As most of you are aware of, there are many instances of people on this site who are suspended because they are caught cheating. Usually they are caught when people tend to notice that their moves match very closely with moves that a computer would make. I think that there are many subtle ways in which a person who cheats can give themselves away before it is finally proven by matching their moves with a chess engine.
I will give you readers one example that I experienced. I was flipping through my friend list yesterday and noticed that one of them was missing. It was a user by the name of BrainFactory. I went to his profile and discovered that his account had been disabled. Personally, I was not very surprised by that. I am ruling out the possibility that he decided to pack up and leave. He was very active on here with about 200 or more games going on at once, and he was very sociable even offering his email address on his profile.
I first played him when his rating was in the mid 1800's. We played a few more games and he was very polite and friendly and invited me to be on his friends list. I didn't think anything was unusual until his rating soared over 2000. When it was at that point, I looked on his profile and saw some very unusual stats. He had about 125 wins, 1 draw, and ZERO losses!!
I noticed that the bulk of his opponents were much lower rated than he was. He probably wasn't getting much more than 2-3 points per win. He would sometimes play opponents that were close to his rating, but he would almost never play anyone above his rating.
By the time that his rating made it to the 2200's, he finally had (in my opinion allowed) his first loss. He had completed 150 or more games by the time this happened. His average opponent rating at that time was in the mid 1600's. This pattern of having a ridiculously high winning percentage coupled with a pitifully low average opponent rating continued until he reached his peak in the mid 2300's. By then, his average opponent rating was in the high 1700's.
There are many instances of players on this site who play mostly lower rated opponents thinking that their overall rating will be higher if they play that way rather than playing opponents who are closer to, or above their rating. At first I thought this might just be a severe case of this, even though the 90+ winning percentage was suspicious. So I asked him casually in the chat box of one of our games why he insisted on playing mostly low rated opponents and why he didn't play more challenging ones. He completely ignored me. I thought something was really fishy at that point. Because as I said, he was very friendly and outgoing, so it struck me that he didn't want to answer a simple question like that.
Looking back at the games we played, which wasn't very many, even though I didn't think about it at the time, (because I had so many games going on at once and didn't play this user very often) I did get a very subtle feeling at some point during a few of our games that I was playing a computer.
Now those of you reading this might wonder what is so suspicious about a very high winning percentage and an extremely low average opponent rating. Well, if you think about it, if you wanted to obtain a relatively high rating by cheating, it would be much easier to get away with it by clobbering mostly low rated opponents. If your rating is only 1700, it isn't going to make a significant amount of difference whether you get beaten by a 2300 player versus getting beaten by a 2000 player. If you, the 1700 player, play both players consistently, you will find that you will win maybe 1/10 games at the most with the 2000 player whereas you would be lucky to win 1/100 games with the 2300 player. But only a couple of games would not be enough for the 1700 player to get a good feel of what the playing strengths are of both players since he will lose almost as badly in each game. Another reason why it would be easier to avoid detection by playing mostly low rated opponents is that higher rated opponents will have a much better ability to tell the difference between computer vs. human moves.
So if a higher rated opponent is being cheated, he will be able to detect it much easier. Even though I can't say for 100% sure that he was a cheater, I suspect that that was the cause of BrainFactory's demise. The evidence that I provided was completely circumstantial. Like I said, the proof is in comparing the moves to what a chess engine recommends.
This might very well be one of those cases where someone decided to make themselves look like a big shot on the internet. What else do you expect from someone named BrainFactory? This user bragged on his profile that he was working on a phd for physics at his university or something to that effect, I don't know, it has been many months since I last looked on his profile. So maybe he thought that that coupled with a very high online rating would make him look respectable on this site.
So I want to know what you, the reader, thinks about this situation. Do BrainFactory's activities on this site seem fishy to you as they did with me? Also, what other subtle cues do you think you can pick up on to spot a cheater?
