside
majority supposed to be an advantage?
ANSWER: I must remind you that is one of the
great shibboleths of the game. A queenside
majority can be an advantage, but only in the
right circumstances, when the majority is
mobile and is well supported by its pieces,
such as in Nimzowitsch-Brinckmann. Here,
the advance of the pawn to c5 already leaves
the white majority rather blockaded (it will be
tough ever to get a pawn to b5), whereas the
black majority on the other wing is not
committed, and can advance more readily.
QUESTION: So is Black better?
ANSWER: No, that would certainly be going
too far and 'annotating by result'. The position
is just equal, but 'equal' is not the same as
'drawn', and there is plenty of scope to play
on and try to make something of one's
position.

23.f3 f6 24.

f2 f7 25.d4 a5 26.g3?!

This is probably White's first misstep. By
allowing Black to fix the queenside pawns with
...a4, he effectively gets his c5-pawn isolated.

[ Nimzowitsch advised 26.b3

d5 27.d3

, when "White's position would have been
difficult to get at". ]

26...a4 27.f4 h5 28.h3

h8 QUESTION:

What is the idea? Does he hope to open the h-
file by ...h4?
ANSWER: Not really. Rather, this is

prophylaxis, anticipating White's playing g4.

29.

d1 g6 30.d4 f5 By holding back

g4, Black has secured his king a nice active
square on f5.

31.

d2 f8 QUESTION: Another funny

move!
ANSWER: Nimzowitsch is preparing to
advance ...e5, which will create a passed
pawn in the centre, and open the f-file; hence,
he puts his rook on that file in anticipation.

32.

e1 e5 33.fxe5 fxe5 So, Black's majority

has yielded a passed pawn, whereas the
white queenside majority seems unlikely ever
to do so.
QUESTION: So Black is better?
ANSWER: He is slightly better, yes, although
it is still not a huge amount. But Black is the
one making progress. The grind is on!

34.

h4? Presumably missing the reply,

although it is pretty obvious.

g5! 35.

b4

[ Of course, 35.

xh5?? drops a rook to

g6+ (incidentally, justifying Black's 31st
move).
QUESTION: But could a master-strength
player such as von Gottschall really miss
something so trivial?
ANSWER: I agree, it is surprising, but if he
didn't, then it is hard to understand why he
would play 34 Rh4. If he did miss it, then
the explanation may lie in a simple failure
to appreciate that there could be any danger
in the position.
One of the things which can happen when a
player is playing for a draw (as von
Gottschall seems clearly to have been doing
since move 1) is that once a seemingly
'dead' position is reached, such as this
endgame, the defender can just switch off
his attention and stop looking for danger.
This is especially true of tactical danger. A
lot of players make the mistake of assuming
that such quiet, technical positions do not
contain any tactical elements, but nothing
could be further from the truth - one only has
to look at the wealth of stunning tactics
that endgame studies reveal, often in
highly simplified positions, to see that.
Telling oneself that the position is a dead
draw and the opponent can do nothing, is a
sure way of losing such a position. ]

35...

e6+ 36.e2 e4 37.f2?!

White continues to mark time, while Black
gradually inches forward.

[ The white rook is doing nothing on the b-file,

so it probably makes more sense to play

37.

d4 , cutting off the enemy king from the

queenside, and preparing possible
harassing tactics with Rd6+ or Rd8-g8, etc. ]

37...

f3 QUESTION: It is obvious that Black

has made a lot of progress, but does he really
have a serious basis to play for a win? After all,
the c5-, b2- and g3-pawns are all easily
defended.
ANSWER: True, but the h3-pawn is also
vulnerable (and on a light square, so not able
to be protected on that square by the white
bishop). Furthermore, such widely-spaced
weaknesses are always a problem, as we
have seen repeatedly in this book.
Nimzowitsch was a past master at switching
the attack from one weakness to another.
Indeed, in his book 'Chess Praxis', he has a
whole chapter on the subject of 'Alternation',
this being one of the games featured.

