If you've played poker online for any time at all, you've certainly heard of the infamous "bot" players. Bots can come in any number of variations, but the bottom line is that they’re computer driven player programs, almost always based on adaptive and/or Artificial Intelligence (AI).

Bots can offer some pretty impressive play, but as of yet they've never been able to fool or beat a pro-level player. One of their main limitations is that they can't bluff.

"Computers are programmed to perform the best strategy, but bluffing is based on unexpected, illogical actions," says Evan Hurwitz, a computer scientist at the University of the Witwatersrand in South Africa.

Enter Hurwitz's research partner Tshilidzi Marwala and the artificial intelligence bot they've developed named Aiden. Aiden is based on a neural network algorithm that usually forecasts stock market fluctuations. That's all very interesting, but the punchline is that Aiden has taught itself some very interesting behaviour while playing a card game called lerpa.

To make a long story short, the researchers played Aiden against three other similarly trained bots to see what would happen. "They began to develop their own personalities—either aggressive or conservative—depending on their past successes," Hurwitz says.

After a length sequence of bad hands, during which it consistently folded, one of the more aggressive bot players suddenly changed tactics: it began to play even when it had poor cards, in other words it had learned to bluff.

"This demonstrates that computers can learn this peculiarly human behaviour," says 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."

Now for part two of our story: after a long run of unsuccessful play, a Full Tilt player named "SukitTrebek" decided to use Poker Tracker—a computer program that provides statistical analysis of Internet poker play—to analyze several player accounts which he suspected of being bots.

The Tracker results revealed that the player accounts in question had virtually identical playing patterns at the Hold'em tables. Such patterns are one of the key tools used to locate and identify automated game play, be it poker or any other turn-based game.

Page 1, 2