Originally Posted by MApoker
For what purpose? If you estimate an hourly win rate of 1 BB/hr, and your hourly SD is 13 BB or less, your win rate is statistically significant here.
I'm not sure if there's a formal meaning of win rate is statistically significant
as used here. But I'll clarify what I think you're saying:
For a 500-hour "session", the SD of the session is equal to sqrt(hrs) * hourly SD, so 22.36 hrs * 13 BB/hr = 290 BB. This means that if we estimated that we were a 1 BB/hr winner, but after 500 hours we're dead even, then we're performing at -1.72 SDs. By dumb luck, we get a result that's 1.72 SDs or more away from the mean in either direction
around 8% of the time (p=.08).
Because p=.05 is the most commonly used definition of statistical significance in academic lit, it makes me wonder if you took the one-tailed value of .04 when the two-tailed value would be more appropriate.
Regardless, in layman's terms, if you play 500 hours at an SD of 13 BB/hr and run dead even, it's only 8% likely that you're "really a 1BB/hr winner"
in those games you played in. That doesn't mean you couldn't actually be beating the game for less than that of course.