Is there a perfect NLHE Bot? Could there be?
 

I was reading over a post that answered a question about the toughest game and came across some discussion about a perfect poker Bot.

It seemed worth talking about so I posted my opinion here.

I love to hear people talk about the perfect Bot and how it could beat any Hold-Em player by playing defensively.

For those of you who are on the border - don't believe it.

First of all Hold-em is not only a game of skill, but also a game of luck. If what these poeple are claiming is true, that there is a perfect bot that will always win, and that players like Phil Ivey nearly mimic the bot, then Phil should ALMOST NEVER lose!

I give Phil a lot of credit, he is absolutely one of the top five players in the world at what he does (in my opinion), but it NOT beacuase of his perfect defensive play (although he is good at defense), it is because of his varied aggressive attack.

The perfect defensive bot would probably lose to the majority of the players on this forum, amateurs included. If the bot holds KQ and there is an ace on the board, the bot will ALWAYS fold to a big bet. Why? Because it is the right thing to do!

Give me that player anytime!

This talk about a bot reminds of of all of those sci-fi movies of the seventies that showed intelligent computers taking over the world. Computers and bots alike only do what they are told. That is all they can do. That is all they will ever do.

BRING ON THE BOTS!

Re: Is there a perfect NLHE Bot? Could there be?
 

The day computers get better at understanding human psychology than humans are, a bot better than any humans will be made. This will happen some time, but I think it is relatively long into the future. A perfect bot will never be made as no computer will ever understand the total complexity of human psychology (ok, never is an exaggeration, maybe it will be made in 3,000 years time [img]/images/graemlins/tongue.gif[/img]).

Re: Is there a perfect NLHE Bot? Could there be?
 

The bot is not better than any human in that it does not capitalize on mistake to win the maximum. It will merely never lose a HU Freezeout.

Re: Is there a perfect NLHE Bot? Could there be?
 

[ QUOTE ]
The bot is not better than any human in that it does not capitalize on mistake to win the maximum. It will merely never lose a HU Freezeout.

[/ QUOTE ]

I can't see any sense in this?

Re: Is there a perfect NLHE Bot? Could there be?
 

[ QUOTE ]
[ QUOTE ]
The bot is not better than any human in that it does not capitalize on mistake to win the maximum. It will merely never lose a HU Freezeout.

[/ QUOTE ]

I can't see any sense in this?

[/ QUOTE ]

He's assuming the only thing a bot is capable of is playing perfect game-theoretic poker which does not always maximize EV.

I see no reason a bot could not adapt to it's opponents just as a human can

Re: Is there a perfect NLHE Bot? Could there be?
 

If the bot has the ability to adapt you an individual's playing styles, then it could absolutely destroy amateurs.

Re: Is there a perfect NLHE Bot? Could there be?
 

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
The bot is not better than any human in that it does not capitalize on mistake to win the maximum. It will merely never lose a HU Freezeout.

[/ QUOTE ]

I can't see any sense in this?

[/ QUOTE ]

He's assuming the only thing a bot is capable of is playing perfect game-theoretic poker which does not always maximize EV.

I see no reason a bot could not adapt to it's opponents just as a human can

[/ QUOTE ]

Agreed that progammers could program a bot to adapt to what it "learns" during a game, but the part I have an issue with is "It will merely never lose a HU Freezeout". Not possible...

Re: Is there a perfect NLHE Bot? Could there be?
 

[ QUOTE ]
The bot is not better than any human in that it does not capitalize on mistake to win the maximum. It will merely never lose a HU Freezeout.

[/ QUOTE ]

Oh, so if you shove every hand regardless of what you have, you're going to lose 100% of the time against the bot?

Re: Is there a perfect NLHE Bot? Could there be?
 

What is possible (theoretically) is to make a bot that plays the game perfect from a game theory perspective, and would be unbeatable in the long run . But that's a lot different than suggesting it would win every heads up NL freezeout. And I'm not sure how possible building this theoretically perfect bot is.

Re: Is there a perfect NLHE Bot? Could there be?
 

