I got 100% accuracy will I get banned?

Sort:
Avatar of Sakura_Virgo

woah

Avatar of asflin

bq

Avatar of nklristic
m_cavalcanti wrote:
boriskravitz wrote:

They give players these high accuracy scores just to bait you into subscribing. Don't believe it. And if you're playing someone who basically cannot play chess and score a high accuracy number, what does that mean? I am sure if I played a chimpanzee, I would have high accuracy.

This is not how it works. The accuracy is mesured by Stockfish (a powerful chess engine). I developed a stockfish based tool to avaluate my matches (chess.com provides analysis only by subscrition) and the level of accuracy is mesure not by the win or lose, but if you chose the best move among all moves possibles.

Actually, his answer is not far from the truth.

Accuracy today is not measured the same way it was measured several years ago on chess com. I had a game with accuracy under 20, I measured it after the changes and it was 50-60. Today it is very difficult to get a very low score.

I had an unrated game against lower rated person that lasted around 10 moves. My opponent blundered 3 times, losing material every time and resigned. The result - his accuracy was 70, which is very high considering how he played the game.

The accuracy has to be taken with the grain of salt. The same goes for rating estimation.

Avatar of m_cavalcanti
nklristic wrote:
m_cavalcanti wrote:
boriskravitz wrote:

They give players these high accuracy scores just to bait you into subscribing. Don't believe it. And if you're playing someone who basically cannot play chess and score a high accuracy number, what does that mean? I am sure if I played a chimpanzee, I would have high accuracy.

This is not how it works. The accuracy is mesured by Stockfish (a powerful chess engine). I developed a stockfish based tool to avaluate my matches (chess.com provides analysis only by subscrition) and the level of accuracy is mesure not by the win or lose, but if you chose the best move among all moves possibles.

Actually, his answer is not far from the truth.

Accuracy today is not measured the same way it was measured several years ago on chess com. I had a game with accuracy under 20, I measured it after the changes and it was 50-60. Today it is very difficult to get a very low score.

I had an unrated game against lower rated person that lasted around 10 moves. My opponent blundered 3 times, losing material every time and resigned. The result - his accuracy was 70, which is very high considering how he played the game.

The accuracy has to be taken with the grain of salt. The same goes for rating estimation.

I agree on the bias accuracy by a website which sells subscribtions, that's why I have my own tool based on stockfish 17 to avaluate my matches. However, the difference is not that big. Trust me.

Avatar of Aismyuser

That's a good trap for the Countrysaurus because he mostly uses the Sicilian

Avatar of thedelcai
m_cavalcanti wrote:
 

I agree on the bias accuracy by a website which sells subscribtions, that's why I have my own tool based on stockfish 17 to avaluate my matches. However, the difference is not that big. Trust me.

I disagree with your first sentence, that there is an accuracy bias. I have my own tool as well and the differences are completely insignificant. Accuracy is just not a difficult number to compute, and skewed results would be very easy to spot, unlike, say, Game Review's rating estimate which is commonly misunderstood and does need to be taken with a grain of salt.

Avatar of magipi
thedelcai wrote:
m_cavalcanti wrote:
 

I agree on the bias accuracy by a website which sells subscribtions, that's why I have my own tool based on stockfish 17 to avaluate my matches. However, the difference is not that big. Trust me.

I disagree with your first sentence, that there is an accuracy bias. I have my own tool as well and the differences are completely insignificant. Accuracy is just not a difficult number to compute, and skewed results would be very easy to spot, unlike, say, Game Review's rating estimate which is commonly misunderstood and does need to be taken with a grain of salt.

In fact, nobody knows for sure how chess.com calculates accuracy, because chess.com won't disclose it. What is known (because they told us), is that accuracy is modified up in low rated games, so that most games give an accuracy number between 50 and 90.

If you have your own method and it gives similar results, good for you.

Avatar of boriskravitz
m_cavalcanti wrote:
boriskravitz wrote:

They give players these high accuracy scores just to bait you into subscribing. Don't believe it. And if you're playing someone who basically cannot play chess and score a high accuracy number, what does that mean? I am sure if I played a chimpanzee, I would have high accuracy.

This is not how it works. The accuracy is mesured by Stockfish (a powerful chess engine). I developed a stockfish based tool to avaluate my matches (chess.com provides analysis only by subscrition) and the level of accuracy is mesure not by the win or lose, but if you chose the best move among all moves possibles.

