Why should you analyse your games?
Analysing your own games is very time consuming and can be daunting, so you might be asking yourself why
you should do it. The simple answer is that it's the most effective way to improve your game.
It's so effective because you are working on the opening, middlegame and endgame in (almost) every game analysis.
Additionally you will find your strengths and weaknesses. This will then help you to plan your future study
accordingly to get the most out of the time you spent on studying.
General advice
When analysing your games, you should only look at the engine after you have finished your analysis. This
is often difficult because it is very tempting to see if you missed a winning move. But as soon as you see
a move from the engine you can't unsee it and the solution seems obvious afterwards, even if it is not so
easy.
I also like to write the analysis by hand in a notebook, because when clicking through the game I find
myself not thinking about the "simple" moves. By writing the moves down by hand, I slow
myself down and pay more attention to the quiet moves.
Step 0: Play a long time control game
To analyse a game, you need to play one first. It is important that the game is a long time control game (30 minutes with increment or more) so that you have enough time during the game to think about your moves and decisions. It makes no sense to analyse a game where you had to rely on your intuition most of the time, since you will not learn much from it and you can't change your intuition easily.
Step 1: Write down your thoughts
The first thing you should do is to write down the thoughts you had during the game. In your notes you
should include the lines you calculated, why you made the decisions during the game and also your
general thoughts on how you felt the game was going.
At this stage you should not add anything new to the thoughts you had during the game. The reason is
that you need to know what you thought about during the game to later analyse and improve your thought process.
Any information from after the game will hide your real thought process and is therefore a hindrance for this
task.
This should be done as soon as possible after the game since your thoughts will be fresh in your memory.
The other steps can be done a few days or weeks after the game was played. It might be even better to
wait a bit with your analysis to gain a new perspective.
Step 2: Analyse the critical lines
Now you revisit the lines you calculated during the game and also analyse other moves. During the analysis you should also try many different moves and try to find interesting options for both sides. It's very easy to go through the game and think to yourself that all moves are obvious and natural and therefore there don't exist playable alternatives. But there are often more options than you suspect, so it is always a good idea to stop and look at the position and try to come up with new ideas.
Step 3: Write comments to the game
I think that it's very important to be able to explain the characteristics of positions with words. To get better
at this, I write comments about the game and try to explain how I evaluate the position. Based on these general
thoughts I try to come up with plans for both sides during the analysis.
At this point you should also compare the thoughts you had during the game about the position with your
assessment afterwards. By examining the thoughts you had during the game you can find faults in your thought
process which is an important part of the analysis.
Step 4: Check your analysis with the engine
Only now it's time to check the game and the lines from your analysis with the engine. But you shouldn't
trust the computer blindly. You should question the evaluations and try your own variations against the
engine. The goal is that you can explain why the computer makes certain moves, not with long variations but
with words and you should only need variations to justify the play tactically.
If the engine finds a forcing tactical line which you hadn't seen during the game or your analysis, you should
ask yourself why you missed this tactic. Did
you miss the idea altogether? Did you see the idea but miss a nuance along the way? Or did you see the tactic
but thought that your opponent had a resource which didn't exist in reality? Knowing the answers to these
questions help you to find the areas you should be working on in your study.
Learning from the engine is an important skill to have. It doesn't help your chess if
you say "Oh, this move would have won" and simply move on. You have to take an active role when using the
engine.
Step 5: Write down 3 takeaways from the game
Thinking of three things you can learn from each game is a very effective way to improve your play. Writing these takeaways down makes it also easier to spot a pattern in your games. After a tournament, you can look back at the takeaways from each game and get a sense of your strengths and weaknesses without going through the whole games.
Example Game Analysis
This this was a 45+45 game in the second round of the Lichess4545 League in season 27.
1.d4 Nf6
This was already a surprise since my opponent had only games in the Dutch on Lichess.
2.c4 g6 3.g3
I usually play the Sämisch against the Kings Indian Defense, but I never felt comfortable with it. Since I only prepared for the Dutch I decided to get both of us out of theory.
3...Bg7 4.Bg2 d6 5.Qa4+?!
During the game I thought that this will cause black some trouble, but this isn't the case. I thought that Nbd7 is forced (which is not bad for black), but black can even play 5...Bd7 and white can't win
the b7 pawn with 6.Qb3 Nc6 7.Qxb7?? Nb4 and black has a fork on c2.
Instead of this, I should have gone for a normal developing move like e4, Nf3 or Nc3.
5...Nbd7 6.Nh3?!
During the game I tried to be creative and I thought that the knight should go to f4 anyway, so this route was the quickest. But I didn't realise that the knight will come under attack as happend in the
game.
6.Nf3 would have certainly been a better move.
Engine note: white has lost his advantage after 5.Qa4+ and black is better after 6.Nh3
6...O-O 7.O-O c5
Black has reached a pleasant position due to white's inaccurate play. Black's knights are better than white's and the white queen is out of play on a4 and would be better if it were back on d1.
8.e3?! Nb6 9.Qb3 Be6?!
Black should have probably played 9...cxd4 10.exd4 Be6 11.d5 Bf5 and compared to 10.d5 after the game move, black's bishop has the great f5-square and black can put pressure on the semi open c-file.
10.Nd2?
I decided to go for a pawn sacrifice, because I didn't like the continuation 10.d5 Bxh3 11.Bxh3 during the game, although white's position looks very good.
