Does group size really matter or, is Mean Radius what we should be aiming for?

Barelstroker · Mar 18, 2024

To be honest, I'm over seeing bullshit 3 or 5 shot groups that mean little more than bragging rights.
If you guys really want to do something meaningful about making your practice rounds count more than anything you've heard of &, something way better than your bullshit group size, C'mon in & learn what we all should be shooting for unless, we are Bench Rest shooters &, don't really count in the normal world.

Supersubes · Mar 18, 2024

Mean radius.

Barelstroker · Mar 18, 2024

Supersubes said:
Mean radius.

Damn if'n you aint good boi.
But, that aint the only thing thou.

FredHammer · Mar 18, 2024

5 O'clock somewhere I see!

Barelstroker · Mar 18, 2024

FredHammer said:
5 O'clock somewhere I see!

Well, do you know or, are you just another 1/4 MOA bullshitter?

Secant · Mar 18, 2024

It's all meaningful, but I like to see mean radius, radial SD, and group ES.

Dstoenner · Mar 18, 2024

Secant said:
It's all meaningful, but I like to see mean radius, radial SD, and group ES.

View attachment 8375411

I agree.

Take 25 rounds and shoot 5 5 shot groups taking at least the velocity off of 5 and then apply the statistics.

We need the sd of velocity to bound our long distance come ups. 3x SD +/- of the average velocity is the min max velocity. How much is that at the distance you are shooting. Then add to that the same +/- 3x SD of the group size and you get a picture to a 95% confidence of what you can hit at distance.

For a lot of us that isn’t a pretty picture

David

Doom · Mar 18, 2024

Group size is determined by 2 shots. Depending on the number of shots it can show how well a shooter can repeatedly put shots in the same spot or how well a load repeats. It does not show anything of statistical value.

Mean radius uses data from every shot and provides usable statistics for predicting how well the load and shooter are performing.

An expert shooting a rifle of known accuracy can tell a lot from either method. Take away one of the knowns then mean radius will probably work better when developing a load.

Bottom line is both have merit. Understanding those merits is the important thing.

Maurygold · Mar 18, 2024

Shooting paper at 100 or 200 yards means way less to me than the chrono. Few people are shooting enough to be statistically significant at that distance. (This is for rifles shooting under 1 moa with multiple loads)

Ledzep · Mar 18, 2024

Up to about 16-18 shots in a group, it really doesn't matter which one you use because the variability group-to-group is so large that either measurement is only a loose ballpark representation of to total population. Both MR and ES have a ton of variability.

At around 18-20 shots mean radius starts leveling out while group size (ES) stays variable. The more you put into the group the more mean radius pulls away for a meaningful repeatable measurement.

The subject of compiling multiple smaller sample sets gets muddy. Changing conditions group-to-group or day-to-day can influence MPOI slightly so even if you correlate MPOA/MPOI of multiple small sample sets it can be a little noisy. If you don't correlate MPOA between smaller subsets then you're at least partly delusional in interpreting the results.

In other words, I don't care how many 0.1-0.3 MOA 3-shot groups you have, your rifle is still probably a .6-.8 MOA setup at best for 20 shots. If you don't correlate POI to POA your tiny group means jack for practical application & hit probability at long range. You can do the math and see how smaller sample sets naturally produce smaller average extreme spreads... That doesn't mean your rifle is more capable. Your "0.1-0.3 MOA" rifle has the exact same capability as someone who shoots 30-shot samples and has a 0.9 MOA system. When you take a 1st shot at a target, you're pooling from the total population of possible dispersion and that's what drives real-world hit probability, not random chance small sample sets where you go back after the fact and make excuses for "fliers" and delete data points.

Also, it turns out that when you zero your rifle off of 20+ shots, the zero is more repeatable and you don't have to dick with it every match at the 100yd zero board.

Grim's Reapers · Mar 18, 2024

Ledzep said:
In other words, I don't care how many 0.1-0.3 MOA 3-shot groups you have, your rifle is still probably a .6-.8 MOA setup at best for 20 shots. If you don't correlate POI to POA your tiny group means jack for practical application & hit probability at long range. You can do the math and see how smaller sample sets naturally produce smaller average extreme spreads... That doesn't mean your rifle is more capable. Your "0.1-0.3 MOA" rifle has the exact same capability as someone who shoots 30-shot samples and has a 0.9 MOA system. When you take a 1st shot at a target, you're pooling from the total population of possible dispersion and that's what drives real-world hit probability, not random chance small sample sets where you go back after the fact and make excuses for "fliers" and delete data points.

