So it boils down to we need enough hash power to expect a result of statistical significance +/- 5%.  In order to acheive that in a 14 day period there has to be a number we need to reach or pass in hash power.  I think we did fine with 1% or more of network speed.  I am sure someone can do the math, but just a guess is 1 to 2% of total should get us within a range that we can honestly say you make the same here or there over time.

I played poker for a living back in the day.  I know AA vs KJ is 80% fav, but I would not want to have to base my poker career of one hand.  I would never want to bet it all on that one hand.  Now if someone gives me those hands and says winner best of 100, I might bet it all on the AA.  If they gave me best of 10,000 it is a no brainer I will take the bet for sure.  So how many hands doe we need to play in 14 days (hash power) to expect a result of any statistical significance?