CB INACCURACY

  • 6.3K Views
  • Last Post 19 June 2014
joeb33050 posted this 25 May 2014

I was thinking, and don't know about CB inaccuracy.Do CBs just make bigger groups?orIs there 1 shot or group that makes the group average bigger?Is there a stranger shot making inaccuracy? I don't know, but... There are 33 sets of 5, five-shot groups in my records for 1 gun/barrel.Does 1 group contribute a lot to large averages? Is 1 of the 5 groups unusually large? Of the 33 sets, 14 sets had one unusually large size group, meaning that the chance of a group that large is 2.5% or less. This suggests that CB group averages are caused by 1 flyer, 1 outlier; and that CB groups may NOT just be large.  Nothing's proven, more in process.

Attached Files

Order By: Standard | Newest | Votes
RicinYakima posted this 25 May 2014

Joe, There was a reason I asked for the best 10-shot group on the other thread. That is the reason I view five 5-shot groups as being less revealing. Ric

Attached Files

joeb33050 posted this 26 May 2014

<user=50>RicinYakima wrote: Joe, There was a reason I asked for the best 10-shot group on the other thread. That is the reason I view five 5-shot groups as being less revealing. Ric Ric, what if we made a box like this, that holds a number of targets. Then we shot at the front target. Every 2 shots we remove and replace the 2nd with a new target.Every 3 shots we remove and replace the 3rd with a new target.Every 4....And so on.Each and every shot goes through each target.After a zillion shots we've got stacks of 2 shot, 3 shot, 4 shot, 5 shot,...10 shot targets.We measure each target and average group sizes.3 shot avg > 2 shot etc, the average group size increases as the number of shots. 10 shot groups are larger than 5 shot. It doesn't matter(much) how many shots are on a target. If we know the averages from our stacks of targets, we know that 7 shot groups are 1.43 times the size of 3 shot targets. NO NUMBER OF SHOTS PER GROUP IS “BETTER” THAN ANOTHER!

Attached Files

joeb33050 posted this 26 May 2014

GROUP SIZE AND THE NUMBER OF SHOTS PER GROUP   Groups can contain any number of shots, from 2 up. The more shots in the group, the larger the average group size. Ten shot group averages are larger than three shot group averages, all shot by the same shooter with the same equipment under the same conditions.             This table shows the ratios of group size with different numbers of shots in the groups.

Ten shot group size is 1.26 times five shot group size; three shot group size .67 times eight shot group size. 

(The table won't print here, see the attached.)

Attached Files

joeb33050 posted this 26 May 2014

Back to the beginning. Added 22 more sets of 5 five-shot groups.Now 27/55 sets seem to contain a largest group larger than 97.5% of those expected.This continues to suggest that CB group size average is affected by outliers/flyers. No proof, and proof ain't coming. But t kinda looks like maybe CB inaccuracy is a product of flyers rather than general inaccuracy.

Attached Files

gpidaho posted this 26 May 2014

Good morning all. I agree with Joeb in that a generally accurate load component wise shows as less so on paper by overlooking some small thing. Gas check not seated square,unbalanced bullet,something small that I overlooked. If eight out of ten are in close proximity I blame me not the load GP

Attached Files

billwnr posted this 26 May 2014

A mathematician I know suggested 26 shot groups for revealing a rifle's accuracy.

I'm not a big fan of 5 shot groups for accuracy/score as the groups seem to have the POI “move around".

Attached Files

fa38 posted this 26 May 2014

http://www.castbulletassoc.org/view_user.php?id=182>billwnr I'm not a big fan of 5 shot groups for accuracy/score as the groups seem to have the POI “move around".

Could part of that be that you move the rifle and rests over to the next target to shoot another group and the rifle and rests are not in the same relationship to the impact area as the previous group?

Attached Files

John Alexander posted this 26 May 2014

Some folks “like” ten shot groups, others “like” five shot groups, a heck of a lot of shooters “like” three shot groups (especially when a writer wants to quote a better accuracy number.) A few people who understand statistics argue that 2 shot groups are best.  It's a hobby so we can “like” what we want to.

But we shouldn't confuse the number of shots we like in a group with the best way to find the truth about the real accuracy which a rifle/ammunition/shooter (R/A/S) can produce. Joe is right, if the results are handled intelligently and honestly, and enough total shots are fired, any size group can be used to find an accurate assessment of the (R/A/S) performance (with the exception of a rifle that doesn't hold its zero as it heats up.)  But enough total shots must be fired and recorded without discarding the ones we don't like.  

