A footnote in the govt response resembles ike's post:

They understand long run luck. But in order to determine whether luck or skill predominates, you have to decide on a timeframe. The question of predominance is certainly hard to define, but that doesn't mean the question doesn't exist. No one is saying it's possible to define a unit of luck, and a unit skill, and come up with a ratio that applies to any competition. But it's still a question. If you were considering betting on Ben Lamb at this year's final table, you'd wrestle with the question of how much skill or luck predominates, for that specific event, with those players and those blinds and stack sizes, and decide on a fair price, even if you never defined some luck units / skill units ratio.

Looking at betting markets across different forms of competition might help. As a baseline, we could look at the betting markets for a sporting event whose outcome we all agree is determined predominantly by skill. I haven't read much on what the lawmakers say here, but I'll assume they'd agree that the NCAA tournament fits. If luck predominated, every team's line would be 64-1. But skill matters, and it's reflected in the lines. Compare that to historical markets for the WSOP final table. If luck predominated, each of the nine players would have odds roughly proportional to their chip stack. But I recall that the year Ivey made the final table, his posted odds were roughly three times as good as his chip stack implied. So the betting market indicated that skill mattered in that case. Did the markets indicate that skill predominated? If you could show that the distribution of odds sufficiently resembled the distribution for the NCAA tournament, then you could say yes.

I obviously meant the win, not the winrate.

The "numerous other" gambling games you've mentioned are not the same type of game as poker because those games are unfair, whereas poker is a fair game. The casino can play a fixed strategy and be certain to win, whereas an optimal strategy in poker is not certain to win.

Also, there's nothing arbitrary about people trying to get poker declared a game of skill. David has presented a reasonable, cogent argument that shows why poker should be considered a game of skill. I don't see any reasonable, cogent argument from people who say it's dominated by skill. There's tons of good evidence that skill dominates over luck, and very little that supports luck over skill, and therefore you can say that by a preponderance of the evidence poker is a game of skill. There's nothing arbitrary about that.

First of all, I would like to take this opportunity to thank Mr. Sklansky for not only this contribution to the whole "skill v. chance" argument, but also for all of his prior work on the subject. DS's books and articles have been frequently cited in PPA legal briefs and many of the arguments we present have been aided by his thoughts and participation in the discussion. Thanks David.

As to this particular post/idea. I do find it very interesting and hopefully useful in the future. The concept is clearly sound as a matter of logic, but my concerns center on the points I have quoted above.

Initially, although I suspect it either already exists or would not be too hard to create, I would want to see some sort of mathematical proof for the proposition that in a game of mixed skill and chance "There has to be an x% chance that the better player will be ahead after y hours." While it seems intuitively correct to me (and probably most of us) I would want to have the back up before making this representation to a court.

Once that is out of the way, the rest seems to flow pretty straightforward until we get to the " an expert does 25% better than the average player in 3-4 hours." Would that not depend, at least to a degree, on the true level of difference between the average player and an expert and how would we measure that?

The final problem, though, is the phrase I have put in bold type. Anyone want to suggest precisely how one would go about establishing the actual proof? Anyone want to volunteer to be the "test" players?

Working through issues like this is a major part of our effort to get poker openly legal. It is far from the only part of the effort and, as has been noted, it could be thwarted by new legislative enactments. But ultimately that is really the point. By pushing these issues in the Courts we threaten to change the status quo. Nothing gets politicians more motivated to act than their fear that the status quo may change without them controlling or influencing that change.

IOW, by continuing to press these issues in law courts and in the court of public opinion, we improve our odds that we will get favorable poker legislation within a reasonable period of time.

And anybody and everybody who contributes to that effort gets my thanks.

Skallagrim

I think the argument in itself "luck vs skill" is just really pretty stupid.

Of course poker players make money off of the "mistakes" of others. Human beings are inherently flawed. If they were perfect no mistakes would be made.

