I'll Post This Here - Hornady's Podcast #50. I thought it was one of their best, but some reloaders might not like what they see....

[JAS-SH]

Hornady podcast #50. Another in their ongoing long series on ballistics. This one focuses on things like SD and Velocity stats that reloaders tend to focus on, and why the statistics tend to prove that those things might not be as relevant as we all think....

A hard pill to swallow for some of us, but here it is.

 

[John Glidewell]

I get the sample size part, simple math. But the rest of it was just mumbo jumbo bs. I do appreciate seeing more of this type of stuff being presented nowadays esp by some top end shooters. The real problem with much of this is most reloaders do not have a seasoned process in the load room or the required skill to shoot the difference. Find a safe max and back off 1.5 to 2 grains. Every rifle will shoot in there. And if you run Quick Load you will see they almost always line up with a barrel time. It just does not need to be a complicated process and require multiple 50 shot strings to proof.

 

[England]

Superb presentation--many thanks. One can develop an accurate load in a quality rifle without any understanding of statistical analysis. But statistical analysis can both make the process more efficient and prevent misconceptions about the potential future accuracy of a load. With chronographs all providing a "velocity standard deviation", one does have to have some understanding of the normal distribution and statistics to appreciate WHAT that standard deviation "means" and that a sample of 20-30 tests is required for the SD to have predictive value. Handloaders really need some understanding of basic statistical analysis, as Bryan Litz has presented in his recent book. With our historical reliance on 3, 4, or 5-shot groups, handloads have long hallucinated meaningful patterns in the statistical noise of data, both with group size and "nodes" in ladder testing. Ultimately we are limited by the cost of testing and using up much of the 3000 round lifespan of many barrels. But the application of appropriate analysis can be a great aid, even without studying Box's "Statistics for Experimenters."

 

[RegionRat]

In a nutshell, the first 38 minutes shows what happens between two different rigs as the number of shots in a group increases and the total number of shots sampled increases.

At 51 mins, they discussed the number of shots used to establish boresight zero.

At 1:00:00 they walk through the cost of making the mistake of only using 5 shots to check a zero, then walk though the stack of errors that creates as the number of shots are wasted at longer distances.

In the above example, the shots were going low at distance, so the ballistic dope was incorrectly adjusted and caused 49 shots to be wasted. The arrows are a decision path that shows what would have been discovered if the 100 yard zero was re-checked... only to discover that the real error was caused by an undersampled zero that was actually lower.

At 1:12, they mention mean radius is applying the weight of all your shots, versus group size being similar to just using the ES.
At 1:17 they each discuss their new rifle load seach/development process.
Miles- Jumps 25 to 35 and "forgets the die is adjustable", meaning he doesn't do depth searches. He then shoots 10 shot groups with powders and bullets, and then goes back and takes the best and shoots 20 shots. If nothing works, he hucks the barrel.
Jayden - Selects bullets for purpose and distance, picks powders that should get the speed required, isn't very concerned with jump as long as it fits the magazine, quickly eliminates components that don't make the cut while looking for the barrel and bullet combo that meets his expectation.
At 1:20, they relate that they do not find seating depth to be a strong lever in group size.

YMMV

 

[M4orturnate]

Anyone tell the guys winning bench and f class matches yet? I really dont buy most of this. I spent a almost a year of weekends at the 100y line testing and seati g depth absolutely makes a huge difference, five shot groups are a pretty good indication of sd(tho they do need to be retested in different conditions). Theres definitely something to be said for buying the best components. It makes it way easier. But you dont need 30 shot groups to know what your loads going to do.

 

[Ledzep]

five shot groups are a pretty good indication of sd(tho they do need to be retested in different conditions)

 

[M4orturnate]

It is. Ill take 10 charges 5 shots each and pick from that. Very rarely do larger strings show much different. Es will be higher but sd will usually be close. I have seen it where it was a lucky 5 that didnt shoot like that later. Im still using my same charge in my match gun i pivked that way. Chrono before every match and sd is always between 4 and 6.

