Also, No fees for rating Rapid and Blitz tournaments

an ordinary chessplayer wrote on 01/08/21 at 02:43:21:


Good point about Horton and the rate of play. In further defense of your position, I note that chess.com looks tough for banning a lot of people regardless of whether those people cheated. But I'm not keen on sanctions against those reporting possible cheating. A public accusation, though, is something else, and Topalov did in fact get sanctioned because he accused Kramnik in an interview after their match.

Yes, but the regulations in the .pdf call for physical measures in online chess, e.g. camera-led remote inspections and searches.

By the way, I suspect chess.com uses a lot of varied data to detect cheating, including a lot of timing data. You know the phenomenon where your opponent blitzes a few opening moves, but when you make a slightly sidelinish move pauses for a long time before resuming fast play again? He's looking up the opening in a book.

So if someone defends himself by showing that his opening moves are in a book that he owns, I'm not particularly impressed with that defense. The timing of the moves matters. If he frequently took the characteristic single long pause, it's reasonable to conclude he was using his library. This hardly bears comparison with engine cheating, and he might be outraged at being accused of cheating just for that, but chess.com forbids it.

Timing data can get a lot more sophisticated than that. Making good analytic moves after a short, always-same-length consult-an-engine pause is different from making the same moves after a deep think. A correlation of thinking time with the correctness or intrinsic rating of the moves could reveal a lot over a large number of games, especially compared to the profile of the user. It's easy enough to get a sample of honest games from known players to profile the spectrum of normal human timing patterns, whether for young Anand or old Grischuk (and their class-player counterparts).

Of course, since chess.com doesn't release its criteria, we can't know all that. If you don't like those arrangements--and I can see why you don't--don't play on chess.com.

The difficulty of this issue is clear. These regulations are highly intrusive, allowing, for example, strip searches online; at the same time, they are not even close to airtight.  Yet that is not the fault of those who wrote them: there is simply no comfortable zone to occupy that is free of both intrusion and cheating. --Do I want to be required to strip-search myself on camera? No. Do I want to lose to a cheater with an undiscovered haptic communication device in his underwear? No. Can I be safe from both? No.

No legal system is or can be free of unjust results, and in general the more guilty people are punished, the more innocent people will be punished as well. Obviously it is a question of balance. In the United States, the standard of proof required in order to be imprisoned or even executed is proof "beyond a reasonable doubt," while that for losing a civil suit, which can, after all, drive the loser to bankruptcy and long-lasting garnishment of pay, is merely a "preponderance of the evidence." Even a lifetime ban from participating in FIDE-rated chess tournaments is less severe than these penalties can be. A standard whose effect is somewhere between those of the civil and criminal standards seems fine to me. 

The published regulations seem like a decent attempt at detecting the means of cheating while saying nothing about the use of statistical means to infer that cheating occurred. For example, the rules stipulate that a determination of cheating by a platform or sponsor, e.g. chess.com, will be taken as suggestive, perhaps drawing an investigation, but not as definitive. That seems reasonable so far. FIDE must adjust on the fly, but they have to start somewhere. 

In this forum, those who emphasize the legitimate fear of false convictions, star chambers, etc., have so far led the conversation. One always looks good defending the innocent, but we have in prospect the destruction of tournament chess by general rot, and one cannot prevent that by defending the innocent alone. I would like to weigh in emphasizing the cost of failing to address cheating. Serious tournament chess simply cannot survive if engine assistance becomes too widespread, any more than any community can preserve its higher functions if it fails to maintain a functioning set of laws, which necessarily involve it in some unjust applications. Laws miscarry and injustice spills out,  but when crime becomes too widespread, the community is deluged with injustice, and only gangsters can act publicly (see: Lance Armstrong).