Tony Boyer's message

  • 861 Views
  • Last Post 17 August 2022
joeb33050 posted this 13 August 2022

In TFS  No. 274, November-December 2021, John Alexander wrote “What Tony Can and Can’t Tell Us”.

John contends that Tony’s sample size; the number of groups shot to make a decision: one three-shot group in Boyer’s “tuning system”, and one or two five-shot groups in “Load Adjustment During a Match …”, is insufficient. Statistics is alluded to.

 

A sample size of one is only sufficient to make a correct decision if the variation in group size is zero.

Group size variation is caused by either the shooter or the equipment.

If the shooter causes no variation; then all variation is caused by the equipment.

If the shooter causes no variation and the equipment causes very little variation; then the variation in group size approaches zero and “correct” sample size approaches 1.

 

If Tony causes little or no variation in group size, and equipment causes little or no variation with “good” loads; then variation in group size is mostly caused by the shooter; and fiddling about with “barrel break in”, weighing powder charges, weighing bullets, using match grade primers, etc. will not reduce variation in group size.

Variation in group size, and group size itself, with “good” loads, is determined primarily by shooter ability. Maybe that’s what Tony is telling us.  

 

 

 

Attached Files

Order By: Standard | Newest | Votes
John Alexander posted this 15 August 2022

I completely agree with Mr. Shaw that nearly all jacketed bullet (JB) benchrest shooters seem to use a tuning method similar to the one described in Mr. Boyer's book. I have been reading of these shooters giving each other the same advice on using such a tuning method for maybe thirty years in the now discontinued Precision Shooting magazine and elsewhere. I have read all the books mentioned by Mr. Shaw which have some form of similar tuning described (although none describe it as clearly as Boyer.)  All of them think they are somehow coming closer to the perfect load for that particular time and place by making minute adjustments to either seating depth, powder charge, or both. All this tweaking is based on the delusion that you can look at a single 3-shot or 5-shot group and from that, and that alone, judge how that load will shoot for the next group or two.  The tweaking doesn't do any harm because it is only done for a very narrow range of seating depths and powder charges. As far as I know there has never been any published test to show how that there is any validity to minor adjustments based on one small group.

I have presented a series of illustrations in my articles why this is impossible. I have suggested that Mr. Shaw shoot a series of three shot groups with the same load to show, that for the small groups he shoots, they are all be alike in both size and shape so the first group could predict what the load would do.  He has ignored both my illustrations and examples and he has yet to produce the string of identical three shot groups. I'm not sure he has read my arguments much less thought about them, Instead, his response has been repeatedly some variation of everybody in JB benchrest does it so it must work. If that depth of reasoning is good enough for the readers here, more discussion probably won't help.

I have done my best to show why you can't depend on a three or a five shot group for making judgements, if readers aren't convinced by now they should perhaps adopt the tuning method Mr. Shaw uses, and see if they can convince themselves that it works.

John 

 

Attached Files

Ken Campbell Iowa posted this 13 August 2022

good stuff, Joeb ...  good because your words agree with me.  ( g ) .

hard to develop the best load when shooting in a extremely gusty day and the guy next bench has a 300 Win. Mag with a muzzle brake.   although you could develop the best * system * for that circumstance.

ken

 

 

Attached Files

  • Liked by
  • Bud Hyett
John F. posted this 14 August 2022

I haven't read the TFS article, but I did dabble with statistics back in the day.  Tony Boyer and his better rifles

constitute a very sensitive instrument for testing ammunition accuracy.  Relatively valid conclusions can be

drawn with smaller sample sizes due to the lack of variance in the population of shots fired, as opposed to, say,

a factory 10/22 shooting various types of Walmart bulk ammo at 100 yards with a 4X scope and 7 lb trigger. 

 

Still, I prefer larger sample sizes, even with my most accurate jacketed-bullet match rifles, but then, I'm not a benchrester

and for sure don't have Tony Boyer's experience or equipment.  FWIW, if using the hand-made, incredibly-high quality bullets

Tony would use in a match, weighing bullets would be a completely irrelevant exercise.  Accuracy variation would come down

to jacket concentricity measured in the 0.0001" range and other aspects of  how it was drawn, core seated, etc.

John

 

Attached Files

  • Liked by
  • Bud Hyett
joeb33050 posted this 14 August 2022

Tippet: n = 5, RANGE = 2.326 standard deviations, stdev, RANGE / 2.326 = stdev.

