Harvard, Professor out to Prove God Exists Poker = Skill

by , May 5, 2007 | 5:00 am

There was a pretty interesting article in Thursday’s Wall Street Journal about a recent gathering at Harvard University, which set out to bring together some of the brightest minds in academia to prove beyond a reasonable doubt that poker is a game of skill. (Uh duh.) Howard Lederer was the main poker dude on hand, along with famed Harvard Law Professor Charles Nesson, Annie Duke, Andy Bloch, PPA honcho Michael Bolcerek, and a number cruncher who wants to run the math on billions of hands.

The WSJ’s unscientific poll is currently running 77-23 in favor of skill. Read the article and you’ll see this isn’t about rehashing old theories and debates … because really, what do you think the poll results would be if we asked: “Is life primarily a matter of skill or luck.” I am guessing 77-23 might be pretty close to the results here, too.

In the article, Lederer espouses a new talking-point argument that I hadn’t yet heard/thought of:

The “vast majority” of high-betting poker hands, he says, are decided after all players except the winner have folded. So if no one shows his cards, Mr. Lederer says, “can you legally argue that the outcome was determined by luck?”

Cool stuff — and good to see, in the ivory towers at least, a growing recognition of how some of what is currently shaking down in the poker world reaches into important future matters of internet law, international law, international business, and economics.

After his strategy session wrapped up, Prof. Nesson led the group to a bar for drinks. He was delighted, he said, at how the group “pushed game theory to the level of metaphor.” Sipping a scotch on the rocks, he tossed out the idea of creating a poker university, with himself as one of its teachers. Then, “we could infuse all levels of education with the skills that come from poker,” he said.

8 Comments to “Harvard, Professor out to Prove God Exists Poker = Skill”

  1. Rudy Stegemoeller

    As much as I love poker and as much as I support making it legal, I still think it’s bullshit when I hear players saying that it “isn’t gambling.” Obviously there is a strong element of skill, but the legal question is not whether skill is involved, but whether skill predominates over luck. I think that for the large majority of players, certainly the people I play with at the $80 buy-in charity tournament at the volunteer fire station, the answer is clearly luck.

    Howard’s argument, that in most hands no cards are shown, overlooks the fact that most of the time, the winning player is able to make the powerful bet because he has cards. And the people who fold their hands don’t have cards. Obviously there is skill in reading, bluffing, knowing when to lay down a hand etc. But how many players have those skills to an extent that can override the effect of bad cards? I think a very small percentage. The notion that luck evens out for everybody in the long run is false. It levels out to some extent, of course, but there will still be a bell curve and everybody will have their place on it. Also, the moments when luck really counts the most — at the final table of a tournament, for example, or in the two or three big hands that get you there — are relatively few, so the percentages have less chance of evening out over that relatively small number of hands.

    For the large majority of players, skill consists of memorizing some rudimentary ideas of which starting hands to play, and maybe even some basic percentges of pot odds on flush draws and straight draws, but that’s about it. That’s roughly the same amount of skill that it takes to play craps for the optimal percentages. 70 million Americans play poker, supposedly. How many read All In, Card Player, or Bluff? How many have bought the strategy books? A lot – millions maybe – but nowhere near 70 million.

    Again, I am completely in favor of legalizing poker, but not because it “isn’t gambling.” I would legalize all gambling. I once heard a pro, I think maybe Phil Gordon, saying “I don’t gamble at the table; I make investment decisions.” Fair enough, but what form of investment leaves you vulnerable to losing your entire stake to a 2-outer on the last card? Investors hedge against risks like that, gamblers revel in them for the adrenaline rush. Which description fits most of the poker players that you know?

    Just my cranky thoughts on a Saturday morning when I should still be asleep.

  2. Tim B.

    Obviously there is a strong element of skill, but the legal question is not whether skill is involved, but whether skill predominates over luck. I think that for the large majority of players, certainly the people I play with at the $80 buy-in charity tournament at the volunteer fire station, the answer is clearly luck.

    all youve really said here is that it isnt a matter of “luck vs skill” but a matter of “SUCK vs skill”… the proportion of skillful people playing a game is in no way connected to whether skill is the predominant factor. there are lots of examples, but the one that springs to mind would be a number of video games. i absolutely SUCK at fighting games like dead or alive or tekken. skill absolutely dominates over luck in those games, hungerfan has mad skills in this respect, and yet even i can beat him probably a quarter of the time or more, just by madly mashing buttons and getting stupidly lucky. as satisfying as it is to do that, i dont believe that the game is in fact driven by luck and that folks who think they are skilled at it are just deluded. just because people like me who suck at it CAN win by being lucky doesnt negate the fact that skill is predominate.

