Good Poker Players Aren't Lucky

http://www.bloombergview.com/articles/2014-05-19/good-poker-players-aren-t-lucky

by andrewljohnson • 12 years ago

88 124 12 years ago

bloombergview.com

sejje8• 12 years ago

Any game where you can play badly on purpose has skill involved.

Some games still involve more luck than skill--blackjack, for instance.

Other games involve more skill than luck, and poker certainly falls into that category.

Poker is a game of skill with an element of luck. It is not a game of chance.

dcsommer• 12 years ago

Poker is a game of skill with an element of luck. It is not a game of chance.

It depends on the level of the players. For beginners and casual players with less strategy, the game is much more about luck.

aetherson• 12 years ago

Sure, but the same is true of every skill game -- even, like, chess, without any explicit luck component. If both players are making essentially random legal moves, the game comes down to luck.

sejje8• 12 years ago

By that reasoning, basketball is a game of chance, depending on the level of the players. They're going to lob it up there and hope it goes in.

But their lob was a skill--one they don't control very well, but a skill nevertheless.

Betting in poker is the same thing, and is the only thing the player controls.

The way the cards come out is always chance, and the level of skill can't influence it.

pdpi• 12 years ago

Isn't that the whole point though? A game is a game of skill iff you can differentiate between beginners and experts.

sliverstorm• 12 years ago

Is it a game of chance? In the words of W.C. Fields, Not the way I play it, no

darasen• 12 years ago

There is a reason that the top players happen to very good at math.

sejje8• 12 years ago

This is not a rule. A lot of the players don't use much math--basic pot odds. They develop a feel for when they're making the right play, and openly admit to being "feel" players.

I do believe they would be very good at doing on-the-spot calculations in their head, they're often just not very trained in math.

Anyhow, surprisingly little in poker is math-dependent, particularly in normal ten-handed games. The cognitive load is mostly about factoring in thousands of variables, past hands, and potential hand combinations of your opponents.

zeppelinnn3• 12 years ago

Your statement is true about the "feel" players, but even those players will perform EV analysis away from the table because by estimating percentages of certain tendencies you can figure out what the right play is in tricky situations, and over the long run that ends up saving more money than anything else

sejje8• 12 years ago

@zep

Of course feel players are doing EV analysis, but that's 99% about estimating hand ranges properly. The math is certainly the most simplistic part of the analysis.

Sambdala3• 12 years ago

This isn't the case at the top end of the field.

There used to be prominent winning players who were unaware of the math behind the game, but the theory nowadays is they mostly just lucked into naturally playing in a style that happened to be better than the field during the period of time they played (i.e., years ago back when the game was less sophisticated).

Nowadays, those players are often the weak players who the players more skilled at theory can comfortably beat.

There's an eerie parallel to the early stock traders here.

whbk• 12 years ago

Well, now that we (nearly) have that strawman out of the way..I guess now we can shift gears and focus on Sheldon Adelson's 'online poker is so much more dangerous for compulsive gamblers and minors' doozy. Oh, and can't forget about Terrorist Hold 'Em!

http://www.nationaljournal.com/tech/what-do-terrorism-and-on...

ngokevin• 12 years ago

WSOP Main Event final tables are often used as an prominent exhibit that poker is not "tru gambools". It is the final 9-10 players of a tournament.

Dan Harrington made the final table in 1987 (152 entrants), won the final table in 1995 (273 entrants), and made back-to-back final tables in 2003-2004 (839 and 2576 entrants). Or in a lesser example, Johnny Chan won back-to-back in 87-88 (152 and 167 entrants).

The Cavs winning the #1 pick three times in the last five years, however. That is rigged.

fragsworth• 12 years ago

There is a lot of nebulous, vague law in the U.S. about games when you have the chance to win real money. Each state has its own self-inconsistent set of laws, usually made to preserve the monopolies of neighboring casinos. There is very little actual reasoning behind the laws - it's corruption, plain and simple.