38.

b6 e5 39.b4 d5 40.h4

White is in a kind of zugzwang.

[ Nimzowitsch gives the variation 40.

b6 h4!

41.gxh4 gxh4 42.

xh4 xc5 , gaining a

tempo on the rook, followed by ...Rxh3, and
Black has won a pawn (although it would still
not be so easy to win the game). Von
Gottschall prefers to avoid this and eliminate
his h3-weakness, but at the cost of freeing
up space on the kingside, for the enemy
king. ]

40...gxh4?

[ Rather amazingly, Nimzowitsch does not

even comment on this, but the alternative

40...g4! looks extremely strong. White would

again be in virtual zugzwang: for example,

41.

b6 ( if 41.d4+ xc5 and there is no

effective discovered check, since 42.

xe4+?

loses to

xf2+ ) 41...e3! (this is the key idea

behind 40...g4: Black trades the e-pawn for
the g3-pawn, obtaining a very strong passed
g-pawn in the process) 42.

xe3 ( or 42.e1

xc5 ) 42...xg3 43.b4 h3 , and Black
looks to me to be simply winning. ]

41.gxh4

h3 42.d4+ e5 43.d8 d5

QUESTION: It still looks good for Black!
ANSWER: It is clear from Nimzowitsch's notes
that he thought he was winning this position.
Indeed, he says the win "is not too difficult
now", but I am not convinced of this.

44.

e8+ e6 Now 45...Rb3 is threatened.

45.

d8 f4 46.f8+ f5 47.f7 h2

And now 48...e3 is the threat. The basic plan
is eventually to drive the bishop away from its
defence of h4.

48.

e7? Again, no comment from

Nimzowitsch, but this is a serious mistake.

[ The computer shows the much tougher

defence 48.

f1 : for example, g4 ( not

48...e3? 49.

xe3+! ) 49.e2 , and it is not

clear how Black can make any progress. ]

48...

g4+ 49.e1

[ Now 49.

f1? loses to h1+ 50.g1 ( if

50.

g2 f3# ) 50...f3 51.f7+ g3

52.

e7 h3+ . ]

49...

f3? 50.f7+ g2 51.d2?

And this finally condemns White to defeat.

[ It was still possible to put up stiff resistance

with 51.

d4! .

A) Now 51...

xh4 leaves the black king

with too little breathing space: 52.

f2+

h1 ( 52...g3?? 53.e5+ h3

54.

h2# is the point ) 53.f1+

forces perpetual.;

B) Likewise, after 51...

h1+ 52.d2

xh4 53.xb7 White is holding.;

C) 51...

h3 maintains some pressure, but

after 52.

f2 h1+ 53.d2 f3 54.e3

xh4 55.f4 it is still not clear that
Black's extra pawn is sufficient for the
win.;

D) 51...-- ]

51...

f1 52.e3 f3 Now White is lost, as

he must surrender the b2-pawn.

53.

g3 xb2 QUESTION: So that pawn,

which he allowed to be fixed backwards on
move 26, finally comes back to haunt White!
ANSWER: Indeed, and it is a very good
lesson on the theme of accumulating small
advantages. When Black played 26...a4, it
was hard to imagine that the b2-pawn would
ever actually prove vulnerable, but, lo and
behold, some 25 moves later, it turns out to
be the crucial weakness.

54.

d6 b3+ 55.d4 f2 And now it is the

passed e-pawn which advances decisively.

56.

g7 e3 57.g3+ f1 58.f7 e2 59.e7

c6 A classic example of an endgame 'grind',
where the patient accumulation of small and
seemingly insignificant advantages eventually
enables Black to win what looked like a
hopelessly drawn position.
0-1

A28

Nimzowitsch,A
Reti,R

Berlin

1928

[Steve Giddins]

1.c4 e5 2.

c3 f6 3.f3 c6 4.e4

This move was Nimzowitsch's own invention in
this position. White gives up control of the d4-
square, but seizes d5 and prevents Black
advancing his d-pawn two squares. Later,
White has ideas of attacking the centre with
f4.

