What Is a Good Accuracy in Chess? A Rating-by-Rating Guide

"Is 85% accuracy good?" is one of the most common questions in chess, and the honest answer is: it depends almost entirely on the game it came from. Accuracy is not a skill rating. It measures how closely your moves tracked the engine's preferences in one specific game, and some games make that far easier than others.

As a rough guide, club players around 800-1000 often land somewhere in the 60-75% range, 1200-1400 players commonly sit around 70-80%, 1600-1800 players around 75-85%, and 2000+ players frequently post 85% or higher. Treat these as loose bands, not targets - the spread within any rating is wider than the gap between the bands.

The reason those ranges overlap so much is that position type drives accuracy at least as much as skill does. A game that simplifies into a quiet endgame offers many moves where several options are near-equal, so almost anything you play scores well. A sharp, tactical middlegame punishes a single imprecise move heavily. This is why a beginner can post 90% in a short, simple game while a strong player grinding out a complicated win posts 78%.

Game length matters too. Accuracy is an average, so a 20-move game gives each move enormous weight, and one blunder can crater the number. In an 80-move game, the same blunder is diluted by dozens of routine recaptures and king moves that are trivially "best". Short games therefore produce far more extreme accuracy scores in both directions.

Is 100% accuracy possible? Yes, and it is not as remarkable as it sounds. Play a six-move miniature where every move is forced or obvious and you can hit 100% without doing anything impressive. Sustained high accuracy over long, complex games is the meaningful achievement - not a single perfect scoreline.

A word on suspicion: a high accuracy score on its own is not evidence of cheating. Fair players hit 95%+ regularly in short or forcing games. Platforms that investigate cheating look at consistency across many games, timing patterns, and whether a player finds unnatural engine-only moves - not one number from one game.

The way to actually use accuracy is to stop comparing yourself to other people's screenshots and start comparing yourself to yourself. Track your own average across many games in the same time control. If it drifts upward over months, you are genuinely making fewer costly mistakes. That trend is worth far more than any single game's percentage.

If you want to see this on your own games, ChessDeeper reviews any Chess.com or Lichess game with Stockfish and shows the accuracy for both players, along with exactly which moves cost you the most.