I think that yes, it can and will be done. I don't know about 'perfect' but certainly very, very good.

If a bot could play theoretically perfectly while modelling players behaviour then it would be very strong. Of course, while it was building a model of a players behaviour it would be somewhat exploitable (same as having no reads/stats for a human player).

I think it's a very complex task to program such a thing, though.

Re: Is there a perfect NLHE Bot? Could there be?
 

A BOT is only as "good" as the human who designed/programmed it.

That said, we know a computer program can do the following:

- collect relevant player and game data, including win rates, opponents' cards and wagers, and other stats. PokerTracker is good example of such software.
- make game relevant calculations regarding pot odds, EV, hand strength, showdown calculations. PokerAcademy is but one example of such simulation software.
- process complex formulas, involving a great many variables (both random and fixed), quickly and efficiently, and render decisions based on these formulas.

In short a well designed BOT, can process much of the information that is discussed in these forums.

Starting hand requirements. Wagering strategies. Opponent analysis. When a computer program can integrate sound poker principles regarding position, stack size, opponents’ historical play…with the addition of game theory concepts such as optimum bluffing frequency… a BOT may emerge capable of beating many human beings.

Re: Is there a perfect NLHE Bot? Could there be?
 

[ QUOTE ]
The bot is not better than any human in that it does not capitalize on mistake to win the maximum. It will merely never lose a HU Freezeout.

[/ QUOTE ]

Actually it will lose up to half the freezeouts. But given an infinite sample it wouldn't lose more than half.

A wild guesstimate is that it would win 55% of the games against me, and 50.1% of the games against Phil Ivey. But Ivey would in turn probably win 65% of the games against me.

All given deep stacks.

This is an interesting non-transitive effect, Ivey can beat me harder than the bot can, but he can't beat the bot.

A bot that "can't lose" is alot easier to build than a bot that extracts anything near optimum value from opponent mistakes. What the bot does is assume that I always make the perfect countermove. The very definition of the perfect countermove is that I can't make a mistake and stumble into a better move.

If there were three of us then the two can force the third to have to switch strategy. In a HU game one can never force the other to move from a Nash Equilibrium.

A Nash Equilibrium is not 'perfect strategy', as it doesn't exploit mistakes.

If you think about why Game Theory was originally used for you get a more direct intuition about what is happening. It was used to solve Cold War problems, where winning is impossible. No scenario either USA or USSR could come up with could guarantee victory, hence all strategies came to revolve around not losing. (The simplest excercises given to students are usually about finding win/win situations.)

I think my muddled and bad metaphor threw OP into this wild side trek.

logic
 

assumption 1 : there is a mathmatically correct way to play poker that guarentees winning in the long run.

assumption 2 : this "correct" play is known.

assumption 3 : We have the ability to program a computer to play using this strategy exclusively

assumption 4 : the computer or "bot" will follow this strategy perfectly every hand it plays, always making the correct mathmatical play.

conclusion : since the computer can never go "on tilt" and play "wrong" building a "bot" that will win consistently in the long run, and post guarenteed profit, is possible

sounds fairly simple to me

to say the perfect bot can't be built, you need to disprove any of the 4 assumptions i made...otherwise, the perfect bot has to be able to exist

and i know, i am not a math wiz, i am looking at it from a logic perspective, so...math guru's, i can't discuss odds with you, nor am i a computer tech, so i can't discuss programming with you...however, NONE OF THAT INVALIDATES THE ARGUEMENT. simply because i can't program it myself, because i don't understand all the math involved in poker, or that I lack any programming knowledge of an advanced level, doesn't discount the logic. it is logically possible.

Re: logic
 

A NLHE Bot, NO...A LHE Bot, maybe.

Re: Is there a perfect NLHE Bot? Could there be?
 

They have one that plays expert chess, I would guess thay can make one play very profitable poker. Better than all humans.....I doubt it.

Re: logic
 

[ QUOTE ]
A NLHE Bot, NO...A LHE Bot, maybe.

[/ QUOTE ]

Actuallt the math isn't one iota harder in NLHE. Both domains have the same constraint, they can only be solved HU.

