Some light reading for those with the inclination
David J. Compton, “An Experimental and Theoretical Investigation of Shot Cloud Ballistics”, 1996http://discovery.ucl.ac.uk/1382490/1/396689.pdf
150 different 10 cartridge batches of loads were tested, and more than 2000 patterns analyzed.These findings reinforce the attitude that many aspects of shotgun ballistics, especially patterns, have no satisfactory theory to predict or explain the effects of the internal ballistics of a gun on the downrange behaviour of a shot cloud.
ie. it is magic
Summary. More detail starting p. 232
p. 155 “Three-Dimensional Representation”
The general shape of the shot cloud revealed that the pellets arriving first (leading edge) are located in the centre of the pattern, and the trailing edge pellets appear at the outer regions. The deformed pellets, collected at the outer parts of the pattern, travel at the trailing edge shot cloud. The well-formed pellets, associated with the pattern centre, are to be found at the leading edge of the shot cloud. With the greater associated deformation on pellets at the rear of the load, caused by the pressure in the barrel compressing them into the pellets above, a similar relationship between the pellets deformation and location in the shot cloud is seen in both experiments.
p. 160 The longitudinal distribution of pellets in the shot cloud at ranges between 20-50m was shown, via shot cloud length, to be unaffected when the internal ballistics of the gun, such as choke, were altered. However, from high speed photography it is known that the initial distribution of pellets is affected by the internal ballistics. Therefore it is assumed that the in-flight effects of the pellet become the more dominant factor, thus masking the internal ballistics effect, at ranges greater than 20m.
From the analysis of the shot cloud profiles it was established that the longitudinal pellet distribution is best described as a Rayleigh distribution.
Analysing the lateral distribution of pellets in the shot cloud it was established that there are two independent distributions, that is the horizontal (x) and vertical (y) pellet distributions. These two distributions were shown to be best described as Gaussian (“bell curve”) distributions.
Neil Winston made the point over and over that since pattern distribution was Gaussian, "patchiness" was random and the primarily determinant of pattern efficiency/quality, with the same choke constriction at the same distance, was simply the number of (high antimony) pellets in the shell.
Previous discussion herehttps://www.doublegunshop.com/forums/ubbthreads.php?ubb=showflat&Number=496511