A better way to evaluate and compare accuracies...

A better way to evaluate and compare accuracies...

This is a discussion on A better way to evaluate and compare accuracies... within the General Firearm Discussion forums, part of the Related Topics category; I know I've been gone for a while - complications and busyness (busy-ness, not business). My wife's brother is having health issues and we're all ...

Page 1 of 2 1 2 LastLast
Results 1 to 15 of 20
Like Tree32Likes

Thread: A better way to evaluate and compare accuracies...

  1. #1
    Lead Moderator
    Array Tangle's Avatar
    Join Date
    Dec 2004
    Location
    Chattanooga
    Posts
    12,935

    A better way to evaluate and compare accuracies...

    I know I've been gone for a while - complications and busyness (busy-ness, not business). My wife's brother is having health issues and we're all he's got, and he lives 100 miles away in Knoxville, so I've become too familiar with I-75 OTOH, the fringe of the trip is I get a McDonald's sausage McMuffin when we go - I think I'm addicted

    In between brother trips, I've spent a lot of time reloading, shooting, and chronographing 223/556 and even some 6.8 SPC - I really like the 6.8 SPC! The chronoing is part of my long term plan to see how extreme spreads (ES) and standard deviations (SD) occur over a large variety of loads and a huge number of rounds. This has generated some very interesting things, which I won't get into here, because you probably would rather not know anyway, no offense but ignorance is bliss in this case.

    Part of my effort is trying to evaluate accuracy. And here I use the term accuracy in its erroneous widespread use of group size. That's not accuracy, that's precision, but seems like just about everybody uses "accuracy" to mean group size - and I will continue the erroneous use of the term here for familiarity's sake.

    The main problem with evaluating accuracy (precision) is, as I'm sure we've all experienced, is how to deal with pesky outliers, or fliers. Traditionally, there have only been four ways to deal with them, ignore them altogether (seemingly most popular), include them as part of the group, explain them as a "called" shot, or in lesser occurrences, average groups to lessen the effect of outliers. The latter is popular with gun writers as a way to make guns, ammo, and shooter look better than it really is.

    Since outliers seem to be an integral part of shooting, let's think for a moment about what causes them and there aren't many things to think about. Let's start with what I consider a legitimate reason for excusing or excluding an outlier - the shooter induced outlier, or a called shot. This does happen, and I believe it is one of the justified reasons for not counting an outlier as a part of the accuracy. This does have the caveat, that if you're assessing gun, ammo, AND shooter, that "called" shot has to be part of the group even if it was shooter induced. OTOH, if we're only assessing gun and ammo, we would want to exclude shooter induced outliers because it would not accurately indicate the performance of the gun.

    The environment can cause outliers too - mainly the wind. Unless we are intentionally evaluating accuracy in the presence of wind, then I believe outliers caused by wind drift due to gusts etc, can be legitimately excluded from the group.

    One other thing that could cause an outlier is imperfect bullets, typically out of balance bullets. Bullets can have cavities in the lead which can cause an off-center balance which would likely produce a spiral type trajectory resulting in outliers random in nature. This effect can clearly be seen in slower moving pellets or slugs photographed with high speed cameras. Because the pellet is moving so much slower than a centerfire rifle bullet, we can see things about the pellet/slug that we couldn't capture with a high speed rifle bullet. I've watched a couple of videos about pellet gun barrel development and certain combinations of barrel configurations and pellet/slug weights. These clearly reveal straight and spiral trajectories. As soon as I saw this, I realized a spiral trajectory could readily explain those heretofore unexplainable, seemingly random outliers. So we have two ammo related things that could possibly produce outliers that are truly due to gun/ammo, out of balance bullet and bad match between barrel configurations and bullet weight.

    Unfortunately, the temptation to use some of these legitimate reasons to excuse all, or most, outliers is great. It's awfully easy to look at that otherwise tight little group with maybe one or two outliers and awfully hard not to start making up reasons to not count the outliers. Yet, sometimes, as discussed above, there are justifiable reasons to exclude an outlier. It is up to us to assess the cause of each outlier honestly. If we don't, we're just fooling ourselves.

