Cheating in chess

[John Upham]

The "computer movishness" of a move can be measured in terms of its deviation from the way a human will decide on a move in the context of a competitive game. Measuring characteristics are necessarily vague but can be worded as the degree of departure from normal human heuristics like safety , tangible purpose, and tactical simplicity. Humans don't play a tactically sound but very complicated move when there's a simple-for-humans alternative move leading to the same result.

Could you provide one (or more than one) concrete example that demonstrates your thinking?

A sequence of "movishness" type moves would be of interest.

[NickFaulks]

I’m not totally buying that the ECF would have had the same reaction as the Polish Chess Federation. You had a player caught cheating and banned from the Glorney Gilbert Online International who was selected less than 2 months later for the European Youth Online Championship.

The English cases were online, the Polish one was OTB. There is, and I hope there will remain, a world of difference.

[Matt Bridgeman]

I’m not totally buying that the ECF would have had the same reaction as the Polish Chess Federation. You had a player caught cheating and banned from the Glorney Gilbert Online International who was selected less than 2 months later for the European Youth Online Championship.

The English cases were online, the Polish one was OTB. There is, and I hope there will remain, a world of difference.

I’m sure you’re right. But I suppose there’s a difference between getting caught out playing chess.com challenge games and say playing for your country in an annual event which dates back to 1949, even if it happened to be the first time the event had been held online. I'd guess that if one of the the England Olympiad team had been caught out some actual action might have stirred into motion, as the event might have attracted a little more media attention.

[Brian Egdell]

Could you provide one (or more than one) concrete example that demonstrates your thinking?

A sequence of "movishness" type moves would be of interest.

This article, though dated, gives a good example along with an entertaining and highly readable account of a computer cheat in a tournament in Germany back in 1999. This was a long time ago in terms of modern computer chess software and cheating, but the principles remain the same.

The game below, featured and annotated in that article, is of particular interest.

[Event ""] [Site ""] [Date "1999.12.30"] [Round ""] [White "Allwermann, Clemens"] [Black "Kalinitschev, Sergey"] [Result "1-0"] 1. e4 c5 2. Nf3 Nc6 3. d4 cxd4 4. Nxd4 e5 5. Nb5 d6 6. c4 Be7 7. Be2 a6 8. N5c3 Nf6 9. O-O Be6 10. Be3 O-O 11. Na3 Nd7 12. Qd2 Nc5 13. Nc2 f5 14. exf5 Bxf5 15. Bf3 Kh8 16. Bd5 Qe8 17. Rad1 Qg6 18. Na3 e4 19. f3 exf3 20. Bxf3 Ne5 21. Nd5 Bh4 22. Nf4 Nxf3+ 23. Rxf3 Qe8 24. Nd5 Ne6 25. Rdf1 Qg6 26. b3 Rf7 27. Nc2 Ng5 28. Bxg5 Bxg5 29. Qf2 Bxc2 30. Rxf7 Bf6 31. Qa7 Rg8 32. Qxb7 Be4 33. Nf4 Qf5 34. Qd7 Qe5 35. Kh1 g5 36. Nh3 g4 37. Nf2 Bf5 38. Nxg4 Be4 39. R7xf6 Bxg2+ 40. Kxg2 Qe4+ 41. Kh3 1-0

The moves 31.Qa7, 36.Nh3 and 38.Nxg4 are all moves with a high "computer movishness" (my own grotesque term!) quotient as the article explains using much better terminology than I do.

[NickFaulks]

What's lacking in current rules, as perhaps illustrated in the Polish case, is a mechanism of warning players who are frequently absent from the playing room that their conduct might be seen as suspicious.

I agree. It is at a minimum unsettling to the opponent when a player leaves the playing hall after every move for a purpose unknown. They would not be allowed to leave the hall after every question in an A level or a professional exam and I have never understood why competitive chess is different.

[NickFaulks]

The moves 31.Qa7, 36.Nh3 and 38.Nxg4 are all moves with a high "computer movishness" (my own grotesque term!) quotient as the article explains using much better terminology than I do.