This stifles innovation in real-money games. No company will risk spending any money on the design/development of a new real-money game when there is a tangible chance of going to jail and losing everything. This frustrates me a lot, because I enjoy real-money games (poker included), and I'd love to play new kinds of real-money games, but these freedoms don't exist.

prestonbriggs• 12 years ago

I wonder when they'll go after people for gambling in the the stock market? How would you defend them?

snorkel• 12 years ago

Good point. Instead of calling the game poker it should be dressed in Wall St jargon, and it'd be legitimate investing. The players can be called brokers, the cards are securities, a hand would be called a compound derivative, the ante is an investment, and the pot is just a leveraged mutual fund. The only problem is good poker players get a higher return than good brokers do.

mcphage• 12 years ago

Actually... building a stock market game around a poker re-theme sounds really freakin' cool.

jqm12• 12 years ago

I agree with you in principal but there is one difference...

Gambling involves the creation of risk to speculate on for no other reason than the thrill of speculation. Participants in financial markets (at least in theory) take positions in risk that already exists. In other words, they assume existing risk for others.

sharkweek• 12 years ago

This is a really interesting read -

I played a lot in college, but never really got "good" per se. We had a group of about 20 or so of us who played semi-regularly. There were two players who were both far superior to everyone else, but they didn't always win. Just like everyone else, they naturally suffered the turmoil of a bad beat here and there. But aggregate their winnings over the course of the two years we all played together, they came out far ahead of everyone else.

And I think that's probably the main point here. These two knew how to grind out the bad beats and come out on top over a long enough time frame.

I also have a good friend who crabs in Alaska (his dad and him are on the Deadliest catch together) - each season is a "gamble" with how long it will take to catch enough crab to meet quota, but it's not like they're putting money into a slot machine and hoping for the best. There's a system that leads to success. Sometimes they get "lucky" and their season ends quick, others it's more of a lengthened grind.

They seem awfully similar.

pitnips4• 12 years ago

And those both seem awfully similar to the stock market. If you're good at poker over the long-term, then you may be better off "gambling" than "investing." And hey, at least when you lose at poker, you may have still enjoyed it -- no one enjoys a stock market crash.

Oh, and don't get me started on naked futures/forward positions...

throwaway2837192• 12 years ago

Okay, I'll bite. Why is having a naked futures contract in the S&P500 (which costs around $100,000 for the e-mini) any more risky than owning $100,000 of stocks?

[0] http://www.cmegroup.com/trading/equity-index/us-index/e-mini...

hueving• 12 years ago

no one enjoys a stock market crash.
Oh, and don't get me started on naked futures/forward positions...

Some of the people that purchase these are plenty happy during a crash. What is it about time-based derivatives that you have a problem with in particular?

ssharp2• 12 years ago

Poker's problems are entirely cosmetic. It uses the same chips and cards and is played in the same building as blackjack, so it's always going to be associated with other forms of house-edged gambling, even though it isn't.

There are other places, like you noted, where people can gather, put money into a zero-sum game, and it's completely legal.

There are also lots of nuances in many of these state laws, where poker actually is legal under the right circumstances (like no house rake). I guess the fees associated trading options aren't considered a rake, so that's cool with the government.

mef• 12 years ago

His study of more than a billion hands of online Texas Hold’em found that 85.2 percent of the hands were decided without a show of cards. In other words, players’ betting decisions were of overwhelming importance in determining the outcome. Of the remaining 14.8 percent, almost half were won by a player who didn't hold the best hand but instead had induced the player with the best hand to fold before the showdown.

Isn't inducing the player with the best hand to fold before showdown the same as a hand decided without a show of cards? Or do they mean that 85.2% of hands were decided without seeing the flop? Both don't sound right.

kevhsu• 12 years ago

In the case of 'almost half' of 14.8 percent where the best hand folded, other people stayed in besides the winner and the best hand.

Fishkins• 12 years ago

I think what that must mean is: Players 1, 2, and 3 have the 1st, 2nd, and 3rd best hand, respectively. During betting, player 1 folds. Neither player 2 nor 3 fold and eventually show their hands. Thus player 2 wins not by having the best hand, but by getting the player with the best hand to fold AND winning the showdown.

sejje8• 12 years ago

I imagine that means that often, someone who would have won folded to a raise pre-flop, or on the flop, etc.

Later their flush (or whatever) came in, but they weren't in the hand any longer.

Two players showed down, but they were the player who won the pot and some third guy.

