How to Analyse Your Chess Games (And Actually Improve)

You finish a game, open the analysis board, and click through the moves. The engine shows you where you went wrong. You nod. You close the tab. Next game, you make the same mistakes.

This is how most chess players "analyse" their games. It feels productive, but it's almost entirely passive. And passive review doesn't build the skills that prevent mistakes — it just shows you where they happened.

The Difference Between Reviewing and Analysing

Reviewing is clicking through moves with an engine running, noting where the evaluation bar dipped. It's watching someone else's surgery on video. You can see what happened, but you haven't learned to hold the scalpel.

Analysing is actively engaging with critical positions. It means covering the engine evaluation, looking at the position yourself, finding candidate moves, choosing one, and only then checking the engine. It means asking why your move was worse, not just that it was worse.

The difference is effort. Analysis requires you to think. Review lets you watch.

A Practical Analysis Method

Here's a five-step process that actually produces improvement:

Step 1: Play through the game without an engine. Note the moments where you felt uncertain, spent a long time, or were surprised by your opponent's move. These are your critical moments.

Step 2: At each critical moment, write down your thought process. What did you consider? What did you reject? What were you afraid of? This is the data that matters — not the engine's evaluation, but your reasoning.

Step 3: Now turn on the engine. Compare your thought process to the engine's assessment. The gap between what you considered and what the engine found is your training signal.

Step 4: For each significant mistake, understand the pattern. Was it a tactical blind spot (you didn't see the fork)? A positional misunderstanding (you traded the wrong pieces)? A calculation error (you saw the move but miscounted)? Each type of error requires different training. For more on this approach, see how to train on your blunders.

Step 5: Drill the positions. This is where most players stop — and it's where the actual improvement happens. Take the positions where you went wrong and solve them as puzzles, without the engine. Repeat until you can find the correct move in under 15 seconds.

Why Most Analysis Tools Fall Short

Chess.com's game review is excellent — but it's behind a paywall. Free users get one game review per day with limited depth. Lichess offers unlimited free analysis, but the interface requires you to drive the process yourself.

Neither platform takes the critical next step: turning your mistakes into drills.

From Analysis to Active Training

Cassandra bridges this gap. When you connect your Chess.com or Lichess account, we analyse every game with Stockfish and extract the positions where you blundered. Each blunder becomes a puzzle — the position before your mistake, with you finding the correct move.

This closes the loop between analysis and training. You don't just see where you went wrong — you actively practice getting it right. And because the puzzles come from your own games, they target patterns you'll encounter again.

The analysis is free. No subscription required, no daily limits. We run Stockfish on every game because we believe game analysis shouldn't be locked behind a paywall.

How Often to Analyse

Analyse at least one game per day, even if briefly. Frequency beats depth. A player who spends 10 minutes analysing every game improves faster than one who does a deep 2-hour review once a month.

The best routine: play a game, analyse it with the five-step method above, and then spend 10 minutes solving your personalised puzzle bank. This creates a tight feedback loop between play and training.

[Get free game analysis →](/connect)