Cheating scandal in Croatia – feedback and analysis

Before we present the very heartfelt feedback that has poured in, we would like to make a fairly obvious point: in our report we did not, as some readers claim, accuse the player in question of cheating, and did not do so with insufficient (circumstantial) evidence, or after he had been "cleared". What we did is report, with screen shots and links, that chess was in the mainstream media in Croatia because suspicion had been expressed by some of the top grandmasters in the region that a relative amateur had been using electronic assistance to play miles better than his nominal playing strength – and crush them as no world class player could have done. A perfunctory examination of the player had been sensationalised (in the Croatian media) as a "strip search" and the live broadcast had been interrupted for one game – which the player lost.

All of the above was in the Croatian press, where chess made unexpected headlines. This we reported, independently of our own (not yet drawn) conclusions on the matter. It is adventurous to suggest that a chess news page has a duty to suppress facts – that chess was generating sensational news stories in the mainstream media – because we deemed the conclusions to be insufficiently substantiated. And with regard to the links we provided at the end of our article – we do this all the time, as a service to often first-time readers who become interested in a subject.

Having said this, on to the feedback, which includes most of the messages we received...

Reader feedback

Pablo Pena, Santa Ana, CA
I would be wary of casting judgement too quickly. I was an expert last year when I took clear first over GM Melikset Khachiyan and IM Jack Peters (beating them both in our individual games). If you look at the wall chart it looks lopsided with a 2100 winning clear first and other places awarded to 2500s, 2400s, etc. I certainly wasn't cheating. Sometimes a player has been studying hard and gets past a certain plateua. Chess learning isn't necessarily incremental.

Rama Gitananda, Phoenix, Arizona USA
You quote: "they were wondering why he would take part in a tournament, which costs a couple of thousands of euro, while the cheating equipment, which can be integrated into contact lenses, for instance, costs thousands of Euros more." The in-ear device which was used to cheat in the 2006 World Open cost less than US $300. Did they even peer into his ear canals? I have read no indication that they did. Ref.: Rooked – The supremely old-school game of chess is dealing with a very avant-garde brand of unsportsmanlike conduct.

Richard Mallett, Eaton Bray, Dunstable, UK
The implication of the article is that Zlatko Klaric said that Ivanov was accused of cheating in Bulgaria and Serbia (presumably during the period when he had won only one rating point) and that this meant that he was cheating at Zadar!

Igor Freiberger, Porto Alegre, Brazil
I cannot understand why organizers still do realtime streaming of the games. In this case, they stoped it in the 8th round and saw the suspect lose in a pale way. So why was streaming back in the last round? These days, the public in chess events is quite sparse, while the real impact in on the web. To delay game broadcast causes no problem in terms of divulgation while it avoids all the damage cheating can cause.

Dr. John O'Connell, Dublin, Ireland
I am appalled at the decision of ChessBase to publish an article alleging that a chess player was cheating in the recent Zadar Open in Croatia. Under the principle of natural justice all people are presumed innocent until proven guilty. Your news article offers no convincing evidence that this player was cheating. Yet, your News Item has seriously damaged the good name and integrity of a young man. Nor have you given this man a forum to defend his performance. There is no justification to say he must have been cheating based on previous performances or because is moves are those of a strong chess computer. I sincerely hope that this individual takes ChessBase.com to the liable court and wins a substantial reward for the damage that you have done to him.

Luis Baquero, Medellin, Colombia
The hypothesis that there was no cheating might be rejected with a low probability of error; but why do people sudenly identify cheating with compure cheating? Have all the traditional forms of cheating disappered? Its not easier to fix a game or a tournament? Work has to be done first against fixing games; then against computer cheating.

Minh Tuan, Hochiminh City, Vietnam
Should we have done the same strip searching with Magnus Carlsen when he was young and beat so many strong GMs already. I was deeeply suspicious about Magnus's ability to play chess too, and this suspicion should be proved right or wrong by thorough strip searching and metal detection in every tournament where Magnus participates. If Magnus Carlsen is proved innocent, then my next suspect would be GM Fabiano Caruana. Please advise.

