Magnus Carlsen’s

Middlegame Evolution
By

Ivan Sokolov

Quality Chess
[Link]
 Contents
Key to Symbols used 4
Introduction 5

1 A Minor Sacrifice 9
2 Pawn Majority in the Centre 41
3 Hanging Pawns in the Centre 71
4 Mastering the Nimzo-Indian Sämisch 99
5 Attack & Prophylaxis 141
6 Knight Dominates Bishop 159
7 Magnus is Human 171
8 Exercises 235
9 Solutions 243

Game Index 270

Opening Index 273
Game Index by Carlsen’s Opponents 274
Index of Photos 275
Name Index 276
 Introduction
Throughout my chess career, I have always found it interesting to study how the world champions
approach the middlegame, and how greatly the approach differs from champion to champion.

Two years ago, I had the idea to write a series of middlegame books on the world champions
and their respective middlegame approaches – to have players such as Spassky, Fischer, Karpov,
Kasparov, Kramnik, Anand and Carlsen in the series.
Whether this idea will ever be fully realized, I frankly do not know, but I have decided to start
with the current World Champion, Magnus Carlsen.

Initially my plan was to write just one book on Magnus, but during the process I realized that in
that case many interesting middlegame themes would have to be omitted. The plan has changed,
and now there will be two books, the second volume coming out in Spring 2022. While writing
this book, I came to some pretty clear opinions about what Magnus likes and dislikes in the
middlegame.

Strengths:
Space advantage – Magnus loves having a space advantage in general.
Pawn breaks – Magnus loves to have pawn breaks available.
Space advantage in a fixed pawn structure – when holding a space advantage, pawn breaks are
not necessarily needed!
Minor sacrifices – you rarely see Magnus ‘sacrificing the house’, usually he opts for a minor
sacrifice. This minor sacrifice, usually a pawn but occasionally more, sometimes corresponds to
the requirements of the position and sometimes aims to unbalance the opponent.
Strong self-confidence – Magnus is excellent at giving his opponents opportunities to go wrong.
However, to take the decision to unbalance the situation, one’s level of self-confidence needs to be
very high. In this respect Magnus has (in my view) ‘stolen’ a lot from Fischer!
Pattern recognition and knowledge application – Magnus is also excellent at this.
Finding the best chances in inferior/difficult situations – Magnus is often in trouble, but rarely
loses.

Weaknesses (all relative to his high-standard!):

Positions with less space in general, and definitely positions with less space and no pawn breaks.
Passive positions.
Chaotic Tal-like positions. In Tal-like positions, pattern recognition has limited to no value.
A person who handles those types of positions well – and is able to impose them on Magnus –
would hold the keys to the World Championship!
6 Ivan Sokolov – Magnus Carlsen’s Middlegame Evolution

This book consists of nine chapters. Four of these chapters are on standard thematic topics
(pawn majority in the centre, hanging pawns in the centre, Sämisch Nimzo-Indian structure,
knight dominating bishop), while two others are on themes which are slightly less typical (minor
sacrifice, attack and prophylaxis).

The seventh chapter is the hopefully intriguing Magnus is Human, which consists of ten games
that Magnus lost, and the patterns and connections between them.

The final two chapters are a rather large selection of exercises and solutions, which are intended
to put the reader to work.

You may be wondering what will be in Volume 2?

I will mention a few likely chapters:

Material imbalance
Ruy Lopez Breyer
Rossolimo Sicilian
Sicilian attacks
Chaotic Tal-like positions
The World Championship match with Nepomniachtchi!

The upcoming World Championship match with Nepomniachtchi will be fascinating. For the
first time Magnus will meet a pure tactician and risk-taker.
Also, should the match go the distance, Magnus will – for the first time in his career – not be
the favourite in the tie-breaks!

I hope the reader will find this book both instructive and enjoyable. I have certainly learned a
thing or two while writing it.