White has a central space advantage and can't be put under pressure from black on the next moves, so I can finish my development. White can start playing on the queenside and try to create a weakness in black's position. Black in return doesn't have an active plan and they also can't play Ne4 because of Qc2 and after f5 white plays Nd2 and gets rid of the knight.
Engine note: The engine says white is much better and even thinks it's necessary to sacrifice the b-pawn to gain counterplay on the queen side
So I don't know which ghosts I saw during the game that prevented me from going for this position.
10...dxe4 11.Nf4 (11.exd4 Qc8) dxe3 12.Nxe6 exf2+ 13.Rxf2 fxe6
I saw this position when I played Nd2 and I thought that I have certainly some compensation. Currently white is 2 pawns down, but will be able to win one pawn back, so the question is what white has in return for the pawn. White has the bishop pair, one less pawn island and black's d- and e-pawns are weak. But white also has to solve the problem of his undeveloped bishop on c1.
14.Qe3 Qc8 15.Bh3 Kh8! 16.Bxe6 Qc5?
I think that an exchange of queens is in white's favour so black should avoid it.
Engine note: black is much better after 16...Qc6 17.Rb1 (to avoid problems along the a1-h8 diagonal) Nfd7 18.Re2 Nc5 19.Bh3
Black has kicked white's bishop back and took control of the f-file and now Stockfish gives the move 19...d5! with the idea to answer 20.b3 with d4 and 20.cxd5 with Nxd5 and white's king is under heavy attack.
17.Nb3? (17.Qxc5! dxc5 18.Nb3 Nbd7 19.Bf4) Qxe3?? 18.Bxe3
This is the position I was dreaming of, the queens are excahnged and I finally developed my dark squared bishop. Here I feel that I have enough compensation for the pawn since my bishop pair is active, my second rook can come into the game and black's knights don't have any good central squares to go to.
Engine note: Stockfish also assesses the ending as equal
18...Ne4?! 19.Re2 Nf6 20.Rd1 Ne8
Black wants to play Nc7 to get rid of my strong bishop on e6.
21.c5 Nc7?
This gives white a very strong pawn on a7. Black should have gone for 21...Na4 where I intended to play 22.cxd6 Nxd6 23.Bd4 Bxd4+ 24.Nxd4 and white has lost his bishop pair but black has an isolated
e-pawn and the game position should be equal.
22.cxb6 Nxe6 23.bxa7 Nd8
Black's intention is to play Nc6 and take the a7-pawn, so during the game I thought of ways to prevent this. The most obvious solution is 24.Na5 but after 24...Nc6 25.Nxc6 bxc6 26.Bb6 Re8 I wasn't sure how to continue and my bishop seems locked in on b6. But here white has 27.b4! to prevent black from playing c5 and white will also pick up the c-pawn. This would probably have been easier than the game continuation.
24.Nd4
During the game I thought that this move was quite clever, but I didn't realise that black's exchange sacrifice puts up so much resistance.
24...Rxa7 25.Ne6 Rxa2 26.Nxf8 Bxf8
This is a very important phase of the game. White has an exchange for two pawns and black's rook isn't very active (it only attacks the a2-pawn), but has to find a way to break through black's position. The most obvious way into black's position is the c-file, but during the game I only saw 27.Rc1 Nc6 and thought that I can't break through. But I can now switch to the a-file with 28.Rc3 Bg7 29.Ra3! The b2-pawn is immune due to 29...Rxb2?? 30.Ra8+ and white wins.
In the game I only spent 3 minutes here (I had 19 minutes left) and I hadn't thought about the problems I was going to phase deeply enough.
27.Bd4+?!
Engine note: This is a mistake and the game is equal afterwards. The engine gives 27.Rc2 Nc6 28.Rd3 Bg7 29.Rb3 and white's rooks got very active
27...Kg8 28.Rf2 Nc6
Black has coordinated their pieces and white can't break thorugh. In the game I was only thinking about the f-file and wheter I had any tricks there.
29.Bc3 Bg7 30.Bxg7 Kxg7 31.g4 Ra4 32.h3 g5?
This gives white the chance to bring a rook to b5 with 33.Rf5/Rd5 h6 34.Rb5 and it's clear that white has good winning chances in this position. However I only noticed this idea after my next move
in the game.
33.Rdf1? Nef 34.Rf5 h6 35.Rc1 Rc4??
A blunder from black since white can now exchange one pair of rooks (which is generaly desirable when you are an exchange up) and enter with the remaining rook on the b-file. Black should have gone for
a line like 35...Rb4 36.Rc7 Ng6 and it's hard to see how white can make any progress.
36.Rxc4 Nxc4 37.Rb5
Now I knew that I had a winning position. White wins the b7 pawn and then his own b-pawn will be passed and win the game.
37...Kf6 38.Rxb7 e6 39.b4 e5 40.Rc7 Ne3 41.b5 Nd5 42.Rc6 Ke6 43.b6 Nb4 44.Rc8 Kd7 45.b7 Na6 46.Ra8
This is cleaner than 46.b8=Q Nxb8 because white picks up the d-pawn immediately, but both (and more) moves win.
46...Nb8 47.Rxb8 Kc7 48.Rd8 Kxb7 49.Rxd6 1-0
Takeaways
- Don't try to be creative in the opening just to do something different
- Evaluate positions I reach in my calculations more carefully
- Pay close attention to piece activity in the endgame