The mechanics of popnfresh's conclusion (less runout leads to smaller groups) makes sense: with the bullet initially aligned perfectly with the bore, there are fewer factors to add to group size, as the bullet should enter the bore and engrave in the rifling as closely to perfect as possible.  But I'm trying to figure out how less runout (very close to zero) can contribute to increased group size.  

Having solid data and sorting it out thoroughly is one thing.  Where's the theory to explain this new interpretation?  Why would certain amounts of runout lead to smaller groups, but other, smaller amounts lead to larger groups?  I surely can't come up with any explanation...

The following comment is NOT a reflection on popnfresh's shooting, testing, reloading abilities.  More a suggestion that maybe the test has to be even more anal to know what's going on.  There is just more than "run-out" that factors into group size.  Bullets used, case neck thickness, distance to the lands (and that optimizes different depending on the rifle), primer used, powder load.  Then the external factors, temperature, wind, atmospheric pressure, human factor.  (ha, for me personally, a single caffeinated Coke in my system makes me suddenly group worse).

My point being, besides breathing rituals, grabbing the right ear once before moving the finger to the trigger or other ritual, benchrest loaders sort their bullets by weight and length even out of the same box.  They turn case necks, measure and sort their cases by volume.  They use bench rest primers........and some powders, they find, are more consistent than others in their particular rifle and that's beyond their ability to weigh a repeatable accurate load....and all that besides using Lapua brass and a straight line arbor seater.

LOL!   Me, either.


What you have to be careful of is running a test assuming some factor is important when it is, in fact, a minor contribution.  Then, when your data does not support your initial hypothesis or is only weakly related,... what then?

Another idea is that the group sizes were limited by some other variable which masked the effect of reduced runout.  Possibilities include shooter or rifle limitations, etc.  

I know one thing from personal experience - shooting below 1 MOA is very difficult, even at 100 yards.

The importance of that sub-plot is to show how difficult it can be to get data that unambiguously shows the effect of runout.  Certainly, it looks like there is some effect (and I want there to be an effect) but it is not a strong effect.

Consider this as an hypothesis, conceived only after looking at the data (which I still think is a terrific test AND good shooting).

Runout has a minor effect on group size which will only be manifest if your groups are already well below 1 MOA.  



After all, what is the theory behind runout causing groups to open up?  The ballistics/trajectory is dominated by the center of mass.  For that reason, I would bet the concentricity and uniformity of the core (inside the jacket) is as important as the concentricity of the jacket, itself.

Error sources tend to be dominated by the shooter, rifle and wind.

Popnfresh, did you weigh the bullets or cases?  Did you measure muzzle velocity for the shots?  Did you track and adjust for  the wind?  Did you call any shots as fliers?     Some groups show what might have been the effects of wind.  

I think you meant popnfresh.  All I did was some data processing and some thinking about what the data "says", what it is telling us.


I have no plans to stop using my Redding Competition bullet seaters!

oops!  I didn't read back far enough.....sorry.  fixed it.

I feel the same way about my RCBS Gold Medal Seaters......

I think that this is correct. The problem is that every .1 MOA once you get below 1 MOA is a huge deal. If we could show a .1 MOA difference statistically, that would be an easy decision for precision rifle shooters to make.

What was your testing methodology?  In either proving or disproving the effect of runout on group size I agree that there are a lot of variables that can have an equal or greater effect when we're talking about sub-2 MOA or sub-MOA sizes.

Popnfresh does a decent job describing his test method in his posts (re-read the first page).  I'm just analyzing the data and reviewing the results.

The thing that is most puzzling to me is the virtual lack of correlation (R2 = 0.05) with low to medium amounts of runout (check my second plot).

Benchrest shooters, good ones, shoot well below 1/2 MOA but they have specialized rigs and supports, wind flags out the wahzoo and even special targets.  Perhaps they can see the difference.  



From this data, I would say if your reloads have less than 5 mils runout, you are fine (apparently the AMU specs 3 mils).  If not, and cannot adjust it to give better results, a Competition seater might help.  In my case, that's the die I start with.

