My proof: sweat the details

[Tcox]

I conducted a bit of a "test" today. I wanted to see if all the small things REALLY matter. This one was very simple and I did not get into advanced techniques (mostly because I am NOT an expert). I conducted an optimum charge weight test. I tested charge weights 41.8-42.2 in .1gr increments. I loaded 3 rounds of each weight. The tests were conducted at 300 yards to magnify problems.

Here are some stats regarding my regular loads:

Variance of +\- .06 grains in charge weight
Variance of +\- .007 inches in brass length
Variance of +\- .008 COAL length
Did not weigh brass or projectiles
45 degree dechamfer/deburr
No pocket uniformed

THEN I duplicated the above test but with the following changed variables:

Variance of +\- .03 grains in charge weight
Variance of +\- .0005 in brass length
Variance of +\- .001 COAL length
Variance of +\- .1 grains brass weight
No variance in projectile length/weight
22 degree dechamfer/deburr
Primer pocket uniformed



I found the results to be absolutely astounding in this paticular test and found an outstanding load for my rifle.

To keep as close to Apples to apples, I'll compare charge weighs 41.9 grains. The first pic is the test with more "slop" to the loads if you will:




To be fair I called a flyer on this group. I BELIEVE that I pulled approximately 1/2inch right on the flyer. I'll take measurements ignoring the flyer to get the best possible data. I'll estimate a .456 center to center group WITHOUT the flyer.

Now this is the 41.9 gr group with the tightest tolerances I could achieve and I had zero flyers. This is the best group I have EVER shot and I'm very proud of myself for achieving it. I estimated a .081 center to center group.



What I accomplished his is raw data that the small things matter even for a novice. If you guys can think of better ways to improve this test, please let me know. I'm going to continue these types of tests and see if I can duplicate that group!

 

[Gus Petch]

I loaded 3 rounds of each weight.

Six rounds is hardly a comprehensive sample size.

I also don't see any chronograph data, nor do I see any standards to which any of your measuring instruments have been calibrated.

You're also the sole loader and shooter; none of your biases can be excluded.

 

[Tcox]

Six rounds is hardly a comprehensive sample size.

I also don't see any chronograph data, nor do I see any standards to which any of your measuring instruments have been calibrated.

You're also the sole loader and shooter; none of your biases can be excluded.

I would agree that 6 total rounds at each charge weight is not ideal, however I would have went though an absolute ton of components for a test in which I didn't have a way of knowing whether or not they would produce a good canidate load. I do feel that it is a somewhat comprehensive test as I performed the same test for each charge weight. I should have elaborated on that more, especially since the only photos that I uploaded were for one charge weight. With each charge weight I received similar results, and my worst group with tight tolerances was smaller than any of my groups with looser tolerances. This test was a bi-product of load development. Maybe for the next test I will only test three charge weights and 12 rounds now that I have found my node?

I do wish I had chronograph data, I do not have a chrono at this time. I absolutely agree that velocity would be outstanding to have in this testing.

Regarding my biases, are you referring to excluding my flyer? If so, given the circumstances, I feel that including a shooters flier would skew the data. If I had felt confident in that shot I think that the flier should be absolutely included, but knowing that I pulled approx 1/2" right I don't think that I could include it on a test that is focused on the loads, not the shooters ability.

I appreciatiate your feedback Gus!

 

[Gus Petch]

Regarding my biases, are you referring to excluding my flyer? If so, given the circumstances, I feel that including a shooters flier would skew the data. If I had felt confident in that shot I think that the flier should be absolutely included, but knowing that I pulled approx 1/2" right I don't think that I could include it on a test that is focused on the loads, not the shooters ability.

Short of using a machine rest, you simply cannot exclude any human error or even reduce it to a negligible effect. Whether you realize it or not (or are willing to admit it), you are already predisposed to the tighter tolerance group producing "better", i.e. more repeatable, results or at least the results you want to obtain. You -- and any shooter, for that matter -- knowing a difference exists within the load makeup is more likely to induce (or reduce) any error within your control, consciously or subconsciously.

This still doesn't address any instrument calibration standards or potential discrepancies. What was used to verify the scale and caliper readings? Were the readings compared to those obtained from other devices?

You're obviously happy with your results, but none of them concludes anything one way or another.