Everyone accepts the fact that Michael Jordan was the best NOT because of his perfection....but because he was simply MORE CONSISTENT than his opposition.

Tiger Woods doesn't birdie every hole and yet we accept him as being the best because he is simply MORE CONSISTENT than his opponents.

In poker, obviously a player who makes the proper decisions more consistently than their opposition will win over time. This is absolutely no different than any accepted sport or intellectual competition.


The problem is predominance. Is Monopoly a game of skill? Well sure...I beat my 7 year old regularly. Is Yahtzee? Same story....beat her regularly. If, however, I were to play a strategy that was optimal against other optimal players the edge is nearly negligible.

Therefore, we must assume that predominance is generated from the size of the edge that is generated by close to optimal play vs average or slightly below average play. I think David has it pretty close to correct in what he is saying.



To simplify, IF a contest where a player playing a near optimal strategy vs an average or slightly inferior strategy results in an insignificant and nearly minute long-term edge over the competition, then we can define the game as primarily luck with an element of skill.

IF a contest where a player playing a nearly optimal strategy vs an average or slightly inferior strategy results in a significant long-term edge over the competition, then we can define the game as primarily skill with an element of luck.


Now, of course, we haggle over significant vs insignificant. Really? This is what we are haggling over?

Come on....poker is OBVIOUSLY defined under any reasonable definition as primarily skill with an element of luck.

The real kicker....our lovely government doesnt give a damn whether it is a game of skill or a game of luck. They couldn't care less. It's just an excuse used to further certain agendas. It's a way of Kyl supporting without supporting, etc...it's an ugly mess of corrupt clowns finding a reason to do something they should have done years ago.

That being said, let's just call a spade a spade. Golf is a game of skill with an element of luck. So is basketball, football, baseball, etc....

and YES....poker is an OBVIOUS similarity....now let's move on and get this done because nobody really cares anyway

"Firstly one must point out to him that expert poker players are usually playing against very good players. So the skill difference is small."



Even he would advise you to walk that one back dummy.

For those who want an idea how judges approach the idea of "predominance" here is a very well known quote from a 1973 opinion by the Alaska Supreme Court (it was not about poker and I have removed the citations to other cases):

(3) Is where things always get bogged down in poker. Sometimes skill clearly controls the final result despite the chance element: namely when you bluff a player who is holding a better hand than you. Sometimes chance clearly controls the final result: namely when you get all-in before a showdown and the underdog hits the unlikely card. Between those 2 extremes are a wide variety of possible hands. Proving which factor, skill or chance, is the controlling or dominating factor in those hands is the task.

At least from a legal point of view.

Skallagrim

David- You aren't the arbiter of common sense. Just because you think it commonsensical that a 75% chance to win over an average player is a good criterion for 'predominantly skill' does not mean that when someone objects that your criterion is not commonsensical you can just re-appeal to common sense in order to vindicate yourself.

All criteria for 'predominantly skill' are going to be conventional, and therefore somewhat arbitrary, including yours.

So the first part of our job is factual- we have to show the courts, statistically, that we can win with frequency X for sample Y in game Z vs opponents Q. The second part of our job is rhetorical- we have to convince the courts to accept a standard for 'predominantly skill' which puts poker in the legal category we decide we want.

I said that there is smaller than a 25 chance that an average player would outperform an expert holding the same cards after three or four hours. Not that the expert would do 25% better. The idea is that that the 75+ percentage is more than halfway betwwenn 50% all luck and 100% all skill. The key here is that the average player should be the "placebo" because when skill is high but skill differential is low it can take a long time for the skill to predominate over luck, even if the game is mainly skill. Professional baseball is almost all skill even though no team is a 3-1 favorite over another.

One way to show my poker contention would ironically not require expert volunteers but rather require beginner volunteers. Have them play their normal style in a tough game. Fewer than 25% of them will be ahead after four hours.

