Litz and Cortina - follow up on barrel tuner discussion

[gunsandgunsmithing]

OBT is wrong.

Agreed, and I've done the testing you referred to in your last post.

 

[jelenko]

A little more on OBT, if you will.

I've been using GRT for a couple years now to find loads for my 223 spacegun to find 'nodes' that are .5 to .75 MOA. All my testing is done from prone with a sling.

After doing the measurements and getting velocities from a sample, I've found GRT's predictions to be right on for 30-40% of it's predicted OBT loads with another 20-30% of predicted loads within .2 - .3 grains. 20 - 30% of the predicted loads aren't near an accuracy node.

=> Is the OBT theory just not describing a physical activity that does affect accuracy or is it just not accounting for everything that does affect accuracy?

Thanks

 

[davidjoe]

Bigger differences require smaller sample sizes. The military is far from perfect too.

I’d say that Brian’s statistically large sample pertains to “tweaking”, not “testing.” The first focuses on improvement of one individual gun’s nominal performance, and the other on how that individual compares to a standard, or to others like it. Testing, like proof marks, assumes a finished product to be either accepted or rejected and tweaking assumes a work in progress.

Brian and Hornady have created a stir, imo inadvertently blending the two concepts going after the . ___ moa labels makers and competitive shooters casually use to relate where their guns stand. For example, that any rifle “X” ships came with a .5 moa or better target, is a passed “test” of being a plucked and shot .5 moa gun. The test is passed if it “can” shoot .5 moa, not if it has been proven to many times. Erik often says he needs a four inch or .4 moa at a thousand, gun, etc.

The problem I have with very large sample sizes as to testing is that it destroys the subject, and as to tweaking, that “real improvements” shouldn’t be that hard to discern, and if so, then maybe they aren’t. Everything has very limited life, in what we do. When tweaking, we truly are testing this “five” rounds against the next “five” (ten individual bullets and powder draws) with a slightly older, hotter barrel, making it hard to even discern what accounts for differences, and likely it’s not any more the rifle, as opposed to the ammo, hold, rest, or environment.

So does large sample tweaking help, if small groups are unreliable? Brian points out that load testing is often not repeatable. The best isn’t the best, next time, and noise can’t be overcome. Theoretically the aging barrel is a moving target, and by way of example every 16 inch naval round was sequentially numbered and modified, to mirror barrel wear. I’ll credit Brian for saying for many years, don’t carry this too far, the wrecking of barrels once differences have become minute.

 

[drop_point]

Normal distribution is a thing. There is variability that is natural no matter how much it is accounted for, even in rifles that make small tiny groups.

 

[davidjoe]

Normal distribution is a thing. There is variability that is natural no matter how much it is accounted for, even in rifles that make small tiny groups.

I going to say something controversial here, but there is no inherent symmetry to the tight side and loose side of the group distribution bell curve.

Many things can account for worse than average groups, and those things “want” to happen.

By comparison, almost nothing accidentally improves group size. The “flinch” that “corrected” a prior mistake just made is a lightening strike.

Group shooting is not an exercise in random dispersion generation; if I can “often or somewhat reliably” hit, - say three free throw shots before I miss, I will absolutely have more examples of strings where I hit less than three, than where I hit more than three, probably by a ratio of 25:1.

 

[drop_point]

I going to say something controversial here, but there is no inherent symmetry to the tight side and loose side of the group distribution bell curve.

Many things can account for worse than average groups, and those things “want” to happen.

By comparison, almost nothing accidentally improves group size. The “flinch” that “corrected” a prior mistake just made is a lightening strike.

Group shooting is not an exercise in random dispersion generation; if I can “often or somewhat reliably” hit, - say three free throw shots before I miss, I will absolutely have more examples of strings where I hit less than three, than where I hit more than three, probably by a ratio of 25:1.

I wholeheartedly disagree with you. Nobody ever suggested accidental improvements.

If there weren't a normal distribution a rifle would ALWAYS shoot 0.200" for example. Not 0.198", 0.200", 0.205", 0.175" etc. The more precise a rifle, the smaller that standard deviation is however that doesn't mean it doesn't exist.

 

[Straightshooter1]

The trouble is quantifying what they do. As you say, it's obvious physics that *something* happens when you put a weight on a muzzle and move it. It's easy to fall into the trap of thinking it's as simple as the barrel flopping about with a frequency you can tune. Of course you can do that. But when you dig into it, it gets really hard to explain why moving a tuner that weighs a few ounces out .003" does anything at all. The frequency shift is tiny. Mike keeps telling me its a phase shift. I can't say it's not, but how that happens is not something I can wrap my head around. At this point, the conversation usually devolves into "look, I don't care how it works, I just know it does". That is unsatisfying, and I think we can do better.