Fernando Semprun, Madrid Spain
It is all rather sad. In Cadiz 1991 I had a performance of 2433 (and was winning vs a Russian 2560, Machulsky) which would have translated into a performance of 2533. Then there were no cases or suspicion. And I remember a surprising performance of Sigurjonsson at some tournament when he was clearly the underdog (late 70's? ), and of course many others. So these results, on their own, should not be proof of cheating. But it will kill chess if they become widespread. So sad and annoying.

Rajko Vujatovic, London
These allegations are clearly nonsense and it is most unfortunate that Ivanov's name has been dragged into the public domain despite the lack of evidence. There are several thousand players with Elo rating 2100-2300. Occasionally, it would be statistically expected that one of these players will have enough lucky breaks to achieve a performance 400 points above their rating. These 'breaks' could take the form of poor play by the opponents that are easy to refute with natural moves; or hard work and good opening preparation that helps to nullify the advantage of the grandmaster opponent. Playing through Ivanov's games without any silicon companion, his play struck me as being entirely human, and could be entirely justified with basic principles and motifs. He made obvious attacking moves where the position often played itself. In some games, his grandmaster opponents made incorrect sacrifices where Ivanov's response was natural, regardless of whether he understood at the time if his position was better. In addition, aspersions are unfairly cast upon Ivanov because he is a programmer – but this is not an unusual profession for a chess player!

Nahim Zahur, Singapore
Does ChessBase have any actual evidence that Mr. Ivanov was cheating? If not, what brilliant editorial judgement led ChessBase to write the article in this manner and tone? Yes, the article does not actually go so far as to state that Mr. Ivanov is cheating (that was a clever stroke). But it quotes verbatim three other articles that have clearly painted Mr. Ivanov as guilty. "One of the tournament participants", "knowledgeable sources" and the disgruntled Mr. Klaric all get to have their say in this piece, but not the accused himself!

Now let's come to what ChessBase themselves added. The 2227 rated player is erroneously referred to as "the unrated Bulgarian player," no doubt to throw further incredulity on his results. [The error was corrected to "untitled" shortly after publication – Ed.] Then, instead of letting the proper authorities analyze Mr. Ivanov's games in a fair, judicial process, ChessBase would like to have a public trial of Mr. Ivanov, complete no doubt with a baying crowd and flagellation as the punishment. Finally, the article ends with the sly comment "Oh dear, we are going to have to complete our History of Chess article series..." and then posts a series of articles on the history of cheating in chess. Shattering any remaining illusions as to what ChessBase really believes. I suspect that the key to all this is that Mr. Ivanov is Bulgarian, and ChessBase has held a grudge against Bulgarian players ever since the Kramnik-Topalov fracas. This is an unsubstantiated allegation on my part, and quite possibly utterly unfounded and frivolous. Nonetheless, by the journalistic standards displayed by ChessBase, I do not feel in the least embarrassed to blare out my opinion from the top of my rooftop.

Euclides, Elmwood Park USA
It doesn't really take much to see he was cheating. Haven't you noticed as black he played the most varied and theory loaded variations. There were no offbeat variations at all. Grünfeld, Benoni, King's Indian etc etc. He played the very heart of all variations. All his games looked like world championship preparation. And by the way his moves weren't Fritzy. My computer says it looked more like Hiarcs.

Antonio Gillot, Guatemala
In 1909 Capablanca swept Marshall off the board, with no previous qualifications. In 1911 he won first in San Sebastián. Fortunately at that time envy and stupidity hadn't taken over the chess world. Or he was also a cheater?

Julian Wan, Ann Arbor, USA
Unfortunately this new cheating allegation scandal just reinforces what should be clear – that there is something going on. Physical sports such as track and field (athletics), American baseball, and most of all cycling has long been dogged by accusations and true scandals. Sadly in the end, nearly all proved to be correct. So I fear I have to be skeptical of any adult who makes a late surge in strength. What is next? Mandated game delay in broadcast – how about a reasonable delay like 30 minutes. People following on the Internet won't really miss much. Of course the games will then have to be played with only a screened set of spectators, and all of the toilets have to be chaperoned. What a sad state of chess!

