Using math to detect cheating

Kenneth Regan, associate professor of computer science at the University at Buffalo, has developed a program that can assist in discovering cheating. This was reported yesterday in the New York Times.

For the New York Times, Dylan Loeb McClain, spoke with Regan, who has been researching the problem for five years. The Toiletgate scandal during the Topalov-Kramnik 2006 World Championship match was a starting point. Back then, Veselin Topalov and his manager Silvio Danailov accused Vladimir Kramnik of cheating, arguing that the Russian grandmaster went to the toilet dozens of times during a game.

I thought that the chess world needed someone to try to help investigate these accusations. I felt I was qualified. I felt some definite responsibility,

Dr. Regan told the Times. Normally, he studies the epic P vs NP problem, but being an International Master himself, it's no surprise that Regan often combines his research with chess topics. 

He started to work on a way to have mathematical proof that someone is cheating. This eventually led to a model of how often the moves of players of varying ability match those of chess programs. To test someone for cheating, he runs that player’s Elo rating against the comparative model.

What Regan used for his model are games all the way back to the 19th century. By now, he’s analysed almost 200,000 games, mostly from top tournaments. His computer program will predict moves, and he can then compare these moves with the ones of the alleged cheater.

At the moment, Regan's model is at a stage where it can be used only as support in cases in which cheating is alleged. It cannot conclusively say whether a player is really cheating; all it can do is say that the move the player chose is strangely similar to those chosen by a chess engine.

The chess world has seen many cheating scandals in recent years, including the French case at the Olympiad in 2010. In February 2009, Shakhriyar Mamedyarov accused Igor Kurnosov of cheating, and based hs claims on the similarity with Rybka's moves.

Dr. Regan's research has great potential value beyond chess, according to Jonathan Schaeffer, a professor of computer science at the University of Alberta and the inventor of Chinook, the computer that solved checkers. The New York Times quotes him as saying

What he is doing, what these people are doing, is they are trying to model how people make decisions.