    So, now that we have a better understanding of outliers, how might we assess accuracy in the presence of outliers that we don't feel like we are justified in excluding?

    Let's look at this target as an example. I shot this just the other day.



    We're looking at the upper group - ignore the writing, I'll get to that later. This was a 10 shot group with some of my reloads shot at 200 yards shot from my BSF Carbon Fiber 14.5" "pistol".

    Clearly, there are two outliers. I'm not sure what caused them so since I can't produce a legitimate reason to exclude the outliers, I must deal with them somehow. Traditionally, you ignore them or you include them, but I believe I have a better way. My way does not exclude fliers but at the same time gives me a better quantitative value. Here's how it works. I measure the tight 8 shot group in the red which measures 1.55" ctc or 0.74 MOA. Now I surely like that group! Eight shots in a 0.74 MOA - I like it! But, I have ignored the outliers, so now I'll take them into mix. Here's what I think is a better way to evaluate accuracy. I shot 10 shots; 8 of them were in a 0.74 MOA group. That means, 80% of my 10 shots hit in a 0.74 MOA group. That doesn't ignore the outliers, and it tells me that I have an 80% chance of getting a hit within 0.74 MOA. It also tells me there's a 20% chance a hit will be outside the 0.74 MOA group. Ten is a statistically small number but there's more to come.

    The overall group, i.e. the full 10 shots measured 1.07 MOA, and that's not bad either, but that alone would not indicate as accurately what I could expect from the gun. I'll need to explain that.

    Now let's look at that lower five shot group. I had three targets up and shot 15 shots of the same loads in three different brands of cases, Lapua, Winchester, and Hornady. Lapua won BTW, more consistent accuracy and more consistent velocity. I shot 5 shots of the first brand, then 5 of the next brand on the next target, then 5 shots of the third brand on a third target, and repeated that 2 more times. That was to try to keep things as even as possible.

    Unfortunately, for some reason that made sense at the time, I changed the elevation on the scope after the first 5 shots - I wish I hadn't done that! Anyway, on the target shown above, that lower five shots is before I changed the elevation and the upper ten shots are after I raised the elevation. However...

    That lower 5 is just a wee bit over 0.74 MOA, it's in fact a 0.78 MOA. So now, I have a 10 shot group that measures 0.74 MOA and a 5 shot group that measures 0.78 MOA. And, notice the horizontal position of the lower 5 lines up almost exactly with the upper 10! So going by the larger group, 0.78 MOA, I have 13 shots out of 15 that hit in a 0.78 MOA group - and they were shot at three different times! Using my method, I would have 13/15 shots in a 0.78 MOA group which indicates 87% of 15 shots landed in a 0.78 MOA group. It also indicates some missed -13% to be exact.

    This doesn't ignore outliers, but rather indicates what I could expect from this gun with this load. And, I should see that 87% of the time, although admittedly 15 shots is still kinda statistically low.

    BTW, I failed to mention this, but these were shot with an ATN X sight 4K Pro digital sight.

    You could use this method as I did or you could do some variations. For example you could set a hopeful group size, shoot a minimum of 10 shots, preferably 15 or even better, 20, and see what percentage landed in the group size hoped for.

    Another variation would be to see what size group you could get for a 95% (or whatever percentage you choose) hit percentage.

    I think the ultimate performance test is to shoot in one group and shoot until the group stops expanding - that's what we can expect! The problem is, how many do we shoot after we don't see any more expansion?

    Whatever, bear in mind, the truth comes in large numbers, not 5 shot groups. Even 10 shot groups are "edgy", what's really needed to really know what your gun/ammo does is 20 shots - even more if you can bear it.
    We don't have a gun problem in the US, We have a people problem.
    The problem we have is people that want to kill large numbers innocent people
    in Gun Free Zones.

