Trying to understand sd and es in load development

[timotheius]

So I recently got a Magnetospeed chronograph and have been using it to try to develop an accurate load for varmint/predator hunting with my AR-15. I am using 50 grain sierras, 8208 XBR powder, lapua brass. I started with 24.5 grains, and I loaded up 4 round groups in .3 grain increments and then tested their accuracy and speed. So what exactly is the role of SD and ES in load development? I guess my question is this: will the 4 round group with lowest Standard Deviation and Extreme Spread always be the best load for the rifle? Or is it possible that groups with higher SD and ES might be the better accuracy load? Would I be better off just firing off rounds looking for the low SD without much regard for where they land on paper initially? I have found a few charge weights that shot great looking sub-moa groups, but the SD was about 30. I had one group where the SD was around 10, but the 4 shots were scattered all over the place. My best group had an SD of 8, but was at 26.5 grains which I feel was probably over max.

 

[dok7mm]

Your target should be your main indicator. SD & ES at low levels is really nice to have, but can be skewed by other factors. Brass prep, powder charge consistency, differences in case hardness/neck tension, your shooting & reloading regime, etc...
these all effect consistent numbers on your chrono.

 

[Bob Wright]

I agree w/ dok7mm. Many guys will have low sd and es and shoot little tiny groups. Some with higher double digits sd and es and shoot little tiny groups. It's a good indicator that load consistency is achieved in single digits but if accuracy is repeatable with a higher es /sd, you can feel comfortable not chasing something that has little effect on that particular recipe. Especially on an AR. I just tried reloading for my factory stock AR trying to replicate some Barnes loaded ammo. It ain't working out to be easy getting .75 in groups with my first experiment. Its a 200 yard gun for coyotes at the moment, lol. Needs much more tinkering.

 

[RegionRat]

An SD is a way of describing the width of a normal distribution. Once you have an average and an SD, you should be able to predict the full width of the tails of that normal distribution, and we call that full width the ES.

If the process is normal, the relationship between the SD and the ES is a factor of six. So if you know the average, you can estimate the high by adding 3xSD and the low by subtracting 3xSD. The difference between them is the ES. So, if for example we have something with an SD of 1 and an average of 10, we can estimate the high will be 13 and the low will be 7, with an ES of 6.

When we shoot much past about 600 yards, the SD begins to matter because it contributes to vertical dispersion just on the basis of the difference in trajectory between the fastest and slowest shots in the group. However, you only had a 4 shot sample...

It takes more than 7 samples to start to really see what the SD will become when we are lucky, and something like 15 shots when we are not so lucky.

So, you don't have enough samples to judge the SD of the loads, but if the SD with 4 samples is large, it will only get bigger with more samples anyway.

It is sometimes possible to see a load with a higher SD and ES give a tighter group than a load with a tighter statistic. This happens for two reasons, the sample size is too small and those stats are not true, and because the shooting is done at short range so the ES hasn't shown up in the groups yet. If you are shooting less than 600 yards, the ES will just being to show.

Keep working your load development. If you find a stable speed node, you can run an development for seating depth and see what happens. If you find a good group, then take that load out to your maximum distance and see if it is good enough. YMMV

 

[Deputy819]

I had this bookmarked in my phone. Thought you might find it of interest...

https://www.google.com/url?q=https:...tics/stddev/&usg=AOvVaw0hlmW6GSC-xzGR9sFLDjKv

 

[bob4]

I had this bookmarked in my phone. Thought you might find it of interest...

https://www.google.com/url?q=https:...tics/stddev/&usg=AOvVaw0hlmW6GSC-xzGR9sFLDjKv

Good read.

 

[timotheius]

Thanks everyone for all your input. That article was interesting and helpful. It sounds like I should be more concerned about how groups are showing up on paper rather than the SD of a small sample size.

 

[Greyfox]

For LR hunting and competition, regardless of how tight our groups are at the shorter ranges(<300 yards), vertical dispersion(high ES/SD) will rear it's ugly head at the longer ranges. It's easy enough to determine the effect of your ES value at your desired maximum range. Once you determine a repeatable average velocity and extreme spread, plug the velocity extremes of the ES value into your Ballistic calculator using the maximum range and conditions you plan to hunt/shoot. For example, if your average velocity is 3000FPS and you have an ES of 40FPS, look at the impact points of 2980FPS and 3020FPS(Extreme spread in velocity), and determine the difference of each impact point compared to the impact point of your average velocity. If the difference takes you out of your target zone you either have to reduce your max distance until it matches your target zone, or work on your load to lower the ES to match the target zone. I will usually work with actual ES values and average velocity, and will test a minimum of 3 or 4, 5 shot groups in varied conditions(temperatures) for reproducibility. Just some thoughts.

