Page 1 of 55

Cheating in chess

Posted: Mon Apr 13, 2015 9:50 am
by Christopher Kreuzer
Starting a new topic on this, with quotes from the Dubai thread.
David Sedgwick wrote:About the Dubai Open [...]:

There has been a cheating scandal at the event.

http://chess-news.ru/en/node/18610
Tim Harding wrote: Shocking indeed. If true the player has been doing this for a long time. Can he/ will he be banned and/or stripped of his rating and GM title like the French players who cheated at the olympiad in 2012?
carstenpedersen wrote:According to the Today programme he's got a 15 year ban. Credit to him for getting Chess mentioned on BBC news :mrgreen:
Christopher Kreuzer wrote:Good. Was that his national federation or FIDE who banned him? Maybe both?

This should have its own thread.

Is this guy (with a name harder to remember than Ivanov) the highest rated player to be caught in this fashion? I suppose the French team chess scandal (I forget the names there) was just as bad, but how long did this guy go undetected for? Will there be a full investigation and report?

Reports (so far) in the Telegraph and Daily Mail.

http://www.telegraph.co.uk/culture/ches ... atory.html
http://www.dailymail.co.uk/news/article ... oilet.html
http://gawker.com/georgian-chess-grandm ... 1697356762
http://www.chessdom.com/georgian-chess- ... ournament/
http://chess-news.ru/en/node/18610
http://www.thetimes.co.uk/tto/puzzles/c ... 409318.ece

EDIT: Adding the guy's name: Gaioz Nigalidze - Georgian GM

http://ratings.fide.com/card.phtml?event=13603078

Re: Cheating in chess

Posted: Mon Apr 13, 2015 10:18 am
by Paolo Casaschi
Christopher Kreuzer wrote: how long did this guy go undetected for? Will there be a full investigation and report?.
I'm sure we'll hear about him from the FIDE cheating committee, however this might make the case that the automatic cheating detection using game analysis techniques might not be reliable, with this guy going undetected since the screening of games started.

Re: Cheating in chess

Posted: Mon Apr 13, 2015 10:22 am
by Roger de Coverly
Regarding fifteen year bans, I doubt there's been any formal announcement by either FIDE or the Georgian Federation. I rather suspect journalists have uncovered the proposals of the Anti-Cheating Committee (ACC) where fifteen years was for the second (detected) offence.

It's all gone rather quiet about the ACC's claim to be able to detect cheats by computer analysis. They now have a useful set of games with which to calibrate their methods. If he had used the same method as other games, by no means all moves would represent an engine's preference.

Re: Cheating in chess

Posted: Mon Apr 13, 2015 10:24 am
by Alex McFarlane
Part of a release from Dubai
Abdul Rahim has removed Nigalidze from the tournament and will send a report about the incident to the International
Chess Federation, which has recently established a commission to deal with cheating in chess competitions. He said players
proven to have committed such an offence will be suspended for three years from all sanctioned tournaments, and up to 15
years in case of a repeat offence.
My reading is that the 15 year ban is press speculation (or sensationalism!). FIDE will decide on the ban but 3 years would be the normal for a first offence. Anything more than this would require earlier cases to be proven.

I see Roger has come to the same conclusions as I typed.

Re: Cheating in chess

Posted: Mon Apr 13, 2015 11:44 am
by MartinCarpenter
You can't really hope to catch an otherwise competent player who is only cheating about every few moves/specific events with game screening. Its such a small signal. Game screening is more there to stop outright cheating by 'random' people.

Looking at his FIDE chart he was 2489 back in 2010 so at that point he'd have been rather stronger than engines on mobile phones. 2566 now. We'll see.

If the description of what he was doing is reasonably true then frankly he was always fairly liable to get caught. Not a good/well thought out way to do it. Someone doing something similar in a much more intelligent manner - not technically trivial but obviously possible enough in principle - is a rather scarier proposition.

Re: Cheating in chess

Posted: Mon Apr 13, 2015 11:51 am
by Paolo Casaschi
MartinCarpenter wrote:You can't really hope to catch an otherwise competent player who is only cheating about every few moves/specific events with game screening. Its such a small signal.
That is exactly my point; game screening would only catch a small number of clumsy cheaters; if you add to that the risk of false positives and the cost of running such a system, the spontaneous question is "why bother at all with game screening?"

Re: Cheating in chess

Posted: Mon Apr 13, 2015 12:02 pm
by MartinCarpenter
Fair enough :)

It seems worthwhile to me. Yes, false positives are a worry but the evidence of history (bridge for instance, also the odd case in chess) is that people are very capable of producing those by themselves! And then of going on witch hunts.

So the main contribution would hopefully to calm that sort of thing down/avoid it happening by reassuring people that the really obvious sorts of cheat are getting caught via the screening. Also hopefully a deterrence effect.

