Monday, April 04, 2011

von Neuman Ree visited

Today I want to return [What Happened Next? XVI; What Happened Next XVI] to John von Neumann. More specifically, I want to take another look at Hans Ree’s account of what happened at the 1993 World Open.

The most interesting thing about Shoes on Fire (reprinted in The Human Comedy of Chess, Russell Enterprises 1999) is not so much the facts of the case per se. As Richard James’ comment to last Monday’s post shows, you can find these, albeit with variations, all over the internet. No, what really stands out to the modern eye is Ree’s insight that von Neumann would come to be seen as just the tip of an iceberg towards which the good ship Chess World was steaming. Allowing ourselves the benefit of the eighteen years’ worth of hindsight unavailable to the Dutchman, we can see that he was entirely right, albeit perhaps in a way that he had not intended.

In the years since Ree’s article, cases of cheating with computers have surfaced with ever-increasing frequency. If von Neumann was a warning then so was the introduction of Pocket PCs running cut down versions of Fritz, and so is the emerging possibility of running the strongest engines on nothing more than a common or garden mobile phone. The alarms have gone unheeded, though, and we find ourselves no better prepared to deal with the spread of the silicon menace than we were in 1993.

With the recent decision in the French Olympiad case and the emergence of a British example of computer cheating (more on that tomorrow), another look at Ree’s two-decade old prophesy seemed more than timely. After reviewing what has happened since Shoes on Fire and the extent (or lack thereof) of the chess world’s readiness to deal with computer cheating, perhaps, like me, you'll be left feeling both concerned and optimistic: troubled at the situation in which we find ourselves, hopeful that the problem can be tackled and yet worried about where we’ll end up if it isn’t.

A superb book -
and Big Malc will flog it to you for a mere £7.50

Revisiting Ree's article to discover how he came to the conclusion that cheating by computer had a big future, we see that his starting point is the rank incompetence of the attempted fraud. von Neumann was so inept – his use of technology so ham-fisted, his choice of pseudonym so transparent – that Ree finds it impossible to take him seriously. von Neumann, he concludes, must have had a bigger plan in mind than the simple attempt to grab a few dollars away from some gullible amateur chessers.

If there was more to von Neumann than met the eye, what could he have been up to? Was he a decoy designed to attract attention away from a colleague who was using equally dubious methods to secure an even bigger prize? (The ‘smokescreen theory’ that David Sedgwick and I briefly discussed at Golden Lane last Monday night.) Was he an agent provocateur? Hoping to clean up in the libel courts after goading somebody into saying something that couldn’t be proved?

Ree’s ‘fiendish plot’ hypothesis could be true, I suppose, but I doubt it. As an erstwhile magician and one-time student of criminology, I can say with some confidence that the methods employed by those who make their living by deception are typically much less intricate than those imagined by the folk left trying to explain what they have seen. In the absence of any further evidence, then, I don’t particularly feel the need to consider von Neumann to be anything other than he appeared to be.

Where Ree was on the money was spotting the future significance of electronic cheating. He was right for the wrong reasons, perhaps - he thought that the ’93 World Open was a trial run for a group of organised criminals who wanted to know whether setting up shop in the chess world was worth their while - but he knew full well that there were many more John von Neumanns around the corner.

“It will probably be quite a while before someone knowing nothing about chess except for its notation will manage to penetrate into the ranks of elite players ...”, Ree said. I wonder if he has been surprised at the rate of change since the late 1990s. If it wasn’t the case that the best of humanity ceased to be a match for Stean’s Bloody Iron Monsters when Kasparov lost to Deep(er) Blue in 1997, it’s certainly true now.

Chess engines have become stronger and computer technology has become more portable and affordable. At the same time those, like von Neumann and the three Frenchies, who require communication with a third-party to enact their dirty deeds find mobile phones ubiquitous in a way that was unimaginable back in 1993. As a result cheating at chess has become democratised. Once restricted to people who were in contact with somebody who could both play the game and be trusted to co-operate/stay quiet, underhand methods to win games of chess are now available to anybody willing to make use of them. No surprise, then, that since Shoes on Fire was written silicon-powered cheating has expanded exponentially.