The test I would have done would be to instrument a gun and measure the changes in barrel motion, not to shoot groups. It wouldn't be terribly interesting reading, but i think it'd have been far more illuminating. The work Geoffrey Kolbe did is the best I've seen on this, but he tested a rimfire with and without a substantial weight - enough to materially change the vibration frequency. The results were what you'd expect - the barrel slowed down and lined up a better launch. We need that same level of effort put into centerfire tuners if we want to understand them better.

Since we can detect gravity waves, I'd think it'd be easy enough to measure changes in barrel motion and exactly how it effects POI's. I'd thing that could involve the use of some type of laser equipment. I'm just a little surprised some engineer/scientist hasn't done this.

 

[BoydAllen]

Just a little reality check. Do you hear match winners going on about statistics? In my experience they have done a lot of work figuring out how to make their rifles shoot, and how to shoot them consistently. Barrels have very limited lives. Because of that we have to learn to deal with the limitations of small sample sizes. The proof that this can be successfully done show up in the performance of top shooters. It is my belief that the vast majority of shooters do not own a set of wind flags. For these same shooters to talk about small variations in accuracy as if they are the result of their reloading just seems silly. There are things going on between them and their targets that they cannot see.

 

[davidjoe]

I wholeheartedly disagree with you. Nobody ever suggested accidental improvements.

If there weren't a normal distribution a rifle would ALWAYS shoot 0.200" for example. Not 0.198", 0.200", 0.205", 0.175" etc. The more precise a rifle, the smaller that standard deviation is however that doesn't mean it doesn't exist.

What I’m saying is that the “distribution” is not symmetrical, Shooting for group is not an activity that lends itself to “normal distribution.” A bell curve is symmetrical on both sides of the peak.

Word choices aren’t the focus. If I could only reliably string together 2 free throw shots through the basket, instead of three, and I had to start over, and every time I start over I talley a mark, (on the side of whatever number applied), I’d have an even more skewed ratio of results on the poor side of 2 or more baskets.

The graphed results of a skill activity like this on an X Y axis is a tiny bump, then a steep downward slope, not a symmetrical bell curve. Now, do a lot of shooters being tested, fall into a bell curve, perhaps, but that’s a separate issue.

 

[RegionRat]

It doesn't help our situation when folks don't use common terminology.

For example, I think the goal of all of this recent Litz-Cortina discussion, is to demonstrate a muzzle tuner 1) will "work", or 2) will not "work".

But, we still saw Bryan and Eric spend several minutes where their definition of "work" was the problem.

I'm not hopeful we will ever get to the goal if we can't agree on the definition of what it means to say a muzzle tuner "works" or it does not.

I think (know) this debate will require setting a very narrow context in terms of the gun, the ammo, and the procedure.

I would like to think it is fair to say we are no longer debating how to measure group dimensions,... but the test duration or number is shots isn't agreed upon or set so I should not even make this assumption.

There is a good reason this will not be "easy". It is a daunting task when the vast majority of shooting facilities are not in a controlled lab or gun tunnel since wind is always a factor. The folks on the guns will have to be proficient.

We have many threads where folks tend to drift on math, statistics, metallurgy, etc., and this is normal for a public forum where anyone can comment regardless of their own personal proficiency with the topic.

However, here we have another chance to limit the context to one where the available experts are free to set the definition of the type of shooting (number of shots), the gun, and the ammo.

The folks who agree to spend the effort should be the ones to have their say in setting up the definitions and test procedures.

Make no mistake, it is predictable that some folks can ruin an anvil... so some folks who load and shoot can follow a procedure and demonstrate tuning, while others cannot.

To end the debates, the results will have to repeat for at least a few proficient shooters willing to run the procedure and share their work publicly.

The climate is hard to arrange and that ends up being the catch.... is part of the tuner discussion that preloaded ammo or ammo loaded in real time needs to "show the best groups" for the tuner demo to "work" and prove the benefit of the tuner is it can improve group size due to climate or ammo changes?

If we can contribute without spending money and ammo.... it would be to help set up the test definition, context, and procedure. I tend to leave this to folks like Mike, Eric, Bryan, etc. who have a stake. After all it is their money and time. I would suggest we don't need to debate statistics or group measurements, but it is an open forum.... On the other hand, a definition of what it means for a tuner to "work" is a start... YMMV

 

[drop_point]

What I’m saying is that the “distribution” is not symmetrical, Shooting for group is not an activity that lends itself to “normal distribution.” A bell curve is symmetrical on both sides of the peak.

Word choices aren’t the focus. If I could only reliably string together 2 free throw shots through the basket, instead of three, and I had to start over, and every time I start over I talley a mark, (on the side of whatever number applied), I’d have an even more skewed ratio of results on the poor side of 2 or more baskets.