  2. #2
    VIP Member Array G26Raven's Avatar
    Join Date
    Mar 2012
    Location
    Southwest
    Posts
    7,893
    Very interesting! If I could just come up with a way to know whether the outliers are me, the wind, or the ammo, I would feel a lot better!
    Rockymonster and Tangle like this.
    You are your own first responder.

  3. #3
    Lead Moderator
    Array Tangle's Avatar
    Join Date
    Dec 2004
    Location
    Chattanooga
    Posts
    12,935
    Quote Originally Posted by G26Raven View Post
    Very interesting! If I could just come up with a way to know whether the outliers are me, the wind, or the ammo, I would feel a lot better!
    It's always the wind, earthquake, wind gust, sneezing just as the shot breaks, wind blowing grit in you eyes, etc. - it's rarely us

    But, yeah, isn't that the big question
    We don't have a gun problem in the US, We have a people problem.
    The problem we have is people that want to kill large numbers innocent people
    in Gun Free Zones.

  4. Remove Advertisements
    DefensiveCarry.com
    Advertisements
     

  5. #4
    VIP Member Array G26Raven's Avatar
    Join Date
    Mar 2012
    Location
    Southwest
    Posts
    7,893
    Quote Originally Posted by Tangle View Post
    It's always the wind, earthquake, wind gust, sneezing just as the shot breaks, wind blowing grit in you eyes, etc. - it's rarely us
    But, yeah, isn't that the big question
    Most of the time, I know when it is me. But then there are those, "Why the hell did it do that?"
    Tangle, Rockymonster and airslot like this.
    You are your own first responder.

  6. #5
    Lead Moderator
    Array Tangle's Avatar
    Join Date
    Dec 2004
    Location
    Chattanooga
    Posts
    12,935
    Quote Originally Posted by G26Raven View Post
    Most of the time, I know when it is me. But then there are those, "Why the hell did it do that?"
    I am privileged to get to shoot so often with a very good PRS (Precision Rifle Shooting) competitor, with all the top tier guns, ammo, and optics. I've seen him shoot several 10 shot, 0.3 MOA groups at 300 yards! He's good! We do a lot of testing at 100 yards and it is not frequent, yet not necessarily uncommon to see one of those "Why the hell did it do that?" shots. It just seems unbelievable and unexplainable how that could happen. Just last week he had a "many in one hole" group going and the next shot landed about 2 inches from the group???

    So, you're in good company about strange outliers. Right now, I believe those are caused by bullet imperfections. My buddy is a meticulous reloader. His scale measures to 0.02 grains! His typical numbers are ES's around 10 and SD's around 5 or less. It could be he does something wrong, but my inclination is that a bullet got produced with a balance point off center and it spiralled.
    G26Raven likes this.
    We don't have a gun problem in the US, We have a people problem.
    The problem we have is people that want to kill large numbers innocent people
    in Gun Free Zones.

  7. #6
    Ex Member
    Array 1942bull's Avatar
    Join Date
    Jun 2018
    Location
    SE PA
    Posts
    2,103
    Hats off to the OP whose analysis seems to be right on target so to speak. Here is something I observed in the Corps as I was assigned to do a photo story on the Corps' new sniper training program. Needless to say the training is intense and only the best pass the tests. Anyway my observation was that candidates that had survived to be near completion of the course were amazing shooters. The was little to none operator error, and yet there were rare outliers (not very often). most of those outliers were on the horizontal center target axis. The drift was mostly (almost always) right to left of the vertical axis. I mentioned that to a trainer and asked if that was the norm, and what did he attribute it to. He said R & L drift were the most common deficiencies, and then he said t was due to setting windage wrong. He went on to give me an explanation of wind effect on a bullet that is losing velocity over a long distance. Anyway it was interesting to see it and get an explanation.

