Rifle load development methods compared

[Slamfire]

I am leery of small sample sizes.I used to shoot ten shot groups on initial testing, now I am getting lazy and testing five shot groups, typically increasing charges by 1/2 grain. If the load is bad, it is usually bad all the way through the sequence. Then, if the load has promise, shoot ten shot groups, sometimes 20 shot groups.

Then go to CMP Talladega and shoot at 200, 300, sometimes 600 yards. You would think that would solve everything, but, often I experience leaking or blown primers, which then cause me to cut the loads.

Shot well expect for the blown primers and the bolt face etch.



And you know, testing the reduced loads, I find they were not so bad after all.

It really takes lots of rounds downrange to have confidence in your ammunition.

 

[MCMXI]

I am leery of small sample sizes.I used to shoot ten shot groups on initial testing, now I am getting lazy and testing five shot groups, typically increasing charges by 1/2 grain. If the load is bad, it is usually bad all the way through the sequence. Then, if the load has promise, shoot ten shot groups, sometimes 20 shot groups.

@Nature Boy regularly shoots 20 shot validation groups in F-Class matches and if you've seen his posts you'll know that he gets very good results, so his "small sample size" method has been proven time and time again to work very well. That's the advantage of a system in which the errors are very small, you don't need to shoot hundreds of rounds to have confidence in your system. Also, the precision bar is considerably higher for F-Class than it is for Service Rifle, and obviously it's even higher for Benchrest, but the gap between F-Class and Benchrest seems to be getting tighter (just a guess). The group you show above is impressive for Service Rifle but would be an average group at best for F-Class at 300 yards. The not very impressive best score I ever got in F-Class at 300 yards was 149-4X for a 15-shot match and the 10-ring is under 3", not 7".

 

[Slamfire]

@Nature Boy regularly shoots 20 shot validation groups in F-Class matches and if you've seen his posts you'll know that he gets very good results, so his "small sample size" method has been proven time and time again to work very well. That's the advantage of a system in which the errors are very small, you don't need to shoot hundreds of rounds to have confidence in your system. Also, the precision bar is considerably higher for F-Class than it is for Service Rifle, and obviously it's even higher for Benchrest but the gap between F-Class and Benchrest seems to be getting tighter (just a guess). The group you show above is impressive for Service Rifle but would be an average group at best for F-Class at 300 yards. The not very impressive best score I ever got in F-Class at 300 yards was 149-4X for a 15-shot match and the 10-ring is under 3", not 7".

The group I attached in the previous post was a M70 hunting rifle, had it recently rebarreled and asked the gunsmith to duplicate the SAKO Finnbear contour


And it shot well for such a lightweight rifle.

I agree that Nature Boy's F class groups, particularly his daugther's 600 yard groups are impressive. Finding a good load takes work, we all have developed our own strategies to find the best load with as little work as possible, but as Yogi said: Prediction is hard, especially about the future.

This is an example of small sample size, giving a misleading idea of a load's potential:

This was not bad for a chrome plated period rifle



But, the 20 shot group really opened up



So what is the truth, sub MOA or two MOA?

at distance, 2.5 MOA



This rifle with 130 grain bullets shoots around 2 MOA. The more you put down range the more you figure out what you and your rifle can do.



I know the F Class guys are all shooting inside anything I can do with a hunting rifle or a sling. The more you eliminate the human, the tighter groups become. F Class shooters I know, they stopped using bullets that used to be fine with sling shooters. The sling shooters could not see the inaccuracy but the bi pod mounted, bag rested, F Class rifles could. And I predict, when artificial intelligence advances enough, Robots will shoot inside anything a human can do, even with a bi pod. That has not ruined chess, hopefully it won't ruin mid range.

 

[MCMXI]

Without more data it's tough to say anything meaningful about your 5 and 20-shot test. It could still be the best load for the rifle but the limitations of the rifle might affect group size as a function of round count. It could be a sub-moa load being shot in a 2 moa rifle. So many always assume that a 5-shot, 10-shot or 20-shot group will be bigger than a 3-shot group. Statistically, if nothing else is changing, the second shot of the group could and should on occasion define the group size. Just as the third shot of a 20-shot group could define the group size. But as I said, this is contingent on nothing else changing. In your case, something could be changing that's affecting group size, a systemic problem.

If the first 5 shots of that 20 shot group are always sub-moa then you probably have a systemic problem that no amount of load development will fix.

