results

Carlsen 3 such games out of 100

Nakamura 6

Caruana, Grischuk 10

Kamsky 15

Kramnik 23 !!

Are you aware that Chess.com changed the calculations for accuracy in 2022? Even lower rated players sometimes score 90%+ now.

It is not true that a 90%+ score means "Top GM level" accuracy any more.

https://support.chess.com/article/1135-what-is-accuracy-in-analysis-how-is-it-measured

How is Accuracy in analysis determined?

Each game review shows you an accuracy score, and classifies all of your moves. But how are the accuracy and classifications decided? 

Accuracy Scores

Your Accuracy is a measurement of how closely you played to what the computer has determined to be the best possible play against your opponent's specific moves. The closer you are to 100, the closer you are to 'perfect' play, as determined by the engine. 

Chess.com’s Accuracy score is now powered by “CAPS2”, an improved version of the original Chess.com “CAPS” (Computer Accuracy Precision Score) algorithm.

Moves are still compared against the top engine recommendations, but the math on how these are calculated has changed. Why? Well, most chess players - even low-rated ones - make a lot of the best moves!

Previously, CAPS (v1) looked to create a 0-100 band within the normal human player range. So, scoring perhaps 40% “Best” moves, which is very low, was equal to single-digits on CAPS. And, scoring a very high number of “Best” moves, but not a perfect game, was often rated 99.9 on CAPS, even though it wasn’t played perfectly.

This made some people feel bad (on the low end), and led to a lot of cheating allegations (on the high end). The new Accuracy scores, based on CAPS2, replicate the feeling of being graded on a test in school. 

You will notice that the majority of scores now fall mostly be between 50 and 95, which provides a more intuitive understanding of how accurately you played in your game. 

The image below shows Accuracy scores from CAPS and CAPS2 for players rated between 1000 and 1500. 

Figure 1: Accuracy scores for players rated between 1000 and 1500. 

How they can cheat?

oh wow

If there is cheaters in Titled Tuesday, that would be a very big deal

justbefair

Thanks for this information but whatever is the system, since it is universal, meaning EVERYTHING and EVERYONE is calculated with the same parameters, how do you explain this difference still 3 against 23, it is unexplainable mathematically,isn't it?

And why if its not that unusual to score 90plus for below 2901 players in games against Magnus or Naka they manage only 4,5 percent in average? I mean glad those players do not have those dubious statistics in their games, do not misunderstand me, its not their fault . Just wondering, to put it mildly, how this might work, their 4,5 and mine 23 percent

would mention an important point, all calculations made with games in recent 3 month, so with the same system

justbefair

Another important point, showing 90plus level against weak opponent who might blunder a piece in the opening and against top GM, two completely different levels required. Showing 90plus against top GM IS TOP LEVEL GM performance

In B4 the lock

What @justbefair wanted to say is that the completely same game done by 800 rated player and by 2 500 rated player will show different accuracies, because since 2022. chess.com decided to make accuracy results kinder to weaker players. In my opinion that is unfortunate as accuracy should just be the cold truth instead of what we have today.

That being said, opponents in Titled Tuesday are all titled and pretty high rated (compared to most players), so what I've just said shouldn't impact the statistics that much.

One thing to consider could be styles of play. One top GM might play more solid chess and have naturally more accurate games (for him and his opponent), while the other might play more wild games, but still, as the difference is drastic, you might have a point (plus they ban some titled players every month, so I am sure some of those are because of TT).

I could imagine one top GM can have less opponents who have >90 accuracy than the other top GM, but still, it is strange that the difference is 3% vs 23% of games. That is a really big difference.

Thank you for your comments. I have been told that the questions have been forwarded to the staff and there will be a staff response.

Vladimir, I'm such a huge fan of yours. I hope ur happy & that all is going well in ur retirement.

And sorry about the tt issues, maybe this site can do better.

Finally nice to see that even top GM's are seeing this unusal stuff on this site, this is rampant and no matter the level ,these days if you play less than 80+ accuracy you rarely win a game at any level, you got 800 blitz players, 1000 rated rapid plauyes blasting out moves out in seconds, all top 3 engine lines, max like 1-3 mistakes a game with few inaccuricies, online chess is pretty much dead if this continues like this

leetsak Dreamscapehorizons

I agree since online chess becoming major thing fighting cheaters is essential for the future of chess in general. And half measures wouldn't prevent. Strong mathematical evidence, with high probability calculated via formulas based on huge dataset, and open 1 year ban in case probability is say 97percent. To make other players understand the consequences in such case. Or we slowly but surely lose fair online chess

I have made various calculations, this is just 1 example, there are MAAANY strange numbers I can deliver). And based on those calculations my conclusion is, the problem is much wider and touches all circles of players, it seems. Therefore feel obliged to start talking about it

Oh he's the messiah all right. The only player to beat Garry Kasparov in a match to win the world title. Anyway, the accuracy % thing needs to be improved greatly. I see no need to mess with it for the benefit of weaker players, that simply defeats the purpose of having it at all.

Как интересно,правда ничего не понятно,ну переводчик в помощь

Владимир Крамник,расскажите пожалуйста,как вас выиграл Магнус Карласен