Ivan Sokolov
Amsterdam, August 2021
Chapter 1
A Minor Sacrifice
 Test Yourself
On this page you will find four positions you can use as exercises. They are mixed and should not
be considered as normal ‘play and win’ exercises, but interesting positions you may enjoy thinking
about before you read the chapter.

 
Ç   Ç  
Æ Æ 
Å   Å  
Ä     Ä    
à   à    
Â    Â  
Á   Á   
À  
 
À  
ÈÉÊËÌÍÎÏ ÈÉÊËÌÍÎÏ
How did Magnus build an initiative? How did Carlsen create
(see page 16) problems for his opponent?
(see page 29)
 
Ç  
Ç    
Æ Æ  
Å   Å    
Ä   Ä    
Ã   Ã    
Â    Â  
Á   Á   
À   
 À   
ÈÉÊËÌÍÎÏ ÈÉÊËÌÍÎÏ
How did Ivanchuk organize his pieces? How should Aronian have defended?
(see page 19) (see page 39)
 Chapter 1 – A Minor Sacrifice 11

Magnus often uses what we will describe as Magnus Carlsen – Richard Rapport
a minor sacrifice (usually a pawn) to achieve
different aims. Wijk aan Zee 2019
Sometimes it’s because he believes this to be
the best move in the position, but often the In this game Magnus will utilize a minor
minor sacrifice is aimed at unbalancing the sacrifice in order to develop an initiative and
position or – even more so – psychologically attack on open files. From the perspective of
unbalancing the opponent! opening theory, I like the plan Magnus goes for:
it looks like an easy way (as no memorization
The three games selected exemplify the of long theoretical lines is required) to achieve
following points. an opening advantage against this often-played
Sicilian line.
In the first game versus Richard Rapport,
Carlsen’s pawn sacrifice was the best move in 1.e4 c5 2.¤f3 e6 3.d4 cxd4 4.¤xd4 ¤c6
the position: by opening files and diagonals, 5.¤c3 £c7 6.g3 a6 7.¥g2 ¤f6 8.0–0 d6
White quickly gets a beautiful and harmonious 
attack. White’s piece harmony throughout
the game makes it all look incredibly easy. I
Ç 
strongly recommend studying this game and Æ 
applying the same method (playing White)
against this line.
Å 
Ä    
In the game against Nepomniachtchi, we have
an exchange sacrifice leading to unbalanced
à   
positions (in which it is easy to go wrong), Â    
while the sacrifice is probably also the Á 
objectively best decision in the position.
À  
In the game versus Aronian, Magnus’s minor ÈÉÊËÌÍÎÏ
sacrifice of a pawn had a purely psychological 9.¤xc6!?
aim! I do not think that Magnus thought his Magnus wants to follow a well-known plan,
15.d5! sacrificial idea was objectively best, but without including the move ¦f1-e1.
but the move looks very surprising and is not
a regular candidate move; it unbalanced the 9.¦e1 is White’s main move. He obtains some
position and certainly unbalanced Aronian! opening advantage with it and I will give some
lines and comments that I think might provide
useful opening knowledge.

a) 9...¥e7 is the old main line for Black,

leading to the following pawn structure:
10.¤xc6 bxc6 11.e5 dxe5 12.¦xe5 We have
seen many top-flight games from this position;
Black’s queenside pawn structure is vulnerable
and White has scored well.
12 Ivan Sokolov – Magnus Carlsen’s Middlegame Evolution