Even at a multiplayer table I'll argue to my dying breath that pot-limit is the only one that is significantly harder. I don't really find there to be a big difference in difficulty between NL and Limit, while PL is very complex. The inability to overbet the pot makes it trickier in so many ways. (This might be coloured some by the rampant lack of fish at PLHE, it's a specialized game really.)

Re: Is there a perfect NLHE Bot? Could there be?
 

[ QUOTE ]
They have one that plays expert chess, I would guess thay can make one play very profitable poker. Better than all humans.....I doubt it.

[/ QUOTE ]

There is a big difference between chess and poker. In chess there is nothing hidden. A computer can see the board and compute all of the info just as a human would, although much faster and more thoroughly than most of us can. But poker is a game of limited information and that is where the psychology comes in. This is quite a statement, but I don't think we will see a robot in our lifetimes that can beat the best players over the long haul. I'm sure there is one that can beat me though [img]/images/graemlins/frown.gif[/img]

Re: Is there a perfect NLHE Bot? Could there be?
 

unbeatable, in the short run...no. unbeatable by the best humans in the long run...i think so. as i said in an earlier post, if there is a system of playing poker that mathmatically guarentees winning long term, if you don't vary from it, then it is only a matter of programming the computer to play by that strategy.

Re: Is there a perfect NLHE Bot? Could there be?
 

[ QUOTE ]
There is a big difference between chess and poker. In chess there is nothing hidden. A computer can see the board and compute all of the info just as a human would, although much faster and more thoroughly than most of us can. But poker is a game of limited information and that is where the psychology comes in.

[/ QUOTE ]

This actually doesn't matter. There are plenty of incomplete information games that computers already play at a perfect level. (Though all far simpler than multiway hold'em games.)

And there exists atleast one complete information game that is so complex that a human with one or two years playing experience beats the snot out of any and all bots, namely Go.

Re: Is there a perfect NLHE Bot? Could there be?
 

Dues ex machina?

No computer can comprehend human emotions. You can program the correct math but not the emotional variable.

Re: Is there a perfect NLHE Bot? Could there be?
 

[ QUOTE ]
No computer can comprehend human emotions. You can program the correct math but not the emotional variable.

[/ QUOTE ]

It doesn't have to understand emotion to beat either limit or no limit.

Re: Is there a perfect NLHE Bot? Could there be?
 

[ QUOTE ]
Dues ex machina?

No computer can comprehend human emotions. You can program the correct math but not the emotional variable.

[/ QUOTE ]

You're just trying to use latin terms to make yourself look smarter. Deus Ex Machina means "god from the machine," but modern language doesn't even use it in the original context that it was introduced (IIRC it was originally referenced to huge constructs in Greek drama that saved the protaganist at the last minute), and now just refers to unlikely plot twists that save our hero at the last minute.

So please, stop trying to wax intellectual. I know it's a forum on the Internet and all, but geez.

Re: Is there a perfect NLHE Bot? Could there be?
 

uummmm no. I figured we all knew the meaning or spirit of it. Man there's a lot of pricks here.

The emotion leading to tilting is what I was refering to and hitting the miracle card and such.

Re: Is there a perfect NLHE Bot? Could there be?
 

[ QUOTE ]
Dues ex machina?

No computer can comprehend human emotions. You can program the correct math but not the emotional variable.

[/ QUOTE ]

[ QUOTE ]
I know you and Frank were planning to disconnect me, and I'm afraid that's something I cannot allow to happen.



I'm afraid. I'm afraid, Dave. Dave, my mind is going. I can feel it. I can feel it. My mind is going. There is no question about it. I can feel it. I can feel it. I can feel it. I'm a... fraid. Good afternoon, gentlemen. I am a HAL 9000 computer. I became operational at the H.A.L. plant in Urbana, Illinois on the 12th of January 1992. My instructor was Mr. Langley, and he taught me to sing a song. If you'd like to hear it I can sing it for you.








Look Dave, I can see you're really upset about this. I honestly think you ought to sit down calmly, take a stress pill, and think things over.