85.2% of hands were decided without players showing their own cards, certainly many of those hands had flops.

pimlottc• 12 years ago

Just because the player with the band hand folds before showdown doesn't preclude other players from staying in, forcing a show of hands.

thetrb• 12 years ago

Yes, but only a show of hands of those still in, not of the player who folded. So how would they be able to analyze that he had the better hand?

steven20122• 12 years ago

85.2% of hands were won without a showdown, ie. getting to the river and everyone remaining showing their hands. This sounds exactly right, and I play a lot of poker. If someone induces someone to fold their hand without showing their hand, it means they were able to out-bet the person, which is skill-based.

tom_jones• 12 years ago

The latest statistical analysis suggests that baseball involves more matters of chance than anyone would like to admit. Beyond home runs and strikeouts, it's mostly random. Luck and skill are all around us to varying unmeasurable degrees, including in a game of poker.

lasermike0262• 12 years ago

I have to disagree. Moneymaker, a complete nobody, winning the World Series of Poker.

sejje8• 12 years ago

Being a complete nobody is not reflective of one's skill.

That said, luck is the primary influence of outcome given a small enough time frame.

It's also possible to have your mistakes "fixed" by luck. Moneymaker was not better than Farha, I'll give you that.

lasermike0262• 12 years ago

Watching Moneymaker win and then a series of others unknowns win convinced me that poker is a game of luck.

Why are real rounders not winning?

ssharp2• 12 years ago

Poker tournaments, especially picking a single one, are not a great way to really illustrate that poker is a game of skill.

A book like Kill Everyone [1] puts forth a pretty reasonable tournament strategy where you try and accumulate a lot of chips early on and then start making risky moves in later rounds based your cards, position, and relative chip count.

Tournaments are played this way because the blinds and antes increase faster than the average chip stack increases. Once you get to the point where you have to make risky moves, you open up the luck angle a lot. The skill comes from knowing the probabilities and making plays with positive expected value over the long run, even if it means eliminating yourself from the tournament due to chance. But in the short-run, a relatively unskilled player could make similar decisions, get lucky and win the risky plays, and do very well.

[1]http://www.amazon.com/gp/product/1935396307/

candybar• 12 years ago

As a casual poker player, I find this line of reasoning specious. Sports betting, blackjack and even roulette with a biased wheel are all games of skill by this definition. Poker games are designed specifically to appeal to our gambling instinct so that better players can play the house while worse players gamble away their chips. It's true that from the perspective of the best players, it's not truly gambling, but they are not the ones gambling laws are written to protect. This isn't too different from a casino owner saying - I'm not gambling, I make money every day. You may not be gambling but your customers are. The same is true of poker - even if you're not gambling, your opponents are.

Edit: Will downvoters explain their reasoning? Btw, someone else said something about how poker is a game of luck at low skill level and game of chance at high skill level but that's not really how it works. Poker is a game of skill if you're playing against bad players because the game is consistently beatable with sufficient skills. The better and more sound your opponents become, not just relative to you but on an absolute basis, the more important the role of luck becomes.

lawncheer• 12 years ago

There is a big difference though, and it revolves around choice. A casino owner can not claim, that a roulette player can choose to get better. At some point, the "house" gets to a theoretically optimal edge (either by fixing the broken roulette wheel, or kicking out card counters). They will always have that mathematical edge.

In a poker game, any player can choose to get better, so much so you get to a a point where both players are playing game theory optimal. At that point, there is NO edge.

PS- support the PPA if you can!

Sambdala3• 12 years ago

There's a pretty fundamental difference between house-banked games where no player has a long term positive expecation and peer to peer skill games where better players will be able to gain an advantage by outplaying their opponents.

Roulette is a particularly bad example as there is no decision (well, there's one decision on American Roulette tables that is worse than every other) that changes your expectation on the amount wagered.

dnautics• 12 years ago

That's exactly the top commenter's point. The law doesn't hinge on the distinction between a house-banked game (like blackjack) and a house-hosted, peer-to-peer game (like poker). The distinction between skill/no-skill is less meaningful than the fact that you point out.

Nontheless, the law (and thus the article) is about the former distinction.

chilldream5• 12 years ago

"Game of skill" vs. "game of luck" is not a binary distinction, but the law pretends it is. I think it's definitely closer to the "skill" end of the continuum than the "luck" end though.

pavelrub• 12 years ago

I wanted to disagree at first, but actually I think you are making a valid point. There are obvious ways to treat poker as a game of skill, but this ignores the fact that gambling laws aren't just about some mathematical question of whether it's possible to think of a strategy that gets you positive EV (for poker, the answer is yes). Poker has certain properties which make it addictive to certain types of players who will most certainly lose, and the question is not whether it is possible to win through skill, but whether the less-skilled players should be protected from the psychologically-deceiving nature of the game. Even if skilled poker players might have the same numerical advantage as skilled basketball players, poker is different in its addictive nature and in the results it produces for the "losers" - which can be destructive.

jqm12• 12 years ago

Roulette is often held up as an example of "random chance" but this is not entirely true.