Consider this roll of ‘honour’: a list of cases in which it is either proven or strongly suspected that cheating by computer took place:-

In the circumstances, it is hardly surprising that unfounded allegations -  the best known examples being Nigel Short at the 2008 Commonwealth Championship and Mamedyarov at the Aeroflot Open 2009 - are on the increase too. To these we should also add the incidents that are difficult to categorise: Topalov at San Luis 2005, Elista 2006 and Wijk aan Zee 2007 for example. What do we call these? Cynical attempts to smear an opponent? Unwarranted suspicions? Unproven suspicions? A player caught, but getting away Scot free? Your guess is as good as mine.

Anyhoo, the increasing frequency of computer-based cheating over the last five years in particular is glaringly apparent. Nevertheless, chess authorities at all levels - international, national and local - are still completely unprepared to handle these incidents.

FIDE’s website includes a page dedicated to news and announcements from our favourite game's governing body. Recent articles include things like “Congratulations to Vasily Filipenko” - “Dear Vasily Aleksandrovich! On behalf of the World Chess Federation and me personally, I cordially congratulate you on your election as Chairman of the Duma and the Head of Khanty-Mansiysk city!” - and “Caution” - “This is to inform all chess players / organisers / officials that any chess event organized under the banner of "Amateur Chess Association" is not recognized by FIDE." - but, even though the Frenchmen were caught cheating at FIDE’s own world team championship, you will search in vain for any comment on either their actions our how FIDE intend to respond. There’s not even so much as a “Do you know, on the whole we kind of tend to feel that secretly using computers to help you play a game of chess is sort of a little bit naughty”. Nada.

As long ago as 2007, the Association of Chess Professionals contacted FIDE to request that something be done about cheating in chess. Unfortunately Kirsan and his chums must have been busy with other matters for the past four years because, clearly, nothing happened as a result of that intervention.

I’m surely not the only one to have been surprised and somewhat concerned at Andrew Farthing’s response to a question posed on the EC Forum:

“The formal advice that I have received indicates that "the ECF does not have any legal mechanism" to punish players for cheating. Sanctions are possible within the specific event (in accordance with the laws of chess), but more far-reaching sanctions - such as a fine or a ban - appear to be problematic.”

Actually, I know I’m not because Andrew continues,

“Speaking for myself, without the benefit of discussion with the Board, this situation is unsatisfactory and needs to be clarified.”

Credit is due to the ECF’s Chief Executive Officer, both for engaging in the debate in the first place, and for acknowledging a poor state of affairs and setting himself about the task of putting matters right. That we are in a far from ideal situation is not Andrew Farthing’s fault. His responses thus far do at least give some cause for optimism that, as far as English chess goes, things will change for the better in the coming months.


Finally, what of your local league? How confident are you that procedures are in place to deal with an occurrence of cheating in your area? To be fair to our league and club officials – who do a fairly thankless job after all – this kind of incident would be very difficult and unpleasant to handle, especially when we factor in the observation that (suspected) cheating is often accompanied by threats of legal action from those accused. All the more reason, I would suggest, for each league/club/chess association to lay down in advance a clear process to be followed as and when the need arises.

Hans Ree
Chess Visionary and first-rate karaoke singer
Photograph from Chess Vibes

Hans Ree closed Shoes on Fire with this conclusion:

“I think it was a test.”

We can see now that he was absolutely right. It wasn’t some Sopranos-like bunch of ne’er-do-wells being examined, though. It was us. Twenty years ago the chess world coped with John von Neumann’s attempts to swindle himself a few bucks. Ree asked if we were prepared for a future when cheating with computers was a considerably less amateurish affair. I’m not sure that we were back then; I don't know that we are now.

Is all this a lot of fuss over nothing? Even with the escalation of cases in the last five years, we're still only talking about an average of one instance anywhere in the world every six months or so after all. That's certainly true, but consider this: these are just the ones that we (or I, anyway) know about. They are the most egregious examples; the ones where the person concerned got caught.