b) If Black prevents the idea we have just seen Where to develop White’s dark-squared
with: bishop? 12.¥f4! White has a strategic
9...¤d7 idea in mind (seen on move 15). 12...¥e7
 13.£d2 0–0 14.¦ad1 e5
Ç  
Æ Ç  
Å  Æ
Ä     Å  
à    Ä    
Â     Ã  
Á  Â    
À    Á 
ÈÉÊËÌÍÎÏ À   
White is probably better by following a plan ÈÉÊËÌÍÎÏ
similar to that of Magnus in our main game. What to trade and what to keep? White
However, White also has an enterprising makes a good strategic trade. 15.¥g5!
option in: ¥xg5 16.£xg5 Black will have problems
10.¤d5!? protecting the d6-weakness. 16...¤f6 17.c4
A typical Sicilian piece sac. However, in c5 18.£d2 ¦ad8 19.¦e3! ¥c6 20.¤c3 £b7
this particular position in GM practice, it 21.b3 Black still has his d6-pawn weakness
has only been seen in one game (meaning to worry about, and no counterplay. White
that it might pay off for readers interested was clearly better and went on to win in
in opening theory preparation to investigate Sutovsky – Movsesian, Ohrid 2001.
this sacrifice further!). Black can quickly get 10...exd5 11.¤xc6 bxc6 12.exd5† ¥e7 13.£g4
into terrible trouble here. 
10.¤xc6 bxc6 11.¤a4 Following a plan Ç 
which is in a way similar to Carlsen’s (White Æ 
pushes c2-c4 and takes space) brings White
an opening advantage. One top-level GM
Å  
game proved quite instructive: 11...¥b7 Ä   
 Ã   
Ç   Â    
Æ Á 
Å  À    
Ä     ÈÉÊËÌÍÎÏ
A critical moment! Black’s best move is not
Ã   obvious.
Â     13...g6
Á  Black’s best was 13...¦g8! and we get
À    complicated play after 14.dxc6 ¤e5 15.£a4
ÈÉÊËÌÍÎÏ ¢f8 16.¥d2. My engine gives zeros but in a
game between humans, I would prefer White.
 Chapter 1 – A Minor Sacrifice 13

14.£d4 ¤e5 15.f4 White can play Safarli’s relatively new idea:
White now gets his piece back and is clearly 11.e5!
better. This temporary pawn sacrifice is unpleasant
15...0–0 16.fxe5 dxe5 for Black. White will regain his pawn at
 some point, while Black will be left worrying
Ç  about his queenside pawn structure.
11.¤a4 with a later c2-c4 pawn push has
Æ   been considered the main move.
Å  11...dxe5 12.£e2 ¥d6 13.b3
Ä    Black cannot prevent White from collecting
à     the e5-pawn, after which White will have an
Â     advantage.
13...¤d5 14.¥b2
Á 
À     
ÈÉÊËÌÍÎÏ Ç  
17.£c3 Æ 
Even stronger for White was 17.¦xe5! ¥f6 Å 
(17...¥c5 loses to 18.£xc5 £xe5 19.¥d2) Ä   
18.d6 with a large, even winning, advantage. Ã    
17...¦b8 18.£xc6 ¥c5† 19.¢h1 £a5 20.¥h6
¦b6 21.£c7 ¥d6 22.£c4
Â   
White is still better, however the nature Á
of the pawn structure offers Black hopes of À    
counterplay. Black indeed turned the tables ÈÉÊËÌÍÎÏ
and won in Negi – Grachev, St Petersburg 14...0–0
2012. 14...f6, trying to hold on to his extra
pawn, is not a good idea for Black, as after
c) 9...¥d7 10.¤xc6 bxc6 15.¤e4 ¥e7 16.c4 ¤b6 Black is behind
If Black plays 10...¥xc6 White has a fairly in development and the position opens in
standard way to obtain an advantage: White’s favour. 17.f4 is strong and 17.£h5†
11.¤d5 ¥xd5 12.exd5 e5 White is better in is also good for White.
such positions. 15.¤a4 ¦ab8 16.c4 ¤b6
 
Ç   Ç   
Æ  Æ 
Å  Å 
Ä     Ä    
à    Ã  
Â     Â   
Á  Á 
À    À    
ÈÉÊËÌÍÎÏ ÈÉÊËÌÍÎÏ
14 Ivan Sokolov – Magnus Carlsen’s Middlegame Evolution

17.c5! ¤xa4 18.cxd6 £b6 19.¥xe5 14.£e1

White has regained his pawn and boasts a We now see the reason Magnus refrained
large advantage, though Black managed to from putting the rook on e1: White wants
survive in Safarli – Gelashvili, Minsk 2017. to have the possibility of ¥c3-a5. Black does
not have an easy life; White builds on his
9...bxc6 10.¤a4 space advantage, while Black struggles to find
 counterplay.

