How Elo actually works
An Elo rating is a prediction, not a score. Before a game, the system converts the gap between two ratings into an expected score — the fraction of a point you should average against that opponent:
E = 1 / (1 + 10^((opponent − you) / 400))
A 400-point gap puts the favourite at roughly 91%. After the game, your rating moves by K × (result − E). The result is 1 for a win, 0.5 for a draw, 0 for a loss. Everything else is bookkeeping.
Why the underdog gains more
Because the change is driven by the surprise, not the win. If you were expected to score 0.15 and you win, you collect K × 0.85. If you were expected to score 0.85 and you win anyway, you collect K × 0.15. This is also why beating much weaker opponents is close to worthless, and why losing to them is expensive.
The K-factor
- K = 40 — a player’s first 30 rated games. Ratings move fast while the system is still finding your level.
- K = 20 — most rated players below 2400.
- K = 10 — players who have reached 2400, where the system assumes the rating is already accurate.
One honest caveat
Chess.com and Lichess do not use plain Elo. Both run Glicko-style systems that also track how confident they are in your rating, so a returning player moves more than this calculator predicts and a very active player moves less. The number here is a good estimate of the size and direction of the change, not a guarantee of the exact points.
Rating points follow the mistakes
Knowing what a game is worth does not change the result. If you want the points to move in the right direction, the useful question is where the last ones went — run a free game review on your recent Chess.com or Lichess games and see which moves actually cost you rating, or read how to review your blunders properly.
Frequently asked questions
How is the Elo rating change calculated?
Elo works in two steps. First it turns the rating gap into an expected score: E = 1 / (1 + 10^((opponent − you) / 400)). That is the share of the point you are "supposed" to take. Then the change is K × (actual score − expected score), where the actual score is 1 for a win, 0.5 for a draw, and 0 for a loss. Beating a stronger opponent gains more because your expected score was low.
Which K-factor should I use?
Under FIDE rules, K is 40 for a player’s first 30 rated games, 20 for most rated players below 2400, and 10 once a player has reached 2400. A higher K makes ratings move faster, which is why new accounts swing so much.
Why does Chess.com or Lichess show a different number?
Neither site uses plain Elo. Chess.com and Lichess both run Glicko-style systems, which also track how certain the site is about your rating (the rating deviation). A player returning after months away moves more than this calculator predicts, and a player with a long stable history moves less. Treat the result here as a close estimate, not the exact number you will see.
Does the time control change the calculation?
Not the formula itself, but each time control has its own rating pool. Your bullet, blitz, and rapid ratings are separate numbers, so use the rating from the same time control you are about to play.