How to analyse your own chess games, step by step

Analyse your own game before you ask an engine: write down what you were thinking, find the turning points, then check with Game Review and practise the mistakes.

Updated 14 November 2026 · 5 min read

To analyse your own chess game, go through it once without an engine and write down what you were thinking at the important moments. Then mark the two or three places where the game turned, check them with an engine, and compare its answer with your notes. Finish by writing one lesson in a sentence and practising those positions again. It takes 10 to 20 minutes, and the step without the engine is the one that teaches you most.

Why look before the engine?

An engine tells you what the best move was. It can't tell you why you didn't see it: a threat you missed, a capture you didn't count, a plan you never considered, or the clock. Only you know that, and only right after the game, while you still remember your thoughts. If you open the engine first, its arrows replace your memory, and you learn the move instead of the habit that caused the mistake.

Step 1: play through the game without an engine

Do this soon after the game, ideally the same day. Open it from My games or paste the PGN into the Analysis board, and step through it from the start without looking at engine lines or evaluations.

At each moment that felt important, note in a sentence what you were thinking:

• what you expected your opponent to play;
• which moves you considered and why you rejected them;
• where you felt unsure, short of time or surprised.

Paper or a notes app is fine. Be honest: "I didn't look at his checks" is more useful than "I played badly".

Step 2: find the turning points yourself

Before any engine, mark the two or three moments where you think the game changed: material was lost or won, an attack started, you missed a chance, or you ran low on time. For each one, try to answer on your own:

1. What was my opponent threatening?
2. Were there checks, captures or threats I didn't look at, for either side?
3. What was the plan, and did my move fit it?

If you lost, also find the first moment you felt worse. The real mistake is often a few moves before the one that lost material.

Step 3: check with an engine

Now run Game Review. On ChessMess it's free and unlimited: Stockfish 19 runs in your browser, and the review shows accuracy, a label on each move, an evaluation graph and the key moments, with an explanation of what went wrong. Chess game review explained shows how to read each part.

Compare the engine's key moments with your own list:

• Moments you both found: good. Read the explanation and play through the better line until you understand it.
• Moments only the engine found: these are your blind spots. Ask why you didn't notice anything there.
• Moments only you marked: often the position was fine and you were simply worried. That's worth knowing too.

Don't try to understand every engine move. If the better line is long and only makes sense to a computer, note it and move on. Spend your time on the moments where a human could have found the better move.

Step 4: write one lesson

Sum up the game in one or two sentences you can use next time, for example:

• "I moved my queen without checking for knight forks."
• "When I'm ahead, I should trade pieces, not start a new attack."
• "I spent 4 minutes on move 12 and had to rush the endgame."

One clear lesson per game beats ten notes you won't read again. If the same lesson keeps coming back, that's what to work on. The 5-question blunder checklist helps with the most common one: pieces left hanging.

Step 5: practise the positions you got wrong

Knowing the right move once isn't the same as finding it next time. On ChessMess, every Game Review adds your mistakes, missed wins and blunders to the Mistake trainer, which brings each position back after 1, 3, 7, 14, 30 and 60 days until you find the better move every time. For general patterns, add a few rated puzzles on the themes your lessons point to.

A routine that works

Review your losses first, and the games where you felt lost even if you won. Short blitz games are worth a quick look too, but slower games give you more to work with, because you had time to think. See Blitz, rapid or classical.

Common mistakes when analysing

• Starting with the engine. You lose the record of what you were thinking.
• Looking only at the move that lost material. The cause is often earlier.
• Reviewing only your wins. They feel better, but losses teach more.
• Chasing the accuracy number. It's a summary, not the lesson. Compare it only between your own games.
• Never practising afterwards. Without step 5, the same mistake comes back.

Play a game against a bot, then review it →

Questions and answers

How do I analyse my own chess games?

Go through the game once without an engine and write down what you were thinking, mark the two or three turning points yourself, then check them with an engine and compare. Write one lesson from the game and practise the positions you got wrong.

Should I use an engine to analyse my games?

Yes, but after your own look, not before. The engine shows the better move; your own analysis shows why you missed it, which is the part you can fix.

How long should a game analysis take?

For most club games, 10 to 20 minutes: 5 to 10 without the engine, then 5 to 10 with Game Review and writing your lesson.

Which games should I analyse?

Your losses first, then games where you felt lost or short of time. Slower games give you more to learn from than bullet or blitz.

Can I analyse games I played on other sites?

Yes. Copy the game's PGN from the other site and paste it into the ChessMess analysis board, which runs Stockfish 19 in your browser for free.