To increase the accuracy of the assessment of (R/A/S) performance, shoot more shots. There is no substitute for more shots.  20 shots (no matter the number of shots per group from 2 to 20) will get you closer to the truth than 10 shots even if those ten shots are all in one group.

Say for example you want to report what size five shot groups the (R/A/S) can be expected to shoot and plan to shoot 30 shots to find out.  It makes no difference whether you shoot 2 shot, 3 shot, 5 shot, 10 shot, or 15 shot groups with your 30 shots IF the final averages are adjusted properly to the equivalent 5 shot group size using the adjustment factors Joe presents. You will get about the same answer no matter which group size you choose.

John

Attached Files

billwnr posted this 26 May 2014

It's because it takes a large group to determine the true POI or center of the group.   5 or 10 shot groups would have sampling error established if one believed that was the true POI. The Heavy class rifles along with proper shooting techniques have a remote chance of the POI moving around when repositioning to other targets.  Bullets with voids, wind and mirage affect bullet impact more.

Attached Files

LWesthoff posted this 26 May 2014

This probably won't add anything to this discussion, but somewhere, quite a few years ago I read that one of the manufacturers of bullets/powder/reloading components in general had come to the conclusion that it took a SEVEN shot group to determine which load was liable to be most accurate. Sorry I can't quote the source; allowances have to be made for a “senior citizen's” memory, but that seven shot rule has stayed with me. I'm sure it was some statistics expert who came up with that number, but somebody paid him to do the research.

Anybody else remember hearing/reading that?

Wes

Attached Files

Chargar posted this 26 May 2014

On another board, there is a fellow that claims almost supernatural accuracy with his rifles and chides others they just don't know how.

A while back, he slipped and talked about his “core” group. Which means he measures the closest ones and throws out all the rest as errors of some sort or another.

I hold the opinion, that you don't throw out wide shots. If there is an error you figure out what it was, cure it and try it again. The best way to short circuit the learning process is to go for “core groups” and consider that the accuracy of your rifle, load and yourself. We learned from out mistakes and if we throw them out, we learn next to nothing.

Attached Files

RicinYakima posted this 26 May 2014

Wes, yes I read the same article and seven gives “reasonable” results.

JOE, I like the box! Shoot 5 shots at the first target then take it off. Shoot 5 at the second target, take it off. Shoot 5 at the third target, take it off. Shoot 5 at the forth target, take it off. Shoot 5 at the fifth target and measure it.

The reason I am a chemist is that “either it is or it is not". There is no manipulation of the data.

Ric

Attached Files

joeb33050 posted this 26 May 2014

RicinYakima wrote: Wes, yes I read the same article and seven gives “reasonable” results.

JOE, I like the box! Shoot 5 shots at the first target then take it off. Shoot 5 at the second target, take it off. Shoot 5 at the third target, take it off. Shoot 5 at the forth target, take it off. Shoot 5 at the fifth target and measure it.

The reason I am a chemist is that “either it is or it is not". There is no manipulation of the data.

RicI don't understand what you wrote.Removing and replacing targets in the box after, say, 50 shots would give us 25 2 shot, 16 3 shot, 12 4 shot, 10 5 shot, 8 6 shot, 7 7 shot, 6 8 shot, 5 9 shot and 5 10 shot targets-all shot with 50 shots. What is this about only measuring the fifth target? 

Attached Files

joeb33050 posted this 26 May 2014

fa38 wrote: http://www.castbulletassoc.org/view_user.php?id=182>billwnr I'm not a big fan of 5 shot groups for accuracy/score as the groups seem to have the POI “move around".

Could part of that be that you move the rifle and rests over to the next target to shoot another group and the rifle and rests are not in the same relationship to the impact area as the previous group? Yes. The POI moves as the point of aim changes. Better rests/adjustment  and more skill reduces or ?eliminates this effect. I have worked on equipment and technique, and almost got rid of it.However, a moving POI has nothing to do with group size. 

Attached Files

joeb33050 posted this 26 May 2014

billwnr wrote: A mathematician I know suggested 26 shot groups for revealing a rifle's accuracy.

I'm not a big fan of 5 shot groups for accuracy/score as the groups seem to have the POI “move around".Would you get and post his argument/reason/thoughts here?Thanks

Attached Files

joeb33050 posted this 26 May 2014

billwnr wrote: It's because it takes a large group to determine the true POI or center of the group.   5 or 10 shot groups would have sampling error established if one believed that was the true POI. The Heavy class rifles along with proper shooting techniques have a remote chance of the POI moving around when repositioning to other targets.  Bullets with voids, wind and mirage affect bullet impact more.Note that the box eliminates this effect.