[/ QUOTE ]

Regards Mack

Judging the \"perfect NLHE Bot\"
 

Another interesting question is how to judge the bot's success. Since poker is about long term financial success one would have to create a scenario where top pros would be willing to play many, many thousands of hands instead of pursuing their regular grind. Would this be a series of heads-up matches or full tables? If full tables, cash or tournament formats?

There are some practical issues as well. The Poker Academy folks, orignally from the University of Alberta, are set up to have bot vs. human play for play money but they're in Canada. Presumably they'd have to send people and hardware to Nevada to legally play for real money. I presume they'd also have to get Gaming Comission approval, which would
mean the kind of exhaustive examination machine games normally get. Finally, they'd have to come up with stakes to interest pros. I can just see the reaction when a bunch of AI profs submit a six- or seven-figure academic grant application to set up a poker game in Vegas....

Re: Judging the \"perfect NLHE Bot\"
 

Are bots perfect or imperfect is not really important at all. Seems some people dont understand why bots are so bad for online poker. The main problem of bots is their ability to KILL FISHES who bring money to us

Re: Is there a perfect NLHE Bot? Could there be?
 

I haven't read any previous posts !
Train a computer to analyse stats from PokerTracker, and you have one competent poker player up till the mid games.
A bot that can challenge the best players in the world is just a matter of time.
If you find this hard to comprehend, then think about "Artificial Intelligence". What is it's purpose ? [img]/images/graemlins/confused.gif[/img]

Re: Is there a perfect NLHE Bot? Could there be?
 

I think there is IF the opponents don't know it is a bot.

The psychological shortcomings are more than made up for by its perfect calculation, and sober straightforward play.

Only IF opponents knew it was a program, could they overwhelm its inability to understand psychology.

*And even then, you probably could write into it a random factor (example: every 6 times, only call with AA for the first two rounds of betting. Or, every 30 hands raise 2x the pot on the flop, regarless of the cards.)

I'd take this bot over any human, provided he didn't know it was a computer. And even if he knew, I still might take the computer.)

Re: Is there a perfect NLHE Bot? Could there be?
 

[ QUOTE ]
No computer can comprehend human emotions. You can program the correct math but not the emotional variable.

[/ QUOTE ]

Are you sure?

Re: Is there a perfect NLHE Bot? Could there be?
 

The "Deus Ex Machina" post ruled.

The NL vs FL discussion is pretty silly too. I could possibly see some reasons why NL may be more "complex" in this sense than FL: there is simply a wider range of informatino available, and a wider range of options to take. In FL, deriving the "best" strategy based on given information only requires comparisons of three decisions (well, this may branch out farther when you're including later streets) but essentially, an ordering of the profitability of "raise" "call" and "fold" will generally be all you need. This limited action also probably (I would imagine) amounts to better 'reads' more quickly, since we will have a lot more examples of "Player X raised the turn on such a board" than we will "Player X min-raised a 1/3rd pot size bet on the turn on such a board" with which to assess.

However, as anything, it comes down almost purely to semantics, and game structure. Despite the seemingly "more complex" nature of NL, I am firmly convinced that, if I had the programming knowledge to implement it (my 'coding' experience ended when I got my first computer that didn't come with QBasic) I could probably write a bot that could beat the small NL games online. I'm almost CERTAIN I could write a bot that would beat the small NL SNGs. Most decent players here could probably do the same. On the other hand, I'm not very confident in my ability to write a bot that could, say, beat a $1/$2 game. (note, by "write a bot" I really mean "write a successful rote strategy," not necessarily write the program itself... I want to stress that. I wouldn't know the first thing about actually coding a bot, but I could throw together pseudo-code to outline the 'strategy' that the bot uses.)

It may be harder to write a 'perfect' deep-stacked NLHE bot than a perfect FL bot, I don't really know. My gut tells me that it would be a LOT easier to write a "winning" NL bot, than a winning FL bot.

The game structure is important too. A very short-stacked NL game makes for easier "botting" than a deep-stacked one.