I'm speculating that the 75% figure would appeal to other people's common sense. Halfway between 50 and 100. And that most people recognize that two highly skilled but equal players depend mainly on luck to win a mainly skill game thus making it more fair to measure expert vs average. I'm guessing it would be THEIR common sense, not mine.

So where is the data to demonstrate this?

If you are looking for data involving "expert" and "average" players it seems that the mounds of internet data wouldn't fit because of the professional skill level of many of the players and the difficulty in determining identiy.

Maybe some of the larger tournaments like WSOP and WPT can be used as data sets to search for this trend as there are "expert" and "average" players in these tournaments.

Sorry David, only had time to rest the first post, so maybe its been brought up, but I think you are going about this ALL wrong.

The correct thing to do is to compare poker to other tasks, in particular tasks that people do for their career.

For example, you could compare a poker player's profit to a restaurant owner's profits on an "hourly basis". I can almost guaruntee you that the restaurant has bigger monetary swings than poker does due to attendance variance, economy variance, and the like. I mean, restauraunts surely dont have a 90% chance to be profitible over a 100 hour period of being open.

And then there is the obvious comparison to stocks, where people go years in the red. Why has Berkshire Hathaway lost money over the course of the past year?

Also, the correct person to use for your judgement of poker is the very top expert, which i guess everyone would agree is Ivey. I mean, if its a question of whether or not skill plays a factor, why wouldnt you use the most skilled player to show just how profitible skill makes you? And in your analysis, you should note that a large portion of his winnings are against other incredibly skilled players.

Also, you should note that rake should be ignored if you are judging skill. it is a matter of the probability of a player winning over other players when the player is more skilled.

Presumably his variance would be unbelievably low if he were to play .25/.50 or something. I am not an excellent player and I won at 15.38 PTBB/100 at .01/.02 for 12,000 hands AFTER rake when I had just started. Clearly my variance would be nearly non existant, especially when I was playing 6+ tables at a time.

You could also use the results of the Polaris bot matches to prove some players are not as skillful as others.
"On July 3-6, 2008, Polaris competed against six human professional poker players in the Second Man-Machine Poker Championship, held in Las Vegas at the 2008 Gaming Life Expo. Polaris defeated the human players with 3 wins, 2 losses and 1 tie. Each of the six sessions was a duplicate match of 500 hands against two different players, resulting in six thousand hands played".
Since each player had precisely the same cards, in the same order, their results can only be different because of their skills, not just luck.
http://en.wikipedia.org/wiki/Polaris_%28poker_bot%29

Actually, thinking about it, the fact that bots exist is pretty much all the evidence you need to prove that poker requires skill. You can't program a "lucky" bot.

Hi David,

I more or less agree with your speculation about people's common sense. I'm just noting that if lots of people disagree with us, we're probably wrong about common sense on this issue!

I was using "your criterion" in my other post in the same way you are using it- as a criterion of common sense for the average person, and not as David Sklansky's intuition or even David Sklansky's imagined intuition before he knew about probability.

There is also various forms of luck in sports leagues similar to poker.

The better basketball team may lose 3 games in a row, but over the course of an entire season they will most certainly make the playoffs. They may lose game 1 of the playoffs, but the better team will most likely come back and win over a 7 game series. A team may have a 54% field goal percentage in basketball against their opponent's 44%, but it just so happens that variance leads the 46% miss rate to a larger frequency than usual, and they thus lose the game. Basketball would generally be considered a game of complete skill, but there are countless ways that variance plays a role within the game. Referee calls cannot be controlled by players. A star player can be injured by an opposing player which was out of the star player's control.

There are countless forms of internet gambling, all of which seem to act with impunity and operate with severely larger percentages of house rake. So many sites charge an $x monthly fee, where the best performers win a cash purse of $y. The sick thing is, on most of these sites, if P=all players registered, $x(P)>>>>>>$y.

