Question about ammo and barrel

[grauhanen]

But let's disregard the actual percentage of barrels equally "liking" the same ammo aside (the above is a hypothetical exercise anyway).

The point is that it should become increasing unlikely that all barrels will shoot the same ammo the same way.

Unless, of course, all good barrels shoot the same lot of match ammo the same way.

What would be the result if a shooter shot the same lot of ammo with two rifles? Would the results be very similar? What would be the result if a third rifle was compared? A fourth? Even a fifth?

Such a comparison might be revealing. Of course it would be of little use to compare only 25 rounds or even 50. Those are too small a sample size. It would be better to compare 100s of rounds with each rifle.

 

[_Shorty]

As I've pointed out many times before, how many shots are in your group determine how confident you can be in your answer. And I've shared the figures involved just as many times. I guess we need to mention it at least once more. 25 shots is what is usually considered the minimum number of shots to be concerned with, usually framed as five 5-shot groups. A single 5-shot group has a confidence window of nearly +/- 44%! Meaning, if you shot a 1.0 MOA group the actual answer for how precise is that ammo on average out of that rifle could be anywhere between 0.56 MOA and 1.44 MOA.

Roughly...
5 shots: +/- 44%
25 shots: +/- 17%
50 shots: +/- 12%
100 shots: +/- 8.2%

You want hundreds of shots instead? How many hundreds is good enough for you? Do you recall roughly how many are required to get an answer that's good enough for you? Is 100 shots already at that mark? Do you have a mark in mind that would satisfy you personally?

250 shots, half a brick: +/- 5.3%
500 shots, a full brick: +/- 3.75%

A brick not enough for you yet? How about two bricks?

1000 shots: +/- 2.65%

That's still more than a 5% swing! Is less than a 5% swing enough for you? What's it going to take to fall under that line?

1100 shots: +/- 2.47%

ok, so 1000 shots wasn't far off from bring it in under a 5% swing. But maybe you're not happy with 5%. That's still kinda sloppy, no? What's a case bring us again?

5000 shots: +/- 1.2%

That's it? How about two cases?

10,000 shots: +/- 0.8%

So two cases is barely bringing the window size under 2%. As you can see, it gets very expensive to get a better answer very quickly. A box of the good stuff is probably $25 even when bought by the case, or 50 cents a shot.

25 shots x 2 rifles x 5 lots = $125
50 shots x 2 rifles x 5 lots = $250
100 shots x 2 rifles x 5 lots = $500
1000 shots x 2 rifles x 5 lots = $5000
5000 shots x 2 rifles x 5 lots = $25,000
10000 shots x 2 rifles x 5 lots = $50,000

You've got to draw the line somewhere reasonable. You can't just get carried away. Unless maybe you're a billionaire. Though it comes to mind that 30 shots seems to be enough for you and your 100-yard challenge. That's just +/- 15.3%. In the Olympics they seem to be happy with 40 shots for each stage in 3P shooting. That's just +/- 13.2% to determine the best in the world. Like I say, gotta draw the line somewhere. For-score smallbore benchrest targets tend to be 25 shots. For me, typically when I lot test I prefer to have two boxes of each lot number, but I will settle for one box. For the test results I shared, that was a decent-sized group of us all lot testing, and we only had so much of the samples to go around, so 25 shots per lot was all we could do. Not ideal, but it can give you some idea of what's what between them. Not enough to definitively say one lot is best, but can definitely tell you whether or not you could settle for more than one lot from the bunch, or whether you want to skip all of them altogether. And for the majority of us in that case, it did indeed tell us the latter. As I mentioned, only one of us bought any that time, despite the smoking deal we would've had if our rifles liked any of it.

 

[grauhanen]

Something that many shooters accept as "fact" because it seems to make sense or it gets repeated often enough is that rounds in a .22LR match ammo lot are all very similar, that one box ought to shoot very much like the next.

When they see different results from one box to the next, or from one day to the next, they assume if it wasn't because of shooter it because they shoot outdoors where, even if winds weren't an issue, something like temperature or humidity explains the different results. Such differences usually are not the reason.

I would expect that almost everyone reading this might admit that more than once they have produced different results with the same good ammo from one box or day to the next.

