Contents

Title page 3
Explanation of symbols 6
Foreword 7
Introduction 9

Chapter 1 Style: from Karpov to Tal? 30

Chapter 2 The opening revolution 40
Chapter 3 Attack: inviting everyone to the party 44
Chapter 4 Defence: the preventive counter-attack 57
Chapter 5 Tactics: ‘les petites combinaisons’ 62
Chapter 6 Exchanges: Carlsen’s main positional weapon 76
Chapter 7 Calculation: keeping a clear mind 83
Chapter 8 Planning: when knowledge brings vision 90
Chapter 9 Pawns: perfect technique and new tips 94
Chapter 10 Pieces: the art of going backwards 102
Chapter 11 Endings: breaking the principles 113
Chapter 12 How to win against Magnus Carlsen: the hidden defects? 118
Chapter 13 Games and solutions 122
Games 1-31 122
Games 32-62 144
Games 63-93 173
Games 94-124 197
Games 125-155 221
Games 156-186 246
Games 187-217 267
Games 218-248 289

Index of names 317

Bibliography 320

2
 Emmanuel Neiman

The Magnus Method

The Singular Skills of the World’s Strongest Chess Player Uncovered and Explained

New In Chess 2021

To Stephanie, with love and gratitude

© 2021 New In Chess

Published by New In Chess, Alkmaar, The Netherlands

www.newinchess.com

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

Cover design: Volken Beck

Typesetting, supervision: Peter Boel
Proofreading: Joe Petrolito, Mariska de Mie
Production: Sandra Keetman, Anton Schermer

Have you found any errors in this book?

Please send your remarks to editors@newinchess.com. We will collect all relevant corrections on the
Errata page of our website www.newinchess.com and implement them in a possible next edition.

ISBN: 978-90-5691-968-9

3
 Explanation of Symbols

The chessboard with its coordinates:

K King
Q Queen
R Rook
B Bishop
N Knight

² White stands slightly better

³ Black stands slightly better
± White stands better
µ Black stands better
+– White has a decisive advantage
–+ Black has a decisive advantage
= balanced position
! good move
!! excellent move

4
? bad move
?? blunder
!? interesting move
?! dubious move
™ only move
# mate
ch championship
zt zonal tournament
izt interzonal tournament
ct candidates tournament
tt team tournament
ol olympiad
m match
cr correspondence

5
 Foreword

Why another book on Magnus Carlsen?

The Norwegian has been the World Champion since 2013, and the undisputed number one in the
rating list since 2010. He is probably the most dominant chess champion ever, together with
Capablanca, Alekhine and Karpov at their best – with the exception of Kasparov, who stayed at the
top for fifteen years (1985-2000).

My goal as a trainer is to use Carlsen’s games, especially the areas in which he is exceptional, to
enable readers to use them for themselves.
I try to answer two questions:
1. What does he bring to the game?
2. What are the specific tools he uses?
Carlsen’s technique often seems deceptively simple: when you play through his games, the moves
seem obvious – you feel that this is the logical way to go. But more often than not, his opponents,
usually the best players in the world (after him), didn’t see them coming!

I have divided this book into twelve chapters, each of them dealing with a specific ‘Magnus’
approach to one of the main phases of the game. In each chapter, we start with an explanation
of Magnus’ approach to the theme at hand. Next (with the exception of the first two chapters), a
series of diagrams is given. These diagrams show positions in which I recommend that you first
try to find the required move yourself – what would you play here if you were Carlsen? The
solutions are given in Chapter 13 with further explanations. The numbers of the diagrams refer
to the game numbers in Chapter 13 and the order of the diagrams within those games
respectively, e.g. the number ‘15.2’ means the second diagram in Game 15.

After you have decided on your move, look at the solution. The level of difficulty varies, some topics
(e.g. calculation, planning and endings) are more demanding than others.
Some of the topics are obvious, like tactics, attack, and endings. Others are less common in exercise
books, such as pawn play, piece play and exchanges, because Carlsen has introduced new trends,
sometimes even new understanding, in these areas. A theme like ‘exchanges’ is clearly important to
all players, but especially in Carlsen’s case we are dealing with one of the basics of his game.
I have used computer engines to assist in analysing the games, mainly Stockfish 10, 11 and 12.
When a move is considered the best, or is given an exclamation mark, it is usually Stockfish’s first
move. This does not mean that it is the absolute best move! The evaluation is an indication, nothing
more. You’d be surprised how many changes I noted between the Stockfish 10 and Stockfish 12
evaluations, for example.
In any case, Carlsen’s conception of what a good move is depends on several factors. In most cases,

6
it’s the absolute best move. In other cases, he will choose another move because it gives him a better
‘chance’ of getting the result he is playing for. Recent advances in artificial intelligence demonstrate
the same approach – the right move is now considered to be the one that gives you the best result
according to the statistics.
In the age of computers, Carlsen plays in a very human and logical style, and according to the
‘sporting’ demands of the position. I believe that examining his specific abilities is the best way for
you to improve, and to understand chess better.

I would like to thank some friends, who helped me by giving me precious advices, tips and help:
Charles Bonati, Pascal Chomet, Damian Justo, Olivier Renet, Jérôme Solakian-Vaneyll, Aldo Haïk,
Stephane Schabanel and Samy Sahraoui – merci les amis!

I wish you lots of pleasure and success with this book!

Emmanuel Neiman
Paris, June 2021

7
 Introduction

In this introduction, we will briefly discuss two aspects of Magnus Carlsen’s play: his technique and
various strong points that are not directly related to technique.
Carlsen’s special technique for winning ‘even’ positions
The main point with Carlsen is his ability to win equal or very slightly favourable positions. He
became World Champion in 2013 thanks to his victory in two totally equal positions in rook
endgames (the World Chess Championship Match vs Anand in Chennai 2013, Games 5 and 6). He
retained his title in 2016 against Karjakin thanks to his win in Game 10 after a long, technical game.
In his win over Caruana, in the rapids of the 2018 World Chess Championship, the first decisive win
was in another drawish endgame.
How does he do it so consistently, against the best-educated players of all time? In addition to his
main qualities, Carlsen has developed a unique technique that has two aspects, one negative and one
positive.

The negative side:

Carlsen, commenting on an online rapid game against Dubov (27.12.2020): ‘I was trying to keep
the position alive, obviously, and it felt like there was nothing, nothing, and just suddenly I managed
to trick him and get a win. It felt a bit accidental, to be fair, but it’s okay, as long as you can keep it
from being a forced draw you can often get something in those games.’ This means that you don’t let
your opponent get a game where he knows what to do. An example of this is his game against
Caruana, where he doesn’t opt for a 4/3 rook endgame that was famously lost by Anand. His
opponent would feel comfortable with the position, even though it is clearly worse, because such
endings have now been studied, and a path to the draw could be known.
This is part of ‘prophylactic thinking’1, when you should not let the opponent get what he hopes
for.
1 Dvoretsky explained in his books that you should always consider what the opponent has in mind,
even when you develop your own plan. Prophylaxis was originally a purely technical concept from
Nimzowitsch, but it is now the way modern players think. As Carlsen has also mentioned another
of Dvoretsky’s concept – the superfluous knight – we can safely assume that he appreciates the
Russian’s legacy.
In a simplified position, you look for a simple way to get the draw, so Carlsen chooses moves that
don’t allow the opponent to get that kind of position, anticipating his wishes and trying to prevent
them. He pushes for victory by allowing the game to continue, and as long as there is a game, there
are possible mistakes to be made.

Magnus Carlsen 2881

Shakhriyar Mamedyarov 2760
Shamkir 2014 (1)

8
 White to move

Here Carlsen played 24.Rd6,

refusing to force the play with 24.Rd8+, which makes it easier for Black, because otherwise the
following moves would be easy.

The positive side:

‘He’s a very special player and he manages to create ten times more problems for you than anybody
else. I felt like I was holding decently and then it was one of his last tricks that finally worked for me’
– Daniil Dubov, 30.12.2020.
Carlsen constantly changes the position. He will go for an attack on the queen’s wing. If the opponent
is capable of (or if he feels that the opponent is capable of) countering it, he will go for an assault on
the king, and then he will change everything to get a favourable ending. Even in an endgame, he will
go through all possible endings, using exchanges, changes in pawn structure, etc., so that he can give
the opponent many chances to go wrong. When the opponent is happy to have solved a problem –
answering a question, as in an exam – then another, probably very different question will be asked
next, and then another, and another... even in a simple position like a rook with two pawns against a
rook with one pawn, he will try many piece formations (with two pieces!) to try to confuse the
opponent. Some games against Karjakin are typical.
There was the important win in the 10th game of the World Championship Match in 2016, and even
more so, the thriller played in Wijk aan Zee, 2013.
In this game (see the diagram on the next page), Carlsen, having tried everything (opening
surprises, middlegame strategy changes, structural modifications, queen exchanges, material
imbalances...), reached one of his favourite endgames (a rook and an opposite-coloured bishop each),

9
but was not able to beat his opponent. As a last chance he tried a tactic, a sacrifice of two pawns, that
finally worked.

Magnus Carlsen 2861

Sergey Karjakin 2780
Wijk aan Zee 2013 (8)

White to move

Here, Carlsen played 67.g4 hxg4 68.h5!?.

This way of constantly changing the course of the game is a known ploy in bad positions; Carlsen
uses it in good positions, a new way of playing to gain an advantage – unlike players like Kramnik or
Fischer, who simply tried to outplay their opponents by playing the ‘best’ moves.
Playing for victory in such thrillers is completely exhausting, so it is not surprising that Carlsen,
even while outplaying his opponents, loses to some of them. The losses against Vachier-Lagrave
(2017 Sinquefield Cup), or Caruana (2012 Grand Slam final) were partly due to fatigue, especially
against strong and resilient defenders.
This way of playing also has consequences for his style. Thanks to this evolution, the maturing
Magnus is able to employ all kinds of openings, pawn structures and playing styles, and now often
uses his encyclopedic knowledge to change the course of a game, because he wants his opponents to
be faced with many different problems to solve.
If we take just one game, the Carlsen-Vachier-Lagrave game from the 2015 Gashimov Memorial is
the most typical of this singular technique – and also one of his best achievements. The game begins
with an opening surprise (4.Na3), which leads to a virtually unknown position that turns out to be a
type of Queen’s Indian Defence, a closed game with some space advantage for White. Here, Carlsen

10
surprisingly decides to ‘lose’ – sacrifice – a pawn in the form of an apparent blunder (14.Bg5!?), a
very typical intuitive sacrifice by the Norwegian (as in the 2019 victory over Giri), and forcing the
opponent to make a choice: ‘Should I take, or should I refuse the pawn?’. Then, after Vachier-
Lagrave’s acceptance, a brief tactical mêlée ensues, followed by a phase of positional play, with
White able to find excellent squares for his pieces (instead of regaining the pawn) and gaining a clear
advantage. However, surprisingly, while his pieces seem to be in perfect positions to attack Black’s
king, Carlsen decides to exchange queens, to play an endgame with an advantage on the queenside.
Again, there is a positional phase, when White’s advantage leads him to win a pawn. When Black
decides to sacrifice a piece to counterattack with an advanced protected passed pawn, Carlsen allows
his opponent to promote, to force a checkmate with the rook and bishop.
So, in one single game, there are lots of surprises, changes of pace, shifts from one side to the other
(from the queenside to the centre, then to the kingside, then a few exchanges, back to the queenside,
and dropping the queenside pawns to checkmate on the kingside). One can imagine how many
different problems the opponent has to solve just to stay in the game.
Apart from the necessary calculations, the strategic changes are constant, even if the pawn structure
seems fixed in the opening. In this game, we can witness one of Carlsen’s strengths, namely the
transition from one phase to another.
By transition I mean not only the classical ‘jumps’ between the opening and the middlegame, and
between the middlegame and the endgame, but also the change between a tactical and a positional
phase and also the problems of exchanging. If we compare this way of playing with Fischer’s, here is
an important point: for Fischer, you have to find the best move all the time – there is no real transition
and you always have to be perfect! According to Spassky, in an interview with me in 1992, this was
Bobby’s only flaw: ‘He didn’t feel the important moments of the game’. On the contrary, this is one
of Carlsen’s strengths – as you can see from his thinking time. In the first rapid game of the World
Championship Match against Caruana (Game 247), White had a clear advantage both on the board
and on the clock before his 24th move. Nonetheless, he used nine of his remaining thirteen minutes to
try to find the forced win he thought existed – and indeed it did, but he was unable to find it. Among
the best players, one can imagine many of them making a random move, to use the time advantage
and hope for a mistake by the opponent.
Carlsen’s strong points
Here we discuss some of Carlsen’s strengths that are not directly related to technique:
1. Evaluation
In any position and at any time, Carlsen is able to assess accurately, and this gives him a huge
advantage over his competitors. This is where Carlsen differs from many of his predecessors. Not so
long ago, a touch of optimism was considered to be the mark of great players. Apart from players like
Alekhine or Botvinnik, who often bluffed in their written analyses, claiming to have devised long-
term plans (both would pretend making long-term plans covering nearly the whole game...), most
modern players are of the optimistic variety, with the notable exception of Petrosian (who was often
too pessimistic!). Tal of course was, and Karpov is the most remarkable. He always feels like he is
winning, not for positional reasons, but because he has great confidence in his own talent. Larsen was

11
most optimistic, and it cost him dearly against Fischer. Fischer and Kasparov were more balanced,
but still had a soft spot for ‘their’ position. Topalov is a great optimist – he wants blood and always
plays as if he is winning. Kramnik’s subjectivity is phenomenal, and in press conferences you can still
hear him explain his games with, ‘I was completely winning, crushing my opponent, when something
extraordinary happened...’. Then you look at the game, and the cold assessment of the computer is a
constant 0.00. Anand, Magnus’ predecessor, is the first modern World Champion who constantly tries
to get an objective assessment of the position before making any decision or plan. But Carlsen has
taken this ability to a new level.
This way of constantly evaluating and re-evaluating positions is openly shown to everyone in the
famous ‘Banter Blitz’ sessions against other strong players or aficionados, when he plays and
comments at the same time. Most often, these sessions are played on his co-owned site Chess 24, and
you can see the replays on YouTube. It’s fascinating, especially when you compare what he sees and
looks for with the thinking of another strong player, as in the six-game match with Svidler (both
players comment live; the comparison is impressive). Apart from the humorous commentary, there’s
a lot to learn from these blitz sessions. Magnus likes to state old chess truths – he is well-read! – and
he doesn’t hesitate to admit his momentary weaknesses, the moves he didn’t see, the calculations he
made before taking this or that decision. What is clear is that he is constantly looking for the truth, the
best moves, the best possible decision according to the situation, taking into account all factors, the
clock times, the score and the opponent. Overall, this ability to objectively evaluate almost every
position and situation is probably his greatest talent. I remember a game in London, one of his
favourite tournaments that he used to win almost every time. He was playing against an outsider, the
Englishman David Howell, and had taken a lot of risks to unbalance the position, boldly sacrificing
two pawns for dynamic compensation, which was quite difficult to evaluate.

Magnus Carlsen 2826

David Howell 2633
London 2011 (1)

12
 Black to move

I remember that immediately after finishing this game, Carlsen went straight from the theatre to the
public conference room and was asked about the evaluation. After all, here he was two pawns down
against an initiative. He said it was all the time very even (before Black’s mistake), which is quite
surprising, and is exactly the evaluation offered by today’s computers! About this ability, Kramnik
commented in 2020: ‘[Magnus] is by far the best in this. His assessments are so precise and so “cool-
blooded”. He never gets emotional... I was always amazed how often he was right, to a millimetre... I
haven’t seen a single player who was so precise and so stable.’
This extraordinary ability to assess any type of position is probably the key to Carlsen’s success. He
is able to anticipate most situations better than his opponents. Good positional assessment is both
proof that his understanding of the game is superior, and the guide that allows him to navigate the
complications more effectively, making his calculations easier. Anand once commented on the
possibility of cheating with computers during games. You don’t even need the right move, just the
evaluation, explained the Indian legend. Then you can understand the importance of Carlsen’s talent
in this aspect.

26...R6d7

26...Qa4 is equal.

26...R6d7 27.Qf3 Qa4 28.Qe2 Re7

28...Qa3 was better.

13
29.Nxf6! gxf6 30.Qe3! Red7

30...Qxc4 31.Qxh6+ Kg8 32.Rb4; 30...Kg7 31.Rb4 Qa5 32.Rf3 Kh7 33.Be2 e4 34.Rf4+–.

31.Qc5+! Rd6

31...Kg7 32.Rf3 (32.Rb4) 32...Rd6 33.Rg3+ Kf8 34.Rxc7+–.

32.Rxc7 Qc2 33.Rc8 R8d7 34.Be6 Ke7 35.Bxd7 Bxd7 36.Rh8 Qd3 37.Rf3 Qd5 38.Qxa7 e4 39.Qb8
Qe5+ 40.Rg3

14
Black resigned.
Being able to know the evaluation of the position is crucial in many positions. The most crucial
situations occur in defence. In these cases, it makes a huge difference whether you are slightly worse,
clearly worse, or lost – to mention just three cases – and there are many more nuances. Let’s look at
two positions:

Vladimir Kramnik 2770

Garry Kasparov 2849
London 2000 (4)

Black to move

White has a spatial advantage, while Black’s position, though solid, is slightly passive. Kasparov
hated this kind of position, and on more than one occasion showed himself unable to hold them.

25...Rc2

This is a very bold counterattack. In itself the move is not a blunder, but it is a rather strange move to
play if you know the evaluation (+0.20 approximately). With all his pieces passively placed,
Kasparov intends to counterattack with his only active piece – not the move a ‘positional’ player like
Karpov or... Carlsen would seriously consider. Yet it is perfectly viable according to the computer.
After a passive move like 25...Rc7, the position is equal. Then, Black will replace the bishop on b7
and wait.
The careful 25...Ke8, to defend the d7-knight, is also very good.

25...Rc2 26.Bxa6! Bxa6 27.Rxd7

15
Now Black is in danger.

Apparently, White is in the lead after this tactical operation. Yet a stricter evaluation of the position
would allow Black to find resources.

27...Rxb2

was played in the game, leading to a losing position. Except for the rook, Black’s pieces are passive
and unable to join the attack. Kramnik played accurately at first, only to spoil the victory later in the
game. After 27...Ke8 28.Ra7 Re2!, Black sacrifices a second pawn but manages to activate all his
pieces. It is interesting to note that Kramnik, in his recent and great videos ‘Vladimir Kramnik
rasskazhivaet o legendarnom matche c Garry Kasparovim! Interview tretye’ on YouTube, does not
mention this possibility and regards 27...Rxb2 as practically forced. 27...Re2 first works too.
What is your assessment of the second position, with Black to play?

Samuel Sevian 2642

Timur Gareyev 2557
St Louis 2019 (4)

16
 Black to move

Black has one more pawn, but the accurate assessment is that his position is hopeless. Black was able
to evaluate correctly, and thus played for an exceptional trick, of which Carlsen would certainly have
been proud:

30...Ra4

This is a fantastic attempt to swindle the opponent. ‘Ra4? No way! The rook is trapped forever!’
exclaimed Yasser Seirawan, the live commentator. But since the game is already lost, what difference
does it make? The mistake becomes more plausible when we realize that Black had only two minutes
left to play and was the underdog of the tournament on the Elo scale.
After the ‘normal’ 30...a4 31.Bf2 Kf8 32.Bh4, then f5-f6, and Black is crushed.

30...Ra4 31.a3

17
31...Rb3

This is the real trick. I checked the video on YouTube, and when 31...Rb3 was played, Sevian
carefully took his pen, wrote the move in a very composed way, looked at the board and, after a
second, played:

32.Rd3

Of course, this is a very natural and technical move. Since the rook on a4 is dead, it makes sense to
exchange the last active enemy piece, right? In a blitz game, most GMs would instantly play 32.Rd3.

18
Yet, with 14 minutes to play, a suspicious player would have wondered why an experienced fighter
and strong GM like Gareyev would suddenly play stupid, nonsensical moves like 30...Ra4, allowing
his rook to be trapped, or 31...Rb3, exchanging his only valuable piece.

32...Rxb2!

After

33.f6+

(the main variation goes 33.Rxb2 Rxc4+ 34.Kf3 e4+) White was hardly able to achieve a draw.
2. Chess knowledge
Carlsen’s chess culture is enormous, and he knows practically everything about chess, having learned
it mainly from books. Great games, famous players, openings, pawn structures, endings... he has
studied everything and remembers it! Like Kasparov, he is said to have a great photographic memory.
He is an exception because many players of his generation have been mainly influenced by
computers, and this general chess culture has helped him on some occasions, like when he encounters
a strange pawn structure, or a rare middlegame theme. Of course, it is of the utmost importance in
technical phases, like in openings or endgames.

Teimour Radjabov 2784

Magnus Carlsen 2835
Moscow 2012 (5)

Black to move

19
35...c5!

This is the right plan, just like the great Capablanca played in exactly the same pawn structure (see
below). One of Carlsen’s secrets is his incredible erudition, although he would have been able to find
the plan without knowing the game. His understanding of any pawn structure, especially rare ones
like in this game, is unparalleled, past or present.

36.Ra1 c4 37.Bc2 Kc5 38.Re1 c6 39.Bb1 Kb6 40.Bc2 Kc7 41.Kf2 Kd7 42.a4 bxa4 43.Ra1 Rb8
44.Ra2 d5 45.exd5 cxd5 46.Bxa4+ Kd6 47.Bc2 d4 48.Be4 Rb6 49.Ke2 g4 50.fxg4 Bxg4+ 51.Kd2
Be6 52.Kc2 Bd5 53.Bxd5 d3+ 54.Kd2 Kxd5 55.Ke3 Rg6 56.Ra5+ Ke6 57.Ke4 Rg4+ 58.Kf3 Rf4+
59.Ke3 Rf1

White resigned.

Ilia Kan
José Raúl Capablanca
Moscow 1936 (10)

Black to move

24...c5

Then follow the same pawn moves that Magnus repeated in his game against Radjabov: ...c5-c4, ...c7-
c6, ...d6-d5-d4-d3 and ...g7-g5, ...h7-h5, ...g5-g4-g3 on the kingside: 0-1 (57).
3. Versatility
This great chess culture and his general disposition – he is more a player than a theoretician – mean
that Magnus can change his openings at any moment, even just before the game, if he suddenly feels

20
that the opponent could be prepared against his original idea. His play can vary from slow chess to
purely attacking games, and he can change his plan very quickly during the game. This is an
important tool in matches, as he can choose the openings and the way of playing that will be most
unpleasant for his opponent at any time. He is able to play almost any type of position and any
opening.
Dubov: ‘You put him in any position and he will stay calm and play. There are players who
specialize in this and that. Carlsen is not like that. He is simply the greatest talent I have seen.’
In his 2018 World Chess Championship Match against Caruana, Carlsen as White played twice
1.e4, twice 1.d4 and twice 1.c4.
Carlsen is also the best player regardless of the time control chosen. He is currently a three-time
World Champion at blitz, rapid, and ‘normal’ time controls. He is also the best player in online
tournaments.
4. Fighting spirit
Gelfand wrote (in Technical Decision Making in Chess, Quality Chess 2020): ‘Karpov wanted to win
every tournament, while Carlsen’s idea is to win every game. He once said: “I want every game to be
an event”.’
This traditional quality of sports champions – and for Carlsen, chess is first and foremost a sport –
is the ability to fight to the limit of their capacity to win. But with the Norwegian, it is raised to a
degree that is unknown among chess players. He loves the fight and seeks it, always trying to win,
even when a draw would be a desirable result, to win a tournament or a title. Karpov or Kasparov
were (are) great fighters, but they knew how to secure a tournament win with a draw. Carlsen made a
huge impression when he only needed a draw against Aronian to win the Sinquefield Cup in St Louis,
2013. It was a tough game; Aronian had some advantage, and then tried to get chances to win from an
even position, and since he couldn’t do it, proposed a draw that would have been immediately
accepted by any other player. Carlsen, against all the odds, refused the proposal and went on to win
the game. It seems that the Armenian player never fully recovered from this tournament, since a few
months later he was replaced by Caruana as the second best player in the world – a place that is an
important achievement in the era of Magnus.
Carlsen also played for a win in the final game of the 2013 World Championship Match against
Anand, when a draw was enough to become World Champion, which was unheard of before him.
Carlsen always wants to play the best players, dangerous or not. He’s always willing to play in an
open tournament (the first world champion to do this in ages). He fights for the win in every game –
something very few major players in recent history have done. Another characteristic he shares with
Kasparov is that in the selection of his best games, most of the games are played against the best
players. Even more so than the great Garry, Magnus excels against the strongest, and his best games
are often played against players like Anand, Kramnik, Caruana or Aronian, that is, the players who
are (were) most likely to be his rivals for the chess throne. He possesses a crucial quality that very
few chess players possess, even at the highest level: he is never afraid of losing a game (see Game
53), and defeats do not affect his level of play as they do for most other players.

21
 We could add to this that Carlsen is an avid sports fan, who enjoys playing soccer and basketball in
particular. He is in good physical shape, which helps him to fight as long as it takes in games. All
great chess professionals make efforts to train physically, but for him it is more a way of being – he is
a real sportsman.
The only competition where he has a confidence problem is the World Chess Championship2.
2 Let’s also mention the team competitions for Norway, and playing at home, in the Norway
tournament. Is it because in these cases, he feels ‘responsible’ and does not ‘play’?
In this event, he doesn’t feel like his usual self. About the 2013 World Chess Championship Match,
he recently said, ‘At the start I was so nervous that I dropped one of my pieces on the first move,
which is sort of fascinating, because even at that point I was the highest-rated player in the world and
I was very experienced. But the task of playing a World Championship was so daunting to me that for
the first four games, or three-four games, I was still unbelievably nervous, because it was something
that was absolutely new to me.’
First, as the No. 1 ranked player in Elo, he refused to play for the title in 2011, unhappy with the
conditions offered by FIDE. Second, he has hinted on several occasions that he might give up the title
under certain circumstances. And third, in his four World Chess Championship matches, he has never
shown his usual level and confidence. It seems that he is under pressure at these events (one can also
recall the painful end of the famous 2013 London Candidates Tournament, where he suffered two
losses in the last three rounds after a brilliant start). Nevertheless, he still fights quite well in this
competition, as the results have shown!
5. Pragmatism and perfectionism
Carlsen is a very pragmatic player. We will see that he has a singular understanding of what a ‘good’
move is, which makes him very different from a ‘purist’ like Fischer. For Carlsen, the goal of chess is
to win, not to play good moves or ‘the’ best move or the ‘God move’!
This is probably why some players and specialists have (had) doubts about Carlsen’s level.
Kortchnoi, Ivanchuk and Aronian are among the authorities who are not convinced of Carlsen’s
greatness, compared to illustrious champions of the past.
As for the practicality of his decisions, we should remember the decision not to play on in a
favourable position in the last long game of the 2018 World Chess Championship Match against
Caruana (although most of the advantage that Carlsen had had before had already disappeared) and to
take a draw, reaching the tie-break. This decision was heavily criticized by ex-World Champions
Kasparov and Kramnik and most of the chess community. This is understandable, because if you
follow a live game with a lot of intrigue, it is always a disappointment to see it end in a premature
draw when everything is yet to be decided. On the other hand, Carlsen was a big favourite in the rapid
games (although Kramnik only gave him a 60-40 advantage before the tie-break) because Caruana is
notoriously weaker in the rapid and blitz games. As a logical player3 he needs more time to calculate
and settle into a new position than an intuitive player like Carlsen.
3 See Dvoretsky’s classical distinction between the logical player, strong calculator, with deep play
based on thorough opening preparation like Kasparov, and the intuitive player who has a talent for

22
 feeling perfect piece placements and quick tactics like Karpov.
Carlsen felt that the risk of losing the 12th game, despite what Kramnik claimed at the time4, by
taking advantage of a favourable position, was not worth taking, and this decision proved to be
correct, because he played excellently in the tie-break.
4 I remember discussing this with GM Kouatly, who estimated the chances around 60% draw, 30%
win for Carlsen and 10% loss.
It was the decision of an experienced gambler, as shown by his results in the game of betting on
English soccer matches, where he has consistent success and was even momentarily No. 1 in a field
of 7 million players. Among chess players, there is a new culture of using a statistical approach for
selecting a move. This approach was first developed for some games like backgammon and poker,
and recently carried out by computers using the Monte Carlo method, when the computer plays
against itself to test a position. Hence, in a given position, instead of looking for the ‘best’ move in
the sense of Fischer, you look for the move that has the best chance of winning/not losing according
to statistics. Incidentally, it seems to me that Carlsen had been waiting and hoping for this conclusion
(reaching the tie-breaking rapid games) for several games already, as evidenced by his solid choices
in his last two white games. In retrospect, and although the ever-surprising Grischuk felt just before
the rapids that ‘Caruana has the initiative in the match’, I believe Caruana should have tried other
openings in the last four games, especially as Black, if he was trying to play for an overall win in the
match.
On the other hand, Carlsen is very critical of his own play and is constantly trying to improve.
Recently, he has changed his style and become a more offensive player and looking for dynamic
positions, as shown by his surprising choice of openings against Caruana, with the Sicilian being his
main defence.
His comments, even in the smallest chess event, show that he always tries to play as well as
possible by his own standards. Perfectionism is not so common among the best chess players these
days. Most of them are satisfied with a good result, and are not obsessed with the idea of perfection.
Giri is such a player, for example (Vachier-Lagrave could also be named here, or even
Nepomniachtchi). He is a consistent member of the top ten, and often the top five, very hard-working,
constantly studying and learning from any game played at the highest level, and a renowned openings
specialist, but he is often unable to reach the top through a purely ‘practical’ approach.

Anish Giri 2752

Anton Korobov 2667
MrDodgy Invitational chess24.com 2020 (21.1)

23
 White to move

In this game, Giri had 2:33 minutes left at this point (with some increment after each move). He
played instantly:

37.h6? ... ½-½ (44)

37.g6 is winning; the ‘rule’ here is to advance the pawn that is closest to our king, so that the king can
approach earlier to help the promotion and/or give mate. The main line goes 37...Kf6 38.Kd3 Kg7
39.Ke4 Kh6 40.b3 Kg7 41.Kf3 Kf6 42.Kf4 Kg7 43.Kg5 d3 44.h6+ Kg8 (44...Kf8 45.Kf6) 45.Kf6
d2 46.h7+ Kh8 47.Kf7 d1=Q 48.g7+ Kxh7 49.g8=Q+ Kh6 50.Qg6#. It’s an elementary piece of
calculation (or knowledge) at this level. We can’t imagine Carlsen missing it – if he did, he would be
ashamed and very unhappy about it. Nor do I think Grischuk would ever miss g5-g6. Yet Giri, in the
same situation, was happy to secure a draw that was favourable to him because of the situation in the
event. This is one of the differences between an excellent top ten player and the absolute best player
since 2010. Carlsen wants to play the best moves at all time controls and under all circumstances, as a
matter of principle. This is a great example for young players, who are often all too happy to relax
when things are going well.
6. Intelligence/psychology
Carlsen’s evolution from a gifted child to one of the best players of all time owes little to coaches or
outside help. Even his father, a good player (around 2000 Elo), did not give him much advice. The
young man always decided on his own, and trained mainly by playing (a lot) and reading books
(most!).
Henrik Carlsen always said of his son that he made his own decisions early on: ‘Here is an example
that proves what I am saying. Magnus had a short cooperation with Kasparov and some of his

24
coaches when he was 14 years old, then Kasparov or his trainer asked Magnus to do some homework.
Magnus agreed. After a few days I asked Magnus if he was going to do the homework. He replied
that he had done a bit of it. “But you had assured him that you would do the homework,” I said.
Magnus was not happy with this. He did a little bit more of his homework and then didn’t complete it.
The cooperation between Magnus and Kasparov stopped because Magnus wanted to be in control of
what he was doing. No pushing! At the point when this happened, I felt like, saying no to a
cooperation with Kasparov when you are 14 doesn’t sound like a good idea. But it’s your hobby, so I
have no comments or complaints. It was all his decision. Gradually, I began to understand that
Magnus seems to understand what is best for himself. So, hands off with that idea [smiles]’ (from an
interview with ChessBase in 2019).
Carlsen had another disagreement with Kasparov, his part-time coach in 2009, after a game against
Kramnik. Kasparov called Carlsen minutes before the game, asking him to change the opening
because he had discovered a hole in the preparation they had made in the morning. Carlsen hated
being disturbed before the game, and he agreed with Botvinnik’s advice about preparation: an hour
before the game, rest, as if you were lying on a bed, and a walk just before the game will put you in
the best condition for optimal concentration on the board. Carlsen was unsettled by this call, lost the
game and immediately asked his father to get rid of Kasparov, although he always kept a good
relationship with him. Both publicly share a mutual admiration, and Carlsen has repeatedly stated that
he considers Garry the GOAT [Greatest Of All Time]. However, their collaboration did not last long.
Benjamin wrote in World Champion Chess for Juniors (New in Chess 2020): ‘They were like
ketchup and ice cream; great by themselves, but not great together.’ Whatever greatness you may find
in ketchup or ice cream, it is an apt description of the difficult collaboration between two strong
minds.
Carlsen has too much personality to be ‘trained’, and from a young age he has had only helpers,
sparring partners at times, but no real teachers – except for the books he has read himself (a lot!). It is
well-known that he was helped by Simen Agdestein, and I believe that he was a very important
influence for him in various areas. Agdestein is known to be an astute and gifted GM, who is
particularly attentive to defence, and also an international soccer player, who played eight times for
his country. Nevertheless, Agdestein did not give private lessons to Carlsen, and the young Magnus
always worked mainly alone, preparing openings with his computer. Today, a rich world champion,
Carlsen has built a strong team of assistants, the most important being Peter Heine Nielsen, his main
second who usually accompanies him to tournaments, and Laurent Fressinet, a long-time assistant
since his first World Championship Match in 2013. But Carlsen is clearly the boss and the decision-
maker on everything chess-related – as he was as a child, according to his oldest second, his own
father.
From an interview with Carlsen (with Sal Khan, 12.03.2021) we can learn about his way of making
choices: ‘[...] Very often people look to see if I have a method or a deeper meaning to what I do, but
what I explained at the beginning was that my method was mainly trial and error, and it still is to
some extent, that I find what works for me at a given moment just by experimenting, not necessarily
by good planning.’ That’s why Carlsen played almost every opening and tried every playing style in
his youth, so that he appears today as the most complete player ever.

25
 Magnus can also use psychology as a chess weapon. For example, he sometimes wrote on his blog
(a few years ago) how good his score against the opponent was just before the game. Before the 2013
World Championship Match, he posted pictures of himself doing intensive sports, and described in
advance his strategy against Anand: playing long games to tire out his aging opponent – more or less
what happened in the match.
Occasionally, he also has the ability to make things happen simply by playing a role during the
game. Some people consider this unfair, but it is clearly not the case with Carlsen – nor was it with
Tal, Kasparov or Najdorf5. Here is a well-known example:
5 Najdorf was also capable of doing some things that Magnus would never do! Playing in a rapid
event in the centre of Buenos Aires (imagine a large room with hundreds of kibitzers), he was being
crushed by the young Claudia Amura, a strong IM. The loss was imminent, and so Don Miguel got
up, shook Claudia’s hand ceremoniously under the applause and said loudly: ‘Well played... I
accept the draw!’ Claudia, who was my teammate around 2000, told me half sorry, half smiling,
‘What to do? He was an idol then, and with so many people around... I had to agree.’

Miguel Najdorf
Svetozar Gligoric 2575
Helsinki ol 1952 (5)

White to move

39.Nxd3

After taking the pawn, Najdorf slapped his forehead in despair. Now White apparently loses a pawn,
and his unfortunate opponent instantly replied:

26
39...Nxe4

39...Nc4=.

40.Qe3

Now the knight is lost (40.fxe4 Qxe4+ 41.Kf2 Qxd3 is what Black hoped for).

40...Qe6 41.Qxe4 1-0

In a recent Internet game against Nakamura, when Carlsen had lost his concentration due to the
resigned look on his opponent’s face, he said, clearly amused: ‘Here I fell for the oldest trick in the
book’. I think I saw some acting from Carlsen at the 2018 World Championship Match in London
(Game 3). After a difficult opening, Carlsen had a very slight advantage at the end of the game and
was pressing, while Caruana was in zeitnot just before reaching the time control at the 40th move.
Then, Carlsen forced his opponent to make a crucial decision about a minor piece exchange. Caruana
made the decision that seemed natural. Carlsen immediately took the piece back, got up from his
chair and almost ran off the playing stage, apparently very happy. Caruana had about a minute left to
reach the time control when he was confronted with the situation. First, he turned very red – ‘Did I
just goof up?’ he seemed to ask himself. After twenty or thirty seconds, he pulled himself together
and regained his composure and confidence. Seeing that his opponent had faced difficulties, Carlsen
had tried to induce him into a certain panic, hoping to cause a mistake just before the time control.
Caruana had probably experienced such attempts before and was ready, but he seemed to be in real
trouble for a few seconds. Another example was in the second game against Anand in the 2014 World
Championship Match, when in a slightly better position, Carlsen got up from his chair and stood
behind it but in front of the board, as if giving a simultaneous display, for about thirty seconds, an
unusual posture in such an event. This time, Anand did not react well, and lost the game.
Raised in what looks like the ideal Northern European educated family, with three sisters, Carlsen
can play tough6.
6 ‘There are tough players and nice guys, and I’m a tough player’ – Bobby Fischer in My 60
Memorable Games.
He even played the ‘bad guy’ in his modelling activities – a unique experience for a chess player!
In other sports, he is known for his determination – don’t miss Grischuk’s hilarious anecdote about a
basketball game against him, and how Magnus was still able to score against a better and much bigger
player: https://t.co/k7d7SiHxDU). At the chessboard, in his statements, in his jokes, Magnus can be
rough (see his famous ‘too weak, too slow’ addressed to Fressinet during a blitz session). ‘My jokes
are drier than wood!’ he declared recently.
As he is entering his 30s, he seems cooler and more relaxed. He is also very generous in sharing his
thoughts frankly during Banter Blitz sessions, or when playing with youngsters, for example in
Internet blitz matches – what an experience for them to be able to challenge the best player in the
world, as they try to imitate him. He also enjoys sharing training sessions with rivals, as he did with
Aronian just before the Armenian returned to the top by winning the 2015 St Louis tournament. He

27
often helps young Norwegians train and get invitations to strong tournaments.
He usually acknowledges defeat without making excuses and can be elegant:

Ding Liren 2791

Magnus Carlsen 2863
Chess Masters Final 2020 (2)

Magnus had won the first game in the final of this online event due to a disconnection by his
opponent in a clearly drawn position. In the second game he reacted as follows with black:

1.c4 e6 2.g3

2...Qg5 3.Bg2 Qxd2+ 4.Qxd2 1-0

In the early 1990s, in a rapid event in Paris, Anand arrived late for a game vs Kasparov due to a
scheduling error. Garry took the point with black, and had to draw with white in the second game to
win the match (he won the game and the match). Perhaps the situation was not comparable, because
the error was (at least partially) Anand’s. Anyway, for the spectators it was a shame that the final was
over almost before it started. However, what Giri said after magnanimous Magnus returned the point
to Ding Liren was interesting: ‘Not many people even thought about that possibility. At the end of the
day, you want your point, at any cost (...) I don’t think that would have happened to anyone else.’
How many times have we heard that about a Magnus game? It also happens in life sometimes:
Carlsen has the ability to find the right move in a situation where no one else can find an acceptable
answer! No doubt he will be imitated, as his decisions on the chessboard are imitated, but kudos to
the originator of the process.
Carlsen always speaks his mind in public and tends to be very frank and direct. If he doesn’t feel

28
like doing this, he doesn’t speak. Most notably, he refused to attend a post-game press conference
during the 2016 World Championship Match, and was given a large fine. When he does speak or
comment on chess, he is incredibly precise and penetrating. Once, Grischuk complained about this, as
he felt that Carlsen sometimes revealed professional secrets. When he talks about certain rivals, he
does not hesitate to say what he thinks.
About himself, what he sees in the game, what he thinks, even personal things sometimes (‘Women
hate me!’ he once said), he is incredibly open – the first world champion ever to share his thoughts in
such a serious, almost naive way. His most famous statement was made during the 2018 World
Championship Match in London, when he was asked who his favourite player from the past was: ‘I’m
not really the person to have idols... and I would say, with that in mind, my favourite player from the
past is probably... myself, three or four years ago.’

Ni Hua (2665)
Magnus Carlsen (2801)
London 2009 (5)

1.e4 c5 2.Nf3 d6 3.Bb5+ Nd7 4.d4 a6 5.Bxd7+ Bxd7 6.dxc5 dxc5 7.Nc3 e6 8.Bf4 Ne7 9.Ne5 Ng6
10.Qh5

I remember seeing this game in London – what a tournament it was back then, with the best players
invited on a modern stage in Kensington. It was a treat for the spectators. I went several times, with
my friend Samy Sahraoui, or with my wife Stephanie Menase and my son Hugo. From the centre of
Paris, where I live, to the centre of London, the train under the Channel takes 2 h 15 m. Then a cab,
direct to the tournament room, and a cup of tea with a cloud of milk. And in the tournament room, a
seat next to Carlsen’s board! It was the first time I saw him play live.
Here, Carlsen took a very long time wondering what to play – more than twenty minutes, as far as I

29
remember. Apparently this is a difficult position for Black. Thanks to White’s advantage in
development, there are lots of problems and it looks like Black will have a hard time fighting for the
draw. I tried to figure out several candidate moves myself, but I did not guess his – the very best one,
a great novelty found ‘at the board’.
What I had already noticed in Carlsen’s attitude was not only a real desire to find the right move
(which all players have), but the intuition that he would not play until he had completely understood
and ‘felt’ the position. Chess players have different attitudes at the chessboard. If you look at
Kramnik or Gelfand, you immediately feel that they are hard workers. On the other hand, Aronian or
Vachier-Lagrave look relaxed, and you will see them walking and smiling very often. Kasparov’s
attitude was impressive, and you could feel both his fighting spirit and his fantastic power of
concentration, isolating him from everything else, as if there was a one-metre wide armour separating
him from the rest of the world. Caruana was showing a lot of energy during the 2018 World
Championship Match. He was sitting very close to the table, far forward in his chair (sitting like a
jockey on his horse, it seemed to me), and constantly looking at the chessboard with his head above
the pieces. He looked like a nervous fighter who was calculating millions of moves, and he was
impressive. On the other hand, Carlsen is sometimes lying down (like on a couch, with his legs
stretched out to the side) rather than sitting, and sometimes stays away from the board, often not
looking at it. This is the attitude of a thinker who is trying to understand the position and looking for
ideas, rather than calculating. I already had this impression in 2009, when I attended this game. The
Norwegian was the youngest player in the event, but he looked more like an old philosopher, who
was not involved in questions of competition, time or position, but rather wondered what was going
on, and one felt that he would take all the time necessary to understand it perfectly. Although Carlsen
is arguably the most practical player of all time – Karpov could be a contender for that title – he also
has a researcher’s attitude at the board. What is often described as phenomenal positional intuition is
the result of a lot of dedication and hard work!

10...Bc6!

This move not only allows Black to fight for equality, but also gives an unbalanced game with
chances to play for a win.

11.Bg3

On 11.Nxc6, there are complications starting with 11...Nxf4 12.Nxd8 Nxh5 13.Nxb7 Rb8. Also
interesting is 11.Nxg6 fxg6 12.Qg4 Qf6 13.h4.

11...Nxe5 12.Bxe5

30
12...c4!

Another great move, in the style of Carlsen. The pawn advance opens two lines for the second player:
the f8-a3 diagonal for the bishop and the fifth rank for the queen. It also announces the pawn attack
on the queenside – ...b7-b5, ...a6-a5 and ...b5-b4 to follow – which will eventually win the game. The
move 12...c4 certainly has a solid positional basis. The point is to play it now, when Black still has
time to develop all his pieces (all but one are on their original squares) and two moves away from
castling, while White is about to complete the full mobilization of his army.

13.0-0 Qa5 14.Qg5 h6 15.Qg3

31
15...f6!

Thinking without prejudice, Carlsen allows his opponent to attack him, so that the king will have to
stay in the middle of the board, preventing the f8-bishop from moving.
The Carlsenian 15...h5 was interesting, when Black intends to develop the rook via h6-g6 and put the
king on f8 (say after ...Be7).

16.Qg6+ Ke7 17.Bf4 Be8 18.Qg3

32
18...Kf7!

Now, Black is safe and has a healthy position.

19.Rad1 Bc6 20.Rd2 e5 21.Be3 Bb4 22.f4 Rhe8 23.f5 Bc5 24.Rfd1 Rad8 25.Rxd8 Bxe3+ 26.Qxe3
Rxd8 27.Rxd8 Qxd8 28.Kf2 Qd6 29.a3 a5

29...b5 30.Qa7+.

30.Kf3 Kg8 31.g3

31.Qe2! b5 32.Qd1=.

31...b5 32.Ke2 b4 33.axb4 axb4 34.Nd1 Ba4 35.b3

35.Qd2! Qd4 36.Qxd4 exd4 37.Kd2 Kf7 (37...Kf8 38.b3) 38.b3! cxb3 39.cxb3 Bxb3 40.Nb2!
should hold.

35...cxb3 36.cxb3 Qa6+ 37.Kd2 Bb5–+

Classical wisdom asserts that Q+N is superior to Q+B, but Carlsen is unimpressed with such ‘ready-
made’ thinking.

38.Qc5 Qa2+ 39.Qc2 Qa7! 40.Qc8+ Kh7 41.Kc1 Qa1+ 42.Kc2 Qd4

And White resigned.

We will study Carlsen’s play through twelve chapters, the first two of them (about style and opening
play) without any positions to solve, then ten chapters with an introduction and a series of diagrams.
The final chapter contains the games and the solutions to the quiz. You can read it independently
like a traditional game selection, if you want to escape the ‘decathlon’ offered in Chapters 3-12. But I
hope that you will accept the challenge!

33
 Chapter 1
Style: from Karpov to Tal?

The style of Carlsen as a young player had the fluidity of the American genius Paul Morphy.
Everything flows smoothly, there are tactical highlights and all the pieces connect easily in harmony
towards the general idea. This style appeals to the public, of course, and this is probably why the
famous chess player and critic, Lubos Kavalek, called the young Magnus ‘the Mozart of chess’.
Occasionally, and when the opponent allows it, this beautiful style comes out, as for example in the
second game of the 2014 World Championship Match against Anand, when Carlsen mounts an attack
from scratch with the moves a2-a4, dxe5, Ra1-a3-g3, Nf3-h4-f5, etc. This ‘natural’ style based on
perfect piece placement and good rhythm perception explains why Carlsen is so good in shorter time
controls. In rapid games and blitz (Internet or live), Carlsen sometimes shows such performances,
which are only possible when the opponent is not completely up to it (or doesn’t have time to be).
It seems that Magnus himself likes this light and airy style. Recently, he praised Fischer and
lamented that he could not play with as much style as the legendary Bobby. ‘What I admired most
about him was his ability to make what was actually so difficult seem easy. I try to emulate him,’ was
his reaction to Bobby’s death. In fact, he is certainly capable of playing with style, but it turns out that
given the level of his opponents, the ‘Mozart’ way is no longer enough to win against strong
defenders.
Thus, Magnus’ style became progressively more complex, as he felt that fluidity was not enough to
defeat strong GMs. As a young and evolving player, he was tempted by the attacking game and then
by the technical game. He tried different repertoires of openings (well, all of them, really!), and in
fact all openings except for losing gambits.
In 2010, when he became the top player in the Elo ranking, Carlsen was known for his ability to
win from nothing and from even positions, especially in the ending. That’s why he was compared to
Karpov in the beginning. His game was considered a bit dry and too dull by many critics. Aronian
said, ‘When I looked at Carlsen’s play I never imagined he’d become a top player... I never liked his
style. He tried more to put pressure on his opponent mentally and didn’t have a particular chess style.’
Kortchnoi even said (in 2008, after Carlsen’s joint victory in Wijk aan Zee): ‘I don’t have a great
opinion of Carlsen. He’s an unusually weak player, just lucky, he doesn’t understand much about
strategy.’
Technique in simple positions remains his strong point. Aronian said in 2017 about Carlsen, ‘It’s the
Petrosian/Fischer approach – let’s manoeuvre and see who turns out to be the best.’ However, his
game has evolved and over the years Carlsen has become more attracted to the attack, and eager to
play more complex and dynamic games.
The closest to perfect description of Carlsen’s style, at least before 2018, is Gary Kasparov:
‘Magnus is a lethal combination of Karpov and Fischer.’
Fischer had the same strengths as Karpov: endgame technique, ability to ‘press’ and fighting spirit,
but Fischer’s style is completely different; he was more of a ‘piece player’, while Karpov is more of a

34
pawn player. Karpov looks for long-term advantages, a safe pawn structure, slow play, and will
arrange his pieces accordingly. Fischer wanted above all active pieces and ‘air’ as Elie Agur
described it so well (‘The poetry of empty squares’) in his remarkable book Bobby Fischer: A Study
of His Approach to Chess (Cadogan 1992), and he liked the initiative and immediate actions. One is
very patient, the other is impatient, always looking for a decisive move. Magnus is more energetic
than Anatoly, but less direct and straightforward than Bobby. In fact, he is a master of alternating
between a slow and a fast pace both in terms of the positions he chooses, the moves he selects and
also the management of his thinking time.
Today (2021), Carlsen is mostly considered a complete player. Benjamin: ‘Magnus is the total
package. Deep intuition of strategy, tremendous endgame play, and consistently reliable calculation.’
Dubov, on Carlsen’s weaknesses: ‘His strength is that he doesn’t particularly have any. And therefore
he can play in any style, and there’s a certain level below which he never falls’ (‘Shakhmatist Dubov:
Titul Chempyona Mira dlya Menya Poka Abstraktnaya Chel’ – Sport RIA Novosty, 2020, ria.ru). The
very conspicuous Vishy Anand had already stated in 2013: ‘He has the ability to play in any position.
He is a complete player in chess. He is simply the greatest talent I have seen.’ Kasparov recently
(2020) added: ‘He is a combination of many great players of the past: he has Karpov’s ability to place
his pieces in perfect positions, Fischer’s rage on the board and maybe also myself/Botvinnik’s, the
willingness to analyse his games and find improvements. He is a universal type of player, and that
probably explains his dominance’ (Chess24 Legends: ‘Carlsen in the final | Giri strikes back’).
The key to Carlsen’s style is that active piece play is always more important to him than material
considerations, and he is always willing to sacrifice long-term advantages such as material (which he
does incredibly often) or pawn structure, in favour of initiative and activity. This remains true in all
three phases of the game, which is why I consider him closer to Fischer than to Karpov stylistically.
He likes the initiative, dictating the game as Bobby did. Philosophically, however, Carlsen is closer to
Karpov, because he is a pragmatist, like Capablanca, rather than a purist like Fischer. Carlsen is also
very versatile, with the feel of a real coffeehouse player. Depending on the situation in a game or a
match, he is ready for any adventure, sacrifice or pawn capture, gamble or trick (see Game 49, for
example). Thus, if he considers a decision to be right, he does not hesitate to play an objectively
dubious move (see Game 53). This corresponds to the way Lasker, Capablanca or Tal thought chess
should be played, rather than the perfect chess advocated by idealists, where one must always play the
best move.
Carlsen recently (2020) compared himself to Reuben Fine, a player whose style is similar to that of
Capablanca and Fischer. The American player was one of the young aspiring champions before
World War II, along with Botvinnik and Keres, but then decided not to become a chess professional
and had a career as a psychoanalyst.
It seems to me that Carlsen is at his best in a certain type of position: rather quiet, with an
asymmetrical structure, with an early pawn exchange, for example, like the Carlsbad structure (1.d4
d5 2.c4 e6 3.Nc3 Nf6 4.cxd5 exd5), or one or two minor piece exchanges at the beginning to have
more room for the remaining pieces, like in the Sicilian with 3.Bb5(+). Of course, the ideal Carlsen
game is presented in the Carlsen-Kramnik 2019 blitz game with both a Carlsbad structure and early
exchanges, which happens quite often in the London System.

35
 One thing that separates Magnus from the other ‘positional/intuitive players’ (Capablanca,
Smyslov, Petrosian and Karpov) and from all the other World Champions except Tal, is the risk he is
willing to take at any time. The fact that he offers large numbers of pawns to his opponents is a
distinctive feature of his game, and is quite original for a ‘positional player’. He tries to win from
almost any position and under any circumstances, be it in a World Championship match or in a casual
game.
Each player is unique, of course, but in a first approach, it makes sense to draw some comparisons.
It is difficult to define Carlsen’s style, because he has studied all the great players and taken the best
from them, such as the classical principles of Philidor and Steinitz, the art of combat from Lasker, the
clarity and vision of Capablanca, the precision of Fischer or the sense of the placement of the pieces
of Karpov7.
7 ‘What makes Karpov exceptional?’ I asked Artur Jussupow in 2002. ‘Karpov is the only player
who, if you give him one or two seconds in any random position, will play a good move,’ he
explained to me.
There is an unusual cocktail inside Carlsen, combining an extreme delicacy to intimately perceive
the secrets of the position, almost any position, and an indomitable personality, which frequently
incites him to go for victory at all costs. Technically, he is able to ‘torture’ (his own words) the
opponent in a slightly better position, but also, in an equal position, to take any risk to unbalance the
game. He is the combination of a cold intelligence, with perfect evaluation abilities, and the
temperament of the ultimate fighter. He is a Petrosian with the soul of a Tal.
Throughout this book, we will examine the specifics of Magnus’ playing style in the main technical
aspects of the game. Nevertheless, we can emphasize some general features.
‘I recognize the style. But you see, when I played a game with Bobby, there was no style. Bobby
was playing perfectly. And perfection has no style,’ said grandmaster Miguel Najdorf. I agree with
Najdorf here. In many positions, all strong players play the same moves. Nevertheless, all players
(even Bobby, as Agur has masterfully shown) have stylistic biases. Some are a consequence of their
strengths, e.g. Carlsen is a fan of active defence, but sometimes (not often) he loses because he
defends too actively.
To observe Magnus’ style, it is not necessary to look at positions where Carlsen, but also Caruana,
Ding Liren, Aronian, So, Vachier-Lagrave, Nepomniachtchi, Giri would play the same move. On
some occasions, Carlsen plays suboptimal moves, but when he has a choice between several playable
moves of equal strength, his choice is more relevant. Thus, we will also examine in this chapter some
positions in which Carlsen’s choice is questionable.
1. The piece player
In general, Carlsen is a ‘piece player’. Since he likes his pieces to be active, this ideally means that
some pieces and pawns are exchanged early. On a full board, with 64 squares available for 32 men, at
least some of them will run out of room to breathe. This is why Carlsen does not like certain closed
openings like the French or the Caro-Kann with black, where some pieces will be passive due to the
lack of space.

36
 He is not at his best as White against the King’s Indian, or even against irregular king’s fianchetto
systems, because here White’s strategy involves strangling Black: pushing his pawns to the centre,
blocking the position and keeping a space advantage. This strategy is a bit passive for his liking, and
White also traps his own men. He often prefers the Black side and attacking the white king. On the
other hand, he plays well against the Grünfeld and the Semi-Tarrasch because there are quick
exchanges, and the pieces enjoy more freedom, while White enjoys an advantage in the centre and
attacking chances. This is very close to his ideal in chess.
A) Pawn sacrifice to give ‘air’ to the pieces
This is probably the most distinctive feature of Magnus’ style. He hates it when one of his pieces is
trapped by one of his own pawns. Thus, he will often give such a pawn away to open a line or square.

Magnus Carlsen 2862

Alireza Firouzja 2749
Wijk aan Zee 2021 (1)

Here, White’s pieces are well-placed, but to give them more room to breathe, Carlsen sacrificed
another pawn with 28.e6.

Magnus Carlsen 2861

Erwin l’Ami 2627
Wijk aan Zee 2013 (10)

37
Here, White benefits from a stable pawn formation and from the bishop pair. Black is, however, very
solid, and Carlsen decided to give his pieces more air with 12.c4, a move that weakened his own
pawn structure (see Game 134).

Magnus Carlsen 2775

Levon Aronian 2737
Bilbao 2008 (6)

Carlsen has already sacrificed a pawn; he decides to give up another one with the move 15.d5,
creating space for his pieces at the cost of the proud d4-pawn.
B) Going for the ending

38
Magnus often goes deliberately for an endgame, hoping to outplay the opponent thanks to his superior
technique.

Magnus Carlsen 2581

Nigel Short 2687
Hoogeveen 2004 (3)

White has a chance to attack if he keeps the queens on the board, because of Black’s weak king
protection. Nevertheless, young Magnus opted for a queen exchange with 29.Qf6 and duly won.

Magnus Carlsen 2843

Fabiano Caruana 2773
Sao Paulo/Bilbao 2012 (6)

39
Carlsen played 21.dxe4, having calculated a likely liquidation (...Bxf3, ...Ne5, ...Nxf3+, ...Qd3 and
...Qxe4). ‘I played this and the next few moves quickly, as I couldn’t wait to get to torture him in the
endgame.’ According to him, Nd4 was better.
C) Backward moves
Magnus likes backward moves, to reorganize his pieces. He also likes paradoxical and fancy moves.

Magnus Carlsen 2863

Jeffery Xiong 2709
Clutch Chess Showdown 2020 (1.10)

40
 White to move

In this rapid Internet game, Carlsen made a typical decision: 7.Ng1!?.

The computer recommends Nc1.

Magnus Carlsen 2877

Radoslaw Wojtaszek 2735
Tromsø ol 2014 (4)

White to move

Here, Carlsen played 7.Nh3. ‘[...] part of the reason I played it is because it looks cool’.
2. The fighter
A) Dictating the game
Magnus wants to impose his choices. This is one of the reasons why he plays ‘irregular’ openings,
from the first move (Scandinavian and Alekhine) or the second (Trompowsky and London). Another
way is to play for the initiative, i.e. to create direct threats, forcing the opponent to react.

Magnus Carlsen
Hannes Stefansson
Moscow 2004 (8)

41
 White to move

Carlsen correctly gave up his queenside pawns and played the excellent 13.Rd1, to castle quickly.
Even when defending, Carlsen almost always chooses an active move:

Magnus Carlsen 2835

Veselin Topalov 2770
Wijk aan Zee 2012 (11)

White to move

42
White is fighting for equality, and could achieve it with 19.Rc1.
Carlsen decided to go all out and played 19.Nh4!?, with the intention of playing Nf5. Black could
and did follow up with ...Nb6-c4xb2. The game went crazy and Carlsen won.

A new habit that Carlsen brought to high-level chess was to play any position to the end.

Ruslan Ponomariov 2737

Magnus Carlsen 2813
Nice rapid 2010 (10)

In a rapid game, Carlsen played on and won this ending!

After he became a contender for the world title, chess became a real sport for top players. Instead of
just playing your novelty at move 22 and accepting a draw at move 25 if the opponent was prepared,
against the young Carlsen you had to be ready for any first move in the opening, and you had to have
the stamina to fight to the last pawn in the endgame. Today, such play is normal, and even legitimate,
but it was a completely new approach in 2010.
B) Risk-taking
Carlsen is willing to play to win in any position, so he accepts almost any challenge. He is ready and
willing to:

i) take ‘poisoned pawns’:

Hikaru Nakamura 2758

Magnus Carlsen 2826
Moscow 2011 (9)

43
 Black to move

Black is fine if he plays a solid move like 15...Rc8,

but Magnus decided to go for the centre pawn with 15...g5 and (16.Be3) 16...g4, a Larsen-style
fighting decision!
Another well-known example is Game 149 against Anand.

ii) give material at any time, to throw the game off-balance:

Magnus Carlsen 2690

Alexander Morozevich 2741
Morelia/Linares 2007 (1)

44
White is fine here, but instead of playing calmly, he decided to play the outrageous 13.Nxd6!?. This
is exactly the kind of speculative sacrifice that Tal, Bronstein and Spassky liked so much!

iii) accept an attack on his king, although he doesn’t like it very much:

Viswanathan Anand 2775

Magnus Carlsen 2870
Chennai Wch m 2013 (9)

45
 Black to move

Closing the centre with 8...c4, as Carlsen did, requires courage, because it means accepting a terrible
attack on the king (see Game 148).

iv) play any spoiled pawn structure, if he sees possible advantages for him:

Anish Giri 2779

Magnus Carlsen 2875
Zagreb 2019 (1)

Carlsen played 7...d6 here, accepting a ruined pawn structure after 8.exd6.
3. Stylistic preferences
Carlsen plays any opening with white or black. However, every player has preferences, and
sometimes even ‘lucky moves’.
For example, we will see that Carlsen clearly favours endgames with opposite-coloured bishops.
We will often see that he prefers to attack the enemy king rather than defend his own, to have activity
rather than material, and asymmetrical rather than symmetrical pawn structures. As Black, he doesn’t
often allow his opponents to engage in a boring Slav or French Exchange, although as White he may
adopt these openings from time to time.
But more specifically, I’ve seen a move that has come up repeatedly in various crucial occasions in
his early years as the World #1. In the following four examples, Carlsen plays a move he particularly
likes: bringing his queen to d3 early in the game in front of the undeveloped bishop on f1. In these
four positions, the move is clearly best in the second case only. This case is also an exception,
because the bishop on f1 has already been developed, but has returned to its original square after

46
White castled kingside. We should note that these games were of crucial importance in Carlsen’s
career, as the first and third are from World Championship Matches against Anand, where the scores
were 0-0 and 1-1, while the second is from the Candidates Tournament in 2013. Since Carlsen won
all four games, we could say it was a lucky move for him! Perhaps in general he has a weakness for
queen moves in front of an undeveloped bishop – see his Qd2 move against Xiong in 2020.

position 1

10.Qd3!? Carlsen-Anand, Chennai Wch m 2013 (5) 1-0

position 2

47
15.Qd3! Carlsen-Gelfand, London ct 2013 (10) 1-0

position 3

7.Qd3 Carlsen-Anand, Sochi Wch m 2014 (6) 1-0

position 4

13.Qd3!?

