With the help of some of you, I have been accumulating data on serial (crane recess) vs. service (butt) numbers for Colt 1917 revolvers. While I will continue to do so, I thought is was a good time to show you what I have so far, at least in a general sense. As you know, there is no absolute correspondence between the two numbers, but there is a general correlation. I was interested to know, among other things, how much difference there was between the correlation and the actual numbers. The data are shown below:
There is a fair amount of data displacement from the regression curve, and it seems more pronounced during the middle of production. The experimental regression curve itself is very close to the theoretical one, the small difference likely due to the finite data set (about 70 values). The maximum deviation is about ±12000, although much of the data is within ±5000 or less. The standard deviation of the absolute deviations would be a better measure of scatter, but I haven't gotten around to a better statistical analysis of the data yet.
Anyway, that where the research stands so far. Please keep the numbers coming.
Buck
There is a fair amount of data displacement from the regression curve, and it seems more pronounced during the middle of production. The experimental regression curve itself is very close to the theoretical one, the small difference likely due to the finite data set (about 70 values). The maximum deviation is about ±12000, although much of the data is within ±5000 or less. The standard deviation of the absolute deviations would be a better measure of scatter, but I haven't gotten around to a better statistical analysis of the data yet.
Anyway, that where the research stands so far. Please keep the numbers coming.
Buck