My opponent in a game had 98.5 percent accuracy

Sort:
Erez_Shmerling

What rating does it correspond to?

He is rated 1330 but played against me like a super GM.

Erez_Shmerling
bbmaxwell wrote:

That's now how accuracy works. Even very low rated players (below 1300) can get 99% accuracy in certain games.

Two things that help high accuracy the most:

1) A short game. In fact if you just play some opening theory and then someone resigns, you can get 100% accuracy

2) An uncomplicated position with a lot of trades. Like an exchange French or exchange Slav

 

Now, if we're talking about some 70 move game out of a complex opening (KID let's say) and the middlegame was very sharp, if the 1300 got 98.5 in a game like that, then sure, it was probably cheating.

I don't know if the player was cheating, I prefer not to make unproven assumptions.

But he played like a super GM and his rating is low.

By the way he played against me he should be one of the site's top rated players.

It is possible that he is drastically underrated.

Erez_Shmerling

I think what is definitely possible, is that this player has GM level opening preparation.

But it still means he is sharply underrated, because with outstanding opening preparation

you can beat most amateurs with ease.

Erez_Shmerling

OK. I have decided not to let the loss deflate me.

I will trust the integrity of my opponent and not accuse him of foul play.

I will go on and try to improve.

I think that I play very well for a person who is untrained( no theory, no puzzles, no opening

preparation) and in good games I can reach 93-95 percent accuracy.

So I will try to train, learn openings, solve puzzles etc in a bid to become a strong player in classical time controls.

nklristic

I saw the game. First of all, it was short. You didn't manage to castle in time and was beaten because of it. On move 9. you could have castled already and proceed from there. Instead, you made too many pawn moves on both sides on the board and pretty much destroyed your position. His opening play had nothing to do with it. He just did normal opening moves and reacted to your pawn moves when he had to. You attack his queen with a pawn - he moves his queen, etc. 

You played the game pretty badly, ignoring basic opening principles, which happens to all of us from time to time. That doesn't prove if someone is cheating or not of course, I am just saying that "his brilliant play" is on you mostly in this game.


Nkai20

Don't feel bad, if someone is playing really good you can only applaud and learn from that game afterwards. If the other player is cheating, then is his or her loss, that's no fun.

I once played a game with 57 accuracy while my opponent had 73. I was being beaten badly but managed to checkmate him, had a lot of fun with that.

iSpaceChess

The CAPS/Accuracy score is not a cheat detection tool, here’s how it works:

TurkeySinatra

Can anyone explain why this guy continues to get 99%?

[removed, if you suspect someone the report them -- VP]

Erez_Shmerling
nklristic wrote:

I saw the game. First of all, it was short. You didn't manage to castle in time and was beaten because of it. On move 9. you could have castled already and proceed from there. Instead, you made too many pawn moves on both sides on the board and pretty much destroyed your position. His opening play had nothing to do with it. He just did normal opening moves and reacted to your pawn moves when he had to. You attack his queen with a pawn - he moves his queen, etc. 

You played the game pretty badly, ignoring basic opening principles, which happens to all of us from time to time. That doesn't prove if someone is cheating or not of course, I am just saying that "his brilliant play" is on you mostly in this game.


But he castled long and I was afraid to castle short because of the positional imbalances this

would create. I don't know how to handle such a game. 

nklristic

Well, king in center was way worse in this case. Especially after you've made so many pawn moves, in closed position it might have been different, but king in center can be trouble very often (I need to tell myself exactly the same thing, trust me happy.png ). Opposite side castling can be sharp, that is true. 

TR0LLKlNG
Erez_Shmerling wrote:

What rating does it correspond to?

He is rated 1330 but played against me like a super GM.

BOZO GOT CLAPPED!!
EVERY MOVE WAS SUPER FINDABLE BRUH. 

SantoshWildlifeArt

Also plenty of my opponents have been banned for fair play violations in the last several months I've been playing on chess dot com and I get their usual notification saying your opponent in a certain game was banned for fair play violation & hence you get X points of ELO... and so I invariably scroll through my games archive to locate the guy banned and browse their games history and in almost all cases, they look absolutely legit to me. Hundreds of games played, roughly even win/lose ratio etc... I'd expect cheaters to have like 90% win rate and more than 80% accuracy in games but these guys straight up have lost 50% of their games. Yet still banned for violating fair play.

