User blog comment:DynamitMC/What the Drop Rates Actually Mean/@comment-26154973-20160220000428/@comment-24921408-20170305083057

Well, what I'm thinking is using the method to calculate the confidence intervals on the Poi statistics to give a band of 95% or 99% drop rates which can then be used to plug into the tables above that you've generated. Because some statistics on Poi may not have a high enough number of reports, that would allow people to calculate worst case scenarios and the number of runs needed accordingly.