48
 Chapter 2
The opening revolution

Henrik Carlsen (ChessBase 2019): ‘You know, but still people would tell us that he is too superficial,
he needs to have a proper repertoire, learn the openings in detail etc., but we heard this at all stages of
his development. It never came in between his growth!’
‘Carlsen’s weakness is above all the opening. At this stage of the game he is not a strong player. He
should be outplayed in the opening, while his opponents do the opposite. At the moment Carlsen
shows good practical play, but if Kasparov won’t coach him with his preparation methods I would
say Magnus’ future doesn’t look so promising. He is surely a very talented player, but only those who
will be making the strongest moves can become the champions. Only the scientific component of
chess has a chance to win nowadays. Not a single practical player has a chance of becoming the
champion without a scientific approach to the subject’ – GM Evgeny Sveshnikov in an article called
‘You should outplay Carlsen in the opening, but his opponents play h7-h5...’ on chess-news.ru in
2013.
Carlsen brought about a complete revolution in this regard. Before his reign, the best players
prepared long opening lines in the hope of surprising their opponents with a ‘theoretical novelty’
(how old-fashioned this term seems today!). Several openings were considered ‘acceptable’ at the
world top level, the strongest for Black being the Queen’s Gambit and the Spanish Opening8, while
many were discarded as not strong enough.
8 By the way, those two openings are still Magnus Carlsen’s main repertoire, when he plays it solid
with black.
Occasionally, a player would attempt a surprise opening, even at the highest level, but the motto
remained the same: pick an opening – a solid one if you had world title ambitions – and dive in and
study the complications that may arise as deeply as possible. The level of preparation was very high,
and it was even deepened with the use of computers. In the early 2000s, all top-level players studied
the same Semi-Slav lines – or the Marshall Gambit, the fashionable lines at the highest level during
the post-Kasparov years. Then came Carlsen, who chose to do things his own way, i.e. change
openings as much as possible, using his encyclopedic understanding of any position (because he had
already tried all openings, and any pawn structure!). His goal was to surprise the opponent by using
‘fancy’ openings, less studied and strong enough not to lose by force. He thus used the Alekhine
Defence to crush Topalov, the Scandinavian (with 3...Qd8!?) against Caruana and the Trompovsky
Opening against Kramnik and Karjakin at the 2016 World Championship Match, and many others,
like the famous ‘Norwegian Dragon’ seen in Adams-Carlsen (1.e4 g6 2.d4 Nf6 3.e5 Nh5, 1-0, 38). At
the top level, it was a real revolution. Suddenly, a new type of player appeared, Carlsen’s children in
a way, who are able to play most openings and most types of positions – So, Giri, Caruana. Great
players like Anand, Kramnik and Topalov would adapt, while some disappeared with the new trend:
they could no longer rely on their superior preparation in long forcing lines. Leko (born in 1979),
Morozevich (born in 1977) and even Shirov (born in 1972) are among the big names that come to
mind. Perhaps they had other reasons for declining so early, but it seems to me that heavily-prepared

49
players like these three were put at a disadvantage by the new way of playing chess in the early
Carlsen era, with the battle beginning on the very first moves. In a 2017 article on ChessPro called
‘Porazhennya perenoshu legche, chem pobeduy’, Levon Aronian commented on how Carlsen
changed chess: ‘He managed to minimize the role played by the opening... Carlsen has changed
modern chess and the majority of players now seek ways to get off the beaten path as soon as possible
in the opening, to get a non-standard position. That’s prolonging the era of classical chess.’
In openings, Carlsen’s second major innovation is a systematic structural approach to positions,
instead of the classic theoretical method of looking for innovations move by move. Carlsen plays a
pawn structure, and often uses it with both colours. For example, in the early 2010s he played the
Dragon as Black, and also the English Opening with g2-g3 as White, i.e. exactly the same variation!
And the biggest opening novelty of the 2018 World Championship Match was unveiled only in the
rapid playoffs:

Magnus Carlsen 2835

Fabiano Caruana 2832
London Wch m playoff rapid 2018 (1)

White to move

Here, White plays 4.e4! (an idea of Fressinet). Carlsen reaches a position he had obtained with black
during the same championship – now with a tempo up!
White obtained an advantage in the opening and eventually won in the endgame (1-0, 55).
The more mature Carlsen is slightly less adventurous in the openings. For one thing, he now has a
great team around him. This is one of the privileges of world champions: you have more money to
pay seconds, and prestige to attract young players to work for you and bring you new ideas in
exchange for getting a solid sparring partner. Dubov became the 2018 rapid World Champion after

50
such a collaboration during Carlsen’s preparation for the 2018 World Championship Match, in part
due to the fact that he played many training games with Magnus.
Second, Carlsen is now one of the oldest players in the elite group. When he started his opening
revolution, his main rivals were Anand, Kramnik, Topalov and Aronian, all of whom were (much)
older than he. Today, he is more inclined to opt for classical play – if not always classical openings –
against the younger generation. He is also willing to play main lines much more than before.

Magnus Carlsen 2881

Ding Liren 2836
Legends of Chess Prelims 2020 (6.2)

White to move

In the diagrammed position, Magnus often played the solid 5.e3, leading to the Meran.
However, more recently he successfully adopted the more ambitious line with 5.Bg5 h6 6.Bh4
(instead of the simple 6.Bxf6, which would probably have been chosen by the young Magnus),
leading to one of the most complex lines in modern opening theory.
Here is another proof of how the experienced Magnus is prepared:

Magnus Carlsen 2872

Maxime Vachier-Lagrave 2780
London 2019 (1)

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Bg5 e6 7.f3 h6 8.Be3!?

51
This is a very strange move; hasn’t White lost a tempo in a classical English Attack with 6.Be3 ?
Probably, the idea is to use Black’s move of the h-pawn to be able to open lines later, playing g4-g5
after adequate preparation. This novelty was an idea found during a match between AI engines.
Nowadays, most of the best players follow these games very seriously to use the novelties in their
own games (½-½, 32).

White to move

Here is a great Carlsen novelty against Kramnik, which was shown to him by Hammer (Game 55):
12.Ne2!

52
The following (diagram right) is probably Carlsen’s main novelty, a very subtle move involving a
loss of tempo.
After ...Bf5 has been played earlier, White decides not to exchange bishops with 6.Be2, and Black
makes room for his bishop with 6...h6.

White to move

Then, White offers the bishop exchange after 7.Bd3!, a Petrosian-like finesse played in Carlsen-
Nakamura at the Gashimov Memorial in 2014 (1-0, 61).

Carlsen is a scholar and it is impossible to surprise him by playing ‘small lines’. After the game,

53
Grandelius was asked about the moment of his decisive error. ‘Probably on move two,’ replied the
Swede (see Game 119).
At the end of 2020, Kramnik (and not only him) estimated that among the strongest chess players, the
best prepared in the openings was Magnus Carlsen.

54
 Chapter 3
Attack: inviting everyone to the party

Carlsen has made a video on attacking on Chess 24 with the interesting title ‘Attacking without
sacrificing’.
There are two sides to the attacking game. You have to be able to bring all the pieces to the
direction of the enemy king. Regarding the ability to ‘invite everyone to the party’, Carlsen has no
equal, see for example his 2019 games against Rapport (104), or Giri (105). Basically, Carlsen likes
to attack on the kingside, and opts more and more for a direct attack as often as possible. He also likes
to quickly surrender a pawn – especially the g-pawn – to open up a file for the coming attack.
However, he sometimes lacks the instinct of patented king hunters like Tal or Kasparov to finish
the game. Let’s first take the example of another player who gave detailed explanations of how he
actually thought in the game (from Gelfand’s Dynamic Decision Making in Chess, Quality Chess
2017):

Boris Gelfand 2740

Vladimir Kramnik 2810
London ct 2013 (7)

White to move

19.Ned2 ... ½-½ (36)

The natural move is 19.Nfg5, threatening to sacrifice the knight on f7 or h7. 19...g6 is the main
defence as White wins quite easily against other attempts (if 19...h6, the simple 20.Qh5 will do, e.g.

55
20...hxg5 21.Nxg5 Nf6 22.Qxf7+ Kh8 23.Nxe6 (23.Qxe6!+– is the computer’s move but a human
player won’t calculate it) 23...Qg8 24.Qxb7 Rb8 25.Qxg7+ Qxg7 26.Nxg7 Kxg7 27.Rc6, and White
will be five pawns against a piece after taking a6 next) 20.Nxf7 Kxf7 21.Ng5+ Kf6 22.Qxe6+ Kxg5.

analysis diagram

Here, you have to see a nice switchback manoeuvre: 23.Qh3!. This is strictly the only move,
threatening mate on h4 (taking h7 would also be a huge advantage for White). Gelfand saw it: 23...h5
24.Qe6!, and there is no defence, as the black pieces on the queenside are unable to help their lone
king against the assault of the remaining white attackers. Gelfand could not see the return of the
queen after 23...h5, so he decided not to play this variation. I gave the position to some of my students
from a group of players with a rating between 2000 and 2300 with the indication ‘White to play and
win’ and some (not many) found the winning line, so I guess it’s not too difficult for a super GM like
Gelfand. The problem is that you have to be sure that White wins to find such moves and Gelfand,
while a strong calculator, is not as strong an attacker as the famous king hunters. This means that his
instincts are not as developed, and his specific attacker’s intuition does not allow him to be 100% sure
that the position is a forced win. Thus, when he faces a problem, like the Qe6-h3-e6 manoeuvre, he
may give up too early.
By the way, 19.Neg5 was also winning; if 19...h6 20.Bg6!+–.
We can recall some Carlsen misfires that are comparable, e.g. his game against Wojtaszek (Game
203), when he misses the standard and thematic attacking move Nd5 in the Sicilian, or even in the
aforementioned 2019 game against Giri, where he is able to spot a beautiful and bold positional
sacrifice, but slightly spoils the finish because he cannot calculate the finishing blow to the end.
Recently, in an Internet game against So, he twice (Games 239, 240) missed the opportunity to
gloriously end the game with a winning sacrifice. ‘I didn’t trust my intuition,’ he said of two

56
positions, in which he saw the winning move but didn’t dare to play it. Here, it’s debatable whether to
consider these examples, due to the lack of time to do the proper calculations, but the process is the
same. Carlsen feels he is winning, finds the right move but is not able to calculate it to the end, and
plays another, weaker move. Of course, players like Carlsen or Gelfand do not try to bluff their
opponents by playing moves they are not sure of. The problem here is that the calculation lacks
precision, because there is some doubt about the final result of the sacrifice, a doubt that would not
enter the mind of a natural attacker. Meanwhile, Carlsen is able to spot incredibly bold positional
sacrifices that are objectively much riskier. Thus, while Carlsen is clearly able to calculate well, while
he is able to find decisive moves at the right time to end the game in endgames, for example, and
while he is able – and often willing – to take risks, he sometimes has a problem accurately calculating
standard sacrificial attacks. It’s probably something he should work on, because stylistically he is
more inclined to play attacking chess recently.
In the following diagrams, you are asked to find the best moves, keeping in mind that the theme of
this chapter is ‘Attack: inviting everyone to the party’. The solutions can be found in the games
section in Chapter 13. In each diagram, I recommend that you first look for a move, as if you were in
a game. Then you compare it to what Carlsen played. Some of these exercises are easy (as in the
chapter on pawns or in the chapter on exchanges – sometimes only a few moves make sense), others
are difficult (they require difficult calculations or a fine understanding of the position); there is no
order of difficulty. Carlsen has not solved them all in ‘real life’. To make the best use of the material,
you should first try to make up your own mind, and ‘play’ your move before comparing it to the
Champion’s.
Remember, the diagram numbers refer to the order within the games as they are given in Chapter
13, e.g. ‘43.3’ indicates the third diagram in Game 43.
You may also want to solve the exercises without knowing what the theme is. In that case you can
go directly to the games in Chapter 13, stop after each diagram and try to find the best move.

43.3 (go to the game)

57
 White to move

67.2 (go to the game)

White to move

67.3 (go to the game)

58
 White to move

75.1 (go to the game)

White to move

76.2 (go to the game)

59
 White to move

81.1 (go to the game)

White to move

82.2 (go to the game)

60
 White to move

84.1 (go to the game)

White to move

88.1 (go to the game)

61
 White to move

88.2 (go to the game)

White to move

90.1 (go to the game)

62
 White to move

90.2 (go to the game)

White to move

90.3 (go to the game)

63
 White to move

93.1 (go to the game)

White to move

94.3 (go to the game)

64
 White to move

104.2 (go to the game)

White to move

108.1 (go to the game)

65
 Black to move

114.2 (go to the game)

White to move

114.3 (go to the game)

66
 White to move

114.4 (go to the game)

White to move

115.1 (go to the game)

67
 White to move

115.2 (go to the game)

White to move

119.2 (go to the game)

68
 White to move

122.2 (go to the game)

White to move

122.3 (go to the game)

69
 White to move

122.4 (go to the game)

White to move

124.2 (go to the game)

70
 White to move

126.1 (go to the game)

White to move

129.4 (go to the game)

71
 White to move

156.2 (go to the game)

White to move

159.3 (go to the game)

72
 White to move

168.3 (go to the game)

Black to move

168.4 (go to the game)

73
 Black to move

182.3 (go to the game)

White to move

196.2 (go to the game)

74
 White to move

196.3 (go to the game)

White to move

196.4 (go to the game)

75
 White to move

196.5 (go to the game)

White to move

196.6 (go to the game)

76
 White to move

198.1 (go to the game)

White to move

200.1 (go to the game)

77
 Black to move

203.2 (go to the game)

White to move

204.2 (go to the game)

78
 White to move

204.4 (go to the game)

White to move

205.1 (go to the game)

79
 Black to move

205.4 (go to the game)

Black to move

205.6 (go to the game)

80
 Black to move

209.3 (go to the game)

White to move

209.4 (go to the game)

81
 White to move

213.2 (go to the game)

White to move

213.3 (go to the game)

82
 White to move

213.4 (go to the game)

White to move

215.1 (go to the game)

83
 White to move

227.2 (go to the game)

White to move

228.3 (go to the game)

84
 White to move

233.2 (go to the game)

White to move

236.1 (go to the game)

85
 Black to move

236.2 (go to the game)

Black to move

238.3 (go to the game)

86
 White to move

238.4 (go to the game)

White to move

239.1 (go to the game)

87
 Black to move

240.1 (go to the game)

White to move

244.4 (go to the game)

88
 White to move

244.6 (go to the game)

White to move

89
 Chapter 4
Defence: the preventive counter-attack

Carlsen has always been a great defender, and it’s a mark of his style: the guy is hard to beat and
almost invincible when he wants to play it safe. In his first four World Championship matches,
including the rapid games, he lost only twice. Apart from his good calculations, strong nerves and, of
course, his ability to assess what is going on at any given moment, Carlsen has several tricks up his
sleeve when all seems lost. He’s also a shrewd swindler when necessary, in the grand tradition of the
best coffeehouse players, back when gamblers made a living (or tried to make a living) playing for
money.
When he defends, Carlsen is capable of saving any position, as his famous games against Van Wely
(Game 60) and Kramnik (Game 26) show. In both cases, he was in a totally hopeless position for
most of the game and most players in the top 10 would give up or commit suicide quickly, but
Carlsen managed to get one and a half points out of two. He mixes passive defence, counterattacking
and tricks, and uses essentially the same strategy as when he is looking to win in a slightly favourable
position: make it difficult for the opponent by not allowing him to develop easy plans to finish the
game. The game against Van Wely is particularly memorable and enjoyable – if you like ugly wins!
The game against Kramnik was very important in the careers of both players. After that incredible
save by Carlsen, who had been outplayed for most of the game, the Russian player was not really
going to push for the win against the Norwegian anymore. This defensive skill has proven decisive at
times in World Championship games, ‘defence wins titles’ as they say in soccer. There is Game 148
against Anand, but especially Games 242 (Caruana) and 243 (Karjakin). In a stressful situation, with
little time on the clock, a worse position and a crucial moment in the game, Carlsen chooses an active
defence, involving material sacrifices (one piece against Caruana, a few pawns against Karjakin)
instead of opting for a passive defence, which would have held, at least according to the computers.
This decision illustrates how Carlsen thinks chess should be played in such circumstances, by going
for complications and calculations (and thus forcing the opponent to go for them as well), instead of
the more typical wait-and-see policy of what we usually consider a defensive game. Nevertheless, it
must be admitted that this active defence cost him a few points (or half points) when a passive
approach would have been more appropriate. The truth is that even when he is worse, Carlsen always
dreams of winning, so he is reluctant to sit and wait, and always hoping that the opponent will make a
mistake and give him a chance to take over. The results show that this is often the case!
Again, the diagram numbers refer to the order within the games as they are given in Chapter 13,
e.g. ‘26.1’ refers to the first diagram in Game 26.

26.1 (go to the game)

90
 Black to move

43.4 (go to the game)

White to move

60.3 (go to the game)

91
 Black to move

61.3 (go to the game)

Black to move

67.4 (go to the game)

92
 Black to move

113.1 (go to the game)

White to move

113.2 (go to the game)

93
 White to move

116.4 (go to the game)

White to move

117.2 (go to the game)

94
 Black to move

135.2 (go to the game)

White to move

138.1 (go to the game)

95
 Black to move

144.1 (go to the game)

Black to move

144.2 (go to the game)

96
 Black to move

168.1 (go to the game)

Black to move

168.2 (go to the game)

97
 Black to move

219.1 (go to the game)

Black to move

221.1 (go to the game)

98
 Black to move

222.1 (go to the game)

Black to move

228.2 (go to the game)

99
 Black to move

228.4 (go to the game)

Black to move

100
 Chapter 5
Tactics: ‘les petites combinaisons’

Carlsen is excellent at finding combinations, as combinations are the product of the position and can
be found through a mixture of intuition and accurate (short) calculation, two areas in which the
Norwegian is at his best. Occasionally he may miss one, because he is out of shape, or because there
is another solution that suits him (like in Svidler-Carlsen 2013), but this is exceptional. In general, he
quickly spots any tactical opportunity, as he is always aware of the positional needs of both sides, and
very attentive to any suspicious move by his opponent.
The style of his combinations is therefore usually crystal-clear, as it is the result of a gross mistake
by the opponent, and the punishment is quick and classical. With some more tactically-oriented
players, the combinations can be rather obscure and hidden, while in Carlsen’s case they are often
almost elementary, reminiscent of Capablanca’s famous ‘little combinations’ that allowed him to win
the game after a process of strategic domination.
Occasionally, Carlsen cannot find the right combinatory move because he is not able to calculate it
accurately: we will see some cases of missed sacrifices in the attack, like also in the famous first rapid
game of the 2018 World Championship Match in London. In this respect, he may be slightly weaker
than Anand9 at his best.
9 Aldo Haïk, who played for Lyon, the same team as Anand in the early nineties, told me about the
Indian’s unique talent and instinct for tactics. During joint analysis Vishy would sometimes stop
and declare: ‘Here White is winning.’ ‘How?’ would the co-analysts ask. ‘I don’t know, but White
is winning,’ and very quickly he would find the winning sequence.
Nevertheless, it must be said that the combination had been detected by the World Champion, as his
post-game interviews and especially his thinking time clearly attest: Carlsen certainly knows when
some accident is happening on the board, even if he is not always able to see exactly what.
In reverse, we can also see that when he makes an irreparable mistake, he is very quickly able to
realize it, as happened in the 2014 World Championship Match against Anand, when he blundered
into a very promising position, and the video attests that he was immediately aware of his mistake –
while keeping a ‘poker face’, as they say.
As always, the diagram numbers refer to the order within the games as they are given in Chapter
13, e.g. ‘11.1’ refers to the first diagram in Game 11.

11.1 (go to the game)

101
 Black to move

12.4 (go to the game)

White to move

12.5 (go to the game)

102
 White to move

20.1 (go to the game)

White to move

20.2 (go to the game)

103
 White to move

29.2 (go to the game)

White to move

31.2 (go to the game)

104
 White to move

32.1 (go to the game)

White to move

40.2 (go to the game)

105
 Black to move

62.1 (go to the game)

White to move

63.2 (go to the game)

106
 White to move

64.1 (go to the game)

White to move

64.4 (go to the game)

107
 White to move

65.1 (go to the game)

White to move

65.2 (go to the game)

108
 White to move

66.1 (go to the game)

White to move

67.1 (go to the game)

109
 White to move

75.4 (go to the game)

White to move

82.3 (go to the game)

110
 White to move

83.3 (go to the game)

White to move

85.1 (go to the game)

111
 White to move

86.6 (go to the game)

White to move

93.2 (go to the game)

112
 White to move

93.3 (go to the game)

White to move

98.1 (go to the game)

113
 White to move

104.3 (go to the game)

White to move

107.2 (go to the game)

114
 White to move

109.1 (go to the game)

White to move

110.1 (go to the game)

115
 White to move

111.2 (go to the game)

White to move

115.3 (go to the game)

116
 White to move

116.5 (go to the game)

Black to move

120.1 (go to the game)

117
 White to move

122.5 (go to the game)

White to move

123.3 (go to the game)

118
 Black to move

139.2 (go to the game)

White to move

144.5 (go to the game)

119
 Black to move

145.4 (go to the game)

Black to move

147.3 (go to the game)

120
 White to move

155.2 (go to the game)

White to move

156.3 (go to the game)

121
 White to move

157.1 (go to the game)

Black to move

162.2 (go to the game)

122
 White to move

163.1 (go to the game)

Black to move

166.1 (go to the game)

123
 Black to move

167.1 (go to the game)

Black to move

170.6 (go to the game)

124
 Black to move

171.1 (go to the game)

Black to move

172.1 (go to the game)

125
 White to move

173.3 (go to the game)

White to move

173.8 (go to the game)

126
 White to move

174.2 (go to the game)

White to move

179.1 (go to the game)

127
 Black to move

181.1 (go to the game)

Black to move

186.1 (go to the game)

128
 Black to move

189.2 (go to the game)

White to move

193.1 (go to the game)

129
 White to move

194.1 (go to the game)

White to move

195.1 (go to the game)

130
 Black to move

199.2 (go to the game)

White to move

201.1 (go to the game)

131
 Black to move

202.1 (go to the game)

White to move

204.5 (go to the game)

132
 White to move

208.1 (go to the game)

Black to move

211.1 (go to the game)

133
 Black to move

215.2 (go to the game)

White to move

216.1 (go to the game)

134
 White to move

217.1 (go to the game)

Black to move

223.2 (go to the game)

135
 Black to move

228.5 (go to the game)

White to move

229.1 (go to the game)

136
 White to move

232.3 (go to the game)

White to move

233.3 (go to the game)

137
 White to move

234.2 (go to the game)

White to move

235.1 (go to the game)

138
 White to move

244.7 (go to the game)

White to move

246.1 (go to the game)

139
White to move

140
 Chapter 6
Exchanges: Carlsen’s main positional weapon

Exchanges are one of the main weapons in the arsenal of any chess player for many reasons –
defensive reasons (exchanging queens, for example), attacking purposes like exchanging the
opponent’s fianchetto bishop, or the f6-/f3-knight that defends the castled king. In the endgame, the
choice of the right exchange is crucial to get the desired result. In the opening, it may be important to
exchange an opponent’s active piece.
Carlsen’s exchanges deal with all these points, but the Norwegian exchanges for another reason: he
likes to change the position as often as possible, and an exchange is a way for him to confront his
opponents with new problems, apart from the objective advantage of the transaction. When Carlsen
fails to gain an advantage with a certain set of material, he provokes exchanges so that there are
different problems to be solved. A classic example is his 2020 victory over Firouzja in the pawn
endgame, after trying many different endings.
In the following diagrams, you must decide whether a possible exchange is favourable or not. The
diagram numbers refer to the order within the games as they are given in Chapter 13, e.g. ‘3.2’ refers
to the second diagram in Game 3.

3.2 (go to the game)

Black to move

4.1 (go to the game)

141
 Black to move

8.1 (go to the game)

White to move

12.1 (go to the game)

142
 White to move

38.1 (go to the game)

White to move

47.1 (go to the game)

143
 White to move

51.1 (go to the game)

White to move

57.1 (go to the game)

144
 White to move

63.1 (go to the game)

White to move

64.5 (go to the game)

145
 White to move

74.2 (go to the game)

White to move

74.3 (go to the game)

146
 White to move

75.3 (go to the game)

White to move

78.1 (go to the game)

147
 White to move

79.1 (go to the game)

White to move

103.1 (go to the game)

148
 White to move

108.2 (go to the game)

Black to move

120.2 (go to the game)

149
 White to move

122.1 (go to the game)

White to move

147.1 (go to the game)

150
 White to move

154.1 (go to the game)

Black to move

156.1 (go to the game)

151
 White to move

173.6 (go to the game)

White to move

175.1 (go to the game)

152
 White to move

177.2 (go to the game)

White to move

182.2 (go to the game)

153
 White to move

192.2 (go to the game)

White to move

203.3 (go to the game)

154
 White to move

206.1 (go to the game)

Black to move

209.1 (go to the game)

155
 White to move

220.5 (go to the game)

White to move

231.3 (go to the game)

156
 White to move

237.2 (go to the game)

White to move

245.2 (go to the game)

157
 Black to move

245.3 (go to the game)

Black to move

245.4 (go to the game)

158
Black to move

159
 Chapter 7
Calculation: keeping a clear mind

Carlsen’s calculation ability is undoubtedly one of the best in the world, but this is where people
believe he has equals, and sometimes even experts think that Caruana might be better than him in this
respect.
The ability to calculate well, efficiently and without errors is the crucial difference between ‘gifted
players’ and the best players in the world. The best players do not make mistakes in calculating – at
least not as many as their opponents. And traditionally, the best player in the world is the most
accurate in his calculations – since Fischer, no exception is known. Isn’t chess about seeing further
than the opponent?
However, there are different ways of doing calculations.
The ‘logical’ player (see Dvoretsky’s famous distinction) tries to calculate as far as possible, while
maintaining accuracy. Kasparov is the best example of this, and his ability to calculate very far and
accurately is still proverbial. Nevertheless, it appears that the Kramnik of 2000, at least in the London
World Championship Match, was able to calculate farther than Kasparov, while Anand was clearly
able to calculate farther than Kramnik in their 2008 meeting.
A ‘natural’ player like Carlsen doesn’t go that far if he doesn’t need to. He prefers to calculate a
few moves deep and then evaluate the position, rather than calculate a long series of forced moves.
Nevertheless, we have evidence of Carlsen calculating very far, even in blitz games, and it’s hard to
find an event where he was out-calculated by an opponent. The young Caruana once said that he was
not afraid of Carlsen in long, complicated games, suggesting that he felt stronger than the Norwegian
in this respect. There is no evidence of this in their games: notably in their 2018 World Championship
Match, it was Carlsen who seemed to look for complications and played rather better in this area
(especially with the black pieces), while Fabiano showed a slight dominance in positional and even
technical play. Carlsen sometimes has problems with calculations in irrational positions or situations
(Carlsen-Svidler, Candidates 2013). He also has difficulty calculating attacks; he only sacrifices when
he is able to calculate the complications to the end, and sometimes refuses to follow his intuition.
Nevertheless, the most important thing in calculation is to know if the forced line you are going for is
– or might be – losing in the end (like Fischer’s famous 29...Bxh2); in this respect, Carlsen’s errors
are impressively infrequent and his calculation skills are remarkable.
As always, the diagram numbers refer to the order within the games as they are given in Chapter 13,
e.g. ‘13.2’ refers to the second diagram in Game 13.

13.2 (go to the game)

160
 White to move

15.3 (go to the game)

White to move

16.2 (go to the game)

161
 White to move

19.3 (go to the game)

White to move

22.3 (go to the game)

162
 White to move

34.7 (go to the game)

White to move

38.2 (go to the game)

163
 White to move

39.1 (go to the game)

White to move

52.4 (go to the game)

164
 White to move

52.6 (go to the game)

White to move

66.2 (go to the game)

165
 White to move

82.1 (go to the game)

White to move

Calculate 21.Qe4.

97.2 (go to the game)

166
 White to move

Calculate 24...Ne4.

100.2 (go to the game)

White to move

105.2 (go to the game)

167
 White to move

136.1 (go to the game)

Black to move

137.1 (go to the game)

168
 White to move

143.2 (go to the game)

White to move

151.1 (go to the game)

169
 White to move

162.1 (go to the game)

White to move

165.3 (go to the game)

170
 White to move

170.5 (go to the game)

Black to move

198.2 (go to the game)

171
 White to move

204.3 (go to the game)

White to move

205.5 (go to the game)

172
 Black to move

213.5 (go to the game)

White to move

214.3 (go to the game)

173
 White to move

Calculate 28.Qg6.

218.1 (go to the game)

Black to move

223.1 (go to the game)

174
 Black to move

229.2 (go to the game)

White to move

230.1 (go to the game)

175
 White to move

230.2 (go to the game)

White to move

231.2 (go to the game)

176
 White to move

231.5 (go to the game)

White to move

244.1 (go to the game)

177
 White to move

247.1 (go to the game)

White to move

178
 Chapter 8
Planning: when knowledge brings vision

In this regard, Carlsen is an old-fashioned player who thinks and plays for long-term plans a la
Capablanca. Of course, the 2020s players he plays against won’t let long-range plans come to fruition
if they think they’re good. Long plan, bad plan is the new belief of the modern game. Remember that
most of Capa’s or Alekhine’s opponents were amateur players. Yet Carlsen believes in the old way of
playing, the classical, logical game – and he does well in positions that allow him to make a long
plan, such as the Carlsbad structure – which could be renamed the ‘Carlsen structure’.
The ability to make plans is clearly visible in the endgame. This aspect of the game is often a matter
of culture, and in chess Carlsen’s culture is enormous. A famous example is his game with Radjabov,
which can be compared with a similar plan used by the great Cuban. One can also see the affiliation
with the Cuban genius (as well as with Karpov, whose model was Capa), with the constant desire to
use both sides of the board, and to play according to the principle of two weaknesses. Overall, the
Norwegian is remarkable in his ability to immediately see the small plans. He has the gift of seizing –
or sensing – any mini-plan opportunity, as soon as a new position appears on the board. That’s why
he’s at his best in rapid games, as he needs almost no time to adapt to a new situation.
Again, the diagram numbers refer to the order within the games as they are given in Chapter 13, e.g.
‘97.1’ refers to the first diagram in Game 97.

97.1 (go to the game)

Black to move

105.1 (go to the game)

179
 White to move

106.1 (go to the game)

Black to move

114.1 (go to the game)

180
 White to move

116.2 (go to the game)

Black to move

116.3 (go to the game)

181
 Black to move

130.1 (go to the game)

White to move

147.2 (go to the game)

182
 White to move

158.1 (go to the game)

Black to move

165.2 (go to the game)

183
 White to move

173.1 (go to the game)

White to move

176.1 (go to the game)

184
 White to move

180.1 (go to the game)

Black to move

189.1 (go to the game)

185
 White to move

192.1 (go to the game)

White to move

192.3 (go to the game)

186
 White to move

196.1 (go to the game)

White to move

207.1 (go to the game)

187
 Black to move

220.1 (go to the game)

White to move

188
 Chapter 9
Pawns: perfect technique and new tips

Pawn technique is not an area that is thoroughly studied in chess. There are some experts, notably
Botvinnik, Petrosian and Karpov, who have developed concepts of how to play with pawns in this or
that structure (for example, the Spassky-Petrosian World Championship Match 1966, the Torre
Attack). Carlsen is the pawn expert par excellence, and he knows all the tricks (see the rapid games
with black against Karjakin and Topalov) and even teaches us new ones, like his famous rook pawn
pushes, some of them with sacrifice (see his game as Black against Aronian, Tarrasch Defence). The
difference between most of Karpov’s games – which often aim at maintaining a superior structure –
and Carlsen’s pawn play is that the Norwegian’s style is more active and he often plays aggressive
pawn moves, with the intention of dismantling the opponent’s position. Carlsen’s favourite technique
is a pawn sacrifice played to free a square or a line (for example, ...e5-e4 in the famous rapid game in
the playoff for the world title against Karjakin). More recently, attacking moves like the f2-f4
sacrifice against Giri, giving up a pawn with check and ruining the structure to gain access to the g-
file, became an important part of his arsenal.
In the following diagrams, you are asked to find a strong pawn move. The diagram numbers refer to
the order within the games as they are given in Chapter 13, e.g. ‘3.4’ refers to the fourth diagram in
Game 3.

3.4 (go to the game)

Black to move

3.5 (go to the game)

189
 Black to move

9.1 (go to the game)

Black to move

10.1 (go to the game)

190
 White to move

12.3 (go to the game)

White to move

14.1 (go to the game)

191
 White to move

16.1 (go to the game)

White to move

17.2 (go to the game)

192
 White to move

34.4 (go to the game)

White to move

40.1 (go to the game)

193
 Black to move

42.1 (go to the game)

White to move

42.2 (go to the game)

194
 White to move

48.1 (go to the game)

White to move

50.2 (go to the game)

195
 Black to move

50.3 (go to the game)

Black to move

52.1 (go to the game)

196
 White to move

52.2 (go to the game)

White to move

64.3 (go to the game)

197
 White to move

69.1 (go to the game)

Black to move

71.1 (go to the game)

198
 White to move

100.1 (go to the game)

White to move

101.1 (go to the game)

199
 White to move

112.1 (go to the game)

Black to move

123.2 (go to the game)

200
 White to move

127.1 (go to the game)

White to move

127.2 (go to the game)

201
 White to move

129.2 (go to the game)

White to move

133.1 (go to the game)

202
 Black to move

161.2 (go to the game)

White to move

164.1 (go to the game)

203
 White to move

169.2 (go to the game)

White to move

170.3 (go to the game)

204
 Black to move

170.4 (go to the game)

Black to move

174.1 (go to the game)

205
 White to move

177.1 (go to the game)

White to move

178.1 (go to the game)

206
 White to move

187.1 (go to the game)

Black to move

205.3 (go to the game)

207
 Black to move

220.2 (go to the game)

White to move

220.3 (go to the game)

208
 White to move

237.1 (go to the game)

White to move

245.6 (go to the game)

209
Black to move

210
 Chapter 10
Pieces: the art of going backwards

Piece-playing, the art of placing the pieces on the best possible squares, is a skill that every GM
possesses to a certain degree. But some of them – players who seek direct aggression, like Kasparov –
don’t have the time to always place their pieces in the best possible positions, because replacing a
piece takes precious time – time they would rather use to mount an offensive to speed up the pace of
play. This is why players who are known for getting the most out of their pieces usually play
positional, technical and slow chess. Capablanca, Smyslov, Fischer, Karpov and Carlsen are the
players who come to mind first. Among them, Carlsen remains exceptional for the value he gives to
an active piece position.
In the game against Firouzja in Wijk aan Zee 2021, we can see how Carlsen gives two pawns in a
tense position where the opponent advances a passed pawn, to gain piece activity. This double pawn
sacrifice is one of his specialties; see the games against Howell and Harikrishna.
But Carlsen is best known for his famous backward moves, the most famous of which are the queen
move against Topalov (Paris, rapid) and the glorious knight manoeuvre against Anand.
The diagram numbers refer to the order within the games as they are given in Chapter 13, e.g. ‘2.2’
refers to the second diagram in Game 2.

2.2 (go to the game)

White to move

3.1 (go to the game)

211
 Black to move

5.1 (go to the game)

Black to move

6.1 (go to the game)

212
 White to move

12.2 (go to the game)

White to move

13.1 (go to the game)

213
 White to move

13.3 (go to the game)

White to move

30.1 (go to the game)

214
 Black to move

33.1 (go to the game)

Black to move

34.5 (go to the game)

215
 White to move

34.6 (go to the game)

White to move

43.1 (go to the game)

216
 White to move

43.2 (go to the game)

White to move

46.1 (go to the game)

217
 Black to move

55.2 (go to the game)

White to move

70.2 (go to the game)

218
 White to move

74.1 (go to the game)

White to move

76.1 (go to the game)

219
 White to move

89.1 (go to the game)

White to move

91.1 (go to the game)

220
 White to move

96.1 (go to the game)

White to move

96.2 (go to the game)

221
 White to move

99.1 (go to the game)

Black to move

99.2 (go to the game)

222
 Black to move

104.1 (go to the game)

White to move

106.2 (go to the game)

223
 Black to move

119.3 (go to the game)

White to move

123.1 (go to the game)

224
 White to move

124.1 (go to the game)

White to move

129.1 (go to the game)

225
 White to move

129.3 (go to the game)

White to move

132.2 (go to the game)

226
 White to move

139.1 (go to the game)

White to move

142.2 (go to the game)

227
 Black to move

155.1 (go to the game)

White to move

159.2 (go to the game)

228
 White to move

159.4 (go to the game)

White to move

159.5 (go to the game)

229
 White to move

160.1 (go to the game)

White to move

165.1 (go to the game)

230
 White to move

173.5 (go to the game)

White to move

173.7 (go to the game)

231
 White to move

176.2 (go to the game)

White to move

182.1 (go to the game)

232
 White to move

187.2 (go to the game)

Black to move

187.3 (go to the game)

233
 Black to move

191.3 (go to the game)

White to move

199.1 (go to the game)

234
 White to move

204.1 (go to the game)

White to move

209.2 (go to the game)

235
 White to move

209.5 (go to the game)

White to move

213.1 (go to the game)

236
 White to move

214.2 (go to the game)

White to move

220.4 (go to the game)

237
 White to move

231.4 (go to the game)

White to move

232.1 (go to the game)

238
 White to move

232.2 (go to the game)

White to move

244.6 (go to the game)

239
 White to move

245.5 (go to the game)

Black to move

248.1 (go to the game)

240
 White to move

248.2 (go to the game)

White to move

248.3 (go to the game)

241
White to move

242
 Chapter 11
Endings: breaking the principles

GM Miguel Illescas wrote in 2013 (about Carlsen playing equal positions): ‘Rules have exceptions
and this kid has specialized in detecting them’.
It is remarkable that Carlsen, arguably the best endgame player of all time, frequently goes against
the generally accepted principles of the endgame, though he knows many of them. There are a few
typical points that might be emphasized.
The propensity to deliberately choose an opposite-coloured bishops ending, when other endings are
possible. This is for example the case in the famous 2011 game against Nakamura (Game 38). Most
players would have dismissed this ending because of its known drawish tendencies, but Carlsen often
went for it and played it successfully in most cases.
The choice of the queen and bishop versus queen and knight endgame is another. Classical
theoreticians (in chess this means old Russian coaches!) have considered that the association of the
queen with the knight is most often an advantage over the queen + bishop pair. Dorfman, when he
sees such a fight, thinks that the player who has queen + knight has a ‘material advantage’. Kramnik,
commenting live on Chess 24, recently acknowledged the ‘rule’. Carlsen apparently disagrees, and
willingly plays with the queen + bishop duo for a win, with much success, aside from a win with
black against McShane in 2012, in his encounters with Jobava (Game 58), Leko (Game 59) and Ni
Hua (Introduction), games in which he provokes exchanges that lead to such endings. Incidentally,
Carlsen clearly stated (in a recent Banter Blitz session) that in his view the bishop is stronger than the
knight, so that if you can take a bishop without any immediate negative consequences, you should do
so – in line with Tarrasch, who felt that the difference between bishop and knight was comparable to
(though less important than) the difference between a minor piece and a rook (the ‘small exchange’).
The choice of playing with a queen against two rooks is also a frequent guest in his games –
perhaps it is a kind of stylistic bias. The modern (= computer) evaluation is that the queen is worth
nine pawns, while two rooks are worth ten pawns. In endgames with symmetrical structures, the
advantage of two rooks is often greater, because they are able to attack a single pawn twice.
Nevertheless, there are cases where the queen is superior (chances of attacking the king, dangerous
passed pawn, poor rook coordination), and Carlsen aims to ‘specialize in such positions’, although it
should be noted that on the whole his record is not as good as usual with this balance of material (see
Game 241).
As always, the diagram numbers refer to the order within the games as they are given in Chapter 13,
e.g. ‘1.1’ refers to the first diagram in Game 1.

1.1 (go to the game)

243
 White to move

15.2 (go to the game)

White to move

18.1 (go to the game)

244
 White to move

19.1 (go to the game)

White to move

22.1 (go to the game)

245
 White to move

23.1 (go to the game)

White to move

24.3 (go to the game)

246
 Black to move

26.2 (go to the game)

White to move

54.1 (go to the game)

247
 Black to move

54.2 (go to the game)

White to move

58.1 (go to the game)

248
 White to move

87.1 (go to the game)

Black to move

131.1 (go to the game)

249
 Black to move

145.1 (go to the game)

White to move

145.2 (go to the game)

250
 White to move

146.1 (go to the game)

White to move

154.4 (go to the game)

251
 White to move

175.2 (go to the game)

White to move

182.4 (go to the game)

252
 White to move

183.1 (go to the game)

Black to move

184.1 (go to the game)

253
 White to move

224.2 (go to the game)

Black to move

226.1 (go to the game)

254
 Black to move

230.3 (go to the game)

White to move

255
 Chapter 12
How to win against Magnus Carlsen: the hidden defects?

‘He who has lived by the sword will perish by the sword.’
Carlsen likes to take risks to win. However, he is sometimes ‘rewarded’ with a loss. For example, two
intentional grinds against Vachier-Lagrave (Sinquefield Cup 2017, which the Frenchman won) and
the young Caruana in 2012 went wrong because, tired from a long fight, the Norwegian blundered
and lost – both times from a winning position and after an excellent game. In that same basket is the
important loss with white against Karjakin in the 2016 World Championship Match, which gave the
Russian a one-point lead late in the contest. Carlsen makes great efforts, but in doing so he also gives
chances to his opponents, especially when the strength of the opponents is below the highest level.
Carlsen has lost several times (often badly) against players who are not usually playing at the level of
the top ten, because his will to win leads him to play unusual or dubious openings – to lead the
opponent into unknown positions – unknown even to himself, sometimes! Against the best players, he
has fewer accidents.
But even in this case, we should note some points of interest: always fighting for victory, Carlsen is
mainly in danger as White when the opponent is able to obtain an even, stable and rather symmetrical
position. Here, in trying to win, Carlsen can take too many risks. Two World Championship Match
games, against Karjakin in 2016 and Caruana in 2018 (a famous game where a supercomputer was
able to find a win for Caruana, although in a totally inhuman way), are perfect examples of his
impatience, especially when he is looking for a first win in a match. The other point of weakness,
aside from the occasional tactical error, is his handling of inferior endgames. An unrivalled master in
grinding slightly better endings, Carlsen is clearly less comfortable when he is facing a grinder
himself. On several occasions, we’ve seen Carlsen unable to defend a drawn rook ending – sometimes
one that is not even very tough, like the 2013 game against Caruana (see diagram 146.1 in ‘Endings’).
The important game against Ivanchuk (Candidates Tournament – also in 2013) sums up most of the
sources of Carlsen’s losses: he is White against a much lower-ranked opponent who was in last place
in the tournament, while Carlsen was easily dominating it at the time. Black equalizes, then White
tries to push, takes some risks, lands in a pawn-down endgame – one that he should draw with best
play – and loses it without finding the best moves. While Carlsen is generally a great defender, and
especially in important games where he is at a disadvantage (often with black), he may occasionally
defend weakly against players who are outside the top ten – in part because he can’t adapt quickly
enough to playing for a draw against an ‘inferior’ opponent.
Another point to examine is his style in defence: we have seen that Carlsen almost always chooses
an active defence instead of a passive one, because he hates being passive. This choice gives him
excellent results, but it sometimes backfires.
Recently, it seems that a new generation of players, playing more aggressively, have had good
results against Carlsen (Dubov, Firouzja and Esipenko). Carlsen never had a problem with style when
meeting any category of players, and he has proved that he is able to fight successfully against all
styles of players, and all players except Ian Nepomniachtchi – his next opponent in a World

256
Championship Match, at least until mid-2021. But he has been outclassed at times by this aggressive
young generation, so we need to see if he can once again adapt to the new situation – as for example
against Firouzja, his possible ‘successor’ according to many followers. Firouzja, born in Iran, who
was in the top 25 at the age of 16, manages to surpass him sometimes in purely tactical play, at least
in rapid games. We will see in the next few years how Carlsen adapts to this new situation, but if
Firouzja progresses, it will be a fantastic rivalry because of the opposition of styles (tactical +
dynamic vs positional + ending mastery), like in the Karpov-Kasparov years – both players have
similar styles and the age difference is the same as with the two Ks.
The diagram numbers refer to the order within the games as they are given in Chapter 13, e.g. ‘29.1’
refers to the first diagram in Game 29.

29.1 (go to the game)

White to move

56.1 (go to the game)

257
 White to move

59.1 (go to the game)

White to move

68.1 (go to the game)

258
 White to move

77.1 (go to the game)

White to move

140.1 (go to the game)

259
 White to move

141.1 (go to the game)

White to move

143.1 (go to the game)

260
 White to move

153.1 (go to the game)

Black to move

261
 Chapter 13
Games and solutions

Game 1
Magnus Carlsen 2179
Lothar Olzem
Bad Wiessee 2000 (4)

White to move

35.g4!

This is the best move and a typical pawn break.

Also winning is 35.Kf4 Ke7 36.Ke5 f6+ (36...h5 37.h4) 37.Kd4 Kd7 38.Kc3, intending b4-b5, Kc3-
b4xa4 etc.

35...Ke7

35...fxg4+ 36.Kxg4 Ke7 37.Kh5+–.

36.gxf5 exf5 ½-½

After 36...exf5, the position is easily winning, and the basic idea is to collect the a4-pawn after b4-b5
and Kb4. White has to take care of Black’s invasion on the kingside. To give just an illustrative line:
37.Kf4 Kf6 38.h4 Ke6 39.h5 Kf6 40.f3 Ke6 41.Ke3 Kf6 42.f4. Then White goes pawn-grabbing
and wins. But White took a draw here! This game shows that at nine, Carlsen already had a great

262
sense of pawn play. It is also given as a curiosity, because we witness two events that will not be seen
again:
- Carlsen takes a draw in a favourable position – this he never did as a professional, even with a world
title at stake!
- he undervalues an ending that is winning for him.

Game 2
Magnus Carlsen
Jon Ludvig Hammer
Norway tt 2001

White to move

28.Nc1! Qf6 29.Nd3! Qxf3 30.gxf3 g5 31.Rbc1 Nb8 32.Ne5 Kg7 33.Rc5 Rec7 34.e4

Technically, since Black is completely passive, it might be better to improve the king position first.
An important technical rule that might apply: if you have a pawn move to make, which the opponent
cannot prevent, you should place all your pieces on their ideal squares before playing it. Tibor
Karolyi explains that Karpov would have played 34.Kg2, and he is certainly right. Probably a more
mature Carlsen would have played the king move (or maybe 34.g4, which is also good), but we can
dissect a difference in style between these two great players: while Karpov is cautious and likes to
prepare, or sometimes over-prepares positional changes in the slow style of a Petrosian, Carlsen is a
straightforward player who likes clear play and direct decisions, in the style of Fischer. Both styles
can earn you the title of World Champion, so it is difficult to say which one is better. It may be that in
a winning position like this, Karpov’s style is more effective.

34...dxe4 35.fxe4 f6

263
Black could have activated his pieces with 35...Re7, attacking the e4-pawn if the e5-knight takes on
c6: 36.f3 Rd8.

analysis diagram

Here, we can see that the white king is too far removed from his pawns, and the e4/d4 duo looks
fragile. Now, Karolyi only analyses 37.Rd1± but the best move is a very Carlsenian one: 37.Ng4!.
Thanks to another knight redirection, White is still winning: 37...Rxd4 38.Ne3.

analysis diagram

264
The triple attack on f5 will gain valuable material: 38...Rxb4 39.Nf5+ Kf6 40.Nxe7 Kxe7 41.Rd1!
Ra4 42.Re5+! Kf6 43.Re8, and the poor defensive knight is trapped. Carlsen’s decision was correct,
but he would have to find a number of only moves to keep full domination of the position, so
practically and technically speaking, the slow preparation may be better.

36.Ng4! h5 37.Ne3 Kf7 38.Nf5 Ke6

39.d5+

The finish is perfect.

39...Kd7 40.d6 Rb7 41.e5 fxe5 42.Rxe5 Re8 43.Re7+ Rxe7 44.dxe7 Rc7 45.Rd1+ 1-0

Game 3
Bjorn Erik Glenne 2235
Magnus Carlsen 2163
Oslo 2002 (1)

265
 Black to move

16...Nd6!

This is an excellent square for the knight in this pawn structure. Magnus rightly wants to prevent it
from being exchanged.
16...c6 was also good, intending, if 17.Nc3, 17...Nd6.

17.Nc3 Nc4

17...c6 18.Rac1 Kd7!? is a computer suggestion.

18.Bc1

266
18...Bb4!

Carlsen looks for an ending with a strong knight against a not so good bishop. He often tries to get
this asymmetrical duel (knight against bishop or bishop against knight) to get more winning chances.

19.Rd3 Bxc3

19...c6.

20.Rxc3 Rde8 21.b3 Nd6 22.Kf1

267
22...h5 23.a4 Nf5 24.Bb2?

24.Ra2!, and if 24...Nxd4, 25.Rd2, was equal.

24...g4!

This is an excellent positional decision. The pawn move is important and is the only one that secures
a large advantage for the second player. If you want to open the h-file, you have to fix the enemy
pawn. This still allows you to keep the enemy rook away from the f3-square, and to secure the
position of the knight on f5.
On 24...h4?, 25.g4! keeps the h-file closed.

25.Rac1 c6 26.f3

268
26...h4! 27.e4

If 27.fxg4, the simplest is 27...Ne3+ 28.Kg1 (on 28.Ke1,

analysis diagram

28...h3!, followed by the capture of the g4- and h2-pawns with the knight) 28...hxg3 29.hxg3 Nxg4,
and Black doubles on the h-file with a winning attack.

27...hxg3 28.hxg3 Nxd4 29.fxg4 Rxe4 30.Rd3 Rh1+ 31.Kg2 Rxc1 32.Bxc1 Rxg4 33.Be3 c5

269
34.Bxd4 Rxd4 35.Rxd4 cxd4 36.Kf2 Kd7 37.Ke2 Kd6 38.b4 Ke5 39.Kd3 f5 0-1

Game 4
Espen Lie 2260
Magnus Carlsen 2385
Fredrikstad ch-NOR 2003 (6)

Black to move

14...Nd4! 15.Qd1 Nxe2+ 16.Qxe2 Bxe4 17.Bxe4 Bxb2 18.Rxb2

The rules on how to fix pawns in minor piece endings are well studied in chess literature, as far as
bishops are involved. In the struggle between two bishops of the same colour, there is only one rule,
with very few exceptions: place your pawns on the opposite colour of your bishop, and try to force
the opponent to play his pawns on the colour of his own bishop. With bishops of opposite colours, the
player playing for the win must place his pawns on a different colour than his bishop, while the side
playing for the draw must place the pawns on the same colour as his bishop. When a knight is
fighting a bishop, there are two methods. One, the more aggressive one, is to place his protected
pawns on the bishop’s colour, so that he can block him; the other, the more solid one, is to place his
pawns on the opposite colour, so that he cannot take anything. As for Carlsen, he has a clear
preference for the first one, as he likes the opponent’s bishop to be blocked by his pawn chain.

18...b5!?

The alternative method is shown after 18...d5 19.Bf3 d4=. Both methods have their advantages and
disadvantages. Here, both have equivalent merits (19...dxc4 is probably better after 20.Qxc4 Rbc8, as
Black has a safer pawn structure).

270
19.cxb5 d5 20.Bc2

On 20.Bxg6!? hxg6 21.bxc5, Black must play 21...d4! to hope for an advantage.

20...Qxb5 21.Qxb5 Rxb5 0-1

Game 5
Vincenzo Adinolfi 2110
Magnus Carlsen 2385
Salzburg 2003 (1)

Black to move

57...Rd3!?

This is an excellent try in this drawish position.

58.Ra5+?


Doing nothing, but preparing a counterattack was enough to gain a draw: 58.Ra8 Kg4.

271
 analysis diagram

White has to be accurate: 59.Rg8 g5 60.Rg6 Kf5 61.Rh6=. Another method is to attack the f pawn:
58.Rb6. Now if 58...Rf3 59.Rb8 (or 59.Rb4) 59...Kg4, then 60.Rg8!.

58...Kg4 59.Ra6 Rf3!

Thanks to this surprising manoeuvre, and White’s lack of precision, Black wins the endgame.

60.Re6 Kh3 61.Rxe4 Kg2 62.g4 Rxf2+ 63.Ke1 Kg3 64.gxh5 gxh5 65.Re3+ Rf3 66.Re8 Kxh4
67.Rg8 Rf5 68.Rg6 Kh3 69.Ke2 h4 70.Ke1 Kh2 71.Ke2 h3 72.Ke1 Kh1 73.Ke2 h2 74.Ke1 Rg5
75.Rxf6 Kg2 76.Rf2+ Kh3 0-1

Game 6
Magnus Carlsen 2567
Severin Papa 2397
Lausanne 2004 (2)

272
 White to move

30.Rf5!

A la Capablanca! His domination of the fifth rank allows White to act on the queenside, while
preventing the break with the f-pawn (see Game 7).

30...Ke7 31.Kf3 Rf8 32.Kg4 Ra8 33.a5 c5 34.axb6 cxb4 35.cxb4 Ra4 36.b5 Ra2 37.h4 Rd2 38.Rf3
Kf7 39.Kh5 gxh4 40.gxh4 Rb2 41.Rf5 Rh2 42.d4 Re2 43.Rf4 Rb2 44.e5 dxe5 45.dxe5 Rxb5
46.Rxf6+ Kg7 47.Rg6+ Kf8 48.Kxh6 Rxe5 49.Rf6+ Ke7 50.Kg6 Re4 51.h5 Rg4+ 52.Kf5 Rh4
53.Rh6 Kd7 54.Kg5 Rh1 55.Rg6 Rg1+ 56.Kf6 Rh1 57.h6

Black resigned.

Game 7
José Raúl Capablanca
Dawid Janowski
New York 1913 (3)

273
 Black to move

Thanks to his last move, 26.Rf5!, White was able to play for breaks on b5 and g5, and to provoke
weaknesses.
Capablanca won the game on the 54th move.

Game 8
Magnus Carlsen 2528
Roy Harald Fyllingen 2395
Sandnes ch-NOR 2005 (7)

274
 White to move

33.Bxe6+

This is a typical Carlsen decision. He likes opposite-coloured bishops endings, and deliberately goes
for them even with a winning advantage as in the current position. White has two bishops with pawns
on both wings and a better pawn structure. He can probably win in several ways, but prefers the most
concrete one: exchange a bishop against a knight, win a pawn and play the opposite-coloured bishops
endgame, a choice few GMs would have made, but one he excels at and knows well. Does the
endgame win by force? Probably, but I’m not 100% sure.

33...Kxe6 34.Bxg7 h5 35.Ke3 Kf5 36.Bd4 Bb5 37.Bc5 Bc4 38.a3 Bb5 39.b3 a5 40.Kf2 a4 41.bxa4
Bxa4 42.h3 Bd1

42...d4 43.g4+ Ke5 44.Kg3 d3 45.Be3 Kf6 46.Kh4 Kg6 47.gxh5+ Kf6 48.Bg5+ Kf5 49.h6 Kg6
50.Bd2 Bd7 51.Kg3 Be6 52.h4+–.

43.g4+ hxg4 44.hxg4+ Kg6 45.Kg3 b5 46.f4 Bc2 47.Be3 d4 48.Bf2 Be4 49.Bxd4 Bc2 50.Bb6 Be4
51.Bd8 Bc2 52.Kf2 Be4 53.Ke3 Bc2 54.Bh4 Bb1 55.Kd4 Bc2 56.Ke5 Bd3 57.f5+ Kf7 58.g5 Bc2
59.Be1 Bd3 60.Bc3 Bc2 61.Kf4 Bb1 62.g6+ Kg8 63.Kg5 Bc2 64.f6 Bb3 65.Bb4 Ba2 66.Kf4

Black resigned.

Game 9
Stellan Brynell 2496
Magnus Carlsen 2570
Gausdal 2005 (3)

275
 Black to move

40...h4! 41.Qe1

On 41.gxh4, 41...f4 is overwhelming, for example, after 42.Qxg4 Qxf2+ 43.Kh3 Qe3+ 44.Kg2 f3+
Black is winning.

41...h3+

This advanced pawn gives Black a winning position.

42.Kg1 Qb2 43.Bc6 Qxa2 44.Bb7 Qb2 45.Bc6 Bd4 46.Be8 e3 47.Kf1 Kf6 48.Qe2 Qc1+ 49.Qe1
Qxe1+ 50.Kxe1 exf2+ 51.Kf1 f4 52.gxf4 Kf5 53.Bf7 Kxf4

White resigned.

Game 10
Magnus Carlsen 2625
Giovanni Vescovi 2633
Wijk aan Zee 2006 (7)

White to move

37.f5!

Not the only move, but it is the most direct and energetic way to victory.

37...gxf5

276
37...Bd7 38.fxg6 fxg6 39.Kf4 Bc8 (39...Kc8 40.g5 h5 41.Bd3 Be8 42.Nxe6; 39...Nc6 40.Nxc6 Bxc6
41.g5 h5 42.Bd3 Be8 43.Ke3 Bf7 44.Kd4 Be8 45.Kc5) 40.g5 h5 41.Bd3.

38.gxf5 Bd7 39.Bh5 1-0

Game 11
Mohamed Tissir 2444
Magnus Carlsen 2625
Reykjavik 2006 (2)

Black to move

56...Ne1!

This is a classic interception.

57.Rxe1 Rg1 58.Nxe5+ Kg7 59.Re2 h1=Q 0-1

Game 12
Magnus Carlsen 2698
Zbynek Hracek 2614
Germany Bundesliga 2006/07 (14)

277
 White to move

20.Bxd7!

This is an excellent exchange, intending to spoil Black’s pawn structure and play against the bishop
with the knight.

20...Qxd7 21.dxc5 bxc5 22.Nb3 Qa4 23.Rxd8 Rxd8 24.Nxc5 Qxa2 25.Qb5 Kh7 26.f3 Rd2 27.Na4
Rd4 28.Nc3 Qc4 29.Qxc4 Rxc4 30.Ra1 Rc7 31.Nb5 Rb7 32.Nd6 Rd7

278
33.Ra6!

A perfect square for the rook; now the bishop is dominated, so Black decides to give up a pawn:

33...Bb7 34.Rxa7 Rxd6 35.Rxb7 Rd1+ 36.Kf2 Rd2+ 37.Kf1 Kg6

38.g4!

The threat was to play 38...h5 (or 38...f5), crippling White’s pawns.

38...Kf6 39.b4 Rb2 40.b5 Rb1+ 41.Kf2 Rb2+ 42.Kg3 e5 43.b6 g6 44.Rb8 Kg7 45.Kh2 Kf6 46.Kg1
Ke6 47.b7 Kf6 48.g3 g5 49.Kf1 Rb1+ 50.Ke2 Rb2+ 51.Kd3 Rb3+ 52.Kc4 Rb1 53.Kc5 Rc1+
54.Kd6 Rd1+ 55.Kc6 Rc1+ 56.Kd7 Rb1 57.Ke8 Kg7

Black must keep the king from f8.

