Cheating in chess

[Chris Rice]

Aren’t we only speculating here, nobody has seen any evidence

If the algorithms/test has been done over multiple games showing a very high correlation of best moves then it is high probability some assistance has occurred

I have hovered around 2200 and OTB had one good tournament where I performed at 2400 over 9 rounds; now suddenly this has occurred in multiple occurrences in one event, coincidence?

It may not be perfect, but we either ignore it and allowing cheating fullstop in the online sphere; or accept the cheating algorithms for what they are and also accept their outcomes

Nobody here is actually coming up with any practical alternatives, just complaining; they are never going to win the online servers

As I see it there are really two courses of action. The first is, what you suggest, that we accept the cheating algorithms and the innocent and the guilty get swept up as one. Nobody complains, we just accept it with no presumption of guilt or innocence. How many of us are willing to do that though? The second course of action is to make a slight rule change or two to the game itself so that computers would suddenly become completely useless.

[Jacques Parry]

At some point we should maybe separate out a discussion about sophisticated cheating, by which I mean partial cheating, employing an engine only now and then

Or employing an engine all the time but making one's own move unless the engine says it's a blunder. We patzers could raise our playing strength a lot by doing this.

[Roger de Coverly]

Or employing an engine all the time but making one's own move unless the engine says it's a blunder. We patzers could raise our playing strength a lot by doing this.

Essentially that's a speeded up version of ICCF rules, or the Kasparov experiment of "advanced chess".
https://en.wikipedia.org/wiki/Advanced_chess

[JustinHorton]

At some point we should maybe separate out a discussion about sophisticated cheating, by which I mean partial cheating, employing an engine only now and then

Or employing an engine all the time but making one's own move unless the engine says it's a blunder. We patzers could raise our playing strength a lot by doing this.

Well yes, this is the sort of thing I'm talking about. How much do we know about how much of this goes on? (I'm guessing, very little.) Do we know how effective it is? (I'm guessing that time limit plays a role here.) Do we know how easy it is to detect? (I'm guessing, not very.)

[Alan Walton]

I am no expert in the matter (in fact most of us here aren't as well); but the only way, without actually having spyware on your computer, is to use the probability of strong moves made against that expected for the players strength

So personally speaking if you want to play online chess you more or less have to abide by the current systems (what did people expect in the 4NCL, they made the cheating conditions quite obvious and people entered on the back of these), maybe with more focus on this in the present climate somebody comes up with a new methodology, but perhaps they have been trying at this is the best they have come up with, especially with the resources they have

[MartinCarpenter]

You couldn't even trust spyware to be honest. Not when its running on my computer in my house. So many really quite straightforward ways to engineer round it.

[Jacques Parry]

Or employing an engine all the time but making one's own move unless the engine says it's a blunder.

Well yes, this is the sort of thing I'm talking about. How much do we know about how much of this goes on? (I'm guessing, very little.) Do we know how effective it is? (I'm guessing that time limit plays a role here.) Do we know how easy it is to detect? (I'm guessing, not very.)

I have no evidence but I'm guessing that it happens a lot in rapid games (probably not blitz), because people may not be desperate to win at all costs but they don't like to spoil an interesting game with a blunder. Having played in a few rapid tournaments while on camera, I reckon it would be easy to do it and get away with it.

[JustinHorton]

I'm also guessing though that having one eye on the computer doesn't do a great deal for your standard of play when you're not cribbing from it. Perhaps you can master it with practice, but it strikes me as an unstable approach, and one likely to slide into greater and greater recourse to the engine.

(Though once again, I don't know that. We really don't know enough about what cheats actually do. Maybe the providers know a little more.)

[John McKenna]

These days the "providers" (of whatever) often get a suspected successful offender to change sides and join their team as an expert advisor. It takes a cheat to catch a cheat.

[Alistair Campbell]

Have any of these platforms tested their anti-cheating measures with in-house tournaments when certain "competitors" are given licence to cheat in one of a variety of ways, to see if they can be detected?