 

[JAS-SH]

Anyone tell the guys winning bench and f class matches yet? I really dont buy most of this. I spent a almost a year of weekends at the 100y line testing and seati g depth absolutely makes a huge difference, five shot groups are a pretty good indication of sd(tho they do need to be retested in different conditions). Theres definitely something to be said for buying the best components. It makes it way easier. But you dont need 30 shot groups to know what your loads going to do.

I get you point but if you look at the COLLECTIVE statistics of those shooters over time the implied assumption is incorrect. I did my own tests similar to Hornady's tests back in February "22 with a rifle that shoots bug holes at 100 yards with amazing frequency. The results were enlightening to say the least.... Here is that post:

The Truth and nothing But the Truth

Oh, and yes, Someone did. His name is Brian Litz and he focused on both disciplines in his latest book using shooter aggregates during actual competition.

 

[secondofangle2]

Thank you so much for posting. This is one of the most fascinating and enlightening things I have seen. B Litz level stuff, from two knowledgeable guys who by all appearances are both shooters and engineers.

And it makes sense. Powder/bullet combinations are well-recognized. The seating depth stuff was a little surprising but not in retrospect - if seating depth mattered so much, factory match ammo would only shoot good in some guns.

 

[secondofangle2]

It is. Ill take 10 charges 5 shots each and pick from that. Very rarely do larger strings show much different. Es will be higher but sd will usually be close. I have seen it where it was a lucky 5 that didnt shoot like that later. Im still using my same charge in my match gun i pivked that way. Chrono before every match and sd is always between 4 and 6.

"If I close my eyes and refuse to get large sample sizes, and ignore the results of those who did, I confirm what I have always known.""

 

[JAS-SH]

.....And it makes sense. Powder/bullet combinations are well-recognized. The seating depth stuff was a little surprising but not in retrospect - if seating depth mattered so much, factory match ammo would only shoot good in some guns.

Excellent point. I hadn't thought about that one.

 

[M4orturnate]

"If I close my eyes and refuse to get large sample sizes, and ignore the results of those who did, I confirm what I have always known.""

I have 2 years worth of samples size.

 

[spife7980]

- if seating depth mattered so much, factory match ammo would only shoot good in some guns.

It does only shoot good in some guns.

 

[918v]

It does only shoot good in some guns.

One thing they did not mention is that in their large sample testing they used new brass which behaves differently than reloaded brass and then case prep becomes a whole different variable.

 

[JAS-SH]

Here is more rain on the parade of the value placed on chrono reported ES and SD data for reloading accuracy, and large vs. small samples - this from an F-class shooter:

 

[918v]

So they want us to get some bullets and do a ladder and pick a safe load and shoot a 30 shot group .030” off the lands and if it’s too big then switch powders and or barrel. What they are suggesting is that it is not worth trying to fine tune a finicky combination. And that is what I took away from it.

 

[RegionRat]

So they want us to get some bullets and do a ladder and pick a safe load and shoot a 30 shot group .030” off the lands and if it’s too big then switch powders and or barrel. What they are suggesting is that it is not worth trying to fine tune a finicky combination. And that is what I took away from it.

Yes and no.
What they said has been common knowledge in ballistics for decades. I was inducted into the defense industry in 1983 and can tell you everything they said was old news back then.

A different way to summarize the first parts of the talk, is 1) that many folks get mislead by dispersion and statistical behavior of things like velocity and group metrics, and especially the problems associated with making assumptions based on small samples. 2) This talk applies very well to velocity stats, group stats, as well as boresighting. 3) They go a step farther and add in their observations with their experience with seating depth.

I strongly agree with the first two observations and can attest to this after a career of managing weapons systems and chaotic topics. This isn't worth a debate and has been proven to many times. However... The last one is a little muddy with respect to seating depth. The context is important to keep in mind.