If 57...Rb2 58.Kf8 Rb3, White wins thanks to 59.Rc8! Rxb7 60.Rc6#.

58.Ke7 Rb2 59.Kd6 Kf6

279
60.f4! exf4 61.gxf4 gxf4

62.Rg8!

Threatening checkmate in one.

62...Rb6+ 63.Kc7 Rxb7+ 64.Kxb7 f3 65.Kc6 Ke5 66.Re8+ Kf4 67.Kd5 f6 68.Rf8

Black resigned.

280
 Game 13
Magnus Carlsen 2693
Levon Aronian 2759
Elista ct sf 2007 (1.4)

White to move

22.h5!

This advance allows the knight to reach the f5-square, while the pawn intends to go forward with
back-rank checkmate threats.

22...Nc4 23.Nh4 Nd6 24.h6!?

Played in the same style.

24.Rac1±.

24...Rc3?!

24...Kg8 25.Rac1 Rfd8 26.Kf3 Kf8 gave better chances of survival.

25.Rac1 Rfc8 26.Rxc3 Rxc3

281
27.Nf5!

This is a well-calculated trade, as White is now winning by force.

27...Nxf5 28.exf5 Kg8

29.Re4! Kf8 30.Rg4 Rc7 31.Rg7 b5 32.Rxh7 Kg8 33.Rg7+ Kh8 34.d6 Rd7 35.Kf3 b4 36.Ke4 Rxd6
37.Rxf7 Ra6 38.g4 Kg8 39.h7+ Kh8 40.g5 fxg5 41.f6

Black resigned.

282
Aronian later said: ‘When I was watching Carlsen’s games, I never imagined that he would become a
high-level player... I thought he would be unable to play against strong players...’ That’s a strange
assessment. In 2007, at 16 years old, Carlsen was already able to rout a top-five player in an
important game, but the most interesting thing is that it was Aronian himself! This Candidates Match
was narrowly won by the Armenian after a tie in both slow and rapid time controls, thanks to a final
blitz win in a theoretically drawn ending.

Game 14
Magnus Carlsen 2693
Levon Aronian 2759
Elista ct sf 2007 (1)

White to move

36.f4!

This is a great move, that Stockfish finds only after a while.

36...Rxc3 37.h5 gxh5 38.Rf8! Ra3 39.f5! Bxf5 40.Rxf7+ Kg8 41.Rg7+ Kf8 42.Rb7 Ra8 43.Kg3 Rd8
44.Kf4 Be4 45.g3 c3 46.Rf7+ Kg8 47.Rg7+ Kf8 48.Nd7+ Rxd7 49.Rxd7 1-0

Carlsen: ‘This was obviously a very important win for me, and actually achieved quite smoothly. It
has to be said, however, that this wasn’t Levon’s best day.’

Game 15
Magnus Carlsen 2714
Zoltan Almasi 2691
Kemer 2007 (4)

283
 White to move

33.Bxg6

This is a very characteristic and stylistic decision: Carlsen is always ready for an ending with
opposite-coloured bishops.
33.e5!, a standard line opener, was thematic and good: 33...fxe5 (33...Nxe5 34.Rxd8 Rxd8
35.Qxh7+–) 34.Bg4 Rxd3 35.Rxd3 Nf8 36.Qb2+–.

33...hxg6 34.e5 fxe5 35.Rxd8 Rxd8 36.Rxd8 Qxd8 37.Qxg6 Qf8+ 38.Kg3 Qf6 39.Qh5 Kd6 40.Qe8
Be4 41.Qb8+ Kd7 42.Qxe5 Qxe5+ 43.Bxe5 g6

This position is difficult to judge in advance. It seems to me that the older Magnus would prefer
33.e5; his style has broadened, even if he still likes favourable opposite-coloured bishops endings!

44.Bf6

Now it should be a draw.

44.Kh3 a4 45.g4 was a better chance.

44...a4! 45.Be5 Bd3?

45...Kd8! 46.Kh3 Bf5+ 47.g4 Be6 holds for Black.

46.Kf4 Bxc4 47.Kg5 Ke6 48.Bc7 Kf7

48...Bxb5 49.Kxg6+–.

284
49.Bxb6 Bxb5 50.Bxc5 Bf1 51.g4 Bd3 52.Kf4 Ke6 53.e4 Be2 54.Bd4 Bd1 55.Bh8 Be2 56.Kg5
Kf7 57.h5 gxh5 58.gxh5 Bf3 59.e5 Ke6 60.Bf6 Be4 61.Kf4 Bc2 62.h6 Kd5 63.Ke3 Bg6 64.Kd2
Ke6 65.Kc3 Kd7 66.Kb4 Bc2

67.Kb5

Zugzwang.

67...Ke6 68.Kc6 Bg6 69.Kc7 Bf5 70.Kd8! Kf7 71.Bg7 Bg6 72.Kd7 Bf5+ 73.Kd6 Bg4 74.Bf6 Bf5
75.Bg5 Bg4 76.Bf4 Bf5

285
77.e6+!

You must calculate a long line to play this move, because once you win the bishop against the e- and
h-pawns, Black can still hope for a draw. White’s last pawn must be promoted to the ‘wrong corner’,
so that if the black king is able to reach a8, there is no win.

77...Bxe6 78.h7 Kg7 79.Kxe6 Kxh7 80.Kd5 Kg6 81.Kc6 Kf5 82.Kb5! 1-0

82...Kxf4 (82...Ke6 83.Kxa4 Kd7 84.Kb5 Kc8 85.Kc6+– as the king cannot enter the saving corner.
Of course, this had to be calculated, or known, before the 77th move) 83.Kxa4 Ke5 84.Kb5 Kd6
85.Kb6+–.

Game 16
Magnus Carlsen 2714
Sergei Tiviakov 2643
Heraklion 2007 (6)

286
 White to move

33.g4!

A technical pawn move played to take space: this is the usual and correct procedure in such positions,
and Carlsen never fails to achieve it on either side of the board. It is also one of the ideas behind his
famous rook pawn moves, and his favourite attempt to win equal or slightly better positions.

33...h6 34.f3 Kh7 35.Kf2 Ree7 36.Ra3 Bb5 37.Rac3 Rd8 38.Kg3 f6 39.f4 h5 40.g5 Kg6 41.Rb1
Rd5 42.Rc8 Be8 43.Rbc1 f5 44.Rb8 Rd6 45.Rcc8 Rde6 46.Be5 Kf7 47.Rd8 Kg6 48.Kf2 Kf7

287
49.e4! Kg6

The key variation is 49...fxe4 50.Bd6 Rd7 51.Rxd7+ Bxd7 52.Rf8+ Kg6 53.f5+, winning the rook and
the game. After 49...g6 50.Bd6 Rd7 51.e5, Black is strangled.

50.Bd6 Rd7 51.Rxd7 Bxd7 52.e5

53.Rc7 is threatened. The loss of the a- or b-pawn cannot be averted, because Black’s pieces are in
poor positions.

52...Bc6 53.Rc8 Be8 54.Ra8 Bf7 55.Rxa7 Re8 56.a3 Bd5 57.Ke3 b5 58.Kd4 Bg2

59.e6! Rxe6 60.Be5 Kh7 61.Rxg7+ Kh8 62.Re7+

Black resigned.

Game 17
Magnus Carlsen 2733
Pavel Eljanov 2692
Wijk aan Zee 2008 (2)

288
 White to move

29.Rc6!?

This is a typical Carlsen move. He does not like to play moves that allow the opponent an easy
(forced) reply. So he likes to make them ‘work’ to find the best moves, from the opening up to and
including the endgame. Stepping up the pressure also makes sense around the 30th move, when the
opponent is often running out of time, and would be happy to have a few ‘free moves’ to reach the
time control scheduled on the 40th move.
This type of practical ‘down to earth’ thinking is inherent in Magnus’ way of playing. Remember
what the great Emanuel Lasker said: ‘Chess is a fight’. Another issue is that Carlsen loves to
dominate the sixth rank – see the game against Mamedyarov.
A forced variation would have won a pawn with a great advantage: 29.Nc6 Nb4 30.Nxb4 Rxb4
31.Rd7 (inserting 31.Rc6!? now is interesting) 31...Re7 32.Rc8+ Kf7 33.Rxe7+ Kxe7 34.Rc7+. After
34...Ke6, White has the choice between the two rook pawns, both with a very favourable, but not
clearly won, ending.

29...Kg7 30.Nb5 R4e7 31.Rdd6 Nc5 32.Nc7 Rf8

289
33.h4!

Keeping it complicated! This space-gaining move is excellent.

33.Rxc5 bxc5 34.Ne6+ is also playable, but more easy to play for the opponent.

33...Rff7 34.Nd5 Rd7 35.Rxd7 Nxd7 36.Kg3 Nc5 37.f3! h6 38.Nf4 g5 39.Nh5+ Kg6 40.f4 gxf4+

40...Ne4+ 41.Kf3 Re7 42.Re6!+–.

41.exf4 Kh7 42.f5 Kg8 43.Kf3 Nd7 44.Ke4 Kf8 45.Rc8+ Ke7 46.Kd5 b5 47.Rh8 Nb6+ 48.Kc6
Nc4 49.Ra8 Ne5+ 50.Kc5 Nd7+ 51.Kxb5 Kd6 52.Rxa7 Rf8 53.Kb4 Nc5 54.Kc4 1-0

Game 18
Magnus Carlsen 2775
Yannick Pelletier 2569
Biel 2008 (1)

290
 White to move

55.Bd3!

The idea is to attack f5.

55...Kf7 56.h5 gxh5 57.Bxf5 Kf6 58.Be4 Kg7 59.Bf3 Kh6 60.Kb5 Kg6 61.Bd1 Kh6 62.Be2 Kg6
63.Bf3 Kh6 64.Bc6 1-0

After 64...Kg6 65.Be8+ Kh6 66.Bf7, another pawn is going to fall.

Game 19
Magnus Carlsen 2786
Teimour Radjabov 2751
Cap d’Agde 2008 (6)

291
 White to move

37.Re2!

This is an important theme in the endgame. We know that the rook is worth five pawns, and a bishop
and knight three, although these are values for the whole game. In many cases, the knight’s value
decreases in the endgame – except in some cases, when Magnus uses them!
The king is a very powerful piece in the endgame and the only one apart from the queen to control all
the squares around him. The estimated value of a king is four pawns. So in many endgames, it is a
major point to be able to activate our king and to keep the opponent’s king passive. This is the
meaning of White’s move, cutting off Black’s king before activating his own. This is not a Magnus
speciality, just good technique. But it’s relevant to see how often, against top players, Magnus’ king
is active all over the board, while his opponent’s king stays in place, trying to defend a few pawns;
this game is no exception.

37...Ne6 38.Bh3 Ng5 39.Bf5 g6 40.h4 Nxf3 41.Kxf3 gxf5 42.Nd4

292
As usual, all of Carlsen’s pieces are perfectly placed. The next step is to bring the king deeper into the
opponent’s position and try to swallow up some enemy queenside pawns.

42...Kg7 43.Nxf5+ Kg6 44.Ne7+ Kg7 45.g4 h6 46.Kf4 h5 47.gxh5 Nxh5+ 48.Ke5 b6 49.Rd2 f6+
50.Kd6 d4 51.Nd5 d3 52.c4 Rc8 53.Rxd3 f5 54.Ke5 f4

55.Nxf4!

White falls for the trick and allows a small combination that wins a knight but loses the game. The
active king is no match for the opponent’s poorly-placed pieces. Magnus loves to sacrifice a knight in

293
the endgame, because he anticipates positions where the proud horse is just worth a pawn, as it has to
be sacrificed against a lone passer.

55...Re8+ 56.Ne6+ Rxe6+ 57.Kxe6 Nf4+ 58.Kd6 Nxd3 59.Kc6 Nc5 60.Kxb6 Nxb3 61.c5 1-0

If 61...Kf7 62.c6 Ke7 63.c7 Kd7 64.h5, and a new queen is going to appear.

Game 20
Magnus Carlsen 2776
Alexander Grischuk 2733
Linares 2009 (12)

White to move

28.Rxf6! gxf6 29.Nd7

Attacking both rooks, as 30.Nxf6+ and 31.Nxe8 is a possible threat.

29...f5 30.c4 a5 31.c5 Bg7 32.Nxb8 Rxb8

294
33.Ba6! Bf6

If 33...bxa6 34.d7 Bf6 35.c6 Rxb6, 36.c7 is simple and good.

34.Bxb7! Rxb7 35.c6

A nice ‘coup de chapeau’ to La Bourdonnais! Magnus knows his classics, and enjoys paying tribute
to his glorious predecessors.

35...Rxb6 36.Rc1!

In tennis, when the young Pete Sampras appeared in a semi-final of the US Open and destroyed the
favourite Andre Agassi, the loser said about the winner: ‘He is cool like a cucumber.’ This could
apply to 18-year-old Magnus, playing the cold, and only, winning move.

36...Bxb2 37.d7 1-0

Game 21
Alexander McDonnell
Louis Charles De Labourdonnais
Match London 1834 (16)

295
 White to move

0-1

Game 22
Magnus Carlsen 2772
Dmitrij Jakovenko 2760
Dortmund 2009 (1)

White to move

296
36.g4! hxg4

Allowing White to get a potential outside passed pawn on the h-file is no fun, but the alternative,
letting the opponent create a weak h-pawn blocked on the ‘bad’ colour, is worse:
If 36...Ke5, 37.gxh5 gxh5 38.f4+ Kf6, preventing Ng7xh5. Here this is a kind of mutual zugzwang,
as both White’s knight and Black’s king must keep the crucial f5-square under control: 39.a4
 A) If 39...Bf1, 40.Kf3! wins (40.Ke4 Bg2+) after 40...Kg6 41.a5 Kf6 42.b5+– Bc4 43.Ke4 Kg6
(43...Ke7 44.Nf5+ Kd7 45.bxa6 Bxa6 46.Ng3) 44.b6;
 B) 39...Bd5 40.Kd3 Kg6 41.a5 Kf6

analysis diagram

42.Nc2!. Thanks to a tactical trick, White manages to replace the knight on e3 to keep control of f5
and allow the white king to invade the queenside via d4 and c5. The game is over because b4-b5 will
create a passed pawn.

37.fxg4 Ke5

Once again, in the ‘money time’, five moves before the time control, Carlsen’s opponent has to take a
crucial decision, requiring deep calculation, with very probably little time left on the clock. It is
important for Magnus to be able to play quickly, so that the pressure he exerts through the strength of
his moves can be increased by the opponent’s time pressure and fatigue after three to four hours of
play. Sixty years ago, people didn’t calculate as much, relying more on positional or tactical intuition.
Nowadays, a strong GM spends most of his time calculating variations, which can be exhausting.
Magnus himself, who calculates a lot, sometimes seems very tired just after just a few rounds in a
strong tournament.
Jakovenko had to calculate the following line: 37...c5! (it is clearly an achievement to get rid of this

297
weakness, the problem being that the defensive king looks very far removed from the h-pawn)
38.bxc5+ Kxc5 39.h5 gxh5 40.gxh5 f6 41.h6 Bg8 (this is clearly forced, but White looks able to win
the bishop) 42.Ne6+ Kd6! 43.Nf8 Ke7 44.h7 Bxh7 45.Nxh7 Kf7 46.a4 Kg7 47.Nxf6 Kxf6 48.Kd4
Ke6 with a draw.

38.Nc6+ Kf6 39.Kf4 Ke6 40.h5 gxh5 41.gxh5 Bd3 42.Ke3 Bf1 43.h6 Kf6 44.Ne5 Bb5 45.Kd4
Ba4

46.h7!

The main problem in a dominant position is how to finish off the opponent.
Every player knows the situation: you enjoy a positional superiority, make the most of your pieces,
force the opponent’s pieces into passive squares and everything is fine! But then you have to finish
the game and to strike. This is the dangerous moment when you risk losing your advantage –
sometimes even worse – in trying to convert a so-called ‘winning position’.
Some players (like Karpov and Petrosian) prefer to wait for the opponent’s mistake and sometimes
strike too late. Others like to strike quickly, even too early and before having made all the preliminary
preparations (put your pieces in the best possible positions and the opponent’s in the worst before
striking). Here we could mention Topalov and Kasparov.
In this regard, Carlsen is close to perfection. He hits as soon as possible – being naturally aggressive
and direct – but is able to wait for the right moment to execute a tactic he has noticed. The only player
I can compare him to, as a perfect striker, is Bobby Fischer.

46...Kg7 47.Nxf7 Kxh7 48.Ng5+ Kg6 49.Ne6 1-0

Both pawns will be attacked and taken by the king and the knight, so the endgame is an easy win.

298
 Game 23
Magnus Carlsen 2813
Levon Aronian 2782
Nice rapid 2010 (2)

White to move

52.Ke2!

Carlsen simply brings back the king to take the h-pawn, after which the execution is smooth.
52.Ke1! also works.

52...h3 53.Kf2 Ra2+ 54.Kg1 Rg2+ 55.Kh1 Ra2 56.f5 Kh7 57.f6 h2 58.Rf8 Rxa7 59.Kxh2 Rb7
60.Kg3 Ra7 61.Kf4 Rb7 62.Ke3 Ra7 63.Kd4 Rc7 64.Re8 Ra7 65.Kd5 Ra5+ 66.Kxd6 Ra6+ 67.Ke7
Ra7+ 68.Kf8 1-0

Game 24
Ruslan Ponomariov 2737
Magnus Carlsen 2813
Nice rapid 2010 (10)

299
 Black to move

The ending is even and symmetri​cal. White may be a bit extended – the e5-pawn will need protection
though of course the position is a draw without it, and a fairly elementary one, even though Kasparov
(not the best endgame player of all time) famously lost the position against Piket in a one-hour game
on the Internet (see Game 37). See, for example, Game 25, where White, a pawn up in a team
competition, offered a draw.

39...Rc4 40.Kg3 e6 41.f4 h6

Now Black intends ...g6-g5, des​troying the defence of the e5-pawn.

42.Kf3 Rc3+ 43.Kf2 g5 44.fxg5 hxg5 45.Ra4 Kg7 46.Rg4 Kh6 47.g3 Kh5 48.Ra4 Kg6 49.Ra5 Rd3

300
Is Black threatening 50...Rd5 ?

50.h4?!

A serious weakening of the position, although it still is an easy draw.

After 50.Rb5 Rd5 51.Rxd5 exd5, follows a complicated pawn ending, with many ‘only’ moves for
both sides: 52.Ke3 f5 (52...Kf5 53.Kd4 g4 54.h4 Ke6 55.h5 f6 56.h6 fxe5+ 57.Kc5 Kf6 58.Kxd5=)
53.exf6 Kxf6 54.Kd4=. The passive 50.Ra2 is enough: 50...Kf5 51.Re2 Rd5 (51...g4 52.hxg4+ Kxg4
53.Re4+ Kf5 54.Rf4+ Kg6 55.Rg4+ Kh7 56.Rh4+ etc.) 52.g4+ Kf4 53.Kg2.

50...gxh4 51.gxh4 Rd7 52.Ke3 Rb7 53.Kf4 Rb4+ 54.Kg3 Kf5 55.Ra7 Rg4+ 56.Kf3 Rg7 57.Ra5
Rg1 58.Rb5 Ra1 59.Rc5 Ra3+ 60.Kf2 Ke4 61.h5 Ra8 62.Kg3 Kf5 63.Kh4 Ra4+ 64.Kg3 Rg4+
65.Kf3 Rf4+ 66.Kg3 Kg5 67.h6 Rg4+ 68.Kf3 Rh4 69.Rc7 Kg6 70.Rc8 Rxh6 71.Kg4 Rh1 72.Rg8+?

72.Kf4!= Rh4+ 73.Ke3 Kf5 74.Rc7.

72...Kh7 73.Ra8

301
73...Rf1!–+ 0-1

Game 25
Michael Adams
Maxime Vachier-Lagrave
Batumi ol 2018 (6)

White to move

Here White simply offered a draw, not wishing to test the Frenchman’s technique: ½-½

302
 Game 26
Vladimir Kramnik 2791
Magnus Carlsen 2802
London 2010 (6)

Black to move

This game came right after a famous interview with Kramnik (with Sport Express), of which here is a
short excerpt: ‘You, Vladimir, are one of those known for regularly “jolting” the Norwegian
phenomenon. It seems as though you’ve become a classic “inconvenient” opponent for Magnus...’ ‘In
terms of results, yes. I’ve got a good score against him, and I’ve also regularly beaten him in rapid
chess. For now I’m a very tough opponent for him. You could even say, using the professional slang,
that for now Magnus is my “client”.’ This game looked no exception.

61...Rc5!

This is a great move, from a defensive and a psychological point of view. Black is crushed, and he
had to give up a piece for a mere pawn just to be able to resist a little longer. Now, White has to finish
the game, but remember that we are at move sixty-one, both players have already played for hours
(probably six already). So if you are in the first player’s head, what do you hope for? Clearly, the
simplest and cleanest way is to exchange pieces, so that the win will be a piece of cake. So Carlsen
suggests to his opponent to play for a ‘little combination’, Capablanca-like. There is also some
pleasure in finishing the game with a small ‘flourish’. By the way, the great Cuban, World Champion
in 1921, was adept at such little tricks, and even explained the procedure in his books (see Game 28).

62.Rxd6+

303
This does not spoil the win, but makes it much more difficult!
62.Rd3! is the safest win: 62...Ra5 63.Kf2 Kc5 64.Ke2, and White simply brings his king toward the
a-pawn: 64...Kc6 65.Kd2. The king will go to b3 and White will take the pawn. If Black tries
65...Nb7, then 66.Nb1 Nc5 67.Rxa3 (67.Rc3+–) 67...Rxa3 68.Nxa3.

62...Kxd6 63.Ne4+ Kc6 64.Nxc5 Kxc5 65.Kf2 Kd4 66.Kf3 Kd3 67.g4 Kd2 68.Be6 Kd3 69.Kg3
Ke3 70.Kh4

Now was the last time to play the move that would have won in a few moves: 70.g5! hxg5 (70...Ke2
71.Kf4 Kf2 72.g3 hxg5+ 73.Kg4 Kg2 74.Ba2 Kh2 75.Bd5 g6 76.Ba2 Kg2 77.Bb3 Kh2 78.Bd5+–)
71.Kg4 Kf2.

analysis diagram

The problem for White is that he can’t protect the g2-pawn with the bishop, because a2 has to be kept
under control. Still, the bishop is much more valuable than a simple passed pawn, because apart from
its power to control many squares, it has the ability to play for zugzwang, which is often the decisive
factor in endings! See Game 27. 72.g3! Kg2 (Black has to keep contact with the g-pawn, of course)
73.Bc4 Kf2 74.Bd5 with zugzwang.

70...Kf2™ 71.Bd5 g6™ 72.Kh3 g5! 73.Kh2 Kf1 74.Be6 Kf2 75.Bc4 Ke3 76.Kg3 Kd4 77.Be6 Ke3
78.Kh2 Kf2 79.Bc4 Ke3 80.Kg1 Kf4 81.Be6 Ke5 82.Bb3 Kf4 83.Be6 Ke5 84.Bb3 Kf4 85.Be6
Ke5 86.Bb3 ½-½

‘There is one game out of one million when I would not win this. But it was exactly this game!’ said
Kramnik on Chess 24 (10.04.2021).
After this, Carlsen and Kramnik played 40 recorded games from 2011-2020 (all time controls). The

304
score in decisive games was 18-2 in Magnus’ favour. ‘Some games are worth more than a point’ –
Larsen.

Game 27
With a knight
(analysis, 2020)

White to move

If the minor piece is a knight, White is not even better, because in this case the knight is not worth
more than the a-pawn!

72.Kxg5

72.g3=.

72...Kxg2 73.Kg6 Kf2 74.Kxg7 Ke2 75.Kf6 Kd2 76.Ke5 Kc2 77.Kd4 Kb2 78.Kd3

78.Nb4=; 78.Nc3=.

78...Kxa2 79.Kc2 Ka1 80.Kc1=

Game 28
Ossip Bernstein
José Raúl Capablanca
Moscow 1914

305
 Black to move

23...Rc5!?

Curiously, the trick begins exactly with the same move. Capablanca explained that he tried to put his
rook at attacking distance of the knight, so that his opponent would be tempted to attack it, and when
for the third time the ‘misplaced’ rook allows a fork, White cannot resist and decides to punish his
stronger opponent.

24.Nb3 Rc6 25.Nd4 Rc7!?

306
26.Nb5 Rc5

27.Nxc3?

Falling into the trap. Even in a winning position, the Cuban, a practical player, liked to tempt his
opponents into little traps, and only if they did not fall for it would he try to reinforce his position in
an ‘academic’ way. Here Black is not winning, just a touch better, and he could have tried to build an
advantage, but he saved time thanks to a nice tactic.
27.Nd4 holds.

27...Nxc3 28.Rxc3 Rxc3 29.Rxc3

307
29...Qb2! 0-1

White was only expecting 29...Qb1+? 30.Qf1. Now if 30...Rd1??, 31.Rc8+. After 29...Qb2, White
resigned because if 30.Qxb2 (30.Qe1 Qxc3), 30...Rd1#.

Game 29
Magnus Carlsen 2814
Erwin l’Ami 2628
Wijk aan Zee 2011 (5)

308
 White to move

72.Rh2

Magnus almost never plays such passive moves.

Among other moves, 72.Rh4+ was winning easily: 72...Kg5 (72...Kg3 73.Nf5+; if 72...Kxh4,
73.Nf5+ wins) 73.Rd4 Re2 (73...Re7 74.Rd5+ Kg4 75.Nf5 Re2 76.Kd3+–) 74.g4 Rg2 75.Ne6+ Kf6
76.Kd5+–.
The game continued.

White to move

Black has just played the king to f3. This is the most natural king move, keeping it as central as
possible to avoid possible checkmates, but also the only losing one:

94.Nh4+! 1-0

94...Ke3 (94...Kf2 95.Rg2+) 95.Rg3+ Kf2 96.Rg2+ Kf1 97.Rxa2.

Game 30
Vugar Gashimov 2746
Magnus Carlsen 2815
Monaco rapid 2011 (2)

309
 Black to move

25...Rg8!?

The position is equal, but Carlsen discovers a nice circuit for his inactive rook with a plan that
worked in the game.

26.g3 Rg6 27.Kg2 Rc6 28.Re2

The prophylactic 28.Nb2! would fight against Black’s idea, e.g. 28...b5 29.a4 a6 30.Nd3 Bd6
31.axb5 axb5 32.Nf2 Kc7∞.

28...c4!

Now Black is better, because the game opens up. Both players will create dangerous passed pawns,
and in the end the bishop will prove much stronger than the knight, thanks to its ability to control the
kingside and queenside simultaneously from the long diagonal (a1-h8).

... 0-1 (49)

Game 31
Magnus Carlsen 2815
Vasily Ivanchuk 2776
Medias 2011 (7)

310
 White to move

13.Qd3!?

How do you judge the style of the great players? Most of the time, they play the best moves, so in a
given situation their choice will be the same. Sometimes, there are no clear ‘best’ moves, or they
decide to play a slightly ‘inferior’ move because they like it. In fact, style becomes visible mainly
when the player plays dubious moves, which is very rare in Carlsen’s case.
Carlsen has a weakness for the early move Qd3, placing the queen in front of the f1-bishop. In
general, he does not hesitate to develop his queen early – another reason why he likes the
Scandinavian Defence. But the Qd3 move is a special one for him. He has used it in some of his most
important games. There is something that he likes about it; maybe d3 is often the most central place
available for the queen at the beginning of the game, affecting the whole board. Notably, Magnus
used this move in two crucial games in his World Championship Match against Anand, and it is a
kind of ‘lucky’ move for him. We can add that he also favours the similar move Rd3. In the famous
Gelfand game from the Candidates, he managed to play both the queen and the rook to this magic
square!
...

311
 White to move

65.Nbxd5!

The usual knight sacrifice!

65...Nexd5 66.Nxd5 Nxd5 67.a7 Nc7 68.Kd4 Kb6 69.Ke5 Kxa7 70.Kxf5 Nd5 71.Kg6 Nxe3
72.Kxh6 1-0

312
 Game 32
Magnus Carlsen 2821
Alexei Shirov 2714
Biel 2011 (2)

White to move

33.Rxd8+ 1-0

This attraction move ends the game: 33...Kxd8 34.Qb8+ Qc8 (34...Ke7 35.Bb4+) 35.Qxa7+–.

Game 33
Yannick Pelletier 2590
Magnus Carlsen 2821
Biel 2011 (1)

313
 Black to move

60...Qd4!

In queen endings, it is important to play the queen in the centre of the board, because the occupation
of one of the four central squares moves the opponent’s queen away from those squares.
Nevertheless, in a recent interview, Maxime Vachier-Lagrave – another great endgame player –
declared that his coach, the late Nikola Spiridonov, a specialist in endgame training, had taught him
that the real best squares are c4-c5/f4-f5. Is it more important to have an eye on the enemy f-pawn?
This is an interesting topic for endgame specialists to discuss, among whom Magnus’ opinion would
be most welcome!
60...Qb4! was even better, e.g. 61.Qe5+ Qb2 62.Qd5 b5 63.Kh2 a4 64.Qd1+ Qb1 65.Qd4+ Ka2
66.Qd5+ Ka3 67.Qc5+ b4 68.f4 Qd3–+.

61.Qf1+ Kb2 62.Qe2+ Kc1

62...Ka3.

63.Qe4 Qc3 64.f4 b5 65.Qe5 Qc4 66.Qa1+ Kc2 67.Qxa5 b4 68.Qa4+ Kb2 69.f5 gxf5 70.Kh2 h5
71.Qd7 Qf4+ 72.Kh1 b3 73.Qd3 Qf2 74.Qd5 h4 75.Kh2 Qf4+ 76.Kh1 Kc2 77.Qc5+ Kb1 78.Qb5
b2 79.Qd3+ Ka2 80.Qa6+ Kb3 81.Qb5+ Ka3 82.Qa5+ Qa4 83.Qc5+ Ka2 84.Qd5+ Ka1 85.Qe5
Qe4 86.Qa5+ Kb1 87.Qc3 f4 88.Qd2 0-1

88...Qe3 89.Qb4 f3 90.gxf3 Kc2 91.Qc4+ Qc3 92.Qe2+ Kc1 93.Qf1+ Kd2 94.Qf2+ Kd1 95.Qb6
Qc1, and wins.

Game 34 French Defence

314
 Magnus Carlsen 2843
Fabiano Caruana 2773
Sao Paulo/Bilbao 2012 (6)

1.e4 e6 2.d3 d5 3.Nd2 Nf6 4.Ngf3 Nc6 5.c3 Bd6 6.Be2

White plays a reverse Philidor, a solid, universal set-up. Carlsen likes to play uncommon openings,
but, as it seems, especially against Caruana, who is known for his opening expertise.

6...0-0 7.0-0 a5 8.Re1 e5 9.exd5 Nxd5 10.Nc4 Re8 11.Bf1 Bg4 12.h3 Bh5 13.g3 Nb6 14.Nxb6
cxb6 15.Bg2 b5 16.a4 b4 17.Be3 Bc7 18.Qb3 h6 19.Qc4 bxc3 20.bxc3 e4

315
21.dxe4

‘I played this and the next few moves quickly, as I couldn’t wait to get to torture him in the endgame.
21.Nd4! would have been even stronger. I would probably have considered it, had I noticed that after
21...Ne5 (21...exd3±) 22.Qb5 Nxd3 the bishop is actually hanging after 23.Qxh5,’ wrote Carlsen in
New In Chess.

21...Bxf3 22.Bxf3 Ne5 23.Qe2 Nxf3+ 24.Qxf3 Qd3

25.Kg2!

316
A vintage Carlsen decision. Now​adays (2021), computers consider it the best move. A surprisingly
high number of Carlsen’s moves that were considered inaccurate by 2010 (or 2015) computers are
now praised as the best possible.
In the variation 25.Rab1 Qxe4 (25...b6!?) 26.Qxe4 Rxe4 27.Rxb7 (27.Bd4!) 27...Be5 28.Bd2 Rxe1+
29.Bxe1 Rc8 30.Rb5 Bxc3 31.Bxc3 Rxc3 32.Rxa5, given by Carlsen in New In Chess 2012/8, White
may press for a win in a 4/3 rook endgame. Carlsen was not interested, considering that his opponent
would have good chances for a draw.
One often hears of today’s players that they are worse in the endgame than their predecessors. I think
this is a misjudgement. Until the beginning of the nineties, when games were still adjourned, players
were able to stop playing for (half) a day, study the endgame intensively, rest and then play out the
game. In the 1990 World Championship Match with Karpov, Kasparov decided to adjourn in a
winning position, and the computer showed a forced win. So gradually, adjournment sessions were
cancelled, and that gave a more accurate idea of the endgame level of the champions after four or five
hours of play.
Compare the position Magnus refused to try against the 20-year-old Caruana because it seemed too
drawish, and look what happened in Leko-Anand (Game 36).

25...Qxe4 26.Bd4 Qxf3+ 27.Kxf3 b6 28.Rab1 Rac8 29.Re4!? g6

30.g4!

This is an excellent pawn move, restricting both the possible moves ...f7-f5 and ...h6-h5, and
preparing the attack by h4-h5.

30...Kf8 31.h4 Rxe4 32.Kxe4 Re8+ 33.Kd3 Re6 34.Be3 Kg7

317
35.Rb5!

A great technical move. Thanks to this ideal position, the advanced rook helps White to prepare h4-h5
while attacking the weakness on b6, hoping to fix the pawns on the right colour and to play for the
second weakness, the pawn on h6.

35...Bd8 36.h5 Rd6+ 37.Kc4 Rc6+ 38.Kd5 Re6 39.Bd4+ Kf8 40.f4 Bc7 41.f5 Rd6+ 42.Ke4 Rc6

43.Rb1!

318
Another brilliant positional move. Once f4-f5 and h4-h5 are achieved, the rook does not have much to
do on the fifth rank. Its retreat is not good news for Black, as h6 is the next target. Jussupow
dedicated a whole series of videos on Chess 24 to Carlsen’s retreating moves, which show his great
flexibility.

43...Ke8 44.hxg6 fxg6 45.Rh1 Kf7 46.Kd5 Rd6+ 47.Kc4 gxf5 48.gxf5 Bd8

49.f6! Bxf6

On 49...Kg6, 50.Rg1+ Kf7 51.Rg7+ would also be fatal: 51...Ke6 52.Rh7.

50.Rxh6 Be7 51.Rxd6 Bxd6

319
The rest is a walk in the park for the Norwegian, who mentioned a 1959 Fischer-Keres game reaching
the same position (Game 35). Once the remaining pawns are taken, the only trick is not to allow
Black to sacrifice the bishop against the c-pawn, which could result in a draw if the black king gets to
a8.

52.Kb5 Ke6 53.Bxb6 Kd7 54.c4 Kc8 55.Bxa5 Kb7 56.Bb4 Bf4 57.c5 Ka7 58.c6 Kb8 59.a5 Ka7
60.a6 Ka8 61.Bc5 Bb8 62.Kc4 Bc7 63.Kd5 Bd8 64.Ke6 Bc7 65.Kd7 Ba5 66.Be7

Followed by Bd8 and c6-c7: 1-0

Game 35
Bobby Fischer
Paul Keres
Zurich 1959 (12)

320
 Black to move

66...Kg7 67.Kg5 Bd3 68.f4 Be4 69.h4 Bd3 70.h5 Be4 71.h6+ Kh8 72.Bf5 Bd5 73.Bg6 Be6 74.Kf6
Bc4 75.Kg5 Be6 76.Bh5 Kh7 77.Bg4 Bc4 78.f5 Bf7 79.Bh5 Bc4 80.Bg6+ Kg8 81.f6 1-0

Game 36
Peter Leko 2736
Viswanathan Anand 2753
Linares 2003 (13)

321
 Black to move

39...Ra1

39...h5 is the technical draw: 40.Ra8 (40.Re5 Ra1 41.a5 f6 42.Rb5 Ra3, and it is difficult to make
progress with white) 40...Ra1 41.a5 Kf6 42.a6 Kf5.

40.g4 Kf6 41.Kg3 Rc1

41...h5 was still possible and good: 42.g5+ Ke6. Black should hold because h4 is weak. If now
43.Ra8, Rg1+ 44.Kf3 Rg4=.

42.Rb5 g5 43.Rf5+

White is winning, having achieved an ideal formation. White’s rook protects both queenside pawns
and the passed a-pawn, allowing the king to go to the queenside.

43...Kg6 44.h5+ Kg7 45.a5 Ra1

45...Rc4!? 46.f3 Rc2! 47.f4! gxf4+ (47...Rc3+ 48.Kf2 gxf4 49.Rxf4 Ra3 50.Rf5 Kf8 51.Rb5 Kg7
52.g5 hxg5 53.Rxg5+ Kf6 54.Rb5 Kg7 55.Ke2+–) 48.Kxf4.

46.Kg2 Re1 47.f3 Re6 48.Kf2 Kf8 49.Rb5 Kg7 50.Rf5 Kf8 51.Rc5 Kg7 52.Rb5 Kf8 53.Rb6 Re5
54.a6 Kg7 55.a7 Ra5 56.Rb7 Ra3 57.Ke2 Kf6 58.Kd2 Ke6 59.Kc2 f6 60.Kb2 Ra4 61.Kb3 Ra1
62.Kb4 Kd6 63.Rh7 Ke5 64.Kb5 Ra2 65.Kb6 Kd5 1-0

Game 37
Jeroen Piket 2633
Garry Kasparov 2851
KasparovChess GP g/60 Internet 2000 (4)

322
 White to move

Another completely drawish ending that was played during a one-hour game on the Internet. Maybe
the 37-year-old Kasparov – the best player ever, according to Magnus – didn’t have a lot of time left.
Nevertheless, Carlsen would probably hold it in a bullet game!

40.f4 g6 41.e5 Rd3 42.Kh3

42...Re3

323
A losing idea where most of the moves would have drawn easily.
42...Rd7! was the simplest. Black can wait (though the aggressive 42...g5!? is possible): 43.Kh4
(43.Ra2 Rb7) 43...Kh6 (43...Kg7 44.Kg5 Ra7 is also a draw); or even 42...Kg7 43.Rc7 Kg8 44.Kh4.
Here, Black has to find the only move 44...Rd5! : 45.Kg5 Kg7 46.Re7 Rc5 47.h3 Rd5 48.g4 hxg4
49.hxg4 Rc5.

43.Kh4

43.Rc7 would have won: 43...Kg7 44.Kh4 Re2 45.Kg5 Re3 46.Re7 Ra3 47.f5! (47.e6 Ra5+ 48.Kh4
Kf6 49.Rxf7+ Kxe6) 47...gxf5 48.e6+–.

43...Kg7

43...Kh6! holds, e.g. 44.Rc7 Re2 45.g4 Kg7 46.h3 hxg4 47.hxg4 Re3 48.Kg5 Re1 49.Re7 Re3 50.f5
Rxe5 51.Rxe5 f6+=.

44.Kg5 Re1

44...Ra3 45.Rc7 Ra6.

45.Rc7 Re2 46.Re7 Ra2 47.f5 gxf5 48.e6 h4 49.Rxf7+ Kg8 50.Kf6

Black resigned.

Game 38
Magnus Carlsen 2861
Sergey Karjakin 2780
Wijk aan Zee 2013 (8)

324
 White to move

35.Qh3!

Exchanging queens allows White to seek more play on the light squares.

35...Qxh3+ 36.Kxh3 h5 37.Rb1 Ra8 38.Kg2 Ra6 39.b3 axb3 40.Rxb3 Bf6 41.Rc4 Rd6 42.Kf1 Kf8
43.a4 Nc3 44.Bf4 Re6 45.e3 Nxa4 46.Bd5 Re7 47.Bd6 b5 48.Bxe7+ Bxe7 49.Rxb5 Nb6 50.e4
Nxc4 51.Rb8+ Kg7 52.Bxc4 Ra7 53.f4 Bd6 54.Re8 Rb7 55.Ra8 Be7 56.Kg2 Rb1 57.e5 Re1 58.Kf2
Rb1 59.Re8 Bf8 60.Rc8 Be7 61.Ra8 Rb2+ 62.Kf3 Rb1 63.Bd5 Re1 64.Kf2 Rd1 65.Re8 Bf8 66.Bc4
Rb1

Manoeuvring has failed to produce any effect, so Carlsen tries the last resort: tactical play!

325
67.g4!? hxg4 68.h5!?

Sacrificing a second pawn.

68...Rh1?

After 68...gxh5! 69.f5, White wins the bishop with f5-f6 and if necessary e5-e6. But Black gains
enough counter​play thanks to 69...h4 70.f6+ Kg6! (70...Kg8 71.e6 fxe6 72.Bxe6+ Kh7 73.Rxf8)
71.Rxf8 Kf5! (the only move, which is not easy to spot and creates check​mating chances) 72.Rh8
(72.Bxf7 g3+ 73.Kf3 Rb2–+; 72.Rg8 also enables White to draw) 72...Rb2+ (72...g3+ 73.Kf3 Rf1+
74.Kg2 Rf2+ 75.Kg1 Kxe5 76.Rxh4) 73.Ke1 Rb1+ 74.Ke2 Rb2+ 75.Kf1 (75.Kd1 g3 is winning for
Black) 75...Rb1+, with a draw.

69.hxg6 fxg6 70.Re6 Kh6 71.Bd5 Rh2+ 72.Kg3 Rh3+ 73.Kxg4 Rxd3 74.f5 Re3 75.Rxg6+ Kh7
76.Bg8+ Kh8 77.Kf4 Rc3 78.f6 d3 79.Ke3 c4 80.Be6 Kh7 81.Bf5 Rc2 82.Rg2+ Kh6 83.Rxc2 dxc2
84.Bxc2 Kg5 85.Kd4 Ba3 86.Kxc4 Bb2 87.Kd5 Kf4 88.f7 Ba3 89.e6 Kg5 90.Kc6 Kf6 91.Kd7
Kg7 92.e7

Black resigned.

Game 39
Magnus Carlsen 2870
Viswanathan Anand 2775
Chennai Wch m 2013 (5)

326
 White to move

In this game, his first as White in the match, Carlsen managed to get ‘his’ type of play after the
opening. In the first three games, Anand dominated, but in some cases seemed to settle for a draw,
even when his position looked favourable. Carlsen was clearly impressed in his first World
Championship Match, and psychologically the match was strange, with the two players – one an
experienced 44-year-old World Champion, the other clearly the dominating player in the world –
looking unusually timid and fearful.
Then in the fourth game, a complicated struggle ensued, with Carlsen taking a ‘poisoned’ pawn in
exchange for a long-term initiative for Anand, with the draw being a fair result after an excellent and
well-played fight.
Here Carlsen played 1.c4. I remember that people were dissatisfied with the beginning of the match,
with Magnus playing 1.Nf3 in his first two white games and not obtaining anything. ‘Play 1.e4, like a
man!’, wrote Aagaard on the Quality Chess blog. In fact, Magnus plays 1.e4 when he wants to play it
‘safe’, like when he needs (or does not object to) a draw. When he plays for a win, he generally
chooses 1.c4 or 1.d4. So after the opening, and a surprise from Anand who played the ‘triangle’ Slav,
Carlsen looked like entering the famous 4.e4 gambit, which was known as the refutation at the time,
but immediately chose a sub-line, and Anand was practically on his own after move 6. His position
was fine, but he had unusual problems to deal with and decisions to take. After an exchange of
queens, the position turned into a complicated middlegame without queens, with Carlsen slightly
pressing and Anand defending actively. Then an equal ending arose, an asymmetric one, with still
many concrete problems to solve, and finally, after the time control, Anand made an important
mistake and entered a losing rook endgame. This is the end of the game:

52.a4!

327
52.Rh5, keeping the two-pawn advantage, was too passive: 52...Rg4 53.a4 Rb4+ 54.Ka3 Rb1 55.h4
Ra1+ 56.Kb3 c4+ 57.Kb4 c3 58.Kxc3 Rxa4 and Black achieves an easy draw, as both White’s pieces
are passive.

52...Kxe5 53.a5! Kd6

54.Rh7!

This is the technical move.

54.a6 also wins; but not the ‘academic’ 54.Ra4?, because Black gets an easy draw after 54...Rxh2
55.a6 Rh8=.

54...Kd5 55.a6 c4+ 56.Kc3 Ra2 57.a7 Kc5 58.h4

Followed by h5-h6, Kg7, h6-h7... and wins: 1-0

Game 40
Viswanathan Anand 2775
Magnus Carlsen 2870
Chennai Wch m 2013 (6)

328
 Black to move

The doubled e-pawns are apparently advantageous:

• they allow White to control many squares in the centre;
• they provide a semi-open f-file that gives play on the kingside. However, Carlsen spots their
possible defects:

25...c5!

This is a great way to play against White’s pawn structure. The problem with doubled pawns is their
immobility. Once they begin to move, they are fragile.
After ...c5-c4, Black threatens to cut the centre in half by taking the d3-pawn, with e3/e4 being
doubled and isolated. If White plays d3-d4 (after ...c5-c4) to keep the structure compact, then e4 is
weakened and could be attacked along the e-file. Such plans come immediately to Carlsen’s mind,
because he knows all of the pawn structures, and what to do, what to hope for in all cases.
...

329
 Black to move

60...h3!

This time the doubled pawns are an asset, forcing the win for Black. As the g-pawn disappears, the
black f-pawn is unstoppable, because White’s queenside pawns don’t allow saving side checks by the
rook (see Game 41).

61.gxh3 Rg6 62.c4 f3 63.Ra3+ Ke2 64.b4 f2 65.Ra2+ Kf3 66.Ra3+ Kf4 67.Ra8 Rg1 0-1

Game 41
Without queenside pawns
analysis

330
 White to move

62.Ra3+ Kd4 63.Ra4+ Ke5 64.Ra5+ Kf6 65.Ra4 Kg5 66.Ra5+ Kh4 67.Rf5=

Game 42
Magnus Carlsen 2881
Sergei Movsesian 2672
Dubai Wch Rapid 2014 (9)

White to move

331
21.e4!

The bishop is sent to a passive position. Once again, Carlsen restricts the enemy bishop by putting
pawns on its colour. This strategy is normal in case of a duel with knight against bishop. Here, with a
light-squared bishop, Carlsen plays in opposition with the principle that with a bishop, you should put
your pawns on the opposite colour. Nevertheless the move is good, because here White’s space
advantage is more important. As a matter of fact, White’s bishop, though theoretically ‘bad’, has
more squares than Black’s.

21...Bg6 22.Bg4 Na7 23.a4

Restricting the knight.

23...Qc7 24.h4 h5 25.Be2 a5 26.Qc4 Qxc4 27.Bxc4 Kf8 28.Bd3 Nc8 29.Nc4 b6 30.f3 Ke7 31.g4 f6
32.Ne3 Kd8 33.Kf2 Be8 34.b3 Ne7 35.Kg3 Kc7 36.Ba6 Kd8 37.g5 Bd7 38.gxf6 gxf6 39.f4 Be8

40.f5!

Having blocked the centre, and with the h-pawn isolated and subject to a future attack by the bishop
and knight, Carlsen decides first to block possible counter​attacks on the queenside (Ne3-d1-c3,
played against a possible ...b6-b5), take the king to a3, open the queenside with b3-b4, and only then
go after the h-pawn. This is a beautiful implementation of the principles of prophylaxis and two
weaknesses, especially considering that this was a rapid game.

40...Ng8 41.Be2 Ne7 42.Kf2 Nc8 43.Nd1 Na7 44.Nc3 Ke7 45.Ke3 Kf8 46.Kd2 Ke7 47.Kc2 Kd8
48.Kb2 Kc7 49.Ka3 Kd8 50.b4 Kc7 51.Nd1 axb4+ 52.Kxb4 Nc8 53.Ne3 Ne7 54.Nf1 Ng8
55.Ng3 Kb7 56.Bxh5 Bd7 57.Bg6 Kc7 58.Nh5 Kd8 59.Bf7 Nh6 60.Be6 Ng4 61.Bxd7 Kxd7

332
62.Kb5 Kc7 63.Ka6 1-0

Game 43
Magnus Carlsen 2863
Viswanathan Anand 2792
Sochi Wch m 2014 (2)

White to move

14.Ra3!

A brilliant manoeuvre that was prepared by the preceding moves, a2-a4, Nc4xb6 and d3-d4 (see
Game 45). Anand does not defend in the best possible way, but up to the 29th move, Carlsen’s play
leaves a fantastic impression of harmony, lightness and a perfect coordination of forces.

14...Nf8 15.dxe5 dxe5

333
16.Nh4! Rd8 17.Qh5 f6 18.Nf5

Here, Carlsen got up and stayed standing behind his chair.

18...Be6 19.Rg3! Ng6

We can see how the undeveloped c1-bishop is on the best possible square, as are all White’s pieces!
Of course, for such a great positional player as Magnus, to know where the pieces should be placed is
pretty basic; yet, to be able to manage such a beautiful attacking position against Anand in a crucial
World Championship game is just amazing.

334
20.h4!

Bringing yet another attacking unit to the kingside.

The only piece that does not attack the king directly is the rook on e1, placed prophylactically on the
e-file, as Black will have to exchange the knight with ...Bxf5, thus opening the e-file after exf5. When
this file opens, the ability to dominate it and to enter the seventh rank will ultimately turn to be the
decisive factor in White’s victory.

20...Bxf5 21.exf5 Nf4 22.Bxf4 exf4 23.Rc3 c5 24.Re6 Rab8 25.Rc4 Qd7

26.Kh2!

Prophylaxis in attack, a la Kasparov! Garry, probably the greatest attacker of all time, showed that
before sending all his forces against the enemy king, a defensive move is necessary, to avoid a timely
check that would help the defence. See the most famous example: Game 44.

26...Rf8 27.Rce4 Rb7 28.Qe2 b5 29.b3

29.Re7! would have crowned Carlsen’s display: 29...Qd6 30.f3 Rxe7 31.Rxe7 bxa4 32.Qe4
(threatening 33.Qb7) 32...Qb8 33.Qxa4 Kh8 34.Qxa5, with a material advantage and domination of
the board.

29...bxa4 30.bxa4 Rb4 31.Re7 Qd6 32.Qf3 Rxe4 33.Qxe4 f3+ 34.g3 h5 35.Qb7

Black resigned.

335
 Game 44
Garry Kasparov 2740
Anatoly Karpov 2705
London/Leningrad Wch m 1986 (16)

White to move

White has an attacked piece on a3, but on the other hand he has a promising attack, although the two
blocked pieces on the queenside cannot participate. He played the only move:

31.Kh2!

A purely prophylactic move, played at the hottest moment of the game!

If 31.Nh6, 31...Qc1+ (31...N7e5 32.Qf6+–); 31.Bxd3, cxd3 32.Qf4 Qxa3 33.Nh6 Qa1+! (another
decisive check!) 34.Kh2 Qf6.

31...Rb3 32.Bxd3 cxd3 33.Qf4 Qxa3 34.Nh6 Qe7 35.Rxg6 Qe5 36.Rg8+ Ke7 37.d6+ Ke6 38.Re8+
Kd5 39.Rxe5+ Nxe5 40.d7 Rb8 41.Nxf7

Black resigned.

Game 45
Sergey Karjakin 2723
Boris Gelfand 2758
Khanty-Mansiysk 2009 (6)

336
 Black to move

The power of an early rook transfer (see Game 43) was shown in this famous game:

11...Ra6!! 0-1

Now the best defence was 12.Bxd5 Qxd5 13.Rxe7 Rg6 14.f3 Bxh3 15.Re2 Qxf3 16.Qf1. Here, Black
has to make a draw with 16...Bxg2 17.Rxg2 Rxg2+ 18.Qxg2 Qd1+ 19.Qf1 (after 19.Kh2? Qxc1,
Black wins because White will lose at least another piece) 19...Qg4+ 20.Qg2 Qd1+.

Game 46
Baadur Jobava 2727
Magnus Carlsen 2862
Wijk aan Zee 2015 (8)

337
 Black to move

This is another favourable ending with queen + bishop against queen + knight. The old wisdom says
that the Q+N couple is stronger and represents an advantage. I don’t see this established by the
statistics. I believe that strong players know a lot of specific situations with those material
imbalances, and they don’t generalize like this, but rather look for specific situations. Often, if you
move a single pawn from one square to the next, the evaluation is completely different. It is one of
Carlsen’s biggest strengths to be able to evaluate such circumstances very far in advance, so that you
feel that he always has the right pieces in the given position! Here, objectively, the situation was
equal before Jobava’s blunder. It is not enough to be a good player; you also have to play good
moves, as they say!

40...Qe4!

‘Central Park Station!’, as we used to announce in such cases during our nightly blitz sessions in the
bars of the Quartier Latin years ago. This is the winning move, since White has just made the fatal
mistake on move 40. ‘As so often happens,’ Magnus could add. The centralized queen effectively
wins the house.

41.f5

41.Nxc4 Qg2+ 42.Ke3 dxc4; 41.Qd2.

41...Qf4+

After 41...g5, White will soon be in zugzwang because the h- and a-pawns are weak, while his king is
in danger along the a8-h1 diagonal. Here is a possible variation: 42.Qd2 Bd3 43.Qe1 a4 44.Qd2 d4

338
45.cxd4 Qxd4 46.Kg1 Kh6 47.Kf2 (47.Qc1 Bb5 48.Kf2 Kh7 49.Kg1 Ba6 50.Qe1 Bb7 51.Qc1
(51.Qe2 Qc5) 51...Qe4–+) 47...Qd6. This double attack wins.

42.Kg1 gxf5 43.gxf5 Bd3 44.Qe1 Bxf5 45.Nxf5 Qxf5 46.Qg3

46...Kh6 47.Qb8 Qf3 48.Qh8+ Kg6 49.Qg8+ Kf6 50.Qd8+ Kf5 0-1

Carlsen made an interesting psychological observation about Jobava, a very imaginative, creative and
fighting player who is not often able to meet (nor beat) the very best: he thought that Jobava played
well, but his problem was that sometimes he was tempted by some daring move, and even if he
spotted the refutation, he was not able to refrain from playing it. Such insights are typical, and
Carlsen is a very deep observer of the other player’s way of thinking.
The young Magnus had a ‘flashy’ period, when he liked to play for brilliancies. The older Carlsen is
still very much attracted by daring and paradoxical moves, but he does not often play them in slow
time controls if they are not the absolute best moves in the position. Another important factor is that
today’s (2021) top players are often great defenders, especially the 1990s generation (Karjakin,
Nepomniachtchi, Vachier-Lagrave and Carlsen himself, among others), and they tend to severely
punish attempts at ‘beauty prizes’.

Game 47
Magnus Carlsen 2863
Peter Leko 2663
Legends of Chess 2020 (3)

339
 White to move

23.f3!

After the following exchange of the rooks, once again Magnus’ queen + bishop duo proves stronger
than Leko’s queen + knight.

23...Nd6 24.Rxe6 fxe6

25.Qa4 Qd7 26.Qb4 Nc8 27.Qb8 Kf7 28.Be5 Kg8 29.b4 Kf7 30.a4 Ke7 31.a5 Kd8 32.a6 Ke8

340
33.Kh2 Kf7 34.Qb7 Ke8 35.g4 Ne7 36.Qb8+ Nc8 37.Kg3 Kd8 38.Qb7 Nd6 39.Qxd7+ Kxd7
40.Bxd6 Kxd6 41.f4 Ke7 42.h4 Kf6 43.hxg5+ hxg5 44.fxg5+ Kxg5 45.Kf3 Kg6 46.Kf4 Kf6
47.g5+ Kg6 48.Ke5 Kxg5 49.Kxe6 Kf4 50.Kd6 Ke4 51.Kxc6 Kxd4 52.b5 1-0

Game 48
Magnus Carlsen 2865
David Baramidze 2594
Baden-Baden 2015 (5)

White to move

35.b4!

This is the best move, an academic one that fixes the pawn structure before attacking the weak b5-
pawn.
The straightforward 35.Rd5 allows 35...b4, with unclear consequences. Black will try to exchange
knights to gain counterplay in the four-rooks endgame: 36.c4 (36.cxb4 Kg7 37.b5 Nf6!) 36...Kg7,
avoiding checks and preparing to take on f6 with the king. Here is a sample variation suggested by
the computer: 37.Rb5 Nf6 38.Nxf6 Kxf6 39.Rxb4 Rd7 40.Re2 Rd3 41.Kg2 Kf5 42.b3 Rff3 43.Rb5
(43.Rb2 Rxh3=) 43...Rxb3, and this should be a draw.

35...Nf6 36.Nd6 Rf3 37.Nxb5 Rxh3 38.c4 Rh4 39.Nd6 Nh5 40.b5 Nf4 41.b6 Rg4+ 42.Kf1 Rh4
43.f3 Rh1+ 44.Kf2 Rh2+ 45.Kg1 Rc2 46.Kh1 Nh3 47.Ne4 Rxc4 48.Rd8+ Kg7 49.Rb1 1-0

341
 Game 49
Leinier Dominguez Perez 2758
Magnus Carlsen 2863
St Louis Rapid & Blitz 2020 (5)

Black to move

36...Kh7!?

‘I went 36...Kh7 and obviously a huge part of this was the trap that he couldn’t go 37.Qe4, which I’d
seen.’
Here you can see how attractive tricks are in rapid time-controls, as the natural 36...Rb3 wins by
force: 37.Qa6 (37.Qe4 Rxb2+ 38.Kxb2 d3+–+) 37...d3 38.Red2 Rxa3 39.bxa3 Qb3+ 40.Kc1 Qc3+
41.Kb1 Qa1#.

Rxb2+ 38.Kxb2 d3+ 39.Kc1 Qc5+ 40.Kd2 Qc2+ 41.Ke1 Bh4+ 0-1

42.Kf1 (42.Qxh4 Qxe2#) 42...Qxd1+ 43.Re1 Qxe1+ 44.Qxe1 Bxe1 45.Kxe1 g5 with a winning
advantage for Black.

Game 50
Fabiano Caruana 2802
Magnus Carlsen 2863
Shamkir 2015 (3)

342
 Black to move

31...Nf3+!

Carlsen plays for a mating net.

32.Bxf3

After 32.Kg2, Black wins a pawn with 32...Nxe5. Still, this might have been a better fighting chance.
White could get some activity after 33.Ra5.

32...exf3 33.h3

343
33...h5! 34.g4

34...fxg4!

And you get a good answer only if you saw the following move for Black:

35.hxg4 h4! 36.Kh2 Rd2

The clinical 36...g5! first was more accurate.

344
37.Kh3

37.g5 had to be tried. Here is what could follow: 37...Rxf2+ 38.Kh3 Kh7 39.Kxh4 Rg2 40.Rf1 Kg6
41.Rxf3 Rxg5 42.e4 Rg2 43.Rf4 Rc2 44.Kg4 Rxc4 45.Kf3 Rc3+! 46.Kf2 Rd3 and Black wins.

37...g5! 38.e4 Rd4 39.Ra8+ Kf7 40.Ra3 Rxc4 41.Rxf3+ Ke7 42.Re3 Rd4 43.f3 c4 44.Ra3 Rd3
45.Ra7+ Kd8 46.Kg2 c3 47.Ra4 c2 48.Rc4 Rd2+ 49.Kh3 Kd7 50.Rc5 Rf2 51.f4 Rf3+ 52.Kh2 Rxf4

And White resigned.

Game 51
Magnus Carlsen 2853
Maxime Vachier-Lagrave 2731
St Louis 2015 (3)

345
 White to move

32.Qe5!

This is a logical decision according to the classical theory about duos: in most positions the queen +
knight duo was considered superior to queen + bishop (this is debatable in my opinion), whereas rook
+ bishop was considered superior to rook + knight (this I feel holds true in most cases). So with this
material, White should try to exchange the queens, while Black should try to exchange the rooks.

32...Qxe5

On 32...Qb6, White can use his central queen for attacking purposes. Here is a possible follow-up:
33.Bd5 Nd6 34.f5 gxf5 (on 34...Nxf5, 35.Qe8+ wins) 35.Bf3 h4 36.gxh4 Kh7 37.h5 Qb4 38.h6 Qd4
(38...Kxh6 39.Qh8+ Kg6 40.Bh5+ Kg5 41.f4+ Kh4 42.Be2#) 39.Qxd4 Rxd4 40.Rc1 Rc4 41.Rd1
Ne8 42.Be2 Rc7 43.Rd5 Kxh6 44.Rxf5. White should win this.