A chronograph would offer some validation, as would a "blind" test involving other shooters (who can obtain repeatable results).

 

[Tcox]

Gus, I do understand what your saying, however, don't you think that some of that is a bit unrealistic? I agree that human error plays its role in all testing, but human error in both shooting/loading methods is a constant. The same human shot the same weapon from the same location and received repeatable, consistent results. In my mind, this yielding in five reasonably consistent targets brings credible data. Is it perfect? Absolutely not, nothing is.

You may or may not have significantly more knowledge than I do in this but when a shooter knows for certain that he/she had a significant fault in shot placement wouldn't it substantially alter your data to include that shot? If a shooter "pulls" a shot The point of aim/point of impact would shift and that shot would be completely inconsistent with atleast the point of aim for the previous rounds? Nonetheless, include the flyer and this test shows an even more substantial increase in accuracy due to tighter tolerances in the loads. The outcome is the same. Omitting the flier I think is important and this test is only to show the difference on paper from precise or less precise loading. This wasn't a shooters ability test. Eliminating conscious human error is important because it is a known. I cannot factor subconscious human error.

I use a micrometer and check with a dial caliper. I have not had to calibrate since I have owned it but I check its calibration by measuring 3 known length cases. The dial caliper acts as insurance.

 

[Gus Petch]

Gus, I do understand what your saying, however, don't you think that some of that is a bit unrealistic? I agree that human error plays its role in all testing, but human error in both shooting/loading methods is a constant. The same human shot the same weapon from the same location and received repeatable, consistent results. In my mind, this yielding in five reasonably consistent targets brings credible data. Is it perfect? Absolutely not, nothing is.

You may or may not have significantly more knowledge than I do in this but when a shooter knows for certain that he/she had a significant fault in shot placement wouldn't it substantially alter your data to include that shot? If a shooter "pulls" a shot The point of aim/point of impact would shift and that shot would be completely inconsistent with atleast the point of aim for the previous rounds? Nonetheless, include the flyer and this test shows an even more substantial increase in accuracy due to tighter tolerances in the loads. The outcome is the same. Omitting the flier I think is important and this test is only to show the difference on paper from precise or less precise loading. This wasn't a shooters ability test. Eliminating conscious human error is important because it is a known. I cannot factor subconscious human error.

I use a micrometer and check with a dial caliper. I have not had to calibrate since I have owned it but I check its calibration by measuring 3 known length cases. The dial caliper acts as insurance.

You've missed the point made. You've also made multiple changes between your loading methods that prevents you from quantifying any actual improvement gains. And you seem to be omitting your scale when discussing measurement devices.

Without a chronograph, a larger sample size, independent calibration or measurement verification, and additional shooters, your data doesn't indicate anything meaningful.

 

[Tcox]

Okay, would you share how you would change the things you would change? Maybe I would learn from it that way?

 

[Gus Petch]

Okay, would you share how you would change the things you would change? Maybe I would learn from it that way?

I've stated it three times at this point.

 

[Tcox]

Alright, my apologies I suppose. I'll test further once I have a chronograph, more rounds, and more ways to to verify measurements.

 

[Gus Petch]

Alright, my apologies I suppose. I'll test further once I have a chronograph, more rounds, and more ways to to verify measurements.

I can appreciate your goal. However, there are problems with your data collection that negate your results.

• You tightened up your charge weight tolerances. How accurate is your scale and how did you verify its accuracy? How much (if any) bearing did this change have on your end results?
• You tightened up your case length tolerances. How much (if any) bearing did this change have on your end results?
• You tightened up your OAL tolerances. How much (if any) bearing did this change have on your end results?
• You sorted brass based on weight. How much (if any) bearing did this change have on your end results?
• You changed the chamfering angle in the deburring step. How much (if any) bearing did this change have on your end results?
• You uniformed primer pockets. How much (if any) bearing did this change have on your end results?
• You didn't chrono any of the loads to check external ballistics. (Chronograph technology has reached a point where they're fairly affordable and a wise investment if you're trying to chase a certain reloading goal.)
• Most importantly, you only fired a total of six rounds for the test. Bump that number up to 15 rounds per load.

I'm not trying to heckle or harangue you, but if you're trying to gather repeatable data from your loading and range efforts, you'll need to make a few changes.