When you examine the details, you can also see that they are, like von Neumann, cases that often involve a fair degree of incompetence from those concerned - absurd tournament performances; multiple unexplained absences from the board; 100s of points of rating increases within a few months; layers of clothing worn in hot weather. Even the French Connection, a cheating ring with a sophisticated coding system, was rumbled because they used a mobile phone that belonged to the French Chess Federation's Vice President. As little as £20 spent on a pay-as-you-go mobile instead and they could still be up and running.

The idea that the dozen or so instances listed here are the only cases of chess cheating beggars belief. There are, I'm quite sure, people out there who were and are getting away with it.  Bearing that in mind, imagine the rapidly-increasing rate of known cheating by computer going on for another 18 years. If nothing at all is done, chess could end up somewhere rather unpleasant indeed. Probably just as well, then, that the ECF, if not FIDE, are starting to give the matter some serious consideration.

Shoes from here.


Jonathan B said...

An anonymous commenter writes:-

"The Australian junior caught cheating in 2009 was [Name given], who was given a two year ban by the Australian Chess Federation."

Hello anonymous, thanks for your comment.

Actually after reading the initial report I discovered the name of the person concerned online. I'm not sure I find it appropriate to publish, though, hence not mentioning it in the article and cutting it out of your comment here.


Jonathan B said...

Not that I'm fully familiar with the circumstances of that case, btw, but from what I know of it 2 years sounds about right. More on that tomorrow.


ejh said...

I mostly agree with this, but I have a couple of caveats.

One is that I'm not convinced that anybody's sucessfully getting away with it long-term, and I don't think I will be until somebody'sm caught and is shown to have been at it long term. I think the problem of hiding your real, much-weaker playing strength is considerable.

The other is that in the case of national federations (and I don't grant FIDE this excuse, because they're much more powerful and wealthier) I think there is a sizeable fear of getting sued if they took drastic action against anybody. We don't really know, for instance, whether a chess federation would really be allowed to ban somebody long-term from playing competitive chess, and we wouldn't until such time as it were tested in a court. And it may be that federations feel that they don't even have the money to contest such a case, let alone risk losing it (because you can never guarantee that it will go your way, however strong you feel your case may be).

I don't suggest this is an adequate reason for not acting, because if federations don't, then ultimately chess will be in a lot of trouble, if indeed it's viable at all. But the problem does at least need to be acknowledged and understood. A lot of things in the sporting world changed after Tottenham Hotspur sued the FA in 1991, and one of them was that we understood that the decisions of sporting federations could be overridden by the law.

Jonathan B said...

An anonymous commenter (our friend from above?) writes:-

"Info on of the Australian junior cheating case were detailed in [URL given] and of the subsequent ban in the Australian Chess Federation newsletter [URL given]. Both sources named the junior involved - why are you so coy? Should the journalist who published his game have omitted one player's name, turning him into a non-person as in the old USSR?"

Thanks, again, for the comment. I'll answer it a little later today.

Jonathan B said...

Until I have the time to respond more fully, today's post - particularly the final paragraph or two - will shed some light on why I'm reluctant to name juniors caught cheating unless it's absolutely necessary.

The question is also discussed on this thread - to which I've contributed.

Jonathan B said...

thanks once again to our anonymous friend for the second comment.

The reason I chose not to name the person involved in the 2009 Australia case in this article (or provide direct links to websites that do so) is simply because I think the situation is different when a junior is involved than when the person found to be cheating is an adult. Unless there's a particular reason to name an under 18 in this sort of case - e.g. if subsequent to the ban they were found to be trying to get into tournaments anyway - then I feel it's best avoided.

Other people evidently take a different view. That's fair enough. I'm not trying to police the whole internet, just run my little bit of it.

Your analogy with the "non-person"s of the old USSR is interesting, but I think flawed. With Korchnoi, for example, the Soviet press refusing to publish his name was a punishment for him (and intended as a warning, I've no doubt, for anybody else who might be thinking about defecting). It was a method of exclusion, then.

In the case of juniors, the intent behind not naming them is inclusive, i.e. it's a recognition of the fact that many of us to stupid things when we are young and perhaps should be allowed to forget them when we become adults. It is very much not a punishment and is, I would say, the opposite of what happened in the old USSR.

