Is 88% Accuracy Good in Chess? What 80%, 92% and 98% Really Mean

Updated November 2026 · 7 min read · Every position checked with Stockfish

Accuracy is a single number that squeezes a whole game into one score. It is useful, but it is easy to misread. A 90% can hide the blunder that lost you the game, and a 75% can be a fine result in a sharp, messy position. This guide gives you a quick verdict for the common scores, then a five-minute method to work out whether your number means you actually played well, and what to do with it.

1. Get the quick verdict for your score

Accuracy measures how close your moves were to the engine's choices, weighted by how much each move changed your winning chances. It is not a percentage of 'correct' moves. Chess.com and Lichess calculate it differently, so never compare a Chess.com score with a Lichess one. For the detail, see what accuracy actually means in chess.

As a rule of thumb for club and online players:

• 80%: a solid game. You probably made one or two real mistakes but mostly kept control.
• 88%: very good for anyone under about 1800. In a long game, that is a strong performance.
• 92%: excellent. Worth checking whether the game was short or one-sided.
• 98%: almost never a 'normal' game. Expect a miniature, a lot of book moves or an opponent who handed you the win.

These are rough bands, not rating thresholds. The rest of this guide is about checking whether your score deserves the label.

2. Check how many moves you actually had to think about

Accuracy is an average, so short games distort it. If the game lasted 15 moves and ten of them were opening theory, the score is mostly measuring your memory of the first few moves, not your chess.

In the shortest possible game, Black plays two moves and wins. Any accuracy number for Black here is based on almost nothing. The same thing, less extreme, happens in any game where your opponent resigns early.

A quick test: count the moves after the opening where you had a genuine decision to make. If that number is small, treat the score as a rough hint, not a verdict.

3. Look at the evaluation graph, not just the score

Open the game review and look at the graph. You are asking one question: when was this game decided?

Once your opponent gives away their queen, the position becomes one-sided. Every later move is easier: you trade pieces, avoid tricks and convert. Engines also judge many moves in a completely winning position as roughly equal, so your accuracy tends to drift upwards. A 92% earned after the opponent blundered on move five tells you much less than an 85% in a game that stayed balanced until the endgame.

The reverse also holds. A sharp, tactical game where both sides are walking a tightrope will usually produce lower scores for everyone, even when you played well.

4. Compare the score with your own baseline

A universal 'good accuracy' number is less useful than your own average. Look at your last 20–30 games in one time control and note your typical score. That is your baseline.

• A game well above your baseline, in a full-length game, is a genuinely good performance.
• A game well below it is worth reviewing, because something broke: time trouble, a missed tactic or an opening you did not understand.
• Blitz and rapid scores should be tracked separately. Faster games produce lower accuracy for almost everyone.

For typical ranges by rating, see what counts as a good accuracy on Chess.com.

How Chessy helps: connect your Chess.com or Lichess username and Chessy imports and analyses your games with Stockfish, then shows your accuracy over time and blunder rate per time control, so you can see your baseline instead of guessing it.

5. Check your opponent's score and the gap

Your accuracy only makes sense next to your opponent's. If you scored 88% and they scored 60%, the game was probably decided by their mistakes, and your number may flatter you. If you scored 82% and they scored 85% and you still won, you likely made the better practical decisions at the critical moment.

This classic trap ends the game on move four. The winner's score says little about their skill: the loser walked into a known pattern. When your opponent collapses early, take the point, but do not take the accuracy number as proof you played brilliantly.

6. Count your blunders and mistakes, not the percentage

The most useful line in a game review is not the accuracy. It is the count of blunders and mistakes. One blunder can drop you from winning to losing while barely moving your average, because it is one move among forty.

After each game, write down:

• How many blunders and mistakes you made.
• The move number of each one.
• What kind of error it was: a hanging piece, a missed tactic, a time-trouble move, an endgame slip.

If you are unsure what separates these labels, read inaccuracy vs mistake vs blunder. Over a few weeks, the same type of error will keep coming back. That pattern, not your accuracy, tells you what to train.

How Chessy helps: Chessy turns your own blunders into puzzles, so you practise the exact positions you got wrong, and the weekly AI coaching report summarises what improved and what keeps leaking. For a fuller routine, see how to analyse your chess games.

Common mistakes

• Treating one high score as proof you have improved. A single 92% after a short or one-sided game proves very little. Look at your average over many games. Improvement shows up as a rising baseline and fewer blunders, not one great number.
• Comparing Chess.com and Lichess accuracy. The two sites use different formulas. A score on one is not directly comparable with a score on the other, so compare games from the same site only.
• Ignoring the result. Accuracy does not care who won. If you scored 88% and lost, find the single move where the evaluation swung. That move matters more than the other thirty-nine.
• Chasing accuracy instead of good decisions. Playing only 'safe' moves to protect your score can make you passive. In practical games, a sound attacking plan that sets problems often wins more points than a tidy, high-accuracy draw.
• Judging blitz by rapid standards. Faster time controls produce lower accuracy for nearly everyone. A blitz score below your rapid average is normal, not a sign that you are getting worse.

FAQ

Your accuracy score is a starting point, not a grade. Check the game length, the evaluation graph and your opponent's score, then compare the number with your own baseline. Most of all, count your blunders and find the patterns behind them. Fix those, and both your accuracy and your rating will rise without you having to chase either.

Sources

• Chess.com Support: What is Accuracy? — How Chess.com defines and presents accuracy in game review
• Lichess: how accuracy is computed — Why Lichess accuracy is calculated differently and is not comparable with Chess.com
• Chess Programming Wiki: Stockfish — Background on the engine that evaluates moves for accuracy scores
• Wikipedia: Blunder (chess) — Definition of a blunder and why single errors decide games