William Shea, Honolulu, Hawaii
I don't see how you can even publish this article. There is zero evidence. A master level player has one tournament where he plays a few hundred points above his rating. Statistically, this is quite possible, I would think. He has two losses in the tournament, including round two, before any changes in broadcasting. Am I wrong to think this borders on irresponsible journalism even to publish such a non-story? A reader cant help doubting the players credibility after your story, until you go back and think – what evidence was there again? No computer, no headphones. No transmitting device. Just suspicions? Come on.

Guy Haworth, Reading, UK
It is clear that game-scores alone cannot tell to what extents the winner played well and the loser played poorly. Benchmarking the 'quality of the moves' is required, though this 'quality' will be an opponent-neutral or 'absolute' quality whereas the winner may have decided to choose moves which most discomfit their opponent. The benchmark is most likely to be one or more chess-engines, their evaluations being taken from one or more search-depths.

Measuring %-coincidence with an chess-engine is the crudest way to do this. No account is taken of the feasibility of moves not chosen by the engine. Further, 'measurement of coincidence' usually involves using only one engine at one search-depth - though better options are possible. Measuring Average Error (a method used by Guid/Bratko et al) is better than coincidence-measurement but still ignores the availability of alternative credible moves. Comparing player-choices with those of stochastic agents (found by Ken Regan and I with a 'best fit' technique, and by G di Fatta and I with a 'most likely', Bayesian Inference technique) brings in consideration of the 'best m' moves available.

However a 'Performance Rating' is inferred, the confidence interval around it can only be halved if the amount of data available is four times greater. For more information on publications, please see:

• Understanding distributions of chess performances (2012)
• Intrinsic chess rating (2011)
• Performance and prediction: Bayesian modelling of fallible choice in chess (2010)
• Skill Rating by Bayesian Inference (2009)
• Gentlemen, stop your engines! (2007)

May I also note that, in the recent London 2012 Olympics, outstanding winning performances were recorded by young, 'first time', swimmers hitherto unrated.

Nate Plapp, Lemon Grove
What is ugly is continued suspicion and defamation after being searched and nothing is found. What is the point of a search, if it does not vindicate when nothing is found? The quality of a game, the score at a tournament, sure suspicion is understandable, though I feel a misattribution of one's loss, unless there is some odd behavior involved, but when found innocent, that should be it – in fact there should be a public apology for the thought-to-be-needed search.

"Implants?" Someone has a screw loose. Clearly, an artifact of a person convinced of the guilt of the 25-year-old and willing to invent any delusion to hold to that position. Good performances can be by chance. It is just statistics. Why it is so important to spoil someone's one success? If one is willing to make an accusation they must also be willing to apologize to that person when nothing is found and drop the conspiracy nonsense. They must have lax defamation laws in Croatia.

Imsn Khandaker, Watford
Since he was not caught cheating, it is a little surprising that the story gets such prominence. Perhaps it should be a general rule that all tournaments have a delay in the live broadcast of moves.

Kees v.d. Doel, Vancouver, Canada
Every single move, without exception, in the third and fifth games by White is identical to those of Houdini 2. How obvious can you be?

Lilov analysis of the Ivanov games

The following Youtube video was produced by Valeri Lilov, a very strong Bulgarian FM and chess trainer, rated 2433. Lilov has recorded a number of training DVDs for us (as well as other chess companies). However: we in no way commissioned his investigation, and in fact only heard of it through a message by a Mexican reader who sent us the link.

The video is over an hour long, but quite gripping. It would be interesting to know what other strong GMs think after they have watched it. They are the real experts, they are the people who can voice a qualified and well-founded opinion. Do they agree with Lilov that the evidence is compelling and that the way in which some of the greatest player of the region were crushed defies belief? Or do they think, as some of our amateur readers seem to believe, that such flashes of chess strength and genius do sometimes occur and that it is entirely possible for an untitled 2200 player to on occasion reach such heights.

