Poker Vs Chess

A computer has beaten top human poker players:

Humanity was dealt a decisive blow by a poker-playing artificial intelligence program called Polaris during the Man-Machine Poker Competition in Las Vegas.  Poker champs fought the AI system to a draw, then won in the first two of four rounds (each round had Polaris playing 500 hands against two humans, whose points were averaged.) But in the final two rounds of the match, Polaris beat both human teams, two wins out of four, with one loss and one draw. IBM’s Deep Blue beat chess champion Gary Kasparov in 1997.

This has gained very little news coverage, in stark contrast to the Chess case, even though far more people play poker than chess.  To explain both facts, note that most people can more easily see that they lack chess than poker expertiSe.  They can watch a poker expert fold or raise and imagine that they would have done the same, but know they have no idea why a chess expert makes his moves.

This fools people into thinking they could be a poker champion, which is why so many more people try at poker.  And being less impressed by existing poker champions, people are less impressed by a computer who beats those champions.  The moral: beware of underestimating computers by underestimating the difficulty of the tasks at which they excel. 

GD Star Rating
loading...
Tagged as:
Trackback URL:
  • http://jamesdmiller.blogspot.com/ James Miller

    This will soon destroy online poker because human players will fear they are playing against a computer.

  • Ian Monroe

    The link to the story is broken.

    I would be interested to find out what the precise methodology was for this competition.

  • http://hanson.gmu.edu Robin Hanson

    James, I agree, but its worth asking: since people haven’t feared playing against human experts – why fear computer experts more?

    Ian, hope its fixed.

  • http://profile.typekey.com/stefankrieger/ Stefan Krieger

    This doesn’t look like it has much statistical significance for the claim that the computer is better than the best human.

  • Michael

    I have completely different reasons for being more impressed with chess playing computers than poker playing one. First, Polaris can only play heads-up. To me this is an enormous defect. It’s like saying that computers could outplay grandmasters only in 4-piece endgames. Second, Polaris is beating players in, effectively, online poker. People do not watch online poker on teevee because we are more interested in the psychological/physical aspect of real-life poker. For most people, poker is not just about making correct decisions, but about trying to judge what another player has based on how they act physically. In that sense, online poker is a different game from what most people think of as poker; computers cannot play the latter.

    I think what is more interesting is that people are not as amazed that computers are much better than humans at checkers. What happens is that once computers are sufficiently better at something, we flip a switch in our brains from “human/creative activity” to “computer/brute force activity.”

  • Schizo

    James: You underestimate the overconfidence and, frankly, stupidity of a lot of people on the internet. The whole point of this post is that people think poker is easy/simple, so they’ll think they can beat bots.

  • http://www.allancrossman.com Allan Crossman

    It appears that the game was Limit Holdem, which is often said to be easier for a computer than the (very popular) No Limit version of the game. In Limit, you have at most 3 options for any decision (fold, call, raise) whereas in No Limit, you also need to choose the value of any raise, so there are a lot more options.

  • Ian C.

    There is an element of chance in poker that is not present in chess, perhaps that is why people are less impressed. I guess they try to use statistical techniques to eliminate the possibility that it was luck, or at least quantify it, but y’know…

  • http://michaelgr.com Michael G.R.

    “James, I agree, but its worth asking: since people haven’t feared playing against human experts – why fear computer experts more?”

    Aside from some psychological factors of “playing against a machine”, I would think that people know that human poker experts are relatively rare, while if poker AIs become mainstream, any 12 years old kid could set it up. It’s hard to make the first one, but after that, you can copy software easily.

  • Carl Shulman

    There’s also the fact that the chess victories happened first in time, and the earlier media coverage desensitized people to the general phenomenon of computers mastering games.

  • http://blog.greenideas.com botogol

    @Michael said “For most people, poker is not just about making correct decisions, but about trying to judge what another player has based on how they act physically. In that sense, online poker is a different game from what most people think of as poker; computers cannot play the latter”

    I’d be very surpised if the software managed to win at poker without taking account of the behaviour of the opponent. Although granted the computer can only take account of revealed betting behaviour, being blind to the supposed ‘tells’ (Perhaps that’s why it is so good)

  • http://profile.typekey.com/michaeljameswebster/ michael webster

    Be interesting to see if the computation time has to increase unrealistically for the ordinary 6-7 player game.

  • Ian Monroe

    After looking a little more closely, I have a few thoughts:

    First, what separates poker from chess is game theory. Chess is a game of perfect information; you can see every available move for either player at any point in time. Poker, however, is a game of imperfect information. Information is hidden about what your opponent holds and what cards will come. This means that you can no longer calculate far into the future, but rather you have to make educated guesses at each individual decision point in the game.

