January 2016 Aligulac balance report - Page 5

On February 10 2016 07:07 todespolka wrote:
I am myself a mathmatician and i would like to see how you calculated the p-values. If it doesnt take much time, could you please post the formulas of your binomial tests. I would like to reexamine it and check its validity. Thank you.

On February 10 2016 07:18 Daimai wrote:


If you're a mathematician, you should be able to do this yourself. He just used a calculator linked in the OP and set 0.5 as the probability to win. The calculator uses a binomial distribution and finds the probability that we will have 43% winrate as protoss assuming the balance is perfect (meaning p = 0.5). This comes out to be really low, indicating the matchup is at least somehow skewed against protoss.

On February 10 2016 13:01 Cascade wrote:

Yes, was about to write exactly this, including the first sentence. thanks.

To be fair, the new TL style makes it very hard to spot links.