NEW HAVEN, CONNECTICUT – The Christmas season may seem an odd time to bring up the subject of cheating. But it’s also college exam season — and besides, some stories are impossible to resist.
So: Last week a chess player named Boris Ivanov was booted from a tournament at Navalmoral de la Mata, in western Spain, on suspicion of cheating. This is hardly newsworthy; players do get kicked out from time to time. Ivanov is a special case: For the past year, the young Bulgarian has been leading European chess authorities a merry chase, winning games his rating predicts he should lose, consistently playing moves that match the “top line” of the leading computer programs, and defying every effort to catch him breaking the rules.
Professional players around the world have come to hate him, but many amateurs seem to regard him as a folk hero. Peruse the comment threads on any chess website where his name is mentioned, and you’ll find plenty of ordinary duffers praising his accomplishments and dismissing the suspicions expressed by some of the best in the world as mere sour grapes.
Is Ivanov cheating? Chess ratings have proved over time to be excellent predictors of results in actual games. On the other hand, it is hardly unheard of for a young player who is improving rapidly to meet with swift and unexpected success. Still, the case against Ivanov is strong. Computer scientist Kenneth Regan of the University of Buffalo, perhaps the world’s leading authority on the use of software to cheat, certainly seems convinced.
Once upon a time, chess was thought to be cheat-proof. In 1929, the playwright Percival Wilde published a short story in which a very weak player cheated by having a top master carve moves into slippery elm tablets, which he would read and then consume, destroying the evidence. But the story was all in fun.
In the 1960s, when the television show “Mission: Impossible” broadcast an episode in which Rollin Hand — played by Martin Landau — played a crucial chess game while receiving computer prompts through a fake hearing aid, serious chess players scoffed: The machines weren’t good enough to beat the best.
Those days are long gone. Now anybody with $20 to spend can download software far more powerful than the strongest players in the world. As the economist Tyler Cowen has pointed out, cheating in chess grows more attractive as the cognitive capacities of machines and humans diverge. That’s why he suggests that the future belongs to “freestyle” chess, where human and computer collaborate.
Maybe so. In the meantime, however, the search for cheats goes on. Internet chess sites make enormous efforts to screen them out — with only partial success. At the recent match for the world chess championship in Chennai, India, the players were scanned for hidden digital devices by a handheld metal detector similar to those used at airports.
Which brings us back to Ivanov. Digital devices of all kinds are banned at tournaments these days, but his opponents say he’s been somehow sneaking one in. In the U.S., where chess remains something of a backwater, this might not be a profitable activity. In Europe, where thousands of international grand masters crowd and swarm, there are big-money tournaments everywhere. A couple of years ago, three French players were banned after one was silly enough to send a text message that read “Hurry up and send me some moves.”
Nobody has quite caught Ivanov red-handed — though not for lack of trying. He has been strip-searched. He has agreed to a rigorously monitored test to see whether he is cheating, then failed to attend. He has announced that he will not play again. He has returned, only to face a boycott by leading grand masters. One top player has accused Ivanov of hiding a digital device in his footwear, and receiving messages via vibrations. (Ivanov, challenged, refused to remove his shoes.)
Last week, boycott or no, Ivanov showed up at the aforementioned tournament in Spain. Once again, he mowed down a list of higher-ranked opponents. Then this happened:
“Ivanov was expelled from the tournament after round 6. His opponent of that round, Guliyev, had asked the arbiter to search Ivanov, and indeed they took the alleged cheater apart. Ivanov was forced to remove his clothes. Two participants of the tournament (who wanted to stay anonymous) told ChessVibes that an electronic device was found, and the arbiters gave him the choice of playing on but showing them the device, or leaving the tournament. Ivanov decided to leave the tournament and to the surprise of the participants, his €40 entry fee was returned, and Ivanov even received an extra €50 for ‘compensation.’ “
Now, I don’t know whether Ivanov is cheating or not, and one should never rely on statistical evidence alone to conclude that a player has cheated. True, it is highly unlikely that a player of any strength will consistently play the best or nearly best move recommended by the strongest computer programs. Yet it is also highly unlikely, as Regan points out, that a weak golfer will make a hole in one. Still, every now and then, one does. Thus, as Regan says, the conclusion that a player has cheated should rest on a combination of statistical and observational evidence.
It’s also important to remember, as economists keep reminding us, that the optimal level of cheating isn’t zero. Sometimes the costs of monitoring can outweigh the benefits, and the value of the underlying activity is reduced. Consider how bored many football fans become when the referee’s decision on an instant-replay review stretches out interminably: At some point, we’re better off letting the error stand.
Or ask yourself how long parents would put up with a regime in which their children were strip-searched for digital devices before being allowed to sit for standardized tests. We might reduce the rate of cheating, but at too great a cost.
Which isn’t to say there isn’t a problem. The U.S. has seen many cases of students caught cheating on important standardized tests — particularly the Law School Admission Test. Now suppose Ivanov’s opponents are correct. Imagine, as one suggested, a camera hidden in a contact lens. A student using this method to cheat might send images of the tough questions to an outside confederate, who would then transmit answers, in a pre-arranged code, via the vibrations of a concealed device.
So far there’s no evidence that test cheaters have become quite that sophisticated. But in an era when the returns to attending the right school are so enormous, sooner or later somebody is bound to try. When young people are raised to let machines do their thinking for them, we should hardly be surprised when they resist rules aimed at making them stop.
None of this should be taken as a defense of cheating by either test takers or chess players. To understand isn’t to excuse. Nevertheless, as technology leaps ahead, we should expect more cheating, not less. The time to decide what to do about it is before we set up the metal detectors.
Stephen L. Carter is a Bloomberg View columnist and a professor of law at Yale University. He is the author of “The Violence of Peace: America’s Wars in the Age of Obama” and the novel “The Impeachment of Abraham Lincoln.” Follow him on Twitter at StepCarter.