So I was just talking with our 3 peat league champion and my mentor about shooting the Satterlee Load Test. So he showed me his last Satterlee round. Just like he explained this showed increases in pretty even steps with about 10-13fps difference between shots and mixed in that group were 4 shots total that were 2X2 that were almost exactly the same. So he explained that is what he was looking for.
I’ve shared this screenshot a lot, because it is visualization of what you’re describing - this is the goal of the Satterlee Velocity Curve method of long range load development. We shoot incremental charge weights and look for plateaus in velocity. You can see here, I shot 3 replicant strings, but I could have done ONE shot per charge weight and found the same node, as each individual string shows the node in the same place - which is largely the point.
It’s EXCEPTIONALLY rare to have these end up a straight line. I’ve ran this test over 50 different times for my match rifles in the last 5 years, before almost every match I shoot, and only ONE test showed a straight line - which really indicated I had insufficient neck tension with the new lot of brass, and changing the bushing in my die brought the node back. I use that same test for all of my barrels now, for a dozen or so other rifles, and I get the same results - if neck tension is good and I’m not WAY out of whack on powder charge for the cartridge, flat spots WILL reveal themselves.
Now after shooting with him and watching his work I know that his gun produced ES in single digits so then I asked what to do if like some of mine there is an ES of 60-80 between all of the different powder loads. he could no give a good answer other than that doesn't work.
300 is right on the ragged edge of “short range,” and with 223, probably really already into mid-range. A LOT of short range matches are won with ammo having 50-100fps ES. But your pursuit of reduced ES is certainly worthy here.
Of note - don’t beat yourself up if you never get to a single digit ED load. While it does occasionally happen, it’s EXCEPTIONALLY RARE to actually find single digit ES loads which persist for 10+ rounds. We can often sneak single digit ES out of 5 round tests, but don’t forget, over 95% of shots will fall within +/-2 standard deviations from average, so we often get false small ES’s. Reminding here - see bell curve below - ES is meant to represent Range, and any time ES isn’t somewhere 4-6x the SD, you should be questioning whether you fired enough shots to be statistically valid. I have new competitors tell me every season, “my load is great, I have a 5SD with a 9ES!” Um… no, you don’t. Go shoot 10-15 across the chrony and see what comes out… when we shoot enough to have a valid data set, where the Sample SD and Sample ES ACTUALLY represent the Population SD and Population Range, we’ll see something around ES = 6xSD.
So think about that a little - hitting a single digit ES means your SD can’t be greater than 1.5 (6x1.5=9), which does happen for individual 10-15 round strings sometimes (dumb luck and coincidence), but I’m pretty comfortable saying is impossible to sustain and repeat single digit ES loads for multiple 10+ round tests, let alone single 20+ round tests. I’ve seen infinitely more guys claim doing so than I’ve ever seen able to prove it.
For instance say a bullet powder combination has a start of 21.1gr and a max of 23.4gr I will start at 21.1 and load up in .3gr increments loading 6 rounds each. in this case for a total of 8 different samples, total of 48 loads. Sure that seems like a lot but I am looking for target loads, not hunting or blasting loads.
I’m cutting this early, acknowledging your method is described below, but to point out here, you could do this “first level” ladder with 3 shots, or really only one shot each. I do 3, personally, but as depicted above, when neck tension is where it should be, every single-shot series points to the same node. All roads lead to the same place, whether it’s single round, or 20 rounds each charge. I choose 3 as a safe error indicator.
I then shoot them into 3 shot separate targets and look for the smallest group. When I find that I go back and load 6 more in .1gr increments above and below that load and many times I will load that same sample load again to reshoot and verify..
Group dependent methods always eat a buttload of ammo. I started my load development method development journey a long time ago with a very similar method - loading 5 rounds of each charge weight in the first layer, then refining and loading 10 rounds each charge, then doing seating depth tests with another 5-10 rounds… and barrels were half gone before I even found a load… now I shoot smaller groups with far, far less rounds fired. As mentioned above, I’ll shoot my first layer as 3 rounds each, then typically I have my node and don’t refine any further than that 0.2grn interval test, and then if I tune seating depth, I shoot 5 rounds each in the first seating test (second layer of tests).
Equally, as mentioned above about statistical validity of small sample sizes - be VERY careful chasing small groups, even at 6 rounds. POI and group size are the least reliable lag measures (output performance metrics), and elicit the highest variability - meaning they require the greatest number of samples to garner validity. In other words, if you print one group as 4 shots into 1/2” with a 5th hole just separated and the 6th hole spread at an inch, then print another group with 6 shots held at 3/4” in a round ragged hole, statistically, there’s absolutely no difference between those groups… but we want to tell ourselves that one shot smaller, and one kinda scattered…
Again I have to perform it this way because I am not shooting a Match Grade barrel with a Match grade Action and Trigger in a Chassis. Sure it is a lot of work and a lot of components but it has proven to produce the best results for me that is the most repeatable.
I really think you’re chasing observation bias, and being way too hard on your rifle. I don’t do anything different between my custom barreled rifles or my factory rifles - because I don’t need to. They act the same. Sure, the custom barrels might shoot a smaller raw score, and it might be easier to find more bullets the rifle will shoot small, but when we have a good bullet and proper powder, consistent brass, and proper neck tension, there’s really no difference in load development whether I’m shooting my FIL’s old 1973 Rem 700 or my 1996 Ruger M77 or my wife’s 2010 Savage 12, or my Proof Research barreled Savage Striker or my Bartlein barreled Defiance or Seekins Havak… the “science” is the same.
I really believe if you shift to these proven methods which limit your sensitivity to variability (Audette Ladder, Newberry OCW, Satterlee Vel Curve) to find your powder charge, THEN only bother with group sizes when tuning seating depth, you’ll be able to find your “perfect loads” much faster, in the same way the rest of us doing these methods find them.