Calculate your new Elo rating after a chess game. Enter your rating, opponent rating, and result to see expected score, rating change, and updated rating.
The Elo rating system, invented by Arpad Elo in the 1960s, is the gold standard for measuring relative skill in chess and has been adopted across many competitive domains from football to video games. Your Elo rating rises when you beat players and falls when you lose, with the magnitude of change depending on the difference between your rating and your opponent's.
Our Chess Elo Rating Calculator lets you input your current rating, your opponent's rating, and the game result (win, loss, or draw) to compute your new rating. It also shows the expected score — your probability of winning based on the rating difference — and how that compares to the actual result, which determines whether you gain or lose points.
Whether you play on FIDE, chess.com, Lichess, or in a local club, understanding the Elo formula helps you set realistic improvement goals, evaluate whether an upset victory or loss was "expected," and choose opponents that will challenge you at the right level.
Knowing how the Elo system works demystifies your rating trajectory. You'll understand why beating a much higher-rated player earns many points, why losing to a lower-rated player costs heavily, and why draws against stronger opponents are actually rating-gaining events. It's also a practical tool for tournament players who want to preview rating changes across different scenarios before or after a competition.
Expected Score: E = 1 / (1 + 10^((R_opponent − R_player) / 400)). New Rating: R_new = R_old + K × (S − E), where S = actual score (1 for win, 0.5 for draw, 0 for loss), K = development factor (10, 20, or 40). The 400 divisor means a 200-point difference gives the higher-rated player approximately a 76% expected win rate.
Result: New rating: 1515 (+15)
Your expected score against a 1700-rated opponent: E = 1 / (1 + 10^((1700−1500)/400)) = 1 / (1 + 10^0.5) = 1 / (1 + 3.162) = 0.240 (24% expected). You won (S = 1), so rating change = 20 × (1 − 0.240) = 20 × 0.760 = +15.2 → new rating 1515. An upset win against a player 200 points above you gains a large amount because the expected score was low.
Arpad Elo, a Hungarian-American physics professor and chess master, developed the Elo rating system in the 1960s for the United States Chess Federation. FIDE adopted it in 1970. The system replaced the older Harkness system and has since been adapted for tennis, football (FIFA), Scrabble, Go, esports, and many other competitive activities.
The expected score formula uses a logistic curve with base 10. The choice of 10 (rather than the natural base e) and the 400 scaling constant were pragmatic decisions to make rating differences intuitive. A 400-point gap means 10:1 odds (91% vs 9%). A 200-point gap means roughly 3:1 odds (76% vs 24%). Equal ratings give 50-50 expected results.
The K-factor is the most debated parameter in the Elo system. A high K makes ratings volatile but responsive to genuine improvement. A low K keeps ratings stable but can leave improving players underrated for extended periods. FIDE's tiered approach (K=40/20/10) is a compromise, but online platforms use various alternatives including dynamic K-factors that change based on rating confidence.
The Elo system assumes performance follows a specific distribution and that all games are equally meaningful. It doesn't account for playing conditions, time control, fatigue in multi-round events, or the home/colour advantage. More sophisticated systems like Glicko (used by Lichess) add a "rating deviation" measure to address confidence intervals.
Beginner: <1200; Club player: 1200–1600; Strong club: 1600–2000; Expert: 2000–2200; FIDE Master: 2300; International Master: 2400; Grandmaster: 2500+; Super GM: 2700+; World Champion level: 2800+. The average rated player on major online platforms is around 1000–1200.
The K-factor controls how sensitive your rating is to a single game result. A higher K (like 40) means your rating changes quickly, suitable for new players whose true strength is still being discovered. A lower K (like 10) means small changes per game, appropriate for established players whose rating is already well-calibrated.
In theory, the Elo formula can produce ratings below zero. In practice, most organisations set a floor (FIDE's minimum published rating is 1000, chess.com uses 100). Provisional ratings may vary, but established ratings rarely drop that low.
Expected score accounts for draws. If your expected score is 0.60, it means you're expected to score 0.60 points per game on average — this could be 60% wins and 0% draws, or 40% wins and 40% draws, etc. It's a combined measure of win and draw probabilities.
The 400 constant was chosen by Arpad Elo so that a 200-point difference corresponds approximately to a 76% expected win rate, which matched observed outcomes in chess tournaments of the era. Different values of this constant would rescale the rating range but not change the underlying logic.
The basic Elo formula does not account for colour. White has a slight statistical advantage (about 55% win rate at top levels), but the Elo system treats all games equally. Some advanced systems (like those used by chess.com) make minor adjustments for colour.
Rating inflation occurs when the average rating in a pool rises over time, making it easier to achieve high numbers. Rating deflation is the opposite. FIDE has debated this topic extensively. The introduction of K=40 for new players and floor ratings has contributed to some inflation in recent decades.