Really dedicated/intelligent cheating is sadly very hard to do much about now and only going to get harder as we go forwards/technology marches on.

Re: Cheating in chess

Posted: Mon Apr 13, 2015 12:03 pm
by Ray Sayers
Why bother cheating anyway? Apart from the immorality of it, at the lower levels of chess, there's not much money to be gained - chess just doesn't pay enough. At the higher levels, you will be watched like a hawk if you behave oddly. Grading points are well overrated - anyone cheating to bump there grade up a few points has serious lack of a life issues.

But I guess it's one of those questions where people with a little common sense will say there is no point cheating and the kind of person who just wants to win for winning sake (even if it's not actually them playing the game - e.g. using an engine for online games) will simply fail to see why they are in the wrong.

Re: Cheating in chess

Posted: Mon Apr 13, 2015 1:28 pm
by MartinCarpenter
Motive seems believable enough in this specific case? A couple of national championships (so prestige etc) and I presume decent money in this even with it being in Dubai and all.

Also strong enough that the results were at least possible in a genuine sense rather than hyper suspicious as per Ivanov.

Re: Cheating in chess

Posted: Mon Apr 13, 2015 2:07 pm
by Matt Mackenzie
MartinCarpenter wrote: Also strong enough that the results were at least possible in a genuine sense rather than hyper suspicious as per Ivanov.
Yes, its a lot like the Sebastien Feller case in that respect.

Re: Cheating in chess

Posted: Mon Apr 13, 2015 2:08 pm
by NickFaulks
Paolo Casaschi wrote: with this guy going undetected since the screening of games started.
I don't think he has played in a single event where game scores have been attached to the tournament reports. Surprisingly, that appears to include the 2014 European Championships.

Re: Cheating in chess

Posted: Mon Apr 13, 2015 2:09 pm
by Ray Sayers
MartinCarpenter wrote:Motive seems believable enough in this specific case? A couple of national championships (so prestige etc) and I presume decent money in this even with it being in Dubai and all.

Also strong enough that the results were at least possible in a genuine sense rather than hyper suspicious as per Ivanov.
Exactly my point - at the higher level, where there is a motive, you get noticed.

Re: Cheating in chess

Posted: Mon Apr 13, 2015 2:11 pm
by Christopher Kreuzer
Some semi-random questions/thoughts:

(i) Can such cheats spot each other? What if they play each other - what happens??
(ii) How widespread is it? The question everyone will keep asking.
(iii) I pity some kid somewhere who may well be pioneering some 'undetectable' method thought up their parents and/or trainer.
(iv) Will it become dangerous to go to the toilet at all now during chess games?
(v) I've asked this previously, but can't remember the answer I got: spectators and/or 'non-playing/finished playing' team-mates following games still in progress on mobile phones and analysing using a chess engine (something I've seen done at the 4NCL) will need to be careful - the practice should maybe be banned if this has not already been done.

The point in part (v) above was raised here:

http://www.ecforum.org.uk/viewtopic.php ... 71#p113271

Roger replied there, but no-one else (still less any arbiters) was willing to discuss that, it seems.

Re: Cheating in chess

Posted: Mon Apr 13, 2015 2:12 pm
by Roger de Coverly
MartinCarpenter wrote: So the main contribution would hopefully to calm that sort of thing down/avoid it happening by reassuring people that the really obvious sorts of cheat are getting caught via the screening.
Insisting as a convention if not a rule, that you stay within eye shot of the opponent and arbiters would help alleviate the obvious suspicion that the opponent had left the board to consult. I would have thought that totally undetectable for games played at FIDE's online server other than to catch those with complete replication of an engine's moves. Ever there, I am reminded of one of the Carlsen - Anand games (the "Adams" Sicilian that started 1. e4 c5 2. Nf3 e6 3. g3), where the engines were coming with as many as half a dozen moves with little difference in assessments.

Re: Cheating in chess

Posted: Mon Apr 13, 2015 2:22 pm
by Roger de Coverly
Christopher Kreuzer wrote: (v) I've asked this previously, but can't remember the answer I got: spectators and/or 'non-playing/finished playing' team-mates following games still in progress on mobile phones and analysing using a chess engine (something I've seen done at the 4NCL) will need to be careful - the practice should maybe be banned if this has not already been done.
Provided there's no contact between players and spectators with analysis, that should not be a problem. That said, the "support camps" that some players, usually the younger ones, have with them, often seem to include a computing device in use while play is in progress. Again not totally a problem provided there is no contact. The visibility of the "support camp" varies with location as well. Atrium based venues such as de Vere Sunningdale and Wokefield Park made contact with spectators more likely.

Here's an example. You've finished your own game, but there's at least one other game going on. It's an ending within the scope of tablebases, so you return to your room and look it up to establish the theoretical verdict. Obviously you don't communicate this to the player, but should you discuss it with the match captain or the rest of your team who have completed their games?