Setting the shoulders back for the first time.

[Bart B.]

My original:

Not too long ago, they (stoolshooters) learned that full length sizing their fired cases made two things happen; case necks were better centered on case shoulders and their largest groups got smaller. Smallest ones stayed about the same size.

If the fact that the larger groups got smaller isn't a good indicator of real accuracy, I don't know what is. If you're gonna quote me, put everything in it so it won't be taken out of context.

There's a limit as to how small a group can be. There's no limit to how big they can be.

There's no easy way to prove what makes the tiniest groups the size they are. It's one of two things. One's when everything is perfect; all variables are at zero. The other's when all those variables are at different amounts but cancel each other out. Tiniest groups happen more often when the largest ones are the smallest. Largest groups happen when some of those variables add up in different directions maximizing the distance bullets strike from group center.

That said, if one thinks best accuracy is the smallest groups fired, shoot a few hundred of them with any rifle and one will be very, very tiny. Doesn't matter if the largest one's 10 inches and they average around 5 inches. Lake City's 7.62 NATO National lots of match ammo's shot sub 1 inch 5-shot groups at 600 yards testing it. The extreme spread of all 54 of them in a 270-shot composite is about 10 inches.

 

[Snyper]

Snyder, forgive, there was absolutely nothing I said that applied to one of your responses. I did not quote you.

Actually, you did, in Post # 70:

Quote:
To whom in may concern:
Quote:
In other words, if the loads aren't being fired from a machine rest, no one will ever know the difference.

 

[Snyper]

If the fact that the larger groups got smaller isn't a good indicator of real accuracy, I don't know what is. If you're gonna quote me, put everything in it so it won't be taken out of context.

If a gun has already fired smaller groups, then your comment is meaningless.,
Unless the smallest groups shrink, accuracy has not changed

If you're gonna quote me, put everything in it so it won't be taken out of context.

Nothing was taken out of context.

It's what you said

If the object isn't to make the smallest groups smaller, then there is no "context" to most of what's been posted in regards to "better accuracy".

 

[Unclenick]

Actually, it's the average group getting smaller that matters. Individual groups can be randomly larger or smaller than the rest. Once the average is smaller, the chance of the largest groups being as big as before will decrease, and the chance of the smallest groups being smaller than before increases. Both happen simultaneously.

What often confuses the situation is that while larger groups are better samples for determining accuracy, a larger sample gives outliers more opportunities to occur, so your average groups are larger when sample size is increased. It's why you do best to use a group measuring method that discounts the influence of outliers on the result in some measure. The military likes mean radius, except for artillery, for which they come up with circular error probable which is more useful for targeting. I find radial standard deviation as stable as CEP for small arms ammunition at normal target ranges. But any of those methods are better than simple diameter because the outlier influence is considerably discounted by them.

So, pick a sample size and stick with it. Keep a running average of the result to test if improvements you are making in your loading practices are causing the average to trend smaller.

 

[Bart B.]

What's a more accurate system (rifle plus ammo plus shooter)?

System A; ten 5-shot groups averaging 5 inches with largest at 8 inches and smallest at 1 inch. 50 shot composite's 12 inches.

System B; ten 5-shot groups averaging 5 inches with largest at 7 inches and smallest at 2 inches. 50 shot composite's 9 inches.

Talking with a Lake City arsenal ballistics man, he told me the reason they use mean radius of small arms test group shot holes to determine the average miss distance a bullet will strike from group center. Group center's assumed to be at the point of aim at testing range. They used 600 yards for 30 caliber ammo.

 

[Unclenick]

Bart,

That's a mathematically true fact about how you can fool yourself with averaging. As a math professor of mine liked to say, the trick with averaging is knowing when you can use it and when you can't. However, the underlying reason averaging doesn't track accuracy in your example is that the two sets of composite targets represent non-normal distributions, rather than the usual bivariate normal distribution a machinerest test of ammo produces. For that normal target distribution, 19 out of 20 five-shot groups are expected to be within a diameter ratio of 2.35:1 and both the 8" and 7" groups should fall within that range an average of 19 out of 20 times. That makes the diameter averages unequal at least 19 out of 20 times.

