University Thesis- Poker A.I.

[Flint{$N}]

Hey all,

Haven't made too many posts, but heres another.

I'm looking into creating Artificial Intellegance for limit hold em. I have just started to think about it as a thesis topic for school recently. I'm actually meeting with a professor this week to disscuss.

I'm pretty sure A.I. has been attempted before. If so, do you know who created it and if it was successful? I'm going to read up on the wilson software and see what sort of things they have done.

So a thesis on strucutred hold em AI, what do you think?

Flint

[Reef]

[ QUOTE ]


I'm looking into creating Artificial Intellegance

[/ QUOTE ]

umm. welcome to the forums.

I would do a google search, you'll find many poker programs out there.

[BluffTHIS!]

Google for University of Alberta and poker. Note that almost all such research involves headsup only as the complexities of many player scenarios are so difficult to analyze.

[Al Schoonmaker]

Darse Billings is a doctoral candidate at The University of Alberta, Canada. He is the primary developer of Poki, and it uses AI to play poker.
Darse is a very helpful person, and I'm sure he would answer your questions. An article I wrote for "Poker Digest" was titled "Can computers beat poker champtions?" It contained the following material.

"Darse Billings is a Ph.D. candidate and the primary designer of Poki, The University of Alberta’s poker playing program. He played poker professionally for a few years after doing his M.Sc. on computer poker. In an email discussion of this topic he commented: "There have been lots of naysayers in other games, but they have been forced to accept computer dominance: in checkers, Othello, backgammon, Scrabble... and the list continues to grow...
"I am not overly optimistic by nature, or quick to make a pronouncement of this kind, but I see the dominance of computers in limit poker to be inevitable — it is only a matter of time.
"Lots of people will disagree — humans are especially proud of their mastery over the dark mysteries of poker, and are dismissive of what ‘stupid’ computers are capable of. But to put it quite bluntly, they don't have a clue.”
He jokingly concluded: "I for one will not be the least bit surprised when the top poker players in the world get spanked like little babies." He put a :-) after that remark to make sure I didn’t take it too seriously. They may not get spanked like babies, but they will be beaten.
You can contact him through U of Alberta's website. I regret that I don't have his current email address.

Feel free to use my name when you contact him. He knows how much I respect his work.

Good luck with your thesis.

Regards,

Al

[gaming_mouse]

Al,

While they may get spanked at some time in the future, Poki is not even close to good enough to spanking them. Poki is quite weak, actually, despite its good performance in the IRC games, and would almost certainly lose even in the PP 2/4 game. Not to say that computer won't get a lot a better, but they have their work cut out for them.

gm

[Flint{$N}]

Dr Al,

Thanks for your input. Greatly appreciated. I will be contacting him within the next week (up to my neck in midterms).

A few people in other forums (mid stakes limit) directed me to U of A's website for POKI.

I know that I won't be able to make a definitive A.I. within a year of undergraduate study (not even purely dedicated to my thesis), but I'd like to further look into subject.

Thanks again

Flint {$N}

[Al Schoonmaker]

I agree that there is a lot of work to be done, but my guess is that they will succeed. Computers now slaughter people at several interative games, most noticably Rock, Scissor, Paper. The computer can keep track of your response tendencies and use them against you.

Unfortunately, I am not at all competent with computers. I know only what other people tell me.

Regards,

Al

[Guruman]

there has been work done lately that takes AI in a different direction than just raw human programming. Many new AI technologies can be given a fundamental set of parameters and set loose to "learn" the best way to do things on their own. Computers are far superior to humans when it comes to identifying patterns, and if you give an AI program enough repetitions against real people it can learn how to beat them over time.