How to improve at chess by analyzing one's own games?

Question

Analyzing one's own games can be a useful tool for improving at chess, as it allows players to identify their mistakes and learn from them. However, it can be challenging to accurately assess one's own play and find the best way to improve.

What are some effective techniques for analyzing one's own chess games? How can players avoid the pitfalls of self-analysis, such as being too harsh or lenient on oneself, or overlooking important mistakes? Is it better to analyze one's own games immediately after playing, or is it better to take some time to let the emotions of the game dissipate first?

Answer 1

How can players avoid the pitfalls of self-analysis, such as being too harsh or lenient on oneself, or overlooking important mistakes?

Those are not the major pitfalls of self-analysis. That is using a computer instead of your own brain. There is a place for the engine but it is right at the end of the process.

Here is the ideal, full process:

1. The day after the game start the process. This allows time for the elation/satisfaction/despair (delete as appropriate according to the result) to die down but the game and your thoughts during the game are still fresh enough to remember. If you leave it too long then narratives you tell yourself to make yourself feel better will start to displace what you were really thinking at the time. Get a clean sheet of paper and a pen/pencil because you are going to record the process. That is a lot more permanent than just taking place in your head. Start by trying to identify the key points in the game. These are the points in the game where the evaluation changed or key decisions were made. Analyse the moves around these key points, writing down your analysis, current thoughts and remembered thoughts from the game.
2. Give a copy of the game to a chess-buddy. This could be your coach, if you have one, it could be a club mate at a similar level to you. Ask them to go through the same analysis process and write their thoughts down.
3. Meet up with your chess buddy and compare and discuss your analyses. Both the stuff you agree on and the stuff you disagree on are important learning points (for both of you).
4. Almost finally run the game through an engine to see what it thinks of the game and your analysis and your chess buddy's analysis.
5. Finally reassess to see what you need to learn from the game and the different analyses of the game.

Answer 2

When playing in a tournament, it is generally not possible to analyse one's games immediately afterward. There simply isn't time. Still, it should be done just after the tournament, when you might still be able to remember what you were thinking. "Well, I considered Bxb5 here, but did not like ...Qe5+. Now I see I could have played...." You get the idea.

Do not use a computer.

I cannot stress that enough. You will not learn to look for better moves, and you will not look for your mistakes. You will just let the computer do all the work for you, and you will learn nothing. Only after you have done a thorough analysis yourself should you submit the game to the computer. And put your analysis in, too, to see whether you missed moves in your analysis.

Classify the moves you missed. When one is a novice, they will be simple things such as hanging pieces and missed checks. Then they will move to missed forks, undefended (vulnerable) pieces, and things of that sort. At advanced levels, it will be moves that influence center control, piece mobility, and pawn structure. And it may just be the loss of a tempo that is the difference between a win (or loss) and a draw. Finding such mistakes is, of course, more difficult. When you have a bunch of those, see what types of moves you are most likely to miss, and look for them in future games.

Get a coach

If you can, get a coach to review your analysis. He may or may not be of any more use than the computer, but he will provide accountability.