Every hand has theoretical winning propability but if you bet exactly according
winning odds you tell too much about your hand,
other players fold and call and raise according your bets, only when they think
they have better cards, that is bad, obivious book play works only starter
levels.
In poker you example should not raise too much, very often opponent has for
example nuts, such cards that they can't lose, so it is stupid to but all your
money in if there is very little money in bot. If you raise too little that is
bad either as you not get full value of your good cards and turn or river can
give better cards for opponents. I don't think have you mist anything, but
_modeling opponents behaviour_, that is quite much ask from bot. So what is
best play, it depends also how you play, but how bot plays, humans are good for
observer simplified behaviour and find weak points from bot.
Other hand I think that it should be possible easily measure bots quality, even
lot faster and easier than in go game as one hand typically can be played very
fast. You could play in one 300 turn go-game 300 poker rounds, so quality of
poker bots should be easily evaluated, in 10000 rounds small differences start
to show up, but that is maybe out of topic.
So following things should fafor bot in poker:
Exact mathematic.
Exact memory is possible. In Prisoners dilemma, atleas if you remember
longer opponent moves than your opponent remembers
your moves, it does not quarantee your win but it makes it easier if you
can use your data right.
Bot don't lose temper, and don't care if it is losing or not, attleast
starter levels thats not case in humans.
Also like money poker is quite much waiting for opportunity and bot should
have time to wait.
Practical issues are imho much more demanding in go-game AI than in poker AI.
For example generating random table in go-game and poker was nice to notice how
easy it is generate random flop in poker...
t. Harri
----- Original Message -----
From: [EMAIL PROTECTED]
To: computer-go@computer-go.org
Sent: Friday, July 27, 2007 4:26 AM
Subject: Re: [computer-go] U. of Alberta bots vs. the Poker pros
I don't understand this. For a given hand the odds of winning can be easily
calculated for poker and the best play can be formulated accordingly. It's like
to program a com[uter to win a coin toss. I would be surprised if any side win
big. The only thing a computer can to is to model opponent's behavior, which
may deviate from the best play. What did I miss?
DL
------------------------------------------------------------------------------
AOL now offers free email to everyone. Find out more about what's free from
AOL at AOL.com.
------------------------------------------------------------------------------
_______________________________________________
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-go/
_______________________________________________
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-go/