Thread: Seeing the light: A load development case study

Originally Posted by gimp

"Insanity is doing the same thing over and over again, and expecting a different result" - Albert Einstein. maybe.


I'd like you all to meet Dennis.

Attachment 279526

Dennis is an older gentleman, based somewhere in the Southern states of the USA, perhaps Georgia, and the mastermind behind the "Everyday Reloading and Shooting" youtube channel.

https://www.youtube.com/@EverydayReloadingandShooting

Dennis owns a Tikka T3 .223, and a Sako S20 6.5 Creedmoor, and Dennis is a load development enthusiast. Dennis has undertaken many thousands of shots of load development, following conventional tuning processes, videotaped the results, and uploaded them to youtube for everyone to enjoy. This creates, if you can ignore his tortuously-absent-any-factual-basis mis-use of mean radius, a treasure-trove of data for a case study on understanding the effectiveness of those tuning and load selection processes for actually identifying whether any of it makes any difference to rifle precision.

It isn't clear for exactly what purpose Dennis is doing all this reloading. He focusses on creating a series of test loads for some particular combination of components and load parameters, shoots those, then reports back on what results were observed, with a wide range of interestingly meaningless metrics.

Let's have a look at some of these. I highly recommend that you absolutely do not watch these videos and instead read my summary.

6.5 Creedmoor Sako S20

Dennis first determines the distance to the lands for his rifle. He then, using what appears to be a somewhat arbitrary charge weight, seats a range of loads at 0.010" increments deeper starting at 0.020" off the lands. He's using high quality components - lapua brass, H4350, 140gr ELDM, and appears to have reasonable quality reloading equipment. 10 groups of 3 - 30 rounds.
https://www.youtube.com/watch?v=-xBHIKYr6GE

Dennis then goes to the range and shoots these rounds !
https://youtu.be/8c0CnAeNgEU?si=UV4Vfq2BsWH8cAzf
Dennis seems fairly happy with his results.
Attachment 279527

What are these results?

Well, his first 3 shots are "barrel warming and fouling". 0.46" group more or less centred on the X- and Y- axis, just a fracco left. He adjusts 2 clicks up from here. When he fires another 3-round group of the same load, it looks about 2 clicks up, but it's moved out to the right about a click. That's pretty normal - POI wanders around a fair bit between 3-round groups, even in a windless indoor range.

Attachment 279517

However enough about accuracy, let's look at the group sizes and talk precision

Attachment 279515

The variation above and below the mean across these is about plus and minus 70%. That's about what you get when shooting 10x 3rd groups of exactly the same load...

Attachment 279516

Regardless of that bit of foreshadowing, Dennis decides that he might be seeing some nodes at 2.872" and 2.812", and decides they require a bit more testing. Fair enough.

So it's off to the range again, this time to "zero in on the node".
https://youtu.be/J_9e6SEOqHo?si=vNFmNiVOgVk48YiA


First up, Dennis re-tests the "node" that produced a 0.24" 3-shot group - 0.020" off the rifling. He does, admirably, test this with a 10 round group.

Attachment 279519

For some reason he's mis-labelled it vs his commentary - data management is a bitch - but unfortunately for Dennis, his tuned node at 0.020" jump doesn't produce what he's looking for - just a 1.28 inch 10 round group. Astute observers at this stage might look at the mean radius of 0.43 inches, multiply it by the correction factor of 4.16, calculate that he should expect a 95% dia of 1.78 inches plus or minus about 20%, note that all his previous shots from testing fall into that diameter centred around a common MPOI, and reconsider the approach or expectations. However, not our man Dennis.

Dennis gets onto his next investigative step - testing a range of seating depths in 0.003" increments either side of the other previously identified, um, node at 2.812". No, it isn't clear why he tested one node with a 10rd group and the other with a series of 3-rd groups. What is clear however, is that he's now 63 shots deep in this load development (including foulers) and hasn't yet worked out what he can expect from any of these loads in the future.

Attachment 279520

For some reason his POI has shifted a bit. Alarming. He doesn't seem terribly concerned that the seating node he was testing (2.812") produced a group 50% larger this time than last time (0.45" vs 0.3"), but gets on with it and decides that in fact, 2.815" is the magic seating depth that requires more testing. Happy days.

