Quote:
Originally Posted by bldeeney
This is very helpful, thank you. If I keep my current pace (45 hours per month), looks like it might take me a few years to get to 2000 hours.
I think 2,000 hours is excessive, depending on how "sure" sure is.
The general rule is this: with a winrate of WR and standard deviation of SD, if you've played N = 4*(SD/WR)^2 hours, you can be "sure" you're some kind of a winner, although you won't really know how big of a winner you are. You may be a marginal winner running good or a solid winner running bad, but you're some kind of a winner.
For someone with SD = 80, WR = 10, that's a mere 250 hours. For SD = 80, WR = 5, that's 1,000 hours.
If you want to know your exact winrate, it's fairly straightforward to calculate off of N. Take the fraction error you're willing to tolerate and square it, and divide N by that number. So if you want to know your exact winrate to +/- 10%, it's N/(0.1^2) = 100*N, if you want your exact winrate to +/- 50%, it's 4*N.
Chasing your exact winrate to high precision is a fool's errand unless you destroy a game for an extended period. For recreational players, playing to 1*N is probably sufficient.