Quote:
Originally Posted by vinny34119
Maybe I didn't ask the right question as that I know there are many factors that will affect your win rate.
A better is
It possible that I am a winning player who is running badly enough to be down $327 (basically break even) over 700 hours?
Or if I am a winning player, is there no possible way that I could have broken even over such a long stretch?
So, the way confidence intervals work is that you can assign a percent chance that your true winrate lies within +/- Xbb/Hr. from your observed winrate.
I went to evplusplus.com and input your stats. They use BB/100 (2 * bb), so keep that in mind when converting back and forth.
You observed a winrate of -0.15 BB/100 over 20,000 hands. According to their model, you are 50% likely to have a true winrate between +/- 2BB/100, and 95% likely to have a true winrate between +/- 6BB/Hr. You are 99.99% likely to be between +/- 12BB/100.
Phrased another way: there is only a 50% chance that your winrate is outside +/- 2BB/100, and a 5% chance that your winrate is outside +/- 6BB/100. There is a .01% chance taht your winrate is outside +/- 12BB/100
Translating that into hourly rates: 2BB/100 is about $7/Hr live.
This means that you are VERY LIKELY to be a modest winner or a modest loser, and unlikely to be a large winner. Given some priors about the distribution of player skill, I would say that it is most likely that you are breakeven to a small loser in your game, but that is assuming I know nothing about you as a player.
Keep in mind that if you're playing in a game with brutal rake, being a breakeven player puts you way ahead of the curve.
Also keep in mind that most players overestimate their relative skill and underestimate the degree to which their mistakes are setting them back.
Last edited by Mismo; 07-24-2012 at 03:12 PM.