There are on occasions patterns in Roulette, especially with certain dealers. In addition, there is a little trick casinos sometimes employ. They loosen the legs on the table to the point it wobbles slightly. The dealer will then lean against the table and watch as the ball is about to drop. They can then pop the ball out of an undesired slot by a well placed shove of the hip. Just something to watch out for if you are gambler.

I firmly believe just like in computing, the entire concept of "random" depends mostly on ones frame of reference.

TylerE3• 12 years ago

That's a load of rubbish submitted completely without any sort of proof. That might have gone on in a back-alley speakeasy in the Prohibition era, but in any sort of reputable casino - not a chance. The casino has no need to cheat at roulette. They have a built in edge.

jqm12• 12 years ago

It absolutely is not a load of rubbish. I have personally observed it many times. And I have spent a LOT of time observing. I don't claim to have "proof" however I know what I have seen. I also didn't say "reputable" (a strange word in this context). Also, every/most Casino does not engage in this but some do.

Your claim "they wouldn't do this because they already have an edge" is just silly. Why wouldn't they take a chance to make more money? Especially since, as you say, "proof" is a bit hard. An accidental bump. That's all. No intentional cheating here.

I would like to re-iterate my main point. This is something to watch for if you are a roulette player. If the table wobbles when you push against it... walk away.

steven20122• 12 years ago

You would be a fool if you believe casinos don't try to go above and beyond their statistical edge to make money.

Case in point:

http://www.marketwatch.com/story/wynns-abnormal-lucky-streak...

I think we all know how a game with a statistical edge to the house has 9 straight quarters of larger-than-expected wins for the casino.

ronaldx5• 12 years ago

His study of more than a billion hands of online Texas Hold’em found that 85.2 percent of the hands were decided without a show of cards. In other words, players’ betting decisions were of overwhelming importance in determining the outcome.

It is worth noting that the other 14.8% of hands, where a showdown is reached, are typically the ones where (much) larger amounts of money are exchanged.

So, betting decisions might decide the outcome of a majority of hands, but it's unlikely this corresponds to a majority of the money flow.

Sambdala3• 12 years ago

This is a very complicated question, and while I understand why they would use that percentage figure, I think it confuses the issue if you look into it more deeply for exactly the reason you stated.

It is true that most of the money changes in hands where you get farther in the hand because pots grow geometrically as you continue on in the hand.

However, the importance of decisions becomes more important as well, and most of the edge a skilled player will have over a weaker player comes about from those larger pots based on decisions made later in the hand.

stygiansonic2• 12 years ago

I suppose, like the article mentions, it depends on the circumstances.

It does appear that serious poker players do better in the long run as they not only consider pot-odds when playing but also management of their bankroll to ensure long-term survival. Is there an element of chance? Of course, but it's not the only factor - few things are black and white.

But if someone choose to go all-in blind pre-flop then they are choosing to play the game as pure gambling and their version of the game would be pure chance.

But again, it's all about the circumstances. Poker is often compared to day trading in the markets and though it's not a perfect comparison, there are some analogies. For examples, options can and often are used in complex strategies to either reduce risk or enhance returns. But that doesn't mean they can't be used to speculate wildly - in these situations, the potential payoff profile begins to look something like a straight up bet.

chilldream5• 12 years ago

But if someone choose to go all-in blind pre-flop then they are choosing to play the game as pure gambling and their version of the game would be pure chance.

This isn't making the game pure chance; it's just cutting off that player from future decisions. The opponents still get to call or fold, which means the opponent is manipulating the odds that the all-in player will win a contested pot.

erikb• 12 years ago

Is the poker boom of ten years ago already forgotten? This fact shouldn't be surprising after tv, radio and internet telling everybody exactly this for years. What about the movie Rounders? Is it not watched in the US any more? I still got the DVD and watch it 4-5 times a year.

skizm• 12 years ago

I love poker. I'm not great at it but it is fun. Also it is a great metaphor for life. If you poke your head in on any given hand you might conclude that the game is all about luck. However if you play long enough good players emerge on top and bad players lose their money. Sometimes terrible players win big and sometimes good players lose everything.