The graphed results of a skill activity like this on an X Y axis is a tiny bump, then a steep downward slope, not a symmetrical bell curve. Now, do a lot of shooters being tested, fall into a bell curve, perhaps, but that’s a separate issue.

And this is why you're misunderstanding what Litz is saying at its very core. He is talking about the normal distribution of a rifle's performance; not human performance.

 

[Keith Glasscock]

If I was in this “conversation”, and mentioned using my tuner to induce about a bullet hole worth of vertical in my 30BR VFS Rifle to get out of a horizontal tune, I bet both of these guys would look at me like I was nuts.

Don't tell them that! I do the same thing with seating depth. It's served me well in F-class.

 

[davidjoe]

And this is why you're misunderstanding what Litz is saying at its very core. He is talking about the normal distribution of a rifle's performance; not human performance.

I tend to credit good groups as indicative, even proof of gun capability, and at least for that moment in time, shooter and ammo as well, with bad groups requiring further exploration. - The recent stir being that we are settling on just a slice of the full range of what that gun will shoot, without knowing where in that range we happened to sample.

My thought would be that if the shots we (shooter and gun) fire aren’t going to necessarily be on that bell curve per se, at least the type that is being presented anyway, then there is less worry that one is seeing an anomaly, and a large sample size shot to try to flesh out exactly where on that bell curve they fell, isn’t necessary.

Even if we are just talking about the gun and not the shooter, I’ll analogize again to try to make it less abstract, - a tight group of a reasonable number of shots, from a quality build, is about the same level of proof as a dynamometer certification is on a car with a custom engine. That test is not a benign operation, and while you might be able to get slightly different readings repeating it, if one were to let that bother them a great deal, they would experience diminishing returns determining the range of readings attainable.

 

[Keith Glasscock]

Well this is one of the most bothersome bits of voodoo and witchcraft that I’ve encountered, but because it strikes me as too asinine for me to waste my time testing, I’ve never bothered to prove it one way or the other and didn’t want to bring it up specifically.

Of course you’ll find a flat spot in velocity if you use small enough increments in powder charge! They usually recommend increments that should cause velocity changes well within a decent ES. Lol. The question is, can you test it 25 times and land on the exact same powder charge almost every time? Maybe. I haven’t tested it because I see absolutely no reason why it should work, and what I do already works as good as I need it to. I suspect it could easily be proven wrong, but I haven’t bothered to try, and can’t promise that it doesn’t work. Even if it does work, I CAN PROMISE, that doing it at just 1-2 shots at each powder charge doesn’t give you the ability to distinguish whether you hit a flat spot or whether your random velocity variation caused it entirely.

Agree,

I've done 10 shots of each charge, and magically, no flat spots! Physics doesn't change just because someone observed something one time with far too small of a sample size.

 

[BoydAllen]

Those who have the most experience with the successful use of tuners, using them where small differences in average accuracy essentially are what the sport is all about, believe that they work. If others choose to question that, it really is not important, unless you are in the business of selling tuners....or books.

 

[Keith Glasscock]

Just a little reality check. Do you hear match winners going on about statistics? In my experience they have done a lot of work figuring out how to make their rifles shoot, and how to shoot them consistently. Barrels have very limited lives. Because of that we have to learn to deal with the limitations of small sample sizes. The proof that this can be successfully done show up in the performance of top shooters. It is my belief that the vast majority of shooters do not own a set of wind flags. For these same shooters to talk about small variations in accuracy as if they are the result of their reloading just seems silly. There are things going on between them and their targets that they cannot see.

^^^^ Pure, unadulterated truth in F-class!

The winners only speak about statistics in an effort to mentor the newer shooters out of the cycle of endlessly trying to improve tune instead of improving themselves.

 

[Straightshooter1]

As mentioned in the Litz/Cortina discussion, there's lot of things involved in what determines group size. When we're getting small group sizes, all those things are working in harmony to that end. You can have all but one thing working well, but it's that one thing that screws it all up. Finding out just what that one thing is that needs the attention is quite a challenge, which Litz rightly point out. If all the components, less a barrel tuner are tuned to give the best results, adding a tuner to further address harmonics seems right as Cortina testifies to.

Given all the variables, setting up testing methods that can quantify what a tuner can do or not, is a tough challenge. But I feel if the right testing technology is used, this issue of how much a tuner can help or not is quantifiable and most likely best done through mathematics. Otherwise, it feels to me like we're all chasing our tales.

