I shot over 3,000 patterns looking for ultimate long range performance back in the 80s. Think how much more we would know if we had your 3000 patterns to go in a common database along with Dr. Jones's 2500.
I do not fault Dr. Jones' efforts, and I apologize if I sound high handed, but other than as a mathematical exercise, I didn't "get it" is the only way I can put it. Analyzing patterns with small shot at Sporting ranges struck me sort of like optimizing a Chevy Impala to circulate LeMans at 60 miles an hour, versus tweaking a Ferrari or Porsche to hit 200 on the straight and average well over a 100. Dr. AJ set out to investigate pattern performance with guns and loads he, himself, was likely to use. His initial work was not intended to answer "all things shotgun patter." The real "it" is the tool we can use as a common analysis for patterns and a demonstration of what it can resolve. Imagine more people adding usable data to a common database. Then, many/most/all of our questions about patterns can be answered.
I make no claims to being an accomplished statistician or mathematician. I did find that when comparing the effect of a variable I did get down to comparing 10 shot averages the get conclusive at the end of each trial. Interesting that you and the good Dr. came to the same conclusion about sample size. As a life-long test engineer I can promise you that drawing conclusions from test data without proper consideration for statistical validation is hazardous to your reputation.
I was quite satisfied with my results, but I am still rather mystified at how George Diggweed manages to do a lot of what he does, with what he uses. That guy just amazes me. Mr. Diggweed knows that a small number of pellets will have no sideways velocity in the pattern and will remain very near the ballistic center line of the pattern all the way to the ground. He is making "rifle shots" and could probably hit targets with a .22LR.
DDA
