Load Development Confirmation

[Sixgun]

Here's some pictures. Do you see what I'm seeing? The first string was shot at 85 degrees the second was shot at 60! These rounds were loaded seperately and shot round robin the first group was shot at 100y, the second was shot at 200. The first string I was doing an OCW test..the second string was more of a ladder test. I apologize in advance I don't have a ShotMarker or a Labradar so I had to do it the ol' fashioned way. Look at the 1st group for 39g in the first string and the 2nd group in the second string. The groups might not show it...but I think the ES and SD numbers are more important than group size. The SD numbers were calculated with Calculator.net SD calculator. Velocity was from a Caldwell chrony attached to my phone. Looks to me 39g of H4350 with 144 Bergers is the one to have. What do you guys think?

 

[foster]

Shoot the 39 and 40.4 again to see if it repeats. Shoot each charge at the same time. You want them in as close to the same condition as you can get.

 

[Sixgun]

Shoot the 39 and 40.4 again to see if it repeats. Shoot each charge at the same time. You want them in as close to the same condition as you can get.

The 39 was shot last week 85 degrees no wind. I loaded up more its not like they were left over from last week. I shot those yesterday..60 degrees one or two of those shots in a slight drizzle. What's really profound to me is the ES and SD numbers are almost identicle everything round robin. In this pic, this is were the 4th string went that is why the pics with the dots do not have a 4th shot. The bottom row dots i was testing the EC tuner brake.

The second dot from the left on top is the group for 39g charge. I guess this my bottom node. I'll definately give the higher another shot.

 

[RegionRat]

I will suggest this in the hope that it doesn't cause controversy or give offense to folks.

TLDR: with skinny short range data, study the shot impacts and groups, not the velocity stats.

There is no statistical strength in your SD/ES estimates, at least for the data you have shown. For the time being, focus on your target data and collect speed if you like but postpone making decisions on the velocity stats just yet.

Calculations of SD on three shot (or even 5 shot) groups is very unreliable and you are fine with just a study of the average and consider the ES, but even that ES isn't meaningful.

You can only use a 3 sample ES to reject a recipe, not to accept one.

If there was a hypothetical perfect shooter with a perfect gun, a giant ES could mean the recipe is a reject when it exceeds your goal standards, however, there is still the possibility that the giant ES was really the best charge and the loader screwed up on a single sample of the brass prep. So small samples of the load are very risky to judge, good or bad.

There is a leap of faith many of us take with the confidence we place in our brass prep. We also use the concept of smooth transitions and the idea that if we are taking fine charge steps it is unlikely that we would get a wild change in one step without seeing it on the neighboring charges. This is still risky.

Your math is correct, but the values have nearly a zero probability they represent an extrapolation that could be used to estimate the ES of a future sample.

For example, just because I run the calculation correctly, doesn't make those 3 sample SD values mean what they represent when there are say 30 samples.

That is not to say if you have a small sample with an ES of 30 versus one with an ES of 6, that the one with 30 is guarantied to be worse than the one with 6, but believe it or not there is still a possibility that with small samples the one with the ES of 6 can throw something larger than the 30 on the very next try and vice versa.

Now that I have made a negative comment on the SD of three shot samples, let me add that if you are forced to take risks on small sample sizes, consider using all of the bullet impacts to compare the group performance and then wait to make a stand on the velocity stats.

If you focus on what you think is a good charge weight and take many more samples of those, then the velocity data can begin to form a better average, ES and finally an SD once you have at least 15 or more samples, and more like 30.

To take risks on small sample sizes, it is best to stick with known pet load recipes in standardized guns. However, if you don't have that as a choice, I will suggest you analyze your target data the way that is shown in this link below.

By taking the impact position data of every shot in the group, and also in the adjacent powder steps, you build at least some probability that you can find the best charge (if there really is one).

https://www.ar15.com/forums/Armory/...-Accuracy-Node-Detection-Technique/42-524007/

The method requires you to analyze the target data and ignores velocity for the time being. It exploits every shot you take, but that also puts the burden of taking good shots on the shooter (all methods do.)

Once you take a decision risk on a charge weight, you should consider taking all the velocity data on it as well as nudge the charge up and down to investigate how well centered you are in the node.

