How to analyse grandmaster games.

Oct 15, 2017, 10:23 PM |

      When I was a kid I had the luck to attend a very interesting lecture.The lecturer described a method to analyse grandmaster games without the use of engine that made quite an impression to me.I decided to share it because I think it will help a lot of you.

   First play the game  3 or 4 times and try to understand why the reasons of the result.

    Why white(or Black) won?Which was the key element of the position that gave him the winning advantage?Determining the reasons of the win is the most important thing in an analysis.Careful in this part because it is not enough to say "white won because he attacked the king".That's too easy.Which was the elements that made the attack possible?

Better placed pieces?Pawn majority?Weaknesses in front of opponent's king?     

    Once this is determined , the second thing to do is determine the plan that exploited the advantage that lead to the win.Now you need to play the moves slow and try to understand them.Exchanges are the first thing you need to examine.Which pieces the winning side tried to exchange and why?Try to find the reasons.Exchanges reveal critical and very important aspects of the position and help you gain a deeper understanding.

     Another thing to do is group the moves that have a common goal(piece manoeuvres combined with pawn advances or exchanges).Always have in mind that  a winning advantage is the accumulation of small advantages acquired  via a number of manoeuvres that have a specific goal.Your mission , if you choose to accept it , is to find the manoeuvres and the goals.

          Once you have done all these , comes the 3rd part.You try to recommend possible defenses for the defending side.What he could do better?What should he prevent or avoid?Certainly the most difficult part.

                  That's it.

Not all games are good for analysis for all levels.If you have never done this then consider yourself a beginner in analysis(your rating doesn't matter).You must start with games that demonstrate simple concepts(for example isolated pawn) that are exploited with typical plans.Avoid draws although they are many times more instructive than wins.You want to see Crime(mistakes) and Punishment(exploitation).That is the best way to learn and since the games are millions you can easily find a lot of very instructive wins.Later you can try draws and more complicated middlegame concepts. 

     Let's see an example:

I will use the very instructive game Pilnik-Geller from the 1955 Gothenburg Interzonal.


If we play the above game 3-4 times fast it becomes obvious already from the first time that Black won because of his k-side attack.That is not enough.We also need to answer what created the attack , what was Black's superiority on the k-side.Clearly it was the 3 2 majority on k-side and the very well placed knight on e5.

     That ends part 1.

We need now to find out which were the key moves/plans that created the winning conditions.

    The position seems difficult for white after  24...f4 and Black's game seems like it is on auto pilot: The knight on e5 , the queen on h4 , the rooks on f-file and the h5-g5-g4-f3 advances tear k-side apart.How much simpler than that can be?Of course Geller makes it seem very simple but it isn't easy  to create a position where everything seems so simple.

    So how did Geller "created" that position.

We need to examine exchanges first.

    Exchanges are the most critical part of any game played in high level.They reveal which are the critical pieces and which are the critical squares.If you don't understand that , you understand nothing.

1) 11...Bxd5 12.exd5


A very important exchange that creates 2 imbalancies.One imbalance is on pawn structure and the other one is on the quality of the pieces.We see Black gaining a central majority(2 to 1) and a k-side majority(4 to 3) but it is difficult to see how he will exploit it.As compensation White gets a q-side majority(4 to 3) and the bishop pair.Black neutralised both of them and it is interesting to see how.It is important here to note thought that whenever a light squared bishop exchanges with a knight , the side that has the knights has an advantage on the darksquares since it has 3 light pieces that control them as opposed to 2 pieces that Black has.Since we have played the game already , we noticed the strong unopposed knight on e5.This was the key exchange that created the necessary conditions for that strong knight.

 2)16.Nxc5 bxc5


What do you think about this exchange?

3)23.Bxf6 Qxf6.

A pretty much self explained exchange that allowed Black to dominate the board with an unopposed knight on e5. 


4)32.Rxf4 Rxf4 33.Rxf4 Rxf4 34.g3

   This sequence starts with a pseudo-sacrifice of the rook and allows white to exchange rooks.White hoped that with the rooks out of the board it will be easier to defend.It didn't work but the combination reveals White's last hope to survive and most important defensive resource: g3.


       Now let's examine the manoeuvres:


Black obviously tries to prevent or delay White's q-pawn majority advance on q-side and at the same time relocates his knight in a very good square.This is something worth remembering as it is very typical and appears a lot in Sicilian defense.


A lot of very important moves here.This is the last preparatory manoeuvre.After that the pawn sacrifice and the k-side attack follows.

   This manoeuvre involves the most critical light piece of the game, the knight that will be placed on e5.

The move 17...Nd7 is self explanatory.Geller liberates f-pawn and prepares to install the knight on e5.

The move 18...a4 though seems not to be part of his plan.Why he felt that it was necessary?

18...a4 comes as an answer to 18.Bd1 and that gives us a clue.Is it possible that Black wanted to prevent 19.Ba4?If yes why?

Geller wanted to prevent the exchange of the knight after  18.Bd1 f5 19.Ba4 Bf6 20.Bxd7 or the activation of the bishop after 20.Bc6. 

 The rest of the moves are self explanatory: 19...f5 was necessary to prepare e4 ,20...g6 protected f5 so that 21...Bf6 is possible and 21...Bf6 prepared e4 and the exchange of dark squared bishops.


These two moves haven't been played  in succession but it is important to put them together as they are the most important part of Black's plan.

22...e4 vacates e5 for the bishop and 24...f4 creates the k-side pawn majority.

This is a positional pawn sacrifice that worths to remember.


First the queen goes to h4 so that g5 follows without blocking her out.Then the rooks double on f-file(with 28...g5 in between) and then the pawns advance.The move order is of course important but this will be examined in the analysis of the game.It is very important here to understand that white can escape with a well timed g3.Note that white manoeuvres his queen to c3 to support that advance.It is white's most important defensive resource.That  advance Black must try to prevent.

      We are ready now to do the analysis of the game:

          A very instructive game.What did it teach us?

It taught us  quite well the pawn structure that occurs in Sicilian after ...Bxd5 e4xd5(a very common exchange).

It taught us what we must do as Black.It taught us that the bishop pair is not important in this pawn structure and knights are. better.

But there is one more lesson that can teach us.

What is the famous grandmaster opening preparation?

Nothing more than analysing your game , find out what went wrong and fix it.

A few months after the game with Geller , Pilnik played the same line against Smyslov.But this time he changed some things.It's interesting to see what.

And this is how 13.Qd3 became and remained till today the main move in this line.

Fascinating , isn't it? 


The last thing that we need to do is  examine what Black could do better.

I will leave that to you as an exercise.

Which were White's critical mistakes?

They have been revealed more or less and you should know if you paid attention.


 I will finish with an exercise that wil help you remember the most critical parts of the game.

Can you play like Geller against a defense suggested by the engine?

Be careful because the engine is sneaky.