If you only watch one youtube video this year, make it this one;
https://www.youtube.com/watch?v=QwumAGRmz2I
If you want more, there is a follow-up at;
https://www.youtube.com/watch?v=6yZyXwy40JM
If you don't have the time to watch them, here is a summary;
- use carefully assembled loads with quality components in a quality rifle system. Choose these from your own experience, or from what is popular in your discipline.
- tune using components, not metrics. In other words, if a load isn't performing to realistic expectations, change your bullet, your powder, or your barrel/chamber. Don't bother chasing seating depth, powder charges etc using ladder tests as potential improvements are minimal at best and unlikely to be statistically discernible.
- collect statistically meaningful data by shooting 20 - 30 shot groups. If you want to calculate your chances of hitting a target, measure mean radius and calculate the standard deviation.
In super-summary, use quality and don't sweat the small stuff.
Does this seem outrageous and contradictory to popular and well accepted practices that notable social influencers espouse, and many shooters follow? Should we give up on finding velocity and elevation nodes using ladder tests? Surely seating depth and powder charges have to be tested in a methodical and didactic fashion? Watch the videos and give us your opinion!
Cheers,
Ian
Load Development
Moderator: Mod
-
- Posts: 79
- Joined: Thu Sep 06, 2018 9:46 pm
Re: Load Development
The best advice is to talk to the winners of major shoots, follow there advice and dont believe to much on the internet, buy all means watch heaps but keep a open mind.
A lot is powerd by money, ie they need the views to make there money
A lot is powerd by money, ie they need the views to make there money
-
- Posts: 614
- Joined: Tue Oct 16, 2018 5:20 pm
- Location: 7321 Tas.
Re: Load Development
It really drives home the lack of value (and waste of components) in developing loads with 5 shot ladder tests.
Pete
Pete
-
- Posts: 1289
- Joined: Mon Dec 29, 2008 10:29 pm
Re: Load Development
At big events the winners are the ones that put in the work to test and practice.
-
- Posts: 614
- Joined: Tue Oct 16, 2018 5:20 pm
- Location: 7321 Tas.
Re: Load Development
This is another longish post: So I watched the Hornady video and the follow up one.
I did a fair bit of stats at Uni a long time ago and knew the rules but I have been also listening to (and wasting my time) watching the various YouTube pundits who espouse all sorts of populist (dis)information, with titles like "Load development in 10 shots" or "Ladder tests the easy way".
The Hornady video basically called out 5 shot group testing as rubbish because the sample sizes are far too small and if you shortcut statistics the results may well be rubbish.
There is no substitute for sample size if you want data that are meaningful.
No one here would be happy taking a drug that had supposedly fabulous results but was only tested on 5 people last Wednesday. Why? - because the sample size was too small to be meaningful and the consequences could be serious.
So why would we accept a five shot group (with the same powder load etc) as being representative of a rifle's performance over the longer term. the consequences may not be life threatening but might be a waste of time and money.
The question I asked was " What are the chances of the velocity of five consecutive shots under the same conditions being close or spread out"
So I set out to get some real numbers and have a look at five shot groups in terms of velocity spread – something the video didn’t do.
I fired 30 consecutive rounds in a single string over about 30 mins:
284Win, 190 Berger moly coated, 51.0gn 2209, 0.018" jump, trimmed, chamfered and moly lubed necks, GM210 primers, Lapua 6.5-284 brass once fired. A Labradar was used to chronograph velocity. The rifle is a Barnard P in a carbon stock with a Bartlein 32” parallel barrel, 5 groove, 8 twist with 50 rounds fired.
I acknowledge that as the barrel wears in the velocity variation might improve, but conversely, load development usually starts with a new barrel.
In a five shot group, most of us would be happy with an ES of 5fps and unhappy with an ES above 10pfs.
I looked at the chances of firing 5 consecutive shots within the string of 30 shots that had an ES of 10pfs or less and the chances of firing five consecutive shots with an ES of greater than 10fps.
The first shot was cold bore and may not be representative as it was 21fps lower than the next lowest shot. If it is eliminated from the calculations, it lowers the overall ES by 21fps and drops the SD by 2pfs.
The graph illustrates how each 5 shot string steps up one shot in the firing order. Shot velocity on the vertical axis and shot number on the horizontal axis.
The summary table below shows the effects of removing shot 1 from the calculations, it lowers the ES dramatically, but doesn't affect the subsequent calculations.
The data in the table show each shot in the order of firing and starting at shot#5, the ES for the previous 5 shots is entered in column 3. So in effect there were 26, 5 shot strings.
8 groups showed 10fps or less ES or 31% of groups
18 groups showed greater than 10fps ES, or 69%. of groups
So if the trend generally holds true for other loads then in a ladder test of say 10 loads, 3 of those likely will be acceptable and 7 won't purely by chance. That seems to be pretty much what I see in published ladder tests, - about a third of 5 shot strings will have low ES.
In my mind this test puts a big doubt in the value of a five shot group as being a predictor of anything substantive.
Pete
I did a fair bit of stats at Uni a long time ago and knew the rules but I have been also listening to (and wasting my time) watching the various YouTube pundits who espouse all sorts of populist (dis)information, with titles like "Load development in 10 shots" or "Ladder tests the easy way".
The Hornady video basically called out 5 shot group testing as rubbish because the sample sizes are far too small and if you shortcut statistics the results may well be rubbish.
There is no substitute for sample size if you want data that are meaningful.
No one here would be happy taking a drug that had supposedly fabulous results but was only tested on 5 people last Wednesday. Why? - because the sample size was too small to be meaningful and the consequences could be serious.
