Thursday, June 30, 2011

Rybka banned from computer champs; stripped of all its titles



In shocking chess news-- Rybka, the famous computer program that has dominated the top of the computer chess championships for years, has been stripped of its titles and banned from future competitions.


The problem stems from the complex question of whether Rybka's programming was taken from other, older software-- in short, whether Ryba's basic functions were stolen off other programs.


On the one hand, the fact that Rybka beats other programs means it must be at least a little different than other engines. On the other hand, this could be just a matter of a few tweaks here and there without any substantially new ideas.



The whole issue of the 'copyright' and 'ownership' of source code is a murky one -- very often, free programs already exist whose main purpose is to be modified, tinkered, and enhanced. Stripping Rybka of the titles, at this point, does not appear to be any kind of legal determination about whether Rybka was stolen from others. Instead, Rybka is in violation of the Computer Championship rules, which requires entrants to be unique creations and any adaptations of other's code must be disclosed.


I don't expect this to go away anytime soon, however! Rybka is a commerical success, while the programs it supposedly is borrowed from, Crafty and Fruit, among others, are not. Like I said, the world of computer program property rights is a murky one, but I wouldn't be surprised, with real dollars on the line, for this issue to make it to court in some form.


At any rate, Rybka isn't considered the strongest program anymore. After trying to find a good upgrade engine to my older copy of Fritz, I discovered Houdini may be the best engine around. A free computer program, I put it to work on some positions I thought Fritz was handling badly-- and I was blown away with the speed and the accuracy Houdini possessed.





But don't expect Houdini to appear at the computer championships anytime soon-- it is also considered a derivative work. Too bad that the strongest computer program may not be called the champion. But that doesn't mean we can't use it to improve our analysis.


One last thought-- my favorite comment about this issue:


"Humans have been using computers to cheat at chess, so it is only natural that computers would start using humans to cheat at chess."


For more detailed discussions about this issue, try this link:



2 comments:

Ken Ho said...

An exchange from the classic movie Capablanca:

"I am shocked, shocked, to learn there is cheating going on in this computer chess competition."

"Your Rybka analysis, sir."

"Oh, thank you, just put it in my smartphone."

CTheil said...

I find it particularly interesting that the mainstream chess websites, like ChessBase and the US Chess Federation, did NOT report on this story.

They ignored it at the same time it was being covered by mainstream media, like the New York Times.

On the day the story broke, ChessBase published two articles, including (ironically) one about cheating.

In the days that followed, ChessBase published articles on VERY important subjects, like Chess on Stilts in Belize and the Turkish Isbank Chess League.

It makes you wonder how much ChessBase makes from the sale of Rybka ...