Chess.com Move Classifications Explained: Brilliant, Great, Best, and Every Label

You finish a game, hit Game Review, and every move you played gets stamped with a label: Best, Excellent, a red Blunder on move 19, maybe a shiny "Brilliant" if you're lucky. It looks precise. But what do these words actually mean, and which ones should you care about?

This is a plain-language guide to every classification Chess.com puts on your moves — what triggers each one, how they're calculated, and how to turn them into something that actually makes you a better player instead of just a scoreboard you glance at and forget.

How does Chess.com classify moves?#

Chess.com doesn't just compare your move to the engine's top choice and grade the raw difference. It uses an Expected Points model — a measure of how likely you are to win the game, on a scale where 1.00 is winning, 0.00 is losing, and 0.50 is dead equal (Chess.com Help Center).

After each of your moves, it checks how much your expected points changed. Drop a little and it's an Inaccuracy; drop a lot and it's a Blunder. This is why a move can be far from the engine's first choice but still be graded "Excellent" — if it barely changed your winning chances, it barely mattered. The label reflects consequences, not just how the move ranks against Stockfish.

That's also why the same eval swing hurts more in a close game than a lopsided one. Going from +8 to +6 barely moves your win probability, so it's forgiven. Going from +0.3 to -1.5 flips the game, so it gets punished.

What are all the move classifications?#

Here's every label Chess.com uses, ranked. The first group are moves you got right; the second group are the ones that cost you.

The good moves:

• Brilliant (!!) — a strong sacrifice you weren't forced to make. More on this below; it's the rarest and most misunderstood label.
• Great (!) — the only move that keeps you in the game. Miss it and your position falls apart.
• Best — the engine's top choice in the position.
• Excellent — not the very top move, but so close it kept everything that mattered.
• Good — a sound, sensible move that gave up a little something but nothing you'll be punished for.
• Book — a known opening move from theory. It isn't "graded" so much as recognized as established.

The costly moves:

• Inaccuracy (?!) — a small slip. You didn't lose the game, but a clearly better move was available.
• Mistake (?) — a real error that measurably worsened your position.
• Miss — you had a concrete winning chance (usually a tactic) and didn't take it.
• Blunder (??) — a serious error that swung the game, often hanging material or walking into a tactic.

If you want the deeper version of what the numbers behind these labels mean — the +1.5 and -3.0 values on the eval bar — we cover that in chess engine evaluations explained.

What actually makes a move Brilliant?#

A Brilliant move (!!) is the label everyone chases, and almost everyone misunderstands. It isn't awarded for playing the best move. It's awarded for finding a good sacrifice you didn't have to make.

Three things generally have to be true at once: you give up material, the sacrifice is actually sound (Stockfish has to approve), and a safe, non-sacrificial move was also available. That last condition is the one people miss. If giving up the piece is the only good move, it's not Brilliant — it's Great, because you were forced into it. Brilliance requires a choice.

There are a couple of guardrails, too: you usually can't earn a Brilliant when you're already completely winning (the game's decided, so the sacrifice risks nothing), and you shouldn't be losing after it. If you want to actually get more of them, the ideas that trigger Brilliant are the same tactical patterns that show up in puzzles — sacrifices with check, forks, smothered mates. We wrote a whole guide on what a Brilliant move is and how to get one.

The one everyone confuses: Great vs. Brilliant#

Great (!) and Brilliant (!!) feel similar because both are rare and both feel like a compliment. But they mean opposite things about your options.

A Great move is the only good move on the board. The position demanded exactly this — a defensive resource, the one path that holds a draw, the sole move that keeps a winning edge. There was no alternative, so the credit is for finding it under pressure.

A Brilliant move is a great move you chose when a comfortable one existed. The credit is for spotting something sharp and correct that you could have safely skipped.

So: Great = "there was one move and you found it." Brilliant = "there was an easy move and you found the beautiful one instead."

Which labels are actually worth studying?#

Here's the honest part. Most players scroll through Game Review hunting for the green Brilliant and wincing at the red Blunders, then close the tab. That's backwards. The labels that actually improve your chess are Miss and Blunder — and it's why they happened that matters, not the badge itself.

A Miss is the most instructive label on the list. It means the position handed you a concrete chance — a winning tactic, a forced sequence — and you didn't see it. That's a direct readout of a hole in your pattern recognition. Every Miss is a puzzle you failed in a real game.

But a label is a verdict, and a verdict is where learning stops, not starts. "Move 23 was a Blunder" tells you the outcome. It doesn't make you generate the better move yourself — and that effort is the entire mechanism of improvement. This is exactly why we built ChessSolve as an alternative to Game Review: instead of stamping a grade on the move, it shows live Stockfish candidate arrows on the board so you predict first, then check. The classification tells you that you erred; doing the work yourself is how you stop erring.

Why the labels aren't gospel#

The classifications come from a strong engine, but one running at a fixed depth so it can score thousands of games fast. That's fine for a rough summary and misleading on a genuinely sharp position, where a deeper search can change its mind about which move was "best."

So a move stamped Mistake might be perfectly reasonable once the engine looks a few moves further, and a "Best" move in a quiet position might be one of three equally fine choices. The label collapses a rich position into a single word. When one specific moment actually matters — a tactic that might or might not work — you want to analyze it yourself at real depth, not trust a one-click batch verdict.

How should you use classifications to improve?#

Treat the labels as a map of where to look, not the lesson itself. A workflow that actually builds skill:

1. Run the review, but ignore the good badges. Best and Excellent are noise for learning — you already got them right.
2. Jump to every Miss and Blunder. These are your real material.
3. Before you look at the answer, go back one move and re-solve it. Cover the engine line. What should you have played? Commit to a move, then check. That gap between your move and the right one is the whole game.
4. Ask why, not what. "The move was a Blunder" is useless. "I stopped calculating after his recapture and missed the intermezzo" is a lesson you can carry into the next game.

Do that on the two or three worst moments of each game and you'll get far more out of a review than someone who admires their accuracy score and moves on.

Chess.com's move classifications are a genuinely useful summary — a fast way to see where a game turned. But a label is a conclusion, and you don't get better by reading conclusions. You get better by doing the thinking that produced them. Use the badges to find your worst moments, then close the report and work those positions out yourself.

Want to review with live engine arrows on the board instead of a grade in a sidebar? Download ChessSolve — free, works on Chess.com and Lichess.