 

[Bill Norris]

Just a little reality check. Do you hear match winners going on about statistics? In my experience they have done a lot of work figuring out how to make their rifles shoot, and how to shoot them consistently. Barrels have very limited lives. Because of that we have to learn to deal with the limitations of small sample sizes. The proof that this can be successfully done show up in the performance of top shooters. It is my belief that the vast majority of shooters do not own a set of wind flags. For these same shooters to talk about small variations in accuracy as if they are the result of their reloading just seems silly. There are things going on between them and their targets that they cannot see.

I believe Boyd has hit the nail on the head. The life of a barrel when opposed to the duration a shooter will spend shooting groups is too short to even worry about statistics. A shooter may carry over a given load from one barrel to the next but will never be able to carry over the characteristics of that barrel to another so why bother with statistics ? A given group from one barrel means nothing to the next barrel. It most often will require some tweaking of the load whether using a tuner or not.

 

[drop_point]

I tend to credit good groups as indicative, even proof of gun capability, and at least for that moment in time, shooter and ammo as well, with bad groups requiring further exploration. - The recent stir being that we are settling on just a slice of the full range of what that gun will shoot, without knowing where in that range we happened to sample.

My thought would be that if the shots we (shooter and gun) fire aren’t going to necessarily be on that bell curve per se, at least the type that is being presented anyway, then there is less worry that one is seeing an anomaly, and a large sample size shot to try to flesh out exactly where on that bell curve they fell, isn’t necessary.

Even if we are just talking about the gun and not the shooter, I’ll analogize again to try to make it less abstract, - a tight group of a reasonable number of shots, from a quality build, is about the same level of proof as a dynamometer certification is on a car with a custom engine. That test is not a benign operation, and while you might be able to get slightly different readings repeating it, if one were to let that bother them a great deal, they would experience diminishing returns determining the range of readings attainable.

The point is, whether or not it is necessary to know where that shot falls on the bell curve doesn't insulate it from the fact that it is still part of the normal distribution. Normal distribution is a statistical tool, whether your belief is there doesn't change things. You're over-emphasizing the need of people to plot it, which isn't the point of Litz or the guys of Hornady. That wasn't the point at all, in fact, the purpose was awareness.

With your example of a car engine, the dynamometer is used to set a benchmark. I don't believe shooting one group would be considered a benchmark in any shooting circle. However, we can all agree that engines and barrels wear and the more precise a rifle is, the smaller the inherent variability of group size.

 

[ImBIllT]

I’d say that Brian’s statistically large sample pertains to “tweaking”, not “testing.” The first focuses on improvement of one individual gun’s nominal performance, and the other on how that individual compares to a standard, or to others like it. Testing, like proof marks, assumes a finished product to be either accepted or rejected and tweaking assumes a work in progress.

Brian and Hornady have created a stir, imo inadvertently blending the two concepts going after the . ___ moa labels makers and competitive shooters casually use to relate where their guns stand. For example, that any rifle “X” ships came with a .5 moa or better target, is a passed “test” of being a plucked and shot .5 moa gun. The test is passed if it “can” shoot .5 moa, not if it has been proven to many times. Erik often says he needs a four inch or .4 moa at a thousand, gun, etc.

The problem I have with very large sample sizes as to testing is that it destroys the subject, and as to tweaking, that “real improvements” shouldn’t be that hard to discern, and if so, then maybe they aren’t. Everything has very limited life, in what we do. When tweaking, we truly are testing this “five” rounds against the next “five” (ten individual bullets and powder draws) with a slightly older, hotter barrel, making it hard to even discern what accounts for differences, and likely it’s not any more the rifle, as opposed to the ammo, hold, rest, or environment.

So does large sample tweaking help, if small groups are unreliable? Brian points out that load testing is often not repeatable. The best isn’t the best, next time, and noise can’t be overcome. Theoretically the aging barrel is a moving target, and by way of example every 16 inch naval round was sequentially numbered and modified, to mirror barrel wear. I’ll credit Brian for saying for many years, don’t carry this too far, the wrecking of barrels once differences have become minute.

“I said larger differences require a smaller sample size” in reference to your comment about the military sorting through shooters rather quickly. I’m sure the top 1% makes itself known rather quickly. They’re also competing against “everyone” not just those interested in competitive shooting, so again, I’m sure sniper material presents itself quite quickly.

I mean large differences in RESULTS require smaller sample sizes, not large differences in METHODS.

I didn’t watch the video. I don’t have to watch the video to know that some things that grab the minds of shooters, particularly short range benchrest shooters, are little more than a figment of their imagination.

I don’t know if tuners “work”. I assumed they did because I’ve never heard anyone at the benchrest club that I’m a member of say anything indicating that they didn’t “work”. I’ve heard discussions of what exactly it is that they do, and how to best make use them, but I just assumed that they are doing something.