Poker, a game of luck or skill?
Is poker a game of luck or skill? Obviously, the answer lies on a continuum, as everything in life involves some luck and some skill. This article compares the level of luck in poker with other activities such as chess, golf and investment management.
I also ask the related question whether poker should be regulated as a gambling game or as a skillful pursuit. Billions of dollars rest on the answer to this question since in many jurisdictions ‘games of chance’ are prohibited or tightly regulated.
What do the lawyers say?
Lawyers are not known for their statistical skills compared to, say, statiticians. Their pronouncements are very important, however, as they decide whether poker can be played in a particular jurisdiction, and the litmus test is often whether the game is ‘more skill than luck’. The legal position has flip flopped on this point, going back a long way.
In 1906 (City of Shreveport v. Bowen) the court determined that, ‘It is a matter of common knowledge concerning which there can be no doubt or dispute that draw poker is a gambling game, pure and simple’. In 2007 an English court ruled that poker is primarily a game of luck in finding the owner of London’s Gutshot Club guilty of violating the Gaming Act, which required a licence to host games of chance but not games of skill.
On the other hand, in 2009 the organiser of a poker club in Colorado was found not guilty of ‘illegal gambling on games of chance’ in a trial where the jury agreed with a statistician’s testimony contending that poker is a game of skill.
In a far reaching judgement, in April 2011 (known as Black Friday in American poker playing circles) the FBI shut down American facing poker sites and blocked their access to American players. Today in the US the legality of live and online poker varies from state to state. Online poker is legal in New Jersey, Nevada and Delaware while at the other end of the spectrum in Texas most forms of gambling are illegal.
The US is generally moving towards being more permissive, mainly due to the growth of online sports betting and the potential for government to make tax revenue. Even in Texas many venues are exploiting legal grey areas to facilitate live poker.
Here in Australia online poker is not legally available to play for real money due to the Interactive Gambling Act 2001, which prohibits the provision and advertising of online gambling services to Australian residents, and the amendment bill that was passed in 2017, which banned online poker for real money outright. Note that, technically speaking, it is illegal for sites to offer poker but not for Australians to play on those sites . . . though beware, you may never get your money back …
The legal situation is often quite cynical. Black Friday referenced above was done ostensibly for consumer protection from the evils of gambling and to curtail money laundering, but the ruling was surely with an eye to protecting tax revenues and the local casino industry. Where I grew up in South Africa, for example, gambling was seen as sinful and was banned, yet horse racing for money was extremely popular and there was a huge casino called ‘Sun City’ a short drive away.
Measures of skill
Returning to the question of whether poker is a game of luck or skill, everything has an element of luck: you could be playing the finals of the World chess championship (a game generally regarded as 100% a game of skill), and the reigning World champion Magnus Carlsen could come down with the flu on the day and give you a walk-over. Cricket is a game of skill, though the weather can mean that the result of the opening coin toss has a sizeable bearing on the result. Rugby is a game of skill, yet referee decisions have an outsized (and seemingly random) influence.
In fact, any game that is perfectly deterministic would hardly be classed as a ‘game’, not to mention be not much fun to play!
It is often easier to recognise unskillful play than its opposite, skillfull play. In poker the player has to make a number of decisions each hand: fold, call or raise, and if so to what bet size. These decisions have a bearing on the final result. We can easily call out eggregiously bad play: the player who folds a royal flush on the river has played badly. This is unlike roulette, for example, where no bet is bad a priori. It is literally impossible to play badly any game that rests on the outcome of a coin flip. The converse is that any game that is possible to play badly should also be possible to play well.
Another measure of skill is whether it is possible to improve at a game, and Dedonno and Detterman have showed that players who were given strategic instruction in the game of poker outperformed those who were given no instruction.
Clearly, though, poker involves a lot more luck than chess. Duersch, Lambrecht and Oechssler quantified this, estimating that poker contains about as much skill as chess when 3 out of 4 chess games are replaced by a coin flip. (The paper constructed a rating system for heads up sit ‘n go tournament results, which produce a winner and a loser like chess does, though the conclusion can be extended to multiplayer cash games). Where the legal criterion is that a game be at least 50% skill to be regarded as a ‘game of skill’, poker would be a game of luck, being only 25% skillful.
