Cheating in chess

[NickFaulks]

I want to highlight that the balance of concern seems to lean heavily towards the rights of those extremely likely to have cheated.

That is entirely your own opinion, based on nothing.

[Richard Bates]

I want to highlight that the balance of concern seems to lean heavily towards the rights of those extremely likely to have cheated.

That is entirely your own opinion, based on nothing.

The thing is that the victim of someone who has actually cheated has lost (particularly whilst online chess remains largely a non-monetary contest)... a game of chess. A “victim” of a mistaken but widely accepted accusation of having “likely cheated” has lost their reputation and quite possibly ostracisation from a game that they might greatly enjoy. I have no problem with the balance of concern leaning heavily to the latter. I think that Nick’s attitude is drawn heavily from his 4ncl experience where it is difficult to come to many alternative explanations for the performance of his first round opponent. But it’s still noticeable that many people, even when convinced that they have been cheated against personally, are prone to expressing scepticism, or at least less certainty about, in cases where they are not directly involved, especially, where they know players personally. I accept much of this can be argued against as ignorance to the claimed robustness of cheating detection methods. But i’m still not convinced.

If online chess as an alternative to OTB were to become anything other than a temporary blip then the balance in my view might switch. It might sadly become necessary to set quite a high level for “collateral damage”. But then I will have given up at that point anyway. I don’t think that online cheating is likely to spill over particularly to OTB cheating. It is just largely a consequence of opportunity, and parallel weakness in the human condition, in my view.

[Richard Bates]

A couple of other thought which perhaps might explain differing views on this (as players).

I think that simplistically there are two basic responses that players have to losing games of chess. One is to focus on their own performance (“I played badly/not well enough”). The other focuses on their opponent’s (“you were too good for me”). Both are comfort blankets in their own way - the first holds out the prospect for personal improvement in the future, the second asserts that you did as well as you could. Maybe they also indicate how people see the game - the first perhaps sees playing more as an objective/personal test - judged against the game itself. The latter more as a relative thing - performance against others.

Anyway, those who tend towards the latter group are perhaps more likely to get worked up about cheating and/or be attracted by cheating as an explanation for defeat. Both groups will have people who are scrupulously careful to have strong evidence before making accusations/accepting they were cheated against. But the latter will also have a preponderance of those who make false accusations/assume cheating where there is none. The former will have the naive “there is none so blind as those who cannot see types”. Or in my own case, perhaps do not really care...


Nothing like 7am for long rambling posts...

[John McKenna]

The whole question for OTB players of playing online can written off, of course, as done so summarily above.

Just a 'blip' in a gap in an enforced interlude of indeterminate duration and of no great consequence for some, and not only the cheats.

The giving of the benefit of any doubt at all - even the inevitable small amount that must always exist even if someone is caught redhanded and confesses that it was all due to some fantastic set of personal circumstances that conspired against them - seems to be the order if the day.

The following is an example of such wishful thinking put into words -

Nevertheless, in the interests of the right to self defence, I have taken steps to inform your opponent of what is happening here.

Well done Michael. I've been uncomfortable with some of the recent posts on this thread. A fellow chess player, not a member of this forum as far as I know, has been publicly named as a cheat. I have no idea whether that's true or not, but it's a pretty strong thing to say and an error of judgement in my opinion. At the very least, mud sticks.

The evidence given? A game was posted where he had a losing position against a stronger player who blundered away the advantage. Doesn't this happen all the time, especially at rapidplay? We are told his account is now suspended for cheating but he is still listed in the tournament results. Cheating or not, better left to let the organisers to decide on the best course of action and to make any public announcements I would have thought.

The point has been made before that offenders identities and evidence against them are not being disclosed in "public announcements" by event organisers nor service providers. Therefore the following can be the only coherent recourse -

Nevertheless, in the interests of the right to self defence, I have taken steps to inform your opponent of what is happening here.

That's great, I will be fascinated to hear how in the following 4th round game he managed to achieve the following:- 12 moves of theory, followed by out of the remaining 39 moves, 36 moves are the 1st choice of Stockfish, 2 moves are the 2nd choice and 1 move is the 4th choice.

[Event "Rated Classical game"] [Site "https://lichess.org/6G0c4hF9"] [Date "2020.07.12"] [White "ds97813"] [Black "kingrich3"] [Result "1-0"] [ECO "B23"] [WhiteElo "2003"] [BlackElo "1644"] 1. e4 c5 2. Nc3 Nc6 3. f4 g6 4. Nf3 Bg7 5. Bb5 Nd4 6. O-O Nxb5 7. Nxb5 d5 8. e5 Nh6 9. d4 a6 10. Nc3 cxd4 11. Qxd4 Be6 12. Qb4 O-O 13. Qxb7 Qa5 14. b4 Qd8 15. Nd4 Nf5 16. Nxf5 gxf5 17. Be3 Rb8 18. Qxa6 Rxb4 19. Rfd1 Qc7 20. Qa3 Rfb8 21. Nxd5 Bxd5 22. Rxd5 Qxc2 23. Rd8+ Rxd8 24. Qxb4 e6 25. a4 Bf8 26. Qb1 Qc3 27. Qc1 Bb4 28. Qxc3 Bxc3 29. Ra2 Ba5 30. Kf2 Kf8 31. Ke2 Ke8 32. Bd2 Bxd2 33. Rxd2 Ra8 34. Ra2 Kd7 35. a5 Kc6 36. a6 Kb6 37. a7 Rxa7 38. Rxa7 Kxa7 39. Kf3 Kb7 40. Kg3 Kc6 41. Kh4 Kd5 42. Kg5 Ke4 43. g3 Kf3 44. Kf6 Kg2 45. Kxf7 Kxh2 46. Kxe6 Kxg3 47. Kxf5 h5 48. e6 h4 49. e7 h3 50. e8=Q h2 51. Qh5 {Normal} 1-0