    Second, No-Limit hold ’em is the standard benchmark for poker, if you believe that human tournaments reflect the state of the game. Limit, as was noted in a previous comment, is in many ways a simplified version of the game.

    Additionally, heads-up play is all fine and good, but it constitutes such a small part of the overall game, and correct strategy is so different, that I’m not sure that saying a computer that can beat a human in heads-up game means that a computer can beat a human at poker is a justified claim.

    I don’t believe that there is currently any software on earth that could sit with top players in a large tournament and come out victorious. If you’ve got a field of 2500 players, increasing blinds, shifting seating, and a variety of table sizes, then you have a far, far more complicated problem to solve. I don’t think it’s been done yet.

    I’ve read that the real challenge for computer game strategy is the game of Go. I understand that there is no computer in the world that has yet posed any serious challenge to top human players, simply because the decision tree is so vast.

  • http://jamesdmiller.blogspot.com/ James Miller

    Robin wrote “James, I agree, but its worth asking: since people haven’t feared playing against human experts – why fear computer experts more?

    Most of the players in online small stakes poker games are not experts because experts can’t play many games at once. But if a computer could play thousands of games at once there can’t be an equilibrium in which a significant number of online players are humans playing without the help of a computer because in such an equilibrium it would be profitable to buy a computer and have it play, say, 10,000 games at a time.

    True, online poker providers could set up some test to make sure their players are human. But in this case it would be profitable to hire lots of low paid workers in some poor country and have them play with the assistance of a computer. And it would also be profitable for low skilled workers to play for themselves with the assistance of a computer.

    I suspect most people wouldn’t find it fun to either know they are playing against a computer or to play by doing exactly what a computer tells them to do.

    Finally, I suspect that the advantage computers have over humans is increasing in the percentage of players who are computers. So as more and more online players are computers, the expected return for humans would fall, which would decrease the number of human players and so…

  • Michael

    botogol writes: “I’d be very surpised if the software managed to win at poker without taking account of the behaviour of the opponent. Although granted the computer can only take account of revealed betting behaviour, being blind to the supposed ‘tells’ (Perhaps that’s why it is so good)”

    Note the use of the word “physically” in quote you excerpted. I am making a distinction between live poker and internet poker. Aspects of live poker include reading other players’ physical tells, as well as controlling one’s own. Even if computers do not need to perform the former, they have an unfair advantage regarding the latter.

  • http://zbooks.blogspot.com Zubon

    Ian Monroe, as you and others have said, heads-up with a limit is a simplified problem, and the more complicated problem appears unsolved. Do you take the further contention that it is not solvable in practice, in theory, absent whole brain modeling, etc.? Or is it just a matter of time, practice, and processing power before we see it fall?

    I toss that to anyone, since it seems like the relevant question. Finding benchmarks not yet reached is not terribly interesting if they obviously will be reached. Some people have described Go as relying on some non-verbalizable Go-sense; a more conservative position is that the brute force look-ahead approach to chess will not work for Go even with hardware that improves by several orders of magnitude. Some approaches will never work for some problems, while other walls can be battered down if you hit your head against them hard and often enough.

    To add non-question content: the question of computers in online games does not require that the computer be the best. If you have a computer/program that is in the 60th percentile of online players, you can profitably let it run wild. You do not need to win it all to make money, and it may even slow down counter-bot efforts if the best players always turn out to be humans.

  • http://yudkowsky.net/ Eliezer Yudkowsky

    I’ve read that the real challenge for computer game strategy is the game of Go. I understand that there is no computer in the world that has yet posed any serious challenge to top human players, simply because the decision tree is so vast.

    Last time I checked, no Go program in the world had posed a serious challenge to decent amateur human players.

  • Nitpicker

    Dude — it’s “expertise”, not “expertize”.

  • Ian Monroe

    @Zubon:

    I would contend that computer players will inevitably become more sophisticated, to the point where eventually they will match top humans. I’m not convinced that there is a “perfect strategy” in poker, even in a theoretical sense, so I would contend that it is not possible to build an unbeatable poker program.

    I say this because the game depends on random occurrences. You can calculate, for instance, what the likelihood of drawing a particular card will be, and thus whether is it statistically beneficial to raise, call or fold, but it still, you cannot anticipate the outcome of the hand perfectly before all the cards come out.

    Consider this – It is possible (I would say common) for two top human players to make statistically correct decisions throughout the course of a hand. It’s possible that neither player makes any tactical mistakes at all, and still, one of the two players will win the hand, and one will lose. It is zero-sum.