If you want to follow Lilov's analysis attentively you should maximize the YouTube broadcast:

Click on the tool icon and select HD (if you have HD resolution) and "Full screen" on the bottom right.

After this you will be able to follow everything quite clearly – and can pause the video when things get too fast.

There is one point we would like to make: Lilov is a strong player with deep understanding of human and computer strategy, but he is not a technician or spyware specialist. The "glasses" he speaks of, costing less than $50, and other inexpensive devices he shows on the web site, mostly record images and video onto an micro SD card. There actually are stealth glasses that transmit video in real time (we will not give links here!), but they cost $150. However: such glasses are not required when games are being broadcast live on the Internet. A helper gets the moves in real time by logging in to the official web site, or going to Playchess or one of the other chess servers. And finally Lilov seems to believe that in order to receive external assistance a player needs to have audio contact. Unfortunately that is not the case: there are a variety of ways to communicate the minute amount of information required for a chess move in form of a Morse code – as was shown in a well-known German cheating scandal back in 1999.

As we said in our previous article: we are going to have to publish the final section of our History of Chess article series (see below), which is long overdue...

Previous article on the Zadar Open scandal

ChessBase series on cheating

A history of cheating in chess (1) 29.09.2011 – Hardly a month goes by without some report of cheating in international chess tournaments. The problem has become acute, but it is not new. In 2001 Frederic Friedel contributed a paper to the book "Advances in Computer Chess 9". It traces the many forms of illicit manipulations in chess and, a decade later, appears disconcertingly topical and up-to-date. We reproduce the paper in five parts.

A history of cheating in chess (2) 04.10.2011 – Coaching players during the game is probably the most widespread form of cheating (rivaled only perhaps by bribery and the throwing of games). Although this practice began long before the advent of chess playing machines, computers have added a new and dramatic dimension to this method of cheating in chess. You will never guess: who were the pioneers of cheating with computers?

A history of cheating in chess (3) 18.12.2011 – In January 1999 the main topic of conversation amongst top players like Kasparov, Anand and others: who was the mysterious German chess amateur, rated below 2000, who had won a strong Open ahead of GMs and IMs, with wonderfully courageous attacking chess and a 2630 performance? How had he done it? Turns out it was with unconventional methods, as subsequent investigation uncovered.

A history of cheating in chess (4) 28.02.2012 – Las Palmas 1996: Garry Kasparov is agonizing over his 20th move against Vishy Anand. He calculates and calculates but cannot make a very tempting pawn push work. Immediately after the game he discovers, from his helpers, that it would have won the ultimately drawn position. The point that became clear to him: a single bit of information, given at the top level in chess, can decide a game.

New in Chess debate

Anti-cheating: the fifteen minute broadcast delay 13.05.2011 – For five years we have been trying to get FIDE to implement a 15-minute delay in the Internet broadcast of important games – to make organised cheating harder. A chess journalist has now pointed out a fatal flaw in the plan: it would force chess journalists to walk many yards to find out the current status of the games. Damn – and we thought it was such a good idea! What is your opinion?

Anti-cheating: the fifteen minute debate continues 29.06.2011 – Our recent reply to stern criticism leveled against us in the Dutch magazine New in Chess resulted, unsurprisingly, in a large number of letters from our readers, many quite effusive. But we decided not to publish any until at least one turned up supporting the views of our NiC critic. Six weeks went by until it at last came, authored by the critic himself. Now we can publish your letters.

Reports on the French cheating scandal

Feller's interview, and a solution to the cheating scandal 23.08.2011 – The French Championship is in its eighth round, with four GMs in the joint lead. The event is marred by continued suspicion and anti-cheating measures, brought on by accusations that one of the participants had in the past engaged in organised cheating. Sébastien Feller has given an interview on the subject, and we have a proposal on how to clear up the matter quickly.

Cheating scandal: Opinions, concerns and revelations 06.04.2011 – In a series of interviews, Robert Fontaine from Europe Echecs, culled the opinions of the players, to get a clearer idea on how players both French and foreign viewed the cheating scandal. A lengthy interview with Jean-Claude Moingt, the president of the French federation, revealed not only the next steps to be taken, but also that confessions were not only made to the players. An eye-opener.