Dennis loads up a few of those, except for some reason seated 0.001" deeper, and heads back to the indoor range. This time he's framing it as instructional - "how to find the node". Good stuff, he's done a lot of shooting so surely he has found it and can tell us how.
https://youtu.be/anwzFsFGiw8?si=QoLYRRqJ2Y1XeTmI

He bangs out a 3 round fouling group, then a 10 round test group. Nice work Dennis. That's shots 64 to 76 on this load development process if anyone is keeping score. Let's have a look.

Attachment 279521

Hmm. At a group size of 1.18 inches and a 0.39 inch mean radius, there's no evidence that it's any different from the other seating depth above. Bugger. Back to the drawing board for Dennis - spoilers - he's going to have a crack at length-sorting bullets.

This next range trip isn't a very long video. Dennis bangs out a few foulers and one 10rd group. And it's looking pretty good, actually, except for....
Attachment 279522
...yes, Dennis is very tempted to call that one a flier. Who wouldn't be, after so much effort and 89 rounds of development? The fact remains that if we don't exclude that shot, it is a 1.35 inch group, and the mean radius is inside the range you might expect based on prior groups. Any of these metrics will see some variation around the mean with only 10-shot samples.

Now the chronology seems to get a bit funky with this series. Dennis refers to a video in which he shoots a great 5-shot group with this load. This actually happened much earlier than the rest of this series and it's an odd anachronism. But it appears that some horrible sceptic named Warren (my name is not Warren) challenged him to shoot another 10-rounder with it and see what it looks like. To his credit, Dennis is not afraid to meet this challenge, although it's sort of odd, as he's done a heap of load development since that.
https://youtu.be/wQdPsXPMq4I?si=ZF8SYzSpZFKMG25x

The 5-shot group that inspires this, is this group - it's a quarter MOA. It really shows what the rifle can do, doesn't it?
Attachment 279524

So off he goes - back to the indoor range, armed with the usual fouling shots and enough finely loaded quarter-MOA precision rounds, length-sorted, for a screamer of a 10-round group. Let's see how that goes.
Attachment 279525

Well, not a quarter MOA. However, Dennis seems pretty philosophical about it and takes it in his stride. Which does make me wonder for exactly what purpose he's doing all this. It's not clear where to from here with this one for Dennis, but we've now got a record of at least 109 shots of load development, a bunch of different tuning and testing nodes. Which has led Dennis to shoot 4x 10 shot groups, averaging 1.28" and varying about .1" above and below that, which is actually more consistent than you'd normally expect for 10rd groups. All the 3-shot groups fired during the testing process fall inside the general cone of fire described by the 10 rounders.

Interesting stuff, isn't it? And I didn't have to spend a single cent on powder.

Who's interested in some more Reloading Adventures with Dennis?

Nice exposition.

I missed the part where our fellow enthusiast saw the light. Unless of course you meant that you saw the light, and are planning to follow Dennis' lead

Dennis is simply enjoying himself. Faair enough imo. Free speech etc. Personally I find the leads provided by expert load developers and projectile manufactures such as Sierra and Nosler, which are freely available. Minor tweaking of these can get you there using this data. Better to lractise shooting than spend a lot of pingers wearing out a barrel for a 1/4 moa gain when using a fully supported rifle. I don' know any hunter who can hold better than a half moa off a pack.

Originally Posted by Friwi

Yes out of the whole content there is about 12 to 15% that does interest me. And I have been playing his videos at 1.25 to 1.5X the speed.
I found interesting the jump of the different groups from one place on the target to a different point of impact despite having kept the same zero. And his explanation to try to find a more stable group that works across a larger range of temperatures ( and/or speed variations)And also his results at 600yrds where his best groups are not necessarily his most consistent in terms of speed. I was interested to see others opinions.

I've just watched the poi video.

Theres a bit to unpack there.

He really overcompkicates things tbh. He also puts significant emphasis on analysis from 3 shot groups. It honestly doesnt stand up.

The poi point where he talks about wandering poi is really just random noise from 3 shot groups inside his cone of fire. Overlay the 3 round groups and compare it to the first 10 round group and I'd put money on it matching the cone of fire with an expected increase in total group size that accounts for the larger sample set.

