How much "Extreme Spread" is acceptable in competition?

[Mulligan]

@Mulligan

Where's the target picture?

Tom

@tom My Dasher at 600 with different weight primers
CW

 

[JimSC]

CW take the cup off the Forester and catch them in your hand, a little more work but I have done this before!

Joe Salt

I color code primers with different colors with a sharpie when loading test rounds in the event I mix up the rounds somehow.

 

[243Mendoza]

how useful is extreme spread: keep shooting more and more shots and your extreme spread with keep spreading.

 

[Ned Ludd]

how useful is extreme spread: keep shooting more and more shots and your extreme spread with keep spreading.View attachment 1079156

Your image is about SD, not so much about ES.

 

[243Mendoza]

Your image is about SD, not so much about ES.

True but it illustrates that the extreme spread is extreme A more useful and obtainable measure of 'extreme spread' is a multiple of sd around the average value.

 

[Evan]

ES tells you what to expect for fliers. That said, ES when you shoot 5 is not valuable. 5 barely even gets you SD; it's the mathematical minimum amount. Shoot 20 and both SD and ES are more descriptive of the load's performance.

 

[243Mendoza]

ES tells you what to expect for fliers. That said, ES when you shoot 5 is not valuable. 5 barely even gets you SD; it's the mathematical minimum amount. Shoot 20 and both SD and ES are more descriptive of the load's performance.

The more shots you shoot, the larger ES becomes. The more shots you shoot, the SD will converge to a steady value.

A better method to check the statistical meaning of your test is the use of confidence intervals. This link, https://www.socscistatistics.com/confidenceinterval/Default2.aspx,
is easy to use and useful.

 

[Evan]

The more shots you shoot, the larger ES becomes. The more shots you shoot, the SD will converge to a steady value.

Not true. If your SD is convergent, you are working with a physical system (not random), and the ES will be asymptotic, not divergent, as you shoot more and more shots.


EDIT: I'm not being totally clear here. I mean to say that in a physically bound system like this where we are actively controlling variables and limiting the system, ES will be asymptotic.

 

[CharlieNC]

Not true. If your SD is convergent, you are working with a physical system (not random), and the ES will be asymptotic, not divergent, as you shoot more and more shots.


EDIT: I'm not being totally clear here. I mean to say that in a physically bound system like this where we are actively controlling variables and limiting the system, ES will be asymptotic.

Yes, but it will take so many shots for the ES to converge that no one will realistically shoot that many shots at any one time to realize that.

 

[Dusty Stevens]

What is the typical winning sd/es score at a match? Ive never seen it on a match report?

 

[CharlieNC]

What is the typical winning sd/es score at a match? Ive never seen it on a match report?

I only know about mine. But I doubt the winners shooting beyond 100 yards would say it doesn't matter at all, and is of course secondary to being in a good node. But you already knew that.

 

[Bart B.]

The more shots you shoot, the larger ES becomes.

Up to a point where the cost gets too high for each X% increase.

Lake City arsenal explained all this to me back in the 1970's. They figured decades earlier that 270 shot test groups for 30 caliber ammo was enough. Then developed a rule of thumb as follows:

* 40% of all shots are in the inner 40% of extreme spread,
* 30% in the next outer 30%,
* 20% in the next band,
* outer 10% of ES has the remaining 10%.

The few 270 shot plots of national match ammo I've seen reflect that very well.

 

[243Mendoza]

Not true. If your SD is convergent, you are working with a physical system (not random), and the ES will be asymptotic, not divergent, as you shoot more and more shots.


EDIT: I'm not being totally clear here. I mean to say that in a physically bound system like this where we are actively controlling variables and limiting the system, ES will be asymptotic.

Yes, ES will be asymptotic but if if you look at the normal distribution curve, the tails which are a measure of spread will take a long time to reach an asymptotic value while the sd will converge much more quickly or in other words, less shots needed to find sd than es.

 

[chkunz]

Up to a point where the cost gets too high for each X% increase.

Lake City arsenal explained all this to me back in the 1970's. They figured decades earlier that 270 shot test groups for 30 caliber ammo was enough. Then developed a rule of thumb as follows:

* 40% of all shots are in the inner 40% of extreme spread,
* 30% in the next outer 30%,
* 20% in the next band,
* outer 10% of ES has the remaining 10%.

The few 270 shot plots of national match ammo I've seen reflect that very well.

Bart B. this is real simple, the more shots you shoot the larger the ES.
Best wishes. Clyde the elder.

 

[chkunz]

Yes, but it will take so many shots for the ES to converge that no one will realistically shoot that many shots at any one time to realize that.

ES will never converge, it will only get larger.

 

[Bart B.]

Bart B. this is real simple, the more shots you shoot the larger the ES.

I agree. But it stops when the shooting stops and the shooter decides more shots isn't worth it. The rate it increases per shot decreases as more are fired

 

[AJC]

Its my first winter on this board and i am guessing that this is what happens when things get frozen? Im just mind blown and wondering if there is a giant pile of did not make the cut primers bullets and cases sitting round. Does the not make the cut stuff turn into hunting rounds

 

[Dusty Stevens]

Its my first winter on this board and i am guessing that this is what happens when things get frozen? Im just mind blown and wondering if there is a giant pile of did not make the cut primers bullets and cases sitting round. Does the not make the cut stuff turn into hunting rounds

Oh it gets worse

 

[6MMsteve]

What is the typical winning sd/es score at a match? Ive never seen it on a match report?

lol

 

[Evan]

ES will never converge, it will only get larger.

So if you shoot long enough would you get one that goes 10000 fps? If you shot long enough would you get one that goes -1fps? Of course not. This is a real, controlled system and as such ES will reach natural limits. I maintain that it would be a reasonable number for lovingly prepared reloads, maybe something like 30 or so, and would represent how well you've culled the outliers ahead of time.