 

[Nature Boy]

I agree that Nature Boy's F class groups, particularly his daugther's 600 yard groups are impressive

I don't think we should use her as an example. She's a just a freak of nature. (Nature Girl? )

I am leery of small sample sizes.I used to shoot ten shot groups on initial testing, now I am getting lazy and testing five shot groups, typically increasing charges by 1/2 grain. If the load is bad, it is usually bad all the way through the sequence. Then, if the load has promise, shoot ten shot groups, sometimes 20 shot groups.

I don't think what I'm showing here in the 100 yard load development method should be considered a small sample size.

Step 1. Load testing - it requires 3 shots per charge but those are not discrete data points. They are part of a total analysis of multiple shot groups. Important to note here. The size of the 3 shot group is irrelevant. It's how the center of the group compares to the groups on it's left and right, therefore, you are never analyzing anything less than 9 total shots.

Step 2. Seating depth - same thing. It’s a continuation of Step 1. Again, multiple 3 shot groups

Step 3. Primer selection- again the same thing, 4 to 5, 5 shot groups

Final validation of all 3 steps - 20 rounds at distance

So when evaluating the method’s statistical significance it's cumulative and you have to take it in its totality.

 

[Nature Boy]

This is an example of small sample size, giving a misleading idea of a load's potential:

This was not bad for a chrome plated period rifle

But, the 20 shot group really opened up

So what is the truth, sub MOA or two MOA?

at distance, 2.5 MOA

So in comparison, I’ve found that I can maintain the same accuracy from 100 through 500 yards using the 100 yard load development method outlined in the preceding posts. Not only that, I’m doing it across multiple rifles

In other words, if my rifle with tuned load consistently delivers 0.1 to 0.2 MOA 5 shot groups at 100 yards, it will also do it at 500 yards.

Naturally, shoot 20 into a target at any distance and that group will open up because you’re introducing more variables into the equation.

So how does a rifle that shoots 5 shot groups in the 1’s and 2’s at 100 perform when shooting 20 into a target at 500? Well, on this day, it was a 0.67 MOA rifle



Good enough for an F Class score of 200 X19. Whether it’s good enough for me to quit fiddling with it is still up in the air.

 

[Varminterror]

I’ve started planning a similar exercise for this summer, comparing Satterlee, Audette, and Newberry methods, as I have a few new barrels which don’t have any significant competition life planned, and could take the extra round count.

Unfortunately, my interest in doing so is often diminished when I’m reminded of these “you didn’t shoot enough to prove anything” posts.

 

[Nature Boy]

I’ve started planning a similar exercise for this summer, comparing Satterlee, Audette, and Newberry methods, as I have a few new barrels which don’t have any significant competition life planned, and could take the extra round count.

Unfortunately, my interest in doing so is often diminished when I’m reminded of these “you didn’t shoot enough to prove anything” posts.

I'm looking forward to that.

(and now that you've said it out loud, I'll be bugging you about it )

 

[Legionnaire]

I'm looking forward to that.

Me too!

 

[Varminterror]

Guess I better get out my pen, stop Storming and start Norming on my experimental plan!

 

[WelshShooter]

I'm going to leave the statistics bit aside for not wanting to wind up @Varminterror

In short, @Nature Boy 's method seems logical and single results from a group of three aren't likely to make the whole experiment void. I had traditionally spent a lot of rounds in load development trying to find a good load, but I'm following @Nature Boy 's three step plan and have good results so far with less rounds fired down range. I'm happy with that!

 

[NATO Reloading]

Personally, I believe the largest mistake people make when finding good loads, is they have no "control" in place. For example. If I just buy a lot of brass and just load it..some brass have less/more capacity than others in same lot. This is the same as loading more or less up to .2 grains of powder. You get the idea.

 

[Nature Boy]

Personally, I believe the largest mistake people make when finding good loads, is they have no "control" in place. For example. If I just buy a lot of brass and just load it..some brass have less/more capacity than others in same lot. This is the same as loading more or less up to .2 grains of powder. You get the idea.

Agreed, the more variables you can isolate leaves the most spurious variable in the equation remaining. The shooter

 

[Nature Boy]

As an aside, if this thread was about the best way to make my brass shiny or what case resizing lube to use it would be 6 pages long.

And, I’d be jumping in the middle of those too, because I have an opinion

 

[Varminterror]

some brass have less/more capacity than others in same lot. This is the same as loading more or less up to .2 grains of powder.

