There is highly likely to be a difference in the style of play between the two levels, a difference that OP's "value-oriented" playing style is ill-suited to adapt to.
But do notice that the difference between the two measured win rates is huge: 20bb/100. No, those sample sizes are not enough to give an indication of one's True Winrate (if such a thing exists, which I have reason to question). Jeebus, it's like no one in this thread has ever heard of Reverend Bayes. Man, part of me wants to actually go for that rather than keep trying to move up, lol. To get a margin of error of 1bb/100, it'll take 2 Million hands. At my StdDev, if I did the math right, it'll take about 500k hands to get my 95% confidence interval to have a margin of error of 2bb/100. But if your variance is like mine, and your true winrate is ~10bb/100, then 20k hands could very well be enough to give you solid evidence as to whether you are actually a winner.īut yeah, getting a precise estimate on your actual winrate takes much much more. Obviously every player is different and I honestly haven't looked at all into to how much everyone's variance varies. But what you can say based on this is that there is evidence at the 95% confidence level that I am truly a winning player, since the lower bound of the confidence interval is above 0. So, obviously there's some chance that my true winrate is outside of these bounds. The confidence interval bounds are at the 95% confidence level. This is, roughly, my most recent 20k hands at $50NL (I filtered by date I couldn't figure out a way to get exactly 20k): Just thought I'd add some data to the discussion. Your most likely range of bb/100 could be -4bb/100 to 24bb/100.
Realistically a player who is beating the game for say, 10bb/100 over a 20k hand sample (this can change for example, pending avg stack depths of tables, etc).
0 kommentar(er)
