And also from Rocketman…
“By the way, most of the claims of magic patterns are rooted in the wide variability of patterns to pattern. Two or three are useful for point of aim/point of impact. Not so for statistical understanding of true pattern performance. It takes ten (or more). Jones's automation of pattern analysis makes this reasonable to actually do; shoot pattern, electronically/digitally photo pattern, feed photo into computer program, repeat nine more times, run analysis program based on ten patterns, and consider the correlation of real life pattern to desired pattern.
Pellets are distributed according to a Rayleigh distribution (there are many statistical distributions based on differing conditions). Rayleigh is based on no/zero probability of any single pellet hitting dead center of the pattern. The normal distribution/bell curve is the most usually assumed for patterns. The variability from pattern to pattern makes the difference between Rayleigh and normal relatively moot. Bear in mind that theoretical/statistical distributions are predictions of future events. In the case of the pattern, the distribution predicts the probability of distance from center pellets will land. Unfortunately the pellets are unaware of said prediction. Fortunately they follow the prediction closely enough for it to be useful.
I have read/heard that "old time" barrel makers made adjustments to choke profile based on single patterns and including said pattern as proof of the barrel's choke. Now, that sounds an awful lot to me like shooting patterns until you get the one you want.”
