The owner of Chess.com, Erik Allebest, announced changes. He realized his site is heavily spoiled by cheats and decided to employ trained statistician to detect them. He estimates about 1000 accounts to be closed.
It was about time. Chess.com had for long a reputation of a site, where it was possible cheat and test engines, and this brought new wave of cheaters, who just cheat without shame and bad conscience. It repulses real chess players of course. Chess.com was banning cheaters, but it took very long time in each case. Considering that average user has no tools, no time and no knowledge to distinguish strong honest human player from a cheater and tournament directors have limited powers to act, this is indeed necessary.
Erik made promises in the past a lot, but nothing significant happened yet. A lot of cheaters were banned, but even more cheaters appeared and despite the fact they were very easy to detect, staff allowed them to stay for very long time. While in the year 2008 the average lifetime of a cheat was about 4 months, it became more than 1 year in correspondence section recently. The site has ten times more members, but limited resources. Staff wastes them to detect cheats in Live Chess section, however there are many cheaters who register for free without any profile in a minute, cheat and leave. This is not a job for a site administrator, but rather for a bot. The correspondence section became sad wasteland, where many good players either left or became cheaters too.
Another bad thing for a community are "useful idiots". I mean group league organizers, who don't care about cheating. Usually they are weak players and they don't bother with checking members. Sometimes they even invite obvious cheats into their groups and tournaments they run. They can't take responsibility and they are lazy to ask for reference or cheat detection. All prestigious tournaments are spoiled by cheaters, but who cares?
Who cares when "nothing" is at stake?
The whole future of correspondence chess is at stake! ICCF didn't even try. Hopefully Erik has become wise and will eventually purge Chess.com from this nasty garbage in the next 6 weeks how he promised.
Edit 08/04/2011:
The rumble of rolling skulls started!
today must be a happy day in the palaver :p
ReplyDeleteWell Polar Bear, looks like you've got a lot of work to do clearing off all the cheaters from the ACTIVE list. I'm sure it's a task you're very happy to perform though. This is a wonderful time for chess.com. May the skulls continue to roll!!!
ReplyDeleteHi Polar Bear,
ReplyDeleteHere is an interesting post on a chess site:
http://chesschat.org/showthread.php?p=316476#post316476
Let me tell you that chess.com's anti-cheating methods are far too slow in catching the cheats. But chess.com is basically Erik's business: he only does things which maximise profit short term. He should be hiring REAL players to catch the cheaters and think long term. Not just hire one statistician. In live chess there are plenty of smart cheaters (priapos23, ratsalad, lamestars) that avoid statistical detection by only using the engine in critical positions. priapos23 for example is already a good bullet player, but MAKES SURE that his blitz rating never drops to a low level, by using the engine in games where he is against other engines. Because he is probably a 2200 player, he uses it much better than some patzer. For example, once he is a rook up, say, he just goes back to human mode.
Chess.com is a joke. It has the potential to be the essential internet site. But short term planning is killing it. Erik just wants his quick buck now. Or else he would hire some strong chess players to police the cheats!
As for Alexander Horvath, I have no idea whether he was cheating or not (did not examine the games), but they were dead right in the case of Yelena Dembo.
I am not an expert in cheating in Live chess. I don't play it much, because I can play such chess OTB. My primary goal is correspondence chess. It is a pity computer-free CC isn't played seriously, because it deserves that at least to the same degree as ICCF play.
ReplyDeleteThe first problem with chess.com: anonymity and easy registration help cheaters very much and it spoils community there. Community rewards cheating instead of punishing it and my efforts to create a core of true serious honest online players (via groups and tournaments) had little response. Simple explanation: chess.com is full of anonymous trolls, brats, cheats and snoopers, while real chess players aren't willing to play seriously online. For many there, the primary goal isn't chess, but social interaction online. For me it is a mystery why Erik has been content with such situation and allows such garbage to register, IMO it must backfire. No wonder chess.com is full of cheaters who register anonymously and immediately are allowed to play rated games and join tournaments. Thus I agree, short term planning is killing chess.com, unfortunately I don't know how could chess players help. Erik has already hired IM David Pruess, but the change in cheat detection's speed wasn't noticed. An experienced player can "smell" cheating and can give opinion, but it is not a proof without proper statistical analysis.
I am sure both Alexander Horvath and Yelena Dembo are cheating bastards. Compared with pre-computer CC control group, they have been engine-positive as hell, far far above any possible fluctuation. They differ from ordinary cheaters in two things: they weren't anonymous teen-aged Rybka robots, they have been proud title-holders, thus they feel entitled to babble cheekily and slander chess.com.