33.fxe5 Re2 34.Rb1

White is winning. The rook + bishop duo is very strong, as is the a6 + Bg2 duo, once a7 has been
taken.

34...Rxe5 35.Rb7 Re1+ 36.Bf1 h4 37.Rxa7 h3 38.Rd7 Ra1 39.g4 Nh4 40.Rd3 Ng2 41.Rxh3 Nf4
42.Rf3 g5 43.Rb3

And Black resigned.

Game 52

346
 Magnus Carlsen 2834
Hikaru Nakamura 2793
London 2015 (7)

White to move

41.f5!

Line and square opening. The h2-b8 diagonal is now available for the bishop, and the f4-square for
the king.
Without any pawn influence in the centre, Black’s knights are unable to occupy a central square to
block the bishops, or to force an exchange. This means that the knights are significantly weaker than
the bishops. In general, there are many cases when the knight appears to be a poor piece in the
ending, and this is one of them. I can add that giving the knight an average value of three pawns is
sometimes confusing: the knight is a great piece in the opening, but its value decreases significantly
in the endgame. On the other hand, the rook has the opposite type of evaluation, i.e. poor in the
opening and great in the endgame.

41...gxf5 42.gxf5 Kd7 43.Kf4 Ne8 44.Kg5 Ke7 45.Bf4 a6

347
This is a very strange pawn move, weakening the position without any necessity.
After 45...Nf6, White should still win in the long run with the same plan as in the game.

46.h4!

White wants to give space to his king. The plan is h4-h5-h6, when after the pawn exchange there are
new roads for a penetration by the king on the dark squares. One of the virtues of having the two
bishops is that they can keep the enemy king at a distance. On the other hand, if White’s king is able
to go to the queenside, for example, the game is over.

46...Kf8 47.Bg3 Nf6 48.Bd6+ Ke8 49.Kf4 Nd7 50.Bg2 Kd8 51.Kg5 Ke8 52.h5 Nf6

348
53.h6!

Creating room for his king to march on.

53...Nh7+ 54.Kh5 Nf6+ 55.Kg5 Nh7+ 56.Kh4 gxh6 57.Kh5 Nf6+ 58.Kxh6 Ng4+ 59.Kg7 Nd4
60.Be4 Nf2 61.Bb1 Ng4 62.Bf4 f6 63.Be4 Nf2 64.Bb1 Ng4 65.Be4 Nf2 66.Bxb7 Nd3

67.Kxf6!!

The ‘normal’ 67.Bd6 is only a draw: 67...Nxf5+ 68.Kxf6 Nxd6 69.Bc6+ Kd8 70.cxd6 Nb2 71.Ke6

349
Nc4 72.a4 a5 73.Kd5 Ne3+ 74.Kc5 Ng4 75.Kb6 Ne5 76.Bb5 Nf7 77.d7 Ne5 78.Kxa5 Nxd7.

67...Nxf4 68.Ke5 Nfe2 69.f6?!

69.c6! wins after 69...Nxc6+ 70.Bxc6+ Kf8 71.a4 a5 72.Bb5 Ng3 73.Bd3 Kf7 74.Bc4+ Kf8 75.f6
Nh5
76.Ke6 (triangulation. There are 11 winning moves with the same idea, but not the immediate 76.f7
because of a typical trick: 76...Nf6! 77.Ba2 (77.Kxf6 is stalemate) 77...Ng4+ 78.Kd6 Nh6 79.Kc5
Nxf7, with a draw) 76...Ng3 77.Ke5 (not 77.Kd6, because of 77...Ne4+ 78.Ke5 Nc5.

analysis diagram

If White’s bishop has to defend it the a-pawn, there is no possible progression any more. As soon as
the white king goes the queenside to attack a5, Black takes a4, and after Bxa4 goes ...Kf7 and
...Kxf6: 79.Bb5 Kf7 80.Kf5 Kf8 81.Ke5 Kf7 82.Bc6 Kf8 83.Kd5 and again: 83...Nd3! 84.Ke6
Nb2=) 77...Nh5 78.Be2 Ng3 79.Bd3 Kf7 80.Bc4+ Kf8 81.f7 Ke7 82.Kd4 Nf5+ 83.Kc5 Ne3
84.Bb3 Nf1 85.Kb6 Nd2 86.Bd5 Nb1 87.Kxa5 Nc3 88.Bc4.

69...a5 70.a4 Kf7 71.Bd5+ Kf8

71...Kg6 72.c6 Nxc6+ 73.Bxc6 Kf7 74.Bg2 Nc3 75.Bc6 Kf8 – this draw has already been explained.

350
72.Ke4!

This is the only winning move, dominating the knights. A lot of Magnus’ games contain a situation
when he fights against one (or here two) knights, and demonstrate how helpless this piece can be in
the endgame, however strong it was in the earlier stages of the game.

72...Nc2

This is the only move that does not lose immediately.

White wins a knight after 72...Ke8 73.Bc4 Kd8 74.Ke3.

73.c6 Nc3+ 74.Ke5 Nxa4 75.Bb3 Nb6 76.Bxc2 a4 77.c7 Kf7 78.Bxa4 1-0

Game 53
Magnus Carlsen 2834
Alexander Grischuk 2747
London 2015 (9)

351
 White to move

30.fxe6!?

This is a bad decision according to the computers, but an excellent one from a sporting point of view.
Carlsen prefers to risk losing, to keep playing for a victory that was crucial for him; both the
tournament and the whole Grand Prix series were at stake.
Sometimes, people commit suicide out of desperation because they absolutely need a win, and in the
end they lose, of course. Grischuk (for example, in the 2018 Candidates against Caruana in the last
round, or against Mamedyarov) or Maxime Vachier-Lagrave (against Jakovenko, in the Palma de
Mallorca GP 2017, last round), or Aronian (many times in Candidates tournaments, for example,
Berlin 2018) are specialists in this type of self-destruction among recent top players. On the other
hand, Carlsen very rarely loses when it comes to such drastic measures (refusing an almost forced
draw in an attempt to win) for several reasons: he takes such decisions quite coldly, and not as a last
resort. When he embarks on such complications, most of the time he has anticipated the situation, so
he makes it look like it is the natural outcome. Many opponents (here we can put aside Caruana and
Aronian, who aim for a win against anyone) are afraid of him and are often happy to take a draw, and
are sometimes disappointed when the ‘boss’ refuses, even if it is the best he can achieve. Thus when
the game continues, they play with a defensive mindset, while they should be thinking about crushing
him!
The third reason is the great insight Carlsen has in the specific psychology, strengths and weaknesses
of his opponents. He always does what the opponent doesn’t like, in the manner of Lasker. As for
Grischuk, Carlsen is used to playing quickly against the great Russian strategist, because he knows
that his opponent manages his time very poorly and often blunders in the last ten moves.
The ‘normal’ move was 30.Nxe5 Qe1+ 31.Kh2 (31.Rf1 Rxg2+ 32.Kxg2 Qg3+=) 31...dxe5; or
30.Nxd6 Bd4+ 31.Ne3 Qe1+ 32.Qf1 Qxd2 33.Ndc4 Qc3 34.Kh2 Be5+ 35.Nxe5 Qxe5+ 36.Kg1

352
Ng3 37.Qd3 Qa1+=.

30...fxe6

30...Rxg4 31.hxg4 Qh2+ 32.Kf2 Nf4 is a forced win for Black. Very likely, Carlsen had seen it (or
felt a refutation existed), but he guessed – or hoped – that Sacha, with hardly any time left, would not
spot it. There is also a psychological aspect that I described (with my friend and co-author Yochanan
Afek) in our book Invisible Moves: if Carlsen, the all-seeing monster, allows you such an easy
combination, it must be wrong! So if you have no time left, you play the safest move, keep the
position and hope for the best. Here there is also a personal factor: while Grischuk is perhaps the
deepest strategical player on the circuit, he is not the strongest tactician – a Nepomniachtchi or a
Vachier-Lagrave would probably have found the winning sacrifice quickly, but then they wouldn’t
have obtained such a promising position with black against Carlsen! 33.Rg3 Nxe6 34.Rf3 Nf4 35.Rg3
Ng6, and ...Nh4 or ...Qf4, with a winning attack.

31.Nxe5 dxe5 32.Qxe6

32...Qe1+

32...Qg5! was the only move to hold, with an equal game.

33.Kh2 Rxg2+ 34.Kxg2 Qxd2+ 35.Kg1 Qe1+ 36.Rf1 Qe3+ 37.Rf2 Qe1+ 38.Kg2

Black resigned.

Game 54
Magnus Carlsen 2834

353
 Maxime Vachier-Lagrave 2773
London 2015 (2)

Black to move

This was a blitz game, and both players were down to seconds here.

51...Rf1

Any other rook move, with the idea of going to the eighth rank and giving vertical checks, would
have drawn easily. This is the classical effect of the ‘checking distance’ (three empty squares between
the pawn and the rook) and the poor placement of the white rook on h3: 51...Ra1 52.Rxh3 Ra8 53.Re3
(53.g5 Rh8+ 54.Kg4 Rxh3 55.Kxh3 Ke6 56.Kg4 Kf7! is a draw) 53...Rh8+ 54.Kg5 Rg8+ 55.Kf4
Rf8+ 56.Kg3 Rg8 (all the moves are forced, but quite standard) 57.Re4 Kd5! 58.Re1 Kd6! (of course,
not 58...Kd4?, because now White is able to use an even stronger device: 59.Re6, cutting off the king
horizontally, wins:
A) 59.Rh1!?, followed by 60.Rh6;
B) 59.Re7 Kd5 60.Kf4 Kd6 61.Rh7 Ke6 (61...Rf8+ 62.Kg5 Rg8+ 63.Kf5 Rf8+ 64.Kg6 Rg8+
65.Rg7) 62.g5;
59...Kd5 60.Rh6 Ke5 61.Kh4™ 61...Kf4 62.Rf6+ Ke5 63.g5 and wins as there is no checking
distance anymore with the pawn on g5).

52.Rxh3 Ke7

If 52...Rf8, 53.Rd3+ Ke7 54.Re3+, and the king cannot cross the f-file, due to the winning rook
exchange.

354
53.Kg6 Rf6+ 54.Kg7 Rf7+ 55.Kg6 Rf6+ 56.Kg5 Ra6

57.Rf3! 1-0

Now Black is one tempo short with 57...Ra8. He would draw with the move thanks to Rg8+, etc., but
here several moves win, like 58.Rf6, 58.Rf4 or the direct 58.Kh6.

Game 55 Queen’s Gambit Declined

Magnus Carlsen 2851
Vladimir Kramnik 2801
Stavanger 2016 (7)

1.d4 d5 2.c4 e6 3.Nc3 Nf6 4.cxd5 exd5 5.Bg5 c6 6.e3 Bf5 7.Qf3 Bg6 8.Bxf6 Qxf6 9.Qxf6 gxf6
10.Nf3 Nd7 11.Nh4 Be7

355
12.Ne2!

This was a novelty from Jon Ludvig Hammer, who worked for Carlsen as his main second during the
World Championship Match in 2013, since Magnus’ usual second Peter Heine Nielsen, who had been
with Anand for years, decided not to take part. The move had been played once before (by
Sommerbauer in 2014), but it was considered an innovation at the highest level, and came as a big
surprise for Kramnik, who is not the easiest guy to outprepare.

12...Nb6

After the ugly 12...f5!, it looks like Black’s position will hold, at least until 2020! 13.g3 Bxh4
14.gxh4 Nf6 15.Nf4 Nh5.

13.Ng3 Bb4+ 14.Kd1 Na4 15.Ngf5 Kd7 16.Rb1 Ke6 17.Bd3 Rhc8 18.Ke2 Bf8 19.g4 c5

356
20.Ng2!

Moving the superfluous knight to a better square. As Giri aptly observed, Carlsen has a special
feeling for the knight. The super​fluous knight is one of Dvoretsky’s many inventions. The ever-
creative Russian trainer identified some positions in which two knights can go to a strong square, but
obviously only one can go there at a time. Dvoretsky’s rule is that the opponent should not trade the
knight if another one can replace it, as the second one is superfluous if the exchange is not made.

20...cxd4 21.exd4 Bd6 22.h4 h5 23.Ng7+ Ke7 24.gxh5 Bxd3+ 25.Kxd3 Kd7 26.Ne3 Nb6 27.Ng4
Rh8 28.Rhe1 Be7 29.Nf5 Bd8 30.h6 Rc8 31.b3 Rc6 32.Nge3 Bc7 33.Rbc1 Rxc1 34.Rxc1 Bf4
35.Rc5 Ke6 36.Ng7+ Kd6 37.Ng4 Nd7 38.Rc2 f5 39.Nxf5+ Ke6 40.Ng7+ Kd6 41.Re2 Kc6
42.Re8 Rxe8 43.Nxe8 Nf8 44.Ne5+ Bxe5 45.dxe5 Kd7 46.Nf6+ Ke6 47.h5 Kxe5 48.Nd7+ Nxd7
49.h7 Nc5+ 50.Ke2 1-0

Game 56
Magnus Carlsen 2851
Pavel Eljanov 2765
Stavanger 2016 (9)

357
 White to move

The most difficult tournament for Carlsen turns out to be Norway Chess, as shown by his results there
over the years. So it looks like Carlsen feels too much pressure when playing at home, or for his
country.

27.Qc6

White has a nice advantage and could have kept it with steady play, e.g. 27.Bxd6 Qxd6 28.Qc6±.

27...Ndf6

Here, Black missed the relatively simple 27...Qa7 28.Bxd6 Qa2 29.Bg2 Qb1+ 30.Bf1 Qa2=.

28.Bxd6 Qxd6 29.Qc8+ Kh7 30.Ne5 Qe7 31.Qc6 Ng4 32.Nxg4 fxg4 33.Bd3 g6 34.Bxe4 dxe4
35.Qxb6 1-0

Game 57
Magnus Carlsen 2855
Vladimir Kramnik 2812
Paris rapid 2016 (9)

358
 White to move

50.Rd7+!

This wins by force. All the pawns are on the same side and the bishop is bad. It would be a draw with
the black pawn on h6 instead of h5.

50...Rxd7 51.Nxd7 Ba6 52.Nc5 Bf1 53.Ne4 Kf7 54.Nd2 Bd3 55.Kf4 Ke6 56.Kg5 Kf7 57.Nb3
Bf1 58.g3 Kg7 59.Nd4 Bc4 60.f4 Bd3 61.f5! gxf5 62.Nxf5+ Kh7 63.Ne3 1-0

The knight will go to f4 or f6 and take h5.

Game 58
Magnus Carlsen 2855
Viswanathan Anand 2770
Leuven rapid 2016 (8)

359
 White to move

31.Rd8!

This is a zugzwang.

31...g6

31...Rc7 32.Ne6 Rce7 33.Nf8+ Kg8 34.Nd7+ Ne8 35.Nc5 Rf8 36.Rxf8+ Kxf8 37.Rb8; 31...Ng4
32.Ne6.

32.hxg6+ Kxg6 33.Rd6! Kg7

33...Rf8 34.g4 Ref7 35.Kg2 Re7 36.Kg3 Rff7 37.Kh4 Rf8 38.Ne6 Rff7 39.Nd8 Rf8 40.Kh3 –
zugzwang.

34.Ne6+ Kg6 35.Nd8 Rf8 36.g4 c5 37.b3 Ree8 38.Nxb7 1-0

Game 59
Magnus Carlsen 2838
Levon Aronian 2774
Karlsruhe/Baden-Baden 2017 (2)

360
 White to move

34.Qxe4

Missing a kind of perpetual. Carlsen can occasionally have a blind spot, in favourable positions, for
drawing chances for his opponent, like counterattacks (see his game against Eljanov) or an
unexpected drawing mechanism (or fortresses, like in the famous World Championship Match game
against Karjakin). Yet, it is difficult to draw any conclusions from a relatively small number of
games, and the Norwegian still has a remarkable ability to transform a favourable game into a win.
It is interesting to note that, in spite of what is said by most of the players and writers about it, his
extraordinary capacity to create chances from equal or nearly equal positions is not matched by an
infallibility in transforming winning positions into victories, as the two World Championship
Matches against Karjakin and Caruana, two great defenders, showed. Karjakin was ‘hopelessly’ lost
in two games at the beginning of the match, and Caruana in one (the first). Yet, they fought
desperately (which not all of Carlsen’s opponents do), trying to make the win as complicated as
possible, and they succeeded! Historically, it reminds us of how Alekhine was stimulated in his pre-
match preparations when he noticed that the great Capablanca was not always as precise as was
generally thought in the technical phase.
34.Nfe5 wins the knight and the game.

34...Nf4+ 35.Kg3 Ne2+ 36.Kg4

36.Qxe2 Qg6+ 37.Kf4 Qf5+ 38.Kg3 Qg6+ 39.Kh2 Qh6+.

36...Qf5+ 37.Qxf5 exf5+ 38.Kxf5 dxc3 39.Rxc5 c2 40.Rc8+ Kg7 41.Rc7+ Kg8 42.Kf6 Rg1
43.Nxg1 c1=Q ½-½

361
The Benko Gambit was a favourite of the young Magnus for a long time. It suited his style perfectly,
as Black gets active piece play in exchange for a pawn, but the state of theory led him to give it up.

Game 60
Loek van Wely 2681
Magnus Carlsen 2733
Wijk aan Zee 2008 (10)

Black to move

Black is already crushed here.

16...Ne8 17.a4 Nc7 18.Re1 Rae8 19.b4 Nd7

362
20.Nb5

Van Wely also enjoys complica​tions. It is easier to dominate tactically from a winning position!
20.b5! Qa5 21.Rb3 e6 22.Ne3 was more positional and still overwhelming.

20...Rc8 21.Bh3

This is good, but White is drifting towards tactical play. A positional player – like Carlsen! – would
rather have tried something like 21.Nxc7 Rxc7, and now 22.b5 is tempting, to be able to coordinate
our pieces with tempo: 22...Qxa4 23.Re3!. This is the key idea. If you see it, you are on the right
track! I admit that this is not an easy solution to come up with, because instead of having a proud
passed pawn duo, we’ve returned a pawn, and Black now enjoys the better structure. But he has a lot
of problems with his pieces; the queen, the light-squared bishop and the queen rook are in precarious
positions. White wants to play Ra3, and on ...Qc4, Ne3 would rearrange his pieces very
harmoniously, with tempi against the queen. Then Ra4 could be a serious threat, or first some
combination of b5-b6, and on ...Rc8, Ra7, e.g. 23...Ra8 (if 23...Qa8 24.b6 Rcc8 25.Ra3 Qb8 26.Ne3;
23...Rcc8 24.Ra3 Qc4 25.Ne3 Qd4 26.Ra7! Rb8 27.Ra4+–; 23...Ne5 24.Ra3 Qc4 25.Ne3 Qd4
26.Qc2 c4 27.Nf5! gxf5 28.Be3) 24.Ra3 Qc4 25.Ba5 Rcc8 26.Ne3 Qd4, and now 27.Bh3+–.

21...f5 22.Bg5

Still the best move.

22...Ne5 23.bxc5 Qxa4 24.Nxc7 Rxc7 25.c6 Bc8 26.exf5

363
White has played well during the complications, and the compu​ter now assesses around +4. Now,
Carlsen starts a desperate defen​sive operation: here, there is no point in looking for the ‘best’ moves.
You just have to keep the position as complex as possible, anticipating the comfortable wins your
opponent is looking for and avoiding them, even if your position gets worse.

26...Rxf5!?

26...Bxf5 27.Bxf5 Rxf5 28.f4 Nf7 29.Ne3; on 26...gxf5, White has many safe moves, like calmly
retreating the bishop with 27.Bg2.

27.f4

Again, best, and again an unneces​sary weakening. But it was already not so easy to find a ‘safe’ win.

27...Nc4 28.Rb4 Qa7+ 29.Kg2 Qc5 30.Rb8

364
30...Nb2!

The idea is to keep the position as complicated as possible – the computer says +6.

31.Qf3!

Excellent.

31...Qxc2+ 32.Re2 Qb1 33.Bxf5

+8.

33...Qxf5 34.g4!

Best by far, but still very weakening.

34...Qf7 35.Bxe7 h5!

Both giving his own king a safe square on h7 and trying to weaken the opponent’s king position.

36.Bxd6 hxg4 37.Qe4 Kh7

+12.

38.Bxc7 Bf5

365
Now Black has four well-placed pieces, including the queen, with chances of an attack against the
white monarch. He couldn’t have hoped for anything more, even if the computer evaluation is around
+24. Of course, Van Wely has spent a lot of time to find all those compli​cated moves, and now it’s
not easy for him with two moves to play.

39.Qe3

The first weak move from the colourful Dutch player.

There were many better choices:
39.Rh8+! is not the computer’s first move – it gives up a rook! – but it wins easily, of course, from
the human point of view. 39...Bxh8 (39...Kxh8 40.Qe8+) 40.Qe7 Qxe7 41.Rxe7+. With the two
passed pawns and the opponent’s poor placed pieces, the ending is very easy. The ‘best’ move was
39.Qe7, though with seconds on the clock it is not pleasant to allow 39...Qxd5+, with other checks to
follow. To be fair, it is unplayable from a human point of view. The computer says White is fine after
40.Kg1 Qd1+ 41.Kf2.

39...Qxd5+ 40.Kg3

White has reached the time control and is winning in all variations but one. Alas for him, Magnus
plays it!
40.Kg1! was the best move. To allow 40...Bd4, you have to see beforehand 41.Rh8+! Kxh8
(41...Bxh8 42.Be5+–) 42.Be5+ Qxe5! 43.fxe5 Bxe3+ 44.Rxe3 Nc4. White is much better here, but
Black has some chances of survival.

40...Nc4 41.Qf2

A better try was 41.Rd8, but Black should win after 41...Qxc6.

366
Now comes a succession of checks.

41...Qd3+ 42.Kg2 Be4+ 43.Rxe4 Qxe4+ 44.Kf1 Qd3+ 45.Qe2 Nd2+ 46.Ke1 Nf3+ 47.Kf1 Nxh2+
48.Ke1 Bc3+ 49.Kf2 g3+ 0-1

This looks like a hold-up from the young Magnus, but it is not. First, the Dutchman did not play in
the spirit of the position. With the enormous advantage he got directly from the opening, he did not
go for a regular conversion, but on the contrary played to crush his opponent – exactly the contrary of
Magnus’ technique, which consists of making things as simple as possible! Van Wely’s moves were
the best from the computer’s point of view, but if you are already winning, it makes no sense at all to
complicate the game! On the other hand, Magnus fought courageously, in the beginning probably
more as a matter of principle than with any real belief in his chances of saving the game. To do so, he
tried to complicate the game as much as possible, so as not to offer a straightforward win to his
opponent. When the blunder came, he took his chance without mercy.

Game 61
Hikaru Nakamura
Magnus Carlsen
Zurich 2014 (3)

Black to move

This looks very bad for Black. Basically, White’s attack is stronger and quicker than his opponent’s.

17...g6

Quite a shocking move. Creating a hook on g6 helps White to develop his standard attack with h3-h4-

367
h5. Probably, Carlsen estimated that this was his best practical chance.
17...a4 18.Nf5 looks no picnic either, and Black will probably have to play 18...g6 anyway. As we
know, a knight on f5 is a terrific attacker.

18.h4 a4 19.Rh2

Excellent play, both attacking – the threat is to double the rooks on the h-file – and defending b2.

19...Qa5 20.Bd2 Qc7 21.g5

This is a demonstration of high-class technical play, in Euwe’s style! Nakamura considers – and
rightly so – that he has time for an academic assault on the h-file. First chasing the knight and fixing
the weakness to be attacked is even better than the standard 21.h5, which was also good obviously.

21...Ne8 22.h5 Rb6 23.Bc1

23...Rb3

A counterattack is his only chance.

24.Qg4 Nb6 25.Be2 Nd6 26.Rdh1

There is already a direct win after 26.hxg6 fxg6 27.Qe6+, followed by doubling the rooks, or 26.Nf5,
with a devastating attack (Stockfish 11 gives +9). But why calculate when simple attacking play will
do? Here, it seems to me that Hikaru made the mistake of trying to win ‘positionally’, exactly what
Van Wely should have done! As this is basically a mating attack, he should have tried at some
moment to calculate a forced win, and this position is a good opportunity to do so. Of course, later,

368
White will have other chances to finish the game, but now was an especially good moment. This is
move 27, so we are still far away from the stress of the time control on move 40. Black has not
developed a counter​attack yet, so it is just attack against defence – the easiest to calculate: 27...Kh8
(27...Qf7 28.Rdh1 Qxe6 29.dxe6 Rxg3 30.Rxh7) 28.Rdh1 Rxg3 29.Rxh7 Qxh7 30.Rxh7 Kxh7
31.Qxd6+–.

26...Bxb2!

You have to try this, even if it does not work. The idea is refreshing: though Black is getting
checkmated, he still is the first to sacrifice! Important to keep in mind is that at this point, Nakamura
had never beaten Magnus in a classical game.

27.Bxb2 Nbxc4 28.Bxc4 Nxc4 29.hxg6 Qb6

369
30.g7

The best move, and the safest.

The forced but slightly unnatural 30.gxf7+ Rxf7 31.Nh5! Rxb2+ 32.Ka1 was winning according to
the computer.

30...Rd8 31.Qh4 Rxb2+ 32.Ka1 Rxh2 33.Rxh2 Qg6 34.Nf5 Re8

Now is the time to finish off the guy.

35.Qg4

370
The threat is 36.Rh6, winning the queen.

35...Qb6 36.Qh3 Qg6

Strictly forced. Now, out of 31 possible moves, 16 are crushing. But alas for him, Nakamura did not
choose one of them:

37.d6

The best move was 37.Qf1!, attacking the knight:

371
 analysis diagram

The knight cannot move without being lost:

 A) 37...Nd6 38.Rh6;
 B) 37...Na5 38.Qb5 Nb3+ 39.Kb2, and now the queen or king is lost;
 C) 37...Qa6 38.Nh6+ (or 38.Rxh7);
 D) 37...b5 and now:
 D1) 38.Qh3 is the sadistic computer win, based on zugzwang. Neither the black king, nor the queen
nor the rook can move!: 38...b4 (38...Na5 39.d6 Nc6 40.Qg4, with Rh6 or d6-d7 to follow) 39.d6
Nxd6 40.Nxd6 Rd8 41.Nf5;
 D2) 38.Rxh7!! (a two-move tactic, but such a nice one!) 38...Qxh7 (or 38...Kxh7 39.Qh1+ Kg8
40.Qh8#) 39.Nh6+ Kxg7 40.Qxf7+ Kh8 41.Qxe8+ Kg7 42.Qf7+ Kh8 43.Qf8+ Qg8 44.Qxg8#.

37...Nxd6 38.Nxd6

38...Rd8! 39.Nc4

39.Nf5 Rd1+ 40.Kb2 Qb6+. By the way, this line is not feasible if Black has played ...b7-b5, and
hence there was the possible win after the 36.Qh3 variation. The best move was 39.Nc8, when after
taking on g7, Black gets three pawns against the knight; the game is unclear and possibly drawn after
best play.

39...Qxe4

372
Now the position is equal for the computer; Hikaru immediately slips, and Magnus finds the best
moves to finish him off clinically.

40.Qh5

40.Ne3∞.

40...Rd3!–+ 41.Rh4 Qf5 42.Qe2 b5 43.Nd2 Qxg5 44.Qxd3 Qxh4 45.Ne4 Kxg7 46.Qf3 Qf4
47.Qg2+ Kf8 48.Kb2 h5 49.Nd2 h4 50.Kc2 b4 51.axb4 cxb4 52.Qa8+ Kg7 53.Qxa4 h3 54.Qb3
h2 55.Qd5 e4 56.Qh5 e3 57.Nf3 e2 58.Kb3 f6 59.Ne1 Qg3+ 60.Ka4 Qg1 61.Qxe2 Qa7+ 0-1

After this game, or the one against Van Wely, Carlsen could have said, à la Bogoljubow: ‘When I am
winning, I win because of the better position. When I am losing, I win because I am Magnus
Carlsen!’

Game 62
Magnus Carlsen 2625
Alexander Beliavsky 2626
Wijk aan Zee 2006 (5)

373
 White to move

17.Bg5

This pin is simply winning the game! As his minor pieces on the queenside are unable to come to the
help of the pinned knight, Black is defenceless.

17...exd4

The main variation is 17...h6

374
 analysis diagram

18.Bh4 g5 19.Nxg5 hxg5 20.Bxg5, with Nh5 to follow.

18.Nh5 dxc3 19.Nh4! Kh8 20.Nf5

Intending calmly to take on g7 before grabbing the pinned black knight.

1-0

375
 Game 63
Magnus Carlsen 2826
Etienne Bacrot 2716
Nanjing 2010 (1)

White to move

11.Qd3!

Not allowing the dark-squared bishop to come back into play. The pin on f6 turns out to be really
unpleasant, as there is no black piece to protect f6. Black does not want to play ...g7-g5, as that would
be too weakening, so he has to sit still while White reinforces his position, brings his rooks into play
and tries to get his knight to attack f6. This Bg5 pin is a favourite tool of Magnus, both a tactical and
a strategic one. To get his queen out of the pin, Black will have to allow White to destroy his pawn
cover, leading to a bad position. In one of the crucial games of his career, Carlsen also elected the
Bg5 manoeuvre (tenth game against Karjakin, World Championship Match 2016, a must-win
situation).

11...Nxb3 12.cxb3 Re8 13.0-0-0 d6 14.Qc2 Bd7 15.Bc4 Be6 16.Rhe1 Qe7 17.e5

Setting up another pin.

17...dxe5 18.Rxe5 Qf8

Black cannot avoid unpinning and allowing his pawn structure to be damaged.

19.Bxf6 gxf6 20.Re2 Qg7 21.Bxe6 Rxe6 22.Rxe6 fxe6 23.Rd3 Kh8 24.Rg3 Qh7 25.Qd2 Bc5

376
26.Ne4 Be7 27.Rh3 Kg7 28.Qd7 Kf7

29.Ng5+!

29.Nxf6! was also good, e.g. 29...Kxf6 30.Rf3+ Ke5 31.Re3+ Kf6 32.Rxe6+.

29...fxg5 30.Rf3+ Kg8 31.Qxe6+ Kh8 32.Rf7 Bd6 33.Rxh7+ Kxh7 34.Qf7+ Kh8 35.g3 Ra6
36.Kb1 Bb4 37.f4 gxf4 38.gxf4 1-0

Game 64
Magnus Carlsen 2646
Michael Adams 2720
Turin ol 2006 (7)

377
 White to move

12.Bh4! e5

If 12...g5, 13.Nxg5! hxg5 14.Bxg5 wins, because after 14...e5 15.d5 Na5...

analysis diagram

16.f4! was decisive.

13.Nxe5 Nxe5 14.dxe5 Qxd1 15.Rfxd1 Nd7 16.f4 Re8

378
17.e6!

This is a Fischer-like move, giving air to his pieces.

17...fxe6

17...Nc5 was better, with approxi​mate equality, e.g. 18.exf7+ Kxf7 19.e5∞.

18.e5 Nxe5

19.fxe5 g5 is the trick. Now Magnus comes up with a brilliant intermediate move, which was not easy
to anticipate:

379
19.Bd8! Nd3 20.Bxc7 Rb7 21.Bxb7 Bxb7

22.a4!

This is an excellent positional decision, giving up a pawn to take the b-file and to force the rook
exchange, thus reaching a favourable ending. It is a known rule that if you have two rooks against
one, you should strive to exchange the opponent’s only rook.

22...bxa4 23.Rdb1 Bc6 24.Rb8 Kf7 25.Rxe8 Kxe8 26.Bd6 a6 27.Rb1 h5 28.Rb6 Kd7 29.Bf8 g6
30.Rxa6 1-0

380
 Game 65 Italian Game
Magnus Carlsen 2840
Sergey Karjakin 2785
Wijk aan Zee 2017 (13)

1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.c3 Nf6 5.d3 0-0 6.Bg5 d6

6...h6 7.Bh4 Be7 is the standard reply, which Karjakin played in the World Championship Match in
2016.

7.Nbd2 h6 8.Bh4 g5

9.Nxg5!

This is a great position for White, provided that he follows up with the best moves.

9...hxg5 10.Bxg5 Kg7 11.Qf3 Be6

381
Now, White has to anticipate Black’s only chance of dealing with the pin: getting the knight back to
b8, to protect the pinned knight via d7.

12.b4 ½-½

12.Bd5! leads to a large advantage. Black still has to play ...Nc6-b8-d7, but this will cost him a pawn
and an exchange, as the a8-rook is trapped.

Game 66
Veselin Topalov 2690
Peter Leko 2701
Dortmund 1999 (2)

382
 White to move

13.Nxg5!

The main tip for the sacrifice to work is to look at the possible supports of the f6-knight (mainly the
queen’s knight and the dark-squared bishop). If they have no chances of protecting the f6-knight, you
can start calculating!
Another detail is the black bishop on d7. It is awfully placed, as it blocks the queen (...Qd6 is not
available after e4-e5 and ...dxe5 has been played). Additionally, there is no hope to come back to the
rescue with the ...Nc6-b8-d7 manoeuvre, which is sometimes feasible in such positions. If this bishop
was on c8, Black would survive.

13...hxg5 14.Bxg5 Kg7

Not only protecting the knight, but also avoiding the e4-e5 and Qg6+ idea, for example: 14...Nxd4
15.e5 dxe5 16.Qg6+; or 14...Bxd4 15.e5 Bxe5 16.Qg6+ Kh8 17.Qh6+ Kg8 18.Ne4.

383
Here,

15.e5!

was winning, liberating the e4-square for the c3-knight.

15.Nb3 was played in the game.

15...dxe5

Now comes the crucial move to win a decisive tempo:

384
16.Rae1!!

Black cannot unpin or protect the knight, and his only useful defensive measure is ...Bf5 – but only
when Ne4 has been played!
Probably, Topalov only saw the normal 16.Ne4 Bf5!. This counter​pin saves the day, e.g. 17.Rae1
Qd6!. Now, White can ‘win’ the queen for three pieces with 18.Bh6+, 19.Qe3+ and 20.Nxd6, but he
isn’t better here – probably even worse.

16...Bxd4 17.Ne4 Bf5 18.Qf3! Bxe4 19.Rxe4

The unavoidable threat is to take on f6 and to check on g4.

19...Kg6 20.Bxf6 Qxf6 21.Rg4++–

Game 67
Magnus Carlsen 2690
Veselin Topalov 2783
Morelia/Linares 2007 (5)

White to move

49.d5!

A brilliant obstruction. The direct 49.h7+ is not good, because after 49...Kh8 50.Rxe6 Rxe6 51.Rxe6,
Black has the winning 51...Qd5!, attacking all the white pieces at the same time! (51...Nxd4=).

49...Qxd5

385
49...cxd5 50.h7+ Kh8 51.Rxe6 Rxe6 52.Rxe6 Ne7 53.Qd3+–. If allowed, White will play Nd4 next,
a beautiful square liberated by the d4-d5 tactic.

50.Rg4 Qd7

51.Rh1!

Threatening 52.Qh7+, followed by 53.hxg7+, etc.

51...Qe8 52.hxg7 Rxg7 53.Qh7+ Kf7 54.Rxg7+ Nxg7

386
55.Rh6

The crucial problem in attack is the eternal choice between two options: should you act quickly (give
checks or create direct threats) or should you first bring more material towards the enemy king?
In the first case, you play on the time factor, as the defence has no time to organize itself. The risk is
that you may lack material when (if) your threats come to an end.
In the second case, you bet on material superiority, but you also leave time for the defence to set up
possible reinforcements.
Here, the second option was better. The arrival of the knight would have been decisive: 55.Nd2! (a
very Carlsenian move!) 55...Qe7 (55...f5 56.Nf3; 55...Qd7 56.Ne4 Qd5 57.f3, and the desirable
57...Qf5 loses to 58.Nd6+) 56.Ne4.

55...Qd8 56.Qg6+ Kg8 57.Qh7+ Kf7 58.Qg6+ Kg8 59.Rh7 Qd7 60.Nd2 f5

As White has slowed down the rhythm of the attack, Black is in time to defend with 60...Rf7,
overprotecting g7 to free the queen and vacating the f8-square for the king, a perfect defensive move.
61.Ne4 (61.Qh6 Kf8 62.Ne4 Qd5) 61...Qd5!= (just in time) 62.f3 (on 62.Kf1 Black replies
62...Qd1+ (not 62...Qf5 because of 63.Rxg7+! Rxg7 64.Nxf6+ and White wins) 63.Kg2 Qd5, with a
repetition: 64.Kh2 Qe5+) 62...Qf5=.

61.Nf3 Rf6 62.Qxf6 Kxh7 63.Ng5+ Kg8 64.Qg6

Topalov resigned here.

Nevertheless, the position was equal, thanks to an ‘invisible’ defence. White’s threat is 65.Qh7+,
66.Qh8+ and 67.Qxg7+. Black can defend thanks to a three-move manoeuvre that is easy to miss
because it contains:
• an intermediate move

387
• a forward move for defensive purposes
• a line-clearing move
• a backward move
• a residual image problem.
Apart from that, there are also psychological reasons. Topalov, who was then one of the three or four
best players in the world, had been pressed and nearly outplayed for 64 moves by a teenager (just
sixteen years old), and he was not surprised to be lost, because he felt that it was the logical result.
64...Qd5+! is the forward move, and an intermediate one. It is difficult to see from the diagram how
the black queen could reach g8, which is currently occupied by the king and all the same seemingly
out of reach (= residual image). 65.Kg1 e5 (another forward move, played to clear the a2-g8
diagonal) 66.Qh7+ Kf8 67.Qh8+ Qg8!. This is the backward move.

Game 68
Viswanathan Anand 2779
Magnus Carlsen 2690
Morelia/Linares 2007 (10)

White to move

27.Ne6!

This tactical motif is a rare one, and possibly it had escaped the attention of the young Magnus.
Shipov estimated in New in Chess in the early 2010s that Carlsen had a problem with certain
unexpected tactical moves. I am not sure if this was ever true (if you compare him with the other
homo sapiens), but Carlsen’s tactical eye is certainly getting even more accurate with time.

388
27...Kh7

After 27...fxe6 28.dxe6 Be8 29.Qxh6 White has a winning attack, as a timely e6-e7 will remove the
possibilities of a black defence by including a major piece along the seventh rank, for example,
29...Nb6 30.e7+ Nc4 31.Qf8+ Kh7 32.f4!, and the rook joins the attack (or even the nice circuit
32.Bd1!?, intending 33.Bg4 and 34.Be6).

28.f4 Qa7+ 29.Kh2 Be8 30.f5 gxf5 31.exf5 f6 32.Re1 Nc7 33.Rc1 Bd7 34.Rc3 e4 35.Rg3 Nxe6
36.dxe6 Be8 37.e7 Bh5 38.Qxd6 1-0

Game 69
Mateusz Bartel 2608
Magnus Carlsen 2714
Germany Bundesliga 2007/08 (9)

Black to move

This is an equal position, but Bartel has just carelessly allowed his opponent to create an
asymmetrical structure that suddenly gives him chances to win:

27...exd4! 28.cxd4 c4 29.f5 c3 30.bxc3 Na4 31.Rde1 Nxc3 32.Rc2 d5 33.exd5 Nxd5 34.Rc5 Rxc5
35.dxc5 Rc8 36.Bf2 b4 37.Bd4 Nc3 38.Re7 Kg8 39.Kf2 Nxa2 40.Ke3

40.Ra7 was better. Here, 40...Nc1 keeps the pawn advantage. The chances of a win are ‘objectively’
around 50%, but here you have to bet on Magnus, statistically!

389
40...Nc1 41.Re6 Nb3 42.Ra6 Re8+ 43.Re6 Kf7 0-1

Game 70
Magnus Carlsen 2765
Vasily Ivanchuk 2740
Foros 2008 (1)

White to move

19.c5!

Giving up a pawn just to open some lines for the bishop. This is an archetypical decision from
Carlsen. ‘The main priority is to activate the bishop’, completely true to his non-materialistic style
and his obsession to keep creating better positions for his pieces.

19...Nxc5 20.Bc4 Bg5 21.Qe2 Qh6 22.Rad1 Rf8 23.a4 b6 24.g3 Qh3 25.Qg2 Qh6 26.Qe2 Qh3
27.Kh1 Nd7

390
28.Ra1!

This is the best move, a brilliant one that enables the rook both to play along the a-file after a4-a5 and
to join the c-file via a2.

28...Qh6 29.Ra2 Nf6 30.Kg2 Nh5 31.Nd5 Nxd5 32.Bxd5 Bf4 33.Qf2 fxe4 34.Bxe4 Qg5 35.Rc2 d5
36.Bxd5 Bxg3 37.hxg3 Nf4+ 38.Kf1 Nxd5 39.Rce2 Qf6 40.Rxe5 Qxf3 41.Qxf3 Rxf3+ 42.Ke2 Rf5
43.Rxf5 gxf5 44.Kd3 c5 45.Re5 Nb4+ 46.Kd2 1-0

Game 71
Magnus Carlsen 2775
Teimour Radjabov 2744
Bilbao 2008 (4)

391
 White to move

27.e5!

This sacrifice frees the e4-square for the knights and cuts the black queen off of the defence of the g5-
pawn. It has both clearance (opening lines/squares) and obstruction (closing lines/squares) effects.

27...dxe5 28.Nxg5 Bf5 29.Nge4 Kh7 30.Qe2 Nh6 31.Rg3 Be6 32.Rg6 Nf5 33.Qg4 Nh4 34.Ng5+
fxg5 35.Qxg5 Nxg6 36.Qxg6+ Kh8 37.Rg1 1-0

Game 72
Magnus Carlsen 2775
Levon Aronian 2737
Bilbao 2008 (6)

392
 White to move

15.d5!

This is a beautiful move, in typical attacking style. White, a pawn down, throws away the pride of his
position, the strong d4-pawn. Then the position is unclear, and Black has a problem with his king (he
will not be able to castle), while White has a serious material handicap.

15...Nxd5

 A) 15...cxd5 loses after 16.Bb5+ Nd7 17.Ne5. The pin is final; if 17...Bc8, 18.Rc1, followed by
19.Rxc8+ and 20.Bxd7+;
 B) After 15...exd5, White is able to use the liberated d4-square with 16.Nd4, and possibly Nf5, Qf3
(or Qa4);
 C) After 15...Qxd5, White might play 16.Ne5, and after 16...Be7, 17.Qa4!, with the intention of
setting up the rooks on the d- and c-files, while if Black castles kingside, the queen can slip into the
king’s flank, a la Petrosian: 17...0-0 18.Rfd1.

16.Ne5 Nf6 17.Qa4 Bb4 18.Nxc6 Bxc6 19.Qxc6+ Ke7 20.Rfd1 Rc8 21.Qf3 Qb6 22.Bd4 Qb8
23.Ba6 Rcd8 24.Bb7 h5 25.h3 h4 26.Rab1 e5 27.Rxb4 axb4 28.Bc5+ Ke6 29.Ra1 Rd6 30.Bxd6
Kxd6 31.Qc6+ Ke7 32.Ra8 Qd6 33.Qxd6+ Kxd6 34.Rxh8 b3 35.Ba6 Nd7 36.Rxh4 Nc5 1-0

Game 73
Tigran Petrosian
Vasily Smyslov
Moscow ch-URS 1961

393
 White to move

18.Qa4!

The shortest route to the kingside. There are many books, videos and articles dedicated to attacking
play. They offer many tricks, tips and advice, so much that you get lost sometimes. I will provide
only one tip to the aspiring player: bring the queen into the attack! If not, there is no attack in most
cases, no matter how many of your pieces are ‘looking’ at the enemy king.

18...Rfd8 19.Qe4 g6 20.Qg4

White has a winning attack, since Black can only prevent sacrifices on g6 by advancing the pawns
that defend the king, after which Petrosian’s attacking play is perfect.

20...h5 21.Qh3 f5 22.Bc4 Rxd1+ 23.Rxd1 Kf7 24.e4 Qf4 25.Re1 Qg4 26.exf5 Qxc4 27.fxg6+ Ke8
28.g7 e5 29.Qxh5+ Kd7 30.Rd1+ Bd6 31.Bxe5 Nd4 32.Nxd4

Black resigned.

Game 74
Magnus Carlsen 2581
Nigel Short 2687
Hoogeveen 2004 (3)

394
 White to move

23.Nd2!

Very delicate positional play! The knight is well-placed on c4, but it is in the way of the light-squared
bishop and blocks the c-file. So this retreat gives air to the rest of White’s army. Harmony is a very
important positional concept, i.e. pieces working without hurting each other, so that each one works
at full strength. One of the keys to understanding Carlsen’s chess is that he hates disharmony, and is
always ready to give tons of material – or tempi – to avoid it. Even in hair-rising complications like
reciprocal attacks with opposite-side castling, he looks for a harmonic position before anything else.
Short was visibly very impressed by the subtle positional feeling the 13-year-old GM was already
displaying, as he went from an almost equal position to a bad one in only a few moves:

23...g5

395
24.Bb5! a6?! 25.Bxc6 bxc6 26.fxg5 fxg5

27.Qf6+! Kg8

27...Qxf6 offered better saving chances.

28.Qxd4 c5

396
29.Qf6

Keeping the queens on with 29.Qc3 or 29.Qe3 promised good attacking chances.

29...Qxf6 30.Rxf6± ... 1-0 (54)

Game 75 Sicilian Defence

Magnus Carlsen 2570
Gata Kamsky 2690
Khanty-Mansiysk 2005 (7)

1.e4 c5 2.Nf3 e6 3.d4 cxd4 4.Nxd4 a6 5.Nc3 Qc7 6.Bd3 b5 7.0-0 Bb7 8.Qe2 Ne7 9.Re1 Nbc6
10.Nxc6 Nxc6

397
11.Nd5! exd5 12.exd5+ Ne7 13.c4 b4

Carlsen strives for direct and forced lines:

14.Bg5

Simple development like 14.Be3 promised more, for example: 14...0-0-0 15.a3 b3 16.Rac1. The
perfect mobilization of White’s forces allows him to look for a quick win. He can already pick up the
b3-pawn (Rc1-c3xb3) and develop a strong attack against the black king, thanks to his domination in
the centre.

398
14...f6 15.Qh5+ g6 16.Qf3 fxg5 17.Qf6 0-0-0 18.Qxh8 Qd6 19.Be4 Kb8 20.Rac1 Qf4 21.g3 Qf7
22.Qd4 d6 23.Qb6 Rd7 24.Qxb4 Nf5

25.Bxf5!?

Search for a clear plan. This exchange has pros and cons, but it shows the practical approach of the
young Carlsen. The exchange helps Black to improve his pawn structure, but there is something more
important in Carlsen’s eyes: the question of what to do next and what is the plan. This issue is solved
after this exchange, and this is very much in Fischer’s style. Carlsen will use the e6-square, double his
rooks and go for an attack on the enemy king.

25...gxf5 26.Re6 f4 27.Rce1 Rd8 28.Qb6 Rc8 29.b4 fxg3 30.hxg3 h5 31.b5 Qc7 32.Qxc7+ Kxc7
33.a4 axb5 34.axb5 Kb6 35.Re8

After 35.Rf6!, Black can hardly move, as his bishops are unable to get some activity, due to his
unfavourable pawn structure. The light-squared bishop is trapped and not worth much more than a
pawn; meanwhile, the dark-squared bishop has to protect d6: 35...g4 36.Re4 Ba8 37.Rfe6 Bb7
38.Rf4+–. Now Re6-g6-g8 is on the agenda, or first moving the king to d3.

35...Bg7 36.R8e6 Rd8

36...Bc3! 37.Rxd6+ Kc5 38.Ree6 Kxc4 is unclear.

399
37.c5+! Kxc5 38.Re7 Bd4 39.Rxb7 Rf8 40.Rc7+ Kxd5 41.Rd1 Rb8 42.Rg7 g4 43.Rg5+ Kc4
44.Rxh5 Bc5 45.Rg5 Rxb5 46.Rxg4+ Kc3 47.Kg2 Ba3 48.Rgd4

Black resigned.

Game 76 Pirc Defence

Magnus Carlsen 2710
Teimour Radjabov 2746
Biel 2007 (9)

400
 White to move

14.Re3!

A beautiful and powerful move that maintains the central tension and solves some problems White
had with the positioning of his pieces. What makes this game so original is that the placement of the
white rook seems audacious, with all the pieces on the board and an explosive situation in the centre.
Carlsen had rightly calculated that Black cannot exploit this factor, and the computer confirms it.

14...dxe4

On 14...exd4, the point is that White wins after 15.Nxd4! Bxf1 16.Nxc6 Qe8 17.exd5.

15.Re1! Qc7 16.Nf5!

The e2-knight is allowed to go to g3, and all of the white army works at full power.

16...Bd8 17.g5 Nd5 18.Rxe4 f6 19.Neg3 g6 20.Nh6+ Kg7 21.dxe5 fxg5

22.e6!

The best move, intending to let the rooks have their say!

22...Kxh6

On 22...N7f6, White’s idea was 23.e7! Bxe7 24.Rxe7+ Nxe7 25.Rxe7+ Qxe7 26.Nhf5+, forking king
and queen.

401
23.e7 Qb6

The main line is 23...Nxe7 24.Rxe7 Bxe7 25.Rxe7 Rad8 26.Qd4!.

analysis diagram

White threatens checkmate on g7 and h4!

24.exf8=Q+ Nxf8 25.c4 Nf4 26.Qd6 Kg7 27.Bxf4 gxf4 28.Re7+

Black resigned.

Game 77
Magnus Carlsen 2776
Levon Aronian 2750
Linares 2009 (8)

402
 White to move

84.Rf1

Carlsen has lost several drawn rook endgames when a pawn down, one against Caruana, two
important games against Aronian – a loss that meant being eliminated – and one against Ivanchuk, a
loss that put his qualification for the World Championship Match in jeopardy.
After 84.Kf2 Rf6 85.Rh1 (also 85.Ra1 Kxh4 86.Rh1+ Kg4 87.Rg1+ Kf4 88.Rg3 Ke4 89.Rg4+ Ke5
90.Rg5+ Ke6 91.Rg3) 85...Rf4 86.Rh3, Black cannot make progress.

84...f2–+ 85.Kh3 Rf3+ 86.Kg2 Kg4 87.h5 Rg3+ 88.Kh2 Kf3 89.Ra1 Rg2+ 90.Kh1 Rg5 91.Ra3+
Kf4 92.Ra4+ Kg3 93.Ra3+ Kh4 0-1

Game 78
Magnus Carlsen 2776
Viswanathan Anand 2791
Linares 2009 (6)

403
 White to move

18.Nc7!?

Carlsen wants to get rid of the ‘bad bishop’, to make his own bishop stronger and getting a favourable
ending with rook(s) + bishop against rook(s) + knight, with the black pawns being weaker without
their main defender. Suba’s immortal joke is to be taken seriously here: ‘Bad bishops protect good
pawns’.

18...Rag8 19.Nxe6 Kxe6 20.Rxg8 Nxg8 21.Ke2 Ne7 22.Kf3 Rc8 23.a4 Rc7 24.a5 h6 25.h4 Kf6
26.h5 Nc8 27.Kf4 Nd6 28.Rg1 Rc8 29.f3 Ke6

29...a6! 30.Rg2 Rc1.

30.Rg7 Rh8 31.Bc2 Rc8 32.Bb3 Rh8 33.Rg1 Rc8 34.Rg7 Rh8 35.Rg2 Rc8

404
36.Rg1 Ne8 37.e4 fxe4 38.fxe4 Nf6 39.e5 Ne4 40.Ke3 b6 41.axb6 axb6 42.Kd3 Nf2+ 43.Ke2 Ne4
44.Ke3 f6 45.Rg6 Rc1 46.Rxh6 Rh1 47.Bc2 Rh3+ 48.Kf4 Rh4+ 49.Kf3 Nd2+ 50.Ke2 Rh2+
51.Kd1 Nc4 52.Rxf6+ Ke7 53.Bg6 Rd2+ 54.Kc1 Rxd4 55.b3 Nxe5 56.Rxb6 Rh4 57.Bf5 Nf3 58.h6
Nd4 59.h7 Nxf5 60.Rb8 Nd4 61.Kb2 Kd6 62.h8=Q 1-0

Game 79
Magnus Carlsen 2484
Jan Werle 2407
Wijk aan Zee 2004 (8)

405
 White to move

17.Bh3!

This is a simple but brilliant exchanging operation (the knight cannot move because of Rc8+). Once
the knight has disappeared, Black is helpless on account of the weaknesses on the light squares.
Before Carlsen’s move, White was clearly better: bishop pair, better pawn structure, better pieces and
an advantage in development.But most of those advantages are rather abstract. After Carlsen’s move,
the game plays itself, and you just invade with the major pieces, without much calculation needed.
Many of Carlsen’s great moves have this quality: they look simple. But often, players of the highest
level are commenting live, and they don’t suggest those ‘simple’ moves because they can’t see them!
(see Game 80 for another example of such clarity) It is often after the game and having seen it, that
commentators and trainers are able to explain how simple the solution was.
This clarity of mind was already Carlsen’s strength when he was aged thirteen, like in the present
game. From Kramnik we heard the following anecdote: ‘I remember when I played with him in the
first Wijk aan Zee tournament, which actually he didn’t finish so well, on -3 or something, when he
was 15. Magnus was in the A-Group, and I played a game with him. It was a draw, and then we just
analysed a little after, post-mortem on the board, just for 20 minutes, and for me everything was clear
after that, because I could feel how delicate his feeling is for very slight positional changes –
something I frankly not always can witness even with top 10 players who are 40 years old. I was
trying this move, that move, and it was amazing to me that a 15-year-old boy, in a very positional
game, would immediately switch his plan or set up his pieces differently in accordance with what I
would do as White, and it was most of the time, in my opinion, totally correct. And then you could
say that to be able to catch it so quickly at 15, that’s a sign of a huge positional talent.’

17...0-0 18.Bxd7 Rxd7 19.Qd5 b5 20.Rc6 Qa8 21.Rfc1 1-0

Game 80
Bobby Fischer 2760
Tigran Petrosian
Buenos Aires ct 1971 (7)

406
 White to move

22.Nxd7+

‘His play is so simple...’ declared one trainer from the Russian team.

22...Rxd7 23.Rc1 Rd6 24.Rc7 Nd7 25.Re2 g6 26.Kf2 h5 27.f4 h4 28.Kf3 f5 29.Ke3 d4+ 30.Kd2
Nb6 31.Ree7 Nd5 32.Rf7+ Ke8 33.Rb7 Nxf4 34.Bc4

Black resigned.

Game 81
Magnus Carlsen 2552
J Deepan Chakkravarthy 2331
Dubai 2004 (2)

407
 White to move

9.d4!

An excellent attacking move. As the black king is on the open e-file, the most important factor here is
time. So White classically gives up a pawn for a tempo. I remember that Carlsen himself offered
some tips about such positions. He said that if you are one tempo from castling, it is usually okay. If
you’re three tempi from castling, it is probably lost. And if you’re two tempi from castling (like in the
present situation), the situation is unclear – sometimes you are lost, sometimes it is okay.

9...cxd4

If 9...Qxd4,

408
 analysis diagram

White is better after any logical queen move. Strangely, Stockfish advocates 10.Qe1!?±. This is a
typical computer move, both ugly (it blocks the f1-rook) and unnatural! (the reason is that on
10.Qe2+= Be6 11.Bg5 h6 12.Bh4, Black can resist by winning a tempo with 12...Qc4!?, attacking
the queen, although White is still better; 10.Qf3+=) 10...Be6 11.Bg5‚ h6 12.c3 (12.Rd1 Qxb2)
12...Qe5 13.f4 Qf5 14.Bh4.

10.Re1 Be6 11.Bg5

Not allowing queen​side castling.

11...Be7 12.f4 Qd5

On 12...Qf5, 13.Ng3 wins a piece, as Bxe7 and f5 will follow.

13.Bxe7 Kxe7 14.f5 Qxf5 15.Qxd4 Qd5

If 15...Rhd8, 16.Qb4+ Ke8 17.Nd6+ wins.

16.Qb4+ c5 17.Nxc5 a5 18.Qa3 Qd4+ 19.Kh1 Qb4 20.Qe3 1-0

Black might try 20...Qc4, but after 21.Qg5+ (other moves also win) 21...Kf8 22.Rad1 he cannot
defend.

Game 82
Magnus Carlsen 2552

409
 Evgeny Vladimirov 2621
Dubai 2004 (4)

White to move

21.Qe4!!

If you have seen this move and why it is playable, you are GM level, at least!

21...Qc6

The main variation which Carlsen had to calculate is 21...Bc6. If you haven’t seen this move, you are
not yet GM level but your intuition is good! 22.Qe2 Bxh1 23.Nxe6! Qe7.

410
 analysis diagram

Here, the only move is the beautiful 24.Bf5!!. There is no satisfactory defence against 25.Rd8+ Rxd8
(25...Qxd8 26.Nc/g7#) 26.Nc7#.

22.Qe2 a5 23.Nh5 a4 24.Rhf1 Rc7 25.Qf2 Bc8 26.Qd4 Rd7 27.Nxg7+ Bxg7 28.Qxg7 Rf8 29.Qxh6
b3 30.axb3 axb3 31.cxb3 Ba6 32.Bxa6 Qxa6 33.Qf4 Ra7

34.Qb8+ Ke7 35.Qb4+! 1-0

411
Nice geometry!
If 35...Ke8, 36.Rd8+ wins by force, as White will take on f8 and include the rook in the attack with
check.

Game 83 Nimzo-Indian Defence

Magnus Carlsen 2581
Nurlan Ibrayev 2476
Mallorca ol 2004 (10)

1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.Qc2

This looks like the usual Carlsen.

4...c5 5.dxc5 Bxc5 6.Nf3 Qb6 7.e3 Qc7 8.b3 b6 9.Bb2 Bb7

10.Nb5!?

Good, but very direct; such knight moves run the risk of ulti​mately resulting in badly-placed pieces,
and are more typical of aggres​sive players like Kasparov or Topalov.

10...Qd8 11.0-0-0

In the mood for an attack!

11...0-0

412
12.Ng5!

This is very strong, and White is already winning.

12...Re8

13.h4

Atypically, Carlsen misses a nice tactic: 13.Nc7! Qxc7 14.Bxf6 g6 15.h4. Now the attack plays itself:
15...h5 16.Bd3 (the sacrifice on g6 is unavoidable; even 16.g4!? Bxh1 17.gxh5 also gives a winning

413
attack). The more sophisticated 13.Bd3!? was also good after 13...h6 14.Bh7+ Kf8 15.Ne4!? Nxh7
16.Nxc5+–.

13...e5 14.Nd6 Bxd6 15.Rxd6 h6 16.Qf5

A rare occurrence when in a winning position, Carlsen spoils his advantage by playing for tricks.
16.Ba3+–; 16.Bd3+–.

16...Nc6

16...Qe7! 17.Rxf6 Qxf6 18.Qh7+ Kf8 19.Ba3+ d6 20.Bd3∞.

17.Rxf6! 1-0

17...Qxf6 (17...gxf6 18.Qh7+ Kf8 19.Qxf7#) 18.Qh7+ Kf8 19.Ne4! (19.Ba3+ Nb4 20.Bxb4++–).

Game 84
Magnus Carlsen 2570
Zurab Azmaiparashvili 2658
Khanty-Mansiysk 2005 (1)

414
 White to move

31.e6! fxe6

31...f6 32.Nxf6+ exf6 33.e7+.

32.Nf6+! Kh8 33.Rd8+ Kg7 34.Rg8+

Black resigned just in time to avoid 34...Kf7 or 34...Kxf6 35.Qxe6#.

Game 85
Magnus Carlsen 2863
Jan-Krzysztof Duda 2757
Stavanger 2020 (6)

415
 White to move

24.Nf6+!

This tactic is reminiscent of Carlsen’s game against Azmaiparashvili, fifteen years earlier.

24...Kh8

24...Nxf6 25.Qxb4; 24...gxf6 25.Qxe6+.

25.Qxe6 Ra8

25...Ne7 26.Ba1+–.

26.Qxd5 1-0

Game 86 Sicilian Defence

Magnus Carlsen 2570
Farrukh Amonatov 2572
Khanty-Mansiysk 2005 (2)

This whole game is a hymn to piece activation and activity. The 14-year-old Carlsen was already
perfect concerning piece placement and piece activation in calm positions with a static centre.
White’s leitmotiv in the game is: how to get the best from my pieces/how to get another piece in play,
while keeping the initiative – in a Morphy- or Fischer-like style.

