How to Analyze Your Chess Games — Step by Step

Your memory is most accurate in the 10 minutes after a game ends. Before you look at an engine or replay the moves, write two sentences: where do you think you went wrong, and what was the key decision point? It can be a note in a document, a phone note, anything.

This is a habit grandmasters use. When you later compare your impression to what the engine says, you'll learn whether your in-game assessment is calibrated. Players who skip this step often have a blind spot — they think they understand their mistakes but are actually guessing.

If you lost, what was the last moment you felt you had a reasonable position? If you won, did you win because you played well or because your opponent blundered? Honest answers to these questions are more valuable than any engine output.

Replay the game in your head or from memory and identify the move you're least happy with. Not just the blunder — the move where you sensed something was off but played it anyway. What were you thinking? What did you miss?

This step is where the real learning happens. If you go straight to engine analysis, Stockfish will tell you that Nxe5 was a blunder, but it won't tell you why you played it. You already know why — you just need to articulate it. Was it time pressure? Did you miss a piece was hanging? Did you calculate one move deep when you needed two?

Give yourself 3–5 minutes of honest self-analysis before touching any tool. Write down your candidate mistake and your best guess at the correct move. Then you'll have something specific to test against the engine.

PGN (Portable Game Notation) is the standard format for chess games — a plain text file that records every move. Both Chess.com and Lichess let you export any game as PGN for free.

On Chess.com: go to your game, click the menu (three dots), select 'Download PGN'. On Lichess: open the game, click 'Export → PGN'. The downloaded file or copied text is what you'll load into your analysis tool.

You can also add comments and variations directly in the PGN text. Comments go in curly braces {like this} after a move. If you want to annotate your own impressions before running the engine, edit the PGN text before pasting it into a viewer.

Load your PGN into a viewer and walk through the game from move 1. Don't rush to the mistake — replay the whole game at a medium pace, especially the opening and middlegame. Seeing the full arc of the position often reveals a slow-motion misunderstanding you missed in the moment.

For each critical position, ask: what was the threat my opponent just made? Did I answer it, or did I play my own plan and ignore it? Most losses under 1600 trace back to missing or ignoring an opponent's threat for 1–3 moves in a row.

The chess.lc PGN viewer lets you paste any PGN, replay with keyboard arrows, flip the board, and copy the FEN of any position for instant Stockfish handoff. No account, no upload limit.

A common mistake is turning on engine analysis for every move. This produces a wall of arrows and evaluations that's impossible to learn from. Instead, identify 3–5 positions where you were unsure or where something went wrong, and analyze only those.

For each critical position: copy the FEN from your PGN viewer, paste it into a Stockfish analysis tool, and see what the engine recommends. The key question isn't just 'what was the best move?' — it's 'why is that move better than what I played, and would I recognize this pattern in a future game?'

Chess.rodeo gives you full Stockfish engine strength for free — no account required, no paywall, no move limit. It also accepts FEN directly in the URL if you want to send a specific position straight from the PGN viewer.

Not all mistakes are the same. A tactical blunder (missing a fork) has a different cause than a positional mistake (trading a good bishop for a bad knight) or a time management error (spending 10 minutes on a normal developing move). Track which type you make most often.

After 10–20 analyzed games, you'll see a pattern. If most of your losses come from tactical blunders, more daily tactics work is the fix. If you keep getting outplayed in equal-looking endgames, targeted endgame study will pay off faster than anything else. If you repeatedly misplay a specific opening, 30 minutes on that opening's plans beats 30 more minutes of theory.

Keep a simple error log: game date, type (tactical/positional/endgame/time), and a one-sentence description. After a month you'll have a personalized curriculum — worth more than any generic improvement plan.