Anybody who wants to find out that Australian kid's name can find it easily enough if they want to - I did without even trying. The same is true of the recent case in England according to folk on the EC Forum.

That's fair enough, but I see no reason to help people along the way. Like I say, with an adult it would be a different matter entirely, but this lad is/was 14. Different standards apply, I think.

Anyhoo, thanks once again for your comment.


Jonathan B said...

One is that I'm not convinced that anybody's sucessfully getting away with it long-term, and I don't think I will be until somebody'sm caught and is shown to have been at it long term.

One day I might get around to writing along the lines of: "What I learned from my MSc in Criminology and how I used it to develop some hypotheses about cheating in the chess world".

Two points to note here:-
(a) It will always be difficult to evaluate the true extent of cheating at chess because, like crime, people who do it have a vested interest in keeping their activities hidden. When they are discovered they also have a vested interested in minimising the extent of their activities.

(b) Full-time criminals are very rare. The vast majority of crime (or at least the crime that we know about - see above) is committed by amateurs and hobbyists. Semi-pros at best.

What would this mean if reflected in the chess world? People won't necessarily cheat all the time. This could potentially make it harder for them to be uncovered and they could well, therefore, be getting away with sporadic cheating on a long-term basis.

Oh, what the hell, I'll give you a third.

With chess cheating, like crime, the perpetrators we catch will be the incompetent ones. The ones who seem to go out of their way to be caught. The ones,
like the German chappie in 1998, who can't resist announcing a forced mate that even Vishy Anand can't spot;
like von Neumann who stare at a position for 40 minutes before playing an obvious forced move

and so on.

Anonymous said...

Computer cheating is a special case of consulting third parties during the game, which has always been banned in over the board chess.

More could be done to prevent casual cheating, for instance non-players should be strongly discouraged from running chess playing or database software on devices within sight of players. Players themselves should be advised not to wander out of sight of their opponent or the arbiter except for limited exceptions and shouldn't be talking during games.

As far as the actual incident in the British event was concerned, precise details are unknown except perhaps to the witnesses.

The punishment, so far, was exclusion from the event and cancellation of all points previously gained. Whether the ECF will impose (or even has powers to) a formal ban remains to be seen. Tournament organisers can run a blacklist even without ECF sanction. They usually reserve the right to refuse entries.

I'm inclined to think that casual cheating (if this is what it was) is less serious than the grand scale attempt alleged as part of the French Connection.

ejh said...

The thing about not getting out of sight is, do we not suspect that a major location for cheating would be the toilet?

Anonymous said...

Commenting on EJH's latest.

I suppose it has always been possible to analyse in the toilet with a pocket set. You wouldn't really get any more than a marginal advantage if at all.

More dangerous is the electronic device. In the same way that they have to be switched off, they also need to be seen to be off and out of bounds to the player. So conventions or rules around ideas that players leave them in full view by the board, that they deposit them with the arbiters, that they disable them by removing batteries etc. may be needed. Reports suggested that the French Connection did just that, with mobile phones being placed in the custody of the match captain. Of course if the match captain is part of a conspiracy ...

ejh said...

I'm still unconvinced by the thesis that it's the incompetent cheaters that we're catching and that we can assume that there are successful long-term cheaters at large. I think it entirely possibly that some people have cheated once or twice and got away with it (hand-held devices in toilets, for instance) but I think that if you're seeking to discover a middle way between that "once or twice", and cheating as a matter of course, it's really hard to see how you would judge when to cheat and when not to.

Again, if and when somebody is caught cheating and it transpires they've been at it for quite a while, I'll revise my view.

Anonymous said...

Re serial cheating

Details are sketchy, but there appear to be suggestions that the French connection method was prototyped at Biel and possibly the Paris championships.

There are suggestions that a non-playing member of the trio would be present at the venue, following the game live on a laptop. Whilst this remains legal, it's also a cause of suspicion if the players are within eye-shot of the non-players, given the revelation of "spectator notation".

The Biel organisers took no direct action but it is speculated that they considered the presence of the trio undesirable and would not have given invitations and conditions for their next event.