How important is sd/es

S

Shadow9mm

New member
How important is sd/es to you when working up a load. And what do you consider to be an acceptable sd/es?

Thinking along the lines of velocity.

Note: we seem to be getting into the weeds on statistics and quantity needed for valid statistics among other things. What i am trying to get at, is, what does it nean to you. When working up a load, is the velocity sd/es an important factor to you? Is it something you even consider? If it is, what are your standards? What do you feel is acceptable or unacceptable? More of your subjective opinion on what you deem acceptable.

What, if anything, does it mean to you?
 
Last edited:
J

Jim Watson

New member
What gun, what caliber, what range? Velocity variation is not something you can see or a game animal will feel, or even show on target at short range. But it starts to matter at longer ranges in vertical spread on target.
 
S

Shadow9mm

New member
Jim Watson said:
What gun, what caliber, what range? Velocity variation is not something you can see or a game animal will feel, or even show on target at short range. But it starts to matter at longer ranges in vertical spread on target.
Click to expand...
Thats the question. Is it important to you, and if so, how important, what sd/es do you consider acceptable? If you wanting to break to down by categories or your uses thats totally fine. Just curious to see how others look at it.
 
B

ballardw

New member
Of what? Group size? Velocity? Bullet weight? Powder weight? Case length? Cartridge overall length?

Lots of things to measure and every one of them has SD and ES. <Just being obnoxious, pay no attention>
 
J

Jim Watson

New member
Oh, it's a quiz, not a question.

My PhD friend concluded that a coefficient of variation of 1% was "match grade" velocity in pistol ammo. My bulk ammo has to get lots worse before I make adjustments.

When I was shooting BPCR, a standard deviation in single digits was not hard to attain and single digit extreme variation was possible. I have heard of carefully loaded smokeless target rifle ammo doing as well, but mine wouldn't.
 
S

Shadow9mm

New member
ballardw said:
Of what? Group size? Velocity? Bullet weight? Powder weight? Case length? Cartridge overall length?

Lots of things to measure and every one of them has SD and ES. <Just being obnoxious, pay no attention>
Click to expand...
I was thinking along the lines of velocity.
 
L

ligonierbill

New member
It's nice to see single digit standard deviation, and if it exceeds about 30, I start wondering why. But for real world results, I have very accurate loads with SD near 30 and nice single digit loads that don't group so well. It's just one more piece of data.
 
M

mehavey

New member
I consider 1/2% SD (0.5%) to be excellent at any speed.
(For the Marines, Mickey's little hand would be at 3,000, big hand at 15)
 
R

Rimfire5

New member
0.5% would be 14 fps SD at 2800 fps. That would be excellent for factory ammo, IMO. Even GMM ammo is from 15 to 17 fps in most of my tests.

For me, I would be shooting to get my best hand loads 'tuned' to produce SDs more like 7-8 fps or 2.5% or 2.6% at 2800 fps.
However, I would expect starting hand loads would be closer to factory ammo.
 
A

Aguila Blanca

Staff
When discussing statistics like SD and ES, the sample size is also a consideration. I don't recall which forum it was, but on "a" firearms-oriented forum many years ago I participated in a discussion regarding how many shots were required before a data set had any real significance.

Back when printed gun magazines used to print range reports on firearms, some of them included in their test results not only the group size but also the average velocity, SD and ES of each ammo type used in the testing. But their velocity numbers were typically based on a set of five or maybe ten shots with each ammo type. Statisticians were all over that, maintaining (rightly, I believe) that even a ten-shot sample is woefully inadequate to obtain meaningful numbers for SD and ES.

IIRC, the consensus was that you need a sample of at least fifty rounds -- chosen at random from a much larger sample -- in order for the numbers to be considered statistically significant.
 
R

rclark

New member
More samples the better. But statistically 10 is the minimum for SD results. I like to shoot 15. I used to do 30 ... but found 15 adequate for my needs. Now I don't test rifle, just revolver. I like to see under 50 fps ES. BTW, you'll find SD is usually close to 1/3 of ES, so I pay more attention to ES. Will it matter at 25 yards? Not much. But more noticeable as you reach out to farther distances, so less ES the better. Does low ES guarantee accuracy? No, but low ES 'never' hurts accuracy. My 2 cents (4 with inflation) .
 
J

Jim Watson

New member
I had an article on Small Sample Statistics that described what you could and could not expect. Lost in The Incident.

Standard deviation sounds more "scientific" but to the working shooter, extreme spread is more important. Because ALL your shots count and the random outlier might cost you points or game.
 
S

Shadow9mm

New member
Jim Watson said:
I had an article on Small Sample Statistics that described what you could and could not expect. Lost in The Incident.

Standard deviation sounds more "scientific" but to the working shooter, extreme spread is more important. Because ALL your shots count and the random outlier might cost you points or game.
Click to expand...
How it was explained to me was that, es is a historical record of what happened. Sd is a estimate of what will happen in the future.
 
J

Jim Watson

New member
People get upset when they learn that, assuming Gaussian distribution, the "bell shaped curve", +/- one SD only includes about 2/3 of your expectations.
 
T

totaldla

New member
ligonierbill said:
It's nice to see single digit standard deviation, and if it exceeds about 30, I start wondering why. But for real world results, I have very accurate loads with SD near 30 and nice single digit loads that don't group so well. It's just one more piece of data.
Click to expand...
Pretty much my thoughts as well. If I was into 1000yd stuff, I would care a lot about velocity variation. But I'm not, so I don't. But I am concerned that my reloading process isn't right when I see big swings.
 
4

44 AMP

Staff
Note: we seem to be getting into the weeds on statistics and quantity needed for valid statistics among other things.
Click to expand...

Often these discussions go beyond the weeds and out into the sticks, and sometimes deep into the woods! :D

What i am trying to get at, is, what does it nean to you. When working up a load, is the velocity sd/es an important factor to you? Is it something you even consider?
Click to expand...

While this may seem like heresy to some here, I don't worry about ES or SD or even much about actual MV.

What I focus on are down range results, and that includes many more factors than just velocity variations and statistical data points.

Everyone has their own "game" and what can be a benefit in one game is of little or no use in a different one.
 
T

tangolima

New member
I basically shoot groups. I learn that small sd doesn't guarantee small group size. MV is stronger function. However once a good MV is located, tighter sd will make the group even smaller.

Absolute value of sd doesn't mean much to me. I like the ratio sd/MV, called coefficient of variance, better. Anything smaller than 1% is not bad. 0.5% is excellent.

Hunters like to say animals don't feel the sd. Actually they do, when the hunter misses the intended poi.

I know bigger sample size is better. But I also know having money is better than not having enough. I simply can't afford shooting 10 shots or more for each load combination. We all have our own little "recipe" to guide us to an optimal load quickly and efficiently, even it is only a local optimum.

-TL

Sent from my SM-N960U using Tapatalk
 
J

jetinteriorguy

New member
I’d say it depends on what your goal is. My personal goal was to determine my actual shooting capabilities within what I consider the limit of my personal use in the real world. I’ve tried almost everything under the sun to determine what is my real capability as a shooter and have come to understand I’m a pretty reliable 1 MOA marksman. Meaning I can shoot sub MOA on that rarely good day when everything aligns, but can pretty consistently shoot 1 MOA. As far as my particular real world needs this means I consider 300yds the limit of anything I would ever really shoot at and I can maintain a 1 MOA skill at this range. Now how I feel my criteria fits in with the OP’s question I quit worrying about all the rabbit holes and just develop loads that meet my needs in actual shooting.
 
You must log in or register to reply here.
Top