[ Although the move has never been as

popular as 4.g3 , it remains one of the main
lines here, and is still a respectable choice
for White. ]

4...

c5

[ 4...

b4 is the alternative, but the text is

also very reasonable. ]

5.

xe5 White exploits the fork trick with d4,

to eliminate Black's centre pawn.

xe5 6.d4 d6?! This move fails to
Nimzowitsch's excellent reply.

[ The main line is 6...

b4 7.dxe5 xe4

8.

d4 xc3 9.bxc3 e7 10.g4 f8

, with an unclear position; Black's
development has been disrupted, but the
white pawn structure is weaker. For more on
this position, and the entire 4 e4 system, I
naturally recommend the reader to consult
that magnificent monument of opening
literature, 'The English Opening: Move by
Move', by, er, Steve Giddins. ]

7.c5! A quaint position! Black is temporarily a

piece ahead, but has two pieces hanging.

xc5 8.dxe5 g8 9.g4 QUESTION:
Hmmm. It looks as though things have gone
badly wrong for Black!
ANSWER: Yes, his 6th move has been
refuted and he is in some trouble. His next
move is practically forced.

f8

[ Since 9...g6 , weakening the dark squares

permanently, would be even worse. ]

10.

g3 This will be forced after ...d6 anyway.

e7 11.c4 g6 12.f4!? QUESTION: It
looks strange to allow the queens to be
exchanged, when the black king is unable to
castle and apparently vulnerable to attack.
ANSWER: It does! Nimzowitsch was very
proud of his last move, and in his notes in the
Danish magazine 'Skakbladet', he describes it
by the rather condescending term "anti-
amateurish".

QUESTION: Meaning what?
ANSWER: Meaning that it is a move that no
amateur would choose. His idea is that the
attack on the black king is not really weakened
by the exchange of queens, since it depends
first and foremost on the pawn structure,
which remains unchanged.

[ Be that as it may, the "amateurish" move
12.

g5 is at least as strong, with the

continuation

e7 13.xe7+ xe7 14.f4

, and now the queen exchange after

h4

15.

xh4 xh4 16.f2 is still no great relief

to Black. ]

12...

h4

[ 12...d6 was the alternative, when

Nimzowitsch gives 13.

e3 xe3 14.xe3

dxe5 15.f5 ( White should prefer 15.

d1!

when he does indeed have a virtually
decisive advantage ) 15...

e7 ( in itself, his

assessment is correct, but he missed the
much stronger defence 15...

d4! ) 16.d1

"with a decisive attack". ]

13.

d5! c6 14.e3! QUESTION: Why is this

so strong?
ANSWER: Nimzowitsch is playing very
consistently – first, he weakens the d6-square,
by forcing 13...c6, and now he exchanges off
the key enemy bishop that defends that
square.

d6

[ Relatively the best try, since 14...

xe3

15.

xh4 xh4 16.xe3 gives White a

crushing positional advantage. ]

15.

xc5 xg3+ 16.hxg3 dxc5 17.e3

QUESTION: So what is the upshot of this
forcing sequence?
ANSWER: Basically, Black is positionally
busted. He has almost no pieces developed,
and the powerful white pawn centre e4-e5-f4
exerts a stifling effect on his position.
QUESTION: But Black does have a 4 versus
2 queenside pawn majority.
ANSWER: Yes, but just as we said in the last
game, that majority is not worth much in itself,
when (as here), it lacks any support from its
pieces.

b5 18.

e2 e7 19.0-0-0

[ 19.a4 is another very strong and tempting

option. But the text is also extremely strong.
How can Black develop his pieces? The
bishop has no squares on the c8-h3
diagonal, and developing it to b7 not only
buries it alive, but also gives the enemy
knight free access to f5. Meanwhile, his
king's rook is tied to defending the h7-

pawn. ]

19...

f8 20.f5+

[ 20.

d6 is another strong move. In fact,

White's main problem in this position is the
embarrassment of riches that he faces. ]

20...

xf5 21.exf5 d7 22.d6 ac8 23.g4

QUESTION: Why not double rooks on the d-
file?
ANSWER: That is, indeed, very strong,
possibly objectively even stronger than
Nimzowitsch's choice.