    I do not shoot rifles, only pistols and mostly indoors. I get an occasional outlier, and I tell everyone it was due to the air current caused by the ventilation fans. No one believes me.

  8. #7
    Lead Moderator
    Array Tangle's Avatar
    Join Date
    Dec 2004
    Location
    Chattanooga
    Posts
    12,935
    Quote Originally Posted by 1942bull View Post
    Hats off to the OP whose analysis seems to be right on target so to speak. Here is something I observed in the Corps as I was assigned to do a photo story on the Corps' new sniper training program. Needless to say the training is intense and only the best pass the tests. Anyway my observation was that candidates that had survived to be near completion of the course were amazing shooters. The was little to none operator error, and yet there were rare outliers (not very often). most of those outliers were on the horizontal center target axis. The drift was mostly (almost always) right to left of the vertical axis. I mentioned that to a trainer and asked if that was the norm, and what did he attribute it to. He said R & L drift were the most common deficiencies, and then he said t was due to setting windage wrong. He went on to give me an explanation of wind effect on a bullet that is losing velocity over a long distance. Anyway it was interesting to see it and get an explanation.

    I do not shoot rifles, only pistols and mostly indoors. I get an occasional outlier, and I tell everyone it was due to the air current caused by the ventilation fans. No one believes me.
    Very good!

    Wind is a problem - even for the best shooter. One technique they use to eliminate the effect of wind is to just measure the vertical dispersion and use that for group size. It seems the wind, unless it has a significant up or down draft component doesn't affect the vertical, at least not much. So when one has to shoot in windy conditions, ignoring horizontal may help give a better idea about accuracy.

    What! They don't believe you about the ventilation fans! Awww, what do they know!
    G26Raven and 1942bull like this.
    We don't have a gun problem in the US, We have a people problem.
    The problem we have is people that want to kill large numbers innocent people
    in Gun Free Zones.

  9. #8
    VIP Member Array Struckat's Avatar
    Join Date
    Feb 2014
    Location
    Lurking
    Posts
    3,893
    I know I wiggle a lot and that my definition of ďAccuracyĒ is different than yours.
    I donít remember you mentioning the 6.8spc. I think itís a fantastic cartridge. I have an AR and a Savage model 10 that I barreled with an AR Performance custom with their 3R rifling.

    Iíve loaded a few with my preferred hunting bullet, 110 Accubond, but havenít worked up any specific loads. I lucked into a mountain of factory ammo, the Federal 90gr Gold Dots which has distracted me from making my own.
    Both rifles shoot that round so well.

    I have accumulated all of the components and am about to embark on making a bunch test rounds for both.
    I say that as the entire pipeline downstairs is full of 38 spl and 357 mag brass. So many distractions.

    Four years ago I went all in on the 6.8 and currently own no rifles in other cartridges. Works just like a .243 on whitetail inside 300 yards, with maybe 10% more recoil than a 5.56
    G26Raven and Tangle like this.
    Read our government's plan to destroy the Constitution.
    https://sustainabledevelopment.un.or...ormingourworld

  10. #9
    Distinguished Member Array CavemanBob's Avatar
    Join Date
    May 2015
    Location
    Kalirado
    Posts
    1,678
    Quote Originally Posted by Tangle View Post

    I think the ultimate performance test is to shoot in one group and shoot until the group stops expanding - that's what we can expect! The problem is, how many do we shoot after we don't see any more expansion?

    Whatever, bear in mind, the truth comes in large numbers, not 5 shot groups. Even 10 shot groups are "edgy", what's really needed to really know what your gun/ammo does is 20 shots - even more if you can bear it.

