Group sizes.  

  • 585 Views
  • Last Post 23 July 2018
  • Topic Is Solved
stevebarrett posted this 12 July 2018

 

I’m getting confused trying to compare my results with what gets reported by way of group sizes. When someone says they got a 2-inch group, does that mean 2 inches was the between the two furthest apart shots, or something else? Looking at Ballistipedia, it seems group size is a term that could be applied to a whole variety of measurements.

Maybe there’s an unspoken convention here that I’ve missed?

Thanks

Steve

Attached Files

Order By: Standard | Newest | Votes
M3 Mitch posted this 23 July 2018

The problem with shooting much less than 5 shots is that you don't really have a statistically significant sample.  The practical problem here is if you are trying to figure out if a change you made in a load has made it shoot better or worse, if you shoot 3 shot groups, the smaller group, if it's not much smaller, could actually be a less accurate load.  You would find that out if you shot more rounds into the group.  If you are shooting a hotter load in a "pencil barrel" rifle, you may want to shoot 3 shots, let it cool, then fire 3 more.  You can overlay targets so you get the 2, 3 shot groups preserved, but the composite of 6 rounds actually tells you more.

Also worth considering if the group is round, or do you have vertical stringing, or horizontal displacement? 

Then consider bedding issues, or how a lever action carbine will sometimes "walk" shots lower on target if the front barrel band is tight on both the magazine tube and the barrel, plus how they will react if you shoot from the magazine, or if you keep the magazine full, or empty, or whatever you want to use to replicate how you would actually use the gun in the field. 

Attached Files

rfd123 posted this 23 July 2018

for working up loads for my bpcr guns i'll do five groups of two shots each @ 200 yards, then average the five group results.  it works well enuf for me.  using the furthest apart target holes for five or ten shots would be very misleading for me.  do whatever you think works best for you.  that's about it. 134

http://www.BuffaloRifles.org

Attached Files

Brodie posted this 16 July 2018

Thanks Pat. I seem eventually  to have got the answer to my original question, which is that quoted group sizes don’t actually tell you anything at all!

Steve

Truer words were never spoken.

While the vast majority of members of this organization would not knowingly mislead us about their prowess as shooters, other people are not so honest.  You have to look at it like a fish or game story with about a cup of salt.

 I'm with you on that; always get a one hole group, plus it saves on components and time.  Makes me feel good about my loading Q and A and rifleman skills.

Especially if the round went where I wanted it to.

All kidding aside use what you are comfortable with.  It's kinda like Dumbos' magic feather, if it works who cares.

 

B.E.Brickey

Attached Files

John Alexander posted this 16 July 2018

"1 five shot group gives you a high probability that the real capability of the load/rifle is between 55% and 145% of the test group size.  If you shoot 4 five shot groups, the range of possibilities  for the real capability narrows to 77% to 123% of the 4 group agg. At 10 groups the range has narrowed to 85% to 115%. 20 groups = +/- 10%"

Thanks for posting that Jim.  Useful information and I didn't have to look it up.

Most of us tend to forget or don't believe how much successive groups of the same load vary and chase our tails by making judgements based on  a single pair of five shot groups.  Two ten shot groups are of course better but can still lead you astray sometimes.

It also depends on the difference in the sizes of the pair of groups.  If one is twice as big as the other you are fairly (but not completely) safe in your judgement of which is the best load.  If the difference is 20% you are on very shaky ground.

John

Attached Files

pat i. posted this 16 July 2018

I think Ed Harris would be the man to talk to about that.

Attached Files

stevebarrett posted this 16 July 2018

If you're not already a member you should think about joining the CBA. There's a lot of good loading information in the military and BR match reports with no fudging the numbers. ShIoulder to shoulder competition tends to keep people honest.