[Roger de Coverly]

Have any of these platforms tested their anti-cheating measures with in-house tournaments when certain "competitors" are given licence to cheat in one of a variety of ways, to see if they can be detected?

If they have, the results have never been made public. Kasparov claimed in his match against Deep Blue that the engine had received human assistance. Can the measures prove that Deep Blue was definitively a computer?

By the 1990s, engines could play at a respectable level. Have they be able to test that their measures could detect these?

[Jonathan Bryant]

Chess.com writing articles saying they're wonderful doesn't count for much. It's a bit like some of the websites advertising nutritional supplements with great medical claims, endorsed by an "expert". When you search for the "expert", you find they have no science qualifications at all, and all the stuff they've published is making dubious claims for their friends.

Some of the websites ... like Chessbase or the twitter account of the former Times Chess correspondent, for example ? Albeit not about 'suplements'


With regard to the question of occasional cheating with engine assistance > several years ago when I used to play fairly regularly online at 3 days a move I played a pair of games against the same opponent that followed exactly the same pattern. (A) My opponent played terrible chess from the opening (b) they went a piece down without much effort from me (C) they suddenly became extremely difficult to put away. I’m as sure I can be in the circumstances that he was cheating and turned an engine on when he was losing.

I think this kind of 'mostly off but turn the engine on in certain circumstances' pattern would be very difficult to sustain OTB. By which I mean you might want to just look once a game at a particular tactical moment. This would definitely be a great benefit and would raise your rating quite a bit I would think. But how long until you start telling yourself, "I’ll just have a second look in this game. It’s OK because ..." and then it’s the third, fourth, fifth look and eventually you’re doing it all the time.

Online, I might say I’ll choose a move and then look at an engine and only switch if my choice is a blunder. That alone would be worth a lot of rating points but how long would it last? Half a dozen moves at most, I’d think. And then you’d just be following the computer every time.

[Richard Bates]

Chess.com writing articles saying they're wonderful doesn't count for much. It's a bit like some of the websites advertising nutritional supplements with great medical claims, endorsed by an "expert". When you search for the "expert", you find they have no science qualifications at all, and all the stuff they've published is making dubious claims for their friends.

Some of the websites ... like Chessbase or the twitter account of the former Times Chess correspondent, for example ? Albeit not about 'suplements'


With regard to the question of occasional cheating with engine assistance > several years ago when I used to play fairly regularly online at 3 days a move I played a pair of games against the same opponent that followed exactly the same pattern. (A) My opponent played terrible chess from the opening (b) they went a piece down without much effort from me (C) they suddenly became extremely difficult to put away. I’m as sure I can be in the circumstances that he was cheating and turned an engine on when he was losing.

I play games online like that all the time! Playing the role of you opponent. Just substitute the word "brain" for "engine"...

[Jonathan Bryant]

I play games online like that all the time! Playing the role of you opponent. Just substitute the word "brain" for "engine"...

Sure. But I would expect you to be able to play to a decent standard. And I would imagine you would have an online rating that would indicate you could play well when brain is on .

Not the case with my opponent.

[Richard Bates]

I play games online like that all the time! Playing the role of you opponent. Just substitute the word "brain" for "engine"...

Sure. But I would expect you to be able to play to a decent standard. And I would imagine you would have an online rating that would indicate you could play well when brain is on .

Not the case with my opponent.

Maybe - although it's all relative. I get my fair share of people accusing me of cheating (although usually after playing very good games).

It's more a general point rather than intended to apply specifically to your belief on this occasion, I just think sometimes you have to set cheating "evidence" against the fact that many people playing on line may 1) care very little about either their rating or the result 2) may play games in wildly different circumstances (in the middle of the night, in different states of inebriation, just to pass the time) 3) may sometimes play good moves by chance or on instinct, with little in the way of calculation. All of which can easily serve to undermine belief in cheating when it involves variable performance during a game or between games.