There's a group at the University of Alberta that does research in poker strategy. They have a program that plays heads-up poker (2 players) as well as the best in the world. Adding more players into the mix significantly complicates things though, so it will be a little while yet.

There are people that run bots on online poker sites, to varying degrees of success.

Note that the person you are replying to specified No-Limit, which is much harder. It's been a while but last I looked the Alberta Group could only do Limit Poker.

UofA has done a lot of work on no limit, but it doesnt get as much publicity as their limit stuff

Those bots have gotten significantly better the last few years. I would say the best bots today are beating nearly all of the recreational players and even the lower rungs of professional players - but the very best players can still beat them, should they be paying attention to their tendencies. The bots have a big advantage over them though in that they are robots: they don't get tired or hungry and their judgment never gets clouded on runs of bad luck.

PokerStars and FullTilt have worked hard to rid their sites of bots but many other networks have turned a blind eye to them, some even explicitly allowing them - since the bots pay rake like any other player. Every so often the players will revolt and the sites will crack down but the bots always find their way back again.

I agree with everything you have said, but it is worth pointing out that the best online players are assisted by bots. The top players use software that can immediately show your last 10 bets and the outcomes as soon as you raise.

My point is that if you play low stakes, you get crushed by patient bots. If you play high stakes, you get crushed by cyborgs.

Im not positive that the current HUDs do that , and even if they did, i dont think its a huge advantage. Players use statistical tools which could be seen as unfair , except they are so common now that a significant percent of players from the lowest stakes to the highest use these tools. Just having this data is not enough to make soemebody a winning player as it does not suggest a specific move, the player still needs to figure that out. High stakes players are just much, much better than lower stakes players, with or without the HUD.

Those tools present their own problems but they aren't bots. Bots are making decisions and acting on them, not simply providing additional information for players to incorporate into their decisions.

Which sites explicitly allow bots?

The thing about poker (versus chess), is the notion of complete versus incomplete information. In chess, you know the location of every piece. This means you can compute possible positions (theoretically) until every possible game ends. Even better, you can play the algorithm against itself, since it can see everything either player knows. This makes chess computationally expensive - there are many positions, and you can do fairly meaningful analysis on every possible move. The best chess algorithms know how to aggressively cull paths which won't work well early on.

In Texas Hold'em, you only know ( initially ) your pocket cards, then as the rest come up you gain a bit more information. But compare this to the number of cards you can't see as a player. Those overlays they show on TV are effectively the best you can do; there's just no more data to work with. It's not like chess where throwing more computing power at it helps anything.

If you played just one hand that would be true.

Poker strategy is based on large sample sizes - you're not trying to figure out the optimal strategy for a single hand, you're trying to figure out an overall strategy for groups of hands against groups of boards.

Winning strategy for limit poker is essentially making sure you're game theory optimal, so that there isn't a strategy someone can apply against you that will in the long run be profitable. You can't apply a simple greedy algorithm hand per hand, you need to be aware of the ranges of hands you could have and what strategies your opponent could be applying against those hands.

That is the not losing part, the winning more part is figuring out where your opponents arent playing GTO and figuring out if you can afford making your strategy exploitable to in turn exploit your opponents (because if you try to be always GTO you may not beat rake etc).

Looking at poker hand by hand is like looking at chess move by move.

Players can try to be GTO, but limit is not solved, so they are simply guessing. This post (and many others across the web) seem to assume that we know GTO or that we can reason it, but twe cant as this isnt a solved game yet

That's silly. You have an opponent whose previous actions you can analyze with an eye to the possible cards to come and the moves you can make. Add more players and the calculations are more complex. There's plenty of data for a computer to crunch on.

In a group setting, computers can beat humans. They just keep track of more stuff.

And if you get 2 computers communicating, sitting at the same table, well, HUGE advantage.

It is very hard to make a bot that doesn't do stupid things. > 90% will fail and > 95% will fail to beat rake. Plus the top online poker players are basically half machine.

They use stuff like Hold'em manager and various sites to analyze a large amount of statistics against their opponents. Ranging from performance, games they tend to play and win, as well as storing a local database of all their hand histories, positions etc. they encounter. They also use the software to go over their own play to look for weaknesses with the computer giving suggestions. The bots would be going against something that is not fully human and similar to advanced chess results, they would stand little chance.