[ The computer suggests 23.

hd1 hd8

24.

h5! (threatening 25 Bxf7) f6 25.e6+

f8 26.f3 , with a totally crushing
position. ]

23...h6 White was threatening g5 and f6+.
24.

f3 c4 QUESTION: What is this about?

ANSWER: Nimzowitsch awarded this move
two exclamation marks and described it as "a
brilliant rescue attempt in an otherwise
hopeless position", but in fact, the computer
points out a simple refutation. However, I
cannot really suggest anything much better
for Black, and as it happens, Nimzowitsch's
subsequent inaccuracies lead to a fascinating
endgame.

25.

xc6?!

[ Obviously very strong, but less decisive

than the computer's 25.

hd1 hd8

26.

xc6 , with the very nasty point that c5

now allows mate in four: 27.f6+ gxf6

28.exf6+

f8 29.xd8+ , etc.

QUESTION: Very nasty indeed.
ANSWER: Yes, computers are totally
ruthless in such positions. It is amazing how
many little tactical points like this that
humans miss, even in games between the
greatest masters. ]

25...

c5 26.xb5 hd8 Even after his

missed chance at move 25, White is still
winning easily, of course.

27.

xd8 Nimzowitsch described this as a

time-trouble blunder and gave it two question
marks, but in actual fact, there is nothing
wrong with the move.

[ His recommendation 27.

hd1 , with the

idea

d3+ 28.b1 xd6 29.exd6+ xd6

30.

xc4 , also wins, but not necessarily any

more easily than the game continuation. ]

27...

xd8 28.d1

[ 28.

xc4 d4 allows Black some

counterplay, although it should still win, ]

[ but the prophylactic 28.

c2 is possibly the

most accurate: for instance,

d3 29.g3

, etc. ]

28...

d3+ 29.b1 b8 30.a4 a6

Black has now managed to drum up some
play, for the first time in the game.

31.

xa6 xb2 32.d4 xa4+ 33.c2

QUESTION: Is this becoming drawish?
ANSWER: White should still objectively be
winning, but the fact that the pawns are now
all on one side always increases the
defender's drawing chances.

b2+ 34.c1 xg2 35.xc4 b6 36.c7+

f8 37.b7 d5 38.c4 QUESTION: Oh! It
looks as though White has re-established a
decisive advantage.
ANSWER: It looks that way, but Réti now tries
a last trick.

g1+ 39.b2? And White falls for it!

[ As Nimzowitsch acknowledges, he should

bring his king in the other direction: 39.

d2!

g2+ 40.e1 g1+ 41.e2! ( 41.f2

xg4 42.xd5 xf4+ transposes to the
game ) 41...

xf4+ 42.f3 h3 , and now

the neat point is that, rather than taking on
f7, White springs a mating net with 43.

b8+!

e7 44.b5! , and Black soon runs out of
checks. ]

39...

g2+ 40.a3 g3+ 41.a4 xg4!

The point. Having driven the enemy king as
far away from the kingside as possible, Réti
sacrifices his knight, in order to eliminate all
of the white kingside pawns.

42.

xd5 xf4+ 43.b5 xf5 The e-pawn is

dropping off as well, albeit for the black f-
pawn.
QUESTION: And rook and bishop versus rook
is a theoretical draw, isn't it?
ANSWER: Yes, unless White can force one of
a small number of winning positions, the main
one being the Philidor position.
QUESTION: Well, then, if rook and bishop
versus rook is a draw, even when Black has
no pawns, then this position, where he has a
couple of pawns, must also be drawn, surely?
ANSWER: Well, you might think so, but it
turns out that Black is actually losing, as the
pawns get in his way.