A Wiseman test barrel in a shooting machine, a test rig that looks like a sporting rig, a test rig that looks like a match gun, or a benchrest rig in the hands of an expert, may all give a different view of the seating depth issue.

Generalities have a way of stirring up a swirl in the forums because the specific context of the gun and type of shooting matter, yet those details are too often brushed over. The folks who shoot in gun tunnels for industrial purposes are worth listening to. Ignore them and you will loose out, but also keep in mind that they may or may not be in the correct context for all shooting or specific shooting.

There are some elements of truth in both sides of the seating depth arguments, so folks would be wise to slow down and pay attention to things like chamber and reamer details, bullet ogives, etc., with respect to the pros and cons of seating depths. YMMV

 

[Rover31]

Everyone is missing the hidden message from Hornady… Buy more of our bullets… Buy more of our ammo… repeat, this is also sponsored by all the barrel makers, gunsmiths, powder and primer sellers.
Now go get that data!

 

[Ledzep]

Everyone is missing the hidden message from Hornady… Buy more of our bullets… Buy more of our ammo… repeat, this is also sponsored by all the barrel makers, gunsmiths, powder and primer sellers.
Now go get that data!

 

[Gemsbok]

Everyone is missing the hidden message from Hornady… Buy more of our bullets… Buy more of our ammo… repeat, this is also sponsored by all the barrel makers, gunsmiths, powder and primer sellers.
Now go get that data!

Actually I think if we followed their recommendations we'd end up buying fewer barrels, and we'd spend less ammo on load development and zeroing and the same amount more for practice/hunting.

 

[Rover31]

The other thing is, if we follow their data, we all would get better!
My post is satire. That is very good info they put out for our learning!

 

[Huckmeat]

Everyone is missing the hidden message from Hornady… Buy more of our bullets… Buy more of our ammo… repeat, this is also sponsored by all the barrel makers, gunsmiths, powder and primer sellers.
Now go get that data!

You forgot the implicit message of "SD doesn't matter in our factory match ammo"

I have some hornady match still left over from 4-5 years ago, it shoots single digit SD's over 10 rounds, the new stuff is like 35 over 10.

 

[Sparkeytj]

I appreciate the info and statistics they provided in the podcast. I dont agree with alot of there conclusions. Much of the info seems to be things others have tested and proved years ago repackaged by them as if they discovered something. It was stated in the podcast that they have never seen a firearm that shoots under 1/2 moa over 50 rounds when I heard that i stopped taking what they have to say seriously. The benchrest record is well under that, Im not sure but I think the 600yd f class record is probably under 1/2 also. Seems to me its all about convincing customers if your gun shoots 1 1/2 moa its ok everyones does no need to buy better components or better reloading practices. Match ammo is just as good. Things will never improve if we all subscribe to there conclusions, which if I remember correctly were pick a safe powder charge 30 tho off the lands and shoot 20-50 rounds if it doesn't group change powders or bullets, and if that doesn't work the barrel will never group and lower your expectations.

 

[918v]

But there IS something to be said about forgiving loads. If you want to shoot a 225 ELD-M at 2850 FPS and one powder puts 30 shots inside a golf ball and another powder 30 inside a softball (using their load development method) would that not steer you in the right direction? 60 shots vs 200 to figure out a particular powder is not for you?

And that’s before tuning.

 

[England]

I appreciate the info and statistics they provided in the podcast. I dont agree with alot of there conclusions. Much of the info seems to be things others have tested and proved years ago repackaged by them as if they discovered something. It was stated in the podcast that they have never seen a firearm that shoots under 1/2 moa over 50 rounds when I heard that i stopped taking what they have to say seriously. The benchrest record is well under that, Im not sure but I think the 600yd f class record is probably under 1/2 also. Seems to me its all about convincing customers if your gun shoots 1 1/2 moa its ok everyones does no need to buy better components or better reloading practices. Match ammo is just as good. Things will never improve if we all subscribe to there conclusions, which if I remember correctly were pick a safe powder charge 30 tho off the lands and shoot 20-50 rounds if it doesn't group change powders or bullets, and if that doesn't work the barrel will never group and lower your expectations.