    While it seems that this particular challenge tries to minimize the “luck” factor by having the same sequence of hands played twice, with the players swapping places between these sets, in real life this isn’t how the game is played. If you re-define the problem to become solvable, it doesn’t mean the original problem was. In this case, special rules were introduced to minimize factors that otherwise would complicate the game, so I think that the best you could say is that a computer may be able to solve this variant more effectively than a human. I don’t think that the results commute to the original problem.

  • http://brokensymmetry.typepad.com Michael F. Martin

    At the very highest level of play, I would guess that most of the edge that human players have would be washed out by the odds. Why shouldn’t an algorithm that does nothing but play the Kelly Criterion win at this level? It’s the theory of runs.

  • burger flipper

    I don’t know how seriously to take it, but there’s been a recurrent series on building a poker bot on reddit for the last few months:
    http://www.codingthewheel.com/archives/how-i-built-a-working-online-poker-bot-4.

    From what I’ve read, detecting physical “tells” is vastly overrated. Detecting tendencies can be invaluable, and already serious online players use software that accumulates stats on all their online opponents.

    That was how the cheating at Absolute Poker was detected. People within the company were using accounts that allowed them to see everyone’s hole cards. Winning (cheating) players seemed ridiculously lucky to some opponents. They were caught out as impossibly lucky by datamining software.

    Those with computer assistance already have a distinct edge. I would think eventually those players with a human component will have a distinct disadvantage.

  • Jeff

    I think Robin’s explanation for why people fail to appreciate the skill of a poker champion is closest. The big difference is the randomness, so even a skilled poker player is going to lose a lot of hands that could have been won with different (and worse!) play. A random spectator has many chances to say, “ha! I would have won that hand.” Of course those hands are outweighed by the cases where the spectator would have lost more money than the expert, but good ol’ confirmation bias takes care of that.

    Also, if you watch TV poker, you get to see the other players’ hands, so when you see a player making a losing move, it looks like bad play, when of course it may very well be the correct play.

    (If you haven’t tried writing a poker AI, you’d be amazed at the complexity of the problem. Take a ridiculously simplified version of poker: two players each draw a card; player one has a blind bet of one chip; player two can fold, call, or raise one chip; player one can call or fold; high card wins. Even that game has remarkable subtleties in the optimal strategy.)

    –Jeff

  • ad

    This has gained very little news coverage

    The ability to play chess well is widely considered to be a sign of intelligence. Poker does not have the same cachet in the popular mind.

    The media would not have paid much attention if a machine had become a top Go player, either, because their audience would know of little reason to regard it as significant.

  • Matt Huang

    There is a very big gap between the “top human poker players” that Polaris has beaten to the actual top human poker players. Stoxtrader would be absolutely crushed by people like Cole South and Phil Galfond.

  • Doug S.

    The poker-playing AI that I heard about a while ago, which gave a human champion a run for his money, used an approximation to the game-theoretic optimum strategy for heads-up limit poker. To clarify, the game-theoretic optimum strategy the is one that is not exploitable; the most any opponent can do against it is break even. For example, in Rock-Paper-Scissors, the “optimum” strategy is to choose randomly, with an equal probability of making each throw. This strategy cannot be exploited. On the other hand, this strategy is also incapable of exploiting an opponent’s mistakes; even an opponent that always chooses Rock will break even.

    In other words, this poker bot, if it’s the one I’ve heard of before, doesn’t play to win. It plays not to lose, using no opponent modeling whatsoever. If its opponent doesn’t make “dominated errors” – moves that are strictly worse than another – it will tend to break even, no matter how predictable its opponent is. Furthermore, a move is a dominated error if and only if an optimal strategy uses it with probability zero. Even if the optimal strategy says “make this move with probability 1/1000” then it won’t do better than break even against a strategy that says “make this move with probability 1.” However, Poker is sufficiently complicated that it’s not apparent whether a move could be a dominated error, so there might still be ways for a human to screw up that an “optimal” bot will exploit.

    (Also, the bot uses an approximation to the optimal strategy, as finding the true optimal strategy is computationally intractable. However, it does seem to be a pretty good approximation, as it has been successful against top human players.)

  • http://www.iki.fi/aleksei Aleksei Riikonen

    As a semi-professional poker player, I’d like to say that the article being quited does not understand the event being reported very well, since they say “Humanity was dealt a decisive blow by a poker-playing artificial intelligence program”.

    This was no decisive blow, it was a single test, where each round consisted of only 500 hands of one of the simplest forms of poker, and the AI come out on top by a rather small margin. Could very well have been luck. Further tests will tell.