44.

c6 xe5 45.xf7+ e8 46.a7!!

The sealed move, and a brilliant one.
QUESTION: Why not take the g-pawn?

[ ANSWER: After 46.

xg7 f5 47.d6 f1

the position is a draw. Nimzowitsch realizes
that he needs to keep the enemy g-pawn on
the board. ]

46...

e1?! This loses relatively easily.

[ The most tenacious defence is 46...

f8

, after which Nimzowitsch demonstrates a

beautiful win, relying on zugzwang: 47.

d6

e8 48.d7!! , and Black is in a remarkable
zugzwang position. Black now has three
choices:

A) 48...

e1 transposes into the game.;

B) 48...g5 loses as follows: 49.

h7

(threatening 50 Rh8+, winning a rook)

e1

( 49...

d8+? 50.e5 and 51 Kf6 wins )

50.

f7+ e8 51.f6 (threat: 52 Bc6+ and

53 Rf8+, mating)

d1 52.e6+ d8

53.

xh6 and we have the Philidor position,

but with the insignificant g5-pawn also on
the board. We will see the Philidor proper
in the game, after White's 55th move.;

C) The toughest defence is 48...h5
49.

f7+ g8 50.e7+ f8 51.d7

(this simple triangulation manoeuvre
simply brings about the same position, but
with Black to move) h4 ( advancing the g-
pawn, either here or move 54, loses in the
same way as we saw at move 48: 51...g5

52.

h7 e1 53.f7+ e8 54.f5 d8

55.

xg5 , etc; once again, it is the Philidor

position, but with an extra black h-pawn,
which makes no difference ) 52.

f7+ g8

53.

e7+ ( after 53.f4+ h7 54.f7

, Black has g5! and he escapes ) 53...

f8

54.

d7 (again repeating) h3 . Now that

the pawn has reached h3, White springs
the trap: 55.

f7+ g8 56.f3+ h7

57.

f7! and White mates with 58 Rxh3

next move.
QUESTION: If Black did not have the g7-
pawn, there would be no mating net?
ANSWER: Precisely. Now we see the full
subtlety of Nimzowitsch's 46th move,
leaving the g-pawn alive.
QUESTION: Very nice! But just one
question – why does White have to wait
for the pawn to reach h3, before springing
the mating net? Why, for example, can't
he do it with the pawn on h4?
ANSWER: Because Black can escape
after ...g5 (see the note to White's 53rd,
above).
QUESTION: So many subtleties in such a
simplified position!
ANSWER: Indeed, which is what makes
endgames so wonderful. It also shows
once again how wrong those players are
who think endgames either boring, or a
part of the game where tactics and exact
calculation are not important – nothing
could be further from the truth.;

D) 48...-- ]

After that lengthy, but (I hope) thoroughly
worthwhile diversion, let us return to the game.

47.

d6 f8 48.f7+ e8 49.f2!

The threat is 50 Bc6+ and 51 Rf8+, mating.

d8

[ Or 49...

d1 50.a2! and there is no

defence to 51 Ra8 mate. ]

50.

f8+ e8 51.f7 As Nimzowitsch puts it,

"The spirit of Philidor rises gradually from the
grave". Basically, this is just the Philidor
position, with the two extra black pawns not
affecting anything.

e1 52.xg7 e2 53.h7 e1 54.xh6
Now it is a 'real' Philidor position.

e8 55.h7 c8

[ The main line of the Philidor analysis

proceeds as follows: 55...

e2 56.b7 c2

57.

a7! (this waiting move is designed to

force the black rook off the seventh rank,
where it is ideally placed; both the sixth and
eighth ranks have drawbacks, as we will
see)

A) relatively best; after 57...

c3

White wins by 58.

d7+ c8 ( or

58...

e8 59.g7 and mates at once,

since the black rook cannot go to f3 - see
the previous note ) 59.

e7 b8 60.b7+

c8 61.b4 d8 62.c4!;

B) 57...