BTW, there is still a possibility that the rig shoots the whole charge range roughly the same. It is entirely possible the gun produces temporary small groups in one test, and doesn't repeat the next, and then produces an equal group throughout the whole charge weight range when you aggregate all the data.

I know that upsets some folks, but the reality is that many rigs would shoot roughly the same no matter what they are fed especially when weather is factored in. That isn't necessarily a bad thing, but I know it upsets those who are being unrealistic about dispersion and statistics.

I hope you get the main point, which is to study the target and postpone the velocity stats as a pivot point on making decisions. In the end, when you take the testing to distance, if the velocity stats are good but the groups are not, those velocity stats are cold comfort. On the other hand, if you get good groups at max distance, then by definition those velocity stats are good enough. YMMV

Good Luck and have fun.

 

[Doom]

OP, I started my original comments but didn't post them. I decided to wait and see if one of the other members more eloquent than I might do so and @RegionRat has gone the route I was headed down. But I want to address this issue in a little less analytical way then I normally do but still from an engineering standpoint.

When developing a load that will be accurate and precise in any firearm we are relying on what engineers refer to as a system. That system is made up of all the components of the rifle, ignition system, sighting system, stock, barrel and chamber, and cartridge (including powder, primer and bullet). Oh, and the rifleman. They all contribute to the final intended result of putting a bullet in the intended spot. Historically in developing a load it was done without a chronograph but loading varying powder charges and finding loads that tend to print to the same point of impact or vertical position on the target. This development method told the reloader that the entire system was "happy" with the load. No chronograph was needed.

Your chronograph is an instrument. It measures velocity and nothing else. Since these instruments are microprocessor based they can be programmed to spit out information such as velocity, standard deviation and extreme spread. This data is useful, to an extent. To understand average it's pretty simple, add velocityes divide by number of shots. That is it. When we start to look at standard deviations and extreme spreads we are now dealing with statistical models of experimental data which is based on probabilities. It is much more complicated than just looking at very limited test data; its trying to understand what that data means for future rounds. Three, five or even ten rounds just doesn't provide enough data to justify decisions between loads in most cases. It is also necessary to understand that a low standard deviation or extreme spread alone does not make an accurate rifle system. It is the system that make it accurate.

The Target will tell you where the system is "happy". Once a known range of loads is identified it then becomes desirable to refine and understand more specific info on the cartridge and this is where the chronograph data can be used to refine and adjust the cartridge to improve the standard deviation and extreme spread. it can be used to identify case preparation, primer selection, load refinement, seating depth, annealing issues, etc. This is where the chronograph can be used to refine the precision of the reloading process. There is no harm done in taking velocity data during initial load development. Knowing approximate velocity is a good thing and sometimes it can explain that random flier that can sometimes be encountered because the cartridge has an issue, like that one FC case in a batch of LC cases. But if the data becomes the object of attention at that point its easy go down the proverbial "rabbit hole" and waste time and energy that is better suited to making sure the system is happy.

 

[Sixgun]

I will suggest this in the hope that it doesn't cause controversy or give offense to folks.

TLDR: with skinny short range data, study the shot impacts and groups, not the velocity stats.

There is no statistical strength in your SD/ES estimates, at least for the data you have shown. For the time being, focus on your target data and collect speed if you like but postpone making decisions on the velocity stats just yet.

Calculations of SD on three shot (or even 5 shot) groups is very unreliable and you are fine with just a study of the average and consider the ES, but even that ES isn't meaningful.

You can only use a 3 sample ES to reject a recipe, not to accept one.

If there was a hypothetical perfect shooter with a perfect gun, a giant ES could mean the recipe is a reject when it exceeds your goal standards, however, there is still the possibility that the giant ES was really the best charge and the loader screwed up on a single sample of the brass prep. So small samples of the load are very risky to judge, good or bad.

There is a leap of faith many of us take with the confidence we place in our brass prep. We also use the concept of smooth transitions and the idea that if we are taking fine charge steps it is unlikely that we would get a wild change in one step without seeing it on the neighboring charges. This is still risky.

Your math is correct, but the values have nearly a zero probability they represent an extrapolation that could be used to estimate the ES of a future sample.

