horseman308
New member
So I read a comment about number of shots per group to gauge accuracy.
From that I ask myself (and y'all) where we got the idea that a given 10-shot, or 5-shot, or even 3-shot group would give a good idea of a rifle's accuracy. What I mean is, why would we start to believe relatively few rounds would be a valid indicator?
In grad school, when I had to learn statistics and how to conduct research, one of the basic things was that you need plenty of observations of a thing before you can say that, on average, your conclusions are valid. Statistically, an average based on a minimum of 30 observations gives tolerable validity. So based on that notion, backed up by Bart's comment, I'd want at least three 10-shot groups, or six 5-shot groups, etc. (or better yet, 30 groups of 10 shots), taking the average group size, to say with real confidence that my rifle has a certain amount of accuracy.
I figure most would agree that more shots is better than fewer when testing accuracy. Obviously, that many rounds gets expensive and time consuming, especially if we're talking about match-grade ammo. Still, is there any other reason that so many people rely on a relatively small number of rounds? Just curious.
. . . your 3-shot group's got no more than 20% confidence of being what all shots fired will do. But it's very good for a factory rifle. Shoot another one with 10 shots and that'll be about 60% on the confidence scale. - Bart B.
From that I ask myself (and y'all) where we got the idea that a given 10-shot, or 5-shot, or even 3-shot group would give a good idea of a rifle's accuracy. What I mean is, why would we start to believe relatively few rounds would be a valid indicator?
In grad school, when I had to learn statistics and how to conduct research, one of the basic things was that you need plenty of observations of a thing before you can say that, on average, your conclusions are valid. Statistically, an average based on a minimum of 30 observations gives tolerable validity. So based on that notion, backed up by Bart's comment, I'd want at least three 10-shot groups, or six 5-shot groups, etc. (or better yet, 30 groups of 10 shots), taking the average group size, to say with real confidence that my rifle has a certain amount of accuracy.
I figure most would agree that more shots is better than fewer when testing accuracy. Obviously, that many rounds gets expensive and time consuming, especially if we're talking about match-grade ammo. Still, is there any other reason that so many people rely on a relatively small number of rounds? Just curious.