c1 58.b3! (now it is the check

on the d-file that the rook lacks, thanks to
its position on the first rank)

c3 59.e6!

d3+ 60.d5 c3 , and now we have the
rook where we want it, on the sixth rank,
the worst rank of all. We have already
seen the win from here: 61.

d7+! c8

( or 61...

e8 62.g7 ) 62.h7 b8

63.

b7+ c8 64.b4 d8 65.c4! c8

66.

e6+ , and wins. ]

56.

c7+

[ 56.

a7! was quicker, although the text

does not spoil anything. ]

56...

b8 57.b7+ c8 58.b4 d8+

59.

c6 e8 60.a4 b8 61.b6 c8

62.

c6 An excellent example of how a

fascinating endgame can result from inferior
middlegame play. After making a complete
pig's breakfast of an overwhelming position,
Nimzowitsch redeems himself with a virtuoso
display over the final 20-odd moves.
1-0

A06

Nimzowitsch,A
Voellmy,E

Bern

1931

[Steve Giddins]

1.

f3 d5 2.b3 f6 3.b2 f5 4.e3

Against Rubinstein in the first chapter, we
saw Nimzowitsch adopt a double fianchetto
against this black London System-type set-up,
but here he chooses a different development
scheme.

bd7 5.c4 e6 6.c3 c6 7.e2
As in Nimzowitsch-Ahues, Nimzowitsch holds
back his d-pawn, and thus avoids a
transposition into any kind of standard Slav/
Queen's Gambit-type position. He was a
natural experimenter anyway, of course, but
given that this was an exhibition game, against
a relatively weak opponent, he was probably
especially keen to reach a more original
position, where his opponent would have to
think for himself.

d6 8.0-0 e7 9.cxd5 exd5 10.c1
QUESTION: It looks like a fairly modest set-up
by White?
ANSWER: Yes, he certainly cannot claim any
advantage at this point. But, as we will later
see, his main middlegame plan is already
taking shape. The exchange on d5, and
Black's recapture with the e-pawn, has
created an asymmetrical pawn formation.
White has the half-open c-file, and if his pawn
were on d4, we would have the so-called
'Carlsbad structure', where the standard white
plan is a minority attack on the queenside,
with b4-b5. Here, the fact that the d-pawn is
on d2 changes things a little, but it soon turns
out that Nimzowitsch has in mind much the
same plan.

e5 11.d4 d7!?

[ QUESTION: Isn't 11...

g6 more natural?

ANSWER: It is, indeed. Black was probably
worried about a possible f4-f5 advance,
although this is not a threat at the moment,
since if 12.f4

d3 . ]

12.

c2 g6 13.f5 xf5 14.xf5 a3

15.

c2 xb2 16.xb2 0-0 17.a4

QUESTION: Black has simplified the game
with the exchanges. Is White better here?
ANSWER: No, I don't think so. It is just equal,
but Nimzowitsch steadily outplays his
opponent from now on. The text is the first
link in the minority attack chain. The knight
takes aim at c5, its typical outpost in such

structures.
QUESTION: But what difference does it make
that the white pawn is on d2, rather than d4?
ANSWER: This is an interesting point, and
worth spending a little time on. The minority
attack basically consists in pushing the white
pawn to b5, and then exchanging on c6,
leaving Black with a pawn weakness on that
square. A crucial precondition for a
successful minority attack is that Black not be
able to answer b5 with the move ...c6-c5,
which would leave White with nothing.
QUESTION: Well, in that case, surely it is not
going to work with the white pawn on d2?
ANSWER: Well, one should not be too
categorical about such things. Having the
pawn on d2 means that White has less control
of c5, but it doesn't mean he cannot control
the square - if he establishes piece control
over c5, then his minority attack with b4-b5
could still work.
QUESTION: OK, but why doesn't he just play
d4 and be done with it?
ANSWER: Because there are other
advantages in keeping the white pawn on d2.
Firstly, he can expel an enemy knight from e4,
by playing d3, if necessary. Secondly, the e3-
square is securely defended, so the counter-
minority attack by Black, with the advance ...
f5-f4 (one of his possible counter-attacking
options against the minority attack) is less
likely to be effective. Finally, the white queen
has a nice open diagonal and control of the
central dark squares from b2.

fe8 18.d4 e5?! A small inaccuracy, I
think.

