Chess Elo Rating Calculator
Calculate expected score, rating change, and your new Elo after a game. Supports provisional, standard, and master K-factors.
Enter both ratings and select a result.
Chess Elo Rating Calculator: Expected Score and Rating Change
The Elo rating system measures the relative skill of chess players and predicts the outcome of a game between any two of them. This calculator takes both players' ratings, the result, and a K-factor, then returns your expected score, how many points you gain or lose, and your new rating, the way FIDE's Rating Regulations work it out for one game. Because the method is universal, the math is the same whether you play in New York, Istanbul, Madrid, or São Paulo.
D beyond ±400 counts as ±400 for a player rated below 2650 (rule 8.3.1)
Rating change = K × (S - PD), rounded to a whole number, 0.5 away from zero (rule 8.3.4)
S = actual score (win 1, draw 0.5, loss 0)
K = K-factor
Example: 1500 vs 1900 → D = -400, PD = 0.08 (8%)
Win with K=20: 20 × 0.92 = 18.4 → +18 points
Elo's continuous curve for comparison: E = 1 / (1 + 10^((Ropp - Ryou) / 400))
What Is the Elo System?
The Elo system was developed by Arpad Elo, a Hungarian-American physics professor and master-level chess player, and it is the method behind the World Chess Federation's official ratings. It replaced older, less accurate methods with a statistical model: the difference between two ratings predicts the expected outcome. A 400-point gap gives the higher-rated player an expected score of about 91 percent on Elo's continuous curve, and 92 percent in FIDE's own conversion table. FIDE's Rating Regulations print no algebraic formula at all: they define the system through that table, and this calculator reads the table directly. For a player rated below 2650, a gap larger than 400 points counts as 400, so a 700-point gap is scored like a 400-point one.
How the Rating Update Works
After a game, your rating moves toward your actual performance. If you score better than expected, you gain points; if you score worse, you lose them. The size of the change is the K-factor multiplied by the difference between your real score and your expected score, rounded to a whole number with 0.5 going away from zero, so -12.5 becomes -13. Beating a much stronger opponent produces a large gain because the win was unlikely, while beating a much weaker one barely moves your rating because you were expected to win.
Why the K-factor Matters
The K-factor decides how reactive your rating is. FIDE gives new players K=40 until the events they have finished add up to at least 30 games, so their rating moves quickly toward their level. Players under 2400 then use K=20, and a player whose published rating has reached 2400 uses K=10 from then on, even if the rating later drops below 2400. FIDE also keeps K=40 for players under 18 as long as their rating stays below 2300. Choosing the right K-factor is the difference between a rating that adapts fast and one that stays steady.
FIDE, Chess.com, and Lichess
FIDE ratings are the official over-the-board standard and are built on Elo's method. Chess.com rates players with the Glicko system and Lichess with Glicko-2; both keep a rating deviation that measures how confident the system is in your rating, so your online numbers will not match your FIDE rating or each other. Each pool is separate, so compare ratings only within the same platform.
Tips & Recommendations
Use K=40 if you are new, K=20 for most players, and K=10 once you reach 2400. The wrong K makes your rating too jumpy or too slow.
An expected score of 0.5 is a coin flip. Below 0.25 you are a clear underdog, so an upset win is worth a lot of points.
On Elo's curve every 400 points multiplies the odds by ten. In FIDE's table a 200-point edge gives the favorite a 76 percent expected score.
FIDE, Chess.com, and Lichess ratings live in separate pools. Only compare numbers from the same platform.
Frequently Asked Questions
Is Elo the same as a FIDE rating?
FIDE ratings are built on the Elo system, but FIDE applies its own rules for K-factors, rating floors and new players, and it reads the expected score from its own conversion table. This calculator follows that table, the 400-point rule and FIDE's rounding for a single game; it does not apply rules that only exist over a whole rating period, such as the cap of 700 on K multiplied by the number of games.
Why is my Chess.com rating different from my FIDE or Lichess rating?
Each platform runs its own rating pool with its own players and starting values, so the same player often has three different numbers. Lichess runs the Glicko-2 system, which adds a rating deviation, and Chess.com runs Glicko, which also tracks a rating deviation. Numbers are only comparable within the same pool, not across platforms.
What is the K-factor in Elo?
The K-factor controls how much a single game can move your rating. A high K (like 40) makes ratings change quickly, which suits new players whose true strength is still unknown. A low K (like 10) makes ratings stable. FIDE uses K=40 for new players, K=20 for most players under 2400, K=10 once you reach 2400, and K=40 for players under 18 whose rating is still below 2300.
How many games until my rating settles?
FIDE keeps a new player on K=40 until the events they have finished add up to at least 30 games; after that K=20 applies while the rating stays under 2400. There is no fixed number of games after which a rating stops moving: every result still changes it by K times the gap between score and expectation.
What does the expected score mean?
The expected score is the share of the points you are expected to take, between 0 and 1, with a draw counting as half a point. It is not the chance of winning, because draws are part of it; FIDE's table calls it the scoring probability. An expected score of 0.75 means that across many games you would average 75 percent of the points against that opponent. A 200-point gap gives the stronger player 0.76 in FIDE's table.
Can my rating go down even if I win?
No. Winning always increases your rating or leaves it unchanged, never decreases it. However, beating a much weaker opponent gains very little because you were expected to win. Losing to a much weaker opponent costs a lot, while losing to a much stronger one costs little, since the upset was unlikely.
Recent Calculations
No calculations yet