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Be3 e5 7.Nf3 Qc7 8.a4 b6

416
9.Nd5! Nxd5 10.Qxd5 Nc6

11.Ng5! Rb8 12.Bc4!

The showy 12.Nxf7 fails to 12...Nb4!.

12...Nd8 13.Qd3 h6 14.Nf3 Ra8 15.Nd2

This is one of only two moves that don’t take or attack something – from move 9 to 31!

417
15...Ne6 16.Bd5 Rb8

17.Qc4!

Now, White wants to improve the d2-knight.

17...Nc5 18.b4 b5 19.axb5 axb5 20.Qc3 Ne6

21.Qxc7 Nxc7 22.Ba7!

418
Winning material at last.

22...Nxd5 23.Bxb8 Nxb4 24.0-0 Nc6

25.Rfb1!

Brilliant! Another piece is activated.

25...Be7

25...Nxb8 26.Ra8+–.

26.Rxb5 0-0 27.Nc4!

27.Bc7 was enough to secure the win, but Carlsen is having fun.

27...Be6

419
Isn’t this a double attack?

28.Bxd6! Bxc4

If 28...Nd4, 29.Bxe7 Nxb5 30.Bxf8.

White is a piece down and the ‘victim’ of a double attack. But

29.Rc5!!

recoups everything with interest.

29...Rc8

29...Bxd6 30.Rxc6.

30.Rxc4 Bxd6 31.Ra6 Rb8 32.g3 Nd4 33.Rxd6 Rb1+ 34.Kg2 Rb2 35.c3 Nc2 36.Kf1 1-0

Game 87
Ivan Cheparinov 2618
Magnus Carlsen 2570
Khanty-Mansiysk 2005 (3)

420
 Black to move

55...Re3! 56.Rfb2

56.Rb4+ (56.Rf3 Rxf3 57.Rxf3 Re2–+) R5e4 57.Rxe4+ fxe4–+.

56...Rxb3 57.Rxb3 Re2+ 58.Kg1 Ra2 59.Rc3 Kh3 60.Rf3 Ra5 61.Kf1 g4 62.Rb3 Ra2 63.Rb5 Kxg3

This is a winning ending.

64.Rxf5 Ra1+ 65.Ke2 Kg2 66.Rf2+ Kh3 67.Rf8 g3 68.Rh8+ Kg2 69.Re8

69.Rg8 Kh2 70.Rh8+ Kg1.

69...Kg1 70.Rg8 g2 71.Rg7 Ra8 72.Rh7 Re8+ 0-1

After 73.Kd2, there follows the well-known Lucena manoeuvre, when Black builds a kind of bridge
against the vertical checks of the white rook: 73...Re5! (73...Kf2 is not enough, because the rook is
able to keep checking from behind and the king cannot escape, because it has to protect the passed
pawn: 74.Rf7+ Kg3 75.Rg7+ Kf3 76.Rf7+ Kg4 77.Rg7+) 74.Rh8 Kf2 75.Rf8+ Kg3 76.Rg8+ Kf3
77.Rf8+ Kg4 78.Rg8+ Rg5, and wins.

Game 88
Magnus Carlsen 2570
Vladimir Malakhov 2670
Khanty-Mansiysk 2005 (6)

421
 White to move

23.h4!

This is the best attacking move, from an overwhelming position. As Black is reduced to complete
passivity, White has time to include the h-pawn among the attacking forces, with the main idea being
to exchange it against an important defender of the castled king, the g6-pawn.

23...Nf8 24.Nc5 Qa7 25.h5 Nc7 26.hxg6 hxg6

422
27.Bxf8!

The bishop is exchanged to let the queen have its say.

27...Rxf8

27...Kxf8 28.Qh6+ Bg7 29.Qh7 (threatening mate in one!) 29...e6 30.Nd7+ Ke7 31.Nf6!
(31.Qxg7+–) 31...Bxf6 32.exf6+ Kxf6 33.Ne5 Rf8 34.Rc3.

28.Qh6 Bg7 29.Qh4 f6 30.Qg4

White is now attacking the kingside on the light squares with the queen and the knights, and Black’s
possible defenders are too far away.

30...Kh7 31.Nh4 g5 32.Qh5+ Kg8 33.Nf5 Ne8 34.Qg6

Black resigned.

Game 89
Magnus Carlsen 2553
Alexandra Kosteniuk 2490
Wijk aan Zee 2005 (2)

White to move

19.Ncd2!

This kind of retreat is a Carlsen special. Jussupow talks about Carlsen’s famous backward moves. We

423
already saw one in the game against Short. The c4-knight was not bad, but it prevents White putting
pressure on the c-file, and Magnus wants a knight on c5, which is a stronger square.

19...Bb5 20.Nc5 Nxc5 21.Qxc5 Ba4 22.Ne4 Rb3 23.Qc4 Nb6 24.Qe2 Bb5 25.Qc2 Ba4 26.Nc5
Rb4 27.Qd2 Rb5 28.Bf1 Qd6 29.Bxb5 cxb5 30.Qa5 e5 31.Nxa4 1-0

Game 90
Magnus Carlsen 2625
Erwin l’Ami 2550
Wijk aan Zee 2006 (1)

White to move

26.g4!

A very courageous, aggressive and enterprising move. It is difficult to imagine Petrosian or Karpov
playing such a move, but here it is the only chance to play for an advantage. White endangers his own
kingside structure, as Black’s army is aiming at the other side of the board.

26...Rh8?!

26...hxg4? loses immediately to 27.Nxg4 Rh8 28.Nxf6 Kxf6 29.Qf4+ (29.Qd4+, with e4-e5 to come,
is even better).
26...h4 was more solid. After 27.Nf3, White is a touch better: 26...Ne6!? 27.Bxe6 fxe6 28.gxh5 gxh5
29.Kh1, with complicated play.

27.Kg2!

424
This is a small but very important move. White’s king slides on to the light squares (out of reach of
the opponent’s queen and bishop); it protects the h3-pawn and keeps closer to its pawn shelter; it
allows the rooks to use the h- and g-files. This move would be played ‘a tempo’ by experienced
players (they wouldn’t comment on it, as it is an automatic one), but if you did not think of it, you
must add it to your positional arsenal.

27...hxg4 28.Nxg4 Nd7

29.Bxf7!

Bingo!

29...Kxf7 30.Qf4 Ke6

After the preliminary 30...g5, a computer idea played to keep White’s queen from attacking d6, play
continues 31.Qf5 Ke7.
White now gets the rooks into action with a kind of move Magnus is not likely to miss: the line
opener...

425
 analysis diagram

32.e5! Qc6+ (32...Bxe5 33.Nxe5 Nxe5 34.Rxe5+ dxe5 (34...Kd8 35.Qf6+ Kc7 36.Re7+ Kb8
37.Rxd6) 35.Rd7+ Ke8 36.Qf7#) 33.Kh2 Bxe5+ 34.Nxe5 dxe5 35.Rxd7+ Qxd7 36.Rxe5+ Kd8
37.Rd5. After winning the queen, White will collect Black’s pawns with checks, as the rooks have no
time to connect.

31.Nxf6 Nxf6 32.e5 Qb7+ 33.Kg1 dxe5 34.Rxe5+ Kf7 35.Rd6 Rxh3 36.Qxf6+ Kg8 37.Rd8+ 1-0

Game 91
Magnus Carlsen 2646
Julio Granda Zuniga 2631
Turin ol 2006 (11)

426
 White to move

14.Nc1

This is good, solid, twentieth-century positional chess! The move was originally praised by
commentators, and Magnus’ erudition was highlighted with this nice classical treatment of the
position, like in a famous Geller-Fischer game (Game 92). Thanks to some players like Kasparov, and
with the help of computers, chess made great progress after the Karpov era. For example, the slow,
positional manoeuvres that were popular at the time proved inappropriate in many positions, because
of dynamic problems. Kasparov said in an interview in the early 1990s that it appeared that chess was
much more tactical than was originally thought. Probably, the 2021 version of Carlsen would discard
such an academic plan; he has transformed himself, at least since 2018, into a much more dynamic
player, although he retains his characteristic style.
White would have done better to wait for a more favourable moment to accomplish the knight
manoeuvre and play 14.Rd2 Re8 15.Kh1. He might continue his long preparations with Be2, followed
by Qf1/f2, etc. Then only when Black is not ready to react with ...d6-d5, should he play the Nc1-a2-
b4-d5 circuit.

14...Nc5

If Black moves his rook with 14...Re8, anticipating his counter-attack, White continues his intended
manoeuvre with 15.N1a2.

427
 analysis diagram

16.Nb4 on the next move would promise White an excellent game, but Black now plays 15...d5!, and
the game is getting complicated, e.g. 16.Nxd5 Nxd5 17.exd5 f5 18.Rd2 (18.Qd3 Nf6 19.Nc3
(19.Qxf5 Nxd5) 19...f4) 18...Bd6 19.g3 e4 20.f4 Nf6 21.c4 Ng4, with counterplay.

15.Rd2 Ne6 16.N1a2 Nd4 17.Nb4 Rd8 18.Kh1 a5 19.Nd3 Ne6 20.Nf2 Nd7

A blunder in a worse position;

20...Bc6.

21.Nb5 Qb8 22.Nxd6 1-0

Game 92
Efim Geller
Bobby Fischer
Stockholm 1962 (18)

428
 White to move

18.Nc1!? Ng6

18...Rd8! 19.N1a2 d5.

19.N1a2!

White is clearly better. Fischer fought well with his next moves and was able to get some counter​play
that was ultimately enough to save the draw:

19...b5 20.axb5 axb5 21.Bxb5 Bxb5 22.Qxb5 Qxb5 23.Nxb5 Ra5 24.Nbc3 Rca8 25.Rdd1

25.b4 Ra3 26.Rb1 Rxa2 27.Nxa2 Rxa2 28.b5 offered better chances of a win.

... ½-½ (52)

Game 93
Magnus Carlsen 2698
Maxime Vachier-Lagrave 2579
Cap d’Agde 2006 (4)

429
 White to move

15.Rf3!?

The move is brilliant if not sound, but it reflects Carlsen’s attacking style: bring on as many pieces as
possible against the enemy king!

15...exf5!

Vachier-Lagrave is certainly one of the best defenders on the circuit.

The idea of White’s sacrifice is to open the g-file, thus enabling the a8-rook to ‘join the party’ as
quickly as possible: 15...Nxf3 16.gxf3 and now:
 A) 16...Qb6 17.Rg1!! Bxd4

430
 analysis diagram

18.e5!. Here the classical line opener is also a line closer, as the black bishop won’t protect f6
anymore: 18...Bxe3 (18...Bxe5 19.Bxb6) 19.Rxg7+ Kxg7 (19...Kh8 20.fxe6+–) 20.exf6+ Kh8
21.fxe6 and White wins.
16...Ne8 is a computer try; after 17.Rg1 g6 18.Qh6, intending Rg1-g4-h4, White has a strong attack,
but Black still has some chances to resist.

16.exf5 Nxf3 17.gxf3 Bb7 18.Rg1 Bxf3+ 19.Nxf3 Bxe3 20.Qxf6 Bxg1 21.Qg5 Qc6 22.Ne4 f6
23.Qxg1

Although this was a rapid game, it appears that since the rook sacrifice, both players – aged 15 –
found the best moves.

23...Rae8 24.Qd4 d5 25.Nc5 Re7 26.Kg2 g6 27.Ne6 Rxe6 28.fxe6 Qxe6 29.a4 bxa4 30.Qxa4 Rb8
31.b3 Rb6 32.Qa5 Kg7 33.Nd4 Qd6 34.c3 h5 35.h4 Kh6 36.b4 Rb8 37.Qxa6 Qf4

431
38.Nf5+! 1-0

Characterizing tactical themes makes sense, especially for aspiring club players, but the task is not
easy because several themes often appear even in very simple tactics. 38.Nf5+ is an obstruction. It
keeps the black queen from protecting f6 at first, but in the end, the combination turns out to win
thanks to a discovered attack that is a clearance, exactly the reverse of the obstruction! 38...gxf5
39.Qxf6+ Kh7 40.Bxf5+ Kg8 41.Be6+.

432
 Game 94 Queen’s Gambit Declined
Magnus Carlsen 2881
Vladimir Kramnik 2756
Legends of Chess Prelims 2020 (9.3)

1.d4 Nf6 2.c4 e6 3.Nc3 d5 4.cxd5

This Exchange Variation is exactly the reason why Kramnik (and many other players) very rarely
played the Queen’s Gambit Declined when White has not committed to Nf3. The best players
generally adopt the Nimzo move-order 1.d4 Nf6 2.c4 e6, and only after 3.Nf3 play the Queen’s
Gambit with 3...d5. The main reason is the plan used in the game and famously inaugurated by
Botvinnik in the 1930s (see Game 95).

4...exd5 5.Bg5 c6 6.e3 h6 7.Bh4 Be7 8.Bd3 0-0 9.Nge2 Re8 10.0-0 Nh5 11.Bxe7 Qxe7 12.Qd2
Nd7 13.Rae1 Nf8 14.Nc1 Nf6 15.f3 Ne6 16.N1e2 b5

This move was criticized by Grischuk.

17.Ng3 a5

18.Qf2!?

Magnus likes to play for a kingside attack and wants to include his best piece as soon as possible.
Keeping in mind that this was a rapid game, and that probably he would have played another move
with more time on the clock, this display of impatience reflects the current evolution of the World
Champion, who is eager to go for the opponent’s throat as soon as possible.

433
The positional move 18.Nce2 was probably stronger.

18...Ba6

18...b4 is better: 19.Nce2 (19.Na4 c5) 19...g6!. Now, if 20.e4 dxe4 21.fxe4 Ng4! (a typical resource),
Black is not worse after 22.Qf3 Qa7.

19.Nf5 Qd8 20.Ne2 c5 21.Qg3 c4 22.Bc2 Nh5 23.Qh3

Black is positionally crushed, as e3-e4 will follow, with the famous pawn offensive in the centre
known since Botvinnik’s most famous victory, against Capablanca in 1938.

23...g6

23...Nf6 24.e4, and White continues with e4-e5, with a large space advantage that could be turned
into a winning attack.

24.Nxh6+ Kg7 25.Nf5+ Kg8

26.Neg3

The computer advocates 26.g4.

26...Nxg3 27.Qxg3 b4

27...Qg5 was the best try, but Kramnik is not fond of playing endgames a pawn down against
Carlsen.

28.Nh6+ Kg7

434
29.Nxf7 Kxf7 30.Bxg6+ Ke7 31.Bxe8 Kxe8 32.e4 1-0

Game 95
Mikhail Botvinnik
José Raúl Capablanca
AVRO Netherlands 1938 (11)

This was Botvinnik’s best game, introducing a plan that is still crucial more than eighty years later.

White to move

435
19.e4!± Qxa4 20.e5 Nd7 21.Qf2 g6 22.f4 f5 23.exf6 Nxf6 24.f5 Rxe1 25.Rxe1 Re8 26.Re6 Rxe6
27.fxe6 Kg7 28.Qf4 Qe8 29.Qe5 Qe7

30.Ba3!+– Qxa3 31.Nh5+ gxh5 32.Qg5+ Kf8 33.Qxf6+ Kg8 34.e7 Qc1+ 35.Kf2 Qc2+ 36.Kg3
Qd3+ 37.Kh4 Qe4+ 38.Kxh5 Qe2+ 39.Kh4 Qe4+ 40.g4 Qe1+ 41.Kh5

Black resigned.

Game 96
Magnus Carlsen 2733
Judit Polgar 2707
Wijk aan Zee 2008 (6)

436
 White to move

From here on, Carlsen outplays his opponent in the manoeuvring game, and gains a winning advan​‐
tage after a series of ‘best’ moves:

16.Kf2!

Out of all the undeveloped pieces, Carlsen prefers to play his king first, a nice touch. There is a logic
in it, as the king is one of the strongest of the remaining pieces. In the endgame, its value is
considered to be in the middle between a bishop and a rook.

16...Ne5 17.Bb2 f6 18.Rd1 Be6

437
19.Ne2!

An excellent sacrifice.

19...Bf7

19...Nexf3 20.Nf4 Bf5 (20...Bf7 21.Kg3±; both knights are attacked by the king, so they cannot
move, and the weakening ...g7-g5 will be forced at some point) 21.Rxd5 and White is clearly better.

20.Rg1 Nc4 21.Bc1

White has four pieces on the first rank, two on the second, and yet he has a big advantage!

... 1-0 (52)

Game 97
Vladimir Kramnik 2799
Magnus Carlsen 2733
Wijk aan Zee 2008 (12)

438
 Black to move

16...Nc7!

Black intends to push ...f7-f5 and ...g7-g5, to get space on the kingside, a strategy that resembles the
Stone​wall Dutch Carlsen would use later in his career, with great success.

17.Nf3 f5 18.Nc3 g5 19.Qd2 g4 20.Ne1 Bg5 21.e3 Rff7 22.Kg1 Ne8 23.Ne2 Nf6 24.Nf4

24...Qe8

439
A good and logical move.
Carlsen explained he had intended 24...Ne4, but feared the queen sacrifice 25.Nxe6 Qe8 26.Nxg5
Nxd2 27.Nxf7, forgetting that after 27...Ne4! (backward knight moves are notoriously difficult to
see) 28.Nh6+ Kf8 29.Nxf5 Qf7!, Black wins. Carlsen certainly is an intuitive player, but he
thoroughly checks forced lines until a certain evaluation. When in doubt, he prefers not to enter
complications if he has less-forced alternatives, being confident to outplay the opponent later in the
game.

25.Qc3 Rg7 26.b4 Ne4 27.Qb3 Rge7 28.Qa4?!

28.h4=.

28...Ne5! 29.Qxa6 Ra7 30.Qb5

On 30.Qxb6? Reb7 31.Qd4, the best move is 31...Bf6!–+.

30...Qxb5 31.cxb5 Rxa2 32.Rc8+ Kf7 33.Nfd3 Bf6 34.Nxe5+ dxe5 35.Rc2 Rea7 36.Kg2 Ng5
37.Rd6 e4 38.Bxf6 Kxf6 39.Kf1 Ra1 40.Ke2 Rb1 41.Rd1 Rxb4 42.Ng2 Rxb5 43.Nf4 Rc5 44.Rb2
b5 45.Kf1 Rac7 46.Rbb1 Rb7 47.Rb4 Rc4 48.Rb2 b4 49.Rdb1 Nf3 50.Kg2 Rd7 51.h3 e5 52.Ne2
Rd2 53.hxg4 fxg4 54.Rxd2 Nxd2 55.Rb2 Nf3 56.Kf1 b3 57.Kg2 Rc2 0-1

Game 98
Magnus Carlsen 2765
Alexei Shirov 2740
Foros 2008 (5)

White to move

440
63.Bg6+ Kg5 1-0

64.h4+ gxh3 65.f4# or 64.f4+ gxf3 65.h4#.

Game 99
Alexander Huzman 2590
Magnus Carlsen 2786
Kallithea 2008 (7)

Black to move

17...Kd7!

Carlsen plays without prejudice.

18.Bb4 b5 19.Qa3 Bxb4 20.Qxb4 a5 21.Qb3 Ne6 22.Rxc6 Qxc6 23.d5 Qc4 24.dxe6+ Kxe6
25.Nd4+

441
25...Kd5!?

Typical enterprising play. Carlsen had calculated that he risked nothing, and once again he gave the
opponent an opportunity to go wrong.

26.Qf3+ Kxd4 27.Qe3+

27.Re1! was a good move, e.g. 27...Kc5 28.Qe3+ Kc6 29.Rc1.

27...Kd5 28.Rd1+ Ke6 29.Rd6+ Ke7 30.Qg5+ Ke8 31.e6 Rc8 32.exf7+ Qxf7 33.Rd3 Qb7+ 34.Kh3
Rf8 35.Qh5+ Ke7 36.Re3+ Kd8 37.Rd3+ Kc7 38.Rd5 Rce8 39.Rc5+ Kd6 40.Rxb5 Qc8+ 41.Kh4
Qc4+ 42.g4 Qd4 0-1

Game 100
Magnus Carlsen 2772
Dmitrij Jakovenko 2742
Nanjing 2009 (10)

442
 White to move

13.e4! 0-0

The standard 13...dxe4 is wrong with the king still in the centre, because of 14.Nxe4 Qc7
15.Qa3!+–.

14.e5 Qe6 15.Rae1 Rfe8

Best was probably 15...f6+ 16.exf6 Qxf6, and White is slightly better; but not 15...c5?!, because of
16.Nb5±.

16.Nh4!

White logically intends to use the f-pawn in Botvinnik style.

16...Ng6 17.Nxg6 Qxg6 18.Qd2

18.Qxg6?! spoils White’s advantage. With more space and the prospect of an attack on the kingside,
keeping the queens is a no-brainer.

18...Nf8 19.f4 Qf5?!

The only possible resource, played by Capablanca in the 1938 game, is 19...f5, when White will put
pressure on the c6-pawn with b2-b4 and doubling rooks on the c-file.

20.Nd1

443
20.Ne2 and Ng3 was an alternative.

20...f6 21.Ne3 Qd7 22.Qd3

Now, Ne3-f5-d6 is on the agenda.

22...fxe5 23.dxe5!

23.fxe5 was also good for White, but not as crushing as the text.

23...Ne6 24.f5

The 4/2 majority in motion.

24...Nc5 25.Qd4 Ne4

26.Nxd5!?

A surprising ‘little combination’.

Maybe 26.f6 was simpler: 26...Re6 27.Nf5 c5 28.Qd3 Rxe5 29.Rxe4 Rxe4 30.f7+ Kh8 (30...Kf8
31.Qc3 d4 32.Qf3 Qc6 33.Nd6+–) 31.Ng3+–.

26...Qxd5

26...Nc5! was a tougher defence. It is not easy to find a forced win for White, but the feeling is that it
exists. Here is the main Stockfish 11 variation, with many difficult moves: 27.f6 Red8 28.f7+ Kf8
29.e6 Nxe6 30.Qb4+ c5 31.Qe4 Qxd5 32.Qxh7 Ke7 33.Rf5! Qd4+ 34.Rf2! Rd6 35.Re4! Qd1+

444
36.Rf1 Qd3 37.f8=Q+ Rxf8 38.Qxg7+ Kd8 39.Rxf8+ Nxf8 40.Qxf8+ Kc7 41.Qe7+ Rd7 42.Qe5+
Kb7 43.Re1. This ending must be winning, and is probably easier to win if the queens are exchanged,
which is not easy to avoid.

27.Qxe4 Rad8 28.e6 Qxe4 29.Rxe4 Rd6 30.g4 Kf8 31.g5 Ke7 32.Kg2 Rd5 33.Kg3 Kd6 34.h4 c5
35.f6 gxf6 36.gxf6 Rd3+ 37.Kh2 Rd2+ 38.Kh1 1-0

Game 101
Magnus Carlsen 2813
Wang Yue 2752
Medias 2010 (4)

White to move

29.d6!

‘Passed pawns must be pushed’ is an old rule, and Carlsen generally follows it faithfully, especially in
rook endgames.

29...Bf5 30.Nb6 Be6 31.d7 1-0

Game 102 King’s Indian Defence

Magnus Carlsen 2690
Alexander Morozevich 2741
Morelia/Linares 2007 (1)

1.d4 Nf6 2.c4 g6 3.g3 Bg7 4.Bg2 0-0 5.Nc3 d6 6.Nf3 c5 7.0-0 Nc6 8.d5 Na5 9.Nd2 e5 10.b3 Ng4

445
11.h3 Nh6 12.Nde4 f6

13.Nxd6!?

A very interesting sacrifice, giving up a knight for two central pawns, with the idea of blocking the
opponent’s pieces with the eight-pawn wall, and to advance in the centre. This is a known theme, but
here what is interesting is that the sacrifice is completely unforced and made ‘for free’, as White has a
small space advantage whatever non-sacrificial move he chooses. Carlsen simply evaluates that on
this day, against this opponent, it is the most practical way to play for a win. It also allows him to play
with a clear plan in mind.
Yet again, apart from the young Tal, I don’t know who among the champions of the past would have
played like this in a favourable position. It gives an indication that Carlsen’s style is more complex
than the Karpov/Fischer hybrid (with a touch of Capablanca) that he is known for, because neither
Karpov nor Fischer would have sacrificed in this position, as both players were quite materialistic and
not fond of unclear sacrifices. It is a characteristic of Carlsen, who is rightly known as a positional
and technical player, that he loves to sacrifice as much as a pure attacker in any phase of the game.

13...Qxd6 14.Ne4 Qd8 15.Nxc5 f5 16.d6 e4 17.d7 Nf7 18.Rb1 Qe7 19.dxc8=Q Raxc8 20.Na4 1-0

Game 103
Magnus Carlsen 2802
Nigel Short 2680
London 2010 (7)

446
 White to move

20.Nb6!

Thanks to his usual open-minded​ness, Carlsen executes a brilliant exchanging operation, allowing
Black to take a valuable bishop and to destroy White’s pawn structure.

20...Nxe3

20...Nxb6 is of no use, as White has a large advantage after 21.Bxb6 e5 22.Rd5.

21.fxe3 Rb8 22.Kf2 e5 23.Rd5+– Be6 24.Rxa5 Rd8 25.Bd3 Bf6 26.e4 Rd4 27.Rb5 g6 28.a5 Kg7
29.Ke2 Bg4+ 30.Kf2 Be6 31.Kf3 h5 32.Nd5 Bd8 33.b4 Bc8 34.Ke2 Bg4+ 35.Ke1 f5 36.Rxb7+
Kh6 37.Ne3 Bf3 38.a6 fxe4 39.Be2 Bg5 40.a7

Black resigned.

Game 104
Magnus Carlsen 2835
Richard Rapport 2731
Wijk aan Zee 2019 (8)

447
 White to move

31.Ne3! Qc6

Carlsen described his way of play​ing the attack in a series of videos on Chess 24 called ‘Attacking
without sacrificing’. All of White’s pieces suddenly attac​k the king after...

32.Rd5!

According to the computer, Black is already lost, in spite of his small material advantage.

448
32...Qe6 33.Rh5 Qf7 34.Qh3 g6 35.Rh4 Rb6 36.f5 Ne5 37.Nd5 Rd6 38.fxg6 Nxg6

39.Bxf6+ Rxf6 40.Rxh7+ 1-0

Game 105
Magnus Carlsen 2845
Anish Giri 2797
Shamkir 2019 (7)

White to move

449
17.f4!

A fantastic attacking idea.

17...exf4 18.gxf4

This is the follow-up Carlsen had in mind, giving up a pawn with tempo and ruining his pawn
structure, to get an attack against the enemy monarch.
This concept is so original and daring that Giri immediately collapses:

18...Qxe3+

18...Kh7!, protecting h6 and thus intending ...g7-g6, was shown by the computers to be the right
defence. It’s an inhuman move, because it looks like White’s pawn mass should be overwhelming.
Nevertheless, it appears that Black is fine after 19.e4 g6.

19.Kh1 Rd8 20.Rce1 Qc5 21.f5

White already has a decisive advantage.

21...Bf8 22.Be4 Rd5 23.Rf3 b5 24.Rg1 Ra7 25.Bf6 g6

26.Qh3

The natural move 26.Rxg6+ was better, but White had to foresee a cunning move in the key variation:
26...fxg6 27.Qxg6+. Now on 27...Bg7 (27...Rg7 is simpler for White: 28.Bxg7 Bxg7 29.Rg3 (29.f6?
Qc1+–+) 29...Rd7 30.f6+–),

450
 analysis diagram

28.Bc3! is the only winning move. Magnus did not see it – another invisible, backwards move. The
bishop’s retreat allows the pawn to go to f6, with devastating effect.

26...Rd6 27.Qh4 Rxf6 28.Qxf6 Be7 29.Qxc6 Qxc6 30.Bxc6 Kg7 31.fxg6 fxg6 32.d4 a4 33.d5 b4
34.Be8 Bg5 35.h4 Bxh4 36.Rxg6+ Kh7 37.Rc6 Bg4 38.Rf4 Rg7 1-0

Game 106
Sergey Karjakin 2753
Magnus Carlsen 2845
Shamkir 2019 (8)

451
 Black to move

20...0-0!

Black sacrifices a pawn to gain some tempi (two) for his attack.

21.Bxh5 Ne5 22.Be2 Qd7 23.Qa4 Qc8 24.c5 dxc5 25.Nxe4 c4 26.Nc3 b5 27.Qd1 b4 28.Na4

28...Be4!

452
Such moves, pure piece-play, are elementary for Carlsen, letting the queen come into play while
advan​tageously repositioning the bishop.

29.Qd4 Qf5 30.f4 Qg6 31.Bf2 Nd3 32.h5 Qf5 33.Bg4 Qxg4 34.Qxe4 Bd6 35.Qg2 Rae8 36.Bd4
Qxh5 37.Qf3 Qg6 38.Kh1 Re4 39.Bf2 Rfe8 0-1

The rooks will penetrate crushingly on the second rank.

Game 107
Magnus Carlsen 2861
Wei Yi 2736
Abidjan 2019 (17)

White to move

25.R1e2

A diabolical attempt in a blitz game.

To the direct 25.Nxb7, Black replies 25...Qc7! (25...Rxb7 26.Qxf5!), and on 26.Nc5, Black has
26...f6 27.Rxf5 exf5, threatening checkmate on e1.

25...Kh8?

Black might have played 25...Nd6 with a solid position, or 25...Qc8!?.

453
26.Nxb7! Rxb7

26...Qc7 27.Nc5 f6 28.Rxf5 exf5 29.Qxc7+–.

27.Qxf5 Rc7 28.Qd3 Qc8 29.R5e3 Rc6 30.Rh3 h6 31.Qd2 Kh7 32.Re5 Re7 33.Reh5 Qf8 34.f4 f5
35.Re3 Qf7 36.Rhh3 Rec7 37.Re5 Qf8 38.Rd3 Qc8 39.Ree3 Rc4 40.h3 Qf8 41.Rxe6 1-0

Game 108
Sergey Karjakin 2772
Magnus Carlsen 2853
New York Wch m rapid playoff 2016 (3)

454
 Black to move

22...a5!?

According to the computers, it was also possible, although much more engaging and less ‘practical’,
to play for an attack with 22...Qh5. After a standard move like 23.Bd2 (23.Qd1 Kh8 24.axb5 axb5
25.Rxa8 Rxa8 26.c4 Ng3! (26...Bg3) 27.Nxg3 fxg3 28.h3 Bg5–+), the idea is 23...Nxe4 24.dxe4.
The idea of 22...Qh5 is to force the d-pawn to take on e4. After 24...Ng3! 25.hxg3 Bxg3, checkmate
follows.

23.axb5 axb4 24.Bd2 bxc3 25.Bxc3 Ne3 26.Rfc1 Rxa1 27.Rxa1 Qe8 28.Bc4 Kh8 29.Nxf6 Bxf6

29...gxf6, threatening 30...Kg8, was strong, e.g. 30.Bd2 Nc2 31.Ra2 Nd4 32.Qd1 Rg8 33.Ba5 Bg3!
34.h3. This final position is considered clearly better for Black by Stockfish 11; it does not seem so
clear, as Black’s attack has been stopped for the moment, and what to do next is not obvious, as
White controls the only open file (34.hxg3 fxg3 35.Kf1 Qh5 36.Ke1 Qh1+ 37.Kd2 Qxg2+ 38.Kc3
Qh3, and the g-pawn is too strong).

30.Ra3

455
30...e4!?

In this crucial rapid game (the score was 1-1) and with little time left for both players, Carlsen is true
to his style and prefers to give up a central pawn, to get more space for his pieces and exchange the
bad bishop. This is an aesthetic and thematic sacrifice, and very unpleasant to face as White. This was
an excellent practical decision by the World Champion.
30...Qb8 is a computer suggestion, when Black is slightly better.

31.dxe4 Bxc3 32.Rxc3 Qe5 33.Rc1 Ra8 34.h3

The computers recommend 34.Qd2 Ra1 35.Be2, and White should hold, though many players would
prefer to be Black here.

34...h6 35.Kh2 Qd4 36.Qe1 Qb2 37.Bf1 Ra2 38.Rxc7 Ra1 0-1

Game 109
Magnus Carlsen 2853
Sergey Karjakin 2772
New York Wch m rapid playoff 2016 (4)

456
 White to move

With two checkmate threats on g2 and f1, Black would be winning were it not for...

50.Qh6+! 1-0

50...Kxh6 51.Rh8# or 50...gxh6 51.Rxf7#.

Game 110
Magnus Carlsen 2861
Bassem Amin 2704
Abidjan 2019 (10)

457
 White to move

Black would be safe if the rook was on a protected square, say b4. But now Carlsen demonstrates
once again how dangerous it is to have an unprotected piece on an open board, thanks to his favourite
tactic: the double attack.

32.e5!

Carlsen hardly ever misses such tactics. Here, the idea of a double attack is linked with two other
themes that are probably his two favourites (tactically and strategically): the pin against the queen and
the line opening (here, the b1-g8 diagonal and the e-file). In other words, how to give more power to
our pieces, while limiting the action of the opponent’s? This is quite an accurate definition of
Carlsen’s style, and generally the style of any strong chess player.

32...Ndxe5 33.Bxg6+ Nxg6

The main line is 33...Kxg6 34.Qc2+ Kf7 35.Qxa4+–.

34.Rxe7 Nxe7 35.Qc2+ Ng6 36.Qxa4 Nxf4 37.Qc6 Ne2+ 38.Kh1 1-0

Game 111
Magnus Carlsen 2823
Vasily Ivanchuk 2765
Sao Paulo/Bilbao 2011 (9)

458
 White to move

10.Bh3!

This is a brilliant and original concept: Black, in this hybrid Queen Indian/Dutch Defence, has control
over the e4-square that White would like to contest with Nd2 and maybe f2-f3. The problem with
having a bishop on g2 is that Black could exchange the bishop, which is interesting for him. For
example, the security of White’s king is lessened. In this kind of position, White will often play
moves (after Bg2, 0-0 and, say, Rfe1 or Rfd1) like Bf1, to be able to keep the light-squared bishops
after he moves the knight. Here, the bishop goes directly to h3, with similar ideas and gaining a
tempo. As we will see, the bishop on h3 will even unexpectedly gain a major role from this
uncommon square!

10...0-0 11.0-0 a5 12.Rd1 Qe8 13.d5 Na6 14.Bf4 exd5 15.Bxf5 dxc4 16.Ng5 Qh5

459
17.Rxd7! Kh8

After 17...Nxd7, 18.Bxh7+ Kh8 19.Bg6 Qg4 (19...Qh6 20.Nf7+) 20.h3 wins the queen.

18.Re7

A strong move that gives White a large advantage.

The stylish 18.Rad1! was even better. It is a very natural move, bringing the last piece into play, but
probably Carlsen was not able to make it work tactically for some reason: 18...Nxd7 19.Rxd7 Nc5
20.Re7 Rae8 21.g4! Qh4 22.Rxe8 Rxe8 23.Qxc4, and White’s attack cannot be stopped.

18...Nd5 19.Bg4 Qg6 20.Nf7+ Kg8 21.Bf5

460
21...Qxf5 22.Qxf5 Nxe7 23.Nh6+ gxh6 24.Qg4+ Ng6 25.Bxh6 Rf7 26.Rd1 Re8 27.h4 Nc5 28.h5
Bc8 29.Qxc4 Ne5 30.Qh4 Nc6 31.Rd5 Ne6 32.Qc4 Ncd8 33.Qg4+ Ng7 34.Qxc8

Black resigned.

Game 112
Vasily Ivanchuk 2765
Magnus Carlsen 2823
Sao Paulo/Bilbao rapid playoff 2011 (2)

461
 Black to move

16...f5!

Black now has an edge, as he can involve all his pieces in the attack against the opponent’s king.

17.exf5 Bxf5 18.f3

18.g4 does not work in view of 18...Nf4 19.gxf5 Qg5+ 20.Ng4 Nxh3+ 21.Kh2 Qxd2.

18...Bg6 19.Rf2 Qh4 20.Nd5 Rf7 21.Qc3 c6 22.Ne3 Nf4 23.Qd2 Rbf8 24.Nd1 h5 25.Nc3 Bf5
26.Ne2 Ne6 27.Qe3 Rf6 28.Kh1 Rg6 29.Rg1 Nc7 30.Rgf1 Nd5 31.Qd2 Rgf6 32.Nc3 Nf4 33.Ne4
Rg6 34.Rg1 Qd8 35.Nf1 Qe7 36.Nfg3 d5 37.Nxf5 Rxf5 38.Nc3 Qh4 39.Kh2 Qg3+

White resigned.

Game 113
Magnus Carlsen 2826
Boris Gelfand 2744
Moscow 2011 (2)

White to move

24.g5!

A brilliant defensive device against 24...Qh4. Carlsen, in typical style, gives up a pawn to mobilize
his queen and give it access to h4.

462
24...Rdd1

24...Qxg5 25.Qe4! holds the h4-square.

25.Ng3™ Bd6 26.Qe2™ Rg1 27.Qd3 Bc7 28.Ne4 Raf1+ 29.Ke2 Rxf3 30.gxf3 f5 31.gxf6 gxf6
32.Bxf6 Qh7

33.Qb5!

This is both a defensive and an attacking move. Once again, the idea is to provide more space for a
piece, here the white king!

33...Rg2+ 34.Kd3 Qd7+ 35.Qxd7 Nxd7 36.Bd5 Be5 37.f4 Bc7 38.Bc6 1-0

Game 114
Magnus Carlsen 2814
Wang Hao 2731
Wijk aan Zee 2011 (12)

463
 White to move

13.g3!

This little pawn move is the beginning of a carefully-planned operation. The alternative is 13.h4,
followed by g2-g3.

13...Be7

14.h4!

464
This attacking construction, announced by White’s previous move, is very strong, allowing him to
open the h-file after playing the king to g2 and the rook to h1. This is a typical way to exploit the
space advantage in this pawn structure.

14...b5 15.a4 Qb6 16.axb5 axb5 17.Kg2 c4 18.Ng5 Qd8 19.Bg4 Bxg5 20.Bxg5 Qc7 21.Rh1 Nb6

22.h5!

Carlsen has often weakened the opponent’s king’s defence with this break, generally after kingside
castling, or like here with the enemy king in the centre.

22...gxh5 23.Bxh5 Na4

465
24.Bxf7+! Kxf7 25.Qf3+ Kg8

On 25...Kg6, 26.Bf6! is decisive.

26.Rxh8+ Kxh8 27.Rh1+ Kg8 28.Qh5

Threatening 29.Bf6, with check​mate. There is no defence.

28...Rf8 29.Bf6 1-0

29...Qa7 30.Qh7+ Kf7 31.Qh5+ Kg8 (31...g6 32.Qg4!+–) 32.Qg6 Rf7 (32...Nxb2 33.Rh7 Rf7
34.Qh5) 33.Qh7+ Kf8 34.Qh8#.

Game 115
Magnus Carlsen 2814
Hikaru Nakamura 2751
Wijk aan Zee 2011 (8)

466
 White to move

20.g6!

This is a typical pawn sacrifice, intended to open as many lines as possible on Black’s kingside.

20...Nec4

After the opening of the h-file, the attack is too strong: 20...hxg6 21.hxg6 fxg6 22.Nxe6 Rxe6
23.Bxe5 dxe5 (23...Rxe5 24.Bg4, and the queen joins the party from h2) 24.Bg4 Rd6 25.Bxc8 Rxd2
26.Be6#.

21.Bxc4 Nxc4 22.Qd3 fxg6 23.hxg6 h6 24.Qg3 Qb6 25.Bc1 Qa5 26.Rdf1 Ne5 27.Nd5 Bxd5
28.exd5 Qxd5

467
29.Bxh6! gxh6

If 29...Qxd4 30.Be3 Qg4, White can execute a well-known checkmating pattern: 31.Rh8+ Kxh8
32.Qh2+ Kg8 33.Qh7#.

30.g7! Be7

30...Bxg7 31.Nf5+–.

31.Rxh6 Nf7 32.Qg6 Nxh6 33.Qxh6 Bf6 34.Qh8+ Kf7

468
35.g8=Q+!

This is the winning tactic, attacking f6 with the queen.

35...Rxg8 36.Qxf6+ Ke8 37.Re1+ 1-0

Game 116
Vladimir Kramnik 2784
Magnus Carlsen 2814
Wijk aan Zee 2011 (11)

Black to move

12...a6!

Carlsen does not believe in the strength of the white knight.

12...Ne4 was possible, but less incisive.

13.Nd6 Qb8 14.Qb4 a5 15.Qa3 Ba6 16.Ne5 b5 17.Qxa5

17.Rc6 b4 18.Qe3=.

17...Qxd6 18.Rc6 Qb8 19.Rxa6 Rxa6 20.Qxa6 Nxe5 21.dxe5 Qxe5 22.Qxb5 Rb8 23.Qd3 Rxb2
24.Qe3 Qxe3 25.fxe3 Rxe2 26.a4 Rc2 27.a5 Rc7 28.a6 Ra7 29.Bf1 Kf8 30.Rb1 Ke7 31.Rb7+ Rxb7
32.axb7 Nd7 33.Kf2 Kd6 34.Bb5 Nb8 35.Be8 Ke7 36.Bb5 f6 37.Kf3 Kd6 38.Be8 Kc7 39.Bf7
Kxb7 40.Bxe6 Kc6 41.Bg8 h6 42.Kg4 Nd7 43.Kf5 Ne5 44.h3 Kc5 45.g4

469
45...Kd6!

The first step is to get rid of the active white king by playing ...g7-g6, so f6 has to be protected.

46.Bh7 Ke7 47.Bg8 g6+ 48.Kf4 Nf7 49.Bh7 g5+ 50.Kg3 Nd6 51.Bg8 Ne4+ 52.Kg2

52...Kd6!

The second step is to go to the queen​side, to find a way to infil​trate with the king to the main weak​‐
nesses, e3 and h3. Having the knight is an advantage here, as White won’t have easy fortresses based

470
on keeping his king on a specific square like e2. The knight can control any square and chase the
enemy king from anywhere.

53.Kf3 Kc5 54.Bh7 Nc3 55.Bd3 Kb4 56.Ba6 Kb3

57.Bb7

Kramnik attacks the d5-pawn and hopes to exchange it for the e3-pawn, but it is a bad defensive plan
as Black’s king will be able to reach h2. Probably, the super-fit Kramnik from the Kasparov match in
2000 would have been able to find the relevant defensive idea, which is a question of energy more
than one of ability here. It seems to me that the capacity and willingness to defend bad positions is a
speciality of young players, and often declines with the years. For example, Kasparov, who was able
to resist in his first match with Karpov against all odds, being down 0-5 in a match to six wins, later
became a poor defender in many positions (Karolyi suggested that he was not used to worse positions
anymore). Even the 30-year-old Carlsen is probably less good a defender in bad positions than he was
ten years ago. However, this is offset by his supreme ability to anticipate the moment when things are
going from slightly worse to bad, and to alter the course of the game before getting into a hopeless
position.
Here, it seems that there was no win had White played the accurate 57.Bd3!, intending to stop the
black king’s penetration thanks to 57...Kb2 58.Kf2 Kc1 59.Ke1. It seems that no progress is
available to Black with the plan he was able to use in the game; in this case, Magnus would probably
try other plans, torture his opponent for hours (going back to the centre with the king, try some pawn
break...), and still would enjoy serious practical chances of winning, especially against an older
opponent. Yet, it seems that Black has no forced win.

57...Kc2 58.Ba6 Kd1 59.Bb7 Kd2 60.Bc6 Ke1 61.Bb7 Kf1 62.Ba8 Kg1 63.Kg3 Ne4+ 64.Kf3

471
Nd2+ 65.Kg3 Nf1+ 66.Kf3 Nd2+ 67.Kg3 Nc4 68.Bxd5 Nxe3

This kind of position is the night​mare of the bishop against a knight, completely winning and easily
so. Black will force the white king to leave g3, then attack h3 and win it.

69.Bb7 Nf1+ 70.Kf3 Kh2 71.Kf2 Nd2 72.Bg2 Nc4 73.Bf1 Ne5 74.Ke3 Kg1 75.Be2 Kg2 76.Ke4
Kxh3 77.Kf5 Kh4 78.Bd1 Nc4 79.Ke4 Nd6+ 80.Kd5

80...f5!

Any move wins, but this knight sacrifice is a Magnus special that he played many times in endings:
the most important thing at a certain point is to advance the passed pawns, as the pieces are worth
less, in particular the short-ranged knight.

81.Kxd6 fxg4 82.Ke5 g3 83.Bf3 Kh3 84.Kf5 g2 85.Bxg2+ Kxg2 0-1

Game 117 Scandinavian Defence

Fabiano Caruana 2801
Magnus Carlsen 2877
Tromsø ol 2014 (6)

1.e4 d5 2.exd5 Qxd5 3.Nc3

472
3...Qd8!?

A passive-looking move – hasn’t Black just lost two tempi? The believer in Black’s position
advocates that ideally White should put pawns on d4 and c4, so the c3-​knight is badly placed in front
of the c2-pawn, and if it has to retreat to play c2-c4, White has also lost tempi. Carlsen uses this
opening at times, and likes to play it against Caruana. He also already played 3...Qa5 and 3...Qd6
against the same Fabiano!

4.d4 Nf6 5.Nf3 Bg4 6.h3 Bxf3 7.Qxf3 c6 8.Ne2 e6 9.g4 Qd5 10.Bg2 Nbd7 11.Qg3 Qc4 12.Qb3
Qxb3 13.axb3 Bd6 14.c4 a6 15.Be3 0-0-0 16.0-0-0 Rhe8 17.Ng3 Nf8 18.Bf3 Ng6 19.h4 Bf4 20.h5
Bxe3+ 21.fxe3 Ne7 22.e4 h6 23.e5 Nh7 24.Ne4

473
24...Rf8!

This is a clever move that tempts the white knight to come to d6 with check. Carlsen feels that the
glorious knight is not to be feared.
24...Kc7 was possible, keeping away from the checks. But if the king move is played, maybe the
knight doesn’t ‘feel’ like going to d6.

25.Nd6+

Kasparov named such a knight an ‘octopus’ in the comments of one of his most famous games
against Karpov (Game 118). The knight has eight possible squares, like the eight limbs of the
mollusc, most of which reach deep into the opponent’s camp. Carlsen understands that without
queens, without mating chances and with many pieces (soon to be) exchanged, the knight might
simply be misplaced! This is a great example of the superiority of the Norwegian in rare, atypical
positions, when his unique positional feeling together with an extended knowledge of the classics
allow him to grasp the essence of the position.
25.Rhe1, anticipating ...f7-f6, was probably stronger, with a tiny plus for White.

25...Kc7 26.Bg2 Ng5 27.Rhf1 f6 28.Kc2 fxe5 29.dxe5 Nc8 30.c5 Ne7 31.b4 Nd5 32.Bxd5 cxd5
33.b5 axb5 34.Nxb5+ Kc6 35.Nd6 Nf3 36.b4 Ra8 37.Ra1 Rxa1 38.Rxa1 Nxe5 39.Ra7 Rb8 40.Ra3
b6 41.Ra7 bxc5 42.Ra6+ Kc7 43.bxc5 Nd7 44.Ra7+ Kc6 45.g5 Nxc5 46.Nf7 d4 47.Ne5+ Kd5
48.Nd7 d3+ 49.Kc1 Nxd7 50.Rxd7+ Ke4 0-1

Game 118 Sicilian Defence

Anatoly Karpov 2720
Garry Kasparov 2700

474
 Moscow 1985 (16)

1.e4 c5 2.Nf3 e6 3.d4 cxd4 4.Nxd4 Nc6 5.Nb5 d6 6.c4 Nf6 7.N1c3 a6 8.Na3 d5 9.cxd5 exd5
10.exd5 Nb4 11.Be2 Bc5 12.0-0

12.Be3! Bxe3 13.Qa4+ Bd7 14.Qxb4 is the refutation of the sacrifice, which was found some months
later, in the absence of computers at the time.

12...0-0 13.Bf3 Bf5 14.Bg5 Re8 15.Qd2 b5 16.Rad1 Nd3

Here is the famous octopus that won the game in glorious fashion.

17.Nab1 h6 18.Bh4 b4 19.Na4 Bd6 20.Bg3 Rc8 21.b3 g5

475
With all the pieces on the board, the d3-knight really is a monster, worthy of an appearance in a
James Bond movie! I considered this game to be the greatest achievement of Kasparov’s career, but
alas, Iosip Dorfman, a second of Garry at the time, declared recently that up to this point at least,
everything was preparation!

22.Bxd6 Qxd6 23.g3 Nd7 24.Bg2 Qf6 25.a3 a5 26.axb4 axb4 27.Qa2 Bg6 28.d6 g4 29.Qd2 Kg7
30.f3 Qxd6 31.fxg4 Qd4+ 32.Kh1 Nf6 33.Rf4 Ne4 34.Qxd3 Nf2+ 35.Rxf2 Bxd3 36.Rfd2 Qe3
37.Rxd3 Rc1 38.Nb2 Qf2 39.Nd2 Rxd1+ 40.Nxd1 Re1+ 0-1

Game 119 Sicilian Defence

Magnus Carlsen 2851
Nils Grandelius 2649
Stavanger 2016 (3)

1.e4 c5 2.Nf3 Nf6 3.e5

When faced with a line of dubious reputation, Carlsen generally goes straight for the refutation.
Actually he knows everything! I remember that before his first match for the title against Anand,
Kasparov and Kramnik gave 50% chances to both contestants, while Carlsen was a huge Elo
favourite. They explained that the advantage the then World Champion had in opening preparation
was very significant in a World Championship match. Sveshnikov even explained that as long as
Carlsen had no real deep opening repertoire, he had no chance to become the World Champion. How
wrong they were! In fact, Carlsen has always been an opening connoisseur. He is said to read
basically all the books on the subject, and thanks to his erudition and memory he knows more than
any player before him. Unlike his predecessors, he generally refuses to engage in long battles with
various classical lines, like we often saw in the Karpov-Kasparov matches (Queen’s Gambit,
Grünfeld, Spanish, etc.) or Kramnik’s and Anand’s matches. As he feels that he is stronger than his

476
opponents, he usually prefers to avoid purely theoretical debates. He is always able to improvise, like
changing his intended first move at the last moment, if he has the feeling that the opponent is well-
prepared for his previously-chosen move. In any case, he takes all the decisions himself, like he
always did, except for some months’ collaboration with Kasparov.

3...Nd5 4.Nc3 Nxc3 5.dxc3 Nc6 6.Bf4 Qb6

7.Qc1!?

A typical decision from Carlsen, played after some thinking. This move loses time, but it keeps the
opponent guessing what White’s plan will be, and enables the first player to keep all his options open.
7.Rb1 makes queenside castling impossible; 7.b3 weakens the pawn structure.

7...f6

This is the best move according to the engines. Black wants to grab the centre and threatens to play
...g7-g5-g4, winning the e5-pawn. Moving the f-pawn so early with the king in the centre is rarely
great, but Black emphasizes the fact that the white queen is not on light squares anymore, and so a
check on h5, for example, is not an issue.

8.Bc4!?

Faced with an unknown position, Carlsen typically plays for an attack and sacrifices a pawn ‘on
sight’. For a long time, Carlsen was called a technician, playing equal positions and pushing them, to
win with his superior technique. He has mainly been compared with Karpov, and Karpov himself
compared him to Petrosian. This last move and the kind of chess he plays in this game strictly have
no relation with any game by Karpov or Petrosian. This is Tal-like chess: intuitive attacking decisions

477
and an unclear sacrifice.
Strangely, no one has ever made any connection between Tal and Carlsen yet, except Smyslov once.
It seems to me that Carlsen sometimes enjoys the kind of ‘jazzy’ chess that is typical of the Latvian
player, unlike Kasparov or Topalov who are always pressing, for example. Unforced sacrifices, a
change of rhythm in the game (a slow move, 7.Qc1, then a sacrifice intended to gain some tempi:
Bc4), alternating positional moves and tactical ones – Karpov sometimes played this kind of chess,
but only on a positional basis.

8...g5 9.Bg3 g4

10.exf6!

Now comes the real sacrifice, a purely intuitive one, because the consequences are hardly fathomable
for a human being. The evaluation is even too difficult for today’s (2021) average computer, by the
way. This is the kind of chess that pure technicians don’t like. Fischer would not have played like this
either (the slow 7.Qc1 to begin with is not a Fischer move at all). But Carlsen loves this and is known
for his ability to win small advantage/no-risk positions. He reminds me of some secret agents in the
movies. Nice suit, nice tie and delicate manners, but always ready for a street fight!

10...gxf3 11.Qf4 fxg2 12.Rg1 Na5 13.f7+ Kd8

478
14.Bd5!

Even in the middle of tactical complications, Carlsen is stylish! This is a typical move of domination
in bishop against knight studies. The rest of the game is perfectly played by the Norwegian champion.
Soon, Black has to give up his material advantage, and with all of White’s pieces in action while part
of the black army is undeveloped, the issue is no longer in doubt.

14...Bh6 15.Qe5 Rf8 16.Bh4 Rxf7 17.Bxf7 Nc6 18.Qg3 Qxb2 19.Rd1 Qxc2 20.Bd5 Qf5 21.Rxg2
Bf4 22.Qf3 Kc7 23.Rg5 Qf8 24.Bg3 e5 25.Rh5 a5 26.Rxh7 Ra6 27.Rf7 Qe8 28.Kf1 Bxg3 29.hxg3
Qh8 30.Kg2 Nd8 31.Rf8 Qg7 32.Rh1 Rh6 33.Rxh6 Qxh6 34.Qf6 Qxf6 35.Rxf6 d6 36.Kf3 b5 37.g4
Kd7 38.Rh6 1-0

Game 120
Magnus Carlsen 2851
Pentala Harikrishna 2763
Stavanger 2016 (1)

479
 White to move

15.Bxc5!

On 15.b4, Black might have inserted the in-between move 15...Bb5 or played 15...Bxe2!? 16.Re1
Qb5.

15...Qa5

15...Qxc5 16.Qxa6+–.

16.Qc2 Bxe2 17.Qxe2 Qxc5 18.Rac1 Qb6 19.b4 h6

480
20.Qe3!?

This move is an echo of the famous Botvinnik game (see Game 121), where he exchanges queens
while allowing doubled pawns on the same e3-square. The idea is similar: exchange Black’s most
active piece, to enjoy an ending with powerful pieces against passive ones.

20...Qb7

The main line is 20...Qxe3 21.fxe3 Ng4. Here, the most energetic is 22.Nd4 (22.Bh3) 22...Ne7
23.Bh3±. Now, if 23...Rxc1?! 24.Rxc1 Nxe3?, then 25.Re1 wins a knight.

21.Bh3 Re8 22.Qc3 Ne7 23.Nd4 Ne4 24.Qc7 Qa6 25.f3 Ng5 26.Bd7 Red8 27.h4 Nxf3+ 28.Nxf3
Qxa3 29.Kg2 Qb2+ 30.Rd2 Qxb4 31.Re1 a5 32.Rde2 Ng6 33.h5 Nh8 34.Bf5 a4 35.Ne5 Qd6
36.Qc2 Re8 37.Bh7+ Kf8 38.Qf5 Re7 39.Bg6 Kg8 40.Nxf7 Rxf7 41.Bxf7+

Black resigned.

Game 121
Mikhail Botvinnik
Nikolay Sorokin
Moscow 1931 (2)

481
 White to move

20.Qe3!

This is Botvinnik’s brilliant concept that we referred to in the previous game!

The computers like the tactical solution 20.Rc2 even more.

20...Qxe3 21.fxe3 Bg4 22.a5 Nc8 23.Rc1 Bxf3 24.gxf3 Ne7 25.Nd5 Nc6 26.Nxf6+ gxf6 27.Rd7
Rab8 28.Kf2 Nxa5 29.Rcc7

Black has a pawn advantage in this position, but White’s superiority is estimated as +6 by Stockfish
12!

... 1-0 (55)

Game 122 Queen’s Gambit Declined

Magnus Carlsen 2815
Hikaru Nakamura 2774
Medias 2011 (1)

1.d4 d5 2.c4 e6 3.Nc3 Be7 4.cxd5

A favourite opening scheme of the Norwegian. It is both a calm position and one that offers strategic
complications with the asymmetric pawn structure. He particularly favours the f2-f3 and e2-e4 plan,
which is also a favourite of Kasparov (‘He was my part-time trainer,’ said Carlsen).

4...exd5 5.Bf4 c6 6.Qc2 Bg4 7.e3 Bh5 8.Bd3 Bg6

482
9.Bxg6!

Doubling Black’s g-pawns. This structure has advantages, as the h7-pawn is ‘promoted’ to the g-file,
a more central square, and the h-file is opened. The disadvantage of the structure is that there may be
some problems with the king’s security should he castle kingside: an attack with Ng5/Qh4 is
sometimes dangerous (...h7-h6 is not possible), and the g6-pawn provides a hook for an h4-h5 pawn
break, enabling White to open files on the kingside.

9...hxg6 10.0-0-0 Nf6 11.f3 Nbd7 12.Nge2 b5 13.e4 b4 14.Na4 dxe4 15.fxe4 Qa5 16.Kb1 0-0

483
17.h4!

Exemplary pawn play by Carlsen, who is ready to push e4-e5 followed by h4-h5, as soon as the f6-
knight has been chased from its defensive post.

17...Rfe8 18.e5 Nd5 19.h5 g5 20.h6 g6 21.Bc1

21.e6!.

21...N7b6 22.Nc5 Bxc5 23.dxc5 b3 24.Qxb3 Qxc5 25.Nd4 Rxe5 26.Nf3 Re2 27.Nxg5 Qe7 28.Qd3
Rf8

29.Rdf1!

A nice move, bringing the last piece into the attack, with the threat of taking on f7.

29...f5

484
30.g4!

Another break, as soon as the new hook on f5 has been created.

30...Na4 31.Qd4 Qe5 32.Qxe5 Rxe5 33.gxf5 gxf5 34.Nf3 Re7 35.Rfg1+ Kh7

36.Rg7+! Kh8

36...Rxg7 37.hxg7+ Kxg7 38.Bh6+ is winning for White.

485
37.Rhg1 Rfe8 38.Nh4 Rxg7 1-0

39.Rxg7 followed by 40.Ng6+ wins.

Game 123
Magnus Carlsen 2821
Fabiano Caruana 2711
Biel 2011 (4)

White to move

19.Nd4!

This is a brilliant sacrifice that promises White an initiative. He has at least adequate compensation
and probably an advantage. Carlsen has managed to disentangle his pieces, and from an equal
position without ideally-placed pieces, has gone to an active one.

19...Bxf1 20.Kxf1 Nb6 21.Nxc6 Rfe8

486
22.a4!±

This is a typical move against the infamous knight on b6, which is ‘always badly placed’ according to
Tarrasch. The advance of the a-pawn is also played to fix the black a-pawn that is isolated and might
be weak. Should this a-pawn fall, then the white a-pawn would be a protected passed pawn, an asset
that would win by itself in most endings.

22...Kf8

22...Nxa4 23.Ne7+.

23.a5 Nc4 24.Bc1

24.Bc3 was better.

24...a6 25.f4?!

25.c3.

487
25...Re6

Caruana had a beautiful saving resource that the players missed, a typical ‘invisible move’:
25...Ne3+!! 26.Rxe3 Bxf4 27.Rf3 Bxc1 28.Bd5∞.

26.Bd5 Rf6 27.Re4 1-0

Game 124
Magnus Carlsen 2826
Hikaru Nakamura 2758
London 2011 (3)

488
 White to move

29.Bd1!

An original place for the bishop.

29...Rfb8

29...h5 30.h4 Qh6 was holding.

30.h4 Qg6

489
31.Rxf6!

A lethal destruction of the black king’s protection. The disappearance of the g-pawn makes f5 and h5
very weak, so that the attack is winning, and Carlsen plays it clinically.

31...gxf6 32.Qf4 Rb2 33.Bh5 Qg7 34.Bf3 Ra8

34...Qg6 35.d5 Bc8 36.Qxd6+–.

35.d5 Bc8 36.Nh5 Qf8 37.Nxf6+ Kh8 38.Rc1 Kg7 39.e5 dxe5 40.Nh5+ Kh7 41.Be4+

Black resigned.
When the king goes to a dark square, the queen will win the rook by taking the e-pawn with check.

490
 Game 125
Magnus Carlsen 2835
Veselin Topalov 2770
Wijk aan Zee 2012 (11)

White to move

19.Nh4!?

