A while back, I was pounced on because I stated that measuring bullet velocity and it's Standard Deviation, was not meaningful with 3 and 5 shot groups.
I recommended 32 shot groups. (been a long time since Stat Class in College)....
So, I did some research with Dr. Ken Oehler's topics concerning SD.... I have a Model 35P.

So, the good Dr. said trying to determine SD from 5 shot groups is "in the same league as trying to determine the accuracy with a 5 shot group". "Likewise, a SD number should be considered only as an INDICATION of uniformity".
He and other statisticians consider the acceptable sample size to be "20 rounds".

Now, I am the first to admit that every time I put a rifle/scope together, if it shoots a 5 round ragged hole at 50 or 100 yds, I'll say 'done deal'.
However, that begs the question that what about the next day, when said load is all over the place?

I have a safe full of Savage rifles and most are Axis rifles with a few M12 in there. NONE of my rifles has ever fired a factory load, except .17 WSM....and we all know THAT story by now....

ron