 

I am a member but evidently haven't managed to find the right bit of the extensive website. Where best to look? (principally interested in historic military handguns)

Thanks

Attached Files

Scearcy posted this 16 July 2018

The second implied part of the question is what are my own group sizes telling me. Many of these posts have been getting at that question. There is a very good article by Wm Davis in the NRA Handloader. He did alot of work regarding statistical probability and group sizes. Here is a very brief exerp:

1 five shot group gives you a high probability that the real capability of the load/rifle is between 55% and 145% of the test group size.  If you shoot 4 five shot groups, the range of possibilities  for the real capability narrows to 77% to 123% of the 4 group agg. At 10 groups the range has narrowed to 85% to 115%. 20 groups = +/- 10%

It takes a whole bunch of shooting to compare several loads or perhaps your good days and bad days are all within the variance that is expected.

Jim

Attached Files

  • Liked by
  • M3 Mitch
pat i. posted this 16 July 2018

If you're not already a member you should think about joining the CBA. There's a lot of good loading information in the military and BR match reports with no fudging the numbers. Shoulder to shoulder competition tends to keep people honest.

Attached Files

stevebarrett posted this 16 July 2018

Thanks Pat. I seem eventually  to have got the answer to my original question, which is that quoted group sizes don’t actually tell you anything at all!

Steve

Attached Files

stevebarrett posted this 16 July 2018

Steve groups are traditionally measured center to center of the farthest holes in the paper and that's what people expect. The thing you have to realize about comparing your results against those quoted on the internet Is a lot of people follow the "at least 50% rule" which is the results they post on the internet are at least 50% smaller than what was shot on the target. There's also the "tunnel vision phenomenum" where people center in on the few holes that landed close together while ignoring the ones scattered all over the paper. Then there's the "one group wonders" who love to post that one group out of hundreds that just managed to land somewhere in the same vicinity. Just enjoy what you're doing and take a lot of what you read with a grain of salt. This is why I like Joe Brennan's posts. He usually posts multiple groups when delving into something.

Attached Files

jchiggins posted this 16 July 2018

Over extended time periods and numerous groups fired I have come to the conclusion that using the one shot group always yields the most consistent results regardless of conditions.

I'm with you on that; always get a one hole group, plus it saves on components and time.  Makes me feel good about my loading Q and A and rifleman skills.

Attached Files

GWarden posted this 15 July 2018

I use to use five shot groups and shoot several then see what the agg. was of the total groups. Lately I have switched to ten shot group, shoot several and then go for the aggregate of the groups. We all know wind conditions, etc. can affect group size. The one thing we don't like to mention is shooter ability, we all have those good days and then those bad days. I have a friend in the fine State of Tenn. that is one of those phenomenal shooters. I see him once a year and we get together for a day of shooting. When I want to see what a rifle is capable of I let him shoot a few groups with it for comparison in my ability. You all have a good day.

Bob

Attached Files

pat i. posted this 15 July 2018

Steve groups are traditionally measured center to center of the farthest holes in the paper and that's what people expect. The thing you have to realize about comparing your results against those quoted on the internet Is a lot of people follow the "at least 50% rule" which is the results they post on the internet are at least 50% smaller than what was shot on the target. There's also the "tunnel vision phenomenum" where people center in on the few holes that landed close together while ignoring the ones scattered all over the paper. Then there's the "one group wonders" who love to post that one group out of hundreds that just managed to land somewhere in the same vicinity. Just enjoy what you're doing and take a lot of what you read with a grain of salt. This is why I like Joe Brennan's posts. He usually posts multiple groups when delving into something.

Attached Files

Wineman posted this 15 July 2018

I think that the amount of recoil can also have an effect, especially over a long string. I helped a friend do some load work for a 338 Win Mag with 250 grain jacketed bullets (Berger Classic Hunter) using Berger's reloading data. Even with a Muzzle Brake and heavy duty slip on recoil pad, it got to the point after five well spaced (time in between) shots, pulling the trigger was not fun.

