3 shot group validity

powdergun

powdergun

CGN Ultra frequent flyer
Rating - 100%
244   0   0
I know this is one of those old debates that has been beaten to death and yes a .5 MOA 10 shot group repeated ten times is definitive but my question is as follows.

Can 3 shot groups be used to exclude a load as not worth trying ? In other words if a load shoots 3 shot groups that are not what you want is that enough feedback to drop that load and try another.
 
I suppose its a question of expectations...

If you have a target gun that just shoots bug holes, 3 shots can tell you a lot.

If you have a Winchester 30-30, well 5 shots wont tell you much.

I was out shooting my 223 the other day with the Labradar... first 3 shots had an ES of 1 FPS. Don't take that to the bank just yet. A few more shots and I was up to 15 FPS... not bad for what I'm after, but not the story the first 3 shots told.

On my F Class gun I start my ladder testing with one shot per powder charge. That's all I need.

Once I get the powder charge close, I start testing seating depth in 3 shot groups per OAL. Always document my speeds... always.

Then for certainty, I repeat best of breed a couple times to refine that last bit.
 
Last edited:
If you consider that the first shot of a new load might deviate from previous loads fired and subsequent shots of the same load, in reality you're only getting 2 useful data points which in my opinion isn't very helpful unless they happen to print like a shotgun or keyhole. That being said, if you're purposefully trying cold bore "field" type groups it can be helpful to reduce the amount of time you spend waiting for your barrel to cool back down.
 
From a statistical standpoint, a sample size of 3 is not really indicative of any valid statistical analysis. Whether that is for testing for precision, testing for ES/SD, testing for an optimal charge weight (if you believe in that), or testing the aggregate tail length of alligators in the titi swamp of Florida.

From my own anecdotal experience, which is worth about as much as you are paying for it right now, I've found the answer to your question depends. If I'm doing bullet seating tests, I've had really good luck with Berger Hybrids shooting only 3 round groups. I've shot a lot of Berger Hybrids that I've come to know what jump they like, and they are very jump tolerant, so whenever I'm doing testing I'm already coming from a background of knowing what they like. As a result, my bullet seating tests of 3 round groups aren't surprising, and the results continue to hold up for 5+ round groups. Berger's are excellent projectiles and hold pretty good tolerances, that I feel comfortable with just 3 round groups for testing bullet seating.

Now, when I played around with Nosler RDF's years ago, I would shoot some amazing 3 round groups with them. Some tiny one group holes, thinking I got some winners. Settled on a load that produced a tiny 3 shot hole, loaded up a bunch of ammo, and I was never able to produce any 5 shot groups below 1 MOA. There may have been a couple of 0.75 MOA groups, but regardless, those 3 round groups weren't indicative of sh*t. I could never get RDF's to shoot with any real precision, so I gave up on those, but those were a prime example of why a sample size of 3 is essentially meaningless.

I will say if it doesn't shoot good with 3 shots, it's not going to all of a sudden shoot better with 5+ rounds. If you are confident that the group size of 3 had no external influences (you the shooter, or something else effecting the precision), then arguably that could be grounds to throw it out. But from a statistical standpoint, a sample size of 3 should have a big asterisk beside it, and may not tell the whole story.
 
Perhaps some confusion, like one hears between a "cause" and an "effect". For your group size, want to read up a bit about statistics - you are trying to get to point that you can confidently predict where the next unfired shot will go. As you mention, after having fired 10 shots, you will be more confident - higher level of confidence - where the 11th shot will likely fall into - how far away from a theoretical centre it might be expected to land - between here and there...

Statistics guru's will tell you that there is a "real" distribution of the shots that you will fire. A sample of three shots will give you part of that picture - you may or may not be all clustered up in one corner of the "real group", or may be spread out to maximum size - no way to know - either result is equally as likely due to chance. A sample of five shots gives a better picture of what the "real" distribution might be. and so on - 4 groups of 10 probably very accurate prediction of what the "real" ultimate group really is.