I've followed precision shooting for years, but I've missed where 50-round groups are shot for record in benchrest or F-class. Might you point me to this data? I'm sure impressive 5 or 10-round groups have been recorded, but not the statistically far more representative 50-round groups.

 

[6.5SH]

I've followed precision shooting for years, but I've missed where 50-round groups are shot for record in benchrest or F-class. Might you point me to this data? I'm sure impressive 5 or 10-round groups have been recorded, but not the statistically far more representative 50-round groups.

Heavy/Unlimited are 10 shot groups, 5x matches would be 50 shots for the aggregate score.

 

[England]

Heavy/Unlimited are 10 shot groups, 5x matches would be 50 shots for the aggregate score.

So you are saying that 50 consecutive shots are used for group measurement? That would be different than averaging five different ten-shot groups.

 

[6.5SH]

So you are saying that 50 consecutive shots are used for group measurement? That would be different than averaging five different ten-shot groups.

5x 10 shot matches on the same day, yep is different than 50 consecutive shots. It still isn't 5 selected "wallet groups" that can't be consistently reproduced.

 

[Sparkeytj]

Seems like semantics at this point with that. They even admitted to letting the barrel cool down after a certain number of rounds before resuming there 50 shot group. So what you are telling me is if I shoot 10 rounds then stop for an hour to let the barrel cool down is different than doing 5 different 10 shot groups over the course of a few days? I know cleaning the barrel factors into this, my point is the statistical info is great but, hornadys spin to sell bullets lowers the bar people should be striving to surpass.

 

[Barelstroker]

At around 29:15 in the podcast vid is where the Elephant in the room is discussed. The subject of "Group Aggregation" needs to be emphasized & is without a doubt the most important measurement that is never considered when measuring the repeatability of any rifle.
Aggregation shows the entire picture instead of tiny 3 or 5 shot close-ups of the picture.
Without group aggregation, each group alone cannot show the ongoing shot to shot VARIATION because the commencement of each successive group ends the the deviation progression.
What the Hornady guys are trying to get across is that their testing records & CORRALATES all the shots & combines that data to show the entire picture.
The Hornady guys are 100% correct & anyone that disagrees is ignorant PERIOD.

 

[Sparkeytj]

Then explain a benchrest rifle that never shoots over .25 over the course of the entire life of the barrel

 

[Ledzep]

Do you have documentation of a barrel that never shoots over .25 MOA?

Do you suspect that if documentation of such a thing exists, that it is the norm? Regularly achievable? Expectable?

 

[secondofangle2]

Then explain a benchrest rifle that never shoots over .25 over the course of the entire life of the barrel

Are you accounting for wandering zero when your status? Like is the average distance from the point of AIM to the point of impact for each round under half MOA for 50 shots ?

 

[Sparkeytj]

I guess never is an unnecessary absolute and not the point. As I stated before the information they provided is useful I just dont believe the conclusions they provided. If you want to do load development as they layed out go for it and explain to me how you can improve the process. While you do that I will find a benchrest shooter that has "proof" of all the groups shot on a specific barrel.

Norm,regularly achievable or expected doesn't matter in this conversation

 

[England]

5x 10 shot matches on the same day, yep is different than 50 consecutive shots. It still isn't 5 selected "wallet groups" that can't be consistently reproduced.

I'm not trying to be tendentious, but for statistical analysis you have to be precise in defining what and how you are measuring. But taking the average of five 10-round groups will discard the "outliers" and not provide the detail/accuracy of a true 50-round group. It "trims off the tail" of the bell shaped normal distribution curve and makes the results appear better than they truly are. What they present in that video was very basic stuff from Statistics 101 and was well proven more than a century ago. Much like the Gambler's Fallacy, it may not be intuitive but is absolutely real.

