Tuomas Sandholm is Professor at Carnegie Mellon University within the Computer Science Department with greater than 450 published papers. During the last decade, he has applied his computational game theory knowledge and optimization algorithms to develop a pc program, or “bot.”
Sandholm’s objective was to design the best heads-up no-limit Texas hold’em player on the planet and, possibly, without equal training tool for that specific type of poker.
When asked if he thought, within the future, if learning from what his program has discovered can be mandatory for all those desirous to compete on the top of the game, Sandholm didn’t mince words.
“I think so. It’s slightly a nuclear weapon for poker. You don’t need to be bringing a knife to a gun fight.”
Certainly a robust statement, but Sandholm’s success along with his program lends weight to the words. This year his bot won both categories of competition for heads-up no-limit on the Association for the Advancement of synthetic Intelligence (AAAI) Annual Computer Poker Competition, beating each opponent along the best way with statistical significance.
The idea of a ‘poker bot’ has captivated the collective consciousness of the poker world because the inception of online poker. For the dubious, the concept they might be playing poker against a machine was disconcerting. For futurists the potential for applying computer science to the sport was exciting. In actual fact that individuals has been creating bots and running them in online games for a minimum of a decade, with the primary commercially available program, called “WinHoldEm,” even being featured prominently in tech magazine Wired.
The Ny Times even wrote a feature on bots in 2011, with the belief of the object being that, for now, most human players can not fear playing against bots. One reason is since the top online poker sites have sophisticated methods of identifying robotic players and actively hunt for them.
But Prof. Sandholm also notes that while the bots in the market could be sophisticated enough to be winning players, they aren’t nearly as strong because the top bots that he has created and competed against. “They aren’t excellent. The present ones which might be available in the market now, a few of them have participated within the AAAI competition and turned out to be not very good.“
So, bots are currently out within the marketplace, but for now researchers at university computer science programs mostly develop the excellent ones. Card Player recently spoke to Sandholm at length in regards to the competition, the advance of his bot, and what’s in store for this actual program within the future.
Designing ‘Tartanian7’ To Dominate
The ninth annual Computer Poker Competition consisted of six categories, two of which featuring no-limit hold’em. Approximately 50 million hands were played. The kinds all used duplicate matches, which relating to two-player games meant that they played a regular heads-up match of a collection collection of hands between two bots, with the precise cards dealt being recorded. On the end of the desired choice of hands the bot’s memories are reset, they switch positions and play the set of hands over with the opposite player’s cards.
“The competition has two categories in heads-up no-limit Texas hold’em,” elaborated Prof. Sandholm. “One is known as ‘total bankroll,’ through which the bots play against one another and whichever bot wins essentially the most virtual money wins. The opposite category is known as ‘instant runoff,’ by which you again have all the bots play one another and throw out the person who loses essentially the most after which play another round and so forth until there's one bot remaining.”
Sandholm said these two categories measure different strengths. “The total bankroll measures mostly how well you do against the weak opponents and the elimination measures mostly how well you do against the strong opponents,” he explained.
Sandholm’s program, officially entered by him and his Ph.D students Noam Brown and Sam Ganzfried, was named “Tartanian7”. Sometimes teams make variations in their program for the 2 separate hold’em categories, however the same version of Tartanian7 won both categories and, impressively, was the most efficient against each one of the crucial opponents with statistical significance.
When asked if the success of his program had anything to do along with his own experience playing strategy games and poker Sandholm quickly noted that, “we didn't program in any human poker playing knowledge. Our approach was purely algorithmic and our algorithms created the program.”
“They key was really our superior knowledge of computational game theory and optimization algorithms,” said Sandholm. “More specifically, the entire top bots were created with the next roughly framework that my student Andrew Gilpin and that i prepare in 2005. You first run an abstraction algorithm that generates a smaller, but strategically similar, game in order that we will then use an equilibrium-finding algorithm that finds a Nash equilibrium or an approximate Nash equilibrium for the smaller game. The unique game (on this case, heads-up no-limit hold’em) is much too large to resolve for Nash equilibrium. So we solve the smaller, abstract game, after which we use a reverse-mapping algorithm to map the solution from the abstract game back to the unique game.”
Proposed by John Forbes Nash, the Nobel Prize winning mathematician who was the topic of the Hollywood movie A GORGEOUS Mind, the concept that of a Nash equilibrium essentially is an answer in a non-cooperative game wherein the player has no reason to diverge from their strategy given an opponent’s approach. In poker, it'd be a game-theoretic strategy that may work against any approach your opponent presents.
Having found what he considers to a minimum of be an approximation of the Nash equilibrium for heads-up no-limit hold’em, Sandholm entered the similar bot into both categories.
“We didn't do any opponent exploitation this year. We've got done research at the subject and feature used it in some previous years, specifically within the total bankroll competition, because there you are able to do well using opponent exploitation by making the most of the weakest opponents. But this year we didn’t do any of that. We just had an approximately game-theoretic solution and we used that to win both categories, fidgeting with a static strategy throughout. You can win both competitions with the similar program, with the similar exact program that plays superbly against both weak and powerful opponents, is something I LOCATE interesting.”
Make certain to check back on CardPlayer.com for the second one a part of this article, through which Sandholm discusses the potential for man vs. machine match with Tartanian7, how the bot may well be used as a teaching tool and more.
Read More... [Source: CardPlayer Poker News]
No comments:
Post a Comment