Why your game review doesn't help you improve

You finish a game, run the review, and get a list back. A few blunders, a couple of inaccuracies, an accuracy percentage. It looks like feedback. It mostly isn’t.

A label tells you a move was bad. It doesn’t tell you why you played it. Maybe you missed a tactic. Maybe you saw it and misjudged the resulting position. Maybe you were low on time and just grabbed the first legal-looking move. Those are three completely different problems, and a report card that says “blunder” treats them as the same thing.

The move isn’t the lesson. Your thinking is.

Improving players don’t get better by memorising which of their past moves were mistakes. They get better by finding the gap between what they thought was happening on the board and what was actually happening. That gap is invisible in a list of symbols. It only shows up in conversation, when someone asks what you were going for.

★A report card tells you the score. It never asks how you got there.

This is also why the labels themselves need to be honest, not just accurate. A move that drops the game from winning to only slightly winning is not the same as a move that loses the game outright, even if both technically “lose material” by some measure. Calling both a blunder teaches you to fear the wrong things. A label should tell you whether the result was actually put at risk, or it isn’t teaching you anything.

What we do instead

Socratic Chess still runs the same kind of engine review under the hood. But the review is a starting point for a conversation, not the end product. You can ask why a move was good or bad, test the move you were actually considering instead of the engine’s top choice, and get an answer checked against the position rather than a guess. If a claim can’t be verified against the board, it doesn’t get said.

The report card was never the point. Understanding your own game was. If you want to try it on a real game, start with your last one.