Prove Poker is a Skill Game?

[Sciolist]

I haven't read up on any of this, so stop me if it's been suggested before...

Look at every (say) PokerStars $215 tournament run. Only include players who have played ten times or more. If the game was random and you have a 15% chance at cashing, then we should know how many players will cash how many times, i.e, x% have cashed 10%, y% 20%, etc.

If poker is a skill game, then we should see that there's a higher percentage of people who cash more often than if it were random. Correspondingly, we should see a higher percentage of people who cash less frequently than if it were random.

Does this work? My knowledge of stats is not great. What sample size of players with > 10 tournaments entered do you need? Do you have to control for the size of the tournaments?

[75s]

Move to Probability forum. And yes, it's skill based on that reasoning.

[Sciolist]

[ QUOTE ]
Move to Probability forum. And yes, it's skill based on that reasoning.

[/ QUOTE ]
I thought it was more of a poker theory question than a probability question, but either is fine with me obviously. Maybe I should do this kind of analysis then, and sell my results to Gutshot or whoever :]

[Phone Booth]

[ QUOTE ]
[ QUOTE ]
Move to Probability forum. And yes, it's skill based on that reasoning.

[/ QUOTE ]
I thought it was more of a poker theory question than a probability question, but either is fine with me obviously. Maybe I should do this kind of analysis then, and sell my results to Gutshot or whoever :]

[/ QUOTE ]

I don't think there are very many people who doubt that poker is a skill game, just those who doubt that skill is the more dominant factor than chance.

[Sciolist]

[ QUOTE ]
I don't think there are very many people who doubt that poker is a skill game, just those who doubt that skill is the more dominant factor than chance.

[/ QUOTE ]
Perhaps you could quantify it? An average player has a 15% chance to cash. If you give 25% of the field a 30% chance to cash and 50% of the field a 7.5% chance, you don't get the same results anymore. You could presumably fiddle with those numbers until you have a rough approximation of the skill involved.

[AaronBrown]

It is easy to show statistically that moderately skilled players are better than average players. What's harder to show is that there is much scope for skill beyond the moderately good point. In other words, there are clear consistent winners in low-stakes on-line poker, but the number of repeat winners at the World Series of Poker is only slightly more than would be expected if the game were completely random.

That doesn't mean there is no difference between a pretty good player and a world champion, only that it's hard to demonstrate convincingly from the available data. Most tournaments don't keep consistent records of all entrants, and cash games keep no records at all.

[Sciolist]

[ QUOTE ]
It is easy to show statistically that moderately skilled players are better than average players. What's harder to show is that there is much scope for skill beyond the moderately good point. In other words, there are clear consistent winners in low-stakes on-line poker, but the number of repeat winners at the World Series of Poker is only slightly more than would be expected if the game were completely random.

That doesn't mean there is no difference between a pretty good player and a world champion, only that it's hard to demonstrate convincingly from the available data. Most tournaments don't keep consistent records of all entrants, and cash games keep no records at all.

[/ QUOTE ]
Were you familiar with the gutshot case? Do you know why they weren't able to produce data showing that the game is based on skill? Or was that not the point?

[doucy]

i remember sklansky making a couple posts pointing out that in poker, if you want to lose, you can do so very quickly. that alone suggests a non-trivial level of skill is involved.

[tarheeljks]

sciolist, i suggest you go read "fooled by randomness" by nassim taleb.

[jogsxyz]

Don't use the $215 tournament for proof. The $200 SnG is better. There will be many players with over 1000 plays in that game.