 

The RANGE = largest – smallest; 1” - .5” = .5” = RANGE

In samples of five, the RANGE / 2.326 = stdev.

If the ratio of sets of five, largest/smallest, 5-shot groups is 2; then

 

If largest = 1” and smallest = .5”, RANGE= .5” / 2.326 = .224”

If largest = .1” and smallest = .05”, RANGE= .05” / 2.326 = .0224”,

 

Variation = standard deviation varies with group size.

 

(The Coefficient of Variation = CV = standard deviation / mean.

If the ratio of sets of five, largest/smallest, 5-shot groups is 2; then CV = .224 over all group sizes. This is new to me, may be interesting/important.)

 joe b.

 

 

Attached Files

  • Liked by
  • Wm Cook
John Alexander posted this 14 August 2022

John F. said: "Relatively valid conclusions can be

drawn with smaller sample sizes due to the lack of variance in the population of shots fired,"

===================

This is true.But the implicit assumption here seems to be that because Boyer shoots aggs. under .2 moa his group to group coefficient of varation must be small. The fact is, that although it SEEMS like that should be true it isn't.  This isn't debatable, if you look at any !BS match report you will see that the average of the ratios  of largest to smallest group (an approximation for coefficient of variation) in a five group aggregate for the top 5 shooters will be near 2.  The average ratio for the bottom 5 shooters will also be near 2.

Worse, if you shoot several strings of five 5-shot groups with an ordinary 22 rim fire auto, the ratios of largest to smallest groups will average near 2. This is true although 22RF groups will be about ten times Boyer's groups.

Whether we like it or not, the group to group  coefficient of variation ( as approximated by the ratio of largest to smallest groups) is independent of group size, and the fewer shots per group the bigger the ratio. The large to small group ratio for five 3- shot groups is larger than for the 2 for five 5-shot groups.  Single three shot groups is what Boyer depends on when making tuning decisions.

John A

PS:  If you would like to read the TFS article, PM me your email address and I will send you a copy.  Or email to jaalex2520@ gmail.com  

Attached Files

4and1 posted this 17 August 2022

 

I am going to break my silence about this subject. joeb33050 has brought up the writings of John Alexander published in two The Fouling Shot issues. joeb makes reasonable comments. But as the normal, Mr. Alexander has upheld his opinion that Mr. Boyer's method doesn't work.

 

After Mr. Alexander's published articles in TFS 274 and 275, I wrote a response and sent it to the Fouling Shot editor, but I have never seen it published. So I am going attempt to copy and paste it here.

 

 

In Response To John Alexander’s Articles in TFS 274 and TFS 275

 

 

 

By: Wayne Shaw

 

 

 

All of this discussion started in a thread on the CBA Forum, found here: https://forum.castbulletassoc.org/thread/loads-shown-in-bold-designate-potentially-most-accurate-load/

 

Started July 2 2021.

 

 

 

Someone asked why “accuracy” loads were max loads, obviously looking at a older Lyman loading manual. In the 7th post, I (under my user name of (4and1) suggested he do his own short load testing using a tried and true method of trying powder charge and seating depth changes. This would give him an idea of what his gun will shoot the best.

 

 

 

The very next post was from John Alexander, which started a marathon of sorts of post which ended up with 56 comments (I think that number is current), and changed the direction of the thread entirely.

 

 

 

I will not type out a blow-by-blow of this whole thing, suffice it to say, Mr. Alexander took exception to what I posted regarding the testing method.

 

 

 

Over the course of that forum thread, then his article in The Fouling Shot 274, and again in The Fouling Shot 275, he named the source of this “disputed” test method, as Tony Boyer’s book. This may be HIS source, but Boyer’s book is not the only resource to read.

 

 

 

This testing/load method has been used for decades by benchrest shooters. Glenn Newick published a book in 1989. Mike Ratigan published a book in 2007, Boyer’s book 2010. There have been numerous posts on Accurateshooter.com which describe this along with pictures (off the top of my head by James Mock). You can go to Youtube.com, and search for videos by Jack Neary. Jack is a Hall of Fame shooter, and he conducts a seminar of sorts at many of the large benchrest matches, for those who want to better their groups. I personally know quite a few HOF shooters, and other very successful shooters, and all of them use the same or very similar method of testing and tuning their rifles. I have traveled to many matches, I try to get there a day ahead of the match, and there is always about half the total amount of competitors there too. They spend the day “before” working on their loads. If anyone takes the time, they can use a spotting scope and see what everyone is doing. All are testing with 3 shot groups. One has to wonder why so many benchrest shooters use it.

