# What does "TPR" mean in chess tournament reports?

I usually find in reports like this (and many others) something like:

... with a TPR of 2551 ...

I would like to know its meaning, and also what it stands for.

It means Tournament Performance Rating.

Very roughly a TPR of 2551 means that the results this player has achieved in this tournament would have been expected of a player rated 2551.

That's always a bit problematic. Say you scored 100% against a group of 1500 players, what kind of player would have expected to score that? Well, a 1900 player probably, but also a 2800 player. And many methods make your TPR go down if you win against a much lower rated player, because they are based on the average rating of your opponents. But it's a rough measure of how well a player did.

There exist various methods to calculate it that I don't have time now to go into now.

• The more varied the people in the tournament were, and the more varied your results were, the less reliable a metric it will be. And like all metrics (including, quite honestly, the typical rating) they should be taken with a grain of salt. Commented Jan 29, 2019 at 17:48
• A 1900 player actually has a 92% expected performance against a group of 1500 players, so if there is enough 1500 players the 1900 is not high enough. Commented Jan 30, 2019 at 16:53

"Tournament performance rating". An approximate measure of the strength that a player performed/played at in the tournament.

The calculation of such a performance rating varies, but one method is as follows:

• If you beat someone at rating X, your performance for that game is X + 400.
• If you lose to this person, your performance for that game is X - 400.
• If you draw this game, your performance for that game is X.

Then the tournament performance rating is calculated by taking the average performance out of all your games.

• Can this be summarized by this formula: `avg. opponent rating + 400 * (number of wins - number of losses) / number of games`? Commented Jan 29, 2019 at 20:05
• @Akavall Yes, but that looks much more mysterious than the explanation given in the answer. Commented Jan 29, 2019 at 20:26