 

[Sparkeytj]

You missunderstand they score benchrest based on agg not average

 

[England]

You missunderstand they score benchrest based on agg not average

Could be. Walk me through it. Let's say a shooter has a ten-shot group 1/2" in spread. The second ten-shot group is also 1/2" in spread, but is offset 1/2" to the right (if the targets were superimposed). Would this be "scored" as a 1.5" group in total (as if the groups were superimposed), or would it be scored as 1" (for 1/2" plus 1/2")?

 

[Sparkeytj]

I dont think they superimpose the groups Im sure someone with much more experience than me can explain it better. So Im sure the spread may be more. From a logical prespective if there was a 1/2 group dead center then an 1/2 group offset it seems to me it isnt the system that caused the offset its more likely its the change in the actual setup of the rifle or the wind nether of which has anything to do with what we are talking about

 

[lumberjack78]

No, they don't superimpose, they just average the size of all the groups.

 

[6.5SH]

Finished the video and honestly am not understanding the hate for what was said. The more measured and recorded data you have and base your adjustments on (in reloading, scope adjustment or a solver) the better and more predictable the results will be over time.

They cover the outliers and getting lucky that you did dial things in exactly and perfectly zeroed in 5 shots. That isn't the norm. Is seen regularly here when people can't figure out why their solver doesn't line up.

We are closer than ever for an interested individual outside of a government laboratory to gather data (velocity, gun movement and position) for every shot they fire to see these larger trends. For fixed distances the addition of an electronic target adds even more information.

 

[RegionRat]

This is a match report that includes many words that are beyond the current questions, but it includes descriptions of the type of groups and scores at a well attended match to help illustrate what some state of the art shooting results are running these days, outside of a gun tunnel at a bullet/ammo factory.

Folks, keep in mind that a shooting machine with a test barrel inside a gun tunnel is one thing, and a highly skilled competition shooter outside in wind is another. Both scenarios have their challenges, but you must study a little of both situations to learn how much of each set of results should be used to judge the potential state of the art.

ETA: I see my links were not working so I have attached the files as PDF.



Let's try and keep it real and civilized. YMMV

 

[kriller134]

How do they account for their exploding bullets?

 

[JAS-SH]

Then explain a benchrest rifle that never shoots over .25 over the course of the entire life of the barrel

That's patently not true.

I shoot in a club where the majority of good shooters shoot benchrest competitively. I have seen their groups. Yes, I have seen 5 shot groups at 100 yards go in the same hole but that is the exception, a big exception at that, rather than the rule. And at 100 yards, with good shooters and calm wind, your .25 number is a somewhat valid number. It's close, but not always. More later.

Moreover, benchrest is a completely different shooting sport, won by precision, not accuracy. Shooters go through great lengths to achieve that precision, including having the absolute minimum interaction with the rifle - a key point.

Later: I mentioned this before above. In his new book Brian Litz went to benchrest and F-class matches and collected and software mapped shooters accuracy and precision, both for wind and elevation. Look at his book, it's in there, it's worthwhile to know.

 

[Sparkeytj]

I stand corrected disregard everything I said

 

[JAS-SH]

I'm not trying to be tendentious, but for statistical analysis you have to be precise in defining what and how you are measuring. But taking the average of five 10-round groups will discard the "outliers" and not provide the detail/accuracy of a true 50-round group. It "trims off the tail" of the bell shaped normal distribution curve and makes the results appear better than they truly are. What they present in that video was very basic stuff from Statistics 101 and was well proven more than a century ago. Much like the Gambler's Fallacy, it may not be intuitive but is absolutely real.

And in support to your post, they did also mention that the US military has been doing large, very large samples for ages. It makes sense when you're average purchase is over a million rounds of a particular ammunition. They do the same for every system too - from pistols to carbines to artillery to missiles. Less numbers for larger stuff obviously.

