Upgrade to Chess.com Premium!

Chess Mysteries - The Cheating Scandal in Zadar

Today’s blog post I have devoted to a sensitive topic in contemporary chess, one that may determine the future of the ancient game and ultimately destroy the beauty and sportsmanship that chess has been known for generation after generation. The matter to be discussed today is namely chess cheating.

After a spike of tournament cheating cases in recent years throughout Europe and around the world, we have become increasingly sensitive to the sudden displays of extreme genius over the board by otherwise consistently average players. A case of similar kind has been repeatedly reported in the media from the recently finished Zadar Open in Croatia. In the A Group of this mid-December tournament, 32 strong grandmasters and other titled players convened to measure their playing strengths and compete for the top prize of EUR 2200. Among those 16 GMs and nearly as many IMs and FMs was a single untitled player previously registered in the B group of the tournament. The Bulgarian player, Borislav Ivanov, who was rated 2227 at the time, had just decided to transfer to the A group of the tournament and try his best against the GMs. Surprisingly, however, once the tournament began, Ivanov started to win game after game beating world-class players, like the renowned GM Kurajica, GM Kožul, and GM Šaric. The tournament organizers became increasingly suspicious but despite the precautions that were undertaken, Ivanov’s innocence persisted long after the end of this event. 

Now, in my video below, I want to cast some light on the reality surrounding this case. Certainly, guilt cannot be unilaterally proven without the relevant FIDE anti-cheating measures being put in place for this type of events, but I can at least demonstrate you how easily one can be caught cheating, given the modern technologies that can both help and deter cheating attempts at our dear chess tournaments! 

The video I am presenting you with below is very instructive from a theoretical, as well as practical point of view, and can also give you some food for thought, as well as the ultimate knowledge to decide for yourself on whether this case was a case of cheating or pure genius.
 

 

Comments


  • 12 months ago

    brankz

    if you think about it, a sophisticated cheat would play more like karpov and petrosian than kasparov or tal. much harder to detect.

  • 20 months ago

    sollevy10

    Yes we can say with high degree of confidence that he used a chess engine in his games based on those statistics. However, I still think that those stats are insufficient evidence on his guilt but sufficient enough to warrant a police or criminal investigation. they like to play chess in prison and he can probably play simul exhibition games with his inmates, just in case.

  • 20 months ago

    gambit-man

    Apologies for the delay...

    I carried out a standard 3-line analysis using Houdini 3 at 25 ply. I arrived at 25 ply because of the time controls, it would be around that point that a decent modern laptop could generate the analysis in that time.

    Opening moves are excluded from the analysis, they are considered theory, so are ignored for this purpose.

    Borislav Ivanov's performance in the whole of the Zadar tournament looks like this:-

    Analysis Houdini 3 Pro x64, 1024 hash, 25 ply:-

    {Top 1 match: 222/294 (75.5%)}

    {Top 2 match: 251/294 (85.4%)}

    {Top 3 match: 268/294 (91.2%)}

     

    Similar analysis carried out on Magnus Carlsen's performance at the 75th Tata Steel:-

    {Top 1 match: 311/512 (60.7%)}

    {Top 2 match: 404/512 (78.9%)}

    {Top 3 match: 443/512 (86.5%)}.

    So he out-performed Carlsen by quite some margin, but let's take a look at Ivanov's performance in closer detail...

    In round 8 the broadcast was removed, when suspicions were raised about his performance. He couldn't have cheated in this round, and went on to lose it.

    In round 2 there was undoubtedly time-pressure issues, but without actually witnessing the game myself, i can only guess as to how much that played a part. The analysied moves from round 2, because it was such a long game, make up 36% of the total from the tournament. If we remove rounds 2 and 8 from the analysis, the figures are thus:-

    {Top 1 match: 140/165 (84.8%)}

    {Top 2 match: 152/165 (92.1%)}

    {Top 3 match: 159/165 (96.4%)}.


    Now, i know there will still be one or two of you who are still not convinced by this, i decided to take a closer look at the moves where Ivanov didn't make Houdini's first choice move.

    I found that the moves he made were often in fact the top choice for Houdini when viewed at a different ply from my original analysis, in all cases, within just a few ply.

    With rounds 2 and 8 removed (i saw no need to look at these games for this purpose), the figures are:-

    {Top 1 match: 158/165 (95.8%)}

    {Top 2 match: 159/165 (96.4%)}

    {Top 3 match: 161/165 (97.6%)}.

    Anyone who still doesn't believe he was cheating is just naive.

    My analysis, where it met the threshold, and where it hasn't since been superseded, is stored on the Fritz cloud. I have also uploaded a small database (48kb) containing all the games and stored analysis for anyone who wants to check it here:- http://www.sendspace.com/file/64cmjz

  • 22 months ago

    IM TigerLilov

    Here is the new investigation on the recent chess cheating scandals - http://www.youtube.com/watch?v=qhfCUdy2Tzk&hd=1

  • 22 months ago

    gambit-man

    @sollevy - forget match up rates with top 3 moves, the incidence of match up with Houdini 3's FIRST CHOICE move is incredible...

    Caught up with other things just now, i will publish my analysis soon.

    I have resisted watching Lilov's video until my own analysis is complete, and AIUI he makes reference to the round 2 game he lost, in a closed position where engines are notoriously weaker, and in a technical minor piece endgame where engines are also notoriously weaker. Add into the melting pot his comparitively poor performance when the games were not broadcast...

