A colleague asked me a few days ago:
“How many rounds would you say make up a legitimate ‘reliability test’ for a pistol?”
My response was 100 because that’s more than 99.9% of people will ever fire a pistol they buy. He was surprised about this answer because he thought it would be considerably more.
There’s a very detailed discussion about it in a post I wrote years ago. Most of the cognoscenti who were responded to the question then felt that 1000 rounds was the minimum desirable number. There were certain aspects of their analyses leading to that conclusion that I felt weren’t explored with enough depth.
https://thetacticalprofessor.net/2017/04/21/reliability/
Numerous justifications for 1000 round torture tests were presented to me by the cognoscenti. One of the mathematical analyses presented in the original Facebook discussion was that 5 malfunctions per 1000 meant more than one malfunction in a 17 round magazine (8.72%). My belief about that obtuse analysis remains the same.
“If I’m going to have at least one malfunction per magazine, I’ll just keep carrying a revolver.”
I’ll write a bit more about my latest revolver work for #wheelgunwednesday next week.
Hahaha. Alt Text autogenerated by Microsoft Word for the above picture:
“A board game with brown squares”
tacticalprofessor 
Good article, as was the link to your older reliability blog.
Many variables come into play as you’ve written about. Generally I’m in the 100-500 round count, as I check each magazine and some JHP rounds. I generally carry .45ACP, so testing JHP is more important for that caliber in a 1911 design gun. Some 1911 mags are fussy as well.
More heavily modified weapons need more rounds. I have an SKS set up to accept 20 round Tapco mags, and it took time and many rounds to test, tune, and retest each magazine. When some mags have failures at near full load, others at near empty load, and some at both ends, you need a test protocol for each failure mode. As the end result was a smooth running weapon, it was worth the test protocol development.
Nice. I have fired my Walther PPS more than 500 rounds during training with zero malfunctions. My H&K VP9 went 1000 rounds at Gunsite and Ridgeline, 2000 rounds with zero malfunctions. All with inexpensive ammo. Quality and skill beats fancy gear every time.
“One of the mathematical analyses presented in the original Facebook discussion was that 5 malfunctions per 1000 meant more than one malfunction in a 17 round magazine (8.72%).”
You need to link to this math because it isn’t correct if you assume independence.
The probability of having a 17 round mag run flawlessly with a reliability of 0.005 is 91.83% by survivor rule. So 8.17% a mags have at least one failure. Of those mags which have a reliability problem, almost all of them will be single failures assuming independence. Independence might be a bad assumption since a bad mag causing a cluster of feed failures is possible.
Still your statement indicates a 2+ round failure probability that is higher than the base 1+ failure probability. That doesn’t sound correct unless they did a very fancy analysis.
Good question, which I have answered in the subsequent post.
[…] Reliability Testing […]