Carlsen goes wild at times. The game will become completely irrational. 19.Rc1 was equal.

19...Nc4 20.Nf5 Nxb2 21.Bh4

White gives up everything for the attack. According to the computers, this is a blunder, but Carlsen
decided to go all-in some moves ago, and the defence is not easy in these positions.

21...Nxd1 22.Bxf6 Nxc3 23.Qg4 Bxe4 24.Nxh6+ Kh7 25.Bxf7

491
25...Qxf7

This queen sacrifice was not necessary. White’s attack is not as threatening as it seems, in spite of the
numerous attackers and the open position of the king. There is no checkmate on the h-file, for
example, thanks to the defence by the light-squared bishop that covers h7. So a steady move like
25...Qa7, intending ...Rc6, or 25...Kh8, intending ...Bh7, was better. The queen sacrifice equalizes the
chances, but from here Carlsen plays excellently and wins a complicated ending.
25...Qa7 26.Re1 Rc6–+; 25...Kh8 26.Qh5 Bh7 27.Bxe8 gxf6–+; 25...Bd3–+.

26.Nxf7 gxf6 27.f4 1-0

When do you look for a forced win in chess? In the following case, White has mobilized all his pieces
against the king as actively as possible. When you come to this point, you have to sacrifice; it’s as
simple as that!

Game 126
Magnus Carlsen 2837
Wang Hao 2739
Biel 2012 (2)

492
 White to move

25.Rxg7! Qxg7

On 25...Rxg7, 26.Qxh6+ wins quickly after 26...Kg8 27.Bh7+! Kh8 28.Bxf6.

26.Rxg7 Kxg7 27.Qf4 Nd3+ 28.Bxd3 cxd3 29.Kd2 Kg6 30.Kxd3 Re6 31.h4 Rfe7 32.h5+ Kf7
33.Qf5 Re5 34.Qg6+ Ke6 35.f4 1-0

Game 127
Magnus Carlsen 2843
Francisco Vallejo Pons 2697
Sao Paulo/Bilbao 2012 (7)

493
 White to move

16.h4!

The Carlsen touch, though it was popularized by Bent Larsen (whose books were an important
inspiration for him – see Game 128), and imitated by AlphaZero and its cousins. No other top player
before him used the rook pawns so often as an offensive tool.

16...Ndf8

Nowadays, nobody makes any mention of moves like 16...f6, winning a piece – 17.Bxg6 fxg5
18.fxg5 Rf8 19.Qh5 is crushing.

17.Rae1 Qc7 18.Bxe7 Rxe7 19.Re5 f6 20.Rxe7 Qxe7

494
21.h5!

This is excellent, as all the white pieces will benefit from the line- and square-openings. Fischer in My
60 Memorable Games explained that the attack with h2-h4-h5 against the Dragon set-up was just
routine for him: ‘I had it down to a science: to pry open the KR-file, sac, sac... mate!’. For Magnus,
exchanging a minor piece on g6, then play h4-h5, to use the h-file and the light squares around the
black king, is another routine.

21...gxh5 22.Qxh5 Qf7 23.Qg4 g6 24.Nf5 Kh8 25.Nh4 f5 26.Qh3 Qe6 27.Kf2 Re8 28.Nf3 Re7
29.Ne5 Nd7 30.Rh1 Nxe5 31.dxe5 c5 32.b3 c4 33.bxc4 dxc4 34.Be2 g5 35.g3 Qb6+ 36.Kg2 Qe3
37.Kf1 Rf7 38.Qh5 Qxg3 39.Qxf7 Qxf4+ 40.Kg2 Qe4+ 41.Bf3 Qxc2+ 42.Kg3 f4+ 1-0

Game 128
Bent Larsen
Svetozar Gligoric
Vinkovci 1970 (12)

495
 White to move

Since recent computers have shown that they like the rook pawn moves, many commentators have
been suggesting that humans learned this trick near the end of the 2010s, thanks to the all-seeing AI
computers. In fact, I believe that Carlsen was always a fan of such pawn moves, clearly influenced by
the games and books of another great Scandinavian player, the one and only Bent Larsen:
‘Polugaevsky has written that I like to push the rook pawns’ (see the chapter ‘About my style’, in
Bent Larsen’s Best Games, New in Chess 2014). This game is a perfect example of Larsen’s taste, as
he first introduces a pawn on h6 (the ultimate fashion, see, for example, the Carlsen-Mamedyarov
game), and then he opens the a-file, thus, as Larsen noticed, activating the queen’s rook without
moving it!

11.h4!? Qc7 12.Bd2 Rbd8 13.Qc1 d5 14.Bf4 Qc8 15.exd5 Nxd5 16.Bh6 Rfe8 17.Bxg7 Kxg7 18.h5
Nf6 19.h6+ Kg8 20.Qf4 Nh5 21.Qd2 f6

496
Now, White shows great efficiency in the exploitation of the second rook’s pawn advance:

22.a5! b5

23.a6!+– Ba8 24.Na5!

As the opening of the a-file has been avoided by Black, Larsen manages to open it anyway, thanks to
a pawn sacrifice that will release the power of the fianchettoed bishop:

24...e5 25.Qc3 Nd4

497
26.Nb7!

White has the edge and duly won.

... 1-0 (61)

Game 129 Sicilian Defence

Magnus Carlsen 2843
Viswanathan Anand 2780
Sao Paulo/Bilbao 2012 (9)

1.e4 c5 2.Nf3 d6 3.Bb5+ Bd7 4.Bxd7+ Qxd7 5.c4 Nf6 6.Nc3 g6 7.d4 cxd4 8.Nxd4 Bg7 9.f3

The old move-order allows a beautiful equalizing method that was discovered by Ivanchuk: 9.0-0 0-0
10.f3 Rc8! 11.b3 d5! 12.exd5 (12.Nxd5 e6) 12...Nxd5 13.Nxd5 e6.

9...Qc7 10.b3 Qa5 11.Bb2 Nc6 12.0-0 0-0 13.Nce2 Rfd8 14.Bc3 Qb6 15.Kh1 d5 16.Nxc6 bxc6
17.Qe1 Rdc8 18.e5 Ne8 19.e6 fxe6 20.Nf4 Bxc3 21.Qxc3 d4

498
22.Qd2!

The computer takes quite a long time to acknowledge this move as the best in the position (along with
22.Qe1). I don’t think it is a difficult move to find for Carlsen. You like the queen on dark squares,
you like the idea of letting the a1-rook participate, you like the idea of leaving d3 free for the knight;
all in all, a quite natural move for an instinctive positional player, because it is the only queen move
that does not ‘hurt’ another white piece. But why not play 22.Qe1, winning the pawn back with a nice
advantage? I suppose Carlsen had the beautiful 25th move manoeuvre in mind already.

22...c5 23.Rae1 Ng7

499
24.g4!

Taking away f5 from the knight.

24...Rc6

25.Nh3!

This is one of Magnus’ most famous backward moves. The g2-g4, Nh3, Qh6, Ng5 manoeuvre may
appear simple and logical once you have seen the game, but I doubt that many players would have

500
spotted it at the board, and at least it is clear that World Champion Anand had not.

25...Ne8 26.Qh6 Nf6 27.Ng5 d3

28.Re5!

With the threat of taking on h7, then 30.Qxg6+ and 31.Rh5.

28...Kh8 29.Rd1 Qa6 30.a4 1-0

Game 130
Magnus Carlsen 2848
Judit Polgar 2705
London 2012 (6)

501
 White to move

23.e5!

A great and, all the same, extremely risky move. It is difficult to imagine cautious players like
Petrosian or Karpov, even at their best, playing such a daring advance that basically ruins the pawn
structure, rendering the e-pawn both isolated and constantly weak. Yet, this was the best way to
proceed in the given position, because the e5-pawn and the open f-file give White a winning attack in
the long term.

23...Bc6 24.Bd4 Red8 25.Bxc6 Rxc6 26.Nf3 dxe5 27.fxe5 Rdc8 28.Ne4 1-0

Game 131
Michael Adams 2710
Magnus Carlsen 2848
London 2012 (5)

502
 Black to move

54...e3!

This is intended to get the oppo​sition, a decisive factor here.

55.Kf3

After 55.Kxe3, 55...Ke5 is winning.

55...Ke6! 56.Ke2 Kf6 57.Kf3 Kg5 58.Kxe3 Kxg4 59.Ke4 Kxh5 60.Kd5 g5 61.Kxc5 g4 62.Kd4 g3
63.Ke3 Kg4

White resigned.

Game 132
Magnus Carlsen 2861
Pentala Harikrishna 2698
Wijk aan Zee 2013 (4)

503
 White to move

21.Nc5

Interestingly, after the knight, the bishop, the rook and the queen will use this ‘strong’ or ‘weak’
square! This is a typical Carlsenian position, with two pawns sacrificed in search for better piece
activity. I don’t remember any other no. 1 player besides Topalov playing this kind of daring, unclear
sacrifice. Two pawns are given up for purely positional compensation without any clear follow-up,
but just on the general evaluation of the position. I can’t remember any game from Fischer like this,
for example; Kasparov did sacrifice, but most often that was opening preparation or using typical
attacking schemes. Here, the position is unclear and completely new. Hence, it requires calcu​la​tion
and a good evaluation of the position, also great confidence and even some courage.

21...Nd6 22.Qf3 Qe7 23.Rf2 Ra5 24.Nb3 Rb5

504
25.Bc5!

The bishop is well-placed on c5, and the knight might enjoy the excellent d4-square.

... 1-0 (49)

Game 133
Hou Yifan 2603
Magnus Carlsen 2861
Wijk aan Zee 2013 (9)

505
 Black to move

33...c3!

An excellent technical move. Black closes the c-file along which White might be able to attack the
black king, while at the same time providing an excellent and safe square for the dark-squared bishop.

34.bxc3 Bc5 0-1

Game 134
Magnus Carlsen 2861
Erwin l’Ami 2627
Wijk aan Zee 2013 (10)

506
 White to move

12.c4!?

A very un-Karpovian move (Fischer might have played it), looking for more dynamics at the cost of a
weakened structure and allowing Black to get a strong square for his knight. Other moves are also
interesting, like 12.h4 or 12.a4, when White is slightly better.

12...bxc4 13.Bxc4 Qb6 14.Qd2 Rd8 15.Rfd1 Nbd7 16.Rac1 Nd5 17.Bd6 N7f6 18.Bc5 Qb8 19.f3
h6 20.Ne2 Nd7 21.Ba3 e5 22.b3 Qb6 23.Nc3 N7f6 24.Bc5 Qc7 25.dxe5 Qxe5 26.Bd4 Qe7 27.Re1
Qd6 28.Qf2 Nxc3 29.Bxc3 Nd5 30.Be5 Qa3 31.h4 f6 32.Bd4 Kh7

507
33.Bxd5

Controversial but also typical: from a position of strength, Carlsen goes for the opposite-coloured
bishops endgame to win a pawn.

33...Rxd5

On 33...cxd5, 34.Rc7 is strong.

34.Rxc6 1-0

Game 135
Magnus Carlsen 2872
Alexander Grischuk 2764
London 2013 (4)

508
 White to move

13.c4!

Energetic and risky play. Carlsen ‘freezes’ Black’s damaged pawn structure. The strategy involves
some risk, as he allows his oppo​nent to get a good grip on the centre and possible attacking chances.
Here, we see an example of Carlsen taking material, and positional advantages, at the risk of coming
under a severe attack. He does not often allow this to happen, but chooses this option for this
important occasion.

13...Qd6 14.Be3 Nd7 15.d4 e4 16.Nfd2 a5 17.a4 f5 18.c5 Qg6

18...Qe6, keeping the c4-square under control, was better, with an unclear game, e.g. 19.Bf4 Nf6
20.Bxc7 Qf7 21.Be5 Be6. I suppose the Norwegian himself would not mind being Black nowadays.

19.Nc4 Nf6 20.Bf4 Nd5 21.Qd2 Be6 22.Nbxa5 Reb8 23.Ne5 Qf6 24.Bh2 Rxa5 25.Qxa5 Rxb2
26.Rab1 Ra2

509
27.Qa6!

A perfect defensive move: the queen controls e2 and is able to come to f1.

27...e3 28.fxe3 Qg5 29.Re2 Nxe3 30.Nf3 Qg6 31.Rxa2 Bxa2 32.Rb2 Bc4 33.Qa5 Bd5 34.Qe1 f4
35.Bxf4 Nc2 36.Qf2 Bxf3 37.Rxc2 1-0

Game 136
Peter Svidler 2747
Magnus Carlsen 2872
London 2013 (6)

510
 Black to move

25...exd3

Here, Carlsen misses a forced win:

25...Bxh3! :
 A) 26.dxe4 Rg5 27.g3™ Bg4! (Black can also play more slowly with 27...Qe6, intending ...Qg6.
His attack is very strong, and I guess the 2021 Magnus would have gone for it, e.g. 28.Bd3 Qg6 29.e5
Rxg3+ 30.fxg3 Qxg3+ 31.Kh1 Bg4) 28.f3

511
 analysis diagram

28...Rb2! 29.Qxb2 Bxf3!. If White moves the rook, 30...Qh3 or 30...Rxg3+ are deadly. The variation
is not obvious, but also not too long and relatively forced;
 B) 26.Qxe4 Re8 27.Qh4 Rg5 is crushing;
 C) After 26.gxh3 Qxh3, there is no defence against the arrival of the rook.

26.Bxd3 Bxd3 27.Rxd3 c5 28.Be5 Rxd3 29.Bxb8 c4

The main reason why Carlsen did not see the 25...Bxh3 sacrifice is that he was very confident here.
Black is clearly better and can improve his position, as White has no coordination or threats. I saw an
article on ChessBase India on this position, comparing it with a nice Fischer miss in analogous
circum​stances (one spectacular win is available, but a ‘technical’ move is played that is enough for a
large advantage, see Game 137). Maybe Carlsen’s advantage is less great, but he used to convert such
positions all the time. Later in his career, in World Championship contests, Karjakin and Caruana
proved tougher defenders than the older generation.

30.Be5

30.Qe5.

30...Bc5 31.Rb1 Qd5 32.Rb8+

32.Re1.

32...Kh7 33.Qh5 Qe4 34.Rb2 Rd5 35.Re2 Qb1+ 36.Kh2 f6 0-1

Game 137

512
 Bobby Fischer
Ismet Ibrahimoglu
Siegen ol 1970 (6)

White to move

31.Qd2

This ensures a large advantage, as White has the bishop pair, an excellent pawn structure and more
active pieces. Fischer’s move is good enough to keep an important edge.
However, 31.Bh6+! was a beautiful forced win and was not hard to calculate, as White’s attack is
quite standard and Black’s army is far from the action, with passive and fragile pieces like the knight
and the bishop which require protection by the queen.
31...Kxh6 32.Rg8.

513
 analysis diagram

Keeping the king from coming back home! Now White threatens a checkmate in two, beginning with
33.Qd2/c1+. There are not many defences here. Black can try to block the diagonal: queen or bishop
to e7, ...f7-f5 to block the check with the pawn, play ...g6-g5, or move the king to g5:
 A) 32...Qe7 33.Qd2+ (33.Bxc6 allows 33...Rb8±) 33...g5 34.h4 (34.Bxc6+–) 34...f6 35.hxg5+ fxg5
36.f3;
 B) 32...f5 33.exf5 gxf5 34.g4 Be7 35.Qd5;
 C) 32...g5 33.Qf3;
 D) 32...Kg5 33.Qf3 f5 34.Qxf5+ Kh6 35.Qf3 Be7 36.h4, etc.;
 E) 32...Be7 33.h4, and now Black can try 33...f5 34.exf5 gxf5 35.g4!, and White wins.

31...h5

31...Nd8 was better, and then Black will try to play ...Rb8, with fighting chances.

32.Bh6+ Kh7 33.Bg5 Rb8 34.Rxb8 Nxb8 35.Bf6 Nc6 36.Qd5 Na7 37.Be8 Kg8 38.Bxf7+ Qxf7
39.Qxd6 1-0

Game 138
Vladimir Kramnik 2810
Magnus Carlsen 2872
London ct 2013 (9)

514
 Black to move

25...Nd5!

Thanks to this active tactical defence, Carlsen saves the draw.

26.Nxd5 Bxd5 27.Qxc7

27.Rxd5 Qe1+ 28.Kg2 Qxe2+ 29.Qf2 Rxe5=.

27...Qc4 28.a3 f6 29.Qxc4 Bxc4 30.Bc3 Rxe2 31.Rd4 Bb5 32.Bb4 Re3 33.Kf2 Re2+ 34.Kg1 Re3
35.f4 Re2 36.Rd6 Rc2 37.g4 Bc6 38.Bd2 Bf3 39.h3 Ra2 40.Bb4 Rg2+ 41.Kf1 Rh2 ½-½

Game 139
Magnus Carlsen 2872
Boris Gelfand 2740
London ct 2013 (10)

515
 White to move

22.Rd3!

For some reason, such rook lifts are called ‘multi-functional’ by commentators nowadays; is it
because they’ve discovered that the rook can also move horizontally? Here, there is a clear threat of
winning a piece with a discovered attack (Nd5/Ne4), and White can push b2-b4 when the queen
moves. The funny thing is that the bishop that was carefully preserved in the opening Bb5, 0-0, Re1
and Bf1) is blocked in by the major pieces or the c-pawn throughout the game. That does not mean
that it is useless – far from it. It safely protects the king, and the control of the f1-a6 diagonal is
important, giving protection to the d3-square and the c4-pawn, and helping the b4-b5 push. Also, in
such configurations, Carlsen often plays f2-f3, to limit the scope of the enemy bishop, and h2-h4, to
avoid a back-rank checkmate. Thanks to the precious f1-bishop, he spares all those tempi, avoids
potential weakenings, and keeps more room for the elegant major-piece manoeuvres that made this
game a classic.

22...Qc2 23.b4 axb4 24.axb4 Nh5 25.Qe5 Bf6 26.Qxh5 Bxd4 27.Rxd4 Qxc3

516
28.Qa5! Rf8 29.Qb6 e5 30.Rd1 g6 31.b5 Be4 32.Qf6 h5 33.h4 Bf5 34.Rd5 Qc1 35.Qxe5 Be6
36.Rd4 Ra8 37.Qe2 Kh7 38.Rd1 Qc3 39.Qe4 Ra1 40.Rxa1 Qxa1 41.c5 Qc3 42.Qxb7 Qe1 43.b6
Bc4 44.Qf3 Qxf1+ 45.Kh2 Qb1 46.b7 Qb5 47.c6 Bd5 48.Qg3 1-0

Game 140
Magnus Carlsen 2872
Vasily Ivanchuk 2757
London ct 2013 (12)

517
 White to move

23.Kxd3?!

Accepting a ruined pawn structure, much like against Karjakin in 2016, when he accepted a weak
pawn on c4.
After 23.cxd3, the position would have been without much risk for White, and the computer thinks he
is even slightly better (23...Nxd4 24.cxd4 Rab8 25.Rb5 Kc6 26.a4). Carlsen, as a matter of principle,
did not want to allow his opponent a quick and easy draw. Such a way of thinking and playing was
rewarded many times, but here it is just bad judgement, because Black will constantly be better, and
Ivanchuk’s technique is one of the best ever. .

... 0-1 (90)

Game 141
Magnus Carlsen 2853
Sergey Karjakin 2769
New York Wch m 2016 (8)

White to move

24.bxc4

Deliberately weakening his pawn structure, just as in the Ivanchuk game, when c4 was also a key
square.
Once again, the desire not to allow the opponent to ‘escape’ with an easy draw prompted Carlsen to
take risks. I don’t see another World Champion who would have played like this. The move is not so

518
bad in itself. In fact, the computer estimates that it is nearly as good as the other ‘safe’ moves; it is
just that it does not give much chances for a win, while it provides long-term hopes for Black.
After this, Carlsen constantly rejected equalizing possibilities, and lost a crucial game in the end.
Both 24.Rxc4 and 24.Bxc4 were safe and fine. White had no grounds to play for an advantage here.

... 0-1 (52)

Game 142 Nimzo-Indian Defence

Teimour Radjabov 2793
Magnus Carlsen 2872
London ct 2013 (13)

1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.Qc2 d6

For what was the most important game of his career at the time, Carlsen chooses a nearly-unknown
line that probably promises White a slight edge, but obliges him to think for himself very early in the
game.
The tournament was already catastrophic for Radjabov, who at this point was only hoping to finish it
as soon as possible. It took him six years to get away from it, when he was able to achieve a glorious
comeback, winning the 2019 World Cup.

5.Nf3 Nbd7 6.g3 0-0 7.Bg2 e5 8.0-0 c6 9.Rd1 Re8 10.dxe5 dxe5 11.a3 Bxc3 12.Qxc3 Qe7 13.b4
Nb6 14.Be3 Ng4 15.Nd2 f5 16.h3 Nxe3 17.Qxe3 e4 18.Rac1 Be6 19.Qc3 Rad8 20.Bf1

519
20...c5

The simple repositioning with 20...Bf7, intending ...Bh5, ...Nb6-d7-e5 and preparing an attack against
the white king, was favourable for Black. But after his defeat against Ivanchuk the day before and the
loss of the leadership of the tournament to Kramnik, Carlsen was not at his best either. Nevertheless,
he fought admirably, even when there was apparently no hope.

... 0-1 (89)

Game 143
Magnus Carlsen 2872
Peter Svidler 2747
London ct 2013 (14)

520
 White to move

30.Bh4

The best move was 30.Bh8!. After 30...Qc6 31.Qe4 f6, the bishop looks trapped, but after 32.Qxc6
Bxc6, White has 33.Ng8! f5 34.Nf6 Nxf6 35.Bxf6, reaching a safe position with no risk of losing.
30.Bg5!? should hold: 30...f6 31.Qxc5+ Bd6 32.Qb6 fxg5 33.Bb5.

30...Qh3

521
31.f3

31.Bd5! was the best move, but the forced follow-up was not easy to foresee: 31...Bxd5 32.Qxc5+
Kg7 33.Qxd5 Kxh6 34.Qxf7 Rc8 (34...Ba5 35.Qxe8 Bxe1 36.Qxe5 Bxf2+! allows Black to make a
draw after 37.Kxf2 Qxh2+) 35.Be7 Qf5 36.Qxf5 gxf5 37.Rc1! (you have to see this move in advance
to enter the variation) 37...Kg6 38.Bd6.

31...Nf4 32.gxf4 Qxh4 33.Nxf7 Bxf3 34.Qf2 Qg4+ 35.Qg3

35.Kf1 Qxf4 36.Ng5! holds.

... 0-1 (48)

Game 144 Ruy Lopez

Sergey Karjakin 2767
Magnus Carlsen 2868
Stavanger 2013 (5)

1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.h3 Nb8

The Breyer Defence is typical of the style of the young Magnus. The main variation was for a long
time the Chigorin (9...Na5), but voluntarily misplacing a knight is not what a smooth positional
player dreams of. Then Karpov, the archetypical Breyer player, also promoted the complicated
9...Bb7 (the Zaitsev Variation, named after Karpov’s main second), but the long forced tactical lines
that most often follow are not to Carlsen’s taste either. This line also cost Karpov dearly in the 1990
World Championship Match with Kasparov. The Breyer Defence leads to slow, logical play where
Black is slightly passive, but should hold if he is capable of reading his opponent’s intentions. White
has a small advantage, and has many different and subtle plans at his disposal.
Of course, Fischer, Karpov and Carlsen love to be on the white side here, and go for the famous
‘Spanish torture’, with them being the torturer! However, the Breyer is a reliable defence, if one is
positionally inclined. The mature Carlsen does not go often for the Breyer for two reasons. With the
years, he is becoming slightly less patient, and now he often prefers more open positions. The Breyer,
while classified under the ‘Open Games’, may lead to the most completely closed game one can
imagine, together with some positions from the French Defence and the King’s Indian. Now an expert
on topical opening lines, his status as World Champion and no. 1 allows him to get great helpers and
to be more ‘offensive’ on the opening front, as his surprising (shocking at the time) choice of the
Sicilian against Caruana testified.

10.d4 Nbd7 11.Nbd2 Bb7 12.Bc2 Re8 13.a4 Bf8 14.Bd3 c6 15.Qc2 Rc8 16.axb5 axb5 17.b4 Qc7
18.Bb2 Ra8 19.Rad1 Nb6 20.c4 bxc4 21.Nxc4 Nxc4 22.Bxc4 h6 23.dxe5 dxe5 24.Bc3 Ba6 25.Bb3
c5 26.Qb2

522
Carlsen has been positionally outplayed, and now has to go into defensive mode. He faces a
fundamental choice between an active defence, which loses material, and a passive one, which wins
some. Faced with this dilemma, his answer nearly always is the active one.

26...c4!

26...Nd7 is the alternative. Now, it is White who is faced with a pleasant choice: 27.Rxd7 (or 27.b5
c4! (27...Bxb5 28.Bxf7++–) 28.Rxd7 Qxd7 29.Bxc4 Qc7 30.Nxe5 Rxe5 31.Bxe5 Qxc4 32.bxa6
Rxa6, and White is a pawn up) 27...Qxd7 28.Nxe5 Rxe5 29.Bxe5. White can follow up with Re1-e3-
g3, with an attack.

27.Ba4 Re6 28.Nxe5

523
28...Bb7!

This is again the active solution. As usual, Carlsen does not look at materialistic considerations, but
just installs his pieces on the most active squares, hoping for a counterattack. Objectively, Black is
probably lost after this move, but from a practical point of view, this is the best chance, and it even
allows winning chances, as the result will show!
After 28...Bxb4?! 29.Bxb4 (29.Qxb4 Rb6 30.Rd7 is also strong) 29...Qxe5 30.Bc3, White’s bishops
are too strong; 28...Rb8 is the ‘safe’ move: 29.Nd7 Nxd7 30.Bxd7 Re7 31.Ba4. Now Black is
passive, a pawn down, and the position is easy to play from White’s viewpoint because of the absence
of counterplay. In a practical game, I think this is a poor choice, and the position may be lost anyway!
Something we have understood in the computer age is that more often than not passive defence is
without hope, unless maybe Karjakin holds the black pieces...

29.Bc2

The normal reply; I guess most players would play like this, but now Black has sufficient counter​‐
play.
The slightly unnatural 29.Bb5! was the way to get a large advantage. This may lead to a forced line
that is very difficult to calculate, and over all, to evaluate: 29...Bxe4 (29...Ba6!? offers some
defensive resources) 30.Bxc4 Rxe5 31.Bxe5 Qxc4 32.Rd4 (32.Bxf6 gxf6 33.Rd4 Qc2!) 32...Qc2
33.Qxc2 Bxc2.

524
 analysis diagram

Apparently, Black has restored material equality, but it seems that he is lost after 34.b5!, as the threat
of 35.b6 is too strong. Here, there are still many tactical issues, like 34...Bc5 35.Rc4 Nd7 36.Rxc2
Nxe5.
Now, if 37.Rxe5 (only 37.b6! Bxb6 (37...Nf3+!? 38.gxf3 Bxb6 offers some hope, but White should
prevail), 38.Rxe5 should win. This ten-move calculation is maybe fathomable by a Karjakin in a
training session, but not often at this point of the game, with limited time on the clock and the
tiredness after preceding effort) 37...Ra1+ 38.Kh2 Bd6.

29...Rae8 30.f4 Bd6 31.Kh2

The computer claims that White holds after 31.Re3 Nh5 32.g3, but 32...g5∞ is unpleasant to face; or
first 31.g3.

31...Nh5! 32.g3 f6 33.Ng6

525
33...Nxf4! 34.Rxd6

Best; 34.gxf4 Bxf4+ 35.Kh1 (35.Nxf4 Qxf4+ 36.Kh1 Qg3! with ...Rxe4 to follow; there is no
defence) 35...Bg3 36.Re3 f5! is deadly, winning the g6-knight before continuing the attack.

34...Nxg6 35.Rxe6 Rxe6 36.Bd4

Here, the unfortunate position of the bishop on c2 allows Black to play for a mating attack:

36...f5! 37.e5 Nxe5! 38.Bxe5 Qc6 39.Rg1 Qd5 40.Bxf5

526
After 40.Bc3, 40...Re2+ is check​mate on the next move.

40...Rxe5 41.Bg4 h5 42.Bd1 c3 43.Qf2 Rf5 44.Qe3 Qf7 45.g4 Re5 46.Qd4 Qc7 0-1

Game 145
Magnus Carlsen 2868
Wang Hao 2743
Stavanger 2013 (8)

White to move

This looks like a tenable endgame, one pawn down but with an active rook, while the enemy king is
unable to join the battle for the time being.

56.Kd2

56.h5 was one way to reach a quite comfortable draw: 56...Rg1 (56...Rh1 57.Ke3 Rxh5 58.Kf4 Ke7
59.Ke5=; or 56...Re6 57.Ra5 Rd6 58.Kd4 Kf6 59.h6 Kg6 60.Ke5 Rd7 61.Ra6+=) 57.Kd4 Rxg3
58.Ke5 Rh3 59.Ra7+ Kf8 60.Kxf5; the logical 56.Kd4 directly should also hold.

56...Rg1 57.Ra3 Rg2+

The simple 57...Rf6 was winning, with ...d5-d4 to come and the activation of the king via e5, e4, etc.

58.Ke1 Kf6 59.Kf1 Rc2 60.Ra6+ Ke5 61.h5 Rh2 62.h6 d4 63.Kg1 Rh3

527
64.Kg2

There was a (very hard to find) road to heaven: 64.Kf2! d3 65.Ra5+ Kd4 66.Ra4+ Kc3 67.Ra3+ Kc2

analysis diagram

68.Ra2+!. White does not have the checking distance here, three ‘full’ squares between the pawn and
rook. But the rook can find an alternative defensive square: 68...Kb3 69.Ra6 Rh2+ 70.Kg1! Rh5
71.Kf2 d2 72.Ke2 Rh2+ 73.Kd1 Kc3 74.Ra3+! Kd4 75.h7 f4 76.h8=Q+ Rxh8 77.Kxd2=.

528
64...d3 65.Ra5+ Kd4 66.Ra4+ Kc3 67.Ra6 d2 68.Rc6+ Kd3 69.Rd6+ Kc2 70.Rc6+ Kd1 71.Rd6 f4
72.gxf4 Ke2 73.Re6+ Re3 74.Rxe3+ Kxe3 75.h7 d1=Q 76.h8=Q Qf3+ 77.Kg1 Qf2+ 78.Kh1 Qf1+
79.Kh2

79...g3+ 0-1

After 80.Kxg3 Qg1+ 81.Kh4, 81...Qh2+ wins the queen.

Game 146
Magnus Carlsen 2864
Fabiano Caruana 2774
Moscow 2013 (3)

529
 White to move

49.Rf8?

This is a strange mistake, and it seems that Carlsen had seen a forced draw in the variation that was
played in the game.
Giri joked about Carlsen knowing ‘too much’ about endings. This mistake in an innocuous position
reflects the downside of the style of the Norwegian, who likes, as Fischer did, to force the issue up to
a mathematical result, which is generally a strong point, as long as your calculations are correct!
The ‘club player’s move’ 49.Rb8, placing the rook behind the pawn and maintaining its offensive
value, was an easy draw in all variations. For example, after 49...Kg7 50.Kf5 b5 51.Rb7, White is not
even worse.

49...Rxh4 50.Rxf7+

This was Carlsen’s idea, but it loses by force.

50.Rb8 still gave chances for a draw, since if White takes the b-pawn against the g5-pawn, he has to
defend a two-pawns-down ending, but a theoretically drawish one! Here is a funny variation: 50...Rb4
(Black tries to keep his passed pawn; the theoretical draw goes like this: 50...Kg6. Just a try, as White
must play the only move here: 51.Rg8+! Kh7 52.Rb8 b5 53.Rxb5 Kg6 54.Rb8 Kxg5 55.Rg8+ Kh6
56.Kf5 Rh1 57.Rh8+ Kg7 58.Ra8, etc.) 51.Kf6 h4 52.Rb7! Rf4+ 53.Ke5 h3 (53...Rb4 54.Kf6=)
54.Kxf4 h2 55.Rxb6 h1=Q 56.Rh6+ Qxh6 57.gxh6 Kxh6 58.Kf5.

50...Kg6 51.Rf6+ Kxg5 52.Rxb6 Ra4

The white king is cut off horizon​tally, which is even stronger than a vertical cut. Thus it is a very

530
simple win for the then Italian player.

53.Rb8 Kg4 54.Rg8+ Kf3 55.Rf8+ Kg3 56.Rg8+ Kh2 57.Kf5 h4 58.Rb8 h3 59.Kg5 Re4 60.Kf5
Re2 61.Rg8 Rg2 62.Rd8 Rf2+ 0-1

Game 147
Magnus Carlsen 2864
Viswanathan Anand 2786
Moscow 2013 (5)

White to move

12.Bb4!

An original placement and exchange that Carlsen is very fond of (see Game 75 against Kamsky and
Game 174 against Mamedov). This is probably the best move.

12...Nf6

The idea after the principled 12...c5 is 13.dxc5 bxc5 14.Bc3+=; 12...Bxb4+ 13.axb4 Nf6 14.0-0 c6
15.Nf4+=.

13.0-0 Re8 14.Rc1 c6 15.Bxe7 Rxe7 16.Re1 Qd6 17.Nf4 Bc8 18.Qa4 Rc7

531
19.f3!±

Back to the Botvinnik plan.

This Rubinstein Nimzo-Indian is very close to Carlsen’s style, an asymmetric structure with chances
of interesting pawn play. It is strange that he did not play it more often. Perhaps the computers, or his
seconds, gave a negative verdict about this opening.

19...Be6 20.e4 dxe4 21.fxe4 Qd7

22.d5!

532
A brilliant combination that you will play only if you see and evaluate well the decisive 25th move!
22.Nxe6 is not bad, but offers Black more options, like the immediate 22...Qxe6 23.d5 Qe5±.

22...cxd5 23.Qxd7 Rxd7 24.Nxe6! fxe6 25.Bh3! Kh8

There is no defence: 25...Re8 26.exd5 Rdd8 27.Bxe6++–; 25...Kf7 26.exd5.

26.e5

26.Bxe6 Re7 (26...Rdd8 27.e5) 27.exd5.

26...Ng8 27.Bxe6 Rdd8

27...Re7 28.Bxd5 Rd8 29.Rcd1.

28.Rc7 d4 29.Bd7 1-0

This was the last game between these two before the World Cham​pion​ship Match, probably another
reason why Anand didn’t look opti​mistic before or during the match.

Game 148
Viswanathan Anand 2775
Magnus Carlsen 2870
Chennai Wch m 2013 (9)

Black to move

8...c4!?

533
Now Black has prospects on the queenside, with ...b7-b5-b4 either creating a passed pawn or
weakening the pawn centre with ...bxc3. But he must be ready to out-calculate his opponent, as the
kingside attack that will follow is extremely frightening. The live commentators estimated at the time
that Black was probably lost. Some people say that Carlsen’s weakness appears to be in calcula​ting
complex variations. Karpov spoke more accurately about a possible weakness in ‘chaotic’ chess.
Fischer’s weakness was said to be in ‘irrational’ games. It is possible, yet not easy, to create ‘chaos’
against a player who is exceptional at refuting attempts to confuse the game. As the game shows,
Carlsen himself is not afraid at all of confronting one of the best attacking players ever, the Madras
tiger, on his own turf! By the way, in the fourth game of the match, he deliberately took an a2-pawn,
not afraid to getting the bishop trapped like Bobby’s in the first game of the famous Fischer-Spassky
1972 match.

9.Ne2 Nc6 10.g4 0-0 11.Bg2 Na5 12.0-0 Nb3 13.Ra2 b5 14.Ng3 a5 15.g5 Ne8 16.e4 Nxc1
17.Qxc1 Ra6 18.e5 Nc7 19.f4 b4 20.axb4

20.f5.

20...axb4 21.Rxa6 Nxa6 22.f5 b3 23.Qf4 Nc7 24.f6 g6 25.Qh4 Ne8 26.Qh6 b2 27.Rf4 b1=Q+
28.Nf1

28.Bf1™ Qd1 29.Rh4 Qh5∞.

28...Qe1 0-1

Game 149
Viswanathan Anand 2775
Magnus Carlsen 2870
Chennai Wch m 2013 (4)

534
 Black to move

In this position, Black is fine and can play any standard move like ...Kb7, with good play. When
White plays f2-f4, intending to follow with g2-g4 and f4-f5, the standard move ...h6-h5 will be an
adequate defence. Nevertheless, as a matter of principle, Carlsen decided on:

18...Bxa2!

Most specialists remembered Fischer’s famous blunder in the first game against Spassky (World
Championship Match 1972, see Game 150). Thus, Carlsen’s bishop can escape, but the operation
gives White tempi to build a dangerous initiative. The Norwegian (rightly) felt that his move was the
best, so he entered the complications.

19.b3 c4 20.Ndc1 cxb3 21.cxb3 Bb1! ½-½

Game 150
Boris Spassky 2785
Bobby Fischer
Reykjavik Wch m 1972 (1)

535
 Black to move

29...Bxh2?

This is a bad blunder. Later, many esoteric reasons were given to justify it. I believe that Fischer, a
very logical man as far as chess was concerned, simply forgot a tactical detail in an important sub-
line. Otherwise, he would never have unnecessarily sacrificed a bishop, in a game that was heading
toward a draw, the result that he had played and hoped for during the game.

30.g3 h5 31.Ke2 h4 32.Kf3 Ke7

Probably, Fischer had missed the last move in the following variation: 32...h3 33.Kg4 Bg1 34.Kxh3
Bxf2 would be fine, if not for 35.Bd2!, trapping the bishop.

33.Kg2 1-0

Game 151
Magnus Carlsen 2881
Ding Liren 2836
Magnus Carlsen Chess Tour Finals 2020 (16.1)

536
 White to move

After 27.Rf3, White is clearly better.

27.Rxf7+ is the natural attacking move and the best one: 27...Kxf7 28.Qe6+ Kg7 29.Ne7. This is a
rather abstract move, as the threat is not immediately obvious. Of course, Carlsen would probably
have found it in a longer time control, after checking that his original idea was not good. In a rapid
time control, you rely more on intuition, and it is difficult to calculate moves that don’t come easily to
you, like Ne7. Carlsen’s intuition told him that he was winning, but with the clock ticking, you also
need intuition to find the right candidate moves to examine, because time is lacking if you want to do
it systematically (Magnus calculated the natural and aesthetic move 29.Qe7+ Kg8 30.Nxe5 dxe5
(30...Rh7!, mentioned by Carlsen, allows Black to make a draw after 31.Qe6+ Kh8 32.Nxg6+ Kg7
33.Qe7+ Kg8) 31.Bb3! Rh7 32.d6+ Kh8 33.Qxh7+ Kxh7 34.dxc7). 29...Qxc4 (the threat was
30.Rf1!, with checkmate to follow. 29...-- 30.Qxg6+ also forces mate: 30...Kf8 31.Rf1+ Kxe7
32.Qg7+ Kd8 33.Qxh8+ Ke7 34.Qg7+ Kd8 35.Qg8+ Nf8 36.Qxf8+ Kd7 37.Rf7#) 30.Qxg6+ Kf8.
After this, White has to find the only move: 31.Nf5!. Black could try to give back the queen with
31...Qe4+ (31...Rg8 allows checkmate with 32.Qxd6+ Kf7 33.Qe6+ Kf8 34.Qe7#) 32.Ka1 Qxf5
33.Qxf5+ Ke7. Material is equal, but 34.g6 wins easily.

... 1-0 (55)

Game 152 Sicilian Defence

Magnus Carlsen 2863
Viswanathan Anand 2792
Sochi Wch m 2014 (6)

537
1.e4 c5 2.Nf3 e6 3.d4 cxd4 4.Nxd4 a6 5.c4 Nf6 6.Nc3 Bb4

7.Qd3!

Here is one of those famous Qd3 moves in front of the f1-bishop!

7...Nc6 8.Nxc6 dxc6 9.Qxd8+ Kxd8 10.e5 Nd7 11.Bf4 Bxc3+ 12.bxc3 Kc7 13.h4 b6 14.h5 h6
15.0-0-0 Bb7

16.Rd3!?

538
This is a typical Carlsen move. The Norwegian has a soft spot for such piece arrangements: putting a
major piece, often the queen and sometimes the rook, in front of the f1-bishop and thus blocking it.
There were other moves, or plans, that are equally good according to the computers, but Carlsen’s
idea is brilliant: use the d3-square for the rook, slide on the third rank before using the square for the
bishop (as first playing the bishop to d3 would clearly impede the rook manoeuvre). Here, Carlsen’s
game is also typical of Larsen’s advice on how to develop a rook without moving it, thanks to the
push of the rook pawn. After Rd3-g3 and then Bd3 (very aesthetic!), White would continue with Rh1-
h4-g4, so Black plays ...g7-g6, and after hxg6, the h1-rook is ‘naturally’ placed on a semi-open file.

16...c5 17.Rg3 Rag8 18.Bd3 Nf8 19.Be3 g6 20.hxg6 Nxg6

21.Rh5!

An excellent technical move, with mainly defensive purposes, as e5 is attacked and 21...h5 was
threatened.

21...Bc6 22.Bc2 Kb7 23.Rg4 a5 24.Bd1 Rd8 25.Bc2 Rdg8

539
26.Kd2

An incredible mistake, losing a pawn and probably the game, while White was in command. The Kd2
idea is good, but not its execution with an unprotected rook on g4.

26...a4?

Played very quickly. Anand does not believe that Carlsen would allow him such an easy tactic, a
typical case of invisibility based on respect for the opponent. Probably Anand felt that he had been
outplayed, and had already ‘accepted’ his task of a defender in a bad position. 26...Nxe5 27.Rxg8
Nxc4+ was a huge advantage for Black.

27.Ke2 a3 28.f3 Rd8 29.Ke1 Rd7 30.Bc1 Ra8 31.Ke2 Ba4 32.Be4+ Bc6 33.Bxg6 fxg6 34.Rxg6
Ba4 35.Rxe6 Rd1 36.Bxa3 Ra1 37.Ke3 Bc2 38.Re7+ Ka6 39.Rxh6 Rxa2 40.Bxc5 1-0

Game 153
Jan-Krzysztof Duda 2757
Magnus Carlsen 2863
Stavanger 2020 (5)

540
 Black to move

21...h4

‘Black’s compensation is not of the kind that allows the position to be played so slowly,’ wrote l’Ami
in New in Chess 2020/7. This is the same kind of problem Karpov encountered at times against
Kasparov – see the crucial last Sicilian game that made Gary the new champion in 1985. The
positional player likes all his pieces to cooperate ideally before striking, while the dynamic, attacking
player strikes when possible. Those are different and nearly irreconcilable conceptions, although both
are good and well-founded. Here, Carlsen’s fundamental philosophy in the game does not apply to
the wild fight that is going on, and only a pure tactical move fits the position: 21...Ng4!. Now, Black
has adequate counterplay: 22.f3 Bxe4 23.fxe4 (23.Rxe4 Nf2=) 23...Be5.

... 1-0 (63)

Game 154
Alireza Firouzja 2728
Magnus Carlsen 2863
Stavanger 2020 (9)

541
 Black to move

Such minor-piece endings are usually drawn with only few pawns left. White needs to exchange one
or two more pawns, and then any piece sacrifice against the last enemy pawn will draw. The bishops
are of opposite colours, so Carlsen tries another ending:

52...Bc3!?

This is the great strength of Carlsen, and the huge difference compared with many a technician from
the past: he does not try to play the ‘best’ moves in such positions. Instead, he concentrates on
making the opponent’s life difficult, by forcing him to take decisions and to make actual choices
rather than just playing good moves and hoping that the opponent won’t, like Fischer or even
Kramnik did (does!) in such positions.
From a drawn bishop and knight endgame, Carlsen drags the game towards a drawn knight against
bishop endgame, and then to a drawn pawn endgame that he wins in the end. The exchange policy
forces the opponent to play another type of position, a different ending with other principles to take
care of, every time. This policy was very efficient against older players like Anand or Kramnik,
because they got tired at the end of the game session. But it is also efficient against a young player,
who by definition lacks experience in endings.

53.h5 Be1 54.Bxg6 Bxg3 55.Kxg3 Kg5 56.Kf3 Nb3 57.Bf7 Nd4+ 58.Kg3 Ne2+ 59.Kf3 Nf4
60.Kg3 Nxh5+

Asking another question: do you want to play the minor-piece or the pawn endgame?

61.Bxh5

542
61.Kf3 is also a draw. Here is a variation given by Erwin l’Ami in New in Chess: 61...Kh4 62.Be6
Nf4 63.Bc8 Nh3 64.Bf5 Ng5+. Now, even if White lets the enemy king approach playing 65.Ke3
(65.Kf2) 65...Kg3 66.Kd2 Kf4 67.Kd3,

analysis diagram

there is no danger, as when Black threatens ...Nf2 or ...Nc5, with the idea of taking on e4, White’s
king goes to c4 or d5, thus protecting the e-pawn again.

61...Kxh5 62.Kh3 Kh6

543
63.Kh4

The technical move 63.Kh2, with the distant opposition, was the alternative. The opposition is the
situation when the kings are separated by an odd number of squares (1, 3 or 5) and this can be either
vertical, horizontal or diagonal. The real opposition is the situation of the game (one king on h4, the
other one on h6), the other kind of opposition is a ‘mathematical consequence’ of the first one.
All those questions, like the theory of the corresponding squares, are mostly aimed at specialists in
rare endgames, but sometimes it also happens in real life...

63...Kg7 64.Kg3 Kf8 65.Kf2 Ke7 66.Ke2 Ke8 67.Ke3 Kd7 68.Kd3 Kd6

69.Kc3

Only 69.Kd2! draws – distant opposition!; 69.Kc4 Kc6–+; 69.Kc2 Kc6–+ (69...Ke7 also wins, as
Black invades the kingside via f7-g6-h5); 69.Ke2 Kc5 70.Ke3 Kc4–+.

69...Kc5

Now Black wins the e4-pawn by force, because he will enter on d4: 70.Kd3 Kb4 71.Ke3 Kc3
72.Kf3 Kd4–+.
White resigned.

Game 155
Magnus Carlsen 2862
Parham Maghsoodloo 2676
Speed Chess.com 2020 (1.13)

544
 White to move

14.Ne2!

The best move. The knight was badly placed on c3, and White gets perfect harmony with this move,
with the bishop aiming at the king and the knight free to go d4, f4 or g3.

14...Ng4

Capablanca preferred 14...Ne8 back in 1916, but White is still better.

15.h3 Ne5

545
16.Bxh7+

This move wins a pawn.

Not 16.Bxe5? Qxe5 17.Bxh7+ Kh8 18.0-0, as Black is much better after 18...g6.

16...Qxh7 17.Qxh7+ Kxh7 18.Bxe5 1-0

546
 Game 156
Magnus Carlsen 2863
Aryan Tari 2633
Stavanger 2020 (8)

White to move

19.Bxb5! Rxb5 20.Bxf6! Bxf6 21.Nd2

This is a textbook example of the good knight vs bad bishop duel.

21...Qc6 22.Qa4 Rfb8 23.Nc4 Be7 24.g3 Qc8 25.Qd1 g6 26.Kg2 Bf8 27.Qf3 Rb3 28.Rec1 Qe6
29.Ra7 R8b7 30.Rxb7 Rxb7 31.Ra1 h5 32.Ra8 Kg7 33.Ne3 Rc7 34.Nd5 Rc8 35.Ra7 Rb8

547
36.h3!

White will gain a decisive attack by playing g3-g4 next, just like in the Wojtaszek game.

36...Rd8 37.g4! hxg4 38.hxg4 Rd7 39.Ra8 f6 40.g5 f5 41.Qh3 Rf7

42.Re8!

An elegant finish!
There also was a logical way to exploit the h-file: 42.Ra1! Kg8 43.Rh1 Rg7 (or 43...Bg7 44.Qh7+

548
Kf8 45.Ra1+–) 44.Qh8+ Kf7 45.Ra1! – alternating the a- and h-files is a typical trick.

42...Qxe8

42...Qd7 43.Qh6+ Kg8 44.Qxg6+ Kh8 45.Nf6+–.

43.Qh6+! Kg8 44.Qxg6+ Kh8 45.Nf6 1-0

After 45...Rxf6 46.Qxe8, another piece will be lost.

Game 157
Aryan Tari 2633
Magnus Carlsen 2863
Stavanger 2020 (3)

549
 Black to move

26...Re1!

Deflecting the defender of the d4-knight. 26...Re4 is not as strong, but still much better for Black.

27.Rxe1

27.Qxe1 Qxb2#.

27...Bxd4 28.Qb4 Qc7 29.d6 Qc6 30.Qa5 Bxb2–+

30...Rxb2+!.

... 0-1 (45)

Game 158
Alireza Firouzja 2728
Magnus Carlsen 2863
St Louis rapid 2020 (7)

550
 Black to move

13...c5! 14.Ne2

After 14.Qxa5 bxa5 15.dxc5,

analysis diagram

White is a pawn up and has the bishop pair, while his opponent has useless doubled a-pawns. Never​‐
theless, this is a great advantage for Black! In fact, the c5-pawn will fall, after which the c4-pawn is

551
very weak, and the black rooks will quickly occupy the central files (c and d), using his advantage in
development. Here is a possible variation: 15...Nbd7 16.Ba3 Rac8 17.Kf2 Red8 18.Nh3 Nxc5
19.Be2 Nd3+ 20.Bxd3 Rxd3 21.Rhd1 Bxc4, and Black is much better.

14...Nc6 15.Kf2 Nb4 16.Bb2 cxd4 17.Nxd4

After 17.exd4, 17...Nxd3+ 18.Qxd3 Qb4 wins the c4-pawn, and also keeps a big positional
advantage.

17...Rad8 18.Bc3 Nxd3+ 19.Qxd3 Qc5 20.Qc2 Bxc4 21.Rhd1 Nd5 22.Bd2 e5 23.Nf5 Nf4 24.Kg3
Nd3 25.Qc3 Be6 26.Qxc5 Nxc5 0-1

Game 159 Sicilian Defence

Magnus Carlsen 2877
Radoslaw Wojtaszek 2735
Tromsø ol 2014 (4)

1.e4 c5 2.Nc3 d6 3.g3

‘I had a discussion with teammates before the game and I asked them if I should play something
harmless or something bad and they replied very clearly that I should play something bad.’ Here, you
have Carlsen’s appreciation of the Closed Sicilian!
‘I felt that Wojta is somebody who does not deal so well with these slow kingside attacks’ – a typical
Carlsen assessment, so that anyone can get the information. This habit was criticized by Grischuk.

3...Nc6 4.Bg2 g6 5.d3 Bg7 6.Be3 e5 7.Nh3

552
‘In going for this I was inspired by games of Spassky.’ Carlsen knows every player and every
important game of the past; recently (in 2020), he had just read a book about Chigorin and was trying
to reproduce his opening set-ups. Not many members of the top 10 have this culture. ‘I have to admit
that I still don’t completely understand the point... part of the reason why I played it is that it looks
cool.’ Compare this to another great contemporary player: ‘I used to study old games but then I
realized they’re getting older and the computers are getting stronger’ – Wesley So, 2018.

7...Nge7 8.f4 Nd4 9.0-0 0-0 10.Qd2 Bd7 11.Nd1 Qc8 12.Ndf2 Ndc6 13.c3 b5 14.fxe5 Nxe5
15.Bh6 N7c6 16.Bxg7 Kxg7 17.Nf4 Qd8 18.Rad1 Rc8 19.Qe2 h5 20.d4 cxd4 21.cxd4 Ng4 22.h3
Nxf2 23.Qxf2 Ne7

24.Rd3!

This is a simple move, once you have seen the game! According to the computers, this beautiful rook
lift, using the advantage in manoeuvring space, is also the best move. The idea is also linked with the
pawn sacrifice that will be played two moves later.

24...b4 25.Rf3 Qe8

553
26.g4!

Opening the h-file and giving more air to White’s pieces. This move is a pawn sacrifice and weakens
White’s king defences, but as Black is not able to attack, it’s no problem.

26...hxg4 27.hxg4 Bb5

If 27...Bxg4, 28.Rg3 Qd7 (28...f5 29.Ne6+) 29.Bh3 Bxh3 30.Nh5+.

28.Re1 Qd8

554
29.g5!

Making room for the bishop, the last passive piece. A player like Carlsen is not happy until all his
pieces are fully working, even in an apparently tense position like this one.

29...Qb6 30.Bh3 Rcd8

31.Be6!

This is stylistically perfect: the useless bishop delivers the decisive blow!

31...Be8 32.Nd5 Nxd5 33.Bxd5

Black resigned. The threat is 34.Rh3, followed by 35.Qf6. Black is completely crushed if he defends
passively, like 33...Rd7 34.Rh3 Rg8 35.Rh7+! Kf8 (35...Kxh7 36.Qh4+ Kg7 37.Qh6#) 36.Rf1 Qc7
37.Qf6 a5 38.Rh8 Rxh8 39.Qxh8+ Ke7 40.Qf6+ Kf8 41.Qxg6, etc.

Game 160
Magnus Carlsen 2872
Boris Gelfand 2777
Zurich 2014 (1)

555
 White to move

21.Ra5!

Vintage Carlsen, keeping the position under control with this technical rook move.
21.Ng5! was also interesting, but led to more complicated play: 21...Bxe5 22.Bd5+ Kg7 23.Ne6+
Rxe6 24.Bxe6 Bxb2 25.Ra5!.

... 1-0 (37)

Game 161 Réti Opening

Magnus Carlsen 2872
Viswanathan Anand 2773
Zurich blitz 2014 (4)

1.Nf3 d5 2.b3 c5

556
3.e4!?

This is a reversed Budapest Gambit!

3...dxe4 4.Ng5 Nf6 5.Nc3 Nc6 6.Bc4 e6 7.Bb2 Be7 8.0-0 0-0 9.Ncxe4 Nxe4 10.Nxe4 e5

11.f4!±

A classical double line opener.

557
11...exf4 12.Qh5

12.Rxf4 is better.

12...Nd4

12...Be6!.

13.Rxf4 g6 14.Qe5 b6 15.Raf1 Bf5 16.g4 Be6 17.Bxe6 fxe6 18.Rxf8+ Bxf8 19.Nf6+ Kh8 20.c3
Nc6

21.Ne8+ 1-0

This mating pattern was repeated in Kramnik-Aronian, Berlin ct 2018.

Game 162
Magnus Carlsen 2872
Fabiano Caruana 2782
Zurich 2014 (4)

558
 White to move

21.bxc4!

21.Rxf7? Qe8.

21...Qxc4

22.Rxf7!

559
A great sacrifice. After the following more or less forced moves, White gets two pawns for the
exchange and a continuing initiative, because of the wandering black king.

22...Bxf7 23.Rxf7 Rd7

23...dxe4 24.Qf5+ Kb8 25.Qe5+ Ka8 26.Qe7 is winning.

24.Rxd7 Kxd7 25.exd5+– 1-0

Game 163 Benoni Defence

Boris Gelfand 2761
Magnus Carlsen 2872
Zurich rapid 2014 (1)

1.d4 Nf6 2.c4 c5 3.d5 b5 4.cxb5 a6 5.bxa6 g6 6.Nc3 Bg7 7.e4 0-0 8.Nf3 Qa5 9.Bd3

9.Bd2 instead was safer.

9...Nxd5!

A typical trick in fianchetto-type positions, possible thanks to the wrongly-placed bishop on d3.

10.exd5 Bxc3+ 11.bxc3

Better was 11.Bd2.

11...Qxc3+ 12.Qd2 Qxa1 0-1

560
 Game 164
Magnus Carlsen 2843
Levon Aronian 2764
Stavanger 2018 (3)

White to move

23.g4!

Dominating the g7-knight, like in the famous Nh3 game against Anand. Thanks to the space
advantage, White’s major pieces are free to attack on the queenside or kingside, sliding on the third
and fourth ranks, while Black’s main pieces are entrenched on the eighth rank.

23...c6 24.c4 Ne8 25.Qf4 Kg7 26.Rb3 Rb8 27.Ng5 Nf6 28.Rf3 h6 29.Ne4 Nxe4 30.Qxf7+ Kh8
31.Qxg6 1-0

Game 165
Magnus Carlsen 2853
Sergey Karjakin 2769
New York 2016 (10)

561
 White to move

27.Nf1!

As soon as the pawn structure has changed, the knight is looking for new pastures.

27...Kf8 28.Nd2 Ke7 29.Re2 Kd6 30.Nf3 Raf8 31.Ng5 Re7 32.Rae1 Rfe8 33.Nf3 Nh8 34.d4 exd4
35.Nxd4 g6 36.Re3 Nf7 37.e5+ Kd7 38.Rf3 Nh6 39.Rf6 Rg7 40.b4 axb4 41.cxb4 Ng8 42.Rf3 Nh6
43.a5 Nf5 44.Nb3 Kc7

562
45.Nc5

The best plan was to prepare a5-a6 by doubling the rooks on the c-file: 45.Rc1 Kb8 46.a6 Rc7
47.Rfc3, with a winning position.

45...Kb8 46.Rb1 Ka7 47.Rd3 Rc7 48.Ra3 Nd4 49.Rd1 Nf5 50.Kh3 Nh6 51.f3 Rf7 52.Rd4 Nf5
53.Rd2 Rh7 54.Rb3 Ree7 55.Rdd3 Rh8 56.Rb1 Rhh7

57.b5!+– cxb5 58.Rxb5

Black is no longer able to cover b7 and e6 simultaneously.

58...d4 59.Rb6 Rc7 60.Nxe6 Rc3 61.Nf4 Rhc7 62.Nd5 Rxd3 63.Nxc7 Kb8 64.Nb5 Kc8 65.Rxg6
Rxf3 66.Kg2 Rb3 67.Nd6+ Nxd6 68.Rxd6 Re3 69.e6 Kc7 70.Rxd4 Rxe6 71.Rd5 Rh6 72.Kf3 Kb8
73.Kf4 Ka7 74.Kg5 Rh8 75.Kf6

Black resigned.

Game 166
Shakhriyar Mamedyarov 2760
Magnus Carlsen 2881
Shamkir 2014 (6)

563
 Black to move

27...Bxe4! 0-1

After 28.Bxe4 Rxd1 29.Qxd1 Nxe4 30.Qd8+ Kh7 31.Ng6, it appears that White has a checkmate
with the queen + knight duo, but so does Black, and he is the one to move!: 31...Qf2+ 32.Kh1
Nxg3#.

Game 167
Loek van Wely 2667
Magnus Carlsen 2862
Wijk aan Zee 2015 (4)

564
 Black to move

44...Nh2! 0-1

After 45.Kxh2, 45...Be5 wins the rook and the game.

Game 168 Sicilian Defence

Fabiano Caruana 2820
Magnus Carlsen 2862
Wijk aan Zee 2015 (6)

1.e4 c5 2.Nf3 Nc6 3.Bb5 g6 4.Bxc6 dxc6 5.d3 Bg7 6.h3 Nf6 7.Nc3 b6 8.Be3 e5?! 9.0-0

9.Nxe5!±.

9...0-0 10.a3 Qe7 11.Qb1

565
11...Nh5!

A classical way of meeting the attack on the queenside is to attack on the kingside. This is typically
Carlsen, counterattacking instead of defending, and more attracted by the king hunt than by some
queenside gains.
11...Nd7 would be the careful set-up against the queenside attack by White that was announced by
the queen move.
Against 11...a5, White can still play 12.b4 axb4 13.axb4 Rxa1 14.Qxa1 cxb4 15.Na4, with a
promising initiative.

12.b4 f5 13.bxc5 f4 14.Bd2 bxc5 15.Qb3+ Be6 16.Qa4

Here, the passive defence is the best choice, as the c6-pawn is essential – d5 has to be kept under
control:

566
16...Rac8! 17.Qa5 g5 18.Na4 g4 19.hxg4 Bxg4 20.Qxc5 Qf6 21.Nh2 f3 22.Nxg4 Qg6 23.Qe7 fxg2
24.Rfb1 Qxg4 25.Qg5 Qe2 26.Qe3 Qg4 27.Qg5 Qxg5 28.Bxg5 Nf4 29.Bxf4 exf4 30.Kxg2 f3+
31.Kf1

31...Rf4!

This is the best move, if not the only winning one.

32.c3 Rd8 33.d4

567
33...Bh6

This wins, as the king won’t be allowed to escape because d2 is controlled. 33...Bxd4! was even
stronger: 34.cxd4 Rh4 35.Ke1 (35.Kg1 Kh8 shows another advantage of the bishop sacrifice: the g-
file is opened and the check from g8 will be fatal for the white monarch) 35...Rxd4, with an
unavoidable checkmate.