[ If Black wanted to exchange queens, then

he would have done better to do it with

18...

e4 , with the point that after 19.xe4

dxe4 , he has a target on d2. The text is a

little cooperative.
QUESTION: But should Black even be
exchanging queens at all? I thought in
these minority attack structures, Black
usually wanted to attack on the kingside?
ANSWER: That is correct, and in general, I
am suspicious of queen exchanges from
Black's point of view in such structures.
However, here, the various exchanges
which have already occurred mean that
Black has fewer chances than usual of
drumming up effective kingside counterplay,
so the queen exchange probably makes
more sense. Of course, there is also the
additional point that Voellmy was probably

happy to make a draw against his mighty
opponent, which usually leads players to
hoover pieces at every opportunity. ]

19.

xe5 xe5 20.c5 QUESTION: So is

White getting on top now?
ANSWER: Slightly, but it is still not a huge
amount. The next key point is to prevent
Black achieving the advance ...c6-c5.

e7

[ The immediate 20...b6?! fails to 21.

d3

, with tempo on the rook, and after

e6

there follows 22.b4 , keeping the c-pawn
backwards. Admittedly, though, even here,
White has little after a5 23.a3 axb4

24.axb4

a2 , with counterplay. ]

21.b4 Now White secures his control of c5,

and intends to prosecute his minority attack,
by preparing b4-b5.

a6 22.a4

e4 23.b3 d6 24.a1

QUESTION: What is the idea of this?
ANSWER: Nimzowitsch is preparing to bring
his other rook over to the queenside, so he
frees the c1-square for it. Meanwhile, he also
anticipates the opening of the a-file, after a
later b5 axb5; axb5.

ae8?! It does not look terribly logical to
remove the rook from the a-file, with the
advance b5 on the way.

[ 24...

f4 was probably better, ]

[ or a move such as 24...f6 , preparing to

bring the king towards the centre. The
position would still be very balanced after
either of these moves. ]

25.

fc1 e5?! Not really solving the problem

on the queenside. Even returning the rook to
a8 was slightly better.

26.b5! White seizes the chance to make his

desired breakthrough, and can now at last
claim a real advantage, for the first time in the
game.

axb5 27.axb5

xb5 Black now faces the

familiar dilemma that always confronts the
defender in these minority attack positions, if
he allows White to get in b5. The text weakens
the d5-pawn seriously, but the alternative of
allowing the exchange bxc6 also saddles
Black with a pawn weakness on the
queenside.

28.

xb5 cxb5 29.c5 Thus, White regains

his pawn, with an obvious positional
advantage, thanks to the weakness of the d7-
and b7-pawns.

d7 30.xb5 QUESTION: How bad is
Black's position?
ANSWER: It is actually very bad. Such

positions are totally joyless for the defender,
who has two weaknesses. At best, he can
expect a very long session of torture, and in
practice, few black players manage to hold
such positions. For White, on the other hand,
there is almost no risk of losing, so it is a free
squeeze.
QUESTION: The way you describe it, I am
surprised anyone ever allows the minority
attack!
ANSWER: Well, objectively, Black is usually
OK in such positions, if he defends very
accurately, but it is true that I think such
structures are not a great practical choice for
him. As this game shows, it only takes one or
two small inaccuracies for Black to fall into a
desperately depressing position. It is
especially not a structure to play as Black
against a stronger player, who is likely to be
able to control the black player's tactical
chances. A player like the English GM Keith
Arkell, who plays most of his chess in
weekend swisses, where most of his
opponents are lower-rated than himself, must
have furnished his house many times over on
the proceeds of the white side of such
minority attack positions.

c6 This loses a pawn at once.

