Computers and Chess

I've been thinking about computers and chess again recently.

There's been a little flurry of negative news about computers in chess in the past few months. Notably, ESPN's Grantland website featured a major article about cheating in chess: Rooked: The Evolution of Cheating in Chess. The article focuses on a major scandal in scholastic chess at Thomas Jefferson High School for Science and Technology, but expands to consider the broader question of how the sport of competitive chess must evolve to address computer cheating:

So folks at the USCF are now trying to determine exactly how Smiley cheated, and how often. The answers could affect not only the severity of his punishment, but the future of computers in competitive chess.

It does seem that there has been a bit of an epidemic of bad behavior in the chess world recently. Earlier this summer, the World Chess Federation handed down its decision in the computer-assisted cheating scandal in France: Three Players Suspended for Conspiring to Cheat.

The cheating occurred during the Paris Open, the Bienne Open and the Chess Olympiad, all in 2010, and was intended to benefit Sébastien Feller, a grandmaster. The other two players being punished, Arnaud Hauchard, a grandmaster and the French national team captain, and Cyril Marzolo, an international master, were part of a conspiracy to help Feller win, the federation announced in its decision on July 30.

The report from the FIDE ethics committee on the investigation is long and detailed, and in the end decisive, but it concerns itself strictly with the incident at hand and not with any of the larger issues about whether the sport should address these issues with procedural changes or other actions, though it concedes that the incident is of the highest concern:

The violation committed by Mr. Sébastien FELLER, Mr. Arnaud HAUCHARD and Mr. Cyril MARZOLO is particularly serious: it was committed during the most important FIDE competition, the Chess Olympiad, by members of a national representative, using sophisticated methods, to gain remunerative advantages.

Note that FIDE handed down the maximum penalty specified in the FIDE Code of Ethics.

But the news isn't all negative, or at least, not so severely negative. Consider, for example, the interesting revelation in Nate Silver's new The Signal and the Noise: Why So Many Predictions Fail-but Some Don't that the crucial tipping point in the Kasparov vs. Deep Blue match of 1997 may have actually involved a bug, not a brilliancy, by the computer: Did a Computer Bug Help Deep Blue Beat Kasparov?.

Either at the end of the first game or the beginning of the second, depending on who’s telling the story, the computer made a sacrifice that seemed to hint at its long-term strategy.

Kasparov and many others thought the move was too sophisticated for a computer, suggesting there had been some sort of human intervention during the game. “It was an incredibly refined move, of defending while ahead to cut out any hint of countermoves,” grandmaster Yasser Seirawan told Wired in 2001, “and it sent Garry into a tizzy.”

Note that, just 15 years ago, the discussion was still all around whether humans were intervening to assist the computers, while now everyone understands that the computers need no help from us.

But I prefer not to think about computers and chess in this way; I don't like to see it as a contest between machine and man, some sort of John-Henry-vs.-the-Steam-Engine of our time.

Rather, I prefer to continue dreaming of the glories that lie ahead, as we continue to learn to use these tools to help ourselves become better thinkers.

Consider, for example, this wonderful story about the recent performance of the U.S. team at the World Chess Olympiad: Chess Grandmasters Take Lessons From … a Windows Machine?.

Elite players often consult a database of more than 5 million games to study an opponent’s style, analyze past games and plan their moves accordingly, much like a football team will review game film. The software the U.S. squad used to prepare for the Olympiad, which concluded earlier this month, is the Houdini Pro chess engine, and you can fire it up on any Windows machine.

The team used a pretty hefty Windows machine. But it was a Windows machine. Fifteen years ago, IBM needed a supercomputer to defeat world chess champion Gary Kasparov. But nowadays, you can push the boundaries of top chess with a run-of-the-mill computer server.

To prepare for the Olympics of chess, the U.S. team called upon the High Performance Computing Center at Texas Tech University, a school known as a chess powerhouse. Such is its stature that U.S. chess team member and grandmaster Alexander Onischuk is headed there to coach the school’s team. So you can see why the school decided to lend some computational heft to the team’s training...

The result? Nakamura's epic win over Kramnik at the Olympiad will surely be on the short list for the top games of the year.

But closer to home, it is softer and subtler aspects of computers and chess that I find most pleasing. For instance, my son, who has always been fascinated by chess but has never seriously taken it up, has become a lover of the Red Hot Pawn website, where he plays game after game with other chess lovers around the world.

These, then, are the sorts of "Computers and Chess" stories I hope to read more of: people using computers to play chess against other people; people using computers to practice their chess skills; people using computers to study and enjoy the superb works of art that are chess games played at the highest level.