34.Ke1

For example, it is not possible to defend the central pawn with 34.e5. After 34...Rh4 35.Kg1 Kh8,
checkmate follows.

34...Rxe4+ 35.Kd1 c5 36.Kc2 cxd4 37.Kd3 Re2 38.c4 Rxf2 39.Rd1 Re2 1-0

This game was the first between the two rivals after the historical triumph of Caruana in the 2014
Sinquefield Cup, when many people thought that the then Italian player could put Carlsen’s
domination in jeopardy. The result, and the way Carlsen played, showed that the Norwegian was not
ready to give up his crown yet. Only three years later, in 2018, ‘Fabi’ mounted another serious assault
on the world crown, and would defy Carlsen in a World Championship Match after a series of
startling results that took him to only two points short from first place on the Elo rating list. The
match was tense, close, and Carlsen won in the tiebreak, after 12 high-quality draws. At the end of
2020, Carlsen was asked about his next challenger, as the 2020 Candidates had been interrupted with
two leaders, Vachier-Lagrave and Nepomniachtchi. He said that both were the favourites, and maybe
also Caruana. ‘By far the best player among them is Caruana.’

Game 169
Magnus Carlsen 2862

568
 Hou Yifan 2673
Wijk aan Zee 2015 (7)

White to move

In this pawn structure, White must try to push c4-c5 (similarly to Anand-Carlsen, 6th game, World
Championship 2013). Carlsen will manoeuvre for 24 moves before being able to play it:
...

White to move

569
44.c5!+– ... 1-0 (54)

The main variation is 44...d5 45.Nd3 Rxe4 46.Rxh6+!.

Game 170 Dutch Defence

Viswanathan Anand 2797
Magnus Carlsen 2865
Baden-Baden 2015 (4)

1.d4 f5 2.g3 Nf6 3.Bg2 e6 4.c4 c6 5.Nf3 d5

The Stonewall is a surprising weapon in Carlsen’s arsenal. This anti-positional set-up was condemned
by Petrosian. Nevertheless, Carlsen has played it with great success against the very best players. It
seems that he likes pawn structures with ‘holes’, like the Botvinnik set-up in the English Opening
(c4/e4, with d4 being weak), that provide space in exchange for potential weaknesses.

6.0-0 Bd6 7.b3 Qe7 8.Ne5 0-0 9.Nd2

570
9...a5!?

This is ‘Magnus style’. The main move is 9...b6, with a fully playable game, but maybe the
Norwegian was reluctant to weaken the pawn structure. Keeping the b7/c6/d5 chain intact is more
solid for long-term prospects.

10.Bb2 Nbd7 11.Qc2 a4 12.Ndf3 Ne4 13.e3

13...a3!

571
Setting up the ‘potential passed pawn,’ as aptly described in the excellent book Techniques of
Positional Play (Bronznik and Terekhin, New in Chess 2013). The a-pawn is very close to the
queening square, if the a2-pawn happens to fall. In many endings, such a pawn is an important asset,
as the game will show!

14.Bc3 Nxe5 15.Nxe5 Bd7 16.Nxd7 Qxd7 17.c5 Bc7 18.b4

Now, some day, White might pick up the a-pawn and generate play along the b-file with b4-b5.

18...h5!

The Carlsen special!

So both rook pawns have been pushed, and as the centre is blocked, both pawns are the most
powerful assets in Black’s position. Since the famous AlphaZero computer arrived in 2018, every
commentator insists on the influence of the push of these pawns nowadays, but Carlsen was a fan of
these pushes long before the program had been conceived, so that we’d rather say that AlphaZero
plays ‘à la Carlsen’.
18...g5! is another typical pawn advance, with ...h7-h5 as a possible follow-up, or even the strategic
...g5-g4, to fix the pawn structure and thus make e4 a ‘strong’ square, as f2-f3 will no longer chase
the knight away, because of ...gxf3.

19.Be1 e5 20.dxe5 Bxe5 21.Rd1 Qe6 22.f3 Nf6 23.Bh3 g6 24.e4 dxe4 25.fxe4

572
This position will turn out to entail hair-raising complications, as Carlsen, instead of keeping it as
controlled as possible, takes a purely tactical approach in search of the win.

25...Bb2!

Carlsen sets up a beautiful mechanism: the bishop blocks the second file, so that the a-pawn will be
attacked by the queen; then, when the queen has taken the a2-pawn, the blocking bishop is
threatening a discovered attack on d4, winning the queen. This is a brilliant tactic: Black makes an
obstruction with the bishop and takes the pawn. It’s like self-pinning the bishop because of the
alignment of the queens, but at the same time, Black will threaten a discovered check on d4. I
described the whole mechanism (auto-pinning + discovered attack) in my book Tune Your Chess
Tactics Antenna under the name of ‘swing doors’, like those saloon doors that you push, and that
come back in your direction like a boomerang.
The more restrained 25...Ng4 is also good according to the computer.

26.exf5

Of course, it allows White to destroy the king’s pawn cover.

26...Qxa2 27.Bf2

Now, the desirable 27.fxg6 fails to 27...Bd4+.

27...g5 28.Rfe1 Qf7 29.Re6 Ng4

‘A daredevil move’ – Carlsen.

The computer slightly prefers 29...Kh8.

573
30.Bxg4 hxg4 31.Rg6+ Kh7 32.Rd7

32.Re6 should hold.

32...Qxd7 33.f6

33...Qd1+!

The less spectacular 33...Rxf6 was also good enough: 34.Rxf6+ Kg8 35.Rg6+ Bg7 and the a-pawn is
a winner.

34.Qxd1 Kxg6 35.Qd3+ Kh6 36.h4 gxh3 0-1

Game 171
Ian Nepomniachtchi 2778
Magnus Carlsen 2881
Skilling Open 2020 (16.1)

574
 Black to move

31...Bxf2 1-0

31...Qc7! is a beautiful double attack on the royal couple. The first threat is the discovered check on
the king. The second one is slightly more hidden: after 32.Kf3, Black traps the queen with 32...Rd8.

Game 172
Magnus Carlsen 2863
Shakhriyar Mamedyarov 2756
Shamkir 2015 (2)

575
 White to move

27.Bxg7+!

A liquidation that wins by force.

27...Kxg7 28.Nxd7 Qxd7 29.Qe5+ 1-0

29...Kg8 30.Rc7.

Game 173 Réti Opening

Magnus Carlsen 2863
Maxime Vachier-Lagrave 2765
Shamkir 2015 (5)

1.Nf3 Nf6 2.g3 b5 3.Bg2 Bb7 4.Na3!?

The kind of funny move that Carlsen is fond of!

4...a6 5.c4 b4 6.Nc2 e6 7.d4 a5 8.0-0 Be7

576
9.d5!

This is a thematic move when both bishops are on the long diagonal. If Black takes, there will be
some pins, while if the pawn is stabilized on d5, Black’s light-squared bishop is walled in. Radjabov,
a fianchetto specialist, remarked once that in such games, when two bishops are opposed on a long
diagonal, the king’s bishop has the advantage because it is protected.

9...Na6

After 9...exd5, White gets a nice version of the known gambit from the Queen’s Indian Defence after
10.Nh4 c6 11.cxd5 Nxd5 12.Nf5 0-0 13.e4 Nc7 14.e5, with a promising attack (Qg4 and Bh6 are on
the agenda) and a large advantage.

10.Nfd4 Nc5 11.Re1 0-0 12.e4 e5 13.Nf5 d6

577
14.Bg5!?

This original move (looking like a blunder, at first) shows a very Carlsenian approach to the position.
White is slightly better but looks for a bigger advantage, and is willing to give up some material to get
a tenser game. This is the move of an enterprising player, like Topalov for example. I don’t know of
any other ‘technician’ from the past – say Capablanca, Petrosian, Fischer, Karpov or Kramnik – who
would play like this, trying to ‘spice up’ an already favourable game. Players from the past would
continue to press, building on their space advan​tage. It is Carlsen’s belief that you have to make the
opponent’s life as difficult as possible at any given moment.
More standard was 14.Be3. If 14...Nfxe4?, 15.Qg4+– Ng5 16.Nxe7+ Qxe7 17.Bxg5.

14...Nxd5

14...Bc8 or 14...Re8 were more solid.

578
15.Bh6!

This is what White’s previous move was about!

15...gxh6 16.Qg4+ Bg5 17.cxd5 Kh8 18.h4 Bf6

19.Nce3!

19.Nxh6 was also fine, but Carlsen prefers to reinforce his position rather than taking a mere doubled
pawn.

579
19...Bc8 20.Qf3 Bg7

21.Bh3!

Again, first looking for the best possible squares for his pieces.

21...Rg8 22.Bg4 Qf6 23.Bh5 Bxf5 24.Nxf5 c6 25.dxc6 Rac8 26.Qd1 Rxc6 27.Qd5 Rgc8 28.Rad1
Bf8

29.Qxf7

580
A typical reaction, not fearing the potential ending with opposite-coloured bishops. Very possible and
also natural was 29.Bxf7, hoping to play with the queens on the board. Here, Black might try to enter
some complications with 29...Na4.

29...Qxf7 30.Bxf7 Na4 31.Re2 Rc1 32.Rxc1 Rxc1+ 33.Kg2 Nc5 34.b3 Rc3

35.Kh3!

This is the best move. Once again in this game, Carlsen’s obsession is the activation of all his pieces.

35...Nd7

This move is played to keep the white king from the g4-square, as ...Nf6+ is a valuable defence.

36.Be6 Nc5 37.Bd5 Nd7 38.Ne3 Nf6 39.Be6 Rc5 40.Nc4 Kg7 41.f3 Ne8 42.Rd2 Nc7 43.Bg4 a4
44.Nxd6 Bxd6 45.Rxd6 a3

A good try, as the a3-pawn is a dangerous weapon, being only two squares from queening if Black is
able to gobble up the a2-pawn.

46.Bd7 Rc2 47.Bc6 Rxa2 48.Rd7+ Kf6 49.Rxc7 Rc2 50.Rxh7 Kg6 51.Rc7 Kf6

51...a2 52.Ra7 Rxc6 53.Rxa2+–.

52.h5 Rc1 53.Rh7 a2

581
54.Bd5! 1-0

54...a1=Q 55.Rf7+ (even the sadistic 55.Kg4 wins; there is a nice attempt with 55...Rc7 56.Rxc7 Qa7,
as now 57.Rxa7 is stalemate, but the crude 57.Rf7+ Qxf7 58.Bxf7 leads to a winning pawn ending)
55...Kg5 56.Rf5#.

Game 174
Magnus Carlsen 2863
Rauf Mamedov 2658
Shamkir 2015 (9)

582
 White to move

16.a5!

This is a technical move, fixing b7 as a permanent weakness.

...

White to move

35.Qf7! 1-0

35.Rd7+–; 35.Qf7 Qe8 (35...f4 36.Qg6+ Kg8 37.Rd8+) 36.Qxf5+.

Game 175
Magnus Carlsen 2876
Alexander Grischuk 2781
Stavanger 2015 (5)

583
 White to move

41.Qxe8+! Bxe8 42.Bxf7 Bc6+ 43.Kf2 fxg3+ 44.Kxg3 Bd7

45.Bg6! 1-0

Black resigned after 45.Bg6, as the d-pawn wins the bishop. Then, the ending with the bishop and the
‘wrong’ rook pawn would be a draw... without the remaining black pawns. Here, the trick is to
stalemate the king, thus forcing the g-pawn to advance, enabling White’s h-pawn to win the game by
taking on g6: 45...Kg8 46.Kf4 Kf8 47.Ke5. The bishop and e7-pawn form a barrier against the

584
intrusion of the black king in the defence. Here is a possible line: 47...Kg8 48.Kd5 Kf8 49.Kc5 Kg8
50.Kb6 Kf8 51.Kc7 Bh3 52.d7 Bxd7 53.Kxd7 Kg8 54.Ke7 Kh8 55.Bf7 Kh7 56.Kf8 Kh8 57.Bg8
g5 58.hxg6 h5 59.g7#.

Game 176
Magnus Carlsen 2853
Wesley So 2779
St Louis 2015 (5)

White to move

22.Rhf1!

White prepares f3-f4, while keeping the enemy bishop restricted by the h4/g5-pawn barrier. 22.Rhg1,
to play h4-h5 and g5-g6, could be met by 22...Kh8 23.h5 h6, and Black holds.

22...Nxd5 23.Nxd5 Qb7 24.f4 f5 25.Qe3 e4 26.h5 Rc5 27.h6 g6 28.Qb3 Rf7 29.a4 Bd8 30.Rd4 Kf8
31.Rfd1 Rc6

585
32.Ne3!

This is a clever reorganization. As White’s major pieces are excellently placed, the knight looks for a
square that does not harm them. This knight retreat was also played in the 2019 game against Rapport
(see Game 104).

32...Bb6 33.Nc4

33.Rxd6+– was simpler.

33...Bxd4 34.Nxa5 Qb6 35.Nxc6 Bc5 36.Qd5 e3 37.a5 Qb5 38.Nd8! Ra7 39.Ne6+ Ke8 40.Nd4
Qxa5 41.Qg8+ Kd7 42.Qxh7+ Kc8 43.Qg8+ Kb7 44.c3 bxc3 45.Qb3+ Qb6 46.Qxb6+ Kxb6
47.bxc3 Bxd4 48.Rxd4 Kc6 49.Kc2 Ra2+ 50.Kd1 Rf2 51.Ke1 Kd7 52.Ra4 Ke6 53.Ra8 Rh2 54.c4
Kf7 55.Rb8 Ke6 56.Rg8 1-0

Game 177
Magnus Carlsen 2850
Dennis Wagner 2575
Berlin rapid 2015 (2)

586
 White to move

31.a4!

This is excellent positional play: as Black is passive, White can still make small adjustments for free.
The good old ‘accumulation of small advantages’ is still a strong tool. The a-pawn march once again
is characteristic of Carlsen’s style, but is also simply ‘good chess’ as Grischuk would put it. As Ne5
will always be available, and the knight is superior to the bishop, why not get the better of the
position before forcing the play?
By the way, this a-pawn move is a nice prophylactic against Black’s plan of a minority attack with
...b7-b5-b4 – see Carlsen-Kramnik, blitz 2019 (Game 209). It seems that Carlsen evaluates this
advance as an important strategic asset in any phase of the game, and this insight enlarges our
comprehension of the Carlsbad structure.
The computer’s original move is the direct 31.Ne5!? Bxe5 32.Rxe5. White will prepare c3-c4 and d4-
d5 (if ...b7-b5, then Qe2 and a2-a4 first), and that will win by force. But the win requires good
preparation and calculation. The latest computer engine prefers Carlsen’s move!

31...Bf6 32.a5 Qh7 33.a6 b6 34.Nb4 Qf7 35.Qd3 Kg7 36.Kf2 Rc8 37.Kf3 Rc4

587
38.Qg6+! Qxg6 39.hxg6 Bxd4 40.Rxe6 Rxe6 41.Rxe6 Bc5 42.Nxd5 Ra4 43.b4 Bf8 44.Rf6 1-0

Game 178
Richard Réti
Peter Romanovsky
Moscow 1925 (12)

White to move

White can use the position of the black bishop to reinforce his pawn structure:

588
31.a4!

This pawn is certainly much stronger on a4 than on a2, from where it could be attacked by the rook.
Réti plays the rest of the game beautifully.
The computer shows another move, which demands a lot of accurate calculation and evaluation:
31.f6! gxf6 (31...g6 is met with 32.g4 h6 33.h4 g5 34.hxg5 hxg5, and the f6-pawn is still difficult to
pick up; if now 31...Bb2, to avoid the move a2-a4, 32.Bxh7 g6. The bishop seems trapped, but the
computer does not worry: 33.Kg4! e4 34.d4 Bc3 35.Kg5 Bd2 36.Kf4! Rc6 37.Kxe4 Rxf6 38.dxc5
Kg7 39.c6) 32.Bf5! (32.Bxh7 Ke7). Now, the rook and king have gained manoeuvring space, and
White is clearly better.

31...Ke7 32.Bd5 Rc7 33.Rh4 h6 34.Ke4 Kf6 35.Rh5 Rd7 36.g4 g6 37.Rxh6 Kg5 38.Rh7 Kxg4

39.Be6! fxe6 40.fxg6

White is winning.
40.Rxd7?? gxf5.

40...Rd8 41.Rxa7 Kg5 42.g7 Kh6 43.a5 Kh7 44.a6 Rd6 45.h4 Be1 46.h5 Bh4 47.h6 1-0

Game 179
Francisco Vallejo Pons 2684
Magnus Carlsen 2850
Berlin Wch Rapid 2015 (3)

589
 Black to move

36...Bxh3! 37.gxh3 Qxh3+ 38.Ke1

38.Kg1 Qg3+ 39.Kh1 Qe1+.

38...Qg3+ 39.Kd2

39.Kf1 h3 40.Qc7 Qg2+! 41.Ke1 Qg6!–+. After the counterattack is taken care of, the h-pawn
queens.

39...h3

One of the problems of these pawn assaults against a castled king is that after the ritual sacrifice on h3
(classical in the King’s Indian Defence), i.e. when the king is able to flee, the h- or g-pawn will often
promote!

40.Qc7

Had the queen been on a4, White would move it to d7 here, with a double attack on the pawns on f7
and h3.

40...Qg6 41.Bc2 h2 42.e5 Qg2+ 43.Kc1 h1=Q+ 44.Kb2 Qh5 45.Qxd6+ Kg8 46.Qe7 Qxc2+
47.Kxc2 Rxe7 48.Ra1 0-1

Game 180
Shakhriyar Mamedyarov 2748

590
 Magnus Carlsen 2834
Doha 2015 (8)

Black to move

14...Ba5!

A nice manoeuvre that allows the bishop to attack d4. About the isolated pawn, Larsen said that while
before Karpov, one thought you had to block it, Karpov taught us to grab it!

15.Ng3 Bb6 16.Qe4 Bd7 17.h4 Be8 18.a3 Rd6

18...Qf6!? 19.Nh5 Qf5 20.Qxf5 exf5 21.d5 Ne7.

19.Qg4

19.d5 was the only move: 19...exd5 20.Rxd5 Qxe4 21.Nxe4 Rxd5 22.Bxd5 Rd8 and Black is very
slightly better, thanks to the bishop pair.

19...Rad8

Now the game is over.

20.d5 Qf8! 21.Qe4 Ne7 22.Bd3 f5! 23.Qe5 Nxd5 24.Bxf5 exf5 25.Nxf5 Rc6! 0-1

A masterclass on the theme of the isolated pawn.

Game 181

591
 Yu Yangyi 2736
Magnus Carlsen 2834
Doha 2015 (2)

Black to move

Here, you can see that the bishop and the knight are unprotected along the fourth rank.

15...Nxd3 16.Qxd3 Re4! 0-1

The double attack is the most important tactical theme in chess.

Game 182
Magnus Carlsen 2844
Evgeny Tomashevsky 2728
Wijk aan Zee 2016 (6)

592
 White to move

12.Rad1!

Black’s move 11...Ne7 signalled his intention to go 12...Ne4, the traditional equalizing method. Yet
Carlsen allows Black to play it.

12...Ng6

12...Ne4? 13.Nxe4 dxe4 14.Bxe4 Bxe4 15.dxc5!, with a big advantage for White. A rook on the
same file as the enemy king or queen may cause serious problems, no matter how many pieces are in
between!

593
On a broadcast hosted by GM Jon Ludvig Hammer, then rated around 2700, and another
commentator, who was not as highly rated as I remember, the second commenta​tor now suggested
taking on g6 with the bishop, and Hammer instantly replied that Magnus would never trade such a
strong bishop for the knight. As the co-commentator was getting ready to apologize for suggesting
such nonsense, Carlsen played:

13.Bxg6!

The important thing about exchanges, as Tarrasch explained a century ago, is what stays on the
board! After the exchanges on g6 and d6, the e5-square is available for White’s knight, and since the
move ...f7-f6 is very difficult to achieve because of the loose g6-pawn, it will ultimately be a fight
between a good knight and a bad bishop. Like many of Magnus’ decisions, this seems crystal-clear –
once you see him play it.

13...hxg6 14.Bxd6 Qxd6 15.Ne5 g5

594
16.f4!

This is a brilliant move, together with the next one. White has an edge, and could have proceeded in a
standard way, but this energetic continuation is far better, opening lines towards the black king
without losing time with constructive moves like knight or queen to f3.

16...gxf4 17.Rf1! Nd7

17...fxe3 18.Rxf6!+– gxf6 (18...exd2 19.Rxd2 gxf6 20.Qg4+ Kh7 21.Rf2) 19.Qg4+ Kh7 20.Qh4+
Kg7 21.Qg3+ Kh7 22.Rf1 e2 23.Rf4 e1=Q+ 24.Qxe1.

18.Qh5 Nf6 19.Qh4 Qd8 20.Rxf4 Ne4 21.Nxe4 Qxh4 22.Rxh4 dxe4

595
23.dxc5! bxc5

If 23...Rfd8, 24.Rxd8+ Rxd8 25.Rh8+ Kxh8 26.Nxf7+.

24.Rd7 Rab8 25.b3 a5 26.Rc7 a4 27.bxa4 Ba8 28.a5 Rb7 29.Rxc5 Ra7 30.Nc4 1-0

Game 183
Magnus Carlsen 2844
Hou Yifan 2673
Wijk aan Zee 2016 (11)

596
 Black to move

45...h5

45...a5! is the only move. It keeps White’s king from invading on the queenside, and Black should
hold the draw without trouble. Against 46.b4, the safest is to wait with a king move like 46...Ke6
47.bxa5 Kd7, and in case of 48.Kb4, Black has to be ready for the sacrifice of the a-pawn, and wait
with 48...Kc7 (48...Kc8) 49.a6 bxa6 50.Ka5 Kb7. A draw will result.

46.Kb4 Kc8 47.Ka5 Kc7 48.h4 Kb8 49.Kb6 Kc8 50.b4 Kb8 51.b5 cxb5 52.axb5 axb5 53.Kxb5
Kc7 54.c3 1-0

Game 184
Magnus Carlsen 2855
Levon Aronian 2792
Paris rapid 2016 (3)

White to move

49.Kf1!

A famous king march that is reminiscent of a classical game and plan in such positions (see Game
185). One of the theoretical bases of White’s plan is that once the king approaches the c-pawn, White
is easily winning after an exchange of queens. But the main thing is that since the queen and knight
paralyse the whole black army, the white king is free to play an active part, as a counterattack is
impossible due to the lack of manoeuvring space the second player suffers of.

597
49...Kh7

49...Ng8 50.Qd7+.

50.Ke2 Kg7 51.Kd3 Ng8 52.Ne8+ Kh8 53.Kc4+– h5 54.gxh5 Qh6 55.Qxc6 Qd2 56.hxg6 Qe2+
57.Kc5 Qxf2+ 58.Kb5 Qxg3 59.Qd7 Qxg6 60.Ka5 Qg3 61.b5 Qc3+ 62.Ka6 Qa3+ 63.Kb7 Qg3
64.b6 Qg6 65.Ka7 f5 66.exf5 Qg3 67.f6 Qa3+ 68.Kb8 1-0

Game 185
Lev Psakhis 2580
Mark Hebden
Chicago 1983 (6)

White to move

43.Kf1! Ba7 44.Ke2 Bb6 45.Kd3 Ba7 46.Kc4 Qc7+ 47.Kb3 Qe7 48.g4 Bb6 49.Kc4 Ba7 50.Kb5
Qe8+ 51.Bc6 Qd8 52.Kc4 Qe7 53.Qd7!

Exchanging the queens is the whole point of White’s plan. Now any ending is winning, due to the
activation of White’s king.

53...Qe6+ 54.Qxe6 fxe6 55.Rxf8 Kxf8 56.Kb5 Ke7 57.Ka6 Bxf2 58.c4 Kd8 59.Kb7 Be1 60.b5
Bf2 61.b6 Bd4 62.Ba4 d5 63.cxd5 exd5 64.exd5 e4 65.Kc6 Kc8 66.d6 e3 67.Bb5 Bf6 68.Ba6+
Kb8 69.Kd7 1-0

Game 186
Fabiano Caruana 2804

598
 Magnus Carlsen 2855
Paris rapid 2016 (4)

Black to move

The lack of protection of the bishop on g5 costs White dearly, as Black is able to snatch a central
pawn thanks to a standard combination:

13...Nxe4! 14.Bxe7 Nc5!

The point.

15.Bd5 Bxd5 16.Qxd5 Qxe7 0-1

599
 Game 187
Veselin Topalov 2761
Magnus Carlsen 2855
Paris rapid 2016 (6)

The comments by Carlsen are from his ‘Attacking without sacrificing’ lessons on a Chess 24 video
about the game.

Black to move

‘I have a typical trick here...,’ says Magnus. ‘There is no way that I can absorb the pressure that he
has on the queenside.’
It looks like White has a solid initiative on the queenside, but Black is able to solve his problems
thanks to one of Carlsen’s favourite counters, which he also used in the crucial – world title worth –
rapid game against Karjakin in the same year:

23...a5!

‘A standard counterattacking motif; you see it in pawn endgames all the time.’

24.Rb1

Now, the intended 24.axb5 loses after 24...axb4 25.Rxa8 Rxa8 26.Nb1 Nxe4.

24...axb4 25.Rxb4 bxa4

‘I manage to exchange the queen​side.’

600
26.Raxa4

‘It has come from pleasant for him to pleasant for me.’

26...Ng4 27.Nf1 h4 28.f3 hxg3 29.hxg3

29.fxg4 gxh2+ 30.Nxh2 Qxc3.

29...Nf6 30.Qg2

30.Qh2 Qd7.

30...Qh6 31.Qh2 Qg5 32.Qd2 Qh5

‘Here he played...’

33.g4

... ‘that I had foreseen in advance, and once I realized the strength of my reply, I became very, very
excited’.
Nielsen later commented that he was analysing the live game on a computer and was very surprised
when he saw the computer’s suggestion, but even more so when Carlsen quickly played it!

33...Qh8!

The point is to avoid a queen exchange and exploit the dark squares after ...Nf6-h7-g5 (or first ...Be7-
g5-f4, etc.). The idea behind this elegant retreat is to keep the queens on the board, as Black is now
able to counter a possible exchange proposal thanks to the interception with the knight on h7. This

601
also enables the bishop to go to g5 and possibly to f4, and then the kingside attack will be
unstoppable. This is one of the most famous of Carlsen’s moves, and very much in his style.
‘The rest of the game was just a pleasure.’

34.Nb5

The main idea is to keep the queens on against 34.Qh2, thanks to 34...Nh7.

34...Nh7 35.Qh2 Rfb8

‘In order to bring the queen to d8.’

36.c4 Qd8 37.Ne3 Bg5 38.Qe2 Bf4 39.Ng2 Bg3 40.Qe3 Ng5 41.Kf1

41...Qh8!

This is a charming switchback to finish the game.

White resigned as after 42.Qxg5 Qh1+ 43.Ke2 Qxg2+ 44.Kd3 Bf4, it’s all over.

Game 188
Anatoly Karpov
Wolfgang Unzicker
Nice ol 1974 (3)

602
 White to move

24.Ba7!

This is Karpov’s signature move. The idea, as in Carlsen’s game, is to keep pieces on the board – here
the rooks, without moving them. Thanks to this shield on a7, White is able to gain control of the a-file
by doubling rooks, and can, whenever he wishes, reopen the file by moving the bishop.

... 1-0 (44)

Game 189 Tarrasch Defence

Magnus Carlsen
Anish Giri
Chess 24 Paris 2020

1.d4 d5 2.c4 e6 3.Nc3 Nf6 4.cxd5 Nxd5 5.e4 Nxc3 6.bxc3 c5 7.Nf3 cxd4 8.cxd4 Bb4+ 9.Bd2
Bxd2+ 10.Qxd2 0-0 11.Bc4 Nd7

This was the original choice of the players when they first played the variation in 1937 (see Game
190).

12.0-0 b6 13.Rad1 Bb7 14.Rfe1 Rc8 15.Bb3 Re8 16.Re3 Nf6

603
17.d5!?

This is not the only idea, but it is the most thematic one.

17...exd5 18.e5! Ne4 19.Qe1!

This subtle idea had been prepared by White’s 16th move.

19...Qc7 20.Nd4 a6 21.h4

On 21.f3, Black can play 21...Nc3 22.Rc1 Qc5!∞.

21...Rcd8 22.f3 Nc5

22...Nc3 is no longer possible, as the rook does not control the c-file.

23.h5 Ne6 24.Nf5 d4 25.Red3 Nc5 26.Rxd4

White has a decisive edge.

26...Rxd4 27.Rxd4 Nxb3 28.Qg3 g6 29.axb3 Rd8

604
30.e6!

A beautiful touch that makes the difference between a good game and a brilliancy.

30...Qc1+

30...Qxg3 31.Rxd8#.

31.Kh2 Rxd4 32.e7! Qc8 33.Qe5 Rh4+ 34.Kg3 1-0

Game 190 Tarrasch Defence

Paul Keres
Reuben Fine
Ostend 1937

1.d4 d5 2.c4 e6 3.Nc3 Nf6 4.cxd5 Nxd5 5.e4 Nxc3 6.bxc3 c5 7.Nf3 cxd4 8.cxd4 Bb4+ 9.Bd2
Bxd2+ 10.Qxd2 0-0 11.Bc4 Nd7

11...Nc6 became the main move for a long time, before modern players and computers recently
redeemed Fine’s idea: 12.0-0 b6 13.Rad1 Bb7 14.Rfe1 Na5 15.Bd3 Rc8

605
 analysis diagram

16.d5 exd5 17.e5! Nc4 18.Qf4 Nb2 19.Bxh7+! Kxh7 20.Ng5+ Kg6 21.h4+– Rc4 22.h5+ Kh6
23.Nxf7+ Kh7 24.Qf5+ Kg8 25.e6 Qf6 26.Qxf6 gxf6 27.Rd2 Rc6 28.Rxb2 Re8 29.Nh6+ Kh7
30.Nf5 Rexe6 31.Rxe6 Rxe6 32.Rc2 Rc6 33.Re2 Bc8 34.Re7+ Kh8 35.Nh4 f5 36.Ng6+ Kg8 37.Rxa7
1-0 Polugaevsky-Tal, Moscow 1969.

12.0-0 b6 13.Rad1 Bb7 14.Rfe1 Rc8 15.Bb3 Nf6 16.Qf4 Qc7 17.Qh4 Rfd8 18.Re3 b5 19.Rde1 a5
20.a4 b4

606
21.d5! exd5 22.e5! Nd7

22...Ne4 23.e6! (23.Rxe4 dxe4 24.Ng5 h6–+) 23...fxe6 24.Rxe4 dxe4 25.Bxe6+ Kh8 26.Ng5 h6
27.Nf7+ Kh7 28.Ng5+=.

23.Ng5 Nf8

23...h6 may hold, though White can provoke complications with 24.e6!∞.

24.Nxh7! Nxh7 25.Rh3 Qc1 26.Qxh7+ Kf8 27.Rhe3 d4 28.Qh8+ Ke7 29.Qxg7 Rf8 30.Qf6+ Ke8
31.e6 1-0

Game 191 Ruy Lopez

Magnus Carlsen 2855
Wesley So 2770
Bilbao 2016 (4)

1.e4 e5 2.Nf3 Nc6 3.Bb5 Nf6 4.d3 Bc5

607
‘On my way to the game, walking around, I thought why not take, and play queen e2. And that’s what
I did.’ A typical Carlsen attitude! At any time and in any circumstances, he is able to follow his
inspiration, even trying an unstudied plan in a topical line. One reason why Carlsen is unpredictable is
that he sometimes changes his mind at the last minute.

5.Bxc6 dxc6 6.Qe2 Qe7 7.Nbd2 Bg4 8.h3 Bh5 9.a3 Nd7 10.b4 Bd6 11.Nc4 f6 12.Ne3 a5 13.Nf5
Qf8 14.bxa5 Rxa5 15.0-0 Qf7

16.a4!

608
This is a technical move and a gambit, as Black is able to gobble up this pawn. The idea is to keep
Black from installing a kind of blockade with ...Ra4, e.g. 16.Bd2 Ra4!² (16...Rxa3? 17.Nxd6+ cxd6
18.Rxa3).

16...Nc5

17.Qe1!?

Winning a tempo on the rook, to get rid of the pin and move the f3-knight to c4. A backward queen
move, followed by a backward knight replacement: this was too tempting for Carlsen!
The computer advocates 17.Bd2 Ra6 (17...Rxa4 18.g4 Bg6 19.d4 exd4 20.Rxa4 Nxa4 21.e5 fxe5
22.Nxe5+–) 18.a5±. The main point is, as usual, 18...0-0?! 19.Bh6!+–.

17...b6 18.Nd2! Rxa4 19.Nc4 Bf8

Best is 19...Be7 20.Be3 Kd7, when White is better but not winning as in the game.

20.Be3 Kd7 21.Qc3 Nxe4 22.Nxb6+ cxb6 23.dxe4 Qc4 24.Qd2+ Kc7 25.g4 Bg6 26.Rfd1 1-0

Game 192
Magnus Carlsen 2855
Sergey Karjakin 2773
Bilbao 2016 (3)

609
 White to move

21.Kh1!

A fine king move, preparing the attack with f2-f4 and evacuating the g-file, for reasons that will soon
become clear. This is both a prophylactic move and an attacking one, as it prepares the whole
attacking plan that follows. It also avoids a small trick: if 21.f4, 21...Rd4! 22.Qxd4? Nf3+.

21...Rac8 22.f4 gxf4 23.Bxf4 Qb6 24.Qh5 Nf6 25.Qf5 Qd8 26.Bb3 Rd4 27.Bxe5 dxe5 28.Rbd1
Qd7

610
29.Qf3!

As White is playing for an attack, he logically keeps the queens on.

29.Qxd7?! would have been a serious mistake, as the endgame is equal after 29...Nxd7.

29...Rb4 30.Rd2 Rf8

31.g4!

Simple chess: White plays for h3-h4 and g4-g5, exploiting the hook on h6. The rooks will slide to the
g-file, with an unstoppable attack.

31...a5 32.Rg2 Nh7 33.h4 Rb6 34.g5 Kh8 35.Rfg1 f5 36.Qh3 Rb4 37.gxh6 Bxh6 38.Qg3 Nf6
39.Qg6 Ng4 40.Rxg4 1-0

Game 193
Magnus Carlsen
Anish Giri
Bilbao 2016 (9)

611
 White to move

38.Nxf5! Qe6

If 38...Qxf5, 39.Qxd5+ wins.

39.Qg5

The attack is too strong now.

39...g6

39...Qxe5 40.Nh6+ Kf8 41.Qxe5.

40.Nh6+ Kf8 41.Ng4 Ke8 42.Nf6+ Kf7 43.Nxh7 Ra4 44.Qd8 Ra2+

44...Qxe5 45.Qf8+ Ke6 46.Qe8+ Kd6 47.Qxa4.

45.Kg1 1-0

Game 194
Magnus Carlsen 2832
Sergey Karjakin 2781
Stavanger 2017 (8)

612
 White to move

44.Rxg8+

This forces a vertical or horizontal alignment of the king and queen, allowing White to execute or
threaten a pin: 44...Kxg8 (44...Qxg8 45.Rh4 Ke7 46.Rh6 Qf8 47.Qf6+ Kd7 48.Rh7+–) 45.Rg4.

1-0

Game 195
Hikaru Nakamura 2829
Magnus Carlsen 2881
Magnus Carlsen Chess Tour 2020 (41.1)

613
 Black to move

29...Qa4! 0-1

There is a double threat against the unprotected rook and mate on d1.

Game 196
Magnus Carlsen 2863
Jeffery Xiong 2709
Clutch Chess Showdown 2020 (1.3)

614
 White to move

12.Qxd2!?

This is a typical Carlsen move. He loves to play an early queen move in front of an undeveloped
bishop. See his Qd3 moves and also Game 203 against Wojtaszek. Playing like this seems to lose
several tempi, but the plan that follows is clear cut: bring the queen to the kingside, and attack the
castled king.

12...Be7 13.Qf4 0-0 14.Qg4 Re8

15.h4!

This is the technical attacking move here. White brings another unit into the attack, and eliminates
future counterchances for the opponent based on the back-rank checkmate.

15...Qc7 16.Bh6 g6 17.Qf4 Qc2

615
18.g4!

Excellent prophylaxis in attack, as Black wanted to go 18...Qf5. The weakening of the castled
position is not important, because for lack of space, Black’s pieces are unable to reach the kingside.

18...b6 19.b3 Ba3

This move is played to enable some ...Re7 defence if f7 is under attack.

20.Kg2!

616
Another common technical attacking move: White liberates the first rank, allowing the rook to
occupy the h-file, which he will open thanks to the h4-h5 break.
The direct 20.h5 is also good, but slightly less Carlsenian. Both moves will probably be played, so
why not play the harmonious one first?

20...Qb2 21.Ng5 Re7 22.Rad1 Rc2 23.Re3 Rc3 24.h5 Qc2 25.Rxc3 Qxc3

26.Nxh7! Kxh7 27.Bg5! Kg8 28.Bxe7 Bxe7 29.Rc1!

Everything runs like clockwork.

29...Qb4 30.Rc8+ Nf8 31.h6 g5

617
32.h7+!

White gives up the pawn to get to f7 with the queen, both cornering the king and attacking the minor
pieces, with devastating threats.

32...Kg7 33.h8=Q+ Kxh8 34.Qxf7 Qxd4

If 34...a5,

analysis diagram

618
35.Rc1!. If the enemy king is in the corner, you have several check​mates: along the seventh and the
eight ranks, or along the h-file. These rook retreats are always spec​tacular and unexpected, as in a
famous Polugaevsky game (Game 197).
34...Qa3 35.Rc3–+.

35.Rxf8+ Bxf8 36.Qxf8+ Kh7 37.Qf7+ Kh8 38.Qh5+ Kg7 39.Qxg5+ Kh7 40.Qe7+ Kh8 41.Qf6+
Kh7 42.Qxe6 Qd1 43.Qf5+ Kg8 44.e6 1-0

Game 197
Lev Polugaevsky
Gyorgy Szilagyi
Moscow 1960 (10)

White to move

32.Bf8+! Rxf8

Black has the choice between two checkmates: 32...Kh5 33.Rxh7#.

33.Rd3 1-0

34.Rh3# will follow.

Game 198
Magnus Carlsen 2690
Vasily Ivanchuk 2750
Morelia/Linares 2007 (11)

619
 White to move

21.Bf6!

An old attacking rule is that when your opponent is weak on the dark squares, the exchange of the
dark-squared bishops favours the attacker, as seen in the standard attacks against a king’s fianchetto.

21...Nc6

22.Qg5!

620
This is a brilliant move that threatens to take on g7, and then play h5-h6+ and Qf6, winning at least
the h7-pawn and with it the game. There is more to it, as Black’s reply is forced. White had
anticipated this and had calculated a forced win.

22...h6 23.Qc1!

Attacking g6, but also pinning the knight on the c-file.

23...g5 24.Bb5!

Attacking the pinned knight.

24...Bd7 25.d5!

Liberating the d4-square.

25...exd5 26.Nd4

White wins a piece.

26...Bxf6 27.exf6 Qd6 28.Bxc6 Qxf6 29.Bxd7 Qxd4 30.g3 Qc5 31.Qxc5 bxc5 32.Bc6 d4 33.Bb5
Kf8 34.f4 gxf4 35.gxf4

Black resigned.

Game 199
Magnus Carlsen 2886
Viktor Laznicka 2678
Moscow Wch Rapid 2019 (8)

621
 White to move

The queen is not ideally placed on d2, and Carlsen would like to place it, say, on e3, where it is more
active. So he uses an intermediate threat, like in the 2007 game against Ivanchuk, also to induce
Black to slightly misplace his rooks and attract them on the kingside.

19.Qg5! Rdg8 20.Rc1!

Carlsen’s idea is to play on the side where the kings are!

20...Rh5 21.Qe3

The replacement has been completed, with Black’s rook temporarily misplaced.

21...e5 22.dxe5 dxe5 23.c4 d4 24.Qd2 Nd7 25.c5 Rh6 26.b4 Qc6 27.Qb2 Qb5 28.Qb3 Ra6 29.Nb2
Rf6 30.a4 Qc6 31.b5 Qe6 32.Qb4 Rc8 33.c6 Nb6 34.a5 Na8 35.cxb7 Kxb7 36.Rxc8 Qxc8 37.Rc1
Qd8 38.Nd3 Nc7 39.Nc5+ Ka8 40.b6 axb6 41.axb6 Na6 42.b7+

Simpler was 42.Nxa6! Rxb6 43.Rc8+ Qxc8 44.Qxb6+–.

42...Ka7

622
43.Nxa6

43.Qa4!, threatening 44.Nd7, was the only clear win: 43...Rb6+ 44.Kc2! (44.Ka2 Qd6) 44...Qd6
45.Kd1. Not an easy manoeuvre to figure out in seconds, and the posi​tion is not easy to assess,
though the engine finds it crushing!, e.g. 45...f6 46.Nxa6 Rxa6 47.Qe8 Qb8 48.Qe7 and 49.Rc8.

43...Rxa6

43...Rb6! 44.Rc8 Kxa6 45.b8=N+!? (the best try) 45...Kb7 46.Rxd8 Rxb4+ 47.Kc2 Rc4+ 48.Kd2
Ra4. Here, the weakness of the g2-pawn and the passivity of the knight offers Black good drawing
chances.

44.Rc8 1-0

Game 200
Magnus Carlsen 2872
Shakhriyar Mamedyarov 2772
Moscow blitz 2019 (8)

623
 Black to move

34...cxb2+

34...Bf5! was winning, threatening, for example after 35.c7, 35...Qxb2+ 36.Rxb2 cxb2+ 37.Ka2
b1=Q+ 38.Rxb1 Rxf2+ 39.Rb2 Rbxb2+ 40.Ka1 Rb1#.

35.Ka2

Now it is White who has a winning advantage.

35...Bf5 36.c7 Rf8 37.Qc4+ Kh8 38.Ne6 Rxa3+ 39.Kxa3 Ra8+ 40.Kb3 1-0

Game 201
Magnus Carlsen 2872
Nodirbek Abdusattorov 2635
Moscow blitz 2019 (6)

624
 Black to move

33...Rxe4! 34.Rf1

34.Rxe4 Qxd1#.
Magnus played very well in this event and won the blitz as well as the rapid. However, as more often,
he had some problems facing young players like Firouzja and Abdusattorov. In this game too he was
the victim of a simple tactic immediately spotted by his opponent. After some mutual mistakes the
game was drawn.

... ½-½ (51)

Game 202
Magnus Carlsen 2872
Bartosz Socko 2618
Moscow blitz 2019 (10)

625
 White to move

21.Nf6+! gxf6 22.Rxd7 b5 23.Qxa7 Ra8 24.Rxd8 1-0

Game 203 Sicilian Defence

Magnus Carlsen 2843
Radoslaw Wojtaszek 2744
Shamkir 2018 (5)

1.e4 c5 2.Nc3 d6 3.d4 cxd4 4.Qxd4 Nc6 5.Qd2

This remarkable opening idea was suggested and sent to Nielsen by the Greek FM Ioannis
Simeonidis.

5...Nf6 6.b3 e6 7.Bb2 a6 8.0-0-0 b5 9.f3 h5

626
10.Nh3!

This is a standard square for the knight when f2-f3 has been played, as Ne2 shuts in the light-squared
bishop. Ironically, Carlsen already sent the king’s knight to h3 in the preceding game against
Wojtaszek.

10...Be7 11.Ng5 h4 12.f4 Bb7 13.Kb1 Rc8 14.Be2 Qc7 15.Rhe1 Nh7 16.Nxh7 Rxh7

17.g4

627
White is much better, but Carlsen missed a nice tactic that would not have escaped the attention of a
seasoned Sicilian killer like, say, Tal.
The thematic 17.Nd5! wins: 17...exd5 18.exd5 Nd8 (18...Nb8 19.Bd3 g6 (19...Rh5 20.Rxe7+ Qxe7
(20...Kxe7 21.Qe2+)) 20.Bf6) 19.Bg4 Rb8 (19...Kf8 20.Re3 Bf6 21.Rde1) 20.Qd3 Rh8 (20...g6
21.Bf6) 21.Bxg7.

17...hxg3 18.hxg3 Bf6 19.Bd3 Rh8 20.g4 Nd4 21.Re3 Kf8 22.Ne2 Nxe2 23.Rxe2 Bc3 24.Bxc3
Qxc3

25.Qe3

Of course, White keeps the queens on the board, as the attack is ferocious with the king still in the
centre.

25...Rc5 26.e5 dxe5 27.fxe5 Rh1 28.Rxh1 Bxh1 29.Rh2 Rxe5 30.Rh8+ Ke7 31.Qa7+ 1-0

Game 204 Scandinavian Defence

Magnus Carlsen 2886
Azamat Utegaliyev 2564
FIDE Wch Rapid 2019 (1)

1.e4 d5 2.exd5 Qxd5 3.Nc3 Qd6 4.d4 c6 5.Nf3 Nf6 6.Ne5 Nbd7 7.Nc4 Qc7 8.a4 a5

628
9.Qf3!

Simple and excellent. The move had been played once before, in an U14 Girls’ championship, and in
such positions, with an advantage in development, this is what chess is all about: direct,
straightforward time-gaining moves – the Morphy legacy. Now, White is ready to play Bf4 and castle
queenside, or at least use his queen’s rook.

9...Nb6 10.Bf4 Qd8 11.Be5 Nbd5 12.Nxd5 Nxd5 13.c3 Nf6 14.Bd3 Be6 15.0-0 Bxc4 16.Bxc4 e6

17.Qg3!

629
White already has a large advantage, but this is the best move. The only interesting idea for Black is
to play ...Be7, so White prevents it, and kingside castling is made difficult, while queenside castling
looks too dangerous with a damaged pawn structure after ...a7-a5.

17...g6

18.f4

This move is fine. However, White had an even stronger one: the beautiful and thematic 18.d5!. The
logical 18...exd5 loses immediately, due to 19.Rfe1 Kd7 20.Qh3+. If 18...cxd5 19.Bb5+ Ke7 20.Bc7
wins the queen; the only way to ‘survive’ is 18...Bg7, giving up two pawns with 19.dxe6 0-0
20.exf7+, but this is a hopeless situation.

18...Bg7

630
19.f5!

This move is a multi-line opener: the f- and e-files, the a2-g8 diagonal and the h2-b8 diagonal. In a
few moves, what seems like a well-protected king will be completely naked!

19...exf5 20.Bc7! Qc8 21.Qd6! Ne4 22.Rae1

As always, like a magician, Carlsen has managed to involve all his pieces into the battle. ‘Help your
pieces, so they can help you’ – Paul Morphy.

22...Qd7

631
23.Rxe4+!

23.Bxf7+ Qxf7 24.Rxe4+ is the same.

23...fxe4 24.Bxf7+! Qxf7 25.Rxf7 Kxf7

26.Qd7+

The clever finishing touch: if 26...Kf8 (26...Kf6 27.Be5+), 27.Bd6+ Kg8 28.Qe6#. ‘That’s all,
folks!’

632
1-0

Game 205 Tarrasch Defence

Levon Aronian 2784
Magnus Carlsen 2886
Moscow Wch Rapid 2019 (12)

1.d4 Nf6 2.c4 e6 3.Nf3 d5 4.Nc3 c5

The old Tarrasch Defence, played with a modern interpretation by Carlsen. The line with ...Bc5 is one
of Dubov’s many opening ideas.

5.cxd5 exd5 6.g3 Nc6 7.Bg2 cxd4 8.Nxd4 Bc5 9.Nxc6 bxc6 10.0-0 0-0 11.Qc2 Bb6 12.Na4 Bd7
13.b3 Re8 14.Bb2

14...h5!

The best move: most of Black’s pieces are ready to attack the white king, but at the moment there is
no clear way to bring them nearer, so Carlsen uses one of his favourite tools: throwing in the h-pawn.
The three important points are:
• the main idea is to weaken the white king’s pawn cover with ...h5-h4 and probably ...hxg3, so that
the king has less defenders around him. Sometimes, playing ...h4-h3 can also be useful:
• it is possible at some moment to use the h-file by ...Re8-e6-h6 or play ...g7-g6, ...Kg7 and ...Rh8;
• while the attack is on, it is important to free our king from possible back-rank checkmates, thanks to
the escape square on h7.
14...Ne4!? is also thematic, but less to the point.

633
15.h4 Ne4 16.e3 Rc8 17.Nxb6 axb6 18.Qd1

18...Bf5!?

One pawn is given up for two tempi; this is pure attacking play.
18...Bg4 19.f3 Nxg3 is a computer suggestion. It leads to unclear, messy play – not at all a Carlsen
move – at least when he feels he is not worse!

19.Qxh5 Qd7 20.Qd1

634
20...c5!

Black takes the centre and also gains manoeuvring space.

21.Re1

21...Rcd8

Black could have got an advantage with the elegant 21...Rc6!, bringing the rook towards the kingside;
or 21...Bh3!?, followed by ...Rc6.

22.Rc1 Bh3 23.Qf3 f6 24.Bxh3 Qxh3 25.Rcd1 b5 26.Qg2 Qe6 27.f3 Nd6 28.g4 d4 29.Qd2 Qd5
30.exd4 c4 31.Ba3 Qxf3 32.Qg2

635
32...Rxe1+! 33.Rxe1 Qc3!

In this game, between the 28th and 43th move, Black has constantly attacked material, or taken or
given check, quite a sequence of ‘initiative play’ – in Fischer’s style.

34.Rd1 b4! 35.Bb2 Qe3+! 36.Qf2 Qh3!–+ 37.Qg2 Qe3+ 38.Qf2 Qe4 39.Qg2 c3 40.Bc1 c2 41.Rf1
Qxd4+ 42.Kh2 Qe5+ 43.Kg1 Qc5+ 44.Kh2 Nf7 45.g5 Qe5+ 46.Kh1

46...Rd1

636
46...Rd4!. This is another ‘horizontal effect’ of the rook: 47.Kg1 Rxh4–+.

47.g6 Nd8 48.Qf3 Rxf1+

48...Rd4! 49.Kg2 Rxh4 50.Rh1 Re4 51.Rh5 Qd4.

It is interesting that Carlsen misses the chance to use the ‘horizontal effect’ of the rooks three times.
Of course, one has to remember that this was a rapid game, so perhaps we could only ‘reproach’ the
champion for the first one. Nevertheless, it is interesting (and even comforting for an average player,
like yours truly!) that even such a player like Magnus, who has a fantastic vision of the whole
chessboard, is subject to some human limitations regarding spatial geometry.

49.Qxf1 Qe4+ 50.Qg2 Qxh4+ 51.Kg1 Qd4+ 52.Kh2 Qd1 53.Qe4 Kf8 54.Qxb4+ Ke8 55.Qb5+
Ke7 56.Qc5+ Ke6 57.Qc8+ Ke5 58.Bb2+ Ke4 59.Qb7+ Nxb7

White resigned.

Game 206
Maxime Vachier-Lagrave 2873
Magnus Carlsen 2886
Moscow Wch Rapid 2019 (11)

Black to move

44...Qd2

Exchanging the queens allows White to escape. Magnus is generally happy to trade queens to get a
favourable ending; in doing so, he misses stronger continuations at times.

637
44...Rd2!, threatening 45...Rxe2 followed by 46...Qxc1, was winning: 45.Rd1 (45.Rc3 Qd6, and
Black dominates the board; he will win thanks to an attack on the king, for example 46.a3 b5 47.Re3
b4–+) 45...Rxd1+ 46.Qxd1 (46.Bxd1 Qc1#) 46...Qxf2.

45.a3 Qxe1 46.Rxe1 Rd2 47.Ka2 Be6+ 48.Kb1 Bb3 49.Kc1 Rc2+ 50.Kb1 Rc6 51.Bd1 Bxd1
52.Rxd1 Rh6 53.Rd7 b5 54.Kc2 Rxh4 55.Kd2 Rf4 56.Ke2 Kh6 57.Ra7 Rf6 58.Re7 Rc6

Draw.

Game 207
Hikaru Nakamura 2829
Magnus Carlsen 2881
Magnus Carlsen Chess Tour 2020 (53.1)

Black to move

12...c5! 13.dxc5 d5!

A beautiful pawn sequence, and Black is already better. Commenting on the match live, Giri was
impressed by the concept: Carlsen’s 12...c5! and 13...d5! was ‘prep for sure, a fantastic idea,’ said
Giri. Magnus: ‘I can reveal 12...c5! was not part of my preparation – it’s a move once you see it you
know that it’s right.’

14.h3

14.e5³.

638
14...Bxf3 15.Nxf3 dxe4

Black is better.

... ½-½ (40)

Game 208
Baadur Jobava 2604
Magnus Carlsen 2872
Moscow blitz 2019 (9)

Black to move

47...Bb4!

The bishop penetrates and is able to attack the pawns from inside the structure.

48.Nc3

48.Kxb4 a2–+.

48...a2 49.Nxa2 Be1 50.Kc4 Bxg3 51.Nb4+ Kd6 52.Kd3 Bxh4 53.e4 fxe4+ 54.Kxe4 Bg3 55.f5 h4
56.Nd5 h3 0-1

Game 209
Magnus Carlsen 2872
Vladimir Kramnik 2753

639
 Moscow blitz 2019 (18)

White to move

18.Bxg6!

The same exchange that proved excellent against Tomashevsky (Game 182). Here, the idea is not to
gain the e5-square, but rather to get an opposition between the knight and the bishop. Once again,
Carlsen likes to fight against the bishop by putting the pawns on its colour.
18.Nc5!? is not bad, but Carlsen’s method is crystal-clear.

18...hxg6 19.Bxd6! Bxd6 20.g3

Now, we have the fight between the knight and a nominally ‘good bishop’ that is limited by the white
pawns placed on its colour. The bishop is not bad, but relatively ineffective, and the knight is better.

20...b5 21.axb6 Qxb6 22.Re2 Rb8

640
23.Nc1!

The knight is directed towards the royal square in this structure: d3, as shown in many games arising
from the Queen’s Gambit (in the same structure with reversed colours, the key square is d6), with the
reference game being from Petrosian (Game 210). Here, the most important point is that the only
target for Black is b2, and protecting it with the knight kills all possible black threats, as only the
major pieces can threaten it. Hence, Kramnik should have tried to change the pawn structure by
playing ...a4-a3 as soon as possible, to have a chance at counterplay. As he did not go for this plan
quickly enough, Carlsen was able to deliver a short masterclass on how to attack in this structure.

23...a5 24.Nd3 Qb5

641
25.h4!

The attacking plan is h4-h5, and if Black takes the pawn, to play along the h-file with Kg2 and Rh1.
If Black plays ...g6-g5, White plays Qg4 and f2-f4, opening the g-file. At some moment, the
‘defensive’ d3-knight can go to e5, turning into a menacing attacking unit.

25...Rfc8

25...Ra8.

26.Rd2 Ra8 27.h5 gxh5 28.Qxh5 g6 29.Qh6 a4

642
30.Kg2!

This is the strongest attacking move in the position.

30...Bf8 31.Qf4 Bg7 32.Ne5 Qe8 33.Rd3 f6 34.Ng4 Qf8 35.Rf3 a3 36.bxa3 f5 37.Qg5 Kf7

38.Rh1!

Including decisively the last piece that was not directly involved in the king hunt.

643
38...Rxa3 39.Rh7 Rcxc3 40.Nh6+ Ke8 41.Qxg6+ Kd8 1-0

Among many wins, the simplest is 42.Rxc3 Rxc3 43.Rxg7.

Game 210
Milko Bobotsov
Tigran Petrosian
Lugano ol 1968 (2)

Black to move

18...Nd6!

The ideal square for the knight. What follows is an exemplary attack by the Armenian Tiger.

19.Na4 Nbc4 20.Nxc4 Nxc4 21.Nc5 Nd6 22.Rac1 Qg5 23.Qd1 h5 24.Kh1 Re7 25.Nd3 Ne4
26.Nc5 Nd6 27.Nd3 Qf5 28.Ne5 f6 29.Nf3 Rg7 30.Nh2 Re8 31.Kg1 Ne4 32.Qf3 Qe6 33.Rfd1 g5
34.Qxh5 f5 35.Re1 g4 36.hxg4 fxg4 37.f3 gxf3 38.Nxf3 Rh7 39.Qe5 Qc8 40.Qf4 Rf8 41.Qe5 Rf5 0-
1

Game 211
Fabiano Caruana 2819
Magnus Carlsen 2875
Zagreb 2019 (3)

644
 Black to move

Black could have set a nice trap:

13...Be6!?

13...a3 was played in the game, with complicated play.

14.Nxa4?

is losing.
After 14.b3 axb3 15.axb3, Black may be very slightly better.

14...Qe8!

This is a double attack on two unprotected pieces, directly on the a4-knight and indirectly on the e1-
rook.

15.b3 Bg4 16.Rxe8 Rfxe8 17.Qe5 Bxe5 18.dxe5 Rxe5–+ ½-½

Game 212 London System

Magnus Carlsen 2863
Jeffery Xiong 2709
Clutch Chess Showdown 2020 (1.10)

1.d4 Nf6 2.Nc3 d5 3.Bf4 e6 4.e3 Bb4 5.Nge2 0-0 6.a3 Be7

645
7.Ng1!?

This is the kind of move Magnus is fond of. 7.Nc1 is a computer recommendation.

... 1-0 (25)

Game 213
Magnus Carlsen 2863
Fabiano Caruana 2835
Clutch Chess Showdown 2020 (3.3)

646
 White to move

20.Ne4! Qd7

After 20...dxe4?! 21.Bxe4, White has the advantage, though everything is not lost for Black after
21...Nd4!.

21.Ng5

The instalment of this knight near Black’s king is a great achievement for White. Three pieces, the
knight, bishop and queen, are dangerously aiming at the light squares around the king.

21...Rad8 22.Bc2 b4

23.Ba4!

The pin on the a4/e8 diagonal is very annoying.

Another powerful solution was the attacking 23.g4! hxg4 24.Qxg4 bxc3 25.bxc3 Bxc3 26.Re3.

23...bxc3 24.bxc3 Qc8

After 24...Re7 White is still much better. Now the winning move is:

647
25.f5! gxf5

26.c4!

26.Qxh5 Re7 is not enough for the thirsty (for blood) Norwegian.

26...Nd4 27.Rxd4! Bxd4 28.Bxe8 Rxe8 29.Qxh5 Qc7

648
30.cxd5!

‘If you see a good move, look for a better one’ is the old advice. This is not strictly necessary, but just
brilliant. In a blitz game, allowing ...Qxg3 with check would have been avoided by many players,
including many of Magnus’ ‘predecessors’!
30.Kg2?! kept good winning chances, but the game would not be over.

30...Re7

After 30...Qxg3+ 31.Kh1, Black can safely resign, as the queen is under attack, and the attack on his
king is raging. If 31...Qc7, 32.dxe6+–.

31.dxe6 f6 32.Qg6+ Kf8 33.Nf3 Bc3 34.Rc1! Rxe6 35.Nd4 Rb6 36.Qh6+ Kg8 37.Qe3 1-0

Game 214
Magnus Carlsen 2863
Levon Aronian 2773
Clutch Chess Showdown 2020 (2.8)

649
 White to move

19.h4!

This gives the bishop an additional diagonal, and the h-pawn may be handy for attacking purposes.

19...Bd7 20.Bh3 Rb7 21.b4!

Keeping the knight from c5.

21...Ne7 22.Ra2 Nc8

650
As Black’s next move is either 23...Ne7 or 23...Nb6, the prophylactic

23.Ra8!+–

is best.

23...Kh8

If 23...Bxb5, 24.Rxc8 Qxc8 25.Ne7+.

24.Qh5

24.Qa4 or 24.Qa1 was winning, intending to challenge the fianchettoed rook with 25.Qa6.

24...Qe8 25.Bf5 h6 26.Bh3

There was a funny variation: 26.Nf6 Qe7 (26...gxf6 27.Qxh6+ Kg8 28.Qh7#) 27.Nxd7 Qxd7

651
 analysis diagram

28.Qg6! fxg6 29.Bxe6 Qxe6 (29...Qe7 30.Rxf8+ Qxf8 31.Rxc8 Qxc8 32.Bxc8+–) 30.Rxf8+ Kh7
31.Raxc8 g5 32.h5 g6 33.Rce8 Qd7 34.Rh8+ Kg7 35.Reg8+ Kf6 36.Rxg6+ Kf7 37.Rh7+.