In fairness, this is a hypothesis which can be tested in a sensitivity matrix study - which is really more of what we do in load development. We aren’t typically testing a discrete hypothesis, but rather investigating the sensitivity of our precision to our inherent variability (not necessarily variables). The hypothesis, for example, in an Audette ladder test is that there will be a “node,” a range of charge weights, where the POI becomes insensitive to changes in charge weight. Or at least less sensitive in that range than other ranges of charge weight. The experiment then focuses upon finding that node, and a defined control group isn’t terribly relevant. Same deal for seating depth tests - we expect one or a few depths (jumps) to shoot smaller than others, without describing a narrow hypothesis that a certain depth is expected to shoot better than a control depth.

In SOME load work, the nature of the hypothesis makes defining a control much easier, and much more relevant; for example, determining whether it is better to change to WSR’s instead of BR4’s. In that experiment, a failure in experimental design would be to simply swap primers without doing the same degree of charge weight sensitivity work done with the “status quo” control BR4’s. Another hypothesis might be determining if RL16 is better than H4350, or Norma brass better than Lapua, or 107 SMK’s better than 105 Hybrids. Of course, sufficiently defining the KPI’s for success, aka, defining “better.”

In other load work, as in what most of us end up doing, it’s much closer to the open hypothesis of “a region of insensitivity exists within this range, and we will find it.” Seating depth, charge weight, neck tension, etc. The control may just be the first one you shoot, since it’s the first data point for comparison, it becomes the bar, and everything after is hypothesized, inherently, to shoot better or worse.

 

[NATO Reloading]

The audette ladder testing method is, let me just say, too simplistic and does not work "repeatably". flat spots in the curve in SOME, are because of issues like case capacities not being the same, or powder, or etc. Then when you do the SAME ladder test again, the flat spots are now in different nodes?

This whole concept is so horribly flawed, it completely doesn't work.

 

[Varminterror]

The audette ladder testing method is, let me just say, too simplistic and does not work "repeatably". flat spots in the curve in SOME, are because of issues like case capacities not being the same, or powder, or etc. Then when you do the SAME ladder test again, the flat spots are now in different nodes?

This whole concept is so horribly flawed, it completely doesn't work.

Kinda odd to read this, since it has been “working” to minimize vertical dispersion in my long range groups for over a decade, and is the standard procedure for long range competitors all over the country.

Repetition after repetition, my nodes don’t move significantly.

ETA: it’s also kind of apparent you’re casting dispersions while you don’t actually understand the method, because “flat spots in the curve” has nothing to do with the Audette Ladder Method at all. I assume you are unfamiliar with both, since you are confusing the Audette method with the Satterlee method.

Equally, I have been using the Satterlee method for the last two years, and my nodes (flat spots in the curve) don’t move from one rep to the next either.

 

[NATO Reloading]

Here you go.

http://2poqx8tjzgi65olp24je4x4n-wpe...audettes-20-round-string-load-development.pdf

I stand by what I said. It totally doesn't work. It goes against the laws of physics. If you load more powder with all same conditions, you will get more pressure. This is common sense. More pressure = more velocity. all those flat spots are because of inconsistent loads. That is a fact, not an opinion.

 

[Legionnaire]

Quoting from the chapter you posted:

"However, the point of impact did not form a linear, vertical line of holes, evenly spaced from low to high. In other words in certain areas a small increase in powder charge did not necessarily produce a corresponding rise in the point of impact on the target. Curious to know whether this was a result of the rifle being sensitive to a particular pressure-velocity combination of if the ammunition would behave differently within certain portions of the overall curve, I tried an experiment similar to Mr. Audette's ..." "As you can see, there are two areas where the velocity curve levels out, even though the powder charge continues to be increased ..."

"Whatever cartridge and load we chose, it makes sense to operate in a stable area that is at a level below the maximum allowed by safety considerations."

"... my experience does not run contrary to these observations."

Sounds to me like the author agrees with Varminterror.

 

[Varminterror]

Here you go.

http://2poqx8tjzgi65olp24je4x4n-wpe...audettes-20-round-string-load-development.pdf

I stand by what I said. It totally doesn't work. It goes against the laws of physics. If you load more powder with all same conditions, you will get more pressure. This is common sense. More pressure = more velocity. all those flat spots are because of inconsistent loads. That is a fact, not an opinion.

Educate yourself a bit about positive compensation, which is often discussed as nauseam in OBT method discussions, to make your searches easier.

As noted in the article - there are spans (colloquially “nodes”) in a powder charge vs. vertical POI dispersion where POI’s “level out”. The physics of it are relatively well described in the OBT/positive compensation discussions.

Despite inputting an increasing range of potential energy (powder), all of the Audette Ladder, Newberry OCW, and Satterlee velocity curve methods reflect “flat spots” where increases in powder charge don’t influence a vertical lift in POI, nor an increase in muzzle velocity. Your contention such internal capacity variations explain away these changes is simply false - because there are many of us who have sorted cases by internal volume, and proven your statement false.