For example, just because I run the calculation correctly, doesn't make those 3 sample SD values mean what they represent when there are say 30 samples.

That is not to say if you have a small sample with an ES of 30 versus one with an ES of 6, that the one with 30 is guarantied to be worse than the one with 6, but believe it or not there is still a possibility that with small samples the one with the ES of 6 can throw something larger than the 30 on the very next try and vice versa.

Now that I have made a negative comment on the SD of three shot samples, let me add that if you are forced to take risks on small sample sizes, consider using all of the bullet impacts to compare the group performance and then wait to make a stand on the velocity stats.

If you focus on what you think is a good charge weight and take many more samples of those, then the velocity data can begin to form a better average, ES and finally an SD once you have at least 15 or more samples, and more like 30.

To take risks on small sample sizes, it is best to stick with known pet load recipes in standardized guns. However, if you don't have that as a choice, I will suggest you analyze your target data the way that is shown in this link below.

By taking the impact position data of every shot in the group, and also in the adjacent powder steps, you build at least some probability that you can find the best charge (if there really is one).

https://www.ar15.com/forums/Armory/...-Accuracy-Node-Detection-Technique/42-524007/

The method requires you to analyze the target data and ignores velocity for the time being. It exploits every shot you take, but that also puts the burden of taking good shots on the shooter (all methods do.)

Once you take a decision risk on a charge weight, you should consider taking all the velocity data on it as well as nudge the charge up and down to investigate how well centered you are in the node.

BTW, there is still a possibility that the rig shoots the whole charge range roughly the same. It is entirely possible the gun produces temporary small groups in one test, and doesn't repeat the next, and then produces an equal group throughout the whole charge weight range when you aggregate all the data.

I know that upsets some folks, but the reality is that many rigs would shoot roughly the same no matter what they are fed especially when weather is factored in. That isn't necessarily a bad thing, but I know it upsets those who are being unrealistic about dispersion and statistics.

I hope you get the main point, which is to study the target and postpone the velocity stats as a pivot point on making decisions. In the end, when you take the testing to distance, if the velocity stats are good but the groups are not, those velocity stats are cold comfort. On the other hand, if you get good groups at max distance, then by definition those velocity stats are good enough. YMMV

Good Luck

Welp, I don't care about velocity.. I don't even care about group size at this point in the process. I never mentioned it. Right now, I'm looking for the best charge weight and there are only two ways to determine OCW. ES and SD numbers or plotting velocities for 'the flat spot'. Group size is then determined by seating depth. Now, if you or anyone else thinks I'm a fool for doing it this way for load development...please tell me because I'd love to hear about it.

Yes, the sample size is small but...it's random. 25-30 temperature difference with wind and drizzle on one of the days, loaded at different times, shot round robin, even shooting on a different range at different distances. I will shoot more groups with this CHARGE WEIGHT to confirm before moving on to seating depth. It's not just shooting 30 five round groups as you suggested...more groups.. but should be shot in various conditions that's what counts. I should be able to shoot..hopefully this load in sub zero temps and get similar results it's just the velocity will be lower. This is just one node too I have another in the 40.2-40.6 range I'll have to work on.

Thank you for your input I really appreciate it.

 

[RegionRat]

Right now, I'm looking for the best charge weight and there are only two ways to determine OCW. ES and SD numbers or plotting velocities for 'the flat spot'. Group size is then determined by seating depth.

If "best charge" means anything, it means the one that delivers the projectile with accuracy and precision.

It is typical that folks who chase "flat spots" in velocity are surprised to find those flat spots do not repeat, so the best advice I can give you is still the same, that is, nothing wrong with taking velocity data, but not at the expense of shot fall. Prioritize the target data for now, and the velocity stats will follow. The opposite will waste your time.

The OCW method doesn't use a chronograph at all.
An Audette Ladder doesn't use a chronograph at all.

Doom gave a very good description of what, if anything, you could glean from the chrono data.

Depending on how much budget you have to burn, I would advise you only take chrono data while focusing on the target data to help indicate if you have a brass prep problem and later on you can collect more once you find any potential accuracy.

Pick a seating depth you can live with for your first charge weight efforts, then you can always try a method like the one I linked to determine if there is any potential benefit to seating depth.

