Chess Elo Calculator

Enter your rating, your opponent's rating and the result, and this calculator gives you the exact rating change under the Elo system — plus your win probability before the game even starts. No signup, no limits, and below the calculator I explain what the numbers actually mean, because as a grandmaster I have watched players misread their own ratings for twenty years.

How the Elo Rating System Works

The Elo system, designed by the physicist Arpad Elo and adopted by FIDE in 1970, rests on one idea: your rating is a prediction machine. The difference between two players' ratings predicts the expected score of a game between them. Everything else follows from that.

The core formula computes your expected score — the average number of points you should score per game, where a win is 1, a draw is ½ and a loss is 0:

E = 1 / (1 + 10^((R_opponent − R_you) / 400))

If you and your opponent have equal ratings, E = 0.5: over many games you should score fifty percent. If you are rated 200 points higher, E ≈ 0.76; 400 points higher, E ≈ 0.91. Note what this does not say: it does not say the 400-points-stronger player wins 91% of games. Draws count as half, so an expected score of 0.91 can be reached by many mixes of wins and draws. That distinction matters more the stronger the players are, because draws become more common as level rises.

After the game, your rating changes by the difference between what you scored and what you were expected to score, scaled by the K-factor:

ΔR = K × (S − E)

Score more than expected and you gain points; score less and you lose them. Beat a much higher-rated player and S − E is large, so the upset pays well. Beat a much lower-rated player and you gain almost nothing — you only confirmed the prediction.

What the K-Factor Means (and Which One Is Yours)

The K-factor is the volume knob on rating changes. Under FIDE rules:

The logic: a new player's rating is a rough guess, so it should move fast toward the truth; a 2500's rating is backed by hundreds of games, so one bad Sunday should not move it much. When I crossed 2400, my rating slowed to K = 10 — beating a peer earned me five points instead of ten. Online platforms use their own variants (Glicko-based systems on chess.com and Lichess behave like a self-adjusting K), which is one reason online numbers move faster than FIDE numbers.

Win Probability by Rating Difference

Reading the expected-score table is the fastest way to develop realistic expectations. From the formula above:

Two honest lessons from that table. First, a 100-point favorite still concedes about a third of the points — upsets are not miracles, they are the system working. Second, nobody is ever a 100% favorite. I have lost rated games to players 400 points below me, and the table says that had to happen eventually.

Online Ratings vs FIDE Ratings

The question I am asked most often: "I'm 1500 on chess.com — what is that in FIDE?" The honest answer is that no exact conversion exists, and any tool that gives you a single converted number is guessing. The pools are different (online pools include millions of casual players; the FIDE pool starts at club level), the rating floors are different, the starting ratings are different, and the time controls you play online are mostly faster than classical chess. As a rough orientation only: club players commonly find their FIDE classical rating lands below their chess.com rapid number and near or below their Lichess number — but the spread between individuals is hundreds of points wide. If you want a FIDE rating, play FIDE-rated tournaments; if you want to track improvement, pick one online pool and watch your trend there. For what the bands mean in practice — what a 1200, 1600 or 2000 actually plays like — see my guide Chess Ratings Explained.

A Worked Example

Say you are rated 1500 with K = 20 and you face a 1700.

Your expected score: E = 1 / (1 + 10^((1700 − 1500)/400)) = 1 / (1 + 10^0.5) ≈ 0.24.

• You win: ΔR = 20 × (1 − 0.24) = +15.2 points.
• You draw: ΔR = 20 × (0.5 − 0.24) = +5.2 points. A draw against a stronger player gains points — the system rewards you for beating the prediction.
• You lose: ΔR = 20 × (0 − 0.24) = −4.8 points. Losing as the underdog costs little, because the loss was expected.

Your opponent's changes mirror yours with the opposite sign. Rating points are never created or destroyed in a game between two K = 20 players — they flow from one player to the other.

How to Actually Gain Rating

The calculator tells you the accounting; it does not tell you how to earn the points. Rating is a consequence, and chasing the number directly — picking weak opponents, dodging strong ones — is the slowest way to raise it. The players I have coached who improved fastest did the opposite: they played slightly stronger opposition, reviewed every loss, and fixed the recurring leak instead of the memorable one. If that is the project you are on, start with How to Get Better at Chess, and when a plateau arrives — it will — this is how you break it.

Frequently Asked Questions

How is a chess Elo rating calculated?

Your rating changes by K × (S − E) after each rated game: K is your K-factor (40, 20 or 10 under FIDE rules), S is your score (1 win, 0.5 draw, 0 loss) and E is your expected score, computed from the rating difference as 1 / (1 + 10^((opponent − you)/400)). Beating expectation gains points; falling short loses them.

How many points do I gain for beating a higher-rated player?

With K = 20, beating an opponent 200 points above you gains about 15 points; 400 points above you, about 18. The maximum possible gain in one game is your full K-factor, which you approach only against enormously stronger opposition.

Is 1200 a good chess rating?

A 1200 FIDE player is already stronger than the large majority of people who know the rules, and 1200 online is a solid casual-player level. Ratings only have meaning inside their own pool — compare yourself to your own last six months, not to someone else's number in a different pool.

Are chess.com ratings the same as FIDE ratings?

No. Different pools, different starting points and mostly different time controls mean the numbers are not interchangeable, and no exact conversion formula exists. Most club players find their FIDE classical rating is lower than their chess.com rapid rating, but the individual spread is far too wide for any converter to be trusted.

Can my rating go down if I draw?

Yes. A draw scores 0.5, so if your expected score was above 0.5 — that is, you were the higher-rated player — a draw costs you points. Draw against someone 200 points below you at K = 20 and you lose about 5 points.