    That said, I will not be terribly surprised even if we get real evidence that AIs can now beat humans at Heads-Up Limit Hold’em. My guess however is that this current AI will not end up demonstrating decisive superiority, but that we might very well see such a superior AI in 1-3 years. Or not.

    And that is if we are talking about Limit poker. In the case of No Limit, I find it very hard to guess when AIs will reach the human level of ability. Won’t be extremely soon, and the problem might even be AI-Complete, requiring essentially human-level ability to model the strategies of another human. No Limit poker is not so much straightforward math as Limit poker is. It is more about being better at modeling your opponent’s decision-making than they are at modeling yours. (A nice shortcut to seeing whether this Polaris AI will eventually end up easily beatable by some pros would actually be to hand over it’s source code for their examination. Even Limit poker might be substantially less mathy at the highest level than I estimate. I am not a Limit pro.)

    By the way, AI bots have already been used for several years to make lots of money playing Limit poker online. Some people have farms of bots playing lots of tables. The poker sites disallow and (some) try to monitor this, but there are always anti-detection solutions. The prevalence of bots in Limit poker has caused a lot of people to shift over to playing No Limit.

  • http://www.xuenay.net/ Kaj Sotala

    Doug S:
    It plays not to lose, using no opponent modeling whatsoever.

    The linked article indicates this is not the case.

    “And secondly, we have added an element of learning, where Polaris identifies which common poker stratagy a human is using and switches its own strategy to counter. This complicated the human players ability to compare notes, since Polaris chose a different strategy to use against each of the humans it played,” Bowling said.”

    Before the Las Vegas match, this newest version of Polaris had only played two matches against champion poker players, resulting in one loss and one victory. Polaris repeated the pattern of improving as it learned, falling to humans in the first two rounds, but defeating them in rounds three and four. “Repeatedly, I heard players exclaim that they had never seen a human do that before,” said Bowling. “Switching strategies really threw the humans for a loop.”

  • dagon

    The moral: beware of underestimating computers by underestimating the difficulty of the tasks at which they excel.

    s/under/over/ Or maybe just “misestimating”. It’s very easy to generalize incorrectly in either direction from this story. In reality, it’s no more (nor less, to be sure) impressive than, say, a medical or financial system with a few hundred variables and a few dozens of models to weigh.

    http://poker.cs.ualberta.ca/papers/Papers/johanson.msc.pdf is a good semi-technical description of how the software works. As you read it, you’re likely to be less impressed with it as a measure of “machine ability”, and perhaps more impressed with the engineering and theory that goes into “machine-creating ability”.

  • M

    Now that the University of Alberta has a robot that can beat the best poker players I think we can start to see the Edmonton campus expand.

  • Tim Tyler

    Re: very little news coverage – the story was mentioned on Slashdot:

    http://hardware.slashdot.org/article.pl?sid=08/07/09/2014228

  • http://www.arcade40.com/cat/117/Poker/p1.htm poker

    in poker there is a lot of luck involved, isn’t it? Then how a computer could excel in poker games?

  • http://www.spaceandgames.com Peter de Blanc

    poker: the computer can simulate the life cycles of 6 million four-leaf clovers per second, making it luckier than the luckiest human.

  • http://www.allancrossman.com Allan Crossman

    in poker there is a lot of luck involved, isn’t it? Then how a computer could excel in poker games?

    It’s not just luck; there are important decisions to make all the time (unlike roulette, craps, etc). But it’s true that luck plays a big part, which is why poker players tend to measure success in terms of how well they do over the course of a year, say.

  • Michael

    The set of those people who are still reading this thread and who do not check in on the Freakonomics blog is probably pretty slim, but for those in that intersection, you might want to check out Levitt’s two cents:

    http://freakonomics.blogs.nytimes.com/2008/07/14/overreacting-to-a-computer-beating-poker-pros/

  • JAMES D. SIMS

    CHESS vs. POKER, LIFE & “End Game Theory”]

    =================================================================

    (1
    of 4).Difference between,: Playing-Chess, Living-Chess?

    (2
    of 4). Chess-Pieces: King, Queen, Bishop, Knight?,{?}, {? } [which can be

    considered the ROOT, in the game of, LIFE & PROFIT?

    (3 of 4). Can Chess be performed without a CHESS-BOARD-FORMAT, has it

    also been represented/called by
    another name?

    Justasking
    pot-p-s? abnjd

    (4 of 4). Can the Origin of the
    Chess Board’s Format & Intent in the 21

    Centennial, can be referenced in relation to ,
    “Common-Sense”, more

    Important,
    as well as a “GAME”?

    justasking pot-p-s? abnjd