Thread:Msmcurious/@comment-24577221-20170313161238/@comment-24577221-20170314162357

I'm not asking for details of your data collection process; I just want to know how large your data set is. On a number of occasions before, we've had people add highly-precise "statistics" that turned out to be completely bogus, based purely on "well, this is what I think the numbers are, I haven't actually kept any records". Or on data sets so small that the stats were essentially useless.

If you look at my blog, you can see that I and a group of "minions" have been collecting data towards getting the probabilities of breeding Ghazts and Entbrats, and the effects of Wishing Torches. Our data sets are on the order of 10,000 data points per set. This lets us determine the probabilities with a quantifiable relative precision of about &pm;1%. I'd be interested in seeing the precision / uncertainty of your fractions of coins / food / shards / diamonds.

Disclaimer: I'm a chemist, not a statistician.