Velocity or SD - Rifle Loads

Status
Not open for further replies.

Allen One1

Member
Joined
Oct 18, 2017
Messages
405
Location
Nashville, TN
If you were going to choose a rifle load and you had two loads that looked good on paper would you go with the high velocity or lower SD load? Loads were chosen using the expanded 10 shot method to find velocity nodes.

Example:
2784 fps 4.7 SD
2736 fps 2.8 SD
 
The most accurate one if there is any appreciable difference otherwise flip a coin. There is not enough difference in velocity or SD to matter between the two.
That's my problem neither of them stands out from the other enough to offer a clear choice. I'm not sure shooting groups is going to help me either however that would be the best way to separate them if it would.
 
What is the use of the ammo? Does the extra speed help in any way? (Less drop/wind or more energy on target etc?)

Is the faster load significantly closer to max or over max #’s and will it wear out your brass faster?

If you gain nothing significant but wear your brass out faster, the choice is easy

If you run the #s and the speed is beneficial, the choice is easy

If the answer to all is no advantage default to the lowest speed, or load some of both and do a more extensive test in a real world application
 
The ammo is used in a 6.5 Creedmoor shooting steel out to 1250 yards. Velocity does help cut the wind but we are less than 50 fps difference. If they both shoot equally well I will probably go with the higher velocity. Brass has so far not been a problem and I have been running at or a little higher than the higher velocity load for a while. I'm just working on a new series of load tests to squeak out the best accuracy that I can get. Highest hit probability.
 
What was your sample size for your SD numbers? Statistics is a complicated animal.
If both loads yield the same precision your lower SD load will result in less vertical stringing. Having two low single digit SD loads is a great problem to have, wish it were mine.
 
If you have equivalent group size, MV within 50fps, both loads are in a velocity or accuracy node (depending upon before which alter you kneel), and SD’s under 10, I wouldn’t sweat a decision between either of those loads.

If I had loads that similar - both in the nodes, both low SD, both shooting small groups, aka statistically identical, then I might even run a test on my powder dispense system to see which charge weight it throws the most reliably - for example, I know one of my Chargemasters throws 41.8grn H4350 with only an error every 20 rounds or so, but 41.6 will overcharge 1 in 5-6, so loading takes considerably longer if I pick 41.6 than if I pick 41.8.

I’d also look at the neighbors. For example, I shot a velocity profile confirmation profile last week, 3 each of 8 progressive charge weights. Two nodes in the same place as my last two runs, but all 3 tests have shown the same larger divergence in SD above the high node than the error spread happening around my low node. Both nodes are wide with small error bars, but things wiggle a lot more if you inch outside of the high node than if you slip outside of the low. So I load the low.

Run your calculator - the difference in wind drift and drop even at 1,000yrds for ~45fps won’t change much for you, a couple tenths at most, and you’ll see absolutely no difference in vertical dispersion for a 2fps SD difference.

That all depends, if course, whether your “expanded” 10 round velocity profile included enough rounds of each to really establish statistically valid SD’s. If those are 3 shot SD’s, the weight I would give them is very different than if they are 10 shot SD’s.
 
If I had loads that similar - both in the nodes, both low SD, both shooting small groups, aka statistically identical, then I might even run a test on my powder dispense system to see which charge weight it throws the most reliably - for example, I know one of my Chargemasters throws 41.8grn H4350 with only an error every 20 rounds or so, but 41.6 will overcharge 1 in 5-6, so loading takes considerably longer if I pick 41.6 than if I pick 41.8.

Yeah, if those really are the numbers and accuracy is equivalent, then you really are down to things like metering, cost per shot, which one burns cleaner, temperature sensitivity, availability of powder over time, etc.
 
With a sample as small as 10, any estimate of standard deviation has very low precision. Without running any numbers, your two SDs are almost surely indistinguishable.
 
What was your sample size for your SD numbers? Statistics is a complicated animal.
If both loads yield the same precision your lower SD load will result in less vertical stringing. Having two low single digit SD loads is a great problem to have, wish it were mine.
Sample size is 5 rounds of each powder weight not a huge sample but much better than running a total of 10 rounds for the entire test in my opinion. This group of 10 powder charges all had single digit SD numbers I was shocked by that.
 
Yeah, if those really are the numbers and accuracy is equivalent, then you really are down to things like metering, cost per shot, which one burns cleaner, temperature sensitivity, availability of powder over time, etc.
Those are real number but I haven't done the accuracy test yet.
 
Here are the complete load test results if you can read them.
 