    The notion that luck evens out for everybody in the long run is false. It levels out to some extent, of course, but there will still be a bell curve and everybody will have their place on it.

    this belies a frankly PROFOUND lack of understanding when it comes to probablity (which is just “luck” by another name)… there absolutely is NOT a bell curve involved in the distribution of poker hands, with a statistically sufficient sample size, you will see EXACTLY the distribution the probability would indicate, within the standard deviation. additionally, when you hear that “luck” evens out over time, what is meant is that if you get your money in as the favorite in a, say, 55%-45% confrontation, you can expect to win 55% of the time, and as the sample size gets larger the more your results will converge at 55%. luck is utterly irrelevant in this regard. and this is the critical factor that leads to skill dominating over luck: the more skillful you are, the more accurately you will be able to identify situations with positive expectation, and the more your wager when you have a positive expectation (“the best of it” as they say) the more you will win. with a small sample size, statistical variance might lead you to believe that the variance (“luck”) is the predominant factor, when in fact it absolutely is not.

    you can easily demonstrate this to yourself if you have a little programming ability, or manually if you have a lot of patience and free time, by simulating the kind of situation that arises in poker all the time… one where the odds are SIGNIFICANTLY against your winning any single trial, but the expectation from each individual trial is net positive. consider a typical, fair (unweighted, etc) die with 6 sides. each toss of the die costs you $1. every time you roll a 6, you will be paid $6. you are significantly more likely to make a losing roll every trial, and yet, if you play this long enough, you ABSOLUTELY POSITIVELY WITHOUT QUESTION will break the bank, presuming you have enough starting cash to outlast variance (“allow the luck to break even”). there is absolutely NO LUCK REQUIRED. why? because this wager has a positive expectation. on average, for any 6 rolls, you will lose 5 dollars, and win 6 dollars, for a net profit of $1 every 6 rolls. each roll is “worth” +16 2/3 cents. roll only 10 times, and you might well be DOWN, even though you have a positive expectation. after 100, you might be down, but not likely. after 1000 rolls, you will almost absolutely be up on average $167. you might be $16 higher or lower, but you wont be too far from there. over 10,000 rolls, and without question youll be up to the tune of $1,667, give or take a hundred bucks. take it up to 100,000 rolls and beyond, and you will get closer and closer to the mathematical expectation. even though you have to be lucky to win, youll still get rich this way. its the closest thing to a money tree youll ever find. not only does luck not predominate in positive expectation situations, it isnt even a factor long-term. at all. period. this is a fundamental principle of probability and statistics.

    so, given that:

    1) luck is irrelevant in positive expectation situations taken over time, and
    2) identifying positive expectation situations in an incomplete information game like poker is a skill-based ability

    then it is absolutely, incontrovertibly clear that skill is predominant over luck in poker. just because few people posess that skill to any significant degree (ie most people suck), or that the skill is HARD and requires a degree of dedication and talent to cultivate, in NO way changes that single simple fact of the game.

    man i hope that made sense. im not at my logical best early in the morning 🙂

  3. Tim B.

    when i said “each toss of the die cost $1” what i mean is, you WAGER a dollar each toss. on a winning toss you are paid 6-to-1, that is, your wager is returned and you are paid $6. just to clarify that the payout in the example does not include your wager.

  4. Tim B.

    consider also that in poker, positive expectation situations are NOT always card-driven. identifying profitable bluffing situations, ones with positive expectation (a common example is betting half the pot when heads up, if you accurately determine that your opponent will fold more often than one time in three, it is a profitable bet, your cards be damned) is another example of the way skill in poker manifests itself, and how it is often utterly disconnected from the cards in play.

  5. Rudy Stegemoeller

    I appreciate your comment.
    I don’t question your analysis, but I question your assumption of how long it would take for the statistical expectations to be realized over a large population of players. If 70 million people play a thousand hands of poker each, the variation in the distribution of good cards among those 70 million will be enormous (no, I’m not going to try to do the math); and more important will be the variation in the distribution of “lucky” cards WHEN THEY REALLY COUNT.
    If we all lived to eternity (maybe there’s poker in heaven?) then I would agree with you. But (again subject to doing the math) I would suspect that the variance among 70 million players would be far greater than what could be resolved within a normal human lifetime. So some people will have relatively better “luck” over a lifetime of playing than others will.

    The same analysis would apply to a tiny sampling, say ten players, if their skill levels were roughly equivalent. In other words, skill predominates over chance in poker only when skilled players are at a table with unskilled players, and even then over a long period of time.

    That leads back to the first point, which I think is the more significnat one; does it matter that the majority of players don’t learn or use the skills that are available to turn poker into a positive expectation game? The context is that people are debating whether poker should be legal, under the assumption that “gambling” is somehow immoral or dangerous and must be banned for the good of the public. Poker advocates argue, “screw the crap shooters and slot zombies, we’re different.” My argument is that a majority of poker players play the game as a gambling proposition, so if gambling is indeed a social evil, then poker doesn’t have a strong claim to being exempted. Of course I reject the whole idea of banning gambling; it should all be legal and regulated.


  6. Lavigne in Austin

    just because poker is a game of skill doesn’t mean it isn’t gambling.

  7. DanM

    true enough, lavigne. because talk to some serious sports betters, and they too will attest that there is skill to it.

    but i thought, as per our bill, poker is considered a special kind of gambling, right … such that we do not need a constitutional amendment like the casino bills that have since died?

    the key difference — from a state legal standpoint — is that in other forms of gambling, the house only wins when the player loses. in poker, we all know, the house has no stake in who wins or loses … and in fact, the idea of a player going broke is one of the worst things possible … so thus they have an incentive for such not to happen.

  8. Marvin C

    The Missouri Supreme Court said live poker, video poker, live blackjack, and video blackjack are games of skill. They allowed those and only those games when Missouri casinos first opened in spite of the Missouri Constitution outlawed “Games of Chance.”

    The Il. casinos fought the Missouri casinos and won. Until there was an ammendment to the constitution, only those games were available. It was the only reason Missouri had live poker rooms.

    That said, even though it’s a game of skill; it’s still gambling.