Just like life: the harder you work, the luckier you get.

eddywebs• 12 years ago

In my experience luck does not exist.

flud434• 12 years ago

I don't even know how people can even argue it isn't a game of skill anymore. Profitable poker players are just little, mini-casinos. That's it. It's not that complex of an idea.

I play heads-up hyper-turbos, as well as other games, on PokerStars. I've been studying poker, reviewing my play, and grinding for years. I'm a mediocre reg and I can grind out consistent profits. Here's my graph of 18,106 $15 heads-up hyper-turbos on PokerStars.

http://i.imgur.com/AQ26oVQ.png

It really, truly, is not that hard to make money at poker if you're willing to invest even a small amount of time doing quality study and review.

ronaldx5• 12 years ago

IMHO, and with respect, poker results have a very high variance which may yet overwhelm your good results.

According to back of an envelope calculation, if there was no skill in poker, and you were just coin flipping, approximately 1 player for every 28 would come out with similar results.

This is a statistically significant result, I suppose, but not very. There are presumably many thousands of players who can present similar results, whether or not they have any skill. The jury's still out over whether you can continue this streak and claim +EV in the long term.

You should have some serious concern about this, since you are subject to self-selection bias in posting these results.

All-in-all, your results don't yet convince me that poker is a skill game, or at least not (yet) that you are skilled. Get back to me after the next 18,000 games, if you feel like it.

I confess to some dubious assumptions but I tend to believe that those assumptions averaged in your favour. This perhaps illustrates why poker is so good at hooking people.

flud434• 12 years ago

You are absolutely correct that poker results have high variance but you only need a sample of around 5000 matches or so before your results begin to closely depict your true winrate. I only showed an image of my $15 heads-up hyper-turbos on PokerStars. Below is a link to my graph for all heads-up hyper-turbos I've played over my life, 50k+. This graph shows my play from the $1.50 games all the way up to the $200 games. If you didn't trust in a 18k sample, maybe this 50k sample will help sway your opinion.

http://i.imgur.com/1uC5y0U.png

Note that I'm also only a mediocre regular in these games. There are MANY people who have far better results than I do. They average a 3-5% ROI which is much higher than mine.

Also note that the reason my graph dipped after the first 10k games or so was because I went from playing the $15s-$30s to playing the $100s/$200s for awhile. Due to the increased skill level I dropped back down. Here's another image showing my stats for each buy-in level.

http://i.imgur.com/D0lHx3H.png

If you still doubt how confident you can be in this sample, you might want to read the following article on confidence intervals on husng.com:

http://www.husng.com/content/given-my-results-what-should-my...

My 50k+ game sample means that there is a far lower than 0.1% chance that my results are due to the good end of variance.

And here's a graph of ALL my poker tournaments. This includes 9-handed multi-table tournaments, 6-handed SNGs, heads-up shootouts, everything.

http://i.imgur.com/6jyCXn0.png

Poker is 100% a game of skill but it really depends on how you approach it. Professional players or serious amateurs such as myself make money at the game simply because people approach it as if it's bingo.

If every recreational player was putting in time studying poker theory and reviewing their play, it would become much, much harder to make money playing. As such, I really don't mind if people want to continue thinking it's a luck based thing. It just means more profit for me since that perspective, oddly enough, doesn't stop people from dropping $$$ in the game, it merely stops them from thinking they can gain an edge by putting in time and effort. Recreational players wouldn't play if there wasn't a certain amount of luck...

EDIT: Also, it's important to note that every poker player has to pay for their education. Look at my last graph. I was an unprofitable player for the first 10-20 thousand games or so. That was when I was just playing micro/low stakes on PokerStars recreationally. Once I gained some knowledge, and decided to take it much more seriously, my profits skyrocketed.

It's far better to learn how to play poker at micro/low stakes online, where you will pay less for your poker education, than it is to start with mid/high stakes online or in your local casino. Either way, you'll be a losing player to start. Choose the more inexpensive option.

splonk3• 12 years ago

Without double checking your math, I'm pretty sure the significant error in your conclusion here is modeling the games he's playing as coin flipping with even payouts. The specific game he's playing is actually dropping a percentage each time and is more accurately modeled as "heads I win .98, tails I lose 1". (Technically he's betting $15 for a chance to win $14.69). With that in mind, many fewer than 1 in 28 players would end up with his results.