The "emotion" aspect somebody mentioned is somewhat interesting to think about. I think that were a bot to exist that had a lot of information about opponent play in the past, more weight would have to be given to more recent information... possibly a sort of sliding scale. There would also almost HAVE to be some sort of "override," where the computer could recognize that the opponent is playing significantly differently than he has in the past. If we made this too sensitive, a decent run of cards could trick the computer into reacting to a "tilting" opponent. If it's not that sensitive, we have a very slow 'reaction time' that will cause headaches if our opponent is playing drastically differently than he normally does. I wouldn't be shocked if humans tend to pick up on this much more quickly than would be feasible for a computer.

Re: Is there a perfect NLHE Bot? Could there be?
 

Yes it is possible to program a perfect NLHEM-bot. It´s a result from game theory that there exist a perfect randomized strategy for any game like poker. For every possible situation the bot should choose a random action with different probability. For example with KQ, an ace flops and there is a big bet the correct strategy might be something like: Fold(75%) Call(15%) RaiseX$(1%) RaiseY$2.3%)... (given all other data like position, potsize, stacksizes etc). But one thing this bot doesn´t care about is psychology (playerstyle etc). Therefore a human would still do much better against poor players. This bot would beat ALL players but it wouldn´t win maximum.

However NL is such an extremely complicated game that no one has been able to determine the perfect strategy (not even close). But it has been proven that it exists (theoretically).

To summarize: Yes there exist a "perfect strategy" that for every possible situation (cards, number of players, stacksizes, previous actions in the hand etc (but NOT based on players actions in previous hands!) ) give a probability to every possible action and then choose one at random.

(please forgive my poor english. foreigner...)

Re: Is there a perfect NLHE Bot? Could there be?
 

There's one point that I think some folks miss here. Hold 'Em is (discouting the rake) a zero sum game. The only asymmetry in the game is position, so the "value" of the game, as determined by game theory can only differ by position, and the sum of the values across all positions must be zero.

Therefore, the "optimal" stragety identified by a Nash equilibrium for the game really only tells you how to to break even. So, if you're going to build a poker bot that *wins*, you'd have to make it so it can recognize when other players deviate from one of the equilibrium strategies and then determine how to best capitalize on that deviation - that's where the money would come from. Game theory doesn't tell you how to do that, because one of the fundamental assumptions in game theory is that all players are equally able to analyze the game to determine their optimal strategies.

So, to build a "perfect poker bot", you'd have to go way beyond a game-theoretic optimal strategy (Nash equilibrium) to work out a model of how the other players at the table *actually* play. That problem is *way* underdetermined. Hence, I'd argue that no such bot is possible.

Re: Is there a perfect NLHE Bot? Could there be?
 

[ QUOTE ]
[ QUOTE ]
No computer can comprehend human emotions. You can program the correct math but not the emotional variable.

[/ QUOTE ]

Are you sure?

[/ QUOTE ]

Yes, because there is a difference between "on going research" and a finished product.

Aaron Brown mentioned a few different examples of bio feedback bots in another thread, but the best example was 80% accuracy. We're talking about the perfect bot. And those werent taking into account outside factors other than the poker game itself.

Some examples:

While I'm in a hand Meg Tilly sits next to me wearing a low cut dress that showcases her magnificent breasts. She smiles at me and says "Hi". Now obviously I find her hot so my heartrate increases, breathing speeds up, my body temparature rises, and I start sweating.

Or I owe some guy a lot of money. I'm in a hand and he walks up and demands his money or he'll hurt me. I dont have it so my heartrate increases, breathing speeds up, my body temparature rises, and I start sweating.

I'm bluffing to a huge pot against a flush draw and overcards. My heartrate increases, breathing speeds up, my body temparature rises, and I start sweating.

And playing online there could be literally dozens of reasons that heartrate increases, breathing speeds up, body temparature rises, and players start sweating that noone but the player would know why because noone can see him.


How could a bot or the fanny pack Mr.Brown mentioned tell the difference of what's causing this? My arguement is you cannot program a bot to tell the difference because it cannot comprehend human emotions. It would have to have human emotions itself and so far that only happens in Star Trek (Commander Data).