Attachments

  • Load Test.jpg
    Load Test.jpg
    112.7 KB · Views: 9
  • 10 Shot Load Development 4-22-2019.xlsx
    16.3 KB · Views: 6
Those are real number but I haven't done the accuracy test yet.
I'm new at the precision game. I understand the velocity node methodology, but why wouldn't you shoot for groups at the same time you gather your V data? I got a labradar just to do this as I understand the magnetospeed may interfere with groups enough to nullify that portion of testing.
It looks like you have a V node at 41.0-41.2, and another at 41.8-42.0. Is that your interpretation as well, and, which would you pick and why?
If it's not too much and appropriate for this thread topic, can you describe your case prep and loading process?
AdApologiesVance for the questions but I'm trying to edumacate myself.

I’d also look at the neighbors. For example, I shot a velocity profile confirmation profile last week, 3 each of 8 progressive charge weights. Two nodes in the same place as my last two runs, but all 3 tests have shown the same larger divergence in SD above the high node than the error spread happening around my low node. Both nodes are wide with small error bars, but things wiggle a lot more if you inch outside of the high node than if you slip outside of the low. So I load the low.
I've read this over and over and don't get it. I would very much like to understand the wisdom you're trying to impart. Perhaps the numbers associated with this might clue me in?
 
Allen One1 said:
This group of 10 powder charges all had single digit SD numbers I was shocked by that.

I want single digit ES numbers and don't pay too much attention to SD. If I shoot 20 rounds and have an ES under 15 fps I'm very pleased, under 10 and I'm ecstatic. A heavier rifle shooting a lower recoiling round already has an advantage since the firearm recoil velocity should be more consistent. Based on your results, I'd shoot the load with the lower ES number, 8 fps I think.
 
Last edited:
I'm new at the precision game. I understand the velocity node methodology, but why wouldn't you shoot for groups at the same time you gather your V data? I got a labradar just to do this as I understand the magnetospeed may interfere with groups enough to nullify that portion of testing.
It looks like you have a V node at 41.0-41.2, and another at 41.8-42.0. Is that your interpretation as well, and, which would you pick and why?
If it's not too much and appropriate for this thread topic, can you describe your case prep and loading process?
paxman
I use a magnetospeed and it can move the point of impact so I do it in two steps to confirm velocity and group size. With that said I shot 45 of the 50 shots into an 1.5 area the exception being the smallest load at 40.2 gr which went crazy.
I'm interpreting it exactly as you have 41.0-41.2 and 41.8-42.0 being the nodes. My current load is at 41.9 and 42 so this is reconfirming the old numbers may have been pretty good.
I can give you the reloading sequence at a high level. De-prime, wet tumble, salt bath anneal, size, trim, deburr, prime. Then each charge is weighed on a Gempro 250 and bullet seated. Nothing to exotic.
 
Magnify your results, if groups are equal at 200 yards are they equal at 500? If so load whats cheaper
Edit- i see youre shooting over 1000 yards, if they are identicle at that distance youre doing something right. I have never had 2 loads perform exactly the same at long distance, theres always some difference even if very minor.
 
Last edited:
Magnify your results, if groups are equal at 200 yards are they equal at 500? If so load whats cheaper
Edit- i see youre shooting over 1000 yards, if they are identicle at that distance youre doing something right. I have never had 2 loads perform exactly the same at long distance, theres always some difference even if very minor.
I've never been accused of doing something right. Cost isn't a big factor between these loads so results should be the driving factor. I really like your idea of magnifying the results to sort out the better load it seems logical but I don't know if I would have thought about it like that.
 
It's almost not worth saying, giving how close these numbers are, but the coefficient of variation for all of the loads in your excel spreadsheet are 0.1% to 0.3%. The three loads at 0.1% are 41.0, 41.2, and 41.6.

So with this limited data set, I'd think 41.0 may be best with its minimal ES. Maybe try 41.1?

FWIW, here's my excel code for solving for coefficient of variation: =IF(COUNT(V48:AO48)>4,(STDEV(V48:BH48)/AVERAGE(V48:BH48))," req sample > 4") .... where V48:AO48 is the range excel checks for each shot's fps. More data points, better data.
 
Last edited:
I can give you the reloading sequence at a high level. De-prime, wet tumble, salt bath anneal, size, trim, deburr, prime. Then each charge is weighed on a Gempro 250 and bullet seated. Nothing to exotic.
Thanks for the info. Given the consistency of your data across all charges, I was interested in what may help to contribute to that. There's a lot of variables, and I don't mind putting in the time but would rather spend the money and time on the right things.
If you do a lands test on your nodes I'd be interested in what you find out. Thanks again.
 
The problem is that statistics tell part of the story but are not the end all be all. I agree that I would choose neither highest velocity, nor lowest SD. Both of those are not as important as ES and ES is not as important as actual group size (especially at known distance targets).

While it takes low ES (consistent ammo) to shoot small groups, the statistically "best" ammo doesn't always translate into the smallest groups. All things being equal, statistics cannot tell you the velocity range that your barrel prefers.

It's the difference between having accurate ammo and having data that says the ammo should be accurate.
 
Status
Not open for further replies.
Back
Top