KVFinn2• 12 years ago

Is this skill-based rule is why every game in a casino is so bad? The article is about how poker is starting to be considered a game of skill and so not under the restrictions of gambling, and thus legal generally. Is there a similar law that is the inverse of this that restricts casinos to games of non-skill instead of actually good games?

I mean, you'd never sit down with friends and play roulette or blackjack or any casino game just for fun, right? It's be incredibly boring because you have zero choices and without the thrill of money being on the line there is no reason to play them. I wouldn't even put them in the 'game' category. You may as well flip a coin over and over. Even card counting BlackJack is just game of not getting kicked out by the casino -- counting cards itself is simply memorizing a table and has no choices either.

I've always been mystified people can spend hours in a casino playing the least interesting games I could imagine.

Poker is the lone exception. Poker with friends for trivial stakes still good. I can think of plenty of games with similar variance so beginners would always do reasonably well against experts and they'd still be be a million times more interesting than everything else in the casino.

You'd think with all the money in the gambling industry that some effort would be put towards making the games themselves intrinsically better.

Could you open up a 'casino' anywhere that instead used games that were obviously skill based because it woudn't be considered gambling? Poker is one of the most popular casino games and it's also the only game of skill Why not more?

tldr: Why are Casino games so bad?

knodi123• 12 years ago

You are insightful in pointing out that I have absolutely zero interest in playing roulette with my friends, but I really mourn the way babies have ruined my group of poker buddies.

But regardless of the luck factors, I think the real distinction is that in poker, each player is competing against the others. But in blackjack, roulette, etc, each player is competing against the house.

rtkwe• 12 years ago

A casino consisting of skill based games probably wouldn't function well, unless it was entirely player vs player with the house only providing game administration and entertainment/drinks/etc. Why? There are two main ways casino's make money: betting against players, i.e. any game where the house pays money beyond bets place by the players (blackjack, roulette, etc) and incidentals, entertainment, drinks, hotel rooms, and the like. For the first you want luck based games where the house has a statistic edge where dealer/game runner skill has no affect on the outcome of the game. These games are great for the casino because all they need is a trained warm body to run the game correctly and a flow of people coming in.

Skill based games on the other hand are generally player vs player, where the house doesn't win or lose any money, like poker. The casino doesn't want to get involved with betting against the players in these types of games because to win consistently they would need to staff skilled people to play the games and pay them enough that playing the other side of the table is less lucrative. That's likely to be quite expensive. So instead casinos make money off of skill based games through the second less lucrative stream, incidentals, providing guests with food, drink, lodging, and entertainment. These are more expensive and aren't as reliable as just churning through people in non-skilled games.

tldr: Why are they so 'bad'? Because luck based games make the best money.

chilldream5• 12 years ago

Gambling games are bad for the exact same reasons that grind-based are bad: people play them for the addictive reward mechanisms. If you're not affected by those mechanisms, you're usually left with garbage.

Variable rewards are also the reason why poker is one of the few games with a heavy skill component that can move large amounts of money directly between players; few people are willing to be a sucker. So losing players have to either get something else from the experience, or be convinced that they aren't losing players.

I said elsewhere in this thread that skill-based vs. luck based is a continuum. The most popular poker games tend to land on the sweet spot where "enough luck that bad players can think they're good" and "enough skill that you can actually be good" meet.

KVFinn2• 12 years ago

Gambling games are bad for the exact same reasons that grind-based are bad: people play them for the addictive reward mechanisms. If you're not affected by those mechanisms, you're usually left with garbage.

Wouldn't these games be even more appealing if had the same reward mechanisms but were also intrinsically interesting?

The most popular poker games tend to land on the sweet spot where "enough luck that bad players can think they're good" and "enough skill that you can actually be good" meet.

You'd think you'd see games along this continuum but poker is alone in the sweet spot. The other casino games aren't even worth considering.

yodsanklai• 12 years ago

I wonder how well computers can play poker and also, how different it is to play online rather than with actual people around a table.

I'm asking this because I once met a guy that was making a living playing online and it made me think. Would a computer program do better than a human in that context? It seems not to be the case, otherwise everybody would be using a program to replace them and humans player couldn't make profit anymore.