With the second 3 shot ladder aggregated the average apoi between the 2 ladders will be very close. The poi isnt actually jumping around, it's just noise from the random placement of the 3 shot groups inside the overall cone of fire. He talks about the Hornady info but really does seem to miss this point.

He draws some odd conclusions analyzing the 3 shot groups, and assigns poor consistency to temp or speed changes as opposed to the much simpler explanation that the rifle just doesn't shoot that combo well, and results will always be poor because of it.

Interestingly he kinda ends up coming to the right conclusion just with pretty flawed logic, and the most long, roundabout component heavy process.

So logically, with a powder and bullet combo that your rifle likes, you will have less dispersion over all, and a smaller cone of fire. So following his method shooting 3 shot groups that fall randomly into this smaller cone of fire you will see Both a smaller average group size, and a a smaller apparent poi shift. Everything is just closer together due to it being a better load, randomly placing holes inside a smaller area.

This appears stable across different loads because it is. Charge weight doesnt have a large effect on group size. And less overall dispersion means lower extreme spreads at both the top and bottom of the table.

He makes it really complicated, adds more variables than needed, makes a bunch of assumptions from 3 round groups then somehow comes to the right conclusion.

You can save yourself all the drama and just shoot a couple of 15 shot groups with known good combos for the cartridge at a grain under max and achieve exactly the same results

I've just realised what the problem is, He's not shooting a 243

Originally Posted by 7mm tragic

I've just realised what the problem is, He's not shooting a 243

Is he looking for 8 inch groups/patterns

That was an interesting case study @gimp.
Do you think you could do the same with that Little Crow Gunworks precision reloading series he just did?

@gimp best read I've had in a while, thks.

I confess not to being Dennis (or Warren), but I can only thank that Hdy article for saving my bacon. For some years I had the expectation/belief that a accurate rifle would consistently print 3 touching holes on paper at 100m any time I did my part of only I could find the right load/node. If such a rifle exists I probably can't afford it. Of the ones I've got, most of them shoot better than I can.

Originally Posted by Jhon

@gimp best read I've had in a while, thks.

I confess not to being Dennis (or Warren), but I can only thank that Hdy article for saving my bacon. For some years I had the expectation/belief that a accurate rifle would consistently print 3 touching holes on paper at 100m any time I did my part of only I could find the right load/node. If such a rifle exists I probably can't afford it. Of the ones I've got, most of them shoot better than I can.

Jhon what's this Hornady article/podcast that everyone keeps referring to?

Sent from my SM-G973U using Tapatalk

This guy is fun to watch https://www.youtube.com/@JohnnysRelo...ench/playlists same sort of thing. It is just Johnny is funny.

I think there is a youtube business model around doing this sort of thing, there have been a number of similar content creators pop up over recent times doing this.

Pretty cool if the only other thing you would be doing was working out what you were going to have for dinner.......

Could be this one @Eat Meater
https://open.spotify.com/episode/4Xe...SSGmTHS5Sp46CQ

Or youtube Ep 50 and 52 of the Hornady Podcast

Dennis is back! and today, Dennis has generously created a series looking at the 60gr Vmax, with varget (2208) in his Tikka T3 .223, mounted in an MDT chassis. Oh, a flat base bullet. Surely this will shoot well.



https://www.youtube.com/watch?v=_JJrqPsLcLA

This time around, Dennis is starting with powder charge tuning. He's started at book max and gone down in 0.3gr increments for powder charge. It isn't clear where he's seated the bullet, perhaps we'll find out in a subsequent video. He's shooting 5 round groups today - let's see how he gets on.



Well, he's seated out to 2.25" OAL, so that answers the seating depth question. He's a bit worried and confused about pressure - loading up to 24.9gr per the Hornady manual, but Hodgdon go up to 27.0! Don't worry about it too much Dennis, it's a 223 with 2208. You run out of case space before you run out of pressure, with any bullet below about 80gr.

So the groups run a 0.7 inch average, with variation about 50% above and below that average - roughly what you'd expect for normal variation in 5 shot groups of a load with identical precision. This is a less precise average as we only have 5 data points to work with, but the principle continues to hold true.



Dennis seems happy enough and is off to look for a node between 23.7 and 24.2 grains, somewhere.


