Unless you have actually done it, no. You will not get banned.
Basically, you had around 96 to 98%, against a 470 rated player, which is normal into such a a big difference of ratings. Moreover,it is normal into minature games,to have a high accuracy
However, if you do it with smn rated let's say, 1500,you will get into chessdotcom's radar.
Brilliant moves do not really mean anything. Don't take it too seriously. It's an AI algorithm.
Basically, if a move has a huge advantage over other moves, like that it is the only winning move while all other moves may lead to losing, it will be considered by AI as "brilliant".
It may be a very obvious move, e.g. if your opponent's queen takes you rook, and you have a fork to take his queen back. If you don't see the fork, you'll lose a rook. If you find the fork, you are winning.
And that fork will be considered by AI as a "brilliant" move, despite that it may be a very obvious fork. Not brilliant at all in human perspective.
From the angle of a human, the move you sacrifice the rook should be "brilliant", not the move you fork the queen.
So, in the end, do not take AI's appraisal too seriously.
I am disappointed you live in a society where you have to worry about being banned for accuracy.
I am disappointed you live in a society where you have to worry about being banned for accuracy.
As a matter of fact,accuracy is just one of the many factors,thus it is not the case for being banned
Thank you
I recently got an email saying:
"We have detected that one or more of your recent opponents has violated our Fair Play Policy. As compensation for potentially unfair rating losses, we adjusted your following ratings:"
...and they put 9 points on my blitz. I don't know what happened to the other person(s) or who they even were. The algorithm must have detected statistically anomalous behavior. IDK. It was a total surprise to me.
I can't imagine that online chess isn't completely filled with cheating. It's just the nature of the beast. How can you assess that someone is cheating and not just playing really well? I guess there are situations where people choose moves that are advantageous 10 or 12 moves deep that only an engine would pick. Could it just be a 'lucky' guess? A sudden stroke of genius by someone who would otherwise never make such a move? Sure. Sure... Man's inhumanity to man, man. As with all things if you just focus on being honest and trying hard you will probably be okay.
I don't understand why people would cheat. Well, at low levels anyway. Why would someone want to cheat against a 500 level blitz player? (I'm getting better don't laugh.) Bragging rights? To make yourself feel better? How would knowing you had to cheat make you feel better? It would make me feel rotten.
Cheating may get you a good rating but as soon as you go to, say a live tournament, you're not going to stand a chance if you went up around 800 points or more
I'd have to think that the algorithm isn't looking at individual moves, but if a 900 level player that was consistently playing with 60-70% accuracy suddenly starts playing with 80-90% accuracy and sees their rating shoot up, that will trigger some alarm bells.
To dingus_finkelstein:
Look over your completed games, and you will find an opponent who has been marked with a crossed out zero symbol, like a "do not enter" sign.
youre 1300. Yeah youre "low rated" but youre high enough rated that occasionally you might play a few 90%+ games, thats pretty normal
If I was trying to catch cheaters here's what I would look at:
1. Establish a baseline level of play for a player. I'm guessing chess.com has some sort of an algo that knows your 'quality' of play independent of the time controls you use. Your quality of play can be represented on a bell curve with a standard deviation. If you're a 1600 player with a standard deviation of 200, you better not play better than 3 st. dev's from your rating more than 1 out of 300 games or it will look VERY suspicious.
2. Look for significant variations outside of that baseline for an entire game or for moves in a game - comparing the play to himself/herself and other similar player types.
3. I'm looking at the newness of the player. Newer account cheat at a much higher rate. Free account cheat at a much higher rate than paid accounts. I'm looking at the ip address of the account to see if it has been banned in the past.
4. I'm looking at the number of times a player has been reported for cheating and even who reported them. I have a high success rate of players getting banned after I report them.
5. Time in-between moves is a dead giveaway. Playing precisely and quickly in complex positions is very hard to do.
6. I'm comparing a suspected cheater's play to a lot of engines and depths to see if the play matches. Just because a player doesn't use the top engine move doesn't mean he/she isn't using the second best move.
If you're cheating, you have to be amazingly consistent with your quality of play which is really difficult for a human to do. A computer can but then it gives the signs that it is a comp playing.
I play a lot of 960 tournaments and cheating is a real problem. It's obviously cheating when a newer player with a 1200 rapid rating after >= 20 games performs at an 1800 chess960 rating level with an average accuracy above 80% or even a single 50 move game with an extremely high accuracy.
