Poker playing computer takes on the pros

[The Buzz]

Phil Laak and Ali Eslami will be playing against the latest in computer poker technology today and tomorrow at the American Association for the Advancement of Artificial Intelligence's annual meeting.

This meeting was in Boston last year, and Casino City's snooptodd took on some of the bots and was pretty unimpressed. Apparently, there's been quite a bit of improvement in the field, and while they still expect the humans to win, it could be a lot closer than some might think ...

http://www.guardian.co.uk/gambling/s...132455,00.html

[ck8795]

See I don't belive that they could ever come near. Bots are programmed to react to particular situation. I may not know much on them but I am assuming it has to do with the numbers and equations, where as a human will always always end up with an actual thought process that will out play any bot. Maybe it might win occasional but I think over the long run t wouldn't last.



Geez....I wonder how many threads will start in the poker forums if players read this. There's a viral marketing campaign initself for all those players who think that online poker can be beaten thorugh a miraculous system.

Nice Find!

[The Buzz]

The two poker pros narrowly beat the bot after two days and four 500-hand sessions. They were actually slaughtered in the second session, when the bot beat the humans by $955 in a $10/$20 limit hold'em game.

Here's a story about the match ...

http://www.metimes.com/storyview.php...5-081854-3130r

And you can read more and see official results here ...

http://www.cs.ualberta.ca/~games/poker/man-machine/

To make the study statistically significant and to avoid card bias, in one match, the players received the same cards that the computer received in the opposing match, and the flops, turns and rivers were identical in each 500-hand session.

[The Buzz]

There's a good article about this match in today's NY Times ...

http://www.nytimes.com/2007/07/26/bu...ZhS7hoYWfrH5dw

[bb1web]

Hi all,

seems like based on the fact the computer will likely base its play on (as someone pointed out) ... the numbers .... odds ... etc.

I bet at some point once the pros get a handle on how the machine works ..... get more familiar ...... that bluffing will play a large part in being able to beat the machine.

just a matter of learning when to do so..... fooling the machine into "thinking" something is there ..... that is not.

that said .. I seldom bluff ... usually wait for good hands .... and playing sit and goes ..... hate getting into the mix before the first 2 or 3 players are gone because those are the ones that will call your raise while holding kings ... and end up catching their 2 pair of jacks and queens to put you out

so I guess I see it possible the machine could be effective.

[pgaming]

The year 1968 Chess Master David Levy no computer would be able to beat him within ten years. Ten years later he won his bet but where talking about 1978. But during this period many Chess Masters were beat.

Finally in 1989 Mr Levy was finally beat by a computer designed to play chess called Deep Thought. Followed up by Deep Blue and so forth.

Here’s the full story http://en.wikipedia.org/wiki/Compute..._versus_humans

I would have to say this scenario is totally plausible. Computers can process information at a much higher rate than humans, thus having an advantage. It will be very interesting to see the outcome of this contest.

Here's another interesting source: http://www.cs.ualberta.ca/~games/poker/

Greek39

[LiveCasinoPartners]

I think that when they advertise "Human vs. Computer" they are inferring that there's a valid comparison, which I don't believe that there is. It should be "Human software programming vs. professional poker players playing poker" lol.. It lacks the finesse doesn't it.

Human vs. Human. period. Chinook was designed to calculate all possible moves on a checker board from a predefined starting position. So from every starting position it will give you the winning move based on all possibilities. Like Google search results, everything is already calculated. So, in a sense checkers is "solved". >>http://www.aaai.org/AITopics/html/checkers.html

From the article they determined that applying the same brutt mathematical compiling is possible for a lot of games, it's just that the amount of calculation required is staggering.

I don't want to play checkers now. It's solved, and where's the fun in that?