Their groups sizes are in the hundreds of rounds. I believe the measure they use is a statistic called CEP - Circle Error Probable.

 

[JAS-SH]

Seems like semantics at this point with that. They even admitted to letting the barrel cool down after a certain number of rounds before resuming there 50 shot group. So what you are telling me is if I shoot 10 rounds then stop for an hour to let the barrel cool down is different than doing 5 different 10 shot groups over the course of a few days? I know cleaning the barrel factors into this, my point is the statistical info is great but, hornadys spin to sell bullets lowers the bar people should be striving to surpass.

You missed something from the podcast. When they said ten shots and let the barrel cool down, it was because 10 shots is the average number of shots in each station for PRS matches, and not mentioned by them is that it's the same # of shots for benchrest scoring. Point was, concentrate on what discipline it is that you're going to shoot, including hunting, which came with a lot of other caveats.

 

[Sparkeytj]

And my point is it seems like a whole lot of gaslighting. I may have used the wrong example to get my point across but I will again state my point which is the accuracy numbers they use in there podcast dont match my experience.

But I can admit when Im wrong you all have converted me all the hard work i have put in is pointless. My hit percentage is because of my rifle not the fact that I suck. It will make my whole life easier when I lower my expectations, and the way they do load development is stupid easy so I should just do that too.

I dont have access to the applied ballistics software right now but if someone does id be intrested to see the difference in hit percentage between a 3/4 moa rifle and a 1 1/2 rifle at 600 with a 10 mph wind. Hornadys may have just single-handedly explained why Im not national champion.

 

[England]

And my point is it seems like a whole lot of gaslighting. I may have used the wrong example to get my point across but I will again state my point which is the accuracy numbers they use in there podcast dont match my experience.

But I can admit when Im wrong you all have converted me all the hard work i have put in is pointless. My hit percentage is because of my rifle not the fact that I suck. It will make my whole life easier when I lower my expectations, and the way they do load development is stupid easy so I should just do that too.

I dont have access to the applied ballistics software right now but if someone does id be intrested to see the difference in hit percentage between a 3/4 moa rifle and a 1 1/2 rifle at 600 with a 10 mph wind. Hornadys may have just single-handedly explained why Im not national champion.

Carl Zant at Precision Rifle Blog already answered your hit percentage question:
https://precisionrifleblog.com/2015/04/15/how-much-does-group-size-matter/

Interestingly, Brian Litz has written some excellent material on how statistical analysis and sample size applies to shooting and load development. And he does have some trophies on the shelf!

 

[6.5SH]

And my point is it seems like a whole lot of gaslighting. I may have used the wrong example to get my point across but I will again state my point which is the accuracy numbers they use in there podcast dont match my experience.

But I can admit when Im wrong you all have converted me all the hard work i have put in is pointless. My hit percentage is because of my rifle not the fact that I suck. It will make my whole life easier when I lower my expectations, and the way they do load development is stupid easy so I should just do that too.

I dont have access to the applied ballistics software right now but if someone does id be intrested to see the difference in hit percentage between a 3/4 moa rifle and a 1 1/2 rifle at 600 with a 10 mph wind. Hornadys may have just single-handedly explained why Im not national champion.

An online tool similar to the WEZ calculator:

Hit Probability Calculator

A free hit probability calculator for long range shooting.

bisonballistics.com

 

[Sparkeytj]

Carl Zant at Precision Rifle Blog already answered your hit percentage question:
https://precisionrifleblog.com/2015/04/15/how-much-does-group-size-matter/

Interestingly, Brian Litz has written some excellent material on how statistical analysis and sample size applies to shooting and load development. And he does have some trophies on the shelf!

I know that article was one of the reasons the conclusions made on the podcast seemed counter intuitive. If you use a calculator hit precentage with a 1 moa rifle is in the 60-75% range under fairly calm conditions. If thats correct a 85% hit precentage at a 100 round match would be an amazing score but its actually fairly common.