26...c6 27.Nxd6 Nxd6 28.Rxe8 Rxe8 29.Nf6 Re7 30.c5 Nc4 31.Nxd7 Rbxd7 32.Rxf7 Rxf7 33.Bxe6
Rf2+ 34.Kh3 1-0

Game 215
Magnus Carlsen 2881
Fabiano Caruana 2773
Clutch Chess Showdown 2020 (16.2)

652
 White to move

16.e6! fxe6

16...f6! was the best defence; after 17.Qe2 White is still clearly better.

17.Ng5 Rf6 18.Qc2 Rg6

18...g6 19.Bb2.

19.Rxe6! Qxe6

653
If Black takes with the rook with his rook, it is checkmate in two moves.

20.Nxe6 1-0

Game 216
Magnus Carlsen 2881
Hikaru Nakamura 2829
Chessable Masters 2020 (6.3)

White to move

Black has a very bad position, and most of his pieces are hemmed in because of a bad opening. So
there are several good moves for White, but the best one was chosen by Carlsen:

16.Ndc6!

16.Nxf7 was strong: 16...Rxf7 17.Nxe6 Qxe6; also strong was the simple 16.Nc4 Bxf4 17.exf4, with
a clear advantage.

16...Qh4

16...bxc6 17.Nxc6 Qf6 18.Bxd6+–.

17.Bg3 Qg5 18.h4 Qf6 19.Ng4 Qxb2 20.Bxd6 bxc6 21.Rc2 Qb5 22.Bxf8 Nxf8 23.h5 h6 24.Qf3
Qxh5 25.Qxc6 Nc7 26.Qf3 Nd5 27.e4 e5 28.Nxh6+ 1-0

654
 Game 217
Hikaru Nakamura 2829
Magnus Carlsen 2881
Carlsen Tour Final 2020 (2.51)

Black to move

40...g5+! 41.hxg5

41.Kxg5 Qh6+ 42.Kf5 g6+.

41...Qe4+ 42.Kg3 Qe1+

42...Qxc2! 43.Qf3 (43.Qxg7 Qxc3+ 44.Kf4 Qd2+ 45.Kf5 c3–+) 43...Qd3.

43.Qf2 Qxc3+ 44.Kf4 Qxa3 45.Ke5 Qe7+ 46.Kxd5 Kb5 47.Qf5 a5 48.c3 a4 49.Qb1+ Ka5 50.Kc6
Qe6+ 51.Kc7 Qe7+ 52.Kc6 Qe8+ 53.Kc7 Qe7+ 54.Kc6 ½-½

655
 Game 218
Daniil Dubov 2770
Magnus Carlsen 2881
Chessable Masters 2020 (9.2)

Black to move

Black is already a pawn up, but the finish is beautiful:

22...Qxb3! 23.axb3 Bxb4

White cannot unpin, because the only other square for the queen to protect the knight is d1, meeting
another pin on the d-file.

24.Bf1

656
24...a5!

White doesn’t have a move, so why not improve the pawn structure, Réti-like (see Game 178)?

25.Qd1

25.bxa6 Bxa6 would have lost more material.

25...h6 0-1

Black can win the knight at any time; for example, by playing ...Rc8-c5-d5.

Game 219
Sergey Karjakin 2720
Magnus Carlsen 2810
Wijk aan Zee 2010 (10)

657
 Black to move

25...Qe7!

This delicate move is a nice piece of prophylaxis. Black is already much better, thanks to the better
coordination of his pieces; the only one that does not work fully is the bishop on b7, and ...d5-d4 begs
to be played. If Black plays 25...d4 at once, the white knight has a good square on e2 (e4 is not stable
enough), so before playing it, Carlsen decides to keep the square under control, as the queen will
stand well on e7.

26.Rxf4 Rxf4 27.Ne2 Rf1 28.Nd4 Rxg1+ 29.Kxg1 Re8 30.h4 Qe1+ 31.Kh2 Qxg3+ 32.Kxg3 Kf7
33.Kf2 Kf6 34.g3 Bc8 35.c3 Bg4 36.Bc2 g5 37.hxg5+ hxg5 38.Bb3 Ke5 39.Bc2 Rf8+ 40.Kg2 Bd7
41.Nf3+ Kf6 42.Bb3 g4 43.Nd4 Ke5 44.Bc2 a5 45.Bd1 Ke4

White resigned.

Game 220 Queen’s Gambit Declined

Magnus Carlsen 2881
Fabiano Caruana 2773
Chessable Masters 2020 (1.1)

1.d4 Nf6 2.c4 e6 3.Nf3 d5 4.Nc3 a6 5.cxd5 exd5 6.Bg5 Be6 7.e3 Nbd7 8.h3 Bd6 9.Bd3 c6 10.Bf4
Qc7 11.Bxd6 Qxd6 12.0-0 0-0 13.Qb3N Rab8

13...Rfc8 14.Qxb7 Rcb8.

658
14.a4!

In this structure, the plan with a2-a4 and possibly a4-a5 makes more sense for White than the
traditional ‘minority attack’ with b2-b4, a2-a4 and b4-b5, as ...a7-a6 has already been played. Thus,
b4-b5 is more difficult for White to achieve, but on the other hand, the dark squares (a5, b6 and c5)
are weakened.
Ironically, Carlsen will later play a classical minority attack anyway, thanks to some tactical details
and some imprecision from his opponent.

14...Rfe8 15.Qa3 Qc7 16.Rfc1 Nb6 17.b4 Ra8

17...Nc4 18.Qa2 b5.

18.a5 Nc8

659
19.b5! axb5 20.Nxb5 Qd8 21.Nc3 Nd6 22.Qb4 Qe7

23.a6!

23.Ne5! was also good.

23...bxa6 24.Rxa6 Rxa6 25.Bxa6 Ra8 26.Bf1

In a typical structure, White has a great position. The main point is that Black has strictly no counter​‐
play, while White has various plans apart from the assault on c6, which is often insufficient to win.

660
26...Nc8 27.Qb2 Nd7

28.Ne2!

A perfect repositioning of the knight, which also opens the c-file;

28.Qb7 Qa3 29.Rc2 Ncb6.

28...Rb8 29.Qc3 Rb6 30.Qa5 Qd6 31.Nf4 Ne7

32.Nxe6! fxe6

661
32...Qxe6 33.Ne5+–.

33.Ne5 Rb8 34.Qa7 Nxe5 35.dxe5 Qd8 36.Be2 Kh8

This is a blunder, as Black is not able to protect e6 with the king anymore; 36...Ra8 37.Qc5 Kf7±.

37.Bg4+– Ng6 38.f4 Ra8 39.Qc5 Qe8 40.Qxc6 1-0

Game 221
Ivan Cheparinov 2670
Magnus Carlsen 2876
Banter Blitz Cup 2019

Black to move

This was a three-minutes game (without increment) where the players had to comment during the
game. Nevertheless, Carlsen’s next move was played instantly:

23...Qh6! 24.Rxh6 Rxh6 25.Qxh6 Bxh6 0-1

Game 222
Hikaru Nakamura 2829
Magnus Carlsen 2881
Carlsen Tour Final 2020

662
 Black to move

This was the final game of the whole Carlsen Tour. Carlsen had to draw, to win the tournament, and it
was an Armageddon game without increments. White had 1:55 left on the clock, Black 1:25, and the
position looks unpleasant for Black and not easy to play with 30 seconds less, as White has
possibilities to make the game complicated and basically to win on time if the position stays alive. So,
it is not only a matter of finding a good move, but also of how not to get flagged by one of the best
on-line specialists in the world. Carlsen made a fantastic decision he had thought about beforehand,
because he played it nearly instantly (in less than two seconds):

40...Rxd3!

The exclamation mark is mine, but of course the computer does not approve of this move, because
some other moves are given as equalizers, instead of giving up some material.
After 40...Qe6, the game is balanced.

41.Qxe7 Rxd1+ 42.Kb2 Bc6

After the exchange, both White’s and Black’s passed pawn will fall, and it should be an easy draw if
Black contains the remaining pawns.

43.e5 f3 44.Qf6 Rd2 45.e6 f2 46.e7 Re2 47.Qf7 Rxe7 48.Qxf2 Rg7 49.Qf4 Rg2 50.Qe5 Rg6 51.c3
Rg2+ 52.Kc1 Rg4 53.Qe6 Re4 54.Qf7 Re1+ 55.Kd2 Re4 56.b4

663
56...axb4! 57.cxb4 b5!

This is the way to a completely ‘dead’ position, a typical fortress. Carlsen’s famous saying ‘I don’t
believe in fortresses’ is to be understood only when he is the attacker! As a defender, he often uses
the trick of simplifying into a simple-to-hold position.

58.cxb5 Bxb5 59.Qd5+ Bc6 60.Qc5 Re6 61.Kc3 Rd6 62.b5 Bd7 63.Qe5 Bc8 64.Kc4 Rb6 65.Kc5
Kb8 66.Qe8 Rd6 67.Qf8 Rb6 ½-½

Carlsen noted that the position is perfect to play with little time, because he could premove ...Rd6-b6
forever, even when White gives check! In fact, the position is an easy draw even without the bishop;
this is one of the most important fortress positions to remember.

Game 223
Jose Cuenca Jimenez 2514
Magnus Carlsen 2876
Banter Blitz Cup 2019

664
 Black to move

22...Ng4! 23.Qd3 b5 24.Nf3

Here, Carlsen had seen in advance and played instantly:

24...Ne5! 25.Nxe5 Bxd4+ 26.Kh1 Bxe5

Black has a large advantage.

... ½-½ (79)

665
 Game 224 Dutch Defence
Anish Giri 2690
Magnus Carlsen 2815
Monaco rapid 2011 (4)

1.d4 e6 2.c4 Bb4+ 3.Bd2 Bxd2+ 4.Qxd2 f5

In his Stonewall series, Carlsen reveals much of his chess tastes:

• ‘It is generally a very good weapon in order to get those complicated strategic positions from the
opening.’
• ‘There are many plans.’
• ‘The pawn structure can change in so many surprising ways.’
• ‘You get chances to play for a king​side attack, which many people enjoy.’
• ’It is one of those openings where usually, the best player wins.’
‘I hope I haven’t given away all my secrets,’ he concluded, smilingly.
After the Wojtaszek game:
• ‘One thing the Stonewall forces you to do is to think in a non-standard way, which is something I
love.’
• ‘It’s one of the reasons why you play chess. You want to think, you don’t want to play the same way
every game, you want fresh positions, you want to show that you’re better than your opponent in a
massive fight.’
• ‘The Stonewall has never been about an opening for me; it is about the concept of forcing your
opponent to think early in the game.’
...

666
 Black to move

47...Rf2+! 48.Kxf2

48.Kd1 Rf1+ 49.Ke2 Re1+ 50.Kd2 Rg1–+.

48...Nd3+ 49.Kf3 Nxb4

The knight ending is an easy win.

... 0-1 (70)

Game 225
Magnus Carlsen 2834
Li Chao 2750
Doha 2015 (5)

667
 White to move

20.g4!

What a move! Carlsen combines attack and defence, the destruction of the enemy castled posi​tion and
the coordination of his forces, with his usual elegance. The queen will be able to join the attack (from
h2), while the bishop will assist in the defence of its king (from d1). Meanwhile, the g-pawn can open
new lines by taking the f-pawn, or being taken by the f-pawn.

20...Bb3 21.Bd1 a4 22.Qh2 Rfd8 23.Qh7+ Kf8

668
24.d5!

A brilliant defensive move, which is based on the necessity of cutting the perfect connection between
Black’s pieces. In particular, the queen must be refused access to the white king’s neighbourhood.
Black to move would win by playing 24.-- a3! 25.bxa3 Rxa3+ 26.Kb1 Ba2+ 27.Ka1 Bb1+ 28.Kb2
Nc4+ 29.Kxb1 Nd2+ 30.Bxd2 Rb3+ 31.Bxb3 Qxb3+ 32.Ka1 Ra8+.

24...Nc4 25.Nxg6+ Ke8 26.e6 a3

Checkmate in two is threatened.

27.exf7+

Not 27.bxa3?? Rxa3+ 28.Kb1 Bc2+ 29.Bxc2 Ra1+ 30.Kxa1 Bxc3+ 31.Kb1 Na3#.

27...Kd7

28.Ne5+!

This is the only winning idea: sacrificing the knight to liberate the white queen. White would be
checkmated without this trick.

28...Bxe5 29.Qxf5+ Kc7 30.Qxe5+ Nxe5

30...Kc8 31.Qe6+ Kb8 32.Bf4+.

31.Bxb3 axb2+ 32.Kxb2 Nbd3+ 33.Kb1 Nxc1 34.Rxc1 Kc8 35.dxc6 bxc6 36.f4 1-0

669
 Game 226
Ding Liren 2791
Magnus Carlsen 2863
Chessable Masters Final 8 2020 (2.16)

Black to move

36...Bb6! 37.Bxb6 Qxb6 38.Qxb6 Nxb6

Black is slightly better ... 0-1 (55)

Game 227 Nimzo-Indian Defence

Magnus Carlsen 2863
Anish Giri 2764
Legends of Chess Prelim 2020 (1.2)

1.d4 Nf6 2.c4 e6 3.Nf3 d5 4.Nc3 Bb4 5.Bg5 h6 6.Bd2 0-0

670
7.g4!?

A novelty. Carlsen likes positions where he provokes a pawn advance with a bishop move, then
retreats the bishop and attacks, thanks to the hook created. An example is the 6.Bg5 h6 7.Be3
manoeuvre in the Najdorf Sicilian in Carlsen-Vachier-Lagrave (see Game 3). This is also something
that is seen in computer matches.

7...Bxc3 8.Bxc3 Ne4 9.Qc2 c5 10.dxc5 b6 11.g5 hxg5

12.h4!± ... 1-0 (34)

671
 Game 228
Magnus Carlsen 2863
Hikaru Nakamura 2736
Carlsen Tour Final 2020 (2)

White to move

16.d5

A bit hasty, but playable according to Elshan Moradiabadi. It is a typical move from the mature
Carlsen, who may be not as patient as he used to be, but also as a rule likes to strike as early as
possible, in the style of Fischer. Especially at short time controls, Carlsen likes to provoke events and
enter the tactical fight early, as a way of pressing and putting pressure on his opponent.
The standard 16.Re1 is the computer’s choice.

16...exd5 17.Re1 Bxc4 18.Ng3 Qd8 19.Bb1

672
The attack certainly looks impressive!

19...b5!

A clever defensive move, anticipa​ting White’s main threat: 20.Qc2 g6 21.Qc3. Here, 21...d4 would
be a good move, if only the c4-bishop was protected.

20.Nf5 d4 21.Qd2 Be6

22.Rc5!?

673
In his typical attacking style, Carlsen involves the only piece that does not threaten the black king.
Now, the rook might go to g5 at some point.

22...a6 23.Nxg7! Kxg7 24.Qg5+ Kh8 25.Qh4

25...Rg8

Black could have set up a successful defence with 25...Re8!. The main idea is to counterattack against
White’s weakness, the unprotected rook on e1. Now, on the logical and apparently winning sequence
26.Rxc6 (26.g3²) 26...Rxc6 27.Bxd4, White threatens to checkmate in one and attacks the knight on
f6, ultimately winning the queen. However, Black has the terrible counter 27...Bf5!, protecting
against every threat, and threatening mate in one. This wins, as after the forced 28.Rxe8+ Qxe8
29.Bxf6+ Kg8, White will lose more material.

26.Rxc6! Rxc6 27.Bxd4 Kg7 28.Qxh7+ Kf8 29.Qh6+ Ke8 30.Bxf6 Qa5 31.Qe3 Qb6

674
Nakamura apparently grinned at this moment.

32.Rd1

‘Here I fell for the oldest trick in the book,’ explained Carlsen, speaking about the resigned attitude of
his opponent. The next trick allows Black to dream about saving the game, but ultimately, White
prevailed.
32.Be4+–.

32...Rd6! 33.Bd4 Qc6 34.Be4 Qc4 35.h3 Kd7 36.Rd2 Re8 37.Kh2 Bd5 38.Bf5+ Be6 39.Bd3 Qa4
40.Be5 Rd5 41.Qa7+ 1-0

Game 229
Magnus Carlsen 2881
Wesley So 2741
Skilling Open 2020 (20.1)

675
 White to move

32.g5+! Kxg5 33.Rxb5+ Rf5

33...Kf6 was a surprising draw that neither the players nor the commentators spotted during the game
(with the players on seconds remaining): 34.Rxb4 axb3 35.Rf4+ Ke7 36.Rxf8 Kxf8 37.axb3 Ke7
38.Kf3 Kxe6 39.Kf4 Kf6™ (39...h6 40.Ke4 Kd6 41.Kd4 Kc6 42.Ke5+–) 40.b4 h6! 41.Ke4 Ke6
42.Kd4 Kd6 43.b5 Kc7 44.Ke5 Kb6 45.Kf6 g5 46.Kg6 Kxb5 47.Kxh6 g4 48.Kg5 Kc5 49.Kxg4
Kd6 50.Kf5 Ke7 51.Kg6 Kf8=.

34.Rxb4 axb3 35.e7 bxa2

35...Re5 36.Rb5!.

36.e8=Q a1=Q

676
Carlsen instantly played the following forced sequence that leads to a win:

37.Qe7+ Kh6

37...Qf6 38.Qe3+ Kh5 39.Qh3+ Kg5 40.Qh4#; 37...Rf6 38.Qe3+ Kh5 39.Qh3+ Kg5 40.Rg4+ Kf5
41.Rd4+ Ke5 42.Qe3+ Kf5 43.Qe4+ Kg5 44.Qh4+ Kf5 45.Qg4+ Ke5 46.Qe4#.

38.Rh4+ Rh5 39.Qf8+ Qg7 40.Qf4+ g5 41.Qd6+ Qg6 42.Qf8+ Qg7 43.Rxh5+

Black resigned.

Game 230
Magnus Carlsen 2385
Christian Laqua 2196
Salzburg 2003 (4)

677
 White to move

21.dxe6!

This queen sacrifice allows White to enter a favourable ending after some complications.

21...Bxe4 22.Bxe4 Rxe6 23.Bxa8 Rxe1+ 24.Rxe1 Rxa8 25.Re7

This is a winning endgame, thanks to the superiority of White’s pieces, and the 12-year-old Carlsen
displays great accuracy.

25...Nf8

678
26.Rb7!

The beginning of a series of forced moves that lead to a winning rook ending.

26...Ne6 27.Bb8! Nd8 28.Re7! Ne6 29.Bxa7 Kf8 30.Rb7 Nd8 31.Rb8 Rxa7 32.Rxd8+ Ke7 33.Rd2

White is a pawn up, his rook is well-placed and his king will be able to reach an active position, so it
is basically winning.

33...Ke6 34.Kg2 h5 35.h4 g6 36.Kf3 Ke5

679
37.a4!

This is the clearest way to stop all Black’s chances of counterplay. Now, the win is just a walk in the
park. The white king invades on the kingside and grabs more pawns.

37...Rc7 38.Rd5+ Ke6 39.Kf4 Rc8 40.Kg5 Ra8 41.f4 f6+ 42.Kh6 Rc8 43.Kg7 Ra8

44.a5!

This is a nice touch, and typical of the practical player that Magnus already was by then. Other moves
are winning, but why allow even a hint of counterplay, when the opponent is reduced to total
passivity?
44.Kxg6 Rg8+ 45.Kxh5+–.

44...Rxa5 45.Kxg6 Ra3 46.f5+ Ke7 47.Rd3 b5 48.cxb5 c4 49.Re3+ Kf8 50.b6

Black resigned.

Game 231
Magnus Carlsen
Hannes Stefansson
Moscow 2004 (8)

680
 White to move

13.Rd1!

This is the kind of ‘silent’ sacrifice Carlsen uses a lot nowadays. White simply intends to castle and
mobilize more forces. Black has the choice of taking the a-pawn, or to play in a more passive fashion,
hoping to develop his queenside pieces.

13...Qc7

After 13...Qxa2 14.0-0 Qxb2 15.Nb3, White has the typical ‘Catalan’ pressure against Black’s
queenside. He will be able to get back his pawns, with a clear edge. I studied the position with the
computer, and a possible variation is 15...Rxd1 16.Rxd1 Nd7.

681
 analysis diagram

Now, the next move is a hard one to find. To keep his edge, White has to find 17.h4!, a great
prophylactic move. As Black has no active move, White improves the position of his king (if
17.Nxc5, Nf8 18.Rd8 Qb1+! 19.Bf1 Qb6 20.Qd4 e5!⇆; 17.h3 is also good) 17...Qc2 18.Rd6 Qxc4
19.Nxc5 Qb4 20.Qd4 Qxd4 21.Rxd4 Nf8 22.Rd8. Black cannot move, and White can recoup at least
an exchange after Nd7.

14.0-0 Bd7

682
15.Qf4! e5

After 15...Qxf4 16.gxf4 Nc6 (16...Bc6 17.Nb3 Na6 (17...Nd7 18.Bxc6 bxc6 19.Na5+–) 18.Bxc6
bxc6 19.Na5+–) 17.Nb3, Black has to play 17...b6 to protect the pawn, but it loses instantly to
18.Rxd7.

16.Qe3 Na6

17.Ne4! Bc6 18.Nc3 Rd4

683
19.Nb5! Bxb5 20.cxb5 Nb4

20...Qd7 21.Qb3+–.

21.Qb3

The knight is dominated.

21...a5 22.Rxd4 cxd4 23.a3 Qc2 24.Qxc2 Nxc2 25.Bxb7 1-0

Game 232
Magnus Carlsen 2553
Predrag Nikolic 2676
Wijk aan Zee 2005 (11)

White to move

15.Qa1!

Magnus loves to place the queen in the corner. From there, it connects on the a-file with the rook, and
also points to the long diagonal.

15...Qc7 16.b4 f6 17.Re1 Kf7 18.b5 Na5

684
19.Qd1!

A beautiful switchback. Now, the action is on the kingside, and the relevant diagonal is d1-h5.

19...Rae8

20.Ng5+!

Opening the queen’s diagonal.

685
20...fxg5 21.Qf3+ Kg8 22.Rxe6 1-0

22...Rxe6 23.Qf8#.

Game 233
Magnus Carlsen 2570
Oystein Hole 2336
Gausdal 2005 (2)

White to move

21.Bd3!

The target is the king.

21...Ba6 22.Bb1! Nc6

686
After Black’s move, there are two weaknesses in the position. Tactically, the equivalent of an
unprotected piece is a piece that is attacked once and defended once. First, there is a possible
checkmate on h7, if the queen attacks it. Second, the c6-knight is also fragile, as it is attacked once
and defended once. So, the main motif of the combination will be a double attack with the queen on
c6 and h7. The third possible weakness is g7 (attacked by the queen and defended by the king).
So, Carlsen played:

23.Bh6! Rg8

The main variation is 23...gxh6 24.Qe4 (24.Nxc6 first is also possible) 24...f5 25.exf6 Nxf6 26.Qxc6
and White wins. After the queen exchange, e6 and a5 are too weak.

687
24.Bxh7! Kxh7 25.Be3

Threatening checkmate in one on h5.

25...Rc8 26.Nxe6! fxe6 27.Qh3+ Kg8 28.Qxe6+ Kf8 29.Bxb6 Nxb6 30.Nxa5 Rc7 31.Nxc6 Bc8
32.Qb3 1-0

32...Rxc6 33.Qf3+.

Game 234
Magnus Carlsen 2072
Gustav Gudbrandsen 2293
Norway tt 2002 (6)

White to move

19.Rac1!

19.Rfc1!.

19...h6

688
Had Black seen the threat from his 11-year-old opponent and played 19...b6, White could still have
won thanks to his better piece mobilization after 20.Rc7! (20.Bxh7 is playable, but less efficient)
20...h6 21.Rfc1 0-0 22.Nh7 Re8 23.Nf6+ gxf6 24.Qxh6 f5 25.R1c3, and White has an unstoppable
attack, e.g. after 25...Qf8 26.Qf4, and say Bb5, the rook goes to g3 and it is over.

20.Qxa7! Rb8

20...Rxa7 21.Rxc8++–.

21.Qxb8! Nxb8 22.Rxc8+ Kd7 23.Rxh8 1-0

Game 235
Magnus Carlsen 2770
Veselin Topalov 2812
Sofia 2009 (1)

689
 White to move

33.Qd3!

A double attack – the knight on d7 is attacked and mate is threatened.

33...Qxe7

33...Qxc7 34.Ng5+ hxg5 (34...Kh8 35.Nf7+ Kh7 36.Qxg6#) 35.Qxg6+ Kh8 36.Qh5+ Bh6
37.Qxh6#.

34.Rxd7 Qh4+ 35.Kf3 Qh5+ 36.Kg3 1-0

36...Kh8 37.Rd8+ Kh7 38.Nf6+ Bxf6 39.Qd7+ Bg7 40.Qf7, and checkmate is near.

Game 236
Levon Aronian 2782
Magnus Carlsen 2813
Nice blindfold 2010 (2)

690
 Black to move

20...Nexd5! 21.cxd5 Nxd5 22.Kg1

22...e4!

Opening the diagonal for the second bishop to join the party!

23.Nxe4 Bd4+ 24.Rf2 Rxf2 25.Nxf2 Qh4 26.Qe1 Rf8 27.Bf3 Rxf3 28.Qe4 Qxf2+ 0-1

691
The whole game – blindfold – cost the Norwegian less than ten minutes thinking time!

Game 237
Magnus Carlsen 2881
Maxime Vachier-Lagrave 2860
Opera Euro Rapid 2021 (12.1)

White to move

26.b4!

An important prophylaxis, maintai​ning a dominating king position for White. Many specialists
discuss how to deal with minor pieces. Both camps should try to fix their pawns on dark squares, as
they have a light-squared bishop, and of course be cautious with pawn moves that might provide a
strong outpost for the enemy knight. But overall, in minor-piece endings, the most powerful piece is
the king!
If 26.g4 c5+! 27.Ke3 h6=.

26...h6 27.a4 g5 28.g3 gxf4 29.gxf4 Nb8

692
30.Nd5+! Bxd5 31.exd5 b6

Or 31...Nd7, allowing a king march to the kingside: 32.Ke4 Kf7 33.a5 Ke7 34.Kf5 Kf7 35.Bh5+
Kg7 36.Ke6 Nf8+ 37.Ke7+–, and to d8, etc.; 31...Kd6 32.Ke4 Nd7 33.a5 Ke7 34.Kf5 Nf8
35.Bc4+– (35.Bf1+–).

32.h5 a5

32...Kd6 33.Ke4 Ke7 34.Kf5 Kf7 35.Bc4+–.

33.bxa5 bxa5 34.Kc5 f5

693
35.Bb5!

This is a textbook domination, even if many other moves also win here.

35...Kd8

35...Kf6 36.d6 cxd6+ 37.Kxd6.

36.Kd4 1-0

Game 238
Magnus Carlsen 2862
Alireza Firouzja 2749
Wijk aan Zee 2021 (1)

694
 White to move

23.d5!?

‘I knew in my heart that once I’d sacrificed a pawn on d5 I wasn’t going to back down later on and
look for equality, so I knew at that point that at least mentally I’d already burned some bridges.’
Often, Carlsen gives important indications about high-level play, including his own psychology. He
explains on occasion how to play against particular players, and doesn’t mince words. During the first
part of the Candidates in 2020, Alekseenko played a natural move against Grischuk, which attacked a
rook. Carlsen, commenting live, said that it was a great move against Grischuk. 99% of players would
have replied instantly with a natural move, and Grischuk played it, but only after ten minutes’
thought! Carlsen had already explained publicly how to deal with Kramnik’s draw offers after a long
game (‘refuse!’). When asked about Kasparov training Nakamura during 2011, he answered that ‘...
right now there are four to five players in the world, including me, who have significantly better
understanding of chess than Nakamura. And I don’t know if even a man like Kasparov can change
that in the short term.’

23...exd5 24.e5 Ne4 25.Qd4 Rdc8 26.Raa1 a5 27.Rab1 Bc6

695
All White’s pieces are active, so...

28.e6!?

A typical line and square opening, and also a second pawn sacrifice! We can see this kind of
speculative double pawn sacrifice against Howell and Harikrishna. In those games, the main objective
was to restrict the enemy army. Here, the concept is more impressive: just giving more air to his own
pieces, while slightly weakening Black’s castled position. Black’s pieces are also quite active, and the
passed a-pawn can advance freely. But the attacking power of the white pieces compensates for the
missing two pawns, and the engines agree.

28...fxe6 29.Ne5 Qf6 30.f3

Here, 30.Bxe4, followed by taking on c6 and e4, would be enough for a draw, but Carlsen plays for
more.

30...Ng5 31.Rb6 Be8 32.Qe3 a4

32...Qd8, threatening 33...d4, would have been better, or at least more ‘practical’, as Carlsen put it
later.

33.Ng4 Qd8

696
34.Rxe6!

Bringing all the pieces on!

34...Nxe6 35.Qxe6+ Bf7

35...Kh8, with the idea 36.Nxh6 Bh5!, defending g8 while keeping g6 under control, had to be
played. Black is not worse.

36.Nxh6+! gxh6 37.Qxh6 Qc7 38.Qh7+ Kf8 39.Qh8+ Bg8 40.Qh6+ 1-0

697
40...Qg7 (40...Kf7 41.Bg6+ Kf6 42.Bh5+ Kf5 43.Bg4#) 41.Qd6+ Kf7 42.Qe7#.

Game 239
Wesley So 2741
Magnus Carlsen 2881
Opera Euro Rapid 2021 (29.1)

Black to move

16...exd4

‘I really need to take on h3, that’s what I feel, regardless of whether it works!’ said the World
Champion. As Magnus pointed out, 16...Bxh3 17.gxh3, and now 17...a4!, rather than the immediate
17...Qxh3, was very strong: ‘The thing is throwing in a4 is not a novel or particularly difficult idea if
you start looking for things there.’ However, his main point was that he should have trusted himself
to go for the sacrifice. He noted a sample line without 17...a4: 16...Bxh3 17.gxh3 Qxh3 18.Bxf4 Nxf4
19.Ne3 Rae8 20.Ng5 Qh6 21.Qg4 (the best move is 21.Bxf7+ Rxf7 22.Nxf7 Kxf7, and now only
23.Qg4! holds after 23...Re6 24.Nf5=) and he stopped calculating here, thinking White was fine.
However, 21...Qxg5 wins for Black.

17.Nxd4 Bxd4 18.Qxd4

698
‘My first reaction was, “what, he can’t take with the queen on d4, this can’t be it!”, and he also did it
very quickly,’ said Magnus. But again, he spent over three minutes before deciding against any of the
available sacrifices. 18...Nxg2! was lethal, but Magnus had missed the crucial 21st move.

18...Bxb3

18...Nxg2 19.Kxg2 Bxh3+ 20.Kh2 Nh4 21.Bd1

analysis diagram

21...Bg2!, followed by the decisive arrival of the queen on h3. By the way, the simple 18...Bxh3 is

699
also very advantageous for Black: 19.Bxf4 Nxf4 20.gxh3 Qxh3 21.Ne3, and now the best move is
21...Qh6!–+ (21...Rae8 is also better for Black, but it allows White to give up the queen with
22.Qxd6 Re6 23.Qxf4 Rg6+ 24.Qg3).

... ½-½ (54)

Game 240
Magnus Carlsen 2881
Wesley So 2741
Opera Euro Rapid 2021 (30.1)

White to move

23.Bd1

‘My intuition was screaming that 23.Kh1 wins, just screaming! The thing is it wins, and it’s not very
difficult: 23.Kh1 axb3 24.Rg1+ Kh8, and here I was looking at all sorts of Qe3s and all this, but the
thing is I just needed to consider e6 and that’s just game over on the spot. So that was really, really
bad as well, because if I see Kh1, I think at least practically speaking the game is just over, because if
he can’t do anything immediately here, he’s just dead. Strategically he’s absolutely busted since my
attack is just unstoppable.’ (24...Bg6 25.Rxg6+ fxg6 26.Qxg6+ Kh8 27.Qh5+ Kg8 28.Rg1+) 25.e6!,
threatening 26.Be5 followed by 27.Qf5, or vice versa: 25...f6 (25...fxe6 26.Be5+ Qf6 27.Qe3+–)
26.Qf5 Be8 27.Qf4+–.

... ½-½ (55)

Game 241

700
 Magnus Carlsen 2876
Alan Pichot 2590
Banter Blitz Cup 2019 (1.8)

White to move

20.Rfb1

The exchange is slightly strange, as it can give Black more counterplay.

After 20.Bg5, White has a clear edge, with a space advantage and better minor pieces.

20...Qxb1+ 21.Rxb1 Rxb1+ 22.Kg2 a4 1-0

Game 242
Magnus Carlsen 2835
Fabiano Caruana 2832
London Wch m 2018 (6)

701
 White to move

44.Bxd5!?

This is very typical of Carlsen’s defensive style. He prefers to provoke complications, rather than
trying to hold passively.
The passive 44.Nd1 should hold, according to the computer. Karjakin (‘the minister of defence’)
would probably play like this: 44...Nxf3 45.Kxf3 d4+ 46.Ke2 Bb5+ (after 46...Be4 47.g4 Bb1, White
has the important 48.Kd2!, and if 48...Bxa2, 49.Kc2) 47.Kd2 Bb4+ 48.Kc2 d3+ 49.Kc1; or 44.Nf1.

... ½-½ (80)

Game 243
Sergey Karjakin 2769
Magnus Carlsen 2853
New York Wch m 2016 (9)

702
 Black to move

38...Ne7

Entails great complications, while 38...Bd8 could hold.

... ½-½ (74)

In the following example, White’s choice is very interesting. Carlsen – at least the 2014 version –
prefers to keep a small edge without counterplay, rather than going for the complications that might
follow after a capture with the pawn.

Game 244
Magnus Carlsen 2881
Shakhriyar Mamedyarov 2760
Shamkir 2014 (1)

703
 White to move

19.Rxd5!?

Computers claim that White should take with the pawn. I guess that the elder Carlsen might play this,
because he would be aware that Black has equalizing chances.
After 19.cxd5, there is a forced line that is neither easy to calculate, nor simple to evaluate: 19...Nxe5
(on 19...c4, intending to liberate the b4-bishop, White has the strong 20.d6, intending a kingside
attack with Rd4, h2-h4, etc.) 20.Nxe5 Rxe5 21.Qxe5 Qxc2. The position is nearly winning for White,
but this is very difficult to estimate initially. After 22.Re1, White is threatening to push the d-pawn:
22...h6, and here, White has to play 23.d6! hxg5 24.d7 Rf8 25.Qd6 Rd8 26.Qe7 Rxd7 27.Qxd7. Even
at the end of the line, the evaluation is not clear-cut at first, but the computers think that White is not
far from winning.

19...Nf8 20.h4 h6 21.Be3 Ng6

According to the computer, the position is nearly equal after 21...Rcd8. Once again that seems
difficult to figure out at first, as White seemingly has an easy game and a space advantage with more
central pieces. In fact, Black’s queen is remarkably placed on a4, and can become active very quickly,
for example, if Black takes on d5 and plays ...Qe4.

22.Qd3 Re6 23.h5 Ne7

704
24.Rd6

Another remarkable decision, which is very typical of the Norwegian’s way of playing. White keeps
it complex, not allowing exchanges that at least will enable his opponent to play some automatic
moves. The rook move is good, of course, but why should one not play the direct 24.Rd8+ ? Well, the
first thing is that by playing this variation (advocated by the computers), Carlsen would go into a
forced line, when his opponent has ‘only moves’ and no difficulty at all in finding the next few
moves: 24...Rxd8 25.Qxd8+ Kh7 26.Rd2 Bc1 27.Rd7 Bxe3 28.fxe3.

analysis diagram

705
From afar it is not easy to under​stand what’s going on, but the computer estimates that White is
winning. For a lesser player, like yours truly, it is not clear what’s going on even after seeing the
diagram position, as White’s king protection and generally the whole pawn structure is spoilt. In fact,
the important point is that thanks to the h5-pawn, the black king runs serious risks of being mated if
the major pieces go along the eight rank. For example, 28...Nc6 29.Qf8 is decisive, with Rxf7 to
come.

24...Bb4

25.Rc1!

This is the human way, cleverly removing the rook from the queen’s sight and intending to move it to
a better and safer location before going for the attack with all due energy. There is a very nice (and
totally crazy, from a human logic!) computer proposal here: 25.Bg5!. It looks like a joke, because
Black obviously won’t take. The following lines are logical and interesting. The main point is that
while the bishop stays on the queenside, White has one more piece to attack. So, thanks to a well-
timed g4-g5, he will open the kingside with an unstoppable attack. This scenario also happened in the
game.
 A) 25...hxg5 26.Rd8+ Rxd8 27.Qxd8+ Kh7 28.Nxg5+ Kh6 29.Nxf7+ Kxh5 30.Qd3 Qa3 31.Qh7+
Kg4 32.f3+ Kf4 33.Qe4+ Kg3 34.Qg4#;
 B) 25...Qe8 26.Bxe7 Qxe7 27.g3 Qe8 28.a3 Ba5 29.Qd5 Rb8 30.Kg2 a6 31.Rc1 Re7 32.Rb1 Kh8
33.a4 (all the queenside moves are aimed at preventing the possible bishop’s liberation after ...b6-b5)
33...Bb4 (or 33...Qxa4 34.e6 fxe6 35.Qd3 Qe8 36.Ne5 Reb7 37.Nc6+–) 34.g4 Qxa4 35.g5 Qc2
36.Rg1;
 C) 25...Re8 26.Rxe6 fxe6

706
 analysis diagram

27.Bf6! gxf6 28.exf6 Nf5 29.g4, and White recoups the knight and enjoys a winning attack. The key
is that apart from Black’s shattered kingside, the bishop is semi-trapped on b4 and unable to defend or
lead any counterattacking attempt.

25...Re8 26.Rxe6 fxe6

27.Nh4! Qc6 28.a3

707
The stunning 28.Bxh6!! was winning! 28...gxh6 29.Rd1 Qc7 (on 29...Qc8, 30.Qg3+ Kh7 31.Rd6,
followed by queen to d3 will also win) 30.Qg3+ Kh7 31.Rd6 Qc8 32.Qd3+ Kh8 33.Ng6+ Nxg6
34.Qxg6 Bc3 35.Qxh6+ Kg8 36.Qg6+ Kh8 37.Qf6+ Kg8 38.Rd3.

28...Ba5 29.Rd1 Qc7 30.Ng6! Nxg6 31.Qxg6 Qf7 32.Rd3 a6

After 32...Qxg6 33.hxg6, Black cannot move because of the back-rank issues, so White will simply
collect the pawns with the rook.

33.a4 Rf8 34.g4 Qe8 35.Rd6 Qxa4 36.Qxe6+ Kh8

37.Bxh6!

With rook and queen in the attack, the destruction of the guards is worth much more than a mere
bishop.

37...Qa1+ 38.Kg2 Rxf2+

Trying for a perpetual check.

39.Kxf2 Qe1+ 40.Kg2 Qe4+ 41.Kh3 Qh1+ 42.Kg3 Qe1+ 43.Kf4 Bd2+ 44.Rxd2 Qxd2+ 45.Kf5
gxh6 46.Qe8+ Kg7 47.Qe7+ 1-0

After 47...Kg8 48.Kg6 Qg5+ 49.Qxg5 hxg5, White queens first: 50.e6 Kf8 51.h6. ‘... my favourite
player from the past is probably... myself, three or four years ago,’ declared Carlsen during his 2018
World Championship Match against Caruana. Was this game part of the admiration Carlsen had for
the 2014/2015 Magnus?

708
 Game 245
Hikaru Nakamura 2758
Magnus Carlsen 2826
Moscow 2011 (9)

Black to move

Here, Black has a comfortable game, thanks to his superior pawn structure, and could continue in a
slow, positional style with a good game. Occupying the c-file and trying to find a square for the queen
made sense. Carlsen won many games like this, by positionally out​playing his opponents. It would be
suited to his supposed ‘positional grinding’ style, as it is described by many commentators. In fact,
the champion is a much more versatile player than most people originally thought.

15...g5!?

This is a very interesting decision, and not an obvious one by all means. Carlsen has two principles in
the early part of the game, values that he constantly insists on, especially when he blitzes and
comments live on the Internet:
1. safety of the king;
2. control of the centre.
Here, there’s a conflict between the two principles: to grab the important central pawn, Black is
‘invited’ by his opponent to weaken his castled position permanently. There is no clear answer as to
whether Black should play ...g5-g4 or not. Black can keep a safe and equal position by playing
developing moves, or he can get a complicated game by playing ...g5-g4 and grabbing the e5-pawn,
but having a permanently weak king.
Ultimately, there is also a psycholo​gical factor. Nakamura gave the choice to his opponent, and kind

709
of asked, ‘Do you dare to take this pawn?’ Certainly, Carlsen will do this most of the time. More than
one century before, Capablanca explained why he took a poisoned pawn against Marshall (in the first
occurrence of the famous Marshall Gambit). He said that he felt ‘honour bound’ to accept the pawn.
Carlsen had another explanation when he was asked about his playing a dubious opening against a
GM (in a Banter Blitz): he said that there was a ‘fuck you factor’ that he liked. I am not sure that it is
fundamen​tally different, from a sporting point of view.
15...Rc8 was a quieter alternative.

16.Be3

16.Bd2 was superior, to position the bishop along the long diagonal: 16...g4 17.Ne1 Nxe5 18.Bc3,
and the position is unclear.

16...g4 17.Nd4 Nxe5 18.Bh6 Re8 19.e4 Bc5 20.Nb3 Rc8 21.Nxc5 Rxc5 22.Qa4 Bc6 23.Qd4 Qf6
24.Bf4 dxe4 25.Bxe4

25...Nf3+!

A very interesting decision, typical of Carlsen’s style and under​stan​ding of chess. Black should be
winning here, on objective grounds. He has a choice: he can play for a winning ending, with
counterplay for the opponent and possible complications, due to the poor position of his king, or he
can play a completely safe position with a very strong probability of winning, but maybe some
chances of a draw for White. Carlsen chooses the second option, probably for practical purposes: in
this position, there are strictly no risks of losing, and he knows that in a completely passive position,
it will be difficult for White to make the best of his chances, if there are chances at all. He took the
same type of decision against Karjakin, in Game 4 of their 2016 match, when, while winning, he

710
closed the position, thus allowing his opponent to build a fortress. Magnus plays it safe, because he is
convinced that he will outplay his opponent, even when the win is not 100% guaranteed. This
approach is remarkable because when the game is unclear, the Norwegian is prone to take lots of
risks, like earlier in this game when he decided to play the unclear 15...g5 instead of the safe 15...Rc8.
Carlsen also has a weak spot for opposite-coloured bishops endings, feeling that they offer much
more winning chances than is generally believed. This is also a personal preference, and he also
frequently likes some endings that are not so popular among the theoreticians, like queen against two
rooks, or queen and bishop against queen and knight.
25...Bxe4 26.Qxe4 Nf3+ would be the standard decision. Black is a pawn up (he might take another
one on b2), has a strong knight on f3 and should be winning. But with the queens on and the
dangerous position of the black king, this is not easy to play: 27.Kg2 Qxb2. This variation was
advocated by Karolyi, but White appears to have serious counterplay after 28.Be3 (28.Bh6!?).
27...Rec8! looks stronger. The difference with Carlsen’s solution is that here, Black’s king is not as
safe with the queens on the board, and Carlsen’s choice looks more practical and possibly better.
Apart from the theoretical evaluation, there is the practical one: if Black plays the position with the
knight against the bishop, the queens are still on the board, the black king is unsafe, and the game is
much more difficult to play. With Carlsen’s solution, there is absolutely no counterplay for White,
just some chances of draw after long sufferings. 28.Rac1 Rxc1 29.Bxc1 h5 30.h3 Qf5!–+. After the
queen exchange, the rook has a good square on c2, and the black king is able to get some activity,
contrary to White’s.

26.Bxf3 Qxd4 27.Rxd4 Bxf3 28.Rd7

Staying put with a move like 28.Bd2 was safer; for example, 28...h5 29.h4 Rec8 30.Bc3. White is still
holding his own, even if he faces a long torture, with many chances to go astray, and is probably lost
anyway. Carlsen does win these endings in nearly all cases. After 30...Rd5!, the winning chances are
99% according to Stockfish 12!

711
28...Rd5! 29.Rxd5

29.Rxa7? Red8 30.Rf1 Rd1 wins for Black. White’s constant problem is the mating pattern Bf3 +
Rd1.

29...exd5 30.Be3 Re4 31.Re1 d4 32.Bd2

32...Rxe1+!

An expert decision: the endgame is winning for Black.

712
33.Bxe1

33...Be2!

Keeping the enemy king from activating. This more or less forces the following pawn sacrifice:

34.f4 gxf3 35.Bf2 d3 36.Be1 Kg7 37.Kf2 Kf6 38.Ke3 Kf5 39.h3 h5 40.Bd2 Bf1 41.Be1

After 41.h4 Kg4 42.Kf2 Be2 Black can win according to Marin, following a long but simple plan:
• to attack the a2-pawn with Bd1-c2-b1;
• to come back to e4 with the bishop, to protect the passed pawns;
• the black king goes to c2, to attack the white queenside pawns.

41...Bxh3 42.Kxd3 Bf1+ 43.Ke3 Kg4 44.Kf2 Bb5 45.Bc3 Bc6 46.Be5 b5 47.Bb8 a6 48.Bc7 f5
49.b3 Bd5 50.Bd6

713
50...f4!

This brilliant breakthrough is Carlsen’s point. It enables Black to create a second passed pawn.

51.gxf4

On 51.Bxf4, 51...h4! wins immediately, as 52.Bc7 (52.gxh4 fails to 52...Kxf4) 52...h3 would allow
one of the pawns to promote.

51...h4 52.f5 Kxf5 53.Ke3 Kg4 54.Kf2 h3 55.Ke3 Be4 56.Kf2 Bb1 57.a3 Ba2 58.b4 Bf7 0-1

The bishop will get to g4 via h5, the king will go to grab the queen​side pawns. As the bishop alone
cannot defend against the two passed pawns on h3 and f3, the white king has to stay on the kingside.
Black’s king will go to b2 and take a3, and then he will play ...a6-a5, thus creating a third (!) passed
pawn to win the game (of course, without pawns on the queenside, it would be a simple theoretical
draw, as the passed pawns are blocked and separated only by one file). For example: 59.Bh2 Bh5
60.Bd6 Kf5 61.Bh2 Bg4 62.Ke3 Ke6 63.Bg3 Kd5 64.Bh2 Kc4 65.Kf2 Kb3 66.Bd6 Kxa3 67.Kg3
Ka4 68.Be7 a5, etc.

Game 246 English Opening

Magnus Carlsen
‘lempire’
Banter Blitz Cup 2020 (12.10)

1.c4 e5 2.Nc3 d6 3.d4 Be7 4.Nf3 exd4 5.Nxd4 Nf6 6.g3 0-0 7.Bg2 c6 8.0-0 Na6 9.e4 Nc5

‘Maybe he knows I am reputed​ly uncomfortable with those positions with a nice space advantage,’

714
Carlsen said in his live comments.

10.Re1 a5 11.h3 Re8 12.Bf4 Bf8 13.Qc2

‘It can be difficult psycholo​gically when you feel you get every​thing from the opening; then what are
you supposed to do? Like when you get a perfect set-up in the centre.’

13...Nh5 14.Be3 Nf6 15.Rad1

‘So now I have centralized all my pieces, I have great control of the centre. That’s when I start
messing it up.’

15...Qc7 16.f4 g6 17.Bf2 Bg7

‘So as I said, I don’t know what to do in these positions, but I do know one trick here, which is...’

18.Ndb5!+– cxb5 19.Nxb5

‘I don’t know if it is great, objec​tively speaking, but I am so uncom​fortable, manoeuvring in this

better position, that I just go for it!’

19...Qe7 20.Nxd6

‘I know I am giving away a lot of information about the way that I think, but it’s okay.’

... 1-0 (47)

Game 247
Magnus Carlsen 2835

715
 Fabiano Caruana 2832
London Wch playoff rapid 2018 (1)

White to move

Here, Carlsen thought for a very long time in a rapid game (9 minutes, as far as I remember), and it is
actually the moment of truth as White has a forced win.

24.Bxe6+

24.Rxd4 Kf7.

716
 analysis diagram

Here, the computer suggests a forced win with 25.Kh1!, to avoid the ...Ne5 trick, as now 25...Ne5
26.Rxd8 Nxf3 27.Re2 is winning.
If White plays 25.Red1, 25...Ne5 26.Rxd8 Nxf3+ 27.Kf1 Nxh2+ 28.Kg1 Nf3+ 29.Kf1= (29.Kh1
Rh2#).

24...Kf8 25.Rxd4 Ke7 26.Rxd7+ Rxd7 27.Bxd7 Kxd7 28.Rd1+ Ke6 29.f4 c5 30.Rd5 Rc2 31.h4 c4
32.f5+ Kf6 33.Rc5 h5 34.Kf1 Rc3 35.Kg2 Rxa3 36.Rxc4 Ke5 37.Rc7 Kxe4 38.Re7+ Kxf5 39.Rxg7
Kf6 40.Rg5 a5 41.Rxh5 a4 42.Ra5 Ra1 43.Kf3 a3 44.Ra6+ Kg7 45.Kg2 Ra2+ 46.Kh3 Ra1 47.h5
Kh7 48.g4 Kg7 49.Kh4 a2 50.Kg5 Kf7 51.h6 Rb1 52.Ra7+ Kg8 53.Rxa2 Rb5+ 54.Kg6 Rb6+
55.Kh5

Black resigned.

Game 248
Magnus Carlsen 2714
Michael Adams 2729
Khanty-Mansiysk 2007 (4)

White to move

15.Nb3!

A brilliant manoeuvre, having in mind his next move. At first, it is not obvious how White will be
able to activate this piece, which is apparently dominated by the b6-pawn.

717
15...Bc4 16.Na1! Ba2 17.Nxc2 Bxb1 18.Na1

Home, sweet home.

18...Nd5 19.Bd2 e5 20.e3 exd4 21.exd4 Nb8 22.f3 Nc6 23.Bc4 Rd8 24.Kf2 Bf5 25.Nb3 Be6
26.Rc1 f6 27.a4 a5 28.Bc3 Bf7 29.Nd2 Nde7 30.Bf1 Nxd4 31.Re1 Ndc6 32.Nc4 Nd5

33.Rb1!

A ‘mysterious’ rook move.

33.Rc1 is the computer’s move, but White is only slightly better after 33...Nxc3 34.Rxc3 Bxc4
35.Bxc4+ Kf8.

33...Kf8

After the intended 33...Nxc3 34.bxc3 White is close to winning, as the rook is ideally placed on b1;
33...Bg6 is more solid, when White is slightly better after 34.Rd1.

34.Be1! Ke7

718
35.Kg1!

This is the absolute best move, enabling White to reposition the bishop while keeping the king at
home, as it has no active squares. This brilliant series of subtle positional moves is outstanding.

35...Nb8 36.Bf2

Excellent; the bishops are beautifully placed.

... 1-0 (77)

719
 Index of names

(numbers refer to pages)

Aagaard 151
Abdusattorov 277
Adams 41, 139, 174, 226, 315
Adinolfi 126
Afek 163
Agassi 136
Agdestein 23
Agur 31, 33
Alekhine 7, 13, 90, 167
Alekseenko 303
Almasi 131
AlphaZero 223, 256
Amin 207
Amonatov 190
Anand 9, 13-14, 19, 24-26, 30-31, 37-39, 41-42, 62, 83, 102, 144, 146, 148, 150-151, 153-154, 164,
166, 178, 185, 215-216, 224-225, 238-241, 243-244, 249, 251, 254-255
Aronian 19-20, 25, 27, 30-31, 33-34, 42, 94, 130-131, 138, 162, 166, 180, 185, 250-251, 266, 280,
287, 302
Azmaiparashvili 190

Bacrot 173
Baramidze 157
Bartel 179
Beliavsky 173
Benjamin 23, 31
Bernstein 141

720
Bobotsov 284
Bogoljubow 172
Bond 215
Botvinnik 13, 23, 31-32, 94, 197-198, 218, 239
Bronstein 37
Brynell 128

Capablanca 7, 17-18, 32, 62, 90, 102, 127, 140-142, 167, 198, 201, 203, 245, 258, 310
Carlsen,H. 22, 40
Caruana 9, 11-12, 18-21, 25, 27, 33, 35, 41, 57, 83, 118, 145-146, 158, 162, 167, 185, 213-214, 220-
221, 229, 234, 237, 250, 253-254, 267, 285, 287, 290, 306, 310, 314
Cheparinov 192, 291
Chigorin 248
Cuenca Jimenez 292

Deepan 187
De Labourdonnais 136
Ding Liren 26, 33, 42, 241, 294
Dominguez Perez 158
Dorfman 113, 215
Dubov 9-10, 18, 31, 42, 119, 280, 289
Duda 190, 243
Dvoretsky 83, 165

Eljanov 133, 165-166

Esipenko 119
Euwe 170

Fine 32, 269-270

721
Firouzja 34, 76, 102, 119, 243, 247, 277, 303
Fischer 11-13, 20-21, 30-32, 83, 102, 123, 138, 147-148, 174, 183, 187, 190, 195-196, 203, 217, 223,
226-227, 229-230, 234, 238, 240-241, 244, 258, 281, 295
Fressinet 23, 25, 41
Fyllingen 127

Gareyev 16-17
Gashimov 143
Gelfand 18, 27, 39, 44-45, 144, 155, 209, 231, 249, 251
Geller 195
Giri 11, 21-22, 26, 31, 33, 38, 41, 44-45, 94, 165, 204, 238, 269, 272, 282, 293, 295
Glenne 124
Gligoric 24, 223
Granda Zuniga 194
Grandelius 43, 215
Grischuk 21-22, 25-26, 136, 162-163, 197, 228, 248, 260, 262, 303-304
Gudbrandsen 301

Hammer 43, 122, 164, 264-265

Harikrishna 102, 217, 226, 304
Hebden 266
Henri 22
Hole 300
Hou Yifan 227, 254, 265
Howell 13-14, 102, 304
Hracek 129
Huzman 200

Ibrahimoglu 229
Ibrayev 189

722
Illescas 113
Ivanchuk 20, 118, 143, 179, 185, 208-209, 224, 231-233, 275-276

Jakovenko 136-137, 162, 201

Janowski 127
Jobava 113, 155-156, 283
Jussupow 147, 193

Kamsky 182, 238

Kan 18
Karjakin 9-11, 41, 57, 94, 118, 149, 155, 157, 166-167, 173, 175, 205-207, 229, 232, 234-236, 251,
267, 271-272, 289, 307, 311
Karolyi 123, 213, 312
Karpov 7, 13, 15, 18-19, 28, 30-32, 90, 94, 102, 119, 123, 138, 146, 155, 193, 195, 203, 212, 214,
216, 226-227, 234, 239, 243, 258, 263-264, 268-269
Kasparov 7, 13, 15, 17, 19-20, 22-24, 26-27, 31, 40, 44, 83, 102, 119, 138-139, 146, 148-149, 154-
155, 189, 195, 212, 214-216, 219, 226, 234, 243, 304
Kavalek 30
Keres 32, 147-148, 269
Khan 23
Korobov 22
Kortchnoi 20, 30
Kosteniuk 193
Kramnik 11, 13-16, 19-21, 23, 27, 32, 41-44, 57, 83, 113, 140-141, 164, 166, 186, 197-199, 211-212,
215-216, 230, 233, 244, 250, 258, 262, 283-284, 304

l’Ami 34, 142, 193, 227, 243-244

Laqua 297
Larsen 13, 37, 141, 223-224, 242, 263
Lasker 32, 133, 163
Laznicka 276

723
Leko 41, 113, 146, 148, 157, 176
Li Chao 293
Lie 125

Maghsoodloo 245
Malakhov 192
Mamedov 238, 260
Mamedyarov 10, 134, 162, 224, 252, 257, 263, 276, 307
Marin 313
Marshall 311
McDonnell 136
McShane 113
Menase 27
Moradiabadi 295
Morozevich 37, 41, 202
Morphy 30, 190, 278-279
Movsesian 152

Najdorf 24, 33
Nakamura 24, 37, 43, 113, 160, 170-171, 211, 218, 221, 273, 282, 288, 291, 295-296, 304, 310
Neiman,H. 27
Nepomniachtchi 21, 33, 119, 157, 163, 254, 257
Nielsen 23, 164, 268, 277
Ni Hua 27, 113
Nikolic 300

Olzem 122

Papa 126

724
Pelletier 134, 144
Petrosian 13, 31-33, 43, 94, 123, 138, 181, 187, 193, 216, 226, 255, 258, 283-284
Philidor 32
Pichot 306
Piket 139, 148
Polgar 199, 225
Polugaevsky 223, 270, 274-275
Ponomariov 36, 138
Psakhis 266

Radjabov 17-18, 90, 134, 180, 184, 232-233, 258

Rapport 44, 203, 261
Réti 262, 289
Romanovsky 262

Sahraoui 27
Sampras 136
Seirawan 16
Sevian 16
Shipov 178
Shirov 41, 144, 200
Short 35, 181-182, 193, 203
Simeonidis 277
Smyslov 32, 102, 181, 216
So 33, 41, 45, 248, 261, 270, 297, 305-306
Socko 277
Sommerbauer 164
Sorokin 218
Spassky 12, 37, 94, 240, 248
Spiridonov 145

725
Stefansson 36, 298
Steinitz 32
Suba 185
Sveshnikov 40, 215
Svidler 13, 62, 83, 228, 233
Szilagyi 275

Tal 13, 24, 32-33, 37, 44, 203, 216, 270, 278
Tari 246-247
Tarrasch 113, 220, 265
Tissir 128
Tiviakov 132
Tomashevsky 264, 283
Topalov 13, 36, 40-42, 94, 102, 138, 176-178, 189, 216, 221, 226, 258, 267, 302

Unzicker 268
Utegaliyev 278

Vachier-Lagrave 11, 21, 27, 33, 42, 118, 139, 145, 157, 159, 162-163, 196, 254, 257, 282, 295, 302
Vallejo Pons 222, 263
Van Wely 57, 167, 169-170, 172, 252
Vescovi 128
Vladimirov 188

Wagner 261
Wang Hao 210, 222, 236
Wang Yue 202
Wei Yi 205
Werle 186

726
Wojtaszek 35, 45, 246-247, 273, 277-278, 293

Xiong 35, 39, 273, 285

Yu Yangyi 264

727
 Bibliography

Books

Simen Agdestein, How Magnus Carlsen Became the Youngest Chess Grandmaster in the World, New
In Chess 2013
Joel Benjamin, World Champion Chess for Juniors, New In Chess 2020
Bobby Fischer, My 60 Memorable Games, London reprint, Batsford 2009
Zenon Franco Ocampos, Magnus Wins With White, Elk and Ruby Publishing House 2020
Zenon Franco Ocampos, Magnus Wins With Black, Elk and Ruby Publishing House 2020
Tibor Karolyi, Endgame Virtuoso Magnus Carlsen, New In Chess 2018
Emmanuel Neiman and Yochanan Afek, Invisible Chess Moves, New In Chess 2011
Emmanuel Neiman, Tune Your Chess Tactics Antenna, New In Chess 2012
David Smerdon, The Complete Chess Swindler, New In Chess 2020
Andrew Soltis, Magnus Carlsen: 60 Memorable Games, Batsford 2020
Renier Vazquez, Alfonso Romero, Dragan Barlov and Luis Bernal, El secreto de Magnus Carlsen,
Editorial Chessy 2016

728
Table of Contents
Title page 3
Explanation of symbols 4
Foreword 6
Introduction 8
Chapter 1 Style: from Karpov to Tal? 34
Chapter 2 The opening revolution 49
Chapter 3 Attack: inviting everyone to the party 55
Chapter 4 Defence: the preventive counter-attack 90
Chapter 5 Tactics: ‘les petites combinaisons’ 101
Chapter 6 Exchanges: Carlsen’s main positional weapon 141
Chapter 7 Calculation: keeping a clear mind 160
Chapter 8 Planning: when knowledge brings vision 179
Chapter 9 Pawns: perfect technique and new tips 189
Chapter 10 Pieces: the art of going backwards 211
Chapter 11 Endings: breaking the principles 243
Chapter 12 How to win against Magnus Carlsen: the hidden defects? 256
Chapter 13 Games and solutions 262
Games 32-62 313
Games 63-93 376
Games 94-124 433
Games 125-155 491
Games 156-186 547
Games 187-217 600
Games 218-248 656
Index of names 720
Bibliography 728

729