People enjoy participating in competition, and usually, they want to feel that achieving excellence in order to win was worth something. Showing the value of excellence is most usually awarded with prizes. Prizes have monetary value, and people prefer money over a pre-selected prize of equal value. Banning internet gambling restricts companies from creating a league that awards prizes. Creating one would be an enormous lawful risk. Yet many do, and they operate with impunity. When the video game for MLB advertised that the first person to pitch a perfect game would be awarded $1M, that is 100% gambling. The prize purse is taken out of the revenue of the video game. Yet, if you were to make a poker site that charged $50 to play on that site indefinitely, and 1 person would eventually be awarded $1M, you might go to jail. Poker is unfairly treated on a different playing field as these other operations.

All the public needs to know is whether someone with better-than-average poker skill will profit over the long run or not when playing weaker players. I am not an expert on other gambling games in Vegas, but my understanding is that even with skill, a better-than-average player will lose in the long run because of the house's edge, etc.

In response to Ike's earlier philosophical post, what we mean when we say "skill predominates" is not that skill and luck are mutually exclusive, but that a player with better-than-average skill will win in the long run should he play against weaker players, iow that skill will "overcome" the luck factor in the long run.

And if a better-than-average poker player cannot win in the long run, wtf are we doing playing this game let alone trying to have its internet form legalized?

This is a great point. In baseball, if you hit 33% of the time you get an official at-bat, you are going to be a Hall of Famer. It's a game where a high level of success is just succeeding 1 in 3 times. Team versus team is your point, and I get that, but it goes even further; pitcher versus batter.

Also, in basketball a marksman from 3-point range is hitting 3's 40% of the time. This means he needs to succeed just 2 in 5 tries to be a Hall of Fame level 3-point assassin.

In football, success is much more dominate statistically. A good completion percentage for a quarterback is over 60%, and often over 65%. A good kicker hits 85%-95% of his tries at field goals, and closer to 100% on extra points. In this game, a dominate percentage proves skill.

My point is that whether the sport (or game) is skill based has little to do with predominance of percentage. In fact, no one is arguing pro basketball players are just lucky, pro baseball players are luckier still, and yet football players are ALL skills. Everyone admits all three sports are skill based games, and the best players among them are the most skilled most consistently, and their skill in the game is purely judged against other players in their game no matter the competition's skill level or if they even compete against each other (as in the D-Leaugue doesn't play the NBA, but is nonetheless accepted as less skilled, the AA baseball teams don't play the MLB teams but are accepted as less skilled, and the CFL doesn't play the NFL but is accepted as less skilled).

So, my conclusion is that your example is simple and valuable for proving the skill to the average person...but for the slightly above average intellect that requires more, continued use and depth of sports analogies works best. The worst player need not play the best player in poker to prove the skill involved or the disparity therein. Cash versus tournaments is like football versus baseball in variance (baseball succeeds just 1 in 3 times while football succeeds closer to 2 in 3; poker cash games success can be measured in hours, while tournament success will be judged in much larger spans of time with low levels of success like 15-20% ITM being really very good).

If no one is arguing professional sports are luck games versus skill games then they have no argument that poker is a luck game versus skill game.

And as far as legal issues...there is no way to read those definitions and not include the housing market, investment in capital (holding cash in your pocket as opposed to investing it in something else), or even buying milk (investing it will not sour before it is consumed) in the definitions of gambling. EVERY SINGLE "investment" is a flowery way of saying GAMBLE or "game of chance". In all investments/gambles information, and who has more of it, influence success, as do random variance in markets (for housing it might be the random variance of neighbors or who they sell to, for capital holding it might be the daily world exchange rates of which other nations actions are relevant, and for milk it can be the random mistakes in the process of production and pasteurization).

There is no way to define porn or gambling, it's purely a subjective value based on indivudal interpretation of values purely made up. "You know it when YOU see it" is the only real definition, hence why both are undefinable. So, if you label poker gambling and not investing then I label Wall Street and pocket change as gambling and not investment...neither of us can be proved wrong, it's all subjective individually. Our courts seem to be made up of control freaks and idiots sometimes (and before you defend them, remember Dred Scott please).

