Stop Reviewing Games One at a Time — Run Them All Through an Engine at Once

Batch analysis means feeding an engine every game you've played, not clicking through them one by one

Here's the short answer. To analyze many chess games at once, you export a PGN file. A PGN is just the text record of your games. It shows the moves, not board pictures. This file can hold 10 games, 50 games, or even 100 games. Then you run a chess engine (a computer program that calculates the best moves in any position) over the whole file in one pass. The engine checks every move in every game. It scores each one.

The engine scores moves using centipawn loss. A centipawn is one hundredth of a pawn's value. Engines use this small unit to grade how good a move is. Centipawn loss is the gap between the move you played and the move the engine says was best. A small gap means a strong move. A big gap means a weak one.

Stockfish is the engine most tools use for this work. It's free. It's also one of the strongest chess engines in the world. Point it at a stack of games instead of one game, and it stops being a single-game review tool. It becomes a pattern finder.

This matters if you're rated somewhere between 1000 and 1600 and you feel stuck. You've probably done the normal thing. You play a game. You lose it. You click through it move by move on Chess.com or Lichess. You sigh. You close the tab. Then you do that again tomorrow. One game at a time, you only see one mistake at a time. You never see the mistake that shows up again and again.

Getting a PGN export is simple in practice. Both Chess.com and Lichess let you download your game history from your archive page. It comes as a plain text file. You don't need any extra software. That file is your raw material for a batch check.

One blunder is noise. The same mistake showing up in 20 games is your actual weakness

A hung piece in one game does not mean much. Maybe you were tired. Maybe you misclicked. That's noise, not signal. But say a fork (one move that attacks two of your pieces at once, so you can only save one of them) beats you three times in your last month of games. If it beats you in game 4, it's not bad luck yet. If it beats you again in game 11, look closer. If it happens a third time in game 19, that's not luck at all. That's a pattern. And a pattern is something you can actually fix.

This is why batch review beats single-game review. When an engine checks a batch of games, it tags every move on a scale. Here are the labels, from best to worst:

• Brilliant (an excellent, hard-to-find move)
• Great Move
• Best Move
• Good
• Inaccuracy (a small error)
• Mistake (a bigger error)
• Blunder (a big mistake that hands your opponent a large advantage, or the whole game)

Each label comes from centipawn loss, the gap between what you played and what the engine wanted. One Blunder tag is just one bad moment. Twenty Blunder tags, all from the same kind of mistake, are a habit.

A single game also can't show you where in the game you're losing the most points. A full batch can. Good tools track your accuracy across the game in parts. First they check the opening, the early set of moves both sides often know from study. Next comes the middlegame (the middle part of the game, after the opening, when the plans get complicated). Last comes the endgame (the last part of the game, when only a few pieces are left on the board). Good tools also track your average centipawn loss, or ACPL (a single number that shows how far off your moves were, on average, across a whole game), for each of these three parts on its own. A lower ACPL means your moves stayed close to what the engine wanted.

Say your opening ACPL is low and clean. But your middlegame ACPL jumps way up. That tells you something clear. Your problem isn't your opening prep. It's what happens once the position gets messy and you're on your own. Without a phase breakdown like this, you'd never see that split. You'd just feel like you lost, and you'd blame the last thing you remember. That's usually the opening. It's rarely the real cause.

Rapid and blitz can hide two different mistakes — check them separately, not pooled together

If you batch-analyze every game you've ever played in one big pile, you can hide a real problem. Here's how that happens. Say your rapid games, the ones with 10 minutes or more per side, are mostly clean. But your blitz games, the fast ones with 5 minutes or less, are full of blunders. Mix them together in one batch. The average centipawn loss for the whole pile comes out as a medium, so-so number. Nothing jumps out. You'd walk away thinking you have no real weakness. You do have one. It's just hiding inside half your games.

Split the batch by time control instead. Run your rapid games as one group. Run your blitz games as a separate group. Compare the two ACPL numbers side by side. If blitz is much worse than rapid, that's not a knowledge problem. You already know the right moves. You just don't have enough time to find them under pressure. That's a clock problem. The fix is clock training, not more opening study. If both time controls score about the same, the mistake is baked into how you think. It isn't about how fast you think.

Most sites let you filter your game export by time control before you download the PGN, your plain text file of games. Filter first. Then batch-analyze each group on its own.

Your own game archive can block batch analysis before it starts

Here's a real problem that can trip you up before you even run a single check. Some chess database programs store your games in their own private file format. If you import a PGN into one of these programs, it often gets locked as read-only. You can look at your games, but you can't easily pull them back out. You can't feed them into a different tool without converting the file first. The very archive you built to help you can end up trapping your games instead.

Skip that trap. Go straight to the source. Both Chess.com and Lichess let you export a plain PGN of your own games right from your game archive page. You don't need extra software. That file stays clean and simple. It moves easily into any batch tool.

One more honest note, this time on speed. A batch check can run at different search depths. Depth means how many moves ahead the engine looks before it settles on an answer. A deeper search gives you a more accurate read on each move, but it also takes more computer time. That's the real reason a batch of 100 games takes longer to finish than a batch of 10. You're not just waiting on a bigger file. You're waiting on the engine to think harder, game after game.

BlackBox Chess runs this batch check for you and names your #1 recurring mistake for free

Everything above is the manual version. Export a PGN. Run an engine over the whole batch. Sort by time control. Check accuracy by game phase. It works. It's also a lot of steps. BlackBox Chess is a tool built to run that same batch process for you, automatically.

Here's what it is, in plain terms. BlackBox Chess runs Stockfish across your last 10 to 100 games. Then it ranks the mistakes that keep showing up, instead of listing every single move you ever made. You don't dig through move-by-move breakdowns yourself. It sorts the patterns for you. Then it tells you which one is costing you the most rating.

The free version checks your last 10 games with Stockfish. It names your top recurring mistake. There's no account and no card needed. You paste in your games, and it tells you what's actually going wrong.

If you want the deeper version, the paid Full Report costs $10. It covers 100 games instead of 10. It ranks every mistake pattern you have, not just the top one, by how much rating each one is costing you. It pulls 3 real examples straight from your own games so you can see each pattern in context. It also comes with a fix course built around your specific weak spot.

To be clear about what this can and can't tell you: it can name your pattern and show you exactly where it happened in your own games. It can't promise a specific rating gain or a set timeline, and nobody honest can make that promise for you. What it can do is replace hours of clicking through games one at a time with one clear answer: here's the mistake you keep making. If your own review sessions keep ending in 'I don't know, I just lost,' start with the free 10-game check and see what it finds.