Rating calculator

Rating change and performance for a whole event

Add a row for each game with your opponent's rating and the result. You get the rating change round by round, the total, your expected score and your performance rating, calculated the way FIDE does it.

Your rating before the event

K-factor

Round	Opponent rating	Result	Change
1			+8.6
2			+2.8
3			-3.8
4			+10.6
5			+4.8

Score

3/5

60%

Expected

1.86

of 5

Rating change

+22.9

new rating ≈ 1823

Performance

1972

linear estimate 1980

Rating change sums K × (score − expected) over every game, which is exactly how FIDE processes a tournament. Performance rating uses the FIDE conversion table from percentage score to rating difference, added to the average opponent rating. A perfect or zero score maps to ±800, so treat performance figures from very short events loosely.

Only one game to check? The Elo calculator shows the change for a win, a draw and a loss side by side, with FIDE K-factor rules built in.

A worked example

A 1800-rated player with K = 20 plays a five-round weekend event. Every number below is computed with the same code as the calculator.

Round	Opponent	Result	Expected	Change
1	1750	Win	0.57	+8.6
2	1900	Draw	0.36	+2.8
3	2050	Loss	0.19	−3.8
4	1820	Win	0.47	+10.6
5	1980	Draw	0.26	+4.8
Total		3/5	1.86	+22.9

Rating change. The player scored 3 against an expected 1.86, so the rating moves by 20 × (3 − 1.86) = +22.9, to about 1823. The same total comes from adding up the per-round changes.

Performance rating. The opponents averaged 1900 and the score was 60%. FIDE's table converts 60% into a rating difference of +72, so the performance is 1900 + 72 = 1972.

Linear estimate. The simpler “rule of 400” adds 800 × (score fraction − 0.5) to the average, giving 1980. It is close to the table in the middle of the range and differs at the extremes.

From percentage to performance

FIDE's conversion table, sampled every ten percent. Add the difference to the average rating of your opponents.

Score	FIDE table	Linear
0%	−800	−400
10%	−366	−320
20%	−240	−240
30%	−149	−160
40%	−72	−80
50%	0	0
60%	+72	+80
70%	+149	+160
80%	+240	+240
90%	+366	+320
100%	+800	+400

Rating change versus performance

The two headline numbers often confuse people because they can point in different directions. Rating change compares what you scored with what your own rating predicted against those particular opponents. Performance rating asks what rating would have made your score exactly par, and ignores your own rating completely.

As a rule of thumb, a performance above your current rating means you gained points and one below it means you lost some. The size of the change depends on K and the number of games, which is why a great performance in a three-round event moves your rating less than a good one over nine rounds.

Short events produce extreme performances. Three wins in three games gives the maximum +800 on FIDE's table whatever the opposition, so judge form over several events, or look at the accuracy of the games themselves with the accuracy calculator.

Rating calculator questions

What exactly is a performance rating?

It is the rating at which your result would have been the expected one. If you scored 50% against opponents averaging 1900, you performed like a 1900 player. Score more and the performance rises above the average; score less and it falls below. It describes one event, not your strength over time.

Why does a perfect score give exactly 800 points above the average?

FIDE's conversion table stops at 100%, where it adds 800 points to the average opponent rating, and at 0%, where it subtracts 800. A perfect score has no upper bound in theory, so the figure is a convention rather than a measurement. The linear estimate shown next to it gives 400 points instead. Treat both loosely for short events.

Does the order of the rounds matter?

No. FIDE rates a tournament using your rating at the start of the rating period for every game, then adds up K × (score − expected) over all of them. That is what the calculator does, so you can enter the rounds in any order.

Why are my performance rating and my rating change pointing different ways?

They answer different questions. The rating change compares your score with what your current rating predicted. The performance rating ignores your current rating altogether. A 2000 player who performs at 1950 still loses points, even though 1950 might be well above the average opponent.

Can I use this for online tournaments?

For an estimate, yes. Chess.com and Lichess use Glicko systems, where the change per game also depends on how settled your rating is, so their numbers will differ. Performance rating is the same idea everywhere, though sites may calculate it with a different formula.