Thanks Sklansky for the great argument...and also for the various books I've enjoyed.

(Ps. For those using the word "luck", I'd like to suggest for the sake of this conversation we drop it. There is no such thing as "luck" scientifically...it's statistical variance and anomaly. If we continue to use these superstitious terminology we play into the critic's hands (pardon the gambling metaphor). The variance is real, the "luck" is perception. It is a game of variance versus skill, not "luck" versus skill. The "luck factor" is exactly proportional to the variance factor. And the square root of variance is standard deviation.)

I think it's pretty clumsily written, worded, and lacks substance.

Since skill and luck are traditionally defined to be fundamentally inherit in poker games, I'm interested to hear the reasons that follow to make you think that it's nonsensical to discuss any such relationship whatsoever. And what willful ignorance are you referring to exactly? That part sounded like an unneeded and out place ad hominem.

By evens out I take it you mean that a roll of the die, or a hand dealt is random. Well of course. However this doesn't follow that 'skill determines an outcome' is a meaningful claim.

Take for example someone playing optimal strategy in Blackjack over someone playing randomly. One has an edge over the other, so skill determines an 'outcome' for their relative win-rate, but the 'outcome' for the casino is that they always win in the long run (minus card counting, etc). So I don't really see what you're trying to say at all here. Luckily for me there's a big wall of text to read, I guess.

No one says there's some soft of correlation between the two, because in general conversations they aren't quantifiable concepts. People are saying that: if something has more luck ('predominately'), then skill will matter less. And vice-versa. I don't see the big deal in pointing out the obvious.

It's almost as if you're saying that if I said: "In Blackjack, the house as 0.x% edge, and in craps it's 1.x% edge, therefore there is no skill edge and we can make a comparison between the two" - is a stretch.

Since you're not making this argument, what you said there is just pure fluff. No one's making a sweeping claim about the exact amount luck or skill there is in poker because poker is a game of incomplete information. The randomness of human decisions highlights this.

More fluff. Ironically your argument is about meaninglessness, but your argument is anything but coherent. Mostly because this mammoth sentence just looks like random words stringed together without saying much of anything which doesn't reword what you said earlier, except in a clumsier way.

So basically: saying that "This game has x% luck" is meaningless. And that it's only meaningful to point out that skill and luck are attributes of the game?

As per your rock example, it's more like: "I see that the rock has both grey and 'relatively large' attributes. The amount of greyness is x% and its largeness has been measured as x diameter." Yet you seem to be content with just pointing out the first steps of dissecting something.

--------

Above is basically just a response to a poorly written argument. My argument against it is just pointing out the obvious:

Whether you like it or not, the fact that a game has x amount of luck versus y amount of skill is pertinent for the courts.

The amount of luck involved in a game is exactly what determines the edge in house games. If the house doesn't have an edge then the player does. And which player has the greatest edge? The most 'skilled' player.I'm sorry that the word 'skilled' is used in this way. I put it in inverted commas to appease you though I hope.

Odds can be calculated, not 'luck'. Your odds might be 12% to win the hand when you shove 72o into pocket aces without looking at your cards, but it's 100% lucky when you do.

No matter what cards you're dealt in poker you might lose. If you win, you're lucky.

"Unless I get you to fold your hand, then it's all skill!"

I think Ike's theorem holds up well.

So can I just replace '100%' with 'positively' ? You were lucky because 88 times out of 100 you would have lost, so one could say that it was '88 units of luck.' There are degrees of luck, of course, which relates to the average likelihood of events. If you're 49% to win, and do so, you were lucky, but only a fraction so.

Its funny how most people on this site believes poker is a game of skill, yet the have neverending discussion on how to demonstrate that to the public.

Just dumb it down to the simplest possible level?