This online poker bot can bluff !

Poker playing computer programs, commonly called bots, are about to sport a potent new Artificial Intelligence capability that may cause frissons of alarm in the online poker industry. Hitherto such programs have enjoyed some success in a statistical calculation and conventional playing sense, but a major drawback has been an inability to combine a key art of the game - bluffing - with statistical fact.

Enter Evan Hurwitz, a computer scientist at the University of the Witwatersrand in Johannesburg, South Africa, who may just have added that essential bluffing element.

"Computers are programmed to perform the best strategy, but bluffing is based on unexpected, illogical actions," says Hurwitz, who heads a team of scientific developers working on the concept.

Based on a neural network algorithm typically used to predict the stock market, Hurwitz's bots weren't pre-programmed with the rules of a card game. Instead, they were pitted against each other and learned to play by inferring the game's rules from their own hands, those of their opponents and the outcome of the games.

Eventually, one of the bots with the unlikely ID of "Randy" suddenly started to bluff, having calculated that it increased its chances of winning against its computer opponents.

"This demonstrates that computers can learn this peculiarly human behaviour," confirms Philippe de Wilde, a computer scientist at Heriot-Watt University in Edinburgh, UK. "They generate the strategy from play, which is a very human way of learning."

Having triggered the development of what could be a formidable poker playing bot, Hurwitz and his team have not yet announced where the project will head from here.