How to Analyze Your Chess Games
Most players either skip game review entirely or skim an eval bar for a minute and close the tab. A review that changes your rating looks like this: find the turning point yourself first, check it against the engine, name the pattern, train that pattern, and re-test it in your next games. Five steps, ten to twenty minutes.
Step 1: Guess the turning point before you turn on the engine
Immediately after the game — while it is still in your head — decide where the game turned and why. Write it down in one line. This is active recall, and it is the step everyone skips because the engine is right there. If your guess is wrong, the gap between what you felt and what happened is itself the lesson: it tells you your internal evaluation is off, which is exactly what training should fix.
Step 2: Run the real analysis
Now check your guess. On Lichess this is free and unlimited; on Chess.com, Game Review gives you the eval graph and move labels. Compare your guessed turning point with the first big evaluation swing. Agreeing with the engine is less important than noticing where your feel diverged from the truth.
Step 3: Name the pattern, not the blunder
One-off blunders are noise. The signal is repetition: you hang pieces to knights on the third move of a combination, or under time pressure, or in a specific opening structure. Group your mistakes by theme — hanging pieces, pins you missed, forks you walked into, panic in equal endgames — and you will usually find that two or three motifs cause most of your losses. Our tactics map is organized around exactly those motifs.
Step 4: Spend the training time you just earned
A review that ends without training is entertainment. Take the motif you found and drill it for ten focused minutes: the matching Lichess puzzle theme, or a study on the structure that keeps hurting you. Resist the urge to start a new opening — the opening was rarely the reason the game was lost.
Step 5: Re-test and log
Over your next five games, watch for the motif you trained. When it appears and you handle it, that is the review paying off; when it appears and you don't, you know what to drill again. Keeping a one-line log per game (result, motif, one lesson) turns scattered reviews into a record of what is actually holding you back.
Where Chessmasters fits
Steps 2 through 4 are mechanical, and they are what Chessmasters automates: paste the finished game and the report tells you where it turned, names the patterns, and hands you the focus areas with training links. Your part — the guessing in step 1 and the training in step 4 — stays yours, which is the point. See what a report looks like before you try it. The first analysis is free.
Frequently asked questions
How many games should I analyze?
One or two deeply beat ten casually. Pick the games you felt unsure about — losses and draws usually teach more than wins — and give each one a real pass through the five steps.
Should I analyze my wins too?
Occasionally. Wins are worth reviewing when they felt shaky — a won game with three blunders is more instructive than a clean loss, because you got away with mistakes you could repeat under pressure.
How long should a review take?
Ten to twenty minutes per game is plenty if you follow the steps: a few minutes finding your turning point, a few on the engine truth, and most of the time on naming the pattern and training it.