Holding on that fuzzy target at 200 yards isn't a snap, either!  

{it's not supposed to be}

Extreme spreads are noted, but if you look close, you can see them.

The Win FMJ is at its max accuracy-wise, as can be noted in that I shot a better group from sitting, elbows rested on knees than I did from prone, rifle rested on rucksack.

Nevertheless, Molon's point about gaps in the spacing being filled in over time and with subsequent shots, always seems to occur with 10 or 20 shot groups, clearly indicating the impropriety of simply "ignoring that pesky flyer" when shooting 3 or 5 shots while attempting to get a really hard zero.

You mean the guy at the pawn shop who said his AR will shoot "half inch MOA" was full of shit?

Damn.

In all seriousness, I can see the point of 10-shot groups, but I also think people need to understand the limitations of a particular barrel.  I run lightweight non-chrome-lined barrels almost exclusively on all of my AR's.  I prefer to do 3-shot groups when I'm zeroing, and let the barrel cool off for a few minutes, with an open bolt, between groups.  If you want to test 10-shot groups, I think a lot of the accuracy (consistency) is going to come down to whichever has the heaviest barrel profile, all other things being constant - ammo, barrel material, rifling type, chambering, etc.

For my purposes (hunting) firing a series of 3-shot groups tells me everything I need to know.  If I need more than 3 shots at one game animal, I've taken a shot I should not have taken. For a HD rifle, I'm not going to be shooting at any distance, so accuracy isn't much a concern - I just want a ballistic tip bullet, and group size is of no consequence.  If I were going into combat (longer engagements on targets that shoot back) then heavy chrome-lined (or nitrided) barrels would be my choice.  

And I have to add...Black Hills ammo (with match HP or ballistic tip bullets) seems to generally group about twice as tight as XM193.  XM193 is good for shooting steel and shooting fast. If you're measuring group size, you're just going to be disappointed.

No, no, HIS rifles always produce 1/2" groups.  It's just mine that don't!!  

Also, there are those rifles that have bedding issues that prevent the use of 10-shot groups which in themselves add a variable the rifle otherwise does not produce.  For example, I have owned many old Mauser commercial rifles, Lees, Winchester lever guns and combination over/under rifles that shot repeatable 3-shot groups but if a guy sat down to get a 10-shot group out of the gun the groups strung terribly.  This may be fine for defining accuracy as that strung group, but does nothing to help the hunter get his rifle ready for deer or elk season.  In that case, the 3-shot group is not just "OK", it is superior to the 10-shot group.  That is, the 10-shot group is proving something different than the 3-shot group is.

For target rifles and others where long shot strings might be needed, the 10-shot group is of course superior.

Maybe a good summation is to say a rifle should be proven in accuracy the way it will be used.

Personally, I don't think 5 shots is overly hard on a govt' profiled barrel and getting it to group.  Especially if you give it a bit between shots.  I agree that heavier barrels will take higher volume shooting better though.  

One thing that gets overlooked and is more important than group size is: where is your first shot going to hit.  Or your first few shots.  My issue with my Colt A2 is that it shifts POI with different brands of ammo so terribly, that my confidence in it is not very good.  Because even if you get the same brand ammo, sometimes different lots will affect it.  It's just annoying.  I would far prefer a barrel that is 3-4 MOA but is more consistent at hitting POI than a barrel that will shoot tiny little groups but travels the POI all over the place with different ammo or different lots of ammo.  Usually a barrel that will group well isn't so erratic.  But my Colt will group really well with some ammo, but it's so screwed up as to changing POI that I just don't like it.  

And then the other issue with that is does your AR have that first shot syndrome that is outside the other ones because it was manually loaded.  I think that's how bolt actions have it all over the semi's. It's just more of a repeatable affair, with less to mess up.  But I know some semi's aren't as goofy as others.  It's just been my experience that they CAN be finnicky.  I suppose a bolt gun can too, I don't have as much experience with them.

Maybe someone has already mentioned this but while 3 shots may be a bit on the light side I'm not sure 10 isn't too many. Why not 15?  20?  The problem is you are given more of a change to introduce outside and environmental variables.

Because if you do three sets of ten shots and even one of the shots is a flyer because of an outside variable then it throws the validity of the test out the window. It's not like it was due to the barrel.  It could even be the bullet, even if it's a precision load. I think Molon is awesome and pretty much on point with this but I'm not sure we could ever get a true measure of precision and/accuracy unless the gun was strapped down by the grip and FF guard, the bullets where varified under a microscope, and all of the shots were conducted in a vacuum.

