EXPERIMENT 1-SHOTS PER GROUP

  • 279 Views
  • Last Post 12 September 2018
joeb33050 posted this 02 September 2018

We see 10 or 5 or 3 shot groups reported. The larger the number of shots per group, the smaller the number of groups-of data points. 100 shots will make 1 100-shot, 2 50-shot, 4 25-shot, 20 5-shot, and 50 2-shot groups.

Brent Danielson has long advocated the 2-shot group, giving the largest number of data points per total number of record shots fired.

We know that average group size varies with n = the number of shots per group, thus:

 If 2 shot groups size = 1; then

3 shot group size = 1.36

5 shot group size = 1.73

10 shot group size = 2.17

(The group size ratio for n >2 ~ n^1/3)

To test the use of 2-shot groups as a measure of accuracy, I will fire 6 5-shot groups and 15 2-shot groups.

The more of you who perform and report the results of this test, the more sure we will be of the results.

Test procedure/conditions:

Any gun that will reliably shoot 100-yard 5-shot groups <2" @ 100 yards. (I'll use a BSA Martini 12/15, 22 LR).

Range: 100 yards

Everything else is the shooter's choice-fouling shots, benchrest equipment, group measuring tools, targets, etc.

I will shoot my test tomorrow.

 

 

Attached Files

  • Liked by
  • Ross Smith
Order By: Standard | Newest | Votes
stevebarrett posted this 02 September 2018

I wasn't aware of that relationship:  The group size ratio for n >2 ~ n^1/3

How is it derived?

Thanks

Steve

Attached Files

joeb33050 posted this 02 September 2018

From 'A GROUP SIZE MODEL" , an excel workbook, that calculates group sizes for any average and stdev of X and Y; for groups from 2 to 10 shots. 

I'd be happy to send the workbook to anyone who wants it, and answer any questions.

 

Attached Files

Ken Campbell Iowa posted this 02 September 2018

hey joe ... i got one for you  >>

i believe you have explained the relationship between (among? ) maximum group size ... and mean radius ..... relative to validity ...   over a large number of shots., the validity ratio approaches 1 ...

but in our usual case of a few dozen shots ....   it makes common sense to me that * mean radius * is much more valid .

if true how many groups do we have to shoot before group size is as valid as mean radius ?? 

********

this in preparation for standardizing analysis methods of the results of the CBA research team .

ken

 

Attached Files

joeb33050 posted this 02 September 2018

hey joe ... i got one for you  >>

i believe you have explained the relationship between (among? ) maximum group size ... and mean radius ..... relative to validity ...   over a large number of shots., the validity ratio approaches 1 ...

but in our usual case of a few dozen shots ....   it makes common sense to me that * mean radius * is much more valid .

if true how many groups do we have to shoot before group size is as valid as mean radius ?? 

********

 this;  in preparation for standardizing analysis methods of the results of the CBA research team .

ken

 Group size X NUMBER = mean radius, where NUMBER varies with the number of shots in the group.

For 2-shot groups, GS X .49 = MR

For 5-shot groups, GS X .34 = MR.

"validity" = ?

We can make statements like this "Based on a buncha stuff, we're 95% sure that GS or MR = sample avg +/-  2.5%." 

There is a finite amount of information in each bullet hole. Some information extraction methods are much more difficult than others, but none that I can find increase that amount of information. A 5 lb bag holds 5 lb, whether you measure it in short tons of kilograms or cubic inches. 

Attached Files

  • Liked by
  • Ross Smith
GWarden posted this 03 September 2018

I agree, but are we talking about everything being perfect, all the BS conditions? It seems that each shooter may be looking for something a bit different  in information on their groups. For some a 3 shot group supplies, other 5 shot, some of us 10 shot.

My biggest variable I run into is the person behind the rifle, me. If I am "on" and everything else has been worked to find the best load my ten shot groups the I use for evaluation can really look good. You know those times when you are in the "groove", all the planets are lined up etc, etc. BUT, give me a day where I am not on, and someone might say what is the problem with your loads or rifle, way to often the problem is me. Just my two cents worth, but I don't feel we put as much emphasis on the shooter- wish it wasn't that way.I not trying to start a hoo rah, just expressing my opinion. You all have a good day at the range.

Bob 

Iowa

Attached Files

Scearcy posted this 03 September 2018

Agree. sometimes it is simply a bad day for the range as well. I shoot at a public range. Yesterday there was never an uninterrupted  stretch long enough to shoot  10 shot group. Some days I just go home and others I try to finish the group after a cease fire. It is hard to make the line safe and then return to finish a 10 shot group.

Jim

Attached Files

joeb33050 posted this 03 September 2018

Today,30 2-shot = 60 shots; and 6 5-shot = 30 shots;  100 yard groups, BSA Martini 12/15, 30X STS, GECO semi-auto ~$30/brick ammo.

2-shot groups avg .604", stdev = .353".

5- shot groups avg 1.171", stdev /240".

Shooting 2-shot groups was no problem, although the groups moved around on the paper as I moved from target to target.

Then to the arithmetic  and a big surprise.

See thread "2-shot group surprise"

Attached Files

Paul Pollard posted this 12 September 2018

Joe,

Where is the thread “2-shot group surprise?” I looked at the excel workbook and it looked confusing. Can we use the T Test to compare 2 sets of results? I used it on some 5 shot groups to find if one set of loads was better than another set of loads, since most of my data is for 5 shots. It cut out even looking at average group sizes. 

Do you just measure groups in the normal way, regardless of number of shots?

Thanks

Attached Files

joeb33050 posted this 12 September 2018

Joe,

Where is the thread “2-shot group surprise?”

I JUST PUT IT UP.

I looked at the excel workbook and it looked confusing.

WHAT WORKBOOK?

Can we use the T Test to compare 2 sets of results? I used it on some 5 shot groups to find if one set of loads was better than another set of loads, since most of my data is for 5 shots. It cut out even looking at average group sizes.

SURE. BUT I DON'T LIKE t, t TESTS, OR STUDENT.  MY PREJUDICE.

 

Do you just measure groups in the normal way, regardless of number of shots?

YES, WITH A PLASTIC RULER GRADUATED IN TEHTHS OF AN INCH.

Thanks

Attached Files

Categories
Search
Popular Tags
This Weeks High Earners
Hot Topics
Close