I don't think a person should take one group/runout combination and try to draw a conclusion from it, look at all the groups and look for a trend. That is why I just averaged.

The days I did these I fired almost 100 rounds as carefully aimed as possible in a couple hours which is not easy so there is some shooter error; so again, looking at one groups extreme spread is not a good for drawing conclusions.

I believe I mentioned this but this was all blind as could be(for being one person) and all fired round robin to be as equal as possible.

I gave the targets to Rocketman to calculate standard deviation from group center as I thought this would better eliminate shooter error but he must not have gotten to it.
I thought average to center moa per group might be better than Extreme spread moa as it would give a better idea of the trend of the group rather than just looking at the 2 worst shots of each.

The average to center MOA numbers are:
.004 and under= 0.2317 MOA
.005 and over = 0.3506 MOA





This is my 175smk load from 500pc lots no weighing of anything is needed for ESs of 35fps and SDs of 12fps and at 100yds these probably wont make much difference. No wind was accounted for, these were round robin over a short period, I do however wait for gusts to subside out of habit. I have the conditions recorded in my data book, I can post them when I get home but I don't recall heavy winds.

I like this thread, so I'll go back and redo the numbers using your ATC and see if there's a difference.





For me, I conclude this - I make the most concentric rounds I can.  I use the Competition dies as these are most likely to give me highest concentricity.  I install and adjust the dies to provide concentric rounds.  I have the gage(s) to measure while adjusting/tweaking.  

I make no attempt to adjust the bullets after seating, even though I have a tool to do it.

In testing, my ammo shoots ~1/2 MOA vertical at 200 yards with horizontal limited by the wind.  I just can't shoot better than 1/2 MOA, even under nearly windless conditions., as I am not a trained benchrester with all that gear.   Under normal shooting conditions (HP matches), I don't shoot anywhere near 1/2 MOA.  so I am not concerned about tiny improvements (<0.1 MOA) from higher concentricity.

Still, I do the best I can within my personal limits of what is a reasonable amount of time to spend reloading.

We all missed the real question - How did you get all those gorgeously colored bullets?

Here are the numerical values using average distance to the center (ATC).  

I'm not going to post plots because there's really no difference from the earlier plots.  If you really want the  plots (for some reason), I have them.  

As you might expect, the nominal ATC values are about 1/2 the group size values because it is the ATC error for each shot, not the sum of the minimum and maximum errors.


Overall, the correlation coefficient (R2)* is even lower: R2 = 0.23.
The slope is also lower: m = 0.021"/mil of runout but that derives from the smaller numbers in the database..  <--- edit in itlics

For the 5 mils runout and less, R2 = 0.035 (essentially no correlation between runout and ATC).

For the 6 mil and greater, R2 = 0.028 (essentially no correlation between runout and ATC).
The slope is very small but still negative (gets slightly better with higher runout).



*  I am using R2 as an abbreviation for the correlation coefficient in lieu of R-squared (R times R) because there's no superscript font option at ARFCOM.  I am using a linear curve fit to the runout data.  

An R2=1, your data fits the trend line perfectly.  R2=0, there is no mathematical correlation between the trend line and your data.  In this thread, in almost all cases, we are flirting with R2 ~ 0.

It is pretty clear that the groups with less runout have smaller averages to center(better potential MOA).

From the data here, what are the odds the groups will be better with runout under .005"? good or bad?



You don't  see the difference here?
.004 and under(not counting the 3 and 4 shot)

.311 MOA
.310
.230
.285
.272
.263
.242
.214
.173
.233

.005 and over

.485
.483
.242
.200
.491
.237
.322
.386
.361
.299


Which runout amount are you more likely to see average to center numbers over .310moa, .004" and under or .005" and over?

Here is some more data, from a primer comparison test, this is all under .004" runout. These ATC numbers from a later test match the .004" and under from the OP.

.131 MOA
.205
.246
.227
.219
.365
.167
.269
.219
.355
.183
.195
-------
2.781 / 12 =.23175 MOA Ha! look at my numbers 5 posts up for .004 and under.