One of the reloading manuals - possibly Sierra? - mentions using 7 rounds as the sample size they use to be satisfied that they are able to predict "close enough". So, take it as a sampling exercise - whether "sighting in", whether "tightness of group", whether "velocity" - without shooting 5,000 rounds, you are trying to predict what the 5,000 group would look like.

If you have ever fired 10 shot groups at a target - say you get a 2" outside to outside grouping - now check how many "1/2" apart triples there are in there. So, did you get a "lucky" 3 shot group, or is that group of three shots actually representative of "real"??

In actual practice, I think I have shot enough over past 45 years to be able to accurately "call my shots". So, when working up a load or sighting in a scope, I want three solid "good shots". I make changes based on that. If my son is shooting, I want to see five shots - he simply does not put in the trigger time that he should. I also confirm results with several 3 shot groups - so my "sighting in" exercise often results in 3 x 3 shot groups, each on a different target - so, I suppose, that is really a 9 shot group. Makes me feel better about it...
 
Last edited:
powdergun said:
I know this is one of those old debates that has been beaten to death and yes a .5 MOA 10 shot group repeated ten times is definitive but my question is as follows.

Can 3 shot groups be used to exclude a load as not worth trying ? In other words if a load shoots 3 shot groups that are not what you want is that enough feedback to drop that load and try another.
Click to expand...

Yes, that is plenty good.

If the goal is 1/4 MOA and the first 3 shots are 1.5 MOA, time to move to another load. I have a post nearby where I show a quick and simple process to narrow down where you want to spend time and resources... and what the groups will look like as you enter and exit a node.

from there, use as many rds as you feel will satisfy your intended role. For a hunter, 3rds might be plenty For an F class shooter, it may be 22rds or more. to fully 'stress test' the set up.

all you want is to have the confidence that the bullet is going to hit the intended target under the tasks desired.

Jerry
 
No way I would argue with Jerry, but on Post #6, I would add one word to his last line - "all you want is to have confidence that the next bullet is going to hit the intended target under the tasks desired". I am sure that is what he really meant...
 
Potashminer said:
No way I would argue with Jerry, but on Post #6, I would add one word to his last line - "all you want is to have confidence that the next bullet is going to hit the intended target under the tasks desired". I am sure that is what he really meant...
Click to expand...

Actually, what would be ideal is... ALL bullets hit the intended target .....

I think that is what all precision shooters strive for.... :)

Jerry
 
If the load won't put 2 or 3 shots close, it certainly won't put 5 or 10 shots closer...
For testing 2 shots out of a fouled barrel are enough to decide whether to tweak it and try again...
 
For an OCW test ( 100 yards ) three will work, small groups are not necessary there, then after an OSD test is picked ( at 2-300 yards ) Shoot a 10 shot group at 300, and see if it passes your current requirements.
 
Potashminer said:
.........

One of the reloading manuals - possibly Sierra? - mentions using 7 rounds as the sample size they use to be satisfied that they are able to predict "close enough". So, take it as a sampling exercise - whether "sighting in", whether "tightness of group", whether "velocity" - without shooting 5,000 rounds, you are trying to predict what the 5,000 group would look like.

If you have ever fired 10 shot groups at a target - say you get a 2" outside to outside grouping - now check how many "1/2" apart triples there are in there. So, did you get a "lucky" 3 shot group, or is that group of three shots actually representative of "real"??

In actual practice, I think I have shot enough over past 45 years to be able to accurately "call my shots". So, when working up a load or sighting in a scope, I want three solid "good shots". I make changes based on that. If my son is shooting, I want to see five shots - he simply does not put in the trigger time that he should. I also confirm results with several 3 shot groups - so my "sighting in" exercise often results in 3 x 3 shot groups, each on a different target - so, I suppose, that is really a 9 shot group. Makes me feel better about it...
Click to expand...

Interesting that it was "7" shots ... It caused me to go searching for something that was at the back of my mind and I found this reference when I went searching for military acceptance standards..

Skennerton, Lee Enfield Story (1993ed.), p212:
“Accuracy tests for the No. 4 sniper rifle involved placing 7 out of 7 shots within a 5 inch diameter at 200 yards & 6 out of 7 shots into a 10 inch diameter at 400yards.”
Click to expand...
 