    I don't know if GM Jovanic has made any comments about their round 2 match, but i'll bet money he was already onto him and deliberately played a closed game where Ivanov's bishop pair were rendered virtually useless

  • 22 months ago

    sollevy10

    i am interested to see this statistical analysis, not only the results but also the methodologies involved. not too long ago, even those highly educated people agreed that certain ugly looking women were practicing witchcraft, i.e., it is not difficult to convince others of one's suspicion. there are a lot of players who don't lose, they always win or they were just cheated.

    why not give the benefit of the doubt that it was simply pure genius when nobody found any other evidence of cheating. an accusation has to be substantiated by more than comparing one-on-one moves with houdini. in certain games, many excellent moves are not difficult to see while some are simply basic openings and others are simply forced. the critical moves are just a few percentage of the whole game. i believe, any statistical test should only look at those critical moves. if a player's move in an OTB game is showing high similarity with houdini, he should be labeled a genius first. if we found a hardware that's helping him generate such high correlation, only then that the label should be changed to a CHEATER. In an online game however, it should be the other way around.

  • 22 months ago

    gambit-man

    Finding the hardware is not important. Finding how the moves were transmitted to him is not important. If i can use a legal analogy, it's not necessary to find a body to prove a murder has been committed.

    I have done my own extensive analysis using Houdini 3, and have a higher match rate that FM Lilov's friend... This is incontrovertible evidence.

    I fear that he will not be banned, leaving tournament organisers with the dilemma as to whether they should accept his entry. God forbid, i hope we never get to the stage where his opponents refuse to play against him, or withdraw from tournaments he has entered.

    I have made similar analysis of Magnus Carlsen's performance at the recent Tata Steel tournament, just for the purposes of comparison. I will produce my report soon...

  • 23 months ago

    Jamalov

  • 23 months ago

    LegoPirateSenior

    Here's a link to formal statistical analysis done by Prof. Kenneth Regan, who has been developing analysis methods for many years:

    http://www.cse.buffalo.edu/~regan/chess/fidelity/ACPcover-and-report.pdf

    Quoting from the conclusions: "...for a 2300 player to achieve the high computer correspondence shown in the nine tested games, the odds against are almost a million-to-one..."

    Acknowledgment: goldendog pointed me to the above article.

  • 23 months ago

    g-levenfish

    Great job! I found this analysis very interesting! Thanks.

  • 23 months ago

    johnk_rm

    Also I certainly don't think the GMs should lose rating points -- it's bad enough that they lost the prize money!

  • 23 months ago

    johnk_rm

    Yes, can he be banned from tournaments or is that not possible? I thought ogranizers should have the discretion to do that.

  • 24 months ago

    dokter_nee

    But how to prevent it, is a much harder question. 15 minutes delay on live transmission seems like an easy rule to implement. No allowing players to have any kind of electronic devices(phones, tablets) on them, unless you have a prescription. Also doesn't seem unreasonable. 

    And maybe an arbiter can also have the right to remove a member from the audience if there is any suspicion.

  • 24 months ago

    sollevy10

    @sisu, the proof provided is not that difficult to comprehend than you have imagined among us the weaker players. i do have a very strong suspicion that there really was cheating that occured. however, there is a big difference between suspicion and accusation, which is what i was trying to point out. 

  • 24 months ago

    sisu

    @sollevy10: It is not the view of any egotistical player, it is reality. Look at the stats of the players commenting in these threads.

    Is there something about the proof provided that you do not understand? Please ask. 

  • 24 months ago

    sollevy10

    @fm lilov, i did not say that "all strong players know he was cheating and many weaker (lower rated) players believe otherwise as they are not capable of understanding the magnitude of the proofs provided." such an assertion can only be made by an egotistic strong player. you may want to read the thread of comments in your article.

  • 24 months ago

    konhidras

    Maybe they should start playing in swimming trunks and  and be guarded during their visit to the CR.Wink

  • 24 months ago

    IM TigerLilov

    @sollevy10, I agree with you that the current evidence may not be enough for a legal lawsuit, but this is not the point of my two videos on that topic. What I am trying to suggest is that if the important people ignore the existance of the actual problem (cheating), there will be many players who will similarly take advantage from this sort of fraud and we will witness many of these scandals in the future. Trying to prove Mr. Ivanov's guilt is completely pointless because, as you mentioned in an earlier post - all strong players know he was cheating and many weaker (lower rated) players believe otherwise as they are not capable of understanding the magnitude of the proofs provided. All of this, however, has little to do with the main question before us: How to prevent people from using electronic assistance during chess tournaments? With the growing speed of electronics and communications nowadays, using computer assistance will become much easier and more available to anyone. This requires a general solution. I hope FIDE will soon come with a decision on a new rule that will prevent this kind of practice for the time being.

  • 24 months ago

    sollevy10

    statistical test is not without a margin of error. to conclude that there really was cheating is tantamount to saying that there is zero margin of error in the analysis. the analyst therefore is outrightly biased.

    if we argue that the margin of error need not be zero to make a conclusion on the person's guilt, what then is an acceptable margin of error? his performance must have been extra-ordinary and without precedent. does that make him guilty of cheating?

    Usain Bolt ran 9.58s in 100m track, should we say that he must have cheated because no other man had ran as fast as him? Sir Isaac Newton came up with some important laws in physics, should we say that he must be a god because no other man came up with the same brilliant ideas?

  • 24 months ago

    IM TigerLilov

    Here is a short video in which I answer some of the questions and comments raised by viewers, and it will serve as my conclusion on the cheating scandal that took place last month in Zadar, Croatia: http://www.youtube.com/watch?v=d7VvvRX-nOQ&hd=1

Back to Top

Post your reply: