The Std. Err. refers to "standard error", which is simply the standard deviation of an estimated parameter and is related to the more colloquial term "margin of error".
Simplifying a bit, we can say that for large data sets, we're 95% confident that the "true" population mean should within 1.96 standard errors of the observed sample mean.
The standard error of the "win frequency" of a random binomial variable is given by sqrt(p*(1-p)/N), where p, is the observed win frequency, and the N the number of trials in the sample.
So in your case of 52 games with a "success" rate of 13.46%, we have a standard error of sqrt(13.46%*(1-13.46%)/52) ≈ 4.733%.
This means that, as a first order approximation, the true push rate for games fitting your criteria is (with 95% confidence) 13.46%±1.96*4.733%, or in other words, between between 4.18% and 22.74%.
(Note that in the interest of illustrating the concept I've simplified the issue.) |