    42 is the number you're looking for.
    M1911A1 and SAXDM9 like this.
    The best citizenry is an informed citizenry

    NRA Patron member
    Gun Owners of America
    Electronic Frontiers Foundation https://www.eff.org/ - defending digital privacy and free speech including forums such as this one

  11. #10
    Lead Moderator
    Array Tangle's Avatar
    Join Date
    Dec 2004
    Location
    Chattanooga
    Posts
    12,935
    Quote Originally Posted by Struckat View Post
    I know I wiggle a lot and that my definition of ďAccuracyĒ is different than yours.
    I donít remember you mentioning the 6.8spc. I think itís a fantastic cartridge. I have an AR and a Savage model 10 that I barreled with an AR Performance custom with their 3R rifling.

    Iíve loaded a few with my preferred hunting bullet, 110 Accubond, but havenít worked up any specific loads. I lucked into a mountain of factory ammo, the Federal 90gr Gold Dots which has distracted me from making my own.
    Both rifles shoot that round so well.

    I have accumulated all of the components and am about to embark on making a bunch test rounds for both.
    I say that as the entire pipeline downstairs is full of 38 spl and 357 mag brass. So many distractions.

    Four years ago I went all in on the 6.8 and currently own no rifles in other cartridges. Works just like a .243 on whitetail inside 300 yards, with maybe 10% more recoil than a 5.56
    Thanks for that! That's very encouraging about the 6.8 SPC!

    My SPC is an AR with a Wilson Combat 8" barrel. I haven't reloaded a lot for it but what I've seen so far is far more impressive in numbers and equal in accuracy to what I see with 223/556. I actually went to the 6.8 SPC because of inherent, apparently irresolvable, issues with the 223/556 caliber. This will probably start WWIII, but I'll share some of those issues of the 223/556 here. And guys, remember, this is based on reloading AND chronoing over 6,000 223/556 rounds, as well as factory rounds. I have shot these in heavy barrels, long barrels, short barrels (as short as 10.3") beefy barrels and skinny barrels. All chronographing was done with either a MagnetoSpeed or Labradar and sometimes both. I've gone from shooting 5 round groups to 15 and 20 round and with the 6.8 SPC two 230 round groups.

    1- The 223/556 is a small case. Even Lapua cases have a lot of case weight variation and while weight is NOT proportional to case volume, and I've confirmed this numerous times I suspect that case volumes do vary quite a bit which varies pressure for a given load, which in turn varies velocity.

    2- The 223/556 is kind of a "no man's land" cartridge in the precision discipline. Very few shoot 223/556 in long range matches. There are few "precision" matches that the 223/556 can compete in. The 6mm and 6.5mm rounds dominate. Where this leaves the 223/556 is the manufacturers put their efforts in to the more popular "precision" calibers and let the 223.556 fall where it may in mass production.

    3- Did I mention the 223/556 is a small case? It is. Aside from the issues described above, that small case means less brass surrounding the primer pocket which makes it weak. I lose about a third of my 223 cases, be it Lapua, Hornady, IMI, whatever to primer pocket enlargement (PPE). I've had primers blown out from shooting, fall out in tumbling, and "mush" out (and in) when decapping or priming. I thought this was just me until I watched a Johnny's Reloading Bench where he was seeing PPE too.

    I used to wonder why my buddy, shooting much more powerful calibers didn't have this problem. The reason is his cases are significantly larger than 223/556 which means far more brass around the primer pockets than a 223/556 has, so they don't stretch nearly as much. Also, the 556 is a high pressure load, which stresses the primer pocket even more.

    4- chambering/dies: Most "precision" calibers are of a single caliber and hence specification. Unfortunately the 223/556 have different chamber specs. The 556 has a longer throat to deal reduce the pressure peak from the higher 556 charge. Plus, the 556 chamber leans more toward reliability than precision. Then there's the Wylde that's supposed to be the best of both worlds. What it really does is accommodate both the 223 and 556 but there are still issues.

    The dies that are available for 223 and 556, for the most part, produce a "loose" chamber fit for reliability. The loose chamber doesn't position the round as precisely as a single caliber spec can so the rounds can assume differing positions in the chamber.

