Accuracy of a sample size

[hmkpoker]

(New to the forum, I apologize if this has been brought up already)

Let's say I am a winning poker player with a decent edge in my 5-10 full ring online poker game; I guess my edge is somewhere around 2BB/100.

Given a sample size of 10k hands, to what degree of reliability can my BB/100 edge be determined within a parameter of +/- .5 BB/100? How about 20k? 40k? What if I want to measure reliability within .25BB/100? (and so on and so forth with tons of variables)

If anyone could offer any statistical information about this, that would be awesome. Thanks! [img]/images/graemlins/smile.gif[/img]