There is a difference between contexts like preloaded ammo, versus real time adjustments as in Benchrest. Without knowing you, the gun, the recipe, or the interests, my advice would be the same and is only offered if it helps. If it doesn't then no worries. YMMV



ETA: I forgot to address the other point, I am not suggesting 30 sets of 5 shot groups.
The number 30 is a typical sample size for gathering a stable SD value. Many times, with a normal distribution or a two lobe distribution, the probability on the SD value being stable is very low till you have over 15 samples and often not till you go past 30.

The other main point you made is very important. Once an "optimal charge weight" is established, it is very important to perturb it up and down to learn how wide the node is, and then plan for weather/climate. It takes above average shooting and planning to anticipate fall/winter shooting conditions when load development is done in a different season and altitude.

None of that is easy unless your charge weight or seating depth is in a wide window to begin with. Preloaded ammo also has a requirement to travel so it must be sturdy. That includes concepts like realistic neck prep, so I would budget some time for tuning your sizing process as well.

Using concepts like CEP (Circular Error Probability) tend to confuse folks, so I tend to mention Mean Radius. Study the use of the MR to help clarify performance changes. Most folks can do that by eyeball, but it still helps to use target analysis software or use a ruler and do the math.

Again, good luck and have fun with the process.

 

[Doom]

Velocity flat spots show up due to sampling probability and instrument error and almost always for small increments (like .2gr increments in 6.5CM) where the expected velocity increment is on the order of 5-10 fps and the best chronograph is +/- 2.7 fps @ 2700 fps.

Scott Satterlee was the originator of the 10 shot ladder. He didn't use it long. Look up Satterlee 2.0 for his later method if interested.

 

[JFrank]

Welp, I don't care about velocity.. I don't even care about group size at this point in the process. I never mentioned it. Right now, I'm looking for the best charge weight and there are only two ways to determine OCW. ES and SD numbers or plotting velocities for 'the flat spot'. Group size is then determined by seating depth. Now, if you or anyone else thinks I'm a fool for doing it this way for load development...please tell me because I'd love to hear about it.

Yes, the sample size is small but...it's random. 25-30 temperature difference with wind and drizzle on one of the days, loaded at different times, shot round robin, even shooting on a different range at different distances. I will shoot more groups with this CHARGE WEIGHT to confirm before moving on to seating depth. It's not just shooting 30 five round groups as you suggested...more groups.. but should be shot in various conditions that's what counts. I should be able to shoot..hopefully this load in sub zero temps and get similar results it's just the velocity will be lower. This is just one node too I have another in the 40.2-40.6 range I'll have to work on.

Thank you for your input I really appreciate it.

There’s always this method’ I would load on the red dot @ 30.7

 

[Sixgun]

There’s always this method’ I would load on the red dot @ 30.7

I did try that there's actually two more shots that are not pictured they were at 40.4 and 40.6g and printed slightly higher still in the orange respectively. I don't think there's enough to go on. On this second test i was only shooting the lowest ES and SD numbers from the first test. 39g at 2700 and 40.5g at 2816. This picture shows 38.8, 39, 39.2, 40, 40.2, 40.4 and 40.6g.

 

[Webster]

Welp, I don't care about velocity.. I don't even care about group size at this point in the process. I never mentioned it. Right now, I'm looking for the best charge weight and there are only two ways to determine OCW. ES and SD numbers or plotting velocities for 'the flat spot'. Group size is then determined by seating depth. Now, if you or anyone else thinks I'm a fool for doing it this way for load development...please tell me because I'd love to hear about it.

Yes, the sample size is small but...it's random. 25-30 temperature difference with wind and drizzle on one of the days, loaded at different times, shot round robin, even shooting on a different range at different distances. I will shoot more groups with this CHARGE WEIGHT to confirm before moving on to seating depth. It's not just shooting 30 five round groups as you suggested...more groups.. but should be shot in various conditions that's what counts. I should be able to shoot..hopefully this load in sub zero temps and get similar results it's just the velocity will be lower. This is just one node too I have another in the 40.2-40.6 range I'll have to work on.

Thank you for your input I really appreciate it.

Don't understand any of your procedure. You never described the rifle and what it's used for. I never looked at stats, all I care about is group size and repeatability.

