How to Analyze Your Chess Games

Playing more games makes you faster. Reviewing your games is what makes you better. The problem is that most players either skip the review or open an engine, watch the evaluation bar jump around, and close it five minutes later without having learned anything they'll remember. This guide gives you a simple routine that takes ten to twenty minutes per game and turns each loss into a lesson.

Step 1: Write down what you were thinking — right away

Immediately after the game, before you look at anything, spend two minutes on quick notes: where did you feel the game turn? Which moves were you unsure about? What was your plan in the middlegame? Your thoughts during the game are the most valuable part of the analysis, and they fade within an hour. Without them, you can see what went wrong but not why you chose it.

Step 2: Play through the game without an engine

Go through the moves on a board and look for the critical moments — positions where the game could have gone very differently. They usually appear:

• right after the opening, when you left familiar territory;
• when material changed hands, or could have;
• when you spent a long time on a move, or played one instantly in a complicated position;
• when the character of the position changed — queens traded, the king came under attack, a pawn structure was fixed.

At each critical moment, stop and try to find better moves yourself. Use the same checks-captures-threats routine you'd use in a game (see Chess Tactics for Beginners). This is the part that trains your calculation. If you let the engine do it first, you skip the exercise.

An example of a critical moment

The position below comes from a well-known opening trap. After 1.e4 e5 2.Nf3 Nc6 3.Bc4 Nd4 4.Nxe5 Qg5, White has just won a pawn and now faces a decision. The black queen attacks both the knight on e5 and the pawn on g2.

This is exactly the kind of position to stop at when you review a game. The greedy move looks strong: it attacks the queen and the rook. But White's checks-captures-threats check must include Black's replies, and …Qxg2 followed by …Qxe4+ and …Nf3# is a sequence of forcing moves that ends in mate. The stronger path is 5.Bxf7+ Ke7 6.O-O: White gives material back on its own terms and comes out with extra pawns and a safe king. The lesson you would write down is not "don't play Nxf7" but "before grabbing material, check what the opponent's queen can do to my king".

Step 3: Check with the engine

Only now open the engine. Compare its verdict with your notes. You're looking for three things:

1. Moments you found yourself. Good — your review is on track.
2. Moments you missed. Where did the evaluation change a lot that you didn't flag? Go back to that position and try again to find the better move before looking at the engine's answer.
3. Moments that weren't critical. Sometimes the move you worried about was fine. That's useful too — it tells you where your judgment is already good.

Ignore small differences in the evaluation. A change from +0.3 to +0.1 means nothing in a practical game. A change of a full point or more, or a swing from winning to equal, is what deserves your attention.

Step 4: Turn mistakes into lessons

For each real mistake, write one sentence that explains the cause, not just the move. Good lessons sound like this:

• "I didn't look at my opponent's checks after my queen move."
• "I traded into a pawn ending without checking the opposition."
• "I attacked on the side where my opponent was stronger."

After ten games, patterns appear. Maybe half your losses start with a hanging piece, or you always get into time trouble. That pattern tells you what to study next far better than any general training plan.

Step 5: Look at the opening, briefly

Check where you left known theory, and whether the resulting position was good for you. Don't memorise long lines. At beginner and intermediate level, it's enough to know the first five or six moves of your openings and, more importantly, the ideas behind them — where the pieces go and which pawn breaks you're aiming for.

A ten-minute version for busy days

1. Two minutes of notes straight after the game.
2. Five minutes finding the single biggest turning point on your own.
3. Three minutes checking that position with the engine and writing one lesson.

Doing this after every game beats a long, thorough review once a month.

Common mistakes when analysing

• Only reviewing losses. Wins contain mistakes too — your opponent just didn't punish them.
• Trusting the engine without understanding. If you can't explain why the engine's move is better, you won't find it next time.
• Analysing when you're upset. A short break after a bad loss makes the review more honest.

Pick the last game you lost and try the ten-minute version now — then do the same after your next game against the bot or an online opponent.