Pokernomics

[Rudbaeck]

http://www.pokernomics.com

Levitt is attacking poker with the brute force approach. It will be interesting to see what he comes up with, and if you haven't read Frekonomics it's a nice chance to get his book signed and free. (You can learn the effect of Roe v Wade on armed crime at the turn of the millenium!)

Too bad he is writing this book, I'd been nourishing the same idea myself. Who wouldn't want to be the next Ed Miller after all? Oh well, now it's actually going to get done instead of dreamt about. [img]/images/graemlins/smile.gif[/img]

[King Yao]

Sounds very interesting. He can probably pull it off because people are less afraid of their stats being used against them or for profit.

Have you tried it? I'm wondering if they give quality, accurate and useful feedback.

[Rudbaeck]

[ QUOTE ]
Have you tried it? I'm wondering if they give quality, accurate and useful feedback.

[/ QUOTE ]

Sent in a database of some 35k hands today, so haven't had a chance to get the analysis yet. I'll get back on it.

[King Yao]

Levitt's paper on the NFL gambling markets was not good. He used poor data to come to incorrect conclusions. So I am not convinced he can do a good job analyzing poker data.

[Ed Miller]

[ QUOTE ]
Levitt's paper on the NFL gambling markets was not good. He used poor data to come to incorrect conclusions. So I am not convinced he can do a good job analyzing poker data.

[/ QUOTE ]

I'm even less than "not convinced." Part of my evidence for this conclusion is that he asks for only 10k or more hands. Frankly, I don't think he has a clue about what he's getting into.

[Jordan Olsommer]

[ QUOTE ]
[ QUOTE ]
Levitt's paper on the NFL gambling markets was not good. He used poor data to come to incorrect conclusions. So I am not convinced he can do a good job analyzing poker data.

[/ QUOTE ]

I'm even less than "not convinced." Part of my evidence for this conclusion is that he asks for only 10k or more hands. Frankly, I don't think he has a clue about what he's getting into.

[/ QUOTE ]

wouldn't this only apply to the "free analysis" you get emailed and not the main point of the project? For example, I have fewer than 10k hands logged at pokerroom, yet I don't consider their excellent EV stats page to be inaccurate because of it.

[Ed Miller]

[ QUOTE ]
wouldn't this only apply to the "free analysis" you get emailed and not the main point of the project? For example, I have fewer than 10k hands logged at pokerroom, yet I don't consider their excellent EV stats page to be inaccurate because of it.

[/ QUOTE ]

Sure, but the fact that he claims to be able to give you a meaningful analysis about "which hands you play best" after only 10k hands tells me either:

1. He's unwilling to pay people for data, so he's going to offer something that sounds good, but that will, in reality, have very little value instead, or

2. He doesn't know what he's getting into.

[Jordan Olsommer]

[ QUOTE ]


Sure, but the fact that he claims to be able to give you a meaningful analysis about "which hands you play best" after only 10k hands tells me either:

1. He's unwilling to pay people for data, so he's going to offer something that sounds good, but that will, in reality, have very little value instead, or

2. He doesn't know what he's getting into.

[/ QUOTE ]

But...but....free t-shirt! [img]/images/graemlins/smile.gif[/img]

[Ed Miller]

For instance, think about how many hands you would need to figure out LOOKING FROM A STATISTICAL PERSPECTIVE ONLY (i.e., making zero judgements about which plays are good or not based on "human" skills like hand reading) if you pay off too often on the river. It's a very large number. 10k doesn't even begin to scratch the surface, yet that is quite a funadmental thing someone might want to know about their play.

[Jordan Olsommer]

[ QUOTE ]
For instance, think about how many hands you would need to figure out LOOKING FROM A STATISTICAL PERSPECTIVE ONLY (i.e., making zero judgements about which plays are good or not based on "human" skills like hand reading) if you pay off too often on the river. It's a very large number. 10k doesn't even begin to scratch the surface, yet that is quite a funadmental thing someone might want to know about their play.

[/ QUOTE ]

Upon rereading the FAQ, apparently the "personal analysis" bit isn't slated as just a gimmick, but part of the overall project. From the FAQ:

"5. How is your free analysis going to differ from statistics given by Poker Tracker or other software?

The advantage of a pokernomics analysis over statistics reported by Poker Tracker is that our analysis incorporates data about how other people play. A naive analysis would look at a Poker Tracker statistic that says "I win 57 % of the time after seeing the flop with hand Y" and conclude that playing more of those hands will be successful. A Pokernomics analysis will test how other players with a similar style to yours who had that same hand but played it greater percentage of time (controlling for relevant factors) actually did. By comparing the play of many different people over the course of hundreds of thousands of hands our analysis is far richer.

Using advanced statistics, the pokernomics team will also be able to control for table conditions, position, and other relevant factors in order to generate more meaningful statistics. "

Although I don't know if that would even make it valuable. To exaggerate for the purpose of clarity, say I were able to send this guy just one hand history of where I played AA in the cutoff in a NL game and only stole the blinds - what could his analysis possibly tell me about my own game, even if it were based on a gazillion hands where other people got AA in the cutoff?

Nothing more than the pokerroom EV-by-hand and position thingamabob could tell me, I'd imagine.