AP said:
Interesting that it was "7" shots ... It caused me to go searching for something that was at the back of my mind and I found this reference when I went searching for military acceptance standards..
Click to expand...

Not exactly the precision standard that most of us would want to replicate...
 
kthomas said:
Not exactly the precision standard that most of us would want to replicate...
Click to expand...

Well I dont know - perhaps not by todays experts ..... but given that there were about 25,000 No.4 T rifles produced under wartime conditions and, on any day, any, random, rifle pulled from the rack and loaded with 7 service issue rounds had to keep all 7 rounds in a 2 1/2 " circle at 100 yds .... today and tomorrow .. and tomorrow. If it didnt it wasnt put back in the rack. Thats not too shoddy IMO. Its not so much that it couldnt (and obviously many did) shoot tighter ... it just wasnt acceptable to shoot worse.
 
There was something about the stats at 7 shots getting much better. The validity of the data improves significantly as you add shots at these low numbers of shots. Several of the surveys we do need a minimum of 36 data points before the gov will acccept the results, doubt that many are going to put that many shots in a group.

Like several have said if your 3 shot group is not at the level you want, it sure won’t get better by adding more shots.
 
IMHO, 3 shot groups will let you know what a "hunting" rifle is doing if they are repeatable and to the same POI on multiple groups.
 
The ability to call your shots, that is to know whether or not the sight picture was correct when the trigger was squeezed plays a large factor in how many shots are required. I shot a police 100 meter qualifying round recently (prone, kneeling, offhand, weak hand and standing rest) and while walking to the target I stated that I had 2 flyers and when we got to the target I had two bullet holes outside the bull. It takes practice to be able to call the shots but when I make a new handload I can generally tell within two or three shots whether the load has potential or not and the loads that do have potential take quite a few more shots to verify that the groups are repeatable.
 
AP said:
Well I dont know - perhaps not by todays experts ..... but given that there were about 25,000 No.4 T rifles produced under wartime conditions and, on any day, any, random, rifle pulled from the rack and loaded with 7 service issue rounds had to keep all 7 rounds in a 2 1/2 " circle at 100 yds .... today and tomorrow .. and tomorrow. If it didnt it wasnt put back in the rack. Thats not too shoddy IMO. Its not so much that it couldnt (and obviously many did) shoot tighter ... it just wasnt acceptable to shoot worse.
Click to expand...

We are talking about military battle rifles of WW2, not precision match rifles however.
My Long Brach will hold 1 1/2MOA very easily, great for service rifle matches but a complete failure when we are talking accuracy for Palma match rifles.
Cat
 
powdergun said:
.....

Can 3 shot groups be used to exclude a load as not worth trying ? In other words if a load shoots 3 shot groups that are not what you want is that enough feedback to drop that load and try another.
Click to expand...

So, your question implies that any issue might be about the load - what you are really up against, though, is there is also the shooter's role in those results. If you have never shot a .5 MOA group, ever, and if your three shot test group is 1.5 MOA, it will be up to you to decide if that was mostly due to the load, or mostly due to you? If your next group with same components is significantly better or worse, than that is kind of pointing at the shooter? At the rifle? At the scope? So, need some sort of base line to start with - this is what the shooter can do with this rifle set-up - therefore this change is (or is not) due to the changes that were made to the load.
I suppose that leads to obvious "observation" that if your free floating barrel is actually bearing on one side of forearm of stock, or if your receiver bedding has gone punky, then playing with the loads used is not going to fix it...
 
Last edited:
My "hunting" accuracy criteria is based on 3-shot groups.
However, the load must always place 3 shots into that
group size from a fouled barrel.

In simple terms, when I pick up that rifle, and fire a 3-shot group
that was .75 moa 2 weeks ago, it must make a .75" group today.
If it consistently does this. it goes hunting. If not, more development
is in order.

For precision rifles [including varminters] I will want more shots in a
group to verify consistency. Dave.
 
You must log in or register to reply here.
Back
Top Bottom