View Single Post
 
Old 05-07-2012, 05:49 PM
rjb1 rjb1 is offline
Member
 
Join Date: Nov 2010
Posts: 109
Likes: 0
Liked 70 Times in 36 Posts
Default

"The 357 with the 125 gr bullet has an excellent track record unequaled even today from all the accounts I've read."

That's true, but there is an aspect of those "accounts" that I noticed a number of years ago that may be worth a comment.
In one of the gun magazines a famous writer who specialized in the "caliber effectiveness" topic that has been going on for years (decades) presented a plot of "% of one-stop" kills for various calibers. One thing that struck me as an engineer was that first, the data was a bit sparse at the high end, and second, he used a polynomial fit to generate the equation that tried to "best-fit" the data. The tendency of a polynomial fit is that there is/was a fall-off of the curve at the upper end, due to the nature of the equation.
This tended to slightly overstate the effectiveness of the .357 compared to .44's and .45's.
I would have liked to see the same data curve-fitted with an exponential fit. This would have allowed the curve to gradually go up and would have a greater tendency to flatten out at the top end.
I always felt that this would be a better way of presenting the data and would not be quite as counterintuitive as to suggest that smaller is better when it come to bullets. (even accounting for velocity differences)
From personal experience, you can generate curve fits that appear at first glance to really fit the data well, but when you look at the output carefully you can have a terrible fit except at a few points.
Reply With Quote