How to analyze your chess games after you play
Go through the game on your own first, then with the engine, and keep one lesson from each. A routine that takes ten minutes and works at any rating.
Go through the game on your own first, then check it with an engine, and finish with one note on what to do differently next time. That takes about ten minutes a game. It also stops you doing what most of us do after a loss: spot the blunder, groan and click New game, when the game usually went wrong a few moves earlier. Chesslord’s game review does the engine part for your chess.com and Lichess games, for free.
Why analyze your own games at all?
Because your own games show the mistakes you make, not the ones a puzzle book picked. A puzzle tells you there is a tactic. A game shows you the moment you didn’t think to look for one, and those moments tend to come back in the same form. Going over your games is how you notice that before it costs you the next one.
How do you analyze a chess game, step by step?
- Go through it on your own first
- Before the engine, replay the game and stop at the moments you remember thinking hard. Write down what you were worried about and what you wanted to play. Thirty seconds per moment is enough.
- Find where it turned
- Ask when the game stopped being even. Usually it is one move, sometimes two. That is where most of your time should go.
- Check with the engine
- Now turn it on. Look at the moments you marked and at the move where the evaluation swung. Compare what you played, what you considered and what the engine prefers.
- Ask why, not just what
- The engine gives the better move. Your job is to work out what you missed, such as a piece left unguarded, a check you didn’t look at or a plan that was too slow. Name it in a few words.
- Keep one lesson
- One per game. “I don’t look at my opponent’s checks after I move the queen” is a lesson. “Play better in the middlegame” is not.
Should you use an engine to analyze your games?
Yes, but after you’ve thought for yourself. If you start with the engine, you read its moves and nod along, and very little of it sticks. If you start with your own ideas, the engine is answering a question you asked.
Watch the evaluation number more than the best move. A swing of two pawns after one move is worth ten minutes. A difference of 0.2 between your move and the engine’s is not.
What should you look for in a lost game?
| Question | What it tells you |
|---|---|
| Where did the evaluation drop the most? | The move that lost the game |
| Was it a tactic or a slow plan? | Whether to do puzzles or study positions |
| How much time did you have left? | Whether it was a thinking problem or a clock problem |
| Did you leave the opening book early? | Whether your opening needs work |
| Did the same thing happen last week? | Whether this is a habit |
Should you analyze your wins too?
Yes, briefly. A win can hide mistakes your opponent didn’t punish. Look for the moment they could have turned the game around. If you won on a blunder of theirs, you were probably worse before it.
How many games should you analyze?
Not many, but properly. One slow game reviewed carefully teaches more than ten bullet games skimmed. A good rule is every rapid or classical game you lose, plus one or two blitz games a week where something felt off.
How do you see patterns across many games?
A single game shows one mistake. To see the ones you keep making, you need to look at many games together. Chesslord’s insights look at every game it has fetched from your chess.com and Lichess accounts: the openings that score, the tactics you find and miss, the pieces you leave hanging, and the phase of the game where you lose most.
Where do you start?
Add your chess.com or Lichess username to the free game review and pick your last loss. Every move gets a label, from brilliant to blunder, and the key moments of the game are marked, so you know where to look first.
How Chesslord helps
Move-by-move game review