 

[Kansaswoodguy]

With an AR if you can get 5 shot groups with SD in the teens or below your doing well. That 8208 works very well in an AR behind 75-77gr bullets at or slightly above book max. For those 50gr I suspect you'll have better luck with Benchmark maybe R-7 or R-10X. I load looking for a flat spot in the velocity where a couple of .3 grain diffrent powder charges in a row have similar average velocity's and the ones on either side make substantial jumps in speed. I then load for the middle of that flat spot these almost always coincide with my best accuracy at extended ranges. This seems to work best with stick powders and ball powders tend to be more linear without the flat spots in load development. Look at Noslers or Seirra's load manuals for your bullet weight and use their accuracy powder may save you some time and money in components.

 

[rwcole]

When reading about SD and ES in contrast to load development, is there one particular way you all that shoot long range will load? In other words once youve got your brass formed to your chamber, do you prep the brass the same way each time? Can you get by with just neck sizing or should you FL size each and every time?

 

[bob4]

I always prep mine the same each time. Things like De-burring flash holes is only done once. I am not as stringent as some and maybe more than others. But these days my ES has shrunk considerably paying close attention to brass prep.

Can you get by with just neck sizing or should you FL size each and every time?

Problem , as I see it with neck sizing only,here is this. Each time you shoot that piece of brass it grows just a bit. Changing volume in the case and fit in your chamber until it becomes hard to chamber. Now you have to full resize it. If you don't adjust your dies to only bump shoulders back .001-.003. ( personal preference) you'll bump them back so far that your pet load will not shoot the same. Not to mention neck may be different if your all of a sudden using a different die. So why not discover your head space for that rifle and full length size it each time just keeping the shoulders back just far enough to chamber nicely and making your brass/ammo the same each time? Making it the same each time means it should shoot the same each time. Again it depends on your application and how accurate you would like/need to be.

 

[Sevy]

Bob4 is right on with neck sizing. I recently started neck sizing and have periodically checked my velocity with once, twice, and 3 time forced brass and it has slowly gotten slower. Actually by quite a bit. Over 50 fps. It seems to be still in am accuracy node but I am done neck sizing for sure. I am gonna go back to fl sizing with a headspace gauge to only bump the shoulder back a couple thous and just stick to that. And imo brass should be prepped the exact same every time! That actually goes for everything in reloading I say.

 

[DartonJager]

Gosh but sometimes I really HATE yu guys.
Just when I start to think after 25+ years I have learned enough and can consider myself above that of a beginner/intermediate reloader I read a posting like this and I learn things I realized should have been common sense all along, especially concerning neck only sizing brass as it applies to affecting case capacity and related pressures.
I have some of the supposedly best books on advanced reloading techniques and none ever have even mentioned let alone explained all the potential short comings of neck only sizing.
I had to find that out from websites like this as well as YouTube.

 

[RegionRat]

When reading about SD and ES in contrast to load development, is there one particular way you all that shoot long range will load? In other words once youve got your brass formed to your chamber, do you prep the brass the same way each time? Can you get by with just neck sizing or should you FL size each and every time?

The answer is all too often... it depends...

If the brass condition in terms of cold work could stay constant, we would get away with just neck sizing. But, Mother Nature is cruel. Brass will cold work and harden with cycles to the point where we will loose that reliable shoulder and body fit. The goal is not to say neck or FL size, but to minimize the case stretch while keeping the case the same size while allowing reliable feeding.

If we had brass that didn't change with cycles, we wouldn't be able to regain the neck tension either. The neck properties are kept within reason by using annealing, and the shoulder and body changes are kept to a minimum while making sure they feed. This way the case volume doesn't stray and we can still open and close the bolt. Many serious shooters have shown that the case doesn't have to be very tight to the chamber to produce the best accuracy.

Clearly you try to avoid over working the case to avoid unnecessary trimming and premature head separation, but bump sizing ends up being about the best of both worlds. You can get the case to feed, yet not wear the brass out by over doing it. Getting the sizing and condition of the neck the same each time ends up being far more important than having a couple thousandths of change in shoulder space.

Hope that is a little more clear. The advanced topic separated the concepts of the neck from the body. Since we don't want to have the ammo get so large the bolt won't close, and it is not reliable to try to measure when that happens by inspecting case dimension, if we just make sure we bump size the body to keep it just under max headspace, we are then no longer concerned with reliable feeding while still avoiding over working the case. This frees you up to focus on the neck tension.