I understand that this is how it should work. But to assume chess sites, not just this one, do not manipulate the final data beyond what Stockfish concludes would be imprudent. I also do some Ai art to pass the time and sometimes get a message about "You created a masterpiece! You are a great artist! Share it with the community! And the picture has people with three arms and seven fingers" this is where Ai is involved, and Ai is active here as well. The more new people playing blitz and rapid on their mobile phones think they are playing good chess and getting encouraging analysis reports, the more likely they are to invest money.

But there is no proof of this really online. So I may be wrong and I do not understand how Stockfish works. Grok adds this:

Stockfish can show high accuracy for low-ranked players in blitz/rapid games because it evaluates moves against optimal lines, not overall game quality. In fast games, simple moves by weaker players might align with Stockfish’s top choices, especially in straightforward positions, leading to inflated accuracy (e.g., 100%) despite poor strategy. This isn’t deception but a limitation of engine analysis, which prioritizes move correctness over positional understanding. Moderately skilled players notice this disconnect because they assess broader game context, which Stockfish doesn’t weigh.

Avatar of alexdvyl
It’s sur
Avatar of m_cavalcanti
boriskravitz wrote:
m_cavalcanti wrote:
boriskravitz wrote:

They give players these high accuracy scores just to bait you into subscribing. Don't believe it. And if you're playing someone who basically cannot play chess and score a high accuracy number, what does that mean? I am sure if I played a chimpanzee, I would have high accuracy.

This is not how it works. The accuracy is mesured by Stockfish (a powerful chess engine). I developed a stockfish based tool to avaluate my matches (chess.com provides analysis only by subscrition) and the level of accuracy is mesure not by the win or lose, but if you chose the best move among all moves possibles.

I understand that this is how it should work. But to assume chess sites, not just this one, do not manipulate the final data beyond what Stockfish concludes would be imprudent. I also do some Ai art to pass the time and sometimes get a message about "You created a masterpiece! You are a great artist! Share it with the community! And the picture has people with three arms and seven fingers" this is where Ai is involved, and Ai is active here as well. The more new people playing blitz and rapid on their mobile phones think they are playing good chess and getting encouraging analysis reports, the more likely they are to invest money.

But there is no proof of this really online. So I may be wrong and I do not understand how Stockfish works. Grok adds this:

Stockfish can show high accuracy for low-ranked players in blitz/rapid games because it evaluates moves against optimal lines, not overall game quality. In fast games, simple moves by weaker players might align with Stockfish’s top choices, especially in straightforward positions, leading to inflated accuracy (e.g., 100%) despite poor strategy. This isn’t deception but a limitation of engine analysis, which prioritizes move correctness over positional understanding. Moderately skilled players notice this disconnect because they assess broader game context, which Stockfish doesn’t weigh.

Indeed it makes sense. Selling is an decision maker in any business.

Avatar of IcanDodgeBullets1999

idk how to tell if a game has 100% accuracy.. can someone please tell me my accuracy on this game?

Avatar of m_cavalcanti
IcanDodgeBullets1999 wrote:

idk how to tell if a game has 100% accuracy.. can someone please tell me my accuracy on this game?

Avatar of thedelcai
magipi wrote:
 

In fact, nobody knows for sure how chess.com calculates accuracy, because chess.com won't disclose it. What is known (because they told us), is that accuracy is modified up in low rated games, so that most games give an accuracy number between 50 and 90.

You're right. It doesn't really work after they changed to "CAPS2", just comparing what was played to what the engine would have played.

"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 between 50 and 95, which provides a more intuitive understanding of how accurately you played in your game."

https://support.chess.com/en/articles/8708970-how-is-accuracy-in-analysis-determined

Avatar of thedelcai

Also this:

"Accuracy (formerly known as CAPS or Computer Aggregated Precision Score) is not cheat detection. Accuracy calculations were invented to give newcomers to chess a guideline for their performance. A player may achieve a high accuracy score in a single game for many reasons that have nothing to do with cheating."

https://www.chess.com/cheating

Avatar of GM-ImJustHim

I didnt get banned its been 10 days, so i think im fine.

Avatar of GM-ImJustHim

15 days now lol.

Avatar of Vocaloid39
GM-ImJustHim wrote:

So I played a game and got 100% accuracy and the game was 9 moves long will I get banned, as I set up a trap and he fell in it pls tell me as I'm very worried https://www.chess.com/game/live/140380866596

If you play fairly, then you have nothing to worry about. Getting 100% accuracy is definitely possible; this is made much easier if all the moves are "book moves" or if the opponent blunders early. This video will help explain CAPS (accuracy).

Avatar of ArcaneTheDragon

You're fine, I got 100% in a game;

(draw by agreement)
Avatar of Blunder_Bot_11
U good as long as u don’t do it a bunch of times un a row
Avatar of GM-ImJustHim

Ayy im not banned been more than a month