“Facts do not cease to exist because they are ignored.” Failing to understand a truth does not diminish its veracity.

 

[NATO Reloading]

HI man. Thanks for defending your position, I respect your thoughts on the matter. I just totally disagree.
I am obviously not going to change your mind, and you won't change mine. I have asked people like you to prove it at my club, many times. Nobody can seem to prove it. Not anyone. They are all good reloaders.

If you were at my club, and I told you to make say 100 bullets, same exact load, everything same. And you did 5, 20 shot ladder tests. The results would not be consistent, and you would be saying X to justify it still works. Most did that, some were humble and accepted that results say it all. Nobody, ever, has shown any of these methods work consistently. As soon as we get to the 2nd and 3rd groups, the theories fall apart and nodes are changing. I just laugh. Barrel too hot, brass different lots, scale broke, used another one, temperature changing. I heard it all.

It doesn't work "better". Its all nonsense. its voodoo. These people that created all this sound smart, they make a good argument, but it doesn't work reliably. So many people across the country trying this stuff. Its hilarious. The religion of ladders/methods. Lots of fake religions out there. haha

Here is some more humor.



all these tests. LOL Watch that video, this is my experience x 20. They have 1 erroneous result, and they ALWAYS, ALWAYS ALWAYS explain it away..like "oh..I loaded those yesterday... I should have loaded this all at the same time..." LOL always an excuse. Always. Then the other guy is like "all these tests, they all lead to the same spot for accuracy" haha? Really? So all these religions do the same thing as traditional load testing? really?

haha comedy.

Here is part 2:


The excuses are FLYING

• I had an issue with rear bag
• the bolt was sticking
• there was always the Human error, it could have been me! HAHA MY FAVORITE ONE
• The flyer was me...I BELIEVE THAT! hahaha!!

Funny no mention of wind being variable. I hear that a lot too. LOL

 

[Demi-human]

So what does work?

 

[Varminterror]

As soon as we get to the 2nd and 3rd groups, the theories fall apart and nodes are changing.

I feel bad for your local guys, but this is absolutely not my experience. One, because if this is true, they aren’t very good reloaders, but worse, they have your bad advice in their lives.

My nodes don’t move from one string to the next.

The more you talk, the more it becomes apparent you are trolling and not at all familiar with these methods. “2nd and 3rd groups, the theories fall apart and nodes are changing.” Satterlee method isn’t shooting groups, nor is Audette method. The string is what defines the nodes, not the groups.

make say 100 bullets, same exact load, everything same. And you did 5, 20 shot ladder tests.

Loading “100 bullets of the same exact load, everything the same” is not a Ladder. 20 shots isn’t a Ladder string either - that’d be 4-6 grains of range in a short action cartridge, 8-10grns in a magnum, overall an absolutely terrible load development design. More evidence you’re just trolling and don’t really know anything about these methods you are criticizing. The first set of a ladder might be 10 shots, replicated a few times. The second stage might be 10 more at a narrower interval. Subsequent validations don’t usually need to be any more than 5-8 charge weights.

So if you or your local club guys can’t replicate nodes, you’re either fabricating an imaginary experience, or not a very good reloader.

 

[JFrank]

The audette ladder testing method is, let me just say, too simplistic and does not work "repeatably". flat spots in the curve in SOME, are because of issues like case capacities not being the same, or powder, or etc. Then when you do the SAME ladder test again, the flat spots are now in different nodes?

This whole concept is so horribly flawed, it completely doesn't work.

Geez
I thought ladder testing for course powder and charge conformation worked pretty darn well.
Hell I don’t even know Audette whatever the names are.
I never use a chronograph for load development only afterwards for calculating bullet drop. (That’s nothing against those who are comfortable doing so)
For me ladder testing is the simplest and quickest method for developing an accurate recipe.
IME it’s very repeatable.
My 2 cents
J

 

[NATO Reloading]

Varminterror, the "groups" means the 20 bullets. If you want to call it "string", go ahead. Its ok if you want to preach your religion. Keep doing it. Since none of us can verify your claims in person, you will always be right.

The guys at my club are bad reloaders? Is that all you got? I am NOT giving them advice, I am asking for ANYONE on earth to prove this ladder method works, or other methods. They simply can't do it repeatably.

20 shot strings are common. Thats .1 grains each. Thats 2 grain spread. You are not more or less an expert by saying you do 10.

If I was with you, you would completely fail, make excuses, and I would laugh.