Wind conditions and shooter error aside, why don't the lots we usually shoot not produce very similar results each and every time? The most basic reason is ammo variation. Despite coming from a brick or box with the same lot number, no two boxes are exactly alike. They may share many very similar individual rounds, but the boxes are not identical.

 

[grauhanen]

Most lots have flaws.

Very nearly perfect lots (similar to the hypothetical lot referred to above) will have the most very nearly identical rounds per box/brick. But that's not what most shooters get. They get lots that are of average quality. A few might get lots that are a little better or worse.

The majority of lots, including the relatively small number of lots we get in Canada, are average lots. They are typically not the best (or most consistent) and so will have flaws.

These flaws may appear as higher than usual variation in box-to-box SD and ES, they may appear as flyers. Some boxes, for example, will have more or fewer flyers than others. The results on target will reflect the variation between one box and another. Identical results are an inevitability.

 

[grauhanen]

Late last year or perhaps in the spring of 2025 I bought a case of the same lot Center X.

After my shooting season began this year, I soon began testing this lot at 100 yards. I shot ten-shot groups with several rifles.

I shot when it was relatively calm as it is very difficult to account for wind or air movement changes over a distance of 100 yards (50 yards alone can be a challenge). Of course, with some targets there may have been more or less air movement between shooter and target than on others.

I compared the same lot at 100 yards with several rifles. At least 300 rounds were tested with each.

Below are the results with two different rifles. The group averages are for five groups of ten shots each, for 50 rounds.



Of course this could be simply a coincidence. Two rifles made by the same manufacturer might by chance shoot the same ammo similarly.

A third rifle can be compared.

 

[grauhanen]

The 100 yard ten-shot averages are very similar.

A fourth rifle can be added to the comparison. They are all virtually the same with this lot of Center X.

The sample size is 1500 rounds of the same ammo. Same ammo, different rifles/barrels, same results.

 

[grauhanen]

Sceptics might think that this only suggests that Anschutz barrels shoot similarly. Perhaps there's some merit to that concern.

I don't have many rifles with barrels of different makes. I have a Bartlein-barreled Vudoo single shot, a V22S. Unless the steel in the barrel originates from Germany, it is unrelated to the Anschutz barrels except for one thing. Like the Anschutz barrels, the Bartlein barrel on the V22S is made to shoot .22LR match ammo well.

Below is how it compares with the Anschutz rifles. Like the Anschutz rifles, it also shot the same lot similarly.

This comparison of the same lot of Center X in five different rifles/barrels was based on 1800 rounds. The five different rifles shot the same ammo the same way. Good barrels and good ammo produce very similar results.

 

[_Shorty]

Something that many shooters accept as "fact" because it seems to make sense or it gets repeated often enough is that rounds in a .22LR match ammo lot are all very similar, that one box ought to shoot very much like the next.

When they see different results from one box to the next, or from one day to the next, they assume if it wasn't because of shooter it because they shoot outdoors where, even if winds weren't an issue, something like temperature or humidity explains the different results. Such differences usually are not the reason.

I would expect that almost everyone reading this might admit that more than once they have produced different results with the same good ammo from one box or day to the next.

Wind conditions and shooter error aside, why don't the lots we usually shoot not produce very similar results each and every time? The most basic reason is ammo variation. Despite coming from a brick or box with the same lot number, no two boxes are exactly alike. They may share many very similar individual rounds, but the boxes are not identical.

Oh, I remember quite well how fixated you were on how one box had different SDs and ESs than the next box from the same lot number. No matter how many times I tried explaining to you that they were correctly considering separate data points from the same set, for whatever reason you just couldn't come to accept that. When you shoot one box past the chrony and it gets 1085 fps avg with an ES of 37 and an SD of 10.5 but the next box shows an average of 1082 with an ES of 33 and an SD of 9.6 it is not correct to consider them seperate datasets because they are from the same lot number. "Shouldn't they be the same?" Um, no. Variance doesn't stop at packaging limitations. The correct thing to do is consider the MVs from every single one of those 100 shots as one dataset and calculate the statistics as such. If they were from different lot numbers, yes, the correct way to handle it would be as two separate datasets. But you have it in your head that your way, considering them different things despite being from the same batch, is the right way. And you either don't understand the math and how it should be applied, or you just ignore how to correctly handle that math because you don't want to do it that way because you just don't want to. Don't forget, 50 shots means a confidence window of rougly +/- 12%. You have to multiply both boxes' numbers by 0.88 and 1.12 to get the confidence window for that number, and if there is ANY overlap of those two windows you cannot say they are different. Any overlap rules out being able to confidently call them different.

