Yes, I recall you mentioned about the SD, which is importan to have an idea of the distribution’s spread for Montecarlo or similar methods. I would guess it would have to be relatively small to capture 2/3 of the data that has a maximum range of about 12 while assuming a normal distribution. I am guessing/speculating 2 to 3 at most based on the differences you described, or maybe a bit lower, wouldn’t you?
In regards to the “unknown unknowns”, there is not too much we can do but wait.
Thank you
News 
Market Data 
Discover 