Thanks to Nick B for posting another link in a chain of evidence of what amounts to a clear case in point of a performance that has to be seen in order to NOT be believed.

I believe the evidence I have seen when I looked in the round, by round, at the tournament performance in question and I do NOT believe it was due to either luck or a purple patch in an Indian summer of the player's chess life, nor a combination of the two. The only other option is a Minor miracle out in the Open, and who would credit that?

Who here does believe it could havd been a genuine individual unaided performance?

A game v. an anonymous opponent was posted in another thread, see -

viewtopic.php?f=2&t=10824

And, comments and opinions asked for as to whether, or not, the opponent was using assistance. Several people replied, including about specific moves.

Nobody is saying anything much at all about the moves of the two games that Nick B has posted. (The opponent's other 2 from the tourney are also available and I think they add to the evidence against a genuine performance. He got a 1/2 point bye in R2 for some reason.)

How could anyone, including the player himself, explain it in specific detail other than - playing a preponderance of moves that are not realistically possible to all be the players own moves, and thereby winning and drawing games in an Open tournament against significantly stronger players than himself?

As pointed out, by Nick B and others, it was one of the leading online hosting platforms that flagged the account in terms of it having been used to "cheat" (or to "rook" if you prefer) opponents in what was a serious competition organised by a prestigious chess league.

Not making it public amounts to sweeping it under the carpet and brings chess and players into disrepute, which is of greater import than the chess reputation of any individual player.

By all means give the benefit of the doubt where it is due but not where it cannot and should not be.

[Matt Bridgeman]

The player in question has a lengthy wiki page;
https://en.m.wikipedia.org/wiki/Derek_S ... _unionist)

[Adam Raoof]

please take advantage of pgnspy to conduct your own research, using a sample of all the games (5 or 6) from a tournament and compare that with a control sample from a reliable top level event over the board. This will give you a feeling for what is normal and what is not, and often my instincts about a game are not reliable, so I err on the side of caution, as arbiters should.

My experience is that people who make accusations of other players during events have the same issue or have other motives. It is far better to let players play chess and examine the games afterwards, unless there is other (e.g. video) evidence.

[Chris Rice]

Great article by a guy who is a GM and has a Masters Degree in Computer Science regarding school children cheating at online chess.

The penultimate paragraph is something I had not thought about at all:

"We may be creating an environment that is naturally tempting young chess players to take the immoral path. There is no easy solution to such problems. My feeling is that when a player sits in front of their computer to play a game of chess, the more informed they are about the vetting process, the less likely they are to cheat. Moreover, if they feel like they are monitored, and there will be a social cost to their actions, they will behave better."

[John Hodgson]

"I will be fascinated to hear how in the following 4th round game he managed to achieve the following:- 12 moves of theory, followed by out of the remaining 39 moves, 36 moves are the 1st choice of Stockfish, 2 moves are the 2nd choice and 1 move is the 4th choice."

(the comment on the kingrich3 game a few posts higher up this thread)

A) what is theory? In this case it seems to be defined as the point at which a position is not recorded in a database (12 in this case). So bad moves played by 'lower-rated' players have now become 'theory' (I do not mean that the moves in this game were weak, but they might have been)

I don't know how to play the classical Kings Indian, for example, but I would expect to be able to play down quite a few moves of 'theory' if weaker moves are included.

B) do the moves in the game suggest computer assistance? It is a good game, but most (all?) of the moves seem fairly normal and human. There is a little flurry around the Rfd1/Nd5 moves but nothing suggesting particularly deep calculation. The ending is one that a human might be expected to see a long way ahead as there are not many deviations from the main line.

I do not know the OTB playing strength of the White player, but on its own I do not think this game is evidence of cheating.

[Roger de Coverly]

I do not know the OTB playing strength of the White player, but on its own I do not think this game is evidence of cheating.

It was in the 1600 to 1700 range and had been for many years. It was someone who usually entered Opens but who had not shown that he could face 2000 players on level terms, let alone take them apart.