I always play flashy moves such as those if I notice that they are legal. Sometimes you've missed something and a winning game is lost, but it just has to be done.

[John Upham]

What sort of rating is required for moves described as computer moves to transition to being human moves?

2500?

2700?

Hopefully this question is not too ambiguous.

In other words, at what rating level would you look at a series of moves and not been concerned a player is using a mobile telephone in a toilet cubicle?

[IM Jack Rudd]

It seems unambiguous, but I'm not sure the underlying assumption is correct.

[NickFaulks]

Hopefully this question is not too ambiguous.

Sorry, but I don't think it is even meaningful.

Playing every AlphaZero move will presumably get you a 3300 rating. Playing some but not others may be of no value at all.

[Roger Lancaster]

What's lacking in current rules, as perhaps illustrated in the Polish case, is a mechanism of warning players who are frequently absent from the playing room that their conduct might be seen as suspicious.

I agree. It is at a minimum unsettling to the opponent when a player leaves the playing hall after every move for a purpose unknown. They would not be allowed to leave the hall after every question in an A level or a professional exam and I have never understood why competitive chess is different.

Agreed. I'm aware that some people have weaker bladders than others but, if someone needs to pay a visit after every - or almost every - move, I think a medical certificate for production to the arbiters is the least one might ask for. Otherwise, that someone is [whether intentionally or not] playing mind games - his or her opponent is left to reflect that there is a possibility, however remote, that something untoward is taking place.

[Andy McCulloch]

Do we really need to get a medical certificate to go to the loo?
Due to 'underlying conditions' I personally have been on diuretics for nearly two decades. Prior to today only I knew that my personal circumstances had been used at a FIDE anti-cheating seminar, as an example of how frequent toilet visits should not always be regarded as suspicious. When asked how he knew I was not cheating, the answer was simple, 'I look at his results'.

[Li Wu]

John- you should stop looking at rating and start looking at moves only. Rating is terrible single point of data compared to a full analysis of someone's moves in every game (not saying current anti-cheating methods go this far, just what's potentially possible). People have various leaks in their game and may have vastly different probabilities of making some moves, even with the same rating.

The more easily caught cheats will have unusually long string of perfect/near perfect engine moves that no human in history has made (when filtered correctly as explained numerous times on these forums). Or in some sample of games, analysis of qualifying moves have low enough P-value.

Also statistical evidence is, in my view, far stronger than physical evidence in some respects that isn't mentioned here. One is that it can be independently verified by anyone (e.g. Professor Regan, from far away, and all of us). Even going as far as being caught with a phone and the position on board in a Stockfish app- what does this mean? Well we have 2-3 live witnesses that know the truth, and everyone else has to rely on credibility of said witnesses.

[Roger Lancaster]

Do we really need to get a medical certificate to go to the loo?
Due to 'underlying conditions' I personally have been on diuretics for nearly two decades. Prior to today only I knew that my personal circumstances had been used at a FIDE anti-cheating seminar, as an example of how frequent toilet visits should not always be regarded as suspicious. When asked how he knew I was not cheating, the answer was simple, 'I look at his results'.

I thought my original post recognised the existence of medical conditions of the type you describe. Overly frequent toilet visits should, in my view, be regarded as suspicious [not conclusive] in the absence of such a condition - after all, in the absence of such a condition, why the need?

[Matt Bridgeman]

I wonder if a bit like other sports the cheaters are starting to outwit the detection methods. A process which has probably been accelerated somewhat by greater numbers playing online post virus. Methods are there, as an easy search on YouTube will show you. There certainly seems to be plenty of players totally exceeding their best ever over the board form week in, week out in the various 4NCL online competitions and seemingly not getting caught.

As an aside, the latest episode of the new series of Billions throws up an interesting method of the future perhaps. Advances in focus drugs by the pharmaceutical companies. Obviously they exist now, but surely in the near future they will become ever more impressive.

[JustinHorton]

Also statistical evidence is, in my view, far stronger than physical evidence in some respects that isn't mentioned here. One is that it can be independently verified by anyone (e.g. Professor Regan, from far away, and all of us)

I'm not sure I follow this. How can I do more than verify the existence of the data?