7273742• 12 years ago

People are working on it (of course, since it could be extremely lucrative). However, the problem is 'hard' because No Limit Hold 'Em is really a game of bluffing and not math. The best human players alive have mastered the skill of determining the cards their opponents hold based on their moves in the current hand, the current game, and previous games.

Online poker lacks the physical 'tells' that occur in a live game, like making an obvious fake reaction to some event. But, there are also some online 'tells' such as reaction speed.

I used to play a lot of online poker.

panorama3• 12 years ago

This is largely incorrect. NLHE can be more about "bluffing" because there's greater variability between ones chance of winning a hand and the potential bet size. But it still comes down to a mathematical decision, just that it's more difficult to accurately calculate the best play given the wide range of variables: your opponent's likely holdings, the bet size, the pot size, your hand.

In Fixed Limit, the bet size and pot size are all controlled and can be pre-determined, leaving fewer permutations to solve. It's easy to calculate that if a bet is a constant, then you can profitably call with holdings from XY and "higher" in a given situation, but it obviously gets harder when the bet size can vary greatly.

To answer OP, Fixed Limit is considered solved (or close to it? Haven't been following the news on this) from a mathematical and game theory perspective, and so AIs have been built to play game theory optimally. NLHE is still far, far from being solved, but there are bots that have been created to churn out small, albeit automatic, wages short-stacking (to mitigate bet-size variability) smaller stakes games where its tendencies aren't as openly exploited.

Poker is very much a mathematical game. And physical tells, reaction-time tells online are pretty insignificant in the longrun.

webreac• 12 years ago

I have seen a world champion of backgammon giving demonstration games with childs. He has won each time, but if the child had made a double at its last move, he would win. There is no random.

elwell• 12 years ago

Should startups be considered illegal gambling too?

programminggeek• 12 years ago

There is one reason to consider making poker illegal - it is addictive and has destroyed the lives and financial situations of many people who play it.

I've seen friends and family throw away thousands of dollars because of their addiction to poker. I'm not sure that poker is a net gain for society or not, but to say that poker is simply a game of skill that is no more harmful or beneficial than basketball or bowling or fishing or chess is disingenuous.

I admit that poker and other forms of gambling are not equally addictive or destructive to all people, just like WoW isn't equally addictive/destructive to all people who play. However, there are far more cases of gambling, specifically games of luck/skill like poker/blackjack, that result in broken lives and broken families due to a gambling addiction than in a traditional "game of skill".

Perhaps society as a whole is willing to believe the positives of poker outweighs the negatives, like it does with drinking alcohol or I guess now maybe smoking marajuana in some states.

We just shouldn't pretend that poker is all positives because it is statistically a game of skill. There is a lot more to the behavior around the game than that and only a fool would believe otherwise.

dllthomas• 12 years ago

People ruin themselves betting on basketball, too. Though I am not sure the degree to which it's legal...

One difference, of course, is that most of the people ruining themselves betting on other sports are spectators not participants.

Your principle point - that "game of skill" doesn't mean "useful" or "productive" or "worthwhile" - certainly very much stands.

vannevar• 12 years ago

"His study of more than a billion hands of online Texas Hold’em found that 85.2 percent of the hands were decided without a show of cards. In other words, players’ betting decisions were of overwhelming importance in determining the outcome." (emphasis added)

This is a tautology; the real question is whether there is significant discretion in making those decisions. The 85% statistic is meaningless without knowing how many of the winners within that segment simply folded weak hands and kept strong ones, in which case it would be clear that luck is the dominant factor. The 7% where the poorer hand won might merely represent the occasional exception to the rule.

dragonwriter• 12 years ago

Also, share of hands isn't share of stake. Sure, lots of hands of Hold'em are folded out -- often before the flop. But what share of the amount wagered is decided that way may be pretty different than what share of the hands are decided that way.

lazyant• 12 years ago

This is easy: there are professional poker players, there are no (long-term or non-cheating) professional roulette/slot machine/blackjack etc player.

omilu• 12 years ago

in boise you can bet on horse racing. Picking a winning horse out of a field of 12 uniformly looking horses and jocks, involves very little skill and boat load of luck. You can gamble money on power ball and scratch offs which is all luck. The ban on gambling has more to do with the inability to collect state taxes on home games etc...

Good Poker Players Aren't Lucky

Related Stories