Then I'd be convinced.

Not trying to start a fight just a random thought from my questioning mind when it comes to this topic.

10 is a good compromise between statistical probabilities and excessive use of resources.  More samples is definitely better, but only to a point - too many samples and you're just polishing extra decimal points.  10 shots, with no "called" or excluded shots, provides a sample size that is both economical and provides sufficient data for a statistic-based conclusion.  

A flyer does NOT throw your validity out the window.  Instead, it can point to a variable that you should control for (wind, getting whacked in the head by an AK case while shooting, etc.).  The important thing about real world testing is that it takes into account real world variables like temperature, shooter fatigue, and even the guy with an AK to your left on the range.

Mark Twain quoted British PM Benjamin Disralie when he said "there are lies, damned lies, and statistics."  When people misuse statistics, they aren't being honest.  Period.  Statistics are a tool, useful for learning something about a large group of (whatever), though not very good at telling us something specific about a single (whatever).  Distrusting statistical information is common because so many people lack a basic understanding of how statistics works - and because so many times others have abused "statistics" to lie to us.  

The difference between those lies dressed as statistics and what Molon presents is that he doesn't hide anything.  He presents the data as he collected it, explains the environment, test conditions, test materials, and test process - which is the appropriate method of presenting the results of any scientific investigation.  Given similar equipment and materials, in similar conditions anyone else should be able to produce similar data, as long as they can follow the same processes to the same degree as Molon.  

You don't just do a bunch of tests and see what you get, you investigate one or more "questions," such as "given this particular load, how large/small a group can I produce with this rifle given these conditions?"  A lot of us do that exact thing without actively thinking about formally stating the question, or the hypothesis that goes with it: "I think this load will produce smaller groups than a previous load" (usually because of different variables such as different bullet, different powder charge, or different loading processes).

If used properly, statistics are very helpful, but if you "cherry pick" your data, you're basically cheating yourself.  Honesty with yourself is the best route in all things (he said philosophically), and in evaluating loads, it's essential to really learning what those loads do.  Stats aren't magic, they're measurement tools. But you need to know what those tools do, how to use them, and what the results mean in order to "see" the results as anything but simple numbers.

First of all, in order to obtain a statistically significant assessment of the level of precision that a rifle/ammunition are capable of, three 10-shot groups should be fired in a row.  I’ve demonstrated this time and time again in my accuracy/precision evaluations:  



Molon Reference Links 216




At the 1992 Barcelona Olympics, USA Shooting Team members Launi Meili and Robert Foth won the gold and silver medals in the three-position rifle events.  The Olympians used the new Federal Gold Medal ammunition to aid them in obtaining their victories.  This was the first time in more than 30 years that an American won an Olympic medal in one of the small bore shooting events while using American-made ammunition.

It’s interesting to note that pertaining to the accuracy/precision development and multifaceted testing of the Federal ammunition that helped the US Olympians win gold and silver medals in Barcelona, Federal’s Director of Product Engineering, Dave Longren, had this to say:

“The standard test string was three 10-shot groups, with the most attention paid to the 30-shot composite.  When you’re working at this level, the traditional five 5-shot group test simply doesn’t give you statistically valid results.”



Hunnicutt, Robert.  “Ammo Good as Gold.”  American Rifleman  Nov. 1992:  32-33, 72-73.  Print.













Secondly, here’s “Why not 20?”.  The group pictured below was fired from one of my precision AR-15s at a distance of 100 yards.  It’s a 20-shot group.









Even though there are 20 shots in the above group, I can only distinguish the outlines of 12 bullet holes.  The other 8 shots went through the ragged hole in the center of the group.  So, while I fired an additional 10 shots more for this group, I only gained an additional 2 shots worth of data and lost 8 shots worth of data.  This is highly inefficient and again demonstrates the efficiency of using three, 10-shot groups fired in a row.









Yet another example of someone posting their ignorant opinion as if it were a fact and failing to post any statistically significant data to support their claim.










Please post your data that was obtained with the gun “strapped down by the grip and FF guard, the bullets where verified under a microscope, and all of the shots were conducted in a vacuum.”



...

Unless of course you’re an Internet Commando who claims that the bullet collided with a butterfly in-flight.  Then of course, we have to ignore that shot.