You could force the averages to be equal, though. You could, say, take the lot of ammunition that produces the smaller largest groups and cull out all the lowest runout cartridges to set them aside for a match, for example, and that would grow the smallest group without growing the larger group by very much. You just have to adjust your trimming carefully. But it would take that sort of intervention to create that situation where averaging doesn't track intrinsic accuracy of the ammo. It's just not very probable that it will occur randomly.

I get the idea of mean radius as the average miss. However, its another case of knowing what purpose you are averaging for. If you take that mean radius and draw a circle around the group center using it, more than half the POI's will be inside it. That's because the outliers, the really far off misses, pull the mean value disproportionately away from the center where most rounds actually fall. That's why the old practice of drawing the smallest circle that included 50% of the bullet holes was once the standard measure. It said what you could hit half the time. It must have been a nuisance to do before computers, though. I'm guessing that's why the went to MR. Today it could be automated, though.

 

[Bart B.]

Unclenick, good set of words. But then, I want to know what the worst of my ammo does. Average group is also very good if you've got enough of them to be statistically significant.

No wonder Lake City shot 270 rounds testing 30 caliber match ammo to see if it met accuracy specs (3.5" mean radius at 600 yards) with a statistically high level of confidence. Their 1965 lot of 7.62 NM ammo had a 1.9" mean radius. All 270 shots went into 10 inches.

That 1000-shot group in your link appears to have about a 3 MOA extreme spread. Lake City's test targets i've seen have about 60+% of the shots inside the mean radius.

 

[F. Guffey]

Snyper, I think you're ignorant (not stupid)

That was Bart B., not me.

And I thought you were talking about recently. Nothing fits a chamber like a fore formed case, there is nothing that fits a chamber better than a fire formed case 'UNLESS!' it is a fired formed case with a partially sized neck. I said it could help align the case.

Then Bart. comes along and disagrees with that. I am beginning to believe he is widely indignant about every thing. He believes a case can fit the chamber like a peanut complete with the hull in a punch bowl, it is all about the bullet by Sierra and the shooter.

F. Guffey

 

[Snyper]

That was Bart B., not me.

That's not the comment you quoted.

This is what you quoted:

Quote:
In other words, if the loads aren't being fired from a machine rest, no one will ever know the difference.

Nothing fits a chamber like a fire formed case, there is nothing that fits a chamber better than a fire formed case 'UNLESS!' it is a fired formed case with a partially sized neck. I said it could help align the case.

On this we do agree
It really doesn't matter if it's "centered on the case" if it's centered in the chamber neck, as a formed case will

 

[Pond James Pond]

Well, I did it!! I wound down the FL die to the shell holder, measured a case at the datum, sized it and remeasured it: 0.004" or 0.1mm. (at least that means that the Lee die had not been not massively over-stressing the brass as much as I had feared).

So would it out and tried another case. After the 4th case I had it down to 0.05mm shorter at the datum.

So I've done it and set back the shoulders. Seemed like such a black art when I first heard about it but really it was a breeze!!
(please don't now tell me I did it wrong!!)

One case refused to change its length to the datum even though all the others I check randomly (about 4 or 5 out of the 30 I resized) were down to the 40.58mm I had aimed for initially.

I will chamber that one to see it if fits in the rifle without too much difficulty.

 

[cw308]

Checking fired cases & adjusting headspace is not that difficult with the right tools. I check the deprimed fired case with a RCBS Precision Mic, with a set of compitition shell holders, adjust my F/L die to contact longest shell holder, then change shell holders until I get the the proper headspace . Rechecking with Mic. All cases are cleaned first before I measure & size.

 

[F. Guffey]

Am I missing something? (Rhetorical question)
What am I missing? (Actual question)
__________________

The chamber is 308 Winchester. 308 W cases can be formed from 30/06 cases. The shoulder of the 30/06 is further out than the 308 W by .389". If a reloader started forming a 30/06 case to 308 W eventually the formed case would chamber. Problem! Knowing when to quit forming.

A reloader could start forming a case but instead of sizing and measuring, sizing and measuring they could measure to determine the amount of sizing necessary to chamber the case.

F. Guffey