What ChessRanger is

Analysis runs Stockfish in your browser to a fixed depth. Within that depth it is far stronger than any human player; past it, it is guessing like everyone else. A move marked a mistake is a move the search liked less at that depth, which is usually the same thing as a mistake and occasionally is not. Where a line is genuinely unclear, prefer your own understanding of the position over a number.

The play and puzzle ratings on this site measure how you do against our opponents and our puzzle set. They are calibrated to be useful for tracking your own progress over time, and they are not a FIDE, Lichess, Chess.com or national rating. Do not expect them to line up, in either direction.

The coach explains positions in words. Those explanations are generated, grounded in the position and in your own games, and they are still capable of being confidently wrong about a subtlety. Treat the coach as a strong club player looking over your shoulder rather than as an oracle, and check anything that matters against the engine line.

Openings, famous games, puzzles, endgame drills and training positions are validated against a chess engine before they ship: every move sequence is legal, every position is real, and every puzzle marked as a forced mate has been confirmed to be one. Names, dates, events and annotations are written by people and models, and those can still carry mistakes. If you find one, tell us and we will fix it.

A plan, a streak and a rising number are motivating, and none of them guarantee that you will gain points over the board. What raises a rating is playing, reviewing honestly, and working on what is actually costing you games. This site tries hard to point you at that; it cannot do it for you.

Anything unclear, email hello@chessranger.com. The terms and the privacy policy cover the rest.