The real challenge is selective cheating...a move or two a game. But I suspect it's too tempting to cheat just a move or two a game. And it may not be the challenge I think it is - I'm not sure. GMs have been banned for cheating on here and I'm sure it's selective move choices. No GM is using all engine moves.
Side note - I doubt chess.com has this but it wouldn't be too difficult to create an algo that can tell if you're playing drunk.
There's a huge difference between accurately taking advantage of obvious weaknesses vs accurately predicting engine-level outcomes 15 moves ahead. The false positive rate is so incredibly low that at this point they just might not happen any more, I'd argue most chatter about unfair bans come from those who did violate fair play in some capacity. You'd basically have to miraculously execute GM level play against a formidable opponent (unlikely due to matchmaking) by complete accident in order to receive an unjust ban.
I am disappointed you live in a society where you have to worry about being banned for accuracy.
I am more disappointed that people cheat and ruin the fun of a tournament for no good reason. And yes accuracy is a strong indicator of cheating i.e. a 400 player winning 9 games in a row with 90+% accuracy every game.
@updog
So I usually have a standard deviation of 300 points. My rating is now 1900. A couple of days ago, my rating was 2000. I was having a really good day. I kept on winning. In one game, I felt really good and checked the bottom of my game review. It said I played like 2600 with 95% accuracy. Would I get reported and banned?
@updog
So I usually have a standard deviation of 300 points. My rating is now 1900. A couple of days ago, my rating was 2000. I was having a really good day. I kept on winning. In one game, I felt really good and checked the bottom of my game review. It said I played like 2600 with 95% accuracy. Would I get reported and banned?
My first question is what is typical for a player as good as you? Secondly, how many of those games did you have such a high accuracy rate? Next, do you really think a 350 player would win, say 8-10 in a row against increasingly tougher players? Lastly most, if not all, of these players opened an account less than a week before. Kind of curious don't you think?
In order to know your std dev, you'd have to analyze every game, get the game review rating and calculate your std dev. I doubt it's 300 points. That feels really high. I suspect for most everyone it's around 150. If it was 300 points, that means your quality of play falls between 1000-2800 99.7% of the time for a 1900 (three std devs) which is a massive range. More likely I suspect it would be closer to 1450-2350 99.7% of the time. But this is a guess. Without real data, I can't quantify it. Also, I don't personally trust the game review rating either...
There's a lot that goes into calculating accuracy. I've personally found it's much easier to get a high accuracy if your opponent it lower rated or the game has fewer moves. I don't know how much accuracy actually factors into cheat detection - if at all. But when I see a lower rated player with a really high accuracy, it's suspicious. At 1900, it's much less so.
Your opp might have reported you. But you clearly haven't been banned.
We have a similar rating and I do get calculated game review ratings of 2500-2600 on occasion so it's definitely possible to get it legitimately at our rating without cheating.
Unless you have actually done it, no. You will not get banned.
Basically, you had around 96 to 98%, against a 470 rated player, which is normal into such a a big difference of ratings. Moreover,it is normal into minature games,to have a high accuracy
However, if you do it with smn rated let's say, 1500,you will get into chessdotcom's radar.
Even 1 game around, here and there will not do anything on its' own. But if your average accuracy is like 99 and you have a 98.7 win percentage. That is where you'll get banned. If the algorithm wrongly bans you, you can appeal, and usually it's at least decent at it's job.
I recently got an email saying:
"We have detected that one or more of your recent opponents has violated our Fair Play Policy. As compensation for potentially unfair rating losses, we adjusted your following ratings:"
...and they put 9 points on my blitz. I don't know what happened to the other person(s) or who they even were. The algorithm must have detected statistically anomalous behavior. IDK. It was a total surprise to me.
I can't imagine that online chess isn't completely filled with cheating. It's just the nature of the beast. How can you assess that someone is cheating and not just playing really well? I guess there are situations where people choose moves that are advantageous 10 or 12 moves deep that only an engine would pick. Could it just be a 'lucky' guess? A sudden stroke of genius by someone who would otherwise never make such a move? Sure. Sure... Man's inhumanity to man, man. As with all things if you just focus on being honest and trying hard you will probably be okay.
I don't understand why people would cheat. Well, at low levels anyway. Why would someone want to cheat against a 500 level blitz player? (I'm getting better don't laugh.) Bragging rights? To make yourself feel better? How would knowing you had to cheat make you feel better? It would make me feel rotten.
same to me they added 9 to my rapid
I've played a few games with brilliant moves or high accuracy games. Would chess.com be suspicious about this since I'm a low rated player and got a high accuracy game or brilliant moves?