How many poker tournaments must I track to obtain a significant sample

[HokieBalla34]

How many poker tournaments must I track to obtain a "significant" sample

Good afternoon. My question pertains to single table poker tournaments (aka Sit 'N Go's) and I hope some of the bright minds of this forum will find the question at least interesting.

The type of Sit 'N Go I play is a 6-handed turbo, in which 6 players start and the blinds escalate every 5 minutes until only one person has all of the chips. The first and second place players win 70% & 30% (respectively) of the prize pool. I've played in roughly 300 of these SNGs since the beginning of 2008 and I've tracked my results in each one to create statistics for things like In the Money %, Return on Investment, and Expected Value per tournament.

I was wondering if 300 was a good sample size for this type of tournament and what "confidence" I could place in my data. Some of the posters on poker forums say that you have to have at least 1000 SNGs under your belt to have a high confidence, but this number seems very arbitrary and I have not seen any math to back it up.

So my question is: How do I determine the amount of samples needed to obtain a high confidence in my data? It has been a long time since I've taken a stats class, but I seem to remember a model for calculating a confidence interval based on samples, but the equations I can find online center around samples from a finite number (e.g. x% of a population of a country, which is known).

Thank you very much for your time and any responses.

 

[EnumaElish]

http://en.wikipedia.org/wiki/Sample_size#Required_sample_sizes_for_hypothesis_tests

 

[CRGreathouse]

It depends on what you're trying to show. If you were trying to show a 0.02% deviation from an expected chance, you might need many samples. If you were trying to show a larger effect, with a wide confidence interval, then you might need just a few dozen games.