In this ongoing series of news coverage at JazzyBets, we’ll explore the Rise of the PokerBots, and their increasing use in online internet poker rooms.
Before we delve into the complexities of the Pokerbot, let’s look at one of the most famous studies of the power of PokerBots produced by the Department of Computing Science, University of Alberta. It was called “Approximating Game-Theoretic Optimal Strategies for Full-scale Poker“. Here are some interesting quotes from the study, that should prime your knowledge of poker bot logic.
Their objective was to “to create substantially improved poker-playing programs, able to defeat strong human players and be competitive against world-class opponents“, which is quite the feat.
“More than 100 participants volunteered to play against the pseudo-optimal players including many experienced players, a few masters, and one world-class player.”
Regarding poker bot logic “we call their solutions pseudo-optimal, to emphasize that there is no guarantee that the resulting approximations will be accurate, or even reasonable. Some will be low-risk propositions, while others will require empirical testing to determine if they have merit”
“… A number of interesting observations were made over the course of these games. It was obvious that most people had a lot of difficulty learning and adjusting to the computer’s style of play. In poker, knowing the basic approach of the opponent is essential, since it will dictate how to properly handle many situations that arise. Some players wrongly attributed intelligence where none was present. After losing a 1000-game match, one experienced player commented “the bot has me figured out now”, indicating that its opponent model was accurate, when in fact the pseudo-optimal player is oblivious, and does no modeling at all.”
When testing the pseudo-optimal poker bot against a world-class player,
Here’s the money quote: “You have a very strong program. Once you add opponent modeling to it, it will kill everyone.”
