A recent
Math binge has brought me back to predicting confidence levels in the montecarlo method approach to assessing what the bias is in a tournament. Previously I had kept a running total of averages as 1000's of randomly seeded tournaments rolled by, but it looks like I can also keep a
running standard deviation. Once I have both the sample standard deviation and the number of tournaments sampled, I should be able to calculate the
95% confidence interval for the trial. Hopefully I can blow the dust off the old code and get it running again.
No comments:
Post a Comment