Statistically significant data for SD
Posted: Thu Apr 22, 2021 3:43 pm
So, I have received my share of flack for calculating SD (stand deviation) based off of small sample sizes like 5 data points. I needed to get some trigger pulls in today and decided to send a bunch of them over the chrono. Like 42 of them. 
Barrel length 3.7"; 9x19mm Luger; 124-gr X-Treme CPRN; 4.2-gr W-231; CCI-500 SPP; Cartridge overall length about 1.145"
43 F; 42% RH; 30.10" Hg
Average (mean) velocity = 989 fps
Standard deviation = 12.0
Extreme Spread = 59.4
Had I just used the first 5 or 10 data points the average velocity would've been off from this by 1 fps and the SD by 0.1. So yeah, I feel validated. Maybe I just got lucky, but ammo (and even reloading components) are too precious right now to run 40+ samples over the chrono to satisfy OCD (which I admit to having more than a twinge of).
Barrel length 3.7"; 9x19mm Luger; 124-gr X-Treme CPRN; 4.2-gr W-231; CCI-500 SPP; Cartridge overall length about 1.145"
43 F; 42% RH; 30.10" Hg
Average (mean) velocity = 989 fps
Standard deviation = 12.0
Extreme Spread = 59.4
Had I just used the first 5 or 10 data points the average velocity would've been off from this by 1 fps and the SD by 0.1. So yeah, I feel validated. Maybe I just got lucky, but ammo (and even reloading components) are too precious right now to run 40+ samples over the chrono to satisfy OCD (which I admit to having more than a twinge of).