 

[Bottom Fodder]

Wow I guess I’ve been doing it all wrong all these years
after all that I would be to embarrassed to say my simple way, I guess I’m behind the times and still low tec

 

[Webster]

Wow I guess I’ve been doing it all wrong all these years
after all that I would be to embarrassed to say my simple way, I guess I’m behind the times and still low tec

There are no flat spots, it's just not accounting for ES. I have been reloading for 50 years for very accurate varmint rifles. Flat spots have been covered 100 times on this website. My rifles shoot small with just about any load.

 

[lawman29]

What do you guys think?

Poorly presented. Targets are not labeled at all with basic information (charge weight, powder, distance, cartridge, primer, bullet...).

 

[quest450]

Broke in my new barrel shooting a match last Saturday, same lot of powder, new brass, new lot of primers and I was testing Alex’s new bullets, BRA barrel shot good with 29.7 grains and projectiles jammed 8 thou
Today I went to my local club and ran OCW test at 300 yards, there might be something between 29.6 - 29.7 and 30.00 - 30.10 looks decent. Had some leftover ammo from last weekend so I shot two 5shot groups
I’m looking for about 0,5” three shot groups at 300 yards then I always confirm and fine tune at 600
I don’t pay much attention to flat spots or chrono #’s I pay attention to what’s the paper telling me

 

[Rat Rifle]

Broke in my new barrel shooting a match last Saturday, same lot of powder, new brass, new lot of primers and I was testing Alex’s new bullets, BRA barrel shot good with 29.7 grains and projectiles jammed 8 thou
Today I went to my local club and ran OCW test at 300 yards, there might be something between 29.6 - 29.7 and 30.00 - 30.10 looks decent. Had some leftover ammo from last weekend so I shot two 5shot groups
I’m looking for about 0,5” three shot groups at 300 yards then I always confirm and fine tune at 600
I don’t pay much attention to flat spots or chrono #’s I pay attention to what’s the paper telling me


View attachment 1471598

Don’t You want to take a break this coming season? You must be exhausted?

By the way, what powder ?

 

[quest450]

Don’t You want to take a break this coming season? You must be exhausted?

By the way, what powder ?

Lol, I got a bug up mya$$ this am and had to go eff around to find out
gotta have plan "B" for the Nationals
H4895 of course

 

[Rat Rifle]

Lol, I got a bug up mya$$ this am and had to go eff around to find out
gotta have plan "B" for the Nationals
H4895 of course

You must have really stocked up a long time ago ???

 

[Sixgun]

Don't understand any of your procedure. You never described the rifle and what it's used for. I never looked at stats, all I care about is group size and repeatability.

I think i was trying to do to much with the number of rounds loaded in an effort to try to 'conserve'. The gun and what I'll be shooting and competing in is PRS. The gun has an MTU contour, 26", 1-8" Bartlein barrel chambered in 6.5x47L built by GA Precision bedded in a McMillan A5 stock and a S&B PM2 5.5x25x56mm scope.

This pic with all the orange squares is my first test doing the OCW method at 100y...all three shot groups. This pic is my results. The forth row of velocities is the average of the previous three. Group 1 starts on the top left and group 4 starts on the bottom left.

 

[Sixgun]

I think i was trying to do to much with the number of rounds loaded in an effort to try to 'conserve'. The gun and what I'll be shooting and competing in is PRS. The gun has an MTU contour, 26", 1-8" Bartlein barrel chambered in 6.5x47L built by GA Precision bedded in a McMillan A5 stock and a S&B PM2 5.5x25x56mm scope.

This pic with all the orange squares is my first test doing the OCW method at 100y...all three shot groups. This pic is my results. The forth row of velocities is the average of the previous three. Group 1 starts on the top left and group 4 starts on the bottom left.

This pic...top row only..was the the best ES and SD numbers of the first set which was 39g and 40.5.
From 39g..groups 1, 2 and 3 are 38.8, 39 and 39.2. Groups 4-7 are 40, 40.2, 40.4, and 40.6g.
The 2 dots in the middle are sighters the bottom row I was testing an EC tuner brake with 140 Berger hybrids over 37.7g Varget with BR4 primers. All the other groups are 144 Berger hybrids with H4350 and cci 450 primers...these are what I'm really testing.