Ç  The plan chosen by Magnus is logical and

Æ   best fits his style. However, he also had a more
straightforward plan at his disposal: 14.f4
Å  White wants to push e5 and it looks as though
Ä     it works.
Ã   a) 14...¦d8 15.£e2 ¥b7 16.¦ad1 ¤d7 It
Â     makes sense for Black to remove his knight
Á  from White’s tempo-gaining e5-push. 17.g4
¥c6 18.¤c3 White plans g4-g5, which would
À   be unpleasant for Black. 18...¥f6
ÈÉÊËÌÍÎÏ 
10...¦b8 Ç   
Both 10...¥e7 11.c4 and 10...¥b7 11.c4 can Æ 
lead to similar play.
Å 
11.c4 Ä    
Magnus follows his space-advantage strategy à 
and avoids ¦f1-e1, as he plans to place a Â   
different piece on that square. Á 
11...c5 12.b3 ¥e7 13.¥b2 0–0
À  
ÈÉÊËÌÍÎÏ
 Now White has the typical Sicilian central
Ç   pawn break: 19.e5! dxe5 20.¥xc6 £xc6
21.fxe5 ¥xe5 Leading to a material imbalance
Æ   that is favourable to White. (21...¥e7 leads
Å   to a clear White advantage after 22.¤e4
Ä     ¤f8 23.£f3 £e8 24.¤d6.) 22.¦xd7 ¥xh2†
23.£xh2 £xd7 24.¤e4 Black has problems
Ã  with his weak king.
Â    b) In the case of 14...¤d7 White can execute
Á   a favourable trade, similar to that seen in
À   Sutovsky – Movsesian in one of our earlier
comments. 15.£g4 ¥f6 16.¥xf6 ¤xf6
ÈÉÊËÌÍÎÏ
 Chapter 1 – A Minor Sacrifice 15

 
Ç   Ç   
Æ   Æ
Å  
Ä     Å  
Ã Ä    
Â    Ã 
Á   Â   
À    Á  
ÈÉÊËÌÍÎÏ
For the side with less space, it is not a bad À  
idea to trade pieces in general, but here Black ÈÉÊËÌÍÎÏ
will have problems protecting his d6-weakness. 16.£c3
17.£e2 ¥b7 18.¦ad1 White could also have gone for an immediate
16.¥c3 ¥c6 17.¥a5 £b7 18.¤c3 with a
c) 14...¥b7 15.e5! follow-up similar to the game.

Ç    16...¥f6
Æ  16...e5 is the kind of move that (for obvious
reasons) Richard wanted to delay for as long as
Å   possible, because after 17.£c2 ¥c6 18.¤c3 a5
Ä     19.a4 Black has an ugly, passive position.
Ã  
Â    17.£d2 ¥e7 18.£c3
Á   It takes Magnus a move repetition to figure
out that 18.f4! is the way forward for White.
À  
ÈÉÊËÌÍÎÏ 18...¥f6 19.£d2 ¥e7
15...¤e8 (15...¤d7? leads to a winning
advantage for White: 16.exd6 ¥xd6 17.¥xb7 
¦xb7 18.£g4 e5 19.¦ad1) 16.¥xb7 £xb7 Ç   
17.£e2 White has good attacking prospects,
while Black has no counterplay. Æ
Å  
14...¤d7 15.¦d1 ¥b7
Ä    
Ã 
Â   
Á  
À  
ÈÉÊËÌÍÎÏ
16 Ivan Sokolov – Magnus Carlsen’s Middlegame Evolution

20.f4!

