-1

My rating on chess.com is of about 1077 on rapid and today when I won a game against a player with slightly lower rating the analysis results said I played with 73 accuracy according to their best computers and made no blunders.

Obviously the match was far from what a GM would play so why does the computer say I played well and what would be the difference between higher level players that also get a score of 73 from the computer? Or would they get even higher accuracy scores?

1
  • 4
    This is not a bad question at all. While I'm quite significantly higher rated on chess.com than 1077, I've also noticed this tendency of the computer giving me unfairly good reviews on my play (for instance, I blunder a piece for no reason and the comp says I afterwards that I didn't make a single blunder in the game). I think the engine is programmed somehow to give people "nice" feedback on their games, but I don't know the details.
    – Scounged
    Mar 13 at 19:23
3

Without seeing the game it is difficult to say exactly why you obtained such moderately high score. But I can give some general you comments based on how accuracy is calculated by chess.com.

Chess.com use a statistical model (not revealed) to calculate accuracy. However, it is more complex than calculating the proportion of best moves. By what they say, probably there is some kind of weighting that takes into account how near your move is to the best computer one.

Of course the higher rated you are, the higher your accuracy. And GM's get accuracies much above 90%. Now, that said, you could have a high score and still lose: a blunder is enough to lose the game.

Finally, I can think of of some factors that might contribute to having relatively high scores:

  • Following correctly long theoretical lines in the opening.
  • In certain endings it is difficult to go wrong.
  • In a quiet position, if you don't blunder a piece, any other move probably will be OK (suggested by @Hauke Reddman).
1
  • Three's a company: In a quiet position, if you don't blunder a piece, any other move probably will be OK (at least in the short run - of course you can make heinous positional mistakes). Mar 14 at 8:40
2

73 is not a high score. I have had several 100% scores by luck when I played games that were 'pianolas'. EG the opponent kept making moves that left me with only one good move.

Not sure if they weigh your rating when determining the accuracy but I do not think so. So for your rating 73 might be good but as I noted with my results above it is mostly meaningless too.

It does give you a very rough idea of how well you played that one game but says nothing about your overall ability.

2

According to chess.com an average score of 76.86 would be expected for a person with an OTB rating around 1200. A 2400 player (which is about IM level) would have an average score of 96.20. So, no, your score is not one that is typical of a GM. It's actually similar to what you'd expect for your rating.

That's not to say that a GM can't have a score below that, of course. I'm looking at a game by Hikaru in which he scored just 70.5. But there are two differences: he was playing blitz and didn't have time to calculate (he had just 20 seconds left at the start of the move where he made the critical mistake), and his opponent was a GM himself, which typically means that the best moves are going to be harder to find than if you're playing a weaker opponent.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.