What is the correlation with accuracy and ELO?
Sort:
So accuracy just measures how accurate your moves are each turn. You can get a very high accuracy by playing someone much lower rated than you.
Dec 21, 2020
0
#2
scratch2432 wrote:
For example, if someone got 90% on chess.com analysis accuracy, what would be their estimated elo? What about 95%? 97%? 99%!?
Understandable misconception for a beginner to misinterpret CAPS, but a 2100+ bullet player? Maybe? I'll give you the benefit of the doubt and redirect you to this video, but I'd expect you to at least know one game is far from conclusive...
I find that my accuracy rises and falls and has much to do with the playing strength of my opponent.
Dec 21, 2020
0
#5
There is probably a correlation between average accuracy and ELO, but with a huge variance. 99% accuracy doesn't mean anything if it was some boring draw where most pieces vanished during the opening, leaving a static symmetrical position (yes, I know that from experience). If, on the other hand, you achieve 99% in a super complex sharp game, it's actually a sign of really strong play. So, accuracy alone doesn't tell you much.
Dec 21, 2020
0
#6
NilsIngemar wrote:
I find that my accuracy rises and falls and has much to do with the playing strength of my opponent.
Yes, if the opponent doesn't challenge you, strong moves are easy to find.
Dec 21, 2020
0
#7
Sred wrote:
... If, on the other hand, you achieve 99% in a super complex sharp game, it's actually a sign of really strong play...
Chess: jovialdick vs blueemu - 200865884 - Chess.com 99.4%
Dec 21, 2020
0
#8
KeSetoKaiba wrote:
Understandable misconception for a beginner to misinterpret CAPS, but a 2100+ bullet player?
He is 2100 bullet while being only 1700 blitz, which indicates that the OP is a child. Which actually makes this misconception expected.
KeSetoKaiba wrote:
scratch2432 wrote:
For example, if someone got 90% on chess.com analysis accuracy, what would be their estimated elo? What about 95%? 97%? 99%!?
Understandable misconception for a beginner to misinterpret CAPS, but a 2100+ bullet player? Maybe? I'll give you the benefit of the doubt and redirect you to this video, but I'd expect you to at least know one game is far from conclusive...
Of course, I am just wondering what correlation ELO has with accuracy. I am not using it for personal benefit but more for curiosity
KeSetoKaiba wrote:
Thank you very much for sharing the video. It's very informative
Dec 21, 2020
0
#11
blueemu wrote:
Sred wrote:
... If, on the other hand, you achieve 99% in a super complex sharp game, it's actually a sign of really strong play...
And yes, I forgot to mention that because accuracy is an average value, you can afford an occasional blunder 
Dec 21, 2020
0
#12
Sred wrote:
NilsIngemar wrote:
I find that my accuracy rises and falls and has much to do with the playing strength of my opponent.
Yes, if the opponent doesn't challenge you, strong moves are easy to find.
i agree
Dec 21, 2020
0
#13
Accuracy tends to be lower in closed positions where lots of moves are possible and finding the absolute best move is tricky. Sometimes the difference between what the computer thinks is best vs. excellent or best vs. good can be like less than 0.1.
But yeah, when your opponent makes obvious mistakes, finding the best move isn’t as difficult.
But yeah, when your opponent makes obvious mistakes, finding the best move isn’t as difficult.
Dec 21, 2020
0
#14
You can also lose a lot of accuracy in endgames by not finding the fastest mate. I tend to simplify and eliminate the chance for counterplay or stalemate and I am punished for not finding the fastest mate.
Every game is different. You're not going to have a consistent accuracy level as it fluctuates with you and your opponent playing at different strengths and accuracy.
Dec 21, 2020
0
#16
Is anyone else not comfortable with using the term blunder to describe a non-losing move? The computer thinks a move that goes from +20 to +10 is a blunder, but historically that’s not really how the term has been used. In the past they might have called it a mistake or even just an inaccuracy.
Dec 22, 2020
0
#17
IMBacon wrote:
"What is the correlation with accuracy and ELO?"
There isn't. Can people please stop obsessing over this accuracy crap????
bro, its not crap, can you actually be nice? christmas is coming
maybe actually help instead
to answer your question, honestly accuracy easily depends on how well your opponent plays as well
Nov 22, 2022
0
#18
Someone asks a simple question, and all they get is a lot of "there's other factors affecting accuracy".
The question was a well defined statistical question: what's the correlation?
Obviously it depends on strength of opponent. Assume the opponent tends to be of similar ability. Chess.com has a massive database. There should be two numbers for a given time control: slope and intercept (and correlation coefficient).