Does this indicate that their much touted statistical algorithms to detect cheating have a disturbingly high false positive cases?

TR0LLKlNG
SantoshWildlifeArt wrote:

Also plenty of my opponents have been banned for fair play violations in the last several months I've been playing on chess dot com and I get their usual notification saying your opponent in a certain game was banned for fair play violation & hence you get X points of ELO... and so I invariably scroll through my games archive to locate the guy banned and browse their games history and in almost all cases, they look absolutely legit to me. Hundreds of games played, roughly even win/lose ratio etc... I'd expect cheaters to have like 90% win rate and more than 80% accuracy in games but these guys straight up have lost 50% of their games. Yet still banned for violating fair play.

Does this indicate that their much touted statistical algorithms to detect cheating have a disturbingly high false positive cases?

Technically we’re not supposed to discuss “alleged cheating” outside cheat forum, however— I too noticed this phenomenon. I dug deeper and discovered something interesting. “Players” will make any number of weird and/or blatant blunders early on to sandbag their accuracy, before suddenly executing a perfect mating sequence. Excepting an over the board game at the park or a smurfing GM this scenario is extremely unlikely to happen often at lower levels.

Kbz10troy
SantoshWildlifeArt wrote:

Also plenty of my opponents have been banned for fair play violations in the last several months I've been playing on chess dot com and I get their usual notification saying your opponent in a certain game was banned for fair play violation & hence you get X points of ELO... and so I invariably scroll through my games archive to locate the guy banned and browse their games history and in almost all cases, they look absolutely legit to me. Hundreds of games played, roughly even win/lose ratio etc... I'd expect cheaters to have like 90% win rate and more than 80% accuracy in games but these guys straight up have lost 50% of their games. Yet still banned for violating fair play.

Does this indicate that their much touted statistical algorithms to detect cheating have a disturbingly high false positive cases?

It's possible that they violated fair play rules against other opponents, but not against you. You still get a refund.

Kotshmot

Nothing looked out of ordinary in that game, every move was pretty intuitive. 

Pretty much forced him to play good moves.

tushnik2007

Can a player who is rated 1050 play with 98.5 percent aaccuracy in a game of 26 moves.I was playing with white I opened with e5 and he opened with sicilian defence.If you have any idea abt it plz answer.

Here is the PGN

[Event "tushnik2007 vs. spyisZer0"]

[Site "Chess.com"]

[Date "2023-08-27"]

[White "tushnik2007"]

[Black "spyisZer0"]

[Result "0-1"]

[WhiteElo "1052"]

[BlackElo "1069"]

[TimeControl "600"]

[Termination "spyisZer0 won by checkmate"]

1. e4 c5 2. Nc3 Nc6 3. Bb5 Nd4 4. d3 a6 5. Bc4 e6 6. Nf3 b5 7. Nxd4 cxd4 8. Ne2

bxc4 9. dxc4 Qa5+ 10. Bd2 Qc5 11. a3 Qxc4 12. b3 Qc6 13. Nxd4 Qxe4+ 14. Be3 Qxg2

15. Ke2 Bb7 16. Rg1 Qxh2 17. Nf3 Qh5 18. Qd2 Bxf3+ 19. Kd3 Be4+ 20. Kxe4 Qf5+

21. Kd4 Bd6 22. Rg5 Be5+ 23. Kc4 Qe4+ 24. Bd4 Rc8+ 25. Kb4 Qb7+ 26. Ka5 Qb5# 0-1

applewine
I have had a few games with high 95% plus accuracy, I have also had plenty at low 30-40 % some days it works, some days it doesn’t, I agree if every game was 90% plus that would be suss at low levels of skill. My thoughts on people who don’t follow the rules are basically it’s their loss, I get to practice against a level far above what I would otherwise do. I’m not losing money, only pride
BlueHen86

Sometimes your opponent is going to play really well. That doesn't mean that they are cheating. Accept it and move on.

KingKen_1

Hello can somebody send me the link of this game everybody was talking about some years ago please