Steve Levitt looked at evidence from the World Series of Poker (WSOP) and showed that while recreational players had a chance of winning the WSOP, and individual professionals could bomb out early, the group of players that could be considered ‘good’ players going into the tournament earned a return on investment (viz. their buy ins) of +30%, versus -15% for everyone else.
A paper by Croson, Fishman and Pope compared poker to golf. Golf is a useful comparison group, as it is an example of a game thought to be primarily skill-based. The paper found the skill differences among top poker players to be similar to skill differences across top golfers, as measured by the propensity of golfers who have top 18 finishes in elite tournaments to finish in the top 18 in subsequent tournaments. I find this result interesting, as recreational poker players are willing to sit down and play professional poker players for even money, but you don’t find recreational golf players challenging the local club pro on the same terms!
An article in the Financial Times (‘why poker can beat investment management hands down’) noted that if a measure of skill is how often good performance repeats itself, poker is a more skilful activity than investment management. Something for my investment banker mates to ponder . .
Looking at the results of my small-stakes online poker group, I’ve estimated that player skill ranges from -40 big blinds (BB) to +40 BB per 100 hands, with a standard deviation of 190 BB/100 hands. This points to an important feature of poker — one can play a lot of hands!
Let’s say we play 100 hands of online poker twice a week, and we’re a weak -40BB/100 player. We may expect to play circa 10,000 hands in a year. Now the probability of winning any particular hand is almost pure chance. The probability of winning a 100 hand session in one evening is 42%, which will still feel psychologically like a coin flip. However, by the time 1000 hands are played the probability we’re ahead will have fallen below 25%, and at about 9000 hands we would have lost money in all but 5% of cases. The weak player will have good weeks and even months but is almost certain to lose over the course of a year.
Finally, we have the existence of professional poker players, suggesting that poker is a game of skill — over the timeframes that count. Being a professional craps or roulette player is statistically impossible, but it does appear possible to be a professional poker player, albeit it’s likely a fragile financial existence (despite the flexible working hours!). Many pros supplement their income by owning a casino (Doug Polk), vlogging (Brad Owen) or cheating (??).
Poker should be regulated as gambling
I didn’t intend to turn this blog into a moral treatise on the vice of gambling and I should declare my personal stance on these things, which errs towards the liberal. Most forms of activity amongst consenting adults should take place with minimal regulation. That’s not the same as no regulation though — whilst I’m not against gambling, per se, I do recognise that gambling addiction is problematic and real.
Every year Australia has a gambling harm awareness week (in 2023 was 16 to 22 October). But it’s a bit like the racing car driver who has his foot down on the pedal but then decides to tap the breaks at the same time. There’ll be an advert about the Melbourne cup horse race, ‘The race that stops the nation!’. But the advert will sign off with a few cautionary words such as: ’chances are you’re about to lose’; ‘what’s gambling really costing you?’; ‘what are you prepared to lose today?’; or ‘think what else you could be buying’.
Australians love their gambling: In February 2023 The Australia Financial review asked, ‘do Australians have a pokie problem’? Yes, indeed they do. (A ‘pokie’, also called a fruit or slot machine in other countries, is a zero skill electronic game of chance).
Australians are the biggest gamblers per person, than any other country. That Australians also lose the most money is an obvious corollary that almost goes without saying. The state of New South Wales alone has 1 poker machine per 81 people, whose residents collectively lose $74bn each year. It’s also highly regressive — this loss falls mostly on the lower socio economic groups.
So: should poker be treated as a gambling game?
There are two questions here. Is poker a game of luck? (No). Should it be treated as a game of luck? (Yes). In my view, there is a lot of skill in the game of poker, yet it should be treated as a gambling game, with whatever the consequences of that treatment entails in that jurisdiction.
I believe this because poker taps into the gambler’s mindset. There are essentially 3 stakeholders in a casino poker game: first is the casino, who earns a rake. Then there are the professional players, who hope to make enough money from the ‘fish’ to beat the rake. And finally there are the said fish, the recreational players who go to a casino to gamble.
For the recreational players it’s like a game where being able to solve a highly complex puzzle gives you a 51% edge, while playing terribly still gives you a 49% chance. The average punter can’t distinguish this from any other game at the casino, and the financial outcome for them is no different in the short run (they might win) or the long run (they will lose).
I would argue the recreational poker player mindset is no different from the blackjack or pokies player. Therefore it should be regulated as such.