Highest Performance Rating ever

I was looking through the chessbase.com site, and ecnoutnered this interesting article, http://chessbase.com/newsdetail.asp?newsid=13, on a rapid team match played between GM's from Europe against Asia. Gary Kasparov played 12 games, won 10 and drew 2 for a performance of 3043.

This got me thinking, what's the highest performance rating in history.

Chessmetrics gives a 2899 by Karpov in Linares, but those ratings don't take into account events after 2005 (Though he did rate the performance of Topalov in San Luis saying it wasn't quite as high, his Elo performance there was 2890)

There is of course Sofia Polgar's famous performance in Rome, the highest ever in an open, up to that point, but I think it has since been eclipsed.

 Gabriel Sargissian once got a 3021 rating in an 8-player round robin.

Anyone know of any higher performance ratings?

3043? Dang! Yea, that Kasparov score most likely is the highest ever recorded.

Fischer's performance rating in his Candidates match with Bent Larsen supposedly eclipses 3000 as well. I don't know what the exact number is

i'll start my own chess federation. it will be composed of 50 6-year-olds, 1 person that can perform at about my level, and 10 senile 80-year-old. i'll use the same scoring system USCF uses...and watch: my rating will be like 3200!!

my point may be obvious to others. but as long as the rating system remains relative and as an "estimate" of 1 players probability of winning against another player ... wouldn't the "highest rating" ever have many answers?

chess rating(the ELO system) will always be relative and based on the probability of a player beating another.Because thats how the human brain works.Sometime you are thinking straigth other times you don't.And performance here Akuni wish to say is something like the "temporary rating" for the tournament. He wants to know not the "best" player but the player who had the best day.

And for your own chess federation it does not work, because you could only compare your level with the others in your federation.The ELO system works because people don't get isolated from the other, and it will be always an estimated level, as it is how probability works.

 Good response!  This is exactly what I was thinking.

Magnus Carlsen's 8/10 score at the Nanjing Pearl Spring Tournament in China with 6 wins and 4 draws and no losses had a perfomrance rating of 3002.

Fischer's 6-0 win over Larsen in the Candidates semi-final in Dallas 1971 gave him a performance rating of 3060 I believe.  His victory over Taimanov in the quarterfinals was 3020.  Ratings are of course designed to compare players who are playing against each other regularly, so they are probably not comparable over long time periods.

Come on,We can't compare bobby fisher times and karpov times when the ratings were much more lower.Just look at this:Top 100 players in 2000 had an average of 2644. Today the average is 2699.This shows how big the inflation is.It is to make a XXI Tournament class then it is before .Back then it was impossible.

So my conclusion is that we can't compare numbers as a discrimant of strengh.The true strength is inside every game in tournament and every opponent

I believe Fischer/Larsen took place in Denver. 

Woops!   I stand corrected.

The Chessmetrics ratings are not standard ELO performance ratings but are specifically designed by statistician Jeff Sonas for comparing players from different ages. More recently in 2009 (http://www.chessbase.com/newsdetail.asp?newsid=5828) Sonas still  ranks Karpov performance as the greatest.

You can always device a formula and perform calculations to get some numbers but what do these really mean when comparing performances made in different ages, in events of different legth etc? Subjectively I tend to also be most impressed by Karpov's 11/13 in Linares 94.

Actually, when someone scores only wins there is no real performance rating. There simply isn't enough information to calculate one, as performance ratings of e.g. 3000, 6000 or 9000 fit a 100% score equally well. Usually these ratings are calculated by taking 99% or something similar.

It's hard to compare performance ratings without taking the level of opposition into consideration. Many players have scored all wins or lots of wins and one draw against comparatively weak opposition. Morozevich performed 3027 in Kishinev 1998, but his strongest opponent was 2620 (and his only draw in nine games came against a 2405 player). It's more difficult to score similar results in a top tournament. For example Carlsen's 3002 in Nanjing 2009 came against a field where his weakest opponent was 2736 (and the strongest 2813).

Yang-fan Zhou (then an FM) scored 9/9 in this tournament for a tpr of 3038.

Carlsen is close in the London Chess Classicof 2012. He has a tournament rating of 3101 and just won another game. He could have the record soon.

And? The result? :)

Carlsen's performance in London was 2994.

Trying to fathom that level of play is like trying to grasp the size of the universe...   

I don't know if it's the highest ever, but Wesley So has a hell of a high one : 3183