Writing a Statistics term paper on poker - ideas?

[RiverFenix]

I have to write a term paper for a stats class Im taking. We can do it on anything I want so I figure ill do it on poker to keep myself interested. The question then arises on what can I do it on that requires a large quantity of data and will have enough content to write 10+ pages on.

My idea was to show that the Party network is indeed the loosest/fishiest network that there is. I figure ill datamine a different amount of other networks and then just crunch numbers. I obviously have choices in limit vs NL, stakes, # players, etc and how to measure exactly what fishiest means but Im not going to worry about that yet and just try and find a good topic for a poker related paper.

[UATrewqaz]

If that is your theory then here are a few stats to test

1. % of players seeing the flop, take 10K hands at PP and 10K hands at PS, same limit (due a few diffenent limits). The PP % should be higher beyond the range of standard deviation if PP is indeed fisher.

2. avg number of players seeing showdown, again PP should be higher

3. % of hands that actually go to showdown, once again PP should be higher

[mmbt0ne]

Prove that they aren't rigged. You'll have to learn so SQL code though to parse all the data.

Bonus points for making bad players want to play more, and not teaching them how to be good players.

[RiverFenix]

bump

[uuDevil]

How about quantifying sources of variance? Like for instance how much variance is associated with playing various hands. Or with different positions (blinds, early, middle, late). Or looking at the correlation between win rate and variance. Or analyzing the distribution of poker results. Or looking for indicators of running bad as opposed to bad play. Lots of ideas along these lines might be interesting.

[RiverFenix]

last bump

[laja]

I applied some linear algrebra to bankroll requirements

http://www.pokermentor.net/hpm_bankroll.html

the results that I got were what I had expected, that it often profitable to push yourself up limits without a proper 300 BB bankroll and that if you lose some you just have to fall limits, nothing brilliant but I wanted to see if it was mathematically correct.

I'd let you plagarize it if you want lol