21.e5 is a threat and Black is now forced to Ç  
play the move he wanted to avoid. Æ
20...e5 Å  
20...¥c6? loses to 21.e5; while 20...¦fd8? Ä    
runs into 21.¥c3.
à  
21.¥c3 Â   
Magnus wants to bring his bishop to a5,
controlling the d-file.
Á  
À  
The simple 21.¤c3 gives White an advantage
as well.
ÈÉÊËÌÍÎÏ
Now Magnus goes for a nice minor sacrifice
aimed at using the open files for his rooks.
21...¥c6 22.¥a5 £b7 23.¤c3
 25.e5! ¥xg2 26.£xg2 dxe5 27.¤d5! e4
Ç    Or 27...¥f8 28.¥c7 wins an exchange, as
the b8-rook cannot move due to ¤f6† with
Æ a discovered attack on the now-undefended
Å   black queen. 28...¦bc8?? 29.¤e7†

Ä     
à   Ç  
Â    Æ
Á   Å   
À   Ä   
ÈÉÊËÌÍÎÏ Ã  
Black’s only possible counterplay is to attack Â   
White’s e4-pawn and this is what Rapport
does.
Á  
À  
23...exf4 ÈÉÊËÌÍÎÏ
Alternatively, 23...h6 24.¤d5 (also 24.£e3 28.¥c3!
and 24.f5 are decent options) 24...¥xd5 Magnus is not interested in recovering his
25.£xd5 £xd5 26.¦xd5 and White has a minor sacrifice: he is only interested in using
strategically winning position. the files for his rooks, with the bishop on c3
being a tremendous attacking piece in the lines
24.gxf4 ¦fe8 that follow.

28...f6
Time now for White to utilize the g-file.
 Chapter 1 – A Minor Sacrifice 17

29.¢h1 ¢h8 30.¦g1 ¥f8 34...g6

 34...h6 loses to 35.¤f5 when White will
break Black’s pawn defences with a sacrifice.
Ç    White has different ways to win – and your
Æ  engine will show multiple ways – but I like the
following line: 35...¢h7 36.¦h4 ¦b6 37.¦hg4
Å    ¦be6 38.¦4g3 ¢h8 39.¦g6 ¢h7
Ä    
à   Ç   
Â    Æ 
Å 
Á   Ä   
À   Ã  
ÈÉÊËÌÍÎÏ Â  
Another rook is needed and I love the way Á   
Magnus employs the rook lift!
À   
31.¤e3! ÈÉÊËÌÍÎÏ
31.¤xf6 ¤xf6 32.¥xf6 is also winning. 40.¤xh6! gxh6 41.£f5! ¢h8 42.£g4 ¢h7
43.¦g8 and Black is soon mated.
31...£c6 32.¦d5!
White attacks with all his pieces (an absolute 35.¦h4 ¦b6
barrage along the files and across the diagonals!) 
and Black cannot defend his kingside.
Ç   
32...£e6 33.¦h5 £f7 34.£h3 Æ 
 Å  
Ç    Ä    
Æ  Ã  
Å    Â  
Ä    Á   
à   À   
Â   ÈÉÊËÌÍÎÏ
Á    Magnus now performs a swift and clean
execution.
À   
ÈÉÊËÌÍÎÏ 36.f5 ¤e5 37.¤d5 ¦d6
37...g5 38.¦xe4 is also hopeless for Black.
Black is now forced to move one of his
pawns, further weakening his king’s position.
38.fxg6 ¤xg6
18 Ivan Sokolov – Magnus Carlsen’s Middlegame Evolution


Ç   
Æ  
Å  
Ä   
à  
Â  
Á   
À   
ÈÉÊËÌÍÎÏ
39.¥xf6† ¦xf6 40.¦xh7†
The game will be decided by a deadly fork.
1–0