.   .   .   .  

.   .   .   .   .   .   .   .  



....

And here I am, hoping that I can keep each of my 10 shot groups tight enough to merge all the holes...  

I sort of skipped over the need for three groups in my post above.  I think I got sidetracked by emphasizing "science" over "that looks pretty good."  In any case, science wins over opinion any time, as long as science can speak up loud enough to be heard over the din of opinions.

I'm currently working on several loads in several calibers.  So far I've fired ONLY single 10 shot groups, but I'm not looking at precision yet.  I'm looking for indications that I'm on the right track with these loads.  This way I hope to be able to assess the most promising loads for their precision potential, while conserving time and resources.  I think I'm close with a 55gr 5.56 load and a 150gr 300 Blackout load, but the others still need some tweaking.  And still, I test each of these loads with a "hypothesis" in the back of my mind: "which load produces the most consistent, smallest group when fired with the same conditions and from the same rifle."

You're right, definitely. I was just curious though because I figured if we are trying to figure out the precision of a barrel or ammo we'd want to not count human error.

So would that be two different things then, right? The theoretical potential for precision (no human error, no outside variables) and actual precision which accounts for those, including fliers.

Or am I just not thinking about this properly?

*golf clap*. <Comment removed. -Z> I wasn't trying to present anything I was saying as fact. I was simply asking because I honestly don't know. I thought it was pretty obvious.





And what are you even talking about with the last quote?  How hard was my post to understand?





There are two possible avenues here. Potential and actual. That's what I was asking about.





Of course your info is spot on in regards to real world testing. My question is how does one reach conclusions without know all of the variables because news flash, you don't.





Do you have a set of calipers and mics with you, checking to see if your gun moved in between shot?  Because theoretically if it moved, the grouping would be off and it wouldn't be the fault of the barrel.





Do you know for a fact a gust of wind didn't push one of the bullets?  No.





That's why I mentioned the vacuum.  





Because unless you've measured and accounted for everything down to the thpusandths, you are getting real world results and not the actual, true potential of the barrel. The only way to do that is with specific scientific equipment control groups, et cetera.





I was just wanted some clarification on real works vs theoretical.





Zhukov and the other poster seemed to understand what I was saying and gave a good explanation.





<Comment removed. -Z>

Potential is only academic.

Actual is the only test that matters. Unless you live in a vacuum, have access to perfect ammo, perfect parts, and perfect conditions. Variables are always changing slightly.


That's why you got the answer you did. You are refuting the validity of the test with no test of your own.

I've chased the zero and lost many times!

As far as shooting sub-moa and 10 shot groups. Before I started to try my hand at precision, I didn't realize how much effort it would take for me. I remember the first time I did shoot a 10 shot group and it was 2 MOA. Not perfect, but I was happy. I could have picked out two 3-shot groups that were bonafide 1/4 MOA, but that wasn't the point anymore. I wish I had more time at the range for this type of work and effort, but I ended up with a new level of respect for those who can consistently shoot lights out 10 shot groups.

Sorry won't happen again :)

I've shot some spectacularly amazing 1 shot groups, what more do I need?  Some of them actually hit what I'm aiming at.

Seriously, I continue to enjoy and learn from Molon's posts and all the comments.

They have helped me better evaluate many groups I've shot and observed over the years.

My patience and endurance to consistently shoot more than 3 to 5 "as perfect as I'm capable" shots in a row have diminished over the years, even from a front rest and rear bag like Molon uses.  Maybe some kind of sled would help.

For most practical purposes, I will shoot a 10 or 20 shot group of specific ammo for evaluation and zeroing, and if 1/2 of them are in a large bug-hole in the center I'll be happy and use that as reference for zeroing adjustments without worrying about the extreme spread much, especially if the first 3 shots are in bug hole.

If I'm evaluating ammo precision I'll do my best to shoot several 5 or even 3 shot groups, waiting longer between shots.   If the ES and  POI vs. POA is relatively consistent between those groups I won't hesitate to use the results for future reference, but I do recognize that I'm cutting corners from more strenuous, lengthy testing.

Sometimes before I decide something is crap ammo I'll look to see what results Molon posted for it, and if it's much better than mine, I'll just assume I suked.   If my results are better, then I assume I'm hallucinating.

Which does raise the question, how do we know when we've reached the maximum accuracy potential for our weapon/ammo combination vs just being a lousy shot?