Here are the plots using Average to Group Center rather than size in MOA.  

Run probabilty calculations. People want to know if straighter ammo will improve groups it is a simple as that.



Posted Via AR15.Com Mobile

And if straighter bullet seating does not lead to smaller groups, I want to know why not.

I took this advice from this thread and now have .223 and .308 Forster dies.  They work very nicely.  Thanks for the effort.  (hope I ain't duping myself)

No one has said your statement is false.

I am beginning to suspect that many of you do not understand the work I've done and already posted.  

Re-read the posts.  

Do the math, yourselves - it does not take that long and you may learn something along the way.  If you are not interested enough to do that much, ...

Do you have a concentricity gage?  If so, how do the new seaters compare to the older ones with respect to runout of the ammo they produce?  Did it change (at all) and if so, how?

How much of the total runout is manifest in your fired cases?  How much is caused by your sizing die?  How much by your seater?

I'm just restating that if there is evidence that minimizing runout doesn't lead to smaller groups, I'd like to know why it doesn't.  Sort of keeping the "counter-theory" in play.  

Logically, either decreasing runout leads to smaller groups through some mechanism (such as how I described it earlier), or it  leads to larger groups through some other mechanism (unknown at this time), or decreasing runout  has no real effect on group size through yet another (also unknown) mechanism.  So whichever of these three statements is true, we need to understand why it is true, whichever one is shown to win out.

Additionally, I'd like to have a synopsis of just what data was used in these analyses, so that we can all see what sort of statistical population these analyses and conclusions are based on.  As (almost) always, a larger statistical universe tends to provide more clarity, so if this whole thing is based on only a few hundred rounds, maybe we just need to add more data and have the anomalous-appearing issues wash out.

I think it is more like mute button than a volume knob.
Please keep it simple are the odd good or bad that my groups will improve with straighter ammo?

Posted Via AR15.Com Mobile

Before you propose that as a "counter theory", consider the following and think about overall error masking the effect popnfresh hoped to find in his experiment.

Are you familiar with error analysis?  Consider this - there are many error sources that determine the size of the group.  It is entirely possible that you could reduce or eliminate one source of error and have no measureable reduction in overall group size.  This would be especially true if the one you eliminate was not one of the dominant error sources in the first place.  I can provide an example but will move on to your other points.

If you go back and read the discussions above, you will see a long laundry list of things that may affect group size.  If runout is reasonably low (e.g., under 5 mils),  the effect of further reductions in runout may be masked by the variability in all those other things.  

In addition to those, there are also issues such as non-concentric chambers, less than perfect barrels, damaged muzzles, etc, each of which can contribute to masking the benefit of low runout.  

Popnfresh has shot many groups below 1/2 MOA!, so I suspect his rifle may not have any of these defects to any great degree but that is just one rifle.  Even so, every rifle has some.  So do the bullets, the cases, the primers,... the shooter


One thing I have done with my concentricity gage is to monitor runout all through the reloading process.  My fired brass comes out of my rifle's chamber with nearly zero runout (it's a bolt action, not semiauto tossing my brass around).  Everything I do in reloading makes it worse (my barrel is better than my dies).

Popnfresh provided photographic data on page one of this thread.  The images were processed (using OnTarget?) to measure group sizes and Average to Center results.  I transcribed these MOA and ATC into a spreadsheet.  All the data from page 1 is plotted in the graphs.

I do not have any info on the individual shots.

Honestly, I think popnfresh did an awesomely good test here.  

Getting a larger sample would be nice but it's not my test.  I'm just the analyst,... the messenger.

<Trollslayer is on mute>

Just wanted to thank op for costing me more money, just got my forster .308 die.

Txl

+1 well worth the money

Bought mine in 223 and a redding bushing fl s-type die, after getting everything set-up 10 rounds of match prep brass ran out less than .002. Anxious to see how they shoot compared to my 30.00 rcbs dies that I've been using since I started reloading 223.