    5- case variations: the 223/556 cases probably vary in weight and volume more than the more popular precision caliber cases. This influences pressure which in turn influences velocity. For short ranges, i.e. under 600 yards, velocity variations are essentially insignificant, but at longer ranges velocity variations can become quite significant.

    6- large velocity variations: I believe this is an accumulation of a number of variables. One is the small case. Small volumetric variations in the small cases have more of an impact than the same variation would in a larger cases. For an AR and any bolt gun that uses AR mags, we are limited to bullet seating depths due to magazine length limits of 2.260". Again, the chamber/die issue - the dies produce a loose fit which is contrary to consistent performance.

    My testing of thousands of rounds shows that typical 223/556 rounds are going to produce spreads from about 50 fps to 80 fps, and SDs in the 20 to 25 range, with some better and some worse. In comparison, my buddy, in his 6mm Dasher, typically sees spreads of 10-15 and SDs of around 5, sometimes even tighter, rarely ever more than those numbers. What's the difference? All this stuff I've been talking about.

    7- bullets: If manufacturers are tuned into the precision/long range rage, they are likely putting most of their research, development, production efforts into the 6mm - 6.5mm calibers and letting the 223/556 kinda slide. I'm not saying we can't get good 223/556 bullets, I'm just wondering if the 223/556 bullet quality is up there with the 6mm and 6.5mm bullets.

    Anyway, that's my story for now, based on thousands of rounds of research and testing trying to figure out why I can't get lower numbers (SD and ES).

    And now returning to the specifics of @Struckat 's post,

    I'm kinda like you about the pipeline but it's 223/556 filling the line. You've encouraged me to do some more with the 6.8 SPC and I will do that!
    Struckat and Marc_Brooks like this.
    We don't have a gun problem in the US, We have a people problem.
    The problem we have is people that want to kill large numbers innocent people
    in Gun Free Zones.

  12. #11
    Lead Moderator
    Array Tangle's Avatar
    Join Date
    Dec 2004
    Location
    Chattanooga
    Posts
    12,935
    Quote Originally Posted by CavemanBob View Post
    42 is the number you're looking for.
    OK, @CavemanBob , you got me, I'll bite - why 42?
    We don't have a gun problem in the US, We have a people problem.
    The problem we have is people that want to kill large numbers innocent people
    in Gun Free Zones.

  13. #12
    VIP Member Array LimaCharlie's Avatar
    Join Date
    Mar 2013
    Location
    Northwest Oregon
    Posts
    14,832
    I blame horizontal outliers on the wind and vertical outliers on myself.
    Second Amendment: The difference between politicians and rulers.
    US Navy - US Army, Retired
    NRA Benefactor Life Member

  14. #13
    VIP Member Array Struckat's Avatar
    Join Date
    Feb 2014
    Location
    Lurking
    Posts
    3,893
    In 2020 the Savage will get a new stock and a more serious scope alignment. I look forward to working up some loads.
    Read our government's plan to destroy the Constitution.
    https://sustainabledevelopment.un.or...ormingourworld

  15. #14
    Distinguished Member Array CavemanBob's Avatar
    Join Date
    May 2015
    Location
    Kalirado
    Posts
    1,678
    Quote Originally Posted by Tangle View Post
    OK, @CavemanBob , you got me, I'll bite - why 42?
    Because...

    https://www.independent.co.uk/life-s...g-2205734.html

    The best citizenry is an informed citizenry

    NRA Patron member
    Gun Owners of America
    Electronic Frontiers Foundation https://www.eff.org/ - defending digital privacy and free speech including forums such as this one

  16. #15
    Lead Moderator
    Array Tangle's Avatar
    Join Date
    Dec 2004
    Location
    Chattanooga
    Posts
    12,935
    Quote Originally Posted by CavemanBob View Post
    Roger that!
    We don't have a gun problem in the US, We have a people problem.
    The problem we have is people that want to kill large numbers innocent people
    in Gun Free Zones.

Page 1 of 2 1 2 LastLast

Sponsored Links

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •