Rybka is clearly the top computer chess program around these days, and it comes as no surprise that the program won the 2008 World Computer Chess Championship. Rybka finished with a score of 8/9, a full point ahead of HIARCS, and two points ahead of Junior. The tournament was held in Beijing.
Rybka won 7 games and allowed just two draws with Black against Junior and the 4th-place finisher, ClusterToga. Junior actually had Rybka on the ropes, but missed a perpetual check that allowed Rybka to escape with the draw. Interestingly, both draws for the champion came in the slightly offbeat O'Kelly Variation of the Sicilian Defense.
It would have been shocking had Rybka not won the tournament, considering it not only leads every computer rating list, but was also running on the best hardware of any tournament program; I think it might be more sporting to provide each program with similar hardware as much as possible, even if this reduces the objective quality of the chess slightly. That said, it's hard to imagine that it would matter in this case.
What always fascinates me about computer chess is that even these superhuman machines still find ways to beat each other. Rybka didn't lose a game this tournament, but did have that one bad game against Junior, and has certainly been beat by other programs at other tournaments before. In any case, it's clear that even these programs aren't playing perfectly, or at least close enough to guarantee at least a draw in every game. How long will it be until we have a truly unbeatable chess machine?
Comments
I agree. It is really no surprise that that Rybka was the winner. It is entirely too strong a competitor.
As for a package that will beat all humans, I certainly hope not. I think that once we have software that can mimic human intuition and the such, we may be in very big trouble as a society.
Other than that, even if human are beat by software, I think that it will do nothing to reduce the quality of interest by chess players across the board.