But then, that's usually how these discussions go. You have your mind made up, and even when it comes to things that can be objectively shown, if it doesn't already agree with your made up mind then it doesn't matter. You'd rather hold onto your idea than change your mind. The fact of the matter is, every single round is different. And the most alike they will ever be is when taken from the same manufacturing lot. That's why the manufacturers change the lot number as soon as they have to change something. Lead wire spool run out but every other component still the same lot of their own? New lot number. New batch of powder but every other component still the same log of their own? New lot number. New batch of priming compound but still the same brass, powder, and bullet wire? New lot number. They keep the same lot number for as long as all components come from the same respective batches. And that's precisely why box 1 and box 2 from the same lot number shouldn't be treated as if they were separate batches. They're not. They're the same batch. And to properly characterize that batch you count every example from that batch as being of that batch.

On the surface, your shooting results here do indeed make it look like all of your examined rifles shoot pretty much the same with that ammo. I'll also point out that you're just counting two shots from each group rather than counting every single shot and going by the mean radius, which would be more precise. And consequently, a much better way to examine things. Going by the two widest shots in a group doesn't give one a very good idea of what's going on in general. Essentially, your 300 round average is really only examining 60 shots. It's the same reason it isn't a good idea to only look at the ES from chronograph results, ignoring the average and the SD. Measuring the two extreme shots from a group is the same thing. You're ignoring all but two shots when doing so. So while your results do show very, very similar extreme spreads of those groups, that's not telling you the more important aspects of what went on with those groups. You need to examine every shot to learn the more important stuff. The mean radius tells you a great deal more about how it shoots, and it is what you should be concerned with over all other group measurements. The ES is the least interesting number and tells you the least about how that lot actually shoots.

What's better, a 0.984 MOA 25-shot group or a 0.730 MOA 25-shot group?





Or 0.210 MOA mean radius versus 0.215 MOA mean radius? Both with scores of 2250? Can you say one is better than the other? +/- 17% window, and window overlap, remember? But the main point in this case being that you can't tell as much from 0.984 MOA vs. 0.730 MOA extreme spreads as you can from 0.210 MOA mean radius vs. 0.215 MOA mean radius. One tells you how close to centre most shots fall, and the other doesn't tell you anything whatsoever about how close to centre most shots fall. If you want to really know how similar or dissimilar your same-lot/multiple-rifle test's results were you'll have to go back over the results and consider every shot from every group, not just the two extremes. The two extremes tell you very little. You'll have to figure out the mean radii involved and go from there. If you're that curious and would like to know, the confidence window for 300 shots is about +/- 4.8%, and for 400 shots it is about +/- 4.1%, and for 500 shots it is about +/- 3.7%. The extreme spreads you already have aren't all that useful in determining general performance, unfortunately, since you're ignoring most of the shots. But if you figure out the mean radii and their respective windows you can see whether or not they overlap. That's a more studious way to go about it. You could still find out there's overlap galore and consequently end up forming the same general opinion in the end, that the performance doesn't really differ, but at least you'd actually have some considerable math backing up that answer. Not that you should actually go spend an insane amount of time to measure all of those shots. Just sayin'.

 

[grauhanen]

Below are results comparing different rifles shooting the same lots of ammo. The sizes shown are the average of ten consecutive 10-shot groups at 100 yards.

To recap the question:

Do good barrels shoot good ammo similarly?

or

Does each barrel shoot different ammo differently? (i.e. each barrel "likes" certain lots).

Readers should draw their own conclusions about the issue.

 

[_Shorty]

It would be a much more interesting case study with mean radii to compare. To repeat, extreme spread doesn’t really tell you much. Your data kind of exposes that fact. I’ll wager you’d see some significant differences in the mean radii, so it is unfortunate that you do not have those. Ignoring 80% of the shots isn’t very sound.

 

[.22LRGUY]

I think my take on this might mirror Mr. Bill's. Even for a guy like me who sort of lives/breathes 22LR when it comes to guns....threads like these^ are a good reminder of how differently people pursue even the same caliber. The pursuit of accuracy is akin to the guys pushing subsonic ammo to 500 yards with scopes so canted it looks like you could only see your barrel looking through it.

