Thanks, Miller, I understand that, and have been aware of the "hot core" of nearly all shotgun patterns for as long as I have been patterning, which has been a long time. I understand the inability to calculate an exact comparison, because of this and because of the way shot are distributed, as Don explained. But, this question still nags me. If this is the case, and we accept that it is, can we not then go forward with a comparison that accepts the error, and the fact that none of this is exact, and have something useful anyway?
When I study published yield comparisons of corn, cotton or peanuts for the previous crop year in order to determine which variety, out of dozens, I will plant this spring, I know that these same varieties could be planted and compared again and the results would not be exactly the same. The top yielder might fall to second, or third place. But, when I look at those yield comparisons for the past three consecutive years I see one or two emerge as the clear leaders, I know I have some information that is useful to me. At the bottom of the page of corn yields there will be info such as this, showing yields at various locations:
Yields ----------297.3, 312.2, 314.2, 265.6
LSD at 10% level ----9.4, 16.1, 14.6, 18.4
Std. Err. of Entry Mean 4.0, 6.8, 6.2, 7.8
I'm sure these two lower lines of numbers mean much more to some of you than they do to me. I understand, by this, that these yields can be given a certain level of credibility. If this is acceptable to me I put my "faith" in these results.
Is there a difference here that I do not see? I certainly am no student of "probability". There may be some of you who are, and understand these things much, much better than I. If so, please comment.
SRH