This is a strange argument, you say that top players are not fully human, maybe 80% human and 20% machine, for example, implying that this makes them better than being fully human, and yet suggest something that is 100% machine could not beat it. How is this so?

Because there are (especially in No Limit) many places human intuition destroys the bots. On top of this, the aids allow them to have some of the advantages of bots including, summary statistics of hands, betting patterns and styles for every player met, post play review tool for learning to plug leaks and weak spots and a bunch of other stuff.

The combination of stats, practice and experience is better than a bot or pure human. I know this because I have a close friend that does this stuff and makes close to a 100K a year and botting might have once been a hobby of mine.

There is no infinite computation, so brute force play is not possible - even in chess (PSPACE). The problem with bots is that they tend to have blindspots they are blind to since they can't very well model their understanding. Human Computer Combinations still win and will probably continue to do so until the day something comes alone that can beat us at the kind of higher order pattern recognition we are so good at. http://www.palantir.com/2010/03/friction-in-human-computer-s...

For poker, only the very small and IMO uninteresting subset of 2 player limit is effectively solved. The combination of human minds + machine computation will be the top intelligences for the foreseeable future.

Im sorry, but the best computer programs available will beat the best chess player in the world 90% of the time[1].

Also, while I dont doubt your friend is good at poker, that does not mean he has an understanding of the algorithmic advances in game theory as they apply to stochastic games. Further, simply knowing him does not make you an expert, even if you had a background in poker botting. How do we know that a game needs to be perfectly solved in order for it to beat the best players? Would the best nlhe hu player in the world beat the best bot at 100bb stacks? Yes. Will this be the case for the foreseeable future? Probably not. My point is that you cannot speak conclusively about a field like this with so little knowledge of what is going on academically to advance it.

1: http://blog.chess.com/Clavius/most-impressive-computer-game-...

There is certainly no argument against the fact that humans are no longer in the same league as chess bots but what is also true is that the combination of bots and humans can beat top specialist chess computers. Bots + Humans > Bots > Human for these games. See what Kasparov has to say about it here: http://www.nybooks.com/articles/archives/2010/feb/11/the-che...

My friend? No he does not know about stochastic game trees or regret minimization or any of that fancy stuff. But that knowledge is not needed for top poker play any more than a Basketball player needs to know about differential equations to put a ball in the net.

And perfect play? I do not believe that perfect play is required to beat the best players, that is not even computationally tractable. Now, while it is true that computers may some day beat top players (in n-player NL), it is also true that we are very far away from that. I am confident that we will reach that point but it is hard to say when. Probably after Go. But already the top players, using computational aides, are not fully human. With those kind of tools they are able to regularly play the game at a level that is far higher than the old unaided standards, raising the stakes even higher for bots.

Little idea? I keep apace with current machine learning though not fully up to speed on poker specific stuff. Would need to spend a month or so to catch up.

Anyways to wrap this up, what I am saying is Human + Bots > Bots and I'm predicting right now that when Poker Bots surpass humans, they still will not be able to beat that combination.

Like this: https://en.wikipedia.org/wiki/Advanced_Chess The human picks the strategy, and the computer makes sure the tactics are perfect.

that would be miles ahead of what the current top players do. the statistical aid that players get simply recounts what has happened in the past "your opponent has raised X% of the time, called Y%, etc." the human doesnt prune the tree of poor decisions and then have the computer simply tell it what to do, the player needs to make all strategy decisions. Further, that type of hybrid "cyborg" situation is decidedly not optimal, it is simply the best way to make use of limited computing ability. Instead of a computer brute forcing every possible situation, the human tells it what ones are bad and to avoid. With infinite computing power, no human intervention would ever be needed, nor would it help. Obviously we dont have that kind of power, but for chess I believe it is already at the point where humans cannot improve upon a bots strategy by advising it. For poker, the issue is that these games have not been studied anywhere near to the level that chess has, so it is very likely that even a top pro would not be able to pick the "best" starting points for the program to then solve [or what ones to ignore, except for the very obvious which wouldnt be helpful]. For nearly all 2-player fixed-limit games, a pure computer strategy would win. When you add more players, the hybrid solution is needed, however, it is still not the theoretical best.