These activities do seem fishy. Very good observation, Sherlock Holmes. LOL
As most of you are aware of, there are many instances of people on this site who are suspended because they are caught cheating. Usually they are caught when people tend to notice that their moves match very closely with moves that a computer would make. I think that there are many subtle ways in which a person who cheats can give themselves away before it is finally proven by matching their moves with a chess engine.
I will give you readers one example that I experienced. I was flipping through my friend list yesterday and noticed that one of them was missing. It was a user by the name of BrainFactory. I went to his profile and discovered that his account had been disabled. Personally, I was not very surprised by that. I am ruling out the possibility that he decided to pack up and leave. He was very active on here with about 200 or more games going on at once, and he was very sociable even offering his email address on his profile.
I first played him when his rating was in the mid 1800's. We played a few more games and he was very polite and friendly and invited me to be on his friends list. I didn't think anything was unusual until his rating soared over 2000. When it was at that point, I looked on his profile and saw some very unusual stats. He had about 125 wins, 1 draw, and ZERO losses!!
I noticed that the bulk of his opponents were much lower rated than he was. He probably wasn't getting much more than 2-3 points per win. He would sometimes play opponents that were close to his rating, but he would almost never play anyone above his rating.
By the time that his rating made it to the 2200's, he finally had (in my opinion allowed) his first loss. He had completed 150 or more games by the time this happened. His average opponent rating at that time was in the mid 1600's. This pattern of having a ridiculously high winning percentage coupled with a pitifully low average opponent rating continued until he reached his peak in the mid 2300's. By then, his average opponent rating was in the high 1700's.
There are many instances of players on this site who play mostly lower rated opponents thinking that their overall rating will be higher if they play that way rather than playing opponents who are closer to, or above their rating. At first I thought this might just be a severe case of this, even though the 90+ winning percentage was suspicious. So I asked him casually in the chat box of one of our games why he insisted on playing mostly low rated opponents and why he didn't play more challenging ones. He completely ignored me. I thought something was really fishy at that point. Because as I said, he was very friendly and outgoing, so it struck me that he didn't want to answer a simple question like that.
Looking back at the games we played, which wasn't very many, even though I didn't think about it at the time, (because I had so many games going on at once and didn't play this user very often) I did get a very subtle feeling at some point during a few of our games that I was playing a computer.
Now those of you reading this might wonder what is so suspicious about a very high winning percentage and an extremely low average opponent rating. Well, if you think about it, if you wanted to obtain a relatively high rating by cheating, it would be much easier to get away with it by clobbering mostly low rated opponents. If your rating is only 1700, it isn't going to make a significant amount of difference whether you get beaten by a 2300 player versus getting beaten by a 2000 player. If you, the 1700 player, play both players consistently, you will find that you will win maybe 1/10 games at the most with the 2000 player whereas you would be lucky to win 1/100 games with the 2300 player. But only a couple of games would not be enough for the 1700 player to get a good feel of what the playing strengths are of both players since he will lose almost as badly in each game. Another reason why it would be easier to avoid detection by playing mostly low rated opponents is that higher rated opponents will have a much better ability to tell the difference between computer vs. human moves.
So if a higher rated opponent is being cheated, he will be able to detect it much easier. Even though I can't say for 100% sure that he was a cheater, I suspect that that was the cause of BrainFactory's demise. The evidence that I provided was completely circumstantial. Like I said, the proof is in comparing the moves to what a chess engine recommends.
This might very well be one of those cases where someone decided to make themselves look like a big shot on the internet. What else do you expect from someone named BrainFactory? This user bragged on his profile that he was working on a phd for physics at his university or something to that effect, I don't know, it has been many months since I last looked on his profile. So maybe he thought that that coupled with a very high online rating would make him look respectable on this site.
So I want to know what you, the reader, thinks about this situation. Do BrainFactory's activities on this site seem fishy to you as they did with me? Also, what other subtle cues do you think you can pick up on to spot a cheater?