Previous |
Next A Senator Prepares to Push Colleges to Use Antipiracy Technology |
July 20, 2007, 03:44 PM ET
A Game That No One Can Win
Computer scientists at the University of Alberta have proved what Marion Tinsley probably realized half a century ago: In a perfectly played game of checkers, nobody wins.
As many as 200 of the university’s computers have spent parts of the last 13 years poring over the game’s 500 quintillion (that’s 500 billion billion) possible board positions. And they’ve now concluded that there are no surefire winning moves in checkers, just mistakes that lead to losses.
Alberta researchers have also worked to teach computers to play poker, and they’ve earned accolades for those efforts. But poker requires decisions to be made based on incomplete information, so skilled human players still have a significant edge on their computerized counterparts. Not so with checkers: “I don’t think a human would have a chance against a computer now,” said Jonathan Schaeffer, Alberta’s chairman of computer science, to the Los Angeles Times. —Brock Read
Categories: Research
Add Your Comment
Commenting is closed.