What a Chess Rating Actually Measures

A chess rating is a prediction, not a score. It estimates how a player will do against another rated player, and it moves after every rated game depending on whether they did better or worse than expected.

The system was designed by Arpad Elo, a Hungarian-American physics professor and chess master, and FIDE adopted it in 1970. Most national federations and online sites use Elo or a close relative of it.

The gap between two ratings is what matters. A player rated 100 points higher is expected to score about 64% against the other. At 200 points the expected score is about 76%.

Beat a stronger player and you gain more than you would for beating a weaker one. Lose to a weaker player and you drop more. A draw against a much stronger opponent still earns points.

How far a rating moves per game depends on the K-factor. It is 40 for new players and for juniors under 18 rated below 2300, so their ratings catch up quickly. It is 20 for players who have never reached 2400, and 10 for those who have.

FIDE keeps separate ratings for classical, rapid and blitz, and publishes all three every month. Our live rankings follow the classical list. Online ratings on Chess.com or Lichess are separate pools and don’t compare directly with FIDE ratings.

In March 2024 FIDE raised the rating floor from 1000 to 1400. Every player rated below 2000 got a one-time increase of 40% of the gap to 2000, so a player at 1500 went up 200 points. The aim was to undo years of deflation at the lower end.

Ratings also set the bar for titles. A grandmaster needs a rating of at least 2500 on top of three tournament norms.