Next time, Dennis is actually not looking for a node between 23.7 and 24.2 grains. Dennis has sacked up about pressure and is shooting charges from 25.3 up to 26.5 grains ! In 0.4gr increments now. Apparently there may be some extra nodes in there.
https://youtu.be/rSMgo99Np-I?si=hZ25FRqSJ2J5dGtX

He's run out of case capacity at 26.5gr, below the book max of 27.0gr. 5 fouling shots and 5x 5rd groups later -



Seems like pressure wasn't really a problem. His groups are ok. Let's add them to the table and see what it looks like



Ok, the average group size across all these 5 shot groups has now increased. Is it because precision is worse with the higher charge weights, or is it that precision is functionally identical across all charge weights, and the larger sample of 5rd groups now has just provided a more precise mean group size? Who knows. What's certain is that if the higher charge weights don't provide as good precision, you'd want to know that pretty early in the process with some confidence, because it's hard to imagine a functional use for a 60gr vmax doing 2600fps at the lower charge weights - so why would you even test them?

Anyhow, Dennis doesn't agree with me here and has decided to write off the higher charge weights and go back to his lower charges to see if he can find that elusive node.

So, back to the range with, once again, some 5 round groups loaded up - testing at 0.2gr increments in the 23.6-24.4gr range, although skipping 24.0 for some reason. Odd.
https://youtu.be/plCTmMwATwk?si=RoY6GYXHgSTvdr9q

Dennis draws his usual pictures, and sits down for a yarn. Note he's now at 80 shots fired for this process.



Dennis is happy with the groups in general, although a little baffled by the low speed. He decides that 23.6 (0.6" group) and 24.2 (0.49" group) both look good, and he's going to have to test with some larger sample sizes. !!

That'll be exciting. Let's see how it goes.

Back at the range, with 10rd each (nice) loaded at 23.6gr, 23.7gr, 24.2gr, and 24.3gr. Plus the obligatory foulers.

https://youtu.be/5sTkLNKMta0?si=04LymfptvnsV5amd

If you haven't been keeping track, I made a neat little table.



Dennis has also switched to Lapua brass. That's an interesting move, to assume that seating depth and powder charge tweaks make a difference to precision, but that effect is disconnected from brass. He's hoping to get his SDs down - improve velocity consistency - although it's not clear why, as velocity consistency at short range is uncorrelated to precision, and he doesn't seem to be developing a long range load.

Never mind - what does it shoot like?


Dennis astutely notes that he doesn't see a tremendous amount of difference in the groups here. Let's have a nosey at the measurements. We'll ignore the fact that his velocity has changed pretty substantially. His SD's look about normal for any .223 load. Lapua brass may not have helped there - but we don't really care about those right now. Maybe another day.



So functionally you'd say that there's little evidence any of these loads produce different precision. The variation of the 10 shot group size around the average is well within what you'd expect from 4 groups of a load with the same precision. The mean radius, being a more efficient statistic, is more stable than the group size - and the variation is still sort of all around the average.

Dennis uses a tortured calculation unnecessarily manipulating the raw mean radius value through doubling, adding to the group size, and dividing in half, to determine that the 2 groups with the smallest MR values are probably the best, and probably indistinguishable.

My conclusion is that any of the 4 loads produces a mean radius of about 0.3 inches (given measurement error it's a bit silly to go below 1 decimal place). Multiply that by the correction factor of 4.16, and you get 1.25 inches as the probable diameter that 95% of shots will fall into for any of the loads. You have to say that's probably plus or minus a bit given that 10rd isn't sufficient to get a precise MR value. So he's got, lets say, a 1.1-1.45 MOA true cone of fire rifle/load system - although it's already shot 10rd groups over 1.1 so we know it's larger than that. All the 5-rd groups tested fall within that diameter around the MPOI.

Does that precision meet his requirements? Did he need to fire 125 shots to get there?

Originally Posted by Roarless20

That was an interesting case study @gimp.
Do you think you could do the same with that Little Crow Gunworks precision reloading series he just did?

It's like 15 hours of him talking very very slowly ! so.... maybe

Originally Posted by gimp

It's like 15 hours of him talking very very slowly ! so.... maybe

And sniffing his own farts, that's the bit that gets a bit painful