Sustained profit

[evain]

Discussion during a hand last week:

What percentage of poker players win over the long term, after the rake etc. We were thinking maybe 10%. Any thoughts?

[pstripling]

If all we are counting are the people who play regularly (at least 6 times a year) then my guess would be 5% or less. I'm only counting the people who play in B&M and online...not the home game crowd. It is very hard to keep the discipline to play winning poker over time. The vast majority of people cannot do it.

[Cosimo]

This has been discussed before on the boards; it seems to pop up with some regularity. Try searching the archives for a bit. Here's one thread:

http://forumserver.twoplustwo.com/showflat.php?Board=smallholdem&Number=810542

[sammy_g]

Was the question really answered in that thread, though? Sure there are winning players, but how many? I see the 10% figure often, but I don't know where it comes from.

It would be interesting to pool together a bunch of PokerTracker databases and see what the real figure is.

[adamstewart]

[ QUOTE ]
It would be interesting to pool together a bunch of PokerTracker databases and see what the real figure is.

[/ QUOTE ]

Something tells me this would be an EXTREMELY biased sample... and that the percentage of winning players derived from such a sample would be significantly over-estimated.

Unless, of course, you are talking about pooling together not just one's own stats, but those of *all* the players in everyone's databases.

Three problems still arise though:
(i) likely sample size - the number of hands for each player - is way too small.
(ii) again, the sample would be biased to only online players
(iii) the issue of sharing databases is a touchy subject and not many would go for it - present company included.

[evain]

Thanks Cosimo

Great link. I'm sort of new here so it's appreciated.

[sammy_g]

[ QUOTE ]
Unless, of course, you are talking about pooling together not just one's own stats, but those of *all* the players in everyone's databases.

[/ QUOTE ]
This was what I was thinking.

[ QUOTE ]
Three problems still arise though:
(i) likely sample size - the number of hands for each player - is way too small.

[/ QUOTE ]
Right. You'd have to have enough data for, say, 10K hands on many players. Even this would be a biased sample, though, since I'm sure many losing players quit before the play this much.

[ QUOTE ]
(ii) again, the sample would be biased to only online players

[/ QUOTE ]
This is what I'm interested in. The figure is probably higher for online players since the rake isn't as brutal online as in B&M cardrooms.

[ QUOTE ]
(iii) the issue of sharing databases is a touchy subject and not many would go for it - present company included.

[/ QUOTE ]
True. I wasn't volunteering to do this or anything. I just said it would be interesting. /images/graemlins/wink.gif