The contest in Canada between two brilliant poker pros and a computer will be held in July 2007 in Vancouver; British Columbia, This contest will be all about how a computer plays against real players and can it vary its game from conservative to aggressive. If the two pros win both of their two games they will get $25,000 for their efforts. The computer named Polaris by its University of Alberta developers was designed with the idea that it would be able to play to the opponents' weakness. Also it would seamlessly mover from aggressive play to conservative. The computer will underbet hands and also bluff at different times. Unpredictable play is a part of a good poker players gaming skill. This is what the developers of this computer program are counting on said Jonathan Schaeffer one of the lead artificial intelligence researchers on this Polaris Poker Project. We can predict the odds without any problem, but we cannot play to conservatively or the opponent will soon take advantage of this way of playing. We need to be unpredictable to beat a good human player. This is why we built in playing conservatively and then aggressive bluffing.
This high-stakes tournament will be part of the Association for the Advancement of Artificial Intelligence conference that is scheduled for the Vancouver hotel. Attendees to this conference will be able to watch Phil Laak and Ali Eslami take on Polaris with the opportunity to win $25,000 each if they win both of the games they are scheduled to play against Polaris. Each player must win two games each against Polaris in order to win their share of the prize money. Laak recently won NBC's Poker After Dark and Eslami plays in some of the most elite poker games in the world. Computers have done exceedingly well in strategic games such as chess or checkers. But poker has the luck element in it and that may be tougher for Polaris in these matches. There is no luck element in chess or checkers. On the other hand if the style of play by the computer, Polaris, can be changed without any tells, then its unpredictability may be difficult to master.
I am not sure who will win this poker match between solid poker players and a computer due to it is so hard to detect the play of an opponent. Oliver Schulte, a cognitive scientist once worked with the Polaris developers. He would not venture a guess about who will win.
But, what if this computer is amazingly good at playing poker and wins this test. If it can emulate the play of a skilled poker pro and play far more superior than the dogs in the famous poker painting, what chance would average online players have against a computer that plays exceptionally well and can also pull off a solid bluff. And, how soon would some poker rooms install these computer bandits to remove the playing stake from their less than knowledgeable opponents in a friendly $2 No Limit game. Would this be akin to the room cheating to set live players up against such opponents.