Anyway, it's a nuanced conversation by guys invested in it=interesting read for this gopher shooter. I've talked as passionately about how high I've launched pcs. of pumpkin with CCI Velocitors. I think in my mind...I'm still the same 7 year old with a slingshot in my pocket looking for frogs. lol

 

[grauhanen]

It would be a much more interesting case study with mean radii to compare. To repeat, extreme spread doesn’t really tell you much. Your data kind of exposes that fact. I’ll wager you’d see some significant differences in the mean radii, so it is unfortunate that you do not have those. Ignoring 80% of the shots isn’t very sound.

Shorty, it almost sounds like you have a case of sour grapes. It was probably inevitable that you now complain that the comparison isn't valid. You object to how the comparison is made -- by measuring group size. Sure, there are other good ways of looking at information. There usually are.

The vast majority of shooters on this forum and others evaluate by measuring group size. Everyone understands it. There are .22LR challenge threads on this rimfire forum based on group size. As a result it seems appropriate to this discussion.

Readers should keep in mind that the data is based on ten-shot groups. Ten-shot groups provide more reliable information than five-shot groups. And the comparisons total thousands of rounds. There's reason to be confident in the data's reliability.

 

[_Shorty]

Shorty, it almost sounds like you have a case of sour grapes. It was probably inevitable that you now complain that the comparison isn't valid. You object to how the comparison is made -- by measuring group size. Sure, there are other good ways of looking at information. There usually are.

The vast majority of shooters on this forum and others evaluate by measuring group size. Everyone understands it. There are .22LR challenge threads on this rimfire forum based on group size. As a result it seems appropriate to this discussion.

Readers should keep in mind that the data is based on ten-shot groups. Ten-shot groups provide more reliable information than five-shot groups. And the comparisons total thousands of rounds. There's reason to be confident in the data's reliability.

You once again failing to understand something doesn't mean I have a case of sour grapes. Just means you once again fail to understand something.

 

[_Shorty]

Rifle 1


Rifle 2


According to your method these groups should be considered the same. They're both 1.000" groups with 100 shots in them. Do you honestly think they're equal? Or is there perhaps even a small possibility that 0.419" mean radius versus 0.080" mean radius gives you a better idea of which one shoots better? Because you are literally saying that both of these groups are equal.

1.000" group vs. 1.000" group: It's exactly the same group size, so there isn't even any point in calculating the confidence window, since it will be identical. Same group size means identical performance.

0.419" mean radius for 100-shot group confidence window: 0.385"-0.454"
0.080" mean radius for 100-shot group confidence window: 0.073"-0.087"

0.087" is definitely smaller than 0.385". So no overlap whatsoever. Groups definitely different. Performance definitely differs. Second group is just 19% the mean radius of the first. We can confidently say the second group is better. As if that weren't quite apparent just from looking at the two. But if we measure and compare your way, we still need to test more, or something, because they're the same. Sorry, but they're not the same. I don't think you'll find anyone that would argue that they are the same. But for some reason, that's precisely what you're trying to argue here. Sorry, but the way you're measuring them is not all that helpful.

 

[grauhanen]

Shorty, the view that different good rifles/barrels shoot the same ammo very similarly is supported by the results.

While small sample sizes are unreliable and can easily result in differences between different rifles and ammos, the data shown is based on over 8000 rounds. It all says the same thing.

That's exactly what should be expected from barrels of similar quality shooting the same ammo. They are doing the same thing with the same ammo.

If you wish to go on about the question, to argue that the evidence isn't real or valid, you will do so. Knock yourself out.

 

[_Shorty]

Shorty, the view that different good rifles/barrels shoot the same ammo very similarly is supported by the results.

While small sample sizes are unreliable and can easily result in differences between different rifles and ammos, the data shown is based on over 8000 rounds. It all says the same thing.

That's exactly what should be expected from barrels of similar quality shooting the same ammo. They are doing the same thing with the same ammo.

If you wish to go on about the question, to argue that the evidence isn't real or valid, you will do so. Knock yourself out.

Your data isn't based on over 8000 rounds. Your method ignores 80% of the shots. The last two images I shared should enlighten you. Should. I mean, any reasonable person should be able to discern what's going on in those two images. Oh, now I see the problem...