French Chess Federation suspends players accused of cheating 21.03.2011 – On Saturday the Disciplinary Committee of the French Chess Federation suspended GMs Sébastien Feller, Arnaud Hauchard and IM Cyril Marzolo, finding them “guilty of a violation of sporting ethics” for allegedly cheating during the Chess Olympiad 2010 in Khanty-Mansiysk. The three received suspensions, after evidence was presented, including a detailed description of how it was done.

FFE accuses its own players of cheating 22.01.2011 – Shocking news: the French Chess Federation (FFE) has announced that it has initiated disciplinary action against three players – one of them one of France's most promising talents – on suspicion of "organized cheating, serious breach of sport ethics, undermining the image of the national Olympic team in Khanty-Mansyik". We are following the investigation. Press release.

Feller replies: 'I completely deny the cheating accusations' 24.01.2011 – Two days ago the French Chess Federation announced the investigation of three French players on suspicion of "organized cheating" at the Chess Olympiad in Khanty-Mansyik. Now one of the three, 19-year-old French GM Sebastien Feller, has replied emphatically, implying that the entire action was a result of his support of the current FIDE president (the FFE supported his rival Karpov). Open letter.

Battesti: 'It's insulting to our president and his federation' 24.01.2011 – Instead of adopting an ostrich position the President of the French Chess Federation and his VP have initiated an investigation of French Olympiad members suspected of cheating. They have appointed Leo Battesti, a Sorbonne-educated lawyer, as the spokesperson for the Federation. Battesti has reacted to the criticism of one of the accused player with an interview in Europe Echecs.

French GMs: ''We express our full support of the FFE 27.01.2011 – Four grandmasters Maxime Vachier-Lagrave, Laurent Fressinet, Vladislav Tkachiev and Romain Edouard have expressed their dismay at the charges brought against three of their colleagues who are accused of cheating. "If the allegations are found to be true, we will condemn them firmly," they write, in this public statement in Europe Echecs.

FFE: cheating not the first time, Biel statement 01.02.2011– The French Chess Federation disclosed they had evidence that the "organized cheating" accusation, which has rocked the chess world recently, is in fact not the first time. They have now mandated the Federal Bureau to take the case to trial in a court of law. Meanwhile the Organisers in Biel have issued a statement on the same players earlier last year in their Master Group. Open letters.

FFE Cheating: Judge rules incriminating SMS inadmissible 11.03.2011– After unearthing a series of SMS messages between players accused of cheating at the Olympiad using a phone lent by the French Chess Federation vice president, the FFE sought to have those messages transcribed and included as evidence in the upcoming Disciplinary Committee. A judge ruled that secrecy could only be waived if the FFE sued in court, as the FFE explains in a public statement.

French Chess Federation suspends players accused of cheating 21.03.2011 – On Saturday the Disciplinary Committee of the French Chess Federation suspended GMs Sebastien Feller, Arnaud Hauchard and IM Cyril Marzolo, finding them “guilty of a violation of sporting ethics” for allegedly cheating during the Chess Olympiad 2010 in Khanty-Mansiysk. The three received suspensions, after evidence was presented, including a detailed description of how it was done.

Cheating in chess: the problem won't go away 30.03.2011 – As you know the recent suspicion of organized cheating during a Chess Olympiad has led to three French players being suspended. One is currently playing in the European Individual Championship, where his colleagues have published an open letter demanding additional security. For years we have been proposing a remedy for this very serious problem. It needs to be implemented now.

Sebastien Feller wins Paris Championship 13.07.2010 – The Paris Championship is the oldest French Open – Abraham Baratz won the first edition in 1925. The leading players in this year's event included GM Tigran Gharamian (2650), GM Alberto David (2622) and GM Sébastien Feller (2611). The Open Tournament took place from July 3rd to 11th, 2010. It was a nine-round Swiss, FIDE rules, 40 moves/1h 30 + 30 sec then 30 mn + 30 sec. Pictorial report.

Copyright ChessBase