 

 

 

With all that said, Mr. Alexander believes it impossible to determine an accurate load shooting 3 shot groups. He has gone into great detail on what his thoughts are on finding what an accurate load is. It’s very extensive and he has put in a great deal of work detailing it. He challenged me and any others to shoot a test target of 10 - 3 shot groups. He did it using a 6PPC rifle of his own, and posted his target. His gun was not in tune, his groups showed it. Yes some were good, but an out of tune gun will shoot a good group once in a while. He made no attempt to adjust his tune, that he revealed.

 

 

 

But here is the difference. The methods that Mr. Alexander describes would be very fitting if he were doing research and development for a powder company, or bullet company, that wanted to publish reloading data. Those people do not publish load data without thorough testing and record keeping, plus these loads are to be safe in any gun produced for that cartridge.

 

 

The method described in Boyer’s book and numerous other sources, is intended for tuning a gun, THAT gun, for a given powder (or two) and bullet, for that upcoming match. The most expensive and painstakingly assembled benchrest rifle can shoot horrible groups, if it is not tuned. You must go through some process to find the proper load, and this 3 shot group method does that. However, this “TUNE” will not likely work all day. You start a match in the morning, the air has more humidity and it’s cooler, and as the day goes on, things change. If you don’t change the load as you go, you will end up in the “also-ran” category. 99.9% of benchrest shooters change their load for each target they shoot. Understand, most readers here are not attempting to shoot against world class benchrest shooters. This is what it takes to maintain cutting edge accuracy. Additionally, people travel to matches. You can tune your gun at home, load up a bunch of ammo, and travel only to find out your gun is no longer in tune. And you don’t have to go across country to see the change. Most take two powders or more to work with at each range, given the conditions, one may appear to work better than the other. I can guarantee, people are not going through the whole process as Mr. Alexander talks about at every weekend match.

 

 

 

This method can also be used to simply find out what your favorite Remchestersalvage shoots best. Somewhere between the starting load and max load, and a jammed bullet or 20 thousands off the lands. And it doesn’t take up an enormous amount of precious components. One has to use some method to find the tune. I encourage people to test different powders. Burn rate charts and reloading manuals are good for a “BALLPARK” to try, but powders have characteristics, with the bore size, weight of bullet, case capacity, etc. what a manual may show that should work, doesn’t always.

 

 

 

Some further comments of this exchange. “...a single group is an unreliable predictor of the true accuracy”. Many shooters will take two loads to the line, shoot their group, then shoot the other in the sighter target to see if there is improvement. It is common practice, and if a shooter does not have the knowledge of how to change his load over the course of the day, they rely on a sighter group to guide them to the next target. The aggregates being shot today are smaller than ever, people are learning more and more about tune. To a benchrest shooter, a tune only needs to last the 20-30 seconds it takes him to shoot his 5 shots. Then change.

 

 

 

I wrote “when shooting a jacketed bullet you can hang your hat on believing where it went”. That sure got attention! My meaning was simple, fliers are an ACCEPTED (albeit cursed) part of cast bullet benchrest. If a 3 shot group has one out and the other two touching, you need to reshoot to see if it repeats. If those blasted fliers could be eliminated, then you could accept it. I don’t accept fliers, I search to eliminate them, but you can’t call a shot back. Those that have figured it out have my respect. Hand swaged jacketed bullets don’t throw wild shots like that, so if a bullet goes out, I look at my flags to see what I missed.

 

 

 

This whole discussion is comparing apples to oranges. If you want to win a match, you need to work your load for that time and date. If you want to use a published load data that could be a couple years old by the time you read it, or an extensive load you developed two years ago, your results may not be as good as it could be. It all depends on how bad you want to shoot well.

 

 

 

I encourage everyone who shoots competitively, or one who just wants to get the most out of their rifle, to do their own research, go to the shooting forums, go to youtube, search tuning a rifle, and I’ll bet you will come away with a different idea than running the method Mr. Alexander suggests.

 

*********************************************************************************

 

It appears I got it pasted in. My apologies to joeb33050 for highjacking your thread.

 

Mr. Alexander's response in this thread to joe33050's first post is saying a winning aggregate of 5-5shot groups has a ratio of 2, and the last place aggregate has a ratio of 2. The winning agg would likely be around .2000", and the last place agg could be .5000" or bigger. Might have the same ratio, but ratios don't win matches.

 

 

 

 

 

Attached Files

Close