Do your choose load B over load A?

[BCoates]

TRex - I've been involve in statistics enough to know you can prove or disprove anything with enough numbers. I've ran studies where we could show a statistical difference between a 1.1 avg and a 1.2 avg. With that said, most people load modifications with at least 20 shots each. I know a few who shoot F class that will shoot one load per 20 shot string to fine tune loads. They mainly measure the vertical since horizontal can be due to atmospheric conditions. I like to test at least 4 groups of 5 rds or 2 groups of 10 rds when testing an A vs B load or one reloading step modification vs my standard reloading sequence. I always alternate between the two loads to negate any outside influences from atmospheric conditions or barrel fouling.

My 2 cents,
BC

That seems reasonable (four 5 shot groups or two 10 shot groups) but does not answer my question of what it takes to get a ninety five percent confidence level. If we knew we might find that your test is adequate or we might decide that it is too difficult to get there or that it was not worth it to us. I see folks making decisions on a lot less data and have to wonder if their test is good enough. At least I would like to know what the science tells us is required.

TRex - If you really want to test load A vs load B in a statistical fashion you will need to 1.) Find enough cases with the exact same case volume and neck thickness, 2.) Make sure the cases have been fired the same number of times, have the same headspace, and have the same level of hardness, 3.)Find enough bullets with the same weight and bearing surface length, 4.) Find enough primers that weight exactly the same, 5.) Make sure the bullets are seated to the exact same base to ogive length, 6.) Make sure powder charges are weighed within 1 kernel, 7.) Make sure loaded round runout is within 0.001" for all test rounds.

Then you would shoot each round at an individual bullseye and measure the offset from the bullseye in two axes after getting an exact zero for each load. Then analyze the data using a T test or ANOVA test. I ran some numbers, and it will take 384 rounds total to achieve a 95% confidence level with a 5% confidence interval. Of course you would need to fire the rounds in a round robin style to negate atmospheric condition changes, barrel heat and fouling changes, and shooter fatigue issues.

The point of scientific experiments is to limit as many variables as possible and then test the variable(