#11
|
|||
|
|||
Re: September Low-Content Thread
Lets consider the result of each 100 hands to be a measurement of your winrate, with a mean x=winrate and sigmax^2=variance. Each measurement is independent.
We want to know the mean and variance of a combination of several samples (X1, X2, X3...) The mean of all samples is X = sum(x) = x1 + x2 + x3 ... + xn The variance of all samples is sigmaX^2 = sum(sigmax)) = sigx1^2 + sigx2^2 + sigx3^2 ... + sigxn^2 In the special case where all xn and sigxn are the same, the formulas reduce to: X = x*n sigmaX^2 = sigx^2*n Total mean is X and total standard deviation is sigmaX. For a normal distribution it is now simple to determine the chance that a random measurement from this distribution will be at a certain distance from the mean. Refer to any source on the normal distribution such as: http://en.wikipedia.org/wiki/Normal_distribution or the much simpler adhoc rule which I applied here: http://en.wikipedia.org/wiki/Empirical_rule |
|
|