[ Black could have avoided that by, for

example, 30...

d3 , but then a simple move

such as 31.

a4 , threatening Rd4, will soon

pick up a pawn anyway. ]

31.

c5 c7 32.xb7 b8 33.d6 d8

34.

f5 g6 35.g3 e7 QUESTION:

Perhaps Black should have tried to exchange
off his weak d-pawn?

[ ANSWER: The move 35...d4 is met simply

by 36.e4 without any real easing of Black's
problems. ]

36.d4 QUESTION: So the d-pawn finally

moves!
ANSWER: Yes, now we at last have the
classical Carlsbad structure.
QUESTION: Does Black have any hope of
holding?
ANSWER: Well, there is always hope, I
suppose, but objectively, he is just lost - he is
a pawn down, and with a notable weakness on
d5. The presence of several pairs of pieces on
the board slows up the winning process, but
White will gradually organize himself to attack
the d5-pawn. Even if he cannot win it, he can
prepare an advance with f3 and e4, obtaining
a passed pawn in the centre.

dc8 37.f1 c2 38.g4 Another typical

endgame idea.

[ QUESTION: What is the point? I thought he

would just make luft with 38.g3 .
ANSWER: He is doing more than just
making luft. He is seeking to create a
second weakness in the black position. A
typical idea would be to push the pawn to g5,
thus fixing the h7-pawn as a potential
weakness, should the white rooks ever get
round the back.
QUESTION: It looks pretty hard to attack the
h-pawn!
ANSWER: True, but as we saw in Von
Gottschall-Nimzowitsch, weaknesses often
look as though they will not be significant,
only to prove vulnerable, sometimes many
moves later. ]

38...

8c7 39.b8+ g7 40.aa8

Now 41 g5 suddenly becomes a serious threat,
when the black king would also be vulnerable
to possible mating threats. He therefore feels
obliged to stop the g-pawn in its tracks, but
this creates other weaknesses, notably on f5
and h5.

g5?!

[ QUESTION: So why didn't he stop the g-

pawn with 40...h6 instead?
ANSWER: He probably should have done,
although it doesn't fully solve the problem
after 41.h4 . ]

41.

g3 f6 42.h5+ e6 43.g7+ d7

44.

d8+ c6 45.e8 The white knight gets

through an awful lot of work in this game. In
general, knights are quite effective pieces in
these Carlsbad structures, as the fixed pawn
formation tends to suit them rather more than
it does bishops.

b7 46.d6+ b5 47.f6 Now the threat of
48 Nd6+ wins another pawn. It is all over.

b6 48.xf7 g6 49.a1 bc6 50.b1+

c4 51.xh7 And another one goes. Black
could resign, but in exhibition and
simultaneous games, the amateurs have
usually paid for the privilege of playing
against the master, and they tend to want to
obtain their money's worth.

h4 52.c7 Exchanging pieces is the
standard way to exploit a material advantage,
and reduce counterplay.

xc7 53.xc7 f3+ 54.g2 d2 55.d1

e4 56.f1 This minor inconvenience is only
temporary.

d2 EXERCISE: Black hopes to alternate his
knight between attacking the white rook and
the f2-pawn. How can White win?

ANSWER: 57.

e6! QUESTION: What is

this?
ANSWER: Black was trying to force a
repetition. Instead of giving up the f-pawn,
which would leave Black with his active rook
and knight pair, Nimzowitsch realizes that he
can surrender the exchange, and thereby
eliminate the counterplay. With three extra
pawns for the exchange, he is winning easily.
Again, being prepared to give back some
extra material, to eliminate counterplay, is a
typical technical device for converting an
advantage.

xf1 58.xf1 Black is helpless, as the g-
pawn is dropping, after which the white
kingside pawns will roll down the board like
Space Invaders.

d3 59.xg5 c1+ 60.g2 e2 61.h3
Killing any hope of Black winning the f-pawn
by ...Rf1.

d3 62.f3
1-0