How to analyse your chess games (a 10-minute method)

You finish a game and there are two popular ways to waste the next ten minutes. You can close the app and queue again, or you can open the engine review and nod along while the eval bar dips, agreeing that the blunder was indeed a blunder, learning nothing.

Every coach says analysing your own games is the highest-value study habit in chess. Almost nobody explains what analysing actually means for an amateur with ten minutes to spare. Here’s a method that fits.

Guess before the engine tells you

Replay the game once with the engine off and mark the moment you think it went wrong. Commit to an answer before you check.

This step feels skippable but it’s actually the whole method. If your guess matches what the engine says, your sense of your own games is well calibrated, and that’s worth knowing. If it doesn’t, you’ve just found out exactly what you can’t see. Either way the engine’s answer will stick now, because it confirms or corrects something you actually believed instead of washing over you.

Find the real turning point

Now turn the engine on and look for the one move where the game actually changed hands. Not every wobble, and not the small inaccuracy on move 9. You’re after the move where a fine position became a losing one. Everything else is noise on a first pass, and ten minutes doesn’t have room for noise.

Interrogate that one position

At the turning point, ask yourself three questions. What did my opponent’s last move threaten? What did I think my move was doing? And what does the engine’s move do that mine didn’t?

The first tells you whether you lost by not looking. The second tells you whether your plan was wrong or just the way you went about it. The third names the thing you couldn’t see. Most of the value of analysis lives in these three answers, and no eval bar can give you any of them.

Leave with one sentence

Finish by writing down a single concrete sentence. Something like “I lost because I pushed the pawn in front of my king while her queen and knight were still pointed at it.” Not “I need to play safer”, which is a mood rather than a finding.

Keep the sentences somewhere. After ten games they start repeating, and whatever repeats is your actual weakness, written in your own words. That list will do more for you than any opening course.

★One game, one moment, three questions, one sentence. Everything else is spectating.

What this replaces

It’s worth saying what this method leaves out, on purpose. Stepping through every move with the engine running mostly produces noise. Reviewing only your losses hides the mistakes in your wins, which are usually the same mistakes, just unpunished. And memorising the engine’s best line is wasted effort, because you’ll never see that exact position again. The method is short precisely because the things it cut don’t transfer to your next game.

The shortcut, if you want one

This loop is what Socratic Chess runs for you. It finds the turning points, asks what you were thinking before it shows you anything, and if you had a different move in mind it grades that move with the engine instead of just flashing the correct answer. It’s the same questions, just automated. Run it on your last loss, it’s free and you don’t need an account.