Attached Files

John Alexander posted this 26 May 2014

We are lucky that Joe has provided us in his book with easy to use tables to determine how many shots it takes to be reasonable sure that we are looking at a real difference when trying a new load or a different loading procedure.  The answer isn't some one specific magical number of shots in a group or even the a magical total number of shots needed with each load. It also depends on the DIFFERENCE in the size of the the two groups with and without the change you are testing to see if it really improves accuracy.

If load B shoots a group only 5% better than load A,  a seven, ten or even a twenty shot group probably doesn't tell you with much assurance that B is really better than A. Instead that small difference may be just the natural variation of group size and the two loads may really be equally accurate.

Worse yet, maybe Load B is less accurate than A but shot a bit better for that one string of shots. That's one of the reasons that the most accurate rifle/shooter on the line doesn't always win the match. Call it luck or whatever but we can't claim it isn't real.

On the other hand if load D produces a group three times as big as load C's group (and we have all seen that happen) then we can be reasonably sure that D is probably a worse load than C.  You may not want to waste more than a couple of 3 shot groups of each load before deciding you have enough evidence to make a reasonable decision.   There is no one specific answer to the question of how many total shots with each load to be reasonable sure you are looking at an improvement -- or not. We all know this instinctively and Joe's tables allow us to estimate the numbers of shots needed to be reasonable sure there is a real difference between the two loads without digging into any complicated statistics.

John

Attached Files

billwnr posted this 26 May 2014

John, there's a real simple way to determine if one load is better than another. Shoot a score match and check the scores and x counts at the end.

Attached Files

RicinYakima posted this 27 May 2014

Joe, Why care about twenty five 2-shot groups? So you have lots of datum points, but that is not why we shoot groups.   The function of shooting any groups at all is to learn if Load A will have the next fired bullet strike closer to point of aim (!) than Load B's next bullet. It does not matter if it is shot #2 or shot #1,789,432! How many shots have to be fired before we can say Load A is better than Load B, 99.99% of the time?   Group matches test two things: can my load shoot closer than your load to point of aim and can I hold and fire the rifle to demonstrate that my load will hit closer to point of aim than your load.   Good discussion for me, making me think till my head hurts.   Ric

Attached Files

John Alexander posted this 27 May 2014

billwnr wrote: John, there's a real simple way to determine if one load is better than another. Shoot a score match and check the scores and x counts at the end. Bill, If I understand you correctly you are saying shoot a score match with one load and at the next match shoot another score match with another load  under perhaps completely different conditions and after having been out partying a little too much the night before one of the matches or after having a fight with your girlfriend the morning of one of the matches.  I think I can see some problems with this approach.

How about this instead -- load 20 rounds of your standard load A and also 20 rounds of variation B (charge weight or type of powder, or seating depth etc but only one factor different.)  You could also load up 20 rounds of variation C and D perhaps each with one factor changed.

Then shoot a group with each load in rotation --  A then B then C then D.  Then do the rotation again but backwards D, C, B, A.  If you choose to shoot two ten shot groups with each load you are done and you have minimized the differences in conditions of wind, mirage, fatigue etc. and given each load as fair chance as possible against the other groups.  If you have chosen to shoot each load in four five shot groups you have two more rotations to shoot.

Then go home measure the groups and find the average group size for each load.  Find the percent improvement, if any, over load A for loads B, C, and D.  Then get out Joe's tables and see if the percent change between the pairs of competing loads is big enough so that 20 shots for each load of the pair is enough shots to be 95% confident that the difference is likely to be a real improvement.  If one of the challenger loads has an improvement big enough to provide 95% confidence that it is better with the 20 shots for each load you have made real progress.

If B, C, or D seems to be a small improvement but 20 shots isn't enough to be sure, you can shoot additional groups with the two loads during your next trip to the range.  These results, along with the first results, will either increase your confidence that the improvement is likely to be real -- or you may find our the first 20 shots were leading you astray and there is probably no difference in the accuracy of the two loads. 

At any rate you have checked out three possible improvements to your load A and done it in one or two short days at the range. John

Attached Files

Show More Posts
Categories
Search
Popular Tags
This Weeks High Earners
Hot Topics
Close