So why would we accept a five shot group (with the same powder load etc) as being representative of a rifle's performance over the longer term. the consequences may not be life threatening but might be a waste of time and money.
The question I asked was " What are the chances of the velocity of five consecutive shots under the same conditions being close or spread out"
So I set out to get some real numbers and have a look at five shot groups in terms of velocity spread – something the video didn’t do.
I fired 30 consecutive rounds in a single string over about 30 mins:
284Win, 190 Berger moly coated, 51.0gn 2209, 0.018" jump, trimmed, chamfered and moly lubed necks, GM210 primers, Lapua 6.5-284 brass once fired. A Labradar was used to chronograph velocity. The rifle is a Barnard P in a carbon stock with a Bartlein 32” parallel barrel, 5 groove, 8 twist with 50 rounds fired.
I acknowledge that as the barrel wears in the velocity variation might improve, but conversely, load development usually starts with a new barrel.
In a five shot group, most of us would be happy with an ES of 5fps and unhappy with an ES above 10pfs.
I looked at the chances of firing 5 consecutive shots within the string of 30 shots that had an ES of 10pfs or less and the chances of firing five consecutive shots with an ES of greater than 10fps.
The first shot was cold bore and may not be representative as it was 21fps lower than the next lowest shot. If it is eliminated from the calculations, it lowers the overall ES by 21fps and drops the SD by 2pfs.
The graph illustrates how each 5 shot string steps up one shot in the firing order. Shot velocity on the vertical axis and shot number on the horizontal axis.
The summary table below shows the effects of removing shot 1 from the calculations, it lowers the ES dramatically, but doesn't affect the subsequent calculations.
The data in the table show each shot in the order of firing and starting at shot#5, the ES for the previous 5 shots is entered in column 3. So in effect there were 26, 5 shot strings.
8 groups showed 10fps or less ES or 31% of groups
18 groups showed greater than 10fps ES, or 69%. of groups
So if the trend generally holds true for other loads then in a ladder test of say 10 loads, 3 of those likely will be acceptable and 7 won't purely by chance. That seems to be pretty much what I see in published ladder tests, - about a third of 5 shot strings will have low ES.
In my mind this test puts a big doubt in the value of a five shot group as being a predictor of anything substantive.
Pete
You do not have the required permissions to view the files attached to this post.
-
- Posts: 7502
- Joined: Wed Jun 15, 2005 8:22 pm
- Location: Maffra, Vic
Re: Load Development
Agreed Peter.
There is however one rare situation where small sample sizes are quite useful, and that is when testing barrel tuner settings for a given load. You can usually eliminate a setting based on the first two shots if the group is unacceptably big. Any further shots at that setting are pointless if the first two are too far apart because you will probably never adopt that setting. On the other hand if the first two shots are good, the question becomes how many more do you need to be sure the setting is good
!
There is however one rare situation where small sample sizes are quite useful, and that is when testing barrel tuner settings for a given load. You can usually eliminate a setting based on the first two shots if the group is unacceptably big. Any further shots at that setting are pointless if the first two are too far apart because you will probably never adopt that setting. On the other hand if the first two shots are good, the question becomes how many more do you need to be sure the setting is good
![Very Happy :D](./images/smilies/icon_biggrin.gif)
-
- Posts: 350
- Joined: Wed May 28, 2014 11:48 pm
Re: Load Development
I actually talked to Jayden at Hornady on the phone last Wed or Thursday about the videos as I had some questions for him. I also talked to Tommy at Sierra on Friday...Tommy and I where talking about a quote I had to do for him but he was on my mind about a bullet question. Tommy works/runs the ballistic lab at Sierra.
You have to keep in mind that these guys are dealing in absolutes when it comes to data.
I know on Snipershide a lot of the guys where giving them a pretty good beating about the podcast #50 and #52.
Hornady or Sierra etc..or even the gov't test facilities...have a budget that the ordinary shooter doesn't have but in the end it goes back to what I always say....
"Know your rifle. It will tell you what is going on but you have to pay attention to it."
In that quote above is the ammo as well as the environment you are shooting in. It all figures in. So having good data that you can fall back on prior to going to a match will help keep you out of trouble.
You have to keep in mind that these guys are dealing in absolutes when it comes to data.
I know on Snipershide a lot of the guys where giving them a pretty good beating about the podcast #50 and #52.
Hornady or Sierra etc..or even the gov't test facilities...have a budget that the ordinary shooter doesn't have but in the end it goes back to what I always say....
"Know your rifle. It will tell you what is going on but you have to pay attention to it."
In that quote above is the ammo as well as the environment you are shooting in. It all figures in. So having good data that you can fall back on prior to going to a match will help keep you out of trouble.
-
- Posts: 13
- Joined: Thu Sep 22, 2016 7:53 pm
Re: Load Development
Frank Green wrote:I know on Snipershide a lot of the guys where giving them a pretty good beating about the podcast #50 and #52.
I've just spent more time than I should reading the Snipers Hide, and Accurate Shooter threads on these podcasts. It's a fascinating, but ultimately sad, insight in to human nature.
-
- Posts: 350
- Joined: Wed May 28, 2014 11:48 pm
Re: Load Development
ilc wrote:Frank Green wrote:I know on Snipershide a lot of the guys where giving them a pretty good beating about the podcast #50 and #52.
I've just spent more time than I should reading the Snipers Hide, and Accurate Shooter threads on these podcasts. It's a fascinating, but ultimately sad, insight in to human nature.
Well said!
Return to “Equipment & Technical”
Who is online
Users browsing this forum: No registered users and 54 guests