Also, it turns out that when you zero your rifle off of 20+ shots, the zero is more repeatable and you don't have to dick with it every match at the 100yd zero board.

AMEN

TheOfficeT-Rex · Mar 18, 2024

Why are you measuring at all?

Just to dick measure at the range or online? Fine, whatever, 3 or 5 shots, outside to outside and subtract bullet diameter and that 'flier' you pulled and covered with a quarter for scale, just like the fudds do.

Trying to make some sort of load development or tuning decision? Then who cares, use whatever sample size and metric you think gives you the cost/benefit/certainty you need - you're not going to convince anyone here no matter what you use.

ma smith · Mar 18, 2024

Barelstroker said:
To be honest, I'm over seeing bullshit ...

If we shoot 30 round groups, each group would cost $45
you'd only get 10 groups per 300 rounds of barrel life, etc

This places huge constraints on what you can do.

ShtrRdy · Mar 18, 2024

I like to use a personal goal of having my hits within 1/2 moa, (without wind effects).

mrcheesemoose · Mar 18, 2024

honestly the group size (ES) is nearly meaningless. The true population group is done with infinite shots in theory. The thing with ES is as you add shots, ES will get bigger. You can always shoot worse, you can always have a crazier shot, you can always have a worse flier as you shoot more shots in the group. But the mean radius, or median radius (CEP circular error probability) will approach a value that represents the true value of the population as you add more shots (won’t constantly increase), which gives you a valid accuracy of your rifle system. So why would you pick a stat that’s never gonna tell you the true accuracy of your rifle?

JR1200W3 · Mar 19, 2024

This topic is like ranting about people that don't know how traffic circles work. Or people that park in the left hand lane for 30 miles. Are they idiots? Yes. But guess what? They just keep coming. And they just keep people'ing. It's what they do. You can choose to have an aneurysm or not . Makes no difference to them. They'll just keep using the left lane for driving instead of passing and they'll stop before entering a traffic circle and they won't signal when they leave it.

It's really about you....

Gunfighter14e2 · Mar 19, 2024

Barelstroker said:
To be honest, I'm over seeing bullshit 3 or 5 shot groups that mean little more than bragging rights.
If you guys really want to do something meaningful about making your practice rounds count more than anything you've heard of &, something way better than your bullshit group size, C'mon in & learn what we all should be shooting for unless, we are Bench Rest shooters &, don't really count in the normal world.

Just so you know, many here don't give a shit about group size, past first hole placement.

WeR0206 · Wednesday at 2:56 PM

Secant said:
It's all meaningful, but I like to see mean radius, radial SD, and group ES.

View attachment 8375411

This ^^^. Each calc can be helpful but if you can look at ES, mean radius, and radial SD together its much more helpful.

Some good discussion here:

mrcheesemoose said:
honestly the group size (ES) is nearly meaningless. The true population group is done with infinite shots in theory. The thing with ES is as you add shots, ES will get bigger. You can always shoot worse, you can always have a crazier shot, you can always have a worse flier as you shoot more shots in the group. But the mean radius, or median radius (CEP circular error probability) will approach a value that represents the true value of the population as you add more shots (won’t constantly increase), which gives you a valid accuracy of your rifle system. So why would you pick a stat that’s never gonna tell you the true accuracy of your rifle?

You make a great point. I never thought of the group size (ES) as a metric that can only increase with round count but it makes sense.

Something that was mentioned upstream that’s important for folks to keep in mind…For normally distributed data the min recommended sample size to make decisions is around 30. But you can save some money by making rough decisions based on 10-15 round groups. If only doing 5 shot groups then mean radius uses more data points (5) than looking at group size (2) so it’s more than 2x the data points compared to group size. If only looking at one or the other for a 5 shot group I’d take mean radius. In other words mean radius levels out sooner and is less noisy sooner than group size.

Emerson0311 · Wednesday at 6:55 PM

My reading, especially the raw data of BR matches, has convinced me that regularly being <1MOA@300+ for 85% of my bench/prone work is about as good as I will get overall. There is always room to improve, but any shooting not prone or off the bench will prove limited by me, not my system.

Having trouble using the site?

Does group size really matter or, is Mean Radius what we should be aiming for?

Major Hide Member

Heathen

Major Hide Member

FAFOIST

Major Hide Member

2.71828

Sergeant of the Hide

Balding Eagle

Supporter

Bullet Engineer

Operator

Small N, small x̄

Old Salt

Gunny Sergeant

Sergeant of the Hide

Supporter

Hunter/trapper of Remora's

Private

Old Salt

Similar threads