I notice this effect using cast in a M1903A3. With Lee's version of the Ed Harris SKS bullet, and 16 grains of AL 2400, I am more comfortable vs 25 grains of H4895 and a Lyman 314299 (no filler in either). I almost always see smaller groups (10 shots slow fire, with a sling 100 yard CMP matches) with the smaller lighter bullet. I never see the 160 grain bullet in CBA match results so I doubt it is a more accurate bullet but in the one application I use it for, it is.

Dave

Attached Files

Scearcy posted this 14 July 2018

Joe

These numbers are very very close to the ratios I calculated using National Match data.

JIm

Attached Files

joeb33050 posted this 14 July 2018

One winter day a couple of years ago I compared the 5 shot groups vs the 10 shot groups for many of the competitors at the 3 years previous National Matches. I don't recall if I included every class of rifle from every year but with 40-50 entrants per year and both 100 and 200 yard groups, it was a respectable amount of data. The ratio was determined by dividing the 10 shot agg (average of 2 groups) by the 5 shot agg (the average of 4 groups) for each participant at each distance. The ratio was just over 1.3. One year had bad weather on the day of the 200 yard match which created a little noise in the numbers but overall, the ratio was fairly steady.

Indeed Ken if ya keep shooting, small groups will get bigger.

Jim

 

 

7.15.7 HOW MUCH DOES GROUP SIZE CHANGE AS RANGE CHANGES?

 

 

 

5 shot 100 yard groups are .45 times as large as 5 shot 200 yard groups.

 

5 shot 200 yard groups are 2.2 times as large as 5 shot 100 yard groups.

 

10 shot 100 yard groups are .43 times as large as 10 shot 200 yard groups.  

 

10 shot 200 yard groups are 2.35 times as large as 10 shot 100 yard groups.  

 

On average

 

See “ACCURACY 100 YARD VS 200 YARD”, in the excel workbooks.

 

Attached Files

  • Liked by
  • RicinYakima
  • M3 Mitch
joeb33050 posted this 14 July 2018

...some old farmer told me  ... " if ya keep shooting, small groups will get bigger but big groups won't get smaller ... " 

I shoot a lot of groups, and I'll tell you that sometimes, groups get smaller. A big 2 shot group can become a small 5 shot group .Nyet?

 

 

 

 

 

 

 

 

what statistical rule applies here ??   i think it means that if you shoot 3 shots and every bullet hole is 2 or 3 inches from the others .... you can quit shooting now ...  what level of validity  have we achieved ?? ...

ken, statistically not sure of anything ... cool

 

Attached Files

Scearcy posted this 13 July 2018

One winter day a couple of years ago I compared the 5 shot groups vs the 10 shot groups for many of the competitors at the 3 years previous National Matches. I don't recall if I included every class of rifle from every year but with 40-50 entrants per year and both 100 and 200 yard groups, it was a respectable amount of data. The ratio was determined by dividing the 10 shot agg (average of 2 groups) by the 5 shot agg (the average of 4 groups) for each participant at each distance. The ratio was just over 1.3. One year had bad weather on the day of the 200 yard match which created a little noise in the numbers but overall, the ratio was fairly steady.

Indeed Ken if ya keep shooting, small groups will get bigger.

Jim

Attached Files

Ken Campbell Iowa posted this 13 July 2018

...some old farmer told me  ... " if ya keep shooting, small groups will get bigger but big groups won't get smaller ... " 

what statistical rule applies here ??   i think it means that if you shoot 3 shots and every bullet hole is 2 or 3 inches from the others .... you can quit shooting now ...  what level of validity  have we achieved ?? ...

ken, statistically not sure of anything ... cool

 

Attached Files

  • Liked by
  • M3 Mitch
John Alexander posted this 13 July 2018

"Maybe with some basic data, a statistician could advise on just how many shots to a group would give just what level of confidence for results? "

Yes statisticians can do that.  You can do it too. Joe Brennan's book explains how to do it better than most statistic books.  The book is on sale by the CBA.  See inside the back cover of any Fouling Shot or our website.  The book is also available on line.

John

 

Attached Files

Show More Posts
Close