[Matt Bridgeman]

The official top three in each section have been announced. They are the same players as shown already on chess results; http://www.4ncl.co.uk/fide/online/winners_1.htm

[John McKenna]

"I will be fascinated to hear how in the following 4th round game he managed to achieve the following:- 12 moves of theory, followed by out of the remaining 39 moves, 36 moves are the 1st choice of Stockfish, 2 moves are the 2nd choice and 1 move is the 4th choice."

(the comment on the kingrich3 game a few posts higher up this thread)

A) what is theory? In this case it seems to be defined as the point at which a position is not recorded in a database (12 in this case). So bad moves played by 'lower-rated' players have now become 'theory' (I do not mean that the moves in this game were weak, but they might have been)

I don't know how to play the classical Kings Indian, for example, but I would expect to be able to play down quite a few moves of 'theory' if weaker moves are included...

Up to and including 10.Nc3 the opening followed the stem game GM Minasian,Artashes (2567) v. IM Froehlich,Peter ((2451) Linares Open 2001, a draw in 58 moves, and then with 9... cd (c4 was the IM's move) 10.Qd4 a6 11.Nc3 Be6 12.Qb4 followed Jotic,Vladimir (2277) v. IM Cosma,Ioan (2461) 0-1 in 48 moves, Subotica 2002.

Black then lost a pawn with 12... O-O? (Qd7 was the IM's move) and with it the game. There's the similarity to the other game Nick B posted - the stronger player cracks first not the weaker.

B) do the moves in the game suggest computer assistance? It is a good game, but most (all?) of the moves seem fairly normal and human. There is a little flurry around the Rfd1/Nd5 moves but nothing suggesting particularly deep calculation. The ending is one that a human might be expected to see a long way ahead as there are not many deviations from the main line...

Perhaps someone will do as Adam R suggests and subject the game(s) to "pgnspy" and indicate here whether or not the moves are superhuman.(I've already given my opinion about that and await a statistical reply - if anyone cares to make one.)

I do not know the OTB playing strength of the White player, but on its own I do not think this game is evidence of cheating.

Current ECF grade(s): A130 slowplay (not over 145 and not graded for rapidplay in available years)

Current FIDE rating(s): 1632 standard (not over 1700 and unrated at rapid & blitz in available years)

Thanks to John H for taking the time to comment on the moves and the game. Having lost to him comprehensively once - a long time ago - I appreciate & value his input, but must say that as a player I am closer to the weaker party and know how difficult it is to do a performance such as the one - achieved over 4 games - in question.

The official top three in each section have been announced. They are the same players as shown already on chess results; http://www.4ncl.co.uk/fide/online/winners_1.htm

That just leaves -

Best performance trophy in each section

Not an easy task either where online chess is concerned.

[Kevin Thurlow]

The player in question has a lengthy wiki page;
https://en.m.wikipedia.org/wiki/Derek_S ... _unionist)

"They live in an £800,000 grace-and-favour house in Berkhamsted, Hertfordshire provided by Unite for the duration of Simpson's entire lifetime or that of his most recent partner, whichever is longer."

I can see why he left the Communist Party...

[John Hodgson]

Reply to John McKenna:

Thank you for your comments. On reflection, I think I may have significantly underestimated the skill level of the White player (human or computer) in the game. I was misled because on their own there was nothing unfathomable about any of the White moves - unlike when I analyse games with a computer - but taken as whole the game had a strong logic and was without obvious mistakes. Clearly, depending on the type of game, there may be no obvious 'engine move', or series of engines moves, but the totality might be beyond most amateur players (so I would have definitely stumbled at some point).

Aside for this game, some moves do stand out as being obviously engine-generated, but I am not sure about how to identify cheating in the opening. Because I play main lines on most occasions, many of my games go past move 15 - and some beyond move 20 - before an original move is played. On one or two occasions I have corrected my own games around the 20-move mark - thus cheating by following my own games and my own memory. And as others have said, engine improvements may have been found at home before the game. Playing '12 moves of theory' is hardly suspicious.

If I wanted to cheat I would move pieces around on a separate board. That would give me some improvement without generating engine moves, and I would morally justify it to myself by saying the moves were all the results of my own efforts. I will never do this.

[Roger de Coverly]

but I am not sure about how to identify cheating in the opening.

I don't believe the various engine detection programs are capable of doing this with any reliability. If playing online without video oversight, only the player will know whether they consulted their bookshelf during the game where a modern book is likely to contain analysis that at the very least has presumably been engine checked. There again the player may have remembered it.

[Ian Thompson]

Clearly, depending on the type of game, there may be no obvious 'engine move', or series of engines moves, but the totality might be beyond most amateur players (so I would have definitely stumbled at some point).

For what it's worth PGNSpy reports that there were 20 or 21 positions where the result of the game wasn't clear. White played the first choice of Stockfish 55-60% of the time, the first or second choice 85-90% of the time and one of the top three choices every time. Suboptimal moves were never more than 0.5 pawns worse than the best, with 3 more than 0.25 pawns and 3 less than 0.25 pawns worse.

I've been a bit imprecise with the figures because I ran PGNSpy twice and got slightly different results. I don't know why. I thought it would always give the same results with the same input parameters.