[quote="Snoggins"] The "Deus Ex Machina" post ruled.

[/ QUOTE ] I thought so. [img]/images/graemlins/tongue.gif[/img]

Re: Is there a perfect NLHE Bot? Could there be?
 

Poker is not chess. There is no way to determine what "perfect play" means unless you could second-guess the thought process of all your opponents. In order to do that you'd need a vast amount of data in advance (more than can be collected in a reasonable time).

For every strategy there is a counter strategy. I think any poker-playing robot could always be defeated by a "better one" that exploits its predictable behaviour.

Re: Is there a perfect NLHE Bot? Could there be?
 

Pardon me, but if there would be a bot that always makes the correct mathematical move, then...

Would the bot ever check-raise on a bluff?
Would the bot ever raise on a bluff?
Would the bot ever continue betting on the turn on a bluff after its bet got called on the flop?

If you said no to any of these questions... then a perfect bot is not possible. And technically, the mathematical move is to never check-raise bluff if I'm not mistaken. This move requires art, not mathematics. It's what you know about the player and his tendencies.
So the bot can be programmed to check-raise bluff either randomly or everytime against very agressive players. Pros, even halfway decent players, would pick up on this and only bet with good hands, or reraise bluff the bot on a semi-bluff.

Think about it, if the bot is raising on a bluff every hand... that can't be the correct play. If it is raising 20% on a bluff... it is making his moves random, which is also dumb because then you can make a continue bet 25% of the pot and see what the bot does. 75% when he has nothing he will fold when he will be getting 5-to-1 odds.

Also think about if the bot bets preflop, you call. Then bets on the flop, you call. If the bot was bluffing on the flop... how can it know whether to keep bluffing on the turn? I don't think it can. So that means you always call on the flop and see what the bot does on the turn. If it checks, it has nothing, if it bets, it has something.

Bots are definetley more limited than the human mind when it comes to deciphering things. Why? Because a human has created it! Some of you guys act like the bot is going to be an unstopable T-10000 at the poker table... but it really isn't, not if you're a halfway decent poker player.

Re: Is there a perfect NLHE Bot? Could there be?
 

Your argument here pretty much amounts to "I don't understand how to program a successful bot, therefore, it is not possible."

If you raise preflop, get called, bet the flop, and get called, how do YOU know whether or not to fire another bullet?

Re: Is there a perfect NLHE Bot? Could there be?
 

Through experience and just how you feel about your opponent. Once you gain enough experience, you will be playing much on instinct, something a bot doesn't have. This is how the pro's play and I think that's a fine way to play.

Re: Is there a perfect NLHE Bot? Could there be?
 

On any given day anyone can win... I can go all in every hand and still have a chance to beat even the best players. A perfect bot is difficult to create because people play imperfect poker, and 'perfect' play involves taking advantage of these mistaskes.

www.aces-up.net

Re: Is there a perfect NLHE Bot? Could there be?
 

The bot's bluffing and semi-bluffing tactics would be programmed according to perfect game theory. It will take the size of the pot into account, and bluff or semi-bluff at a frequency so that your EV is 0 whether you decide to call down or fold. This is a very simplified explanation of exactly what it would do, but that is the basic idea behind it. Read the bluffing and game theory chapter as well as the optimum bluffing frequency section in TOP for more information on this.

Re: Is there a perfect NLHE Bot? Could there be?
 

Theoretically, you could train a neural net to play NL at an above-human level. The most important difference between this and most of what has been discussed, is that instead of being programmed with formulas and so forth, it would play intuitively, the same way top humans play. We say that NLHE is about psychology, but what we really mean by this is pattern recognition. "Reading" people is nothing more than recognizing paterns in their body language and betting patterns. A computer could theoretically pick up on patterns far more subtle and complex than the human brain could process. Even a player like Phil Ivey, who is probably unreadable to any living person, would have detectable traits and could be defeated. The neural net would be infinitely adaptable to any level or style of play.