Perfect Poker Bots

[Bunch]

I agree it would not be as good as the player that programmed it however im pretty sure there is allready some winning bots playing on party, stars , and paradise. check out pokerbot.com, it is a program that lets you enter code for how you want your bot to play. there is a forum there and people claim they have winning low limit bots allready,im sure it is a ton of work to program. has anyone ever used that before?

[TheRonin]

The problem with your logic is logic itself. The poker bot would try to use its math skills and logic to play perfect poker. Only to be devestated by people calling down a hand w/ the bottom pair or a pair of 2's when there are / 5 overs, staight and flush possibilites on the board. The bot would also realize that its not mathmatically possible to get sucked out 80% of the time with the best hand on the river. The bot would go crazy and blow up or commit botsuicide by running a tub of water and jumping in it. IMO

[sully4321]

1) there's no such thing as perfect... it's like infinity. 2) you especially can't be perfect at something with so much luck involved, because there is no "mathematical" way to determine luck. that's why it's called "luck." -- it goes against mathematical odds.
3) i don't think a bot would be a very good bluffer.
4) if it played "perfectly" it could not beat fish, because it would be on such a higher level of thinking that it would try to out-strategize somebody who wasn't strategizing in the first place.

[sully4321]

[ QUOTE ]
Agree totally, in a heads up at the moment the perfect player would win every time.

[/ QUOTE ]

perfect player would not win everytime.

ex.: if you go all in with 72o and he calls with AA -- mathematically knowing he has the best hand -- is it possible that the board could come 3456X? or 777XX? or 77XXX? or 22XXX? or 89TJX? or..? (i think you get the point). there is something called Luck in poker, so no possible way to be a stagnant winner. you can win more than you lose, but nobody is "undefeated" in poker.

[mikewvp]

I think you are all underestimating what a bot can do.

[Samsonite]

and overestimating what it would need to do. I'm not sure if any of you have spent much time playing low limit but you can pretty much expect people to call down with mid-pair. Which means that there is no reason for a bot to bluff because even if a fish has a hand that should be folded, most of the time they wont fold. A bot could still be a big winner in low limit without even factoring in the possibility of bluffing.

[tshak]

The answer is a clear "no" when it comes to perfect poker. A bot could be profitable at the lower limits, but not trivially mind you (and I am a programmer and I have researched this to a degree). It's only profitable because of how weak the players are though, not because of how strong the bot is. A strong SSH player could easily outplay these bots. The problem, however, is that strong SSH players are playing 4-8 tables at once, so their ability to outplay these bots greatly diminishes. Eventually the bots will become sophisticated enough to outplay average players, and even some pro's. But from what I've seen it will be a while, and again programming one is much easier said than done.

[tshak]

We can already simulate in the millions of games per second. A lot of the other calculations involved are subsecond on a modern CPU, even with 100,000+ hand histories. Computing power is not the limiting factor here. The limiting factor is the software.

[PhantomeX]

Can a "perfect" bot be made absolutely I wouldn't be surprised if there were already several playing in various online rooms. Would a "perfect" bot make money, very likely. In fact against bad players it would almost certainly make money. The computing power doesn't need to go up at all; a 52! is entirely possible to brute force there are networked systems today that are closing in on being able to semi brute force Igo (Go, Paduk) which is 220! but simple hueristics can knock it down to 90! unfortunatly the limit so far outside of government circles is around 62! (note that's still 10 permutations over a deck of cards) but that's not even half of it. In point of fact the brute force only has to be done once for the full 52! and then stored in a SQL database, and only the first 30(2 for each player 3 for burns 5 for board) cards of each permutation need be stored unless you are planning on setting this thing loose on a table of 12 or more (which seems very unlikely). The "bot" need only look at his opening to cards to bring the possible variations to 50! (in other words it could eliminated the VAST majority of the possible combinations at that point) then it just plays the statistics of how often it can win with its hand (including its calculations based on saved data on players on how to treat them), the programer of said bot would have to know poker theory inside and out, probably best to work with a team of several serious programer freaks and a couple of extremely competent players. Determination of play after flop would be trivally easy (but extremely important as thats where it would make all it's money). At that point it has seen 5 cards that it KNOWS what position they are in the deck this leaves it with only 45! combinations possible (and MANY of these are duplicated). The only possible way to beat it would be if it had to come in to more hands than it had mathematical certainty to win and it never got action on any bets it made on the board, and of course if that happened it could easily adjust to betting the board more often to steal the pot more often.


Now the downside, any master of the game who KNEW it was a bot would destroy it.
Many strong amature players who KNEW it was a bot would destroy it.
Any Master who played it long enough to read it would DESTROY it.

So you would need to program it to leave the table if a player that had beat it before showed up.

BTW, I am fairly certain that said bot would be VERY illegal in the US and certainly would be liable to get banned from any reputable online casino when it was discovered (which shouldn't take too long assuming that thier screens for flagging "unusual" betting patterns includes perfect play.

[PhantomeX]

[ QUOTE ]

perfect player would not win everytime.

ex.: if you go all in with 72o and he calls with AA -- mathematically knowing he has the best hand -- is it possible that the board could come 3456X? or 777XX? or 77XXX? or 22XXX? or 89TJX? or..? (i think you get the point). there is something called Luck in poker, so no possible way to be a stagnant winner. you can win more than you lose, but nobody is "undefeated" in poker.

[/ QUOTE ]

he original post was speficially postulating a limit poker situation. Anywhere that the player can control the pot odds to a nearly unlimited degree (NL and PL games) the bot would be easily beat out. And in structured limit games there would be opportunities to beat it out by manipulating pot odds but it wouldn't be enough to offset the "perfect" play of the machine.

By the way anyone who thinks that a computer cannot perform mathematically perfect play has no understanding of how computers work. Anyone who wants to seriously discuss the possibility that mathematically perfect play would be below average in terms of $ won per hand average over a statistically meaningful period I would be willing to explore that possibility my "math gut" tells me that it would be signifigantly above average considering the "typical" online players it would be facing, but I haven't actually done the functions up to say that for sure.

Regardless I can be absolutely certain that a pro that knew he was playing a bot would in short order be able to walk all over it, although if you added in some variance to "perfect play" via a hardware RNG and some serious player tracking it could probably hold it's own against even some of the very strongest players at times, it would most likely go back and forth with the strongest players pulling ahead eventually. Now where it would run into some problems is if it ran into an entire table of people who knew it was a bot and worked together to take it down (not actual collusion though that would obviously work also) in the same way that the "crazy better" that bets at every opportunity might be able to walk all over an average to weak player in headsup play (in limit hold'em) he will always loose to a table of solid players.