Chess.com FAQs and Discussion on Cheating

I totally agree with the comments about perfect endgames being a clear sign of someone using a computer.

If you use the 'computer workout' function in the 'learn' menu then you get the chance to see how the computer would go about mating in the most efficient way. My method of mating K+R vs K is foolproof and works perfectly well, but the computer disagrees with me. Perhaps I'll take 1 or 2 extra moves, but my own approach seems more logical and is less susceptible to errors if I play quickly. I think that this silicon approach to perfect endgames extends into more complex positions too.

I have several games where the chess.com analysis suggests I have no blunders, mistakes or inaccuracies, but these all finished before move 30 (mostly resignations or early mating attacks). As soon as we get into endgame territory the inaccuracies start to mount up. I think I'm OK at endgames, but the computer analysis never, ever agrees with more than approx 60% of my moves. I haven't enough games that went into very close endgames (even material and a tense position) in which each move was completely vital for the outcome of the match, so this may have an impact on the weight of the argument.

I see that your list of cheaters has not been updated for 20 days.Maybe you've cleared the site of this menace but I hope you won't stop with the investigations :))

One thing they've never made public (and rightly so in my opinion) is how long the queue is, and how long it takes to investigate someone once they get to the front of the queue.

Let me totally make up some numbers and show why this can frustrate the user base somewhat. If the queue is 9 days long, and it takes 6 hours to investigate one player, then if you report someone an entire week can pass by without any response at all. One can then make the mistake of "deducing" that the reported individual was investigated and cleared.

Haha, I just came up with a scheme to thwart them! Find someone that you are SURE is a computer cheat (don't know how you would do that) - and then report that person AFTER you report the suspected cheat. Then once the actual computer cheat has their account closed, you can be pretty sure that the other individual has made their way through the queue. :-)

I guess that doesn't work either, right, because someone else could have reported the actual computer cheat. Better to just put our trust in Erik and the mgmt team.

I also suspect the duration of the investigation is highly variable.

I reported a month ago a cheater in Live chess. He played with me, 5 minutes games, while showing a playing strength superior to a GM in a standard game. However, he makes sure to play some games without the computer from time to time (and lose them miserably). Therefore, chess.com finds that he does not cheat. His blitz rating increased dramatically: when he start playing here, he lost (not by disconnection) to players like 1200. Now he is 2100+.

The learning materials on this site must be really amazing, if players can get from beginner to 2200-2300 Fide at least in just a month. Wow ! :)) I am also 100% confident that Erik and his team do fight against cheating. They are soo good at this that me reporting other players is simply silly. 

Yes Richie, they banned some idiots. One of them is this guy:

http://www.chess.com/echess/profile/siahkal

A 2700+ player here for at least 6 months. Really, what is the puropose of banning a cheater after he cheated for such a long time. 

By the way, look how this losers play without the computer: they are checkmated in just 4 moves, as they have no idea about opening theory. Probably they did not yet learned how to castle:

http://www.chess.com/echess/game.html?id=7515587

I'm unclear -- if someone's gotten away with it for a certain amount of time we should give them credit and let them continue to cheat?

Costelus, I think that if you put half the effort into trying to quietly working with the site as opposed to working very publicly against it that you could conceivably make a material positive impact instead of being the PR nightmare that you are now.

If you don't catch a cheater within a reasonable amount of time, what is the purpose of banning him after he cheated in hundreds of games, after so many people lost confidence and went to play on other sites?

Yes, I had a positive impact on this site. Much more than you for sure :)) "PR nightmare"? Common, these are the facts, I did not invent anything from what I said here.

You've repeatedly suggested that there is no value in reporting suspected cheaters, for example.  This runs compeltely at odds with the stated policy of the site, and this is why it is a PR problem.

If you've reported a half a dozen cheaters that did get banned, but discouraged a dozen others from reporting a single suspected cheater each then I say your impact is net-negative.

If I report a cheater and he gets banned MONTHS later, one can be sure that he was not banned because of my report.

Common, I did not report anybody who played strong, I only reported 2700+ FIDE. Obvious things, either you are Kramnik or a computer. It should not take months to settle such cases.

Also, in one case (Chessnut), chess.com found, after my report, that he was not cheating. He was banned long after, when they re-examined his case, but that's another story. Definitely he was not banned because I reported him.

All of these things I agree are valuable, and have also conceded that he is, in fact, correct about a number of points he's raised.

As I've pointed out before, however, it is the way in which he goes about performing this role that has a detrimental impact that significantly undermines any benefit that he might otherwise provide in those capacities.  The general disingenuous tone, suggesting that there is no value in reporting cheaters and at one point (earlier in this thread) nearly causing another user to leave Chess.com because of his misrepresentation of the scale of this problem.

He could easily fulfill the role you've described without these negative side effects but he chooses not to.

Well, I just spoke out obvious things for many persons who played on ICC. Any such person, when coming here and seeing some of the games of the top players (live or correspondence) smiles and says "this site does fight against cheaters? Yeah, right :)" . I bet that many leave chess.com after seeing such "high level" of play, never to return.

I am not a watchdog, I have the info about that banned player from somebody else :)

Thanks richie_and_oprah -- you can trust that the sentiment is mutual.  I will say, though, that we would disagree less often if you were right more often....

I saw on a T-shirt recently - "I could agree with you, but then we'd both be wrong"

I think it's very important that the investigations for the supposed cheater should be done faster especially If that player is playing in a tournament, because he will win his games from his group by the time he is caught and the other players from that group can't advance in the next round.This is a real problem,it happened to my cousin and you can read the same story almost on every tournament forum

Interesting point, I agree that prioritization for tournament games is a good idea.

chess.com is making money from computer assisted chess, eg different classes of database available via the game explorer dependant upon the persons subscripton rate. What next, use of rybka for gold category members, fritz avaiable for silver subscribers, etc etc. 

The use of databases and books is allowed and is not cheating.

It is annoying that chess.com does not have a clear position about what databases can be used. As it became obvious in another topic, a database of 10-15 millions of cyborg games is not quite a database in the classical sense, but rather computer-assisted play. Particularly in some sharp openings. 

Surely you've not changed your position from zero tolerance to advocating throwing in the towel altogether.

It's important to bear in mind, as another user pointed out in this forum, that Chess.com is a two-year old site that has already surpassed all of the competition in many, many areas.  Timely detection of engine use, however, is an area where there is still room for improvement as evinced from Costelus' repeated running-up-the-flagpole of any misses he can find.  Sisyphysian though it may seem, there are precedents out there, not only from other sites, but also Costelus' analysis, that demonstrate that it is in fact achievable.  Highlighting a user who got away with it for six months before being banned doesn't tell me that the detection algorithms don't work -- clearly they do, as the user got his account closed.  What I take away from such evidence is that the initial indicators that put someone in the queue for analysis need improvement.  This is why I find the repeated discouragement of abuse reports that to be so counterproductive.

Incidentally, I fully suspect that as the detection capabilities of Chess.com improve that the users massive engine databases will actually flag as positive results for engine use, as they should have an almost idendical signature.