stability of winrates over given number of hands

[Abbaddabba 0]

Though im familiar with the way one would go about calculating precise confidence intervals given the sample standard deviation, sample size and sample win rate, i thought it was interesting as to how fast the BB/100 figure stabilized. (might now show because.. here's the link http://img291.imageshack.us/img291/7149/bb1009dq.gif )After about 5k hands, my BB/100 winrate for .50/1 stabilized at roughly 4.0BB/100 - which is where it would remain (give or take a VERY small amount) all the way through 12k hands.After 20k and 30k hands im going to check up on it again, but it seems pretty stable given the data up to this point. Maybe that's on account of the fact that .50/1 is a low variance, where it's fairly loose and passive. My preliminary figures at 1/2 show similar results (after 6k hands).