I agree that 10 shot groups tell you more however I go a few steps further.
Always and I mean always note where your first cold bore shot goes compared to the rest of the group. No fouling shots! Also note each shot 1-10 on a target or in a shooting log next to you at the bench.
Usually I will shoot one 10 shot group at a deliberate unhurried pace......Then I will shoot another ten after the barrel has cooled at a rapid pace to see how it varies with less chance of the barrel cooling between shots.
This is done with bulk ammo as well as match ammo and will tell me all I need to know.
This is done benched/prone/sitting/kneeling/standing.
Usually I will also do the above as well after a hike or run to get my adrenaline going to somewhat simulate a stress situation like anyone will feel if they are in a SHTF scenario.

The bench baseline will show which shooting position I need to work on the most.
Usually I do this at 50,100,300,500 yards with targets, as well as a long range stage at a 12" plate with no target  at 720 yards with a spotter verifying hits or misses if I cannot make them out with my scope.

Covers-
Hunter/ Sniper- You know where your 1st round is going to hit.
In case of a miss-You know where the following rounds are going to strike.
Prolonged engagement- You know where your going to hit.
Physically stressed and winded- how it changes your accuracy compared to leisurely shooting at the range.

It may sound like a lot but it helps one get to know there weapon and most importantly my own limitations or what I need to work on.
Try it for yourself as it may help you when you need it most.

The next step involves shooting at unknown distances without a spotter or a range finder.....This helps one learn to not depend on technology and to still be effective in a rapid engagement.

You don't want to use one of those.


...

Is there another thread where you discuss why?  You're not the first one to make this suggestion, but I just can't get my head wrapped around how taking the human element out of the equation is not an inherent improvement in the process.

I've never used one before, but I have a friend who does with good results afaik.  Maybe if I tried a few groups with it I would experience something different than what I expect.

I'd like to know as well, please.

Ask your friend to post some pics of his 10-shot groups fired from an AR-15 at a distance of 100 yards using his "sled."


....

100 round groups are good.  1000 round groups  are better.
10 round groups are for suckers.

Heh, fair enough.  You likely have experience with one and I don't, but I'm kinda' stubborn and will still try one out when I get a chance just to see for myself.

My friend is currently evaluating hunting loads for a bolt gun, not an AR.  I don't think he even cares about groups for his AR ammo as he uses them mostly for relatively close range (< 100yds), rapid fire, tacticool fun shooting.

I can conceive of how an AR in a sled could be inherently different from a bolt gun with the bolt hand cycling action vs. semi-auto gas.  My current round evaluations are also for a bolt gun.

I think he's hinting.

"Sled"

Sleds move.

Im sure we all you know mean and I'd say he's joshing with you but I've yet to see that aspect of his personality. Lol

Excellent.  Be sure to share your results with us.


...

Original post updated with abridged version.


....

updated

Thanks again for one of the greatest threads in arfcom history man. Seriously.

This is the GREATEST explanation of the correct method to measure firearms precision ever written. Yet sadly, 15 years after it was written, 99.9% of the Internet Commandos are still dumb and still using 3-shot groups. Sight.

It's like the Regular Car Review of the S2000. It has pOtEnTiaL.

10 cold bore shots.

Discuss...

2020er pulling a zombie thread for a second post!

Sounds like the "Carlos Hathcock method."

I want to try it someday. Time-consuming though.

https://americanshootingjournal.com/carlos-hathcock-method-of-sighting-in-a-rifle/amp/

This "Zombie" thread leads off with one of the most valuable contributions on this website.  It probably ought to be resurrected regularly as old threads rotate to archive and can't be located by normal means.

All of us once were first year members.  Length of board membership may not translate into firearms experience.  For all we know, he might be a NRA certified instructor or a top flight benchrest shooter.

And?


...

I have a Caldwell Lead Sled w/3 25 lb lead bags.  After a few range sessions, I found it more of a hassle to bring out than the mechanical regular rests.  Besides, I'm more a minute of man shooter.

...

still true today!

Fantastic tread.

So, if you shoot a 10 shot group and calculate the center of the group, followed by adjusting your "Zero" to move your point of aim to the center of the group, is it reasonable to expect the next 10 shot group to be more centered around your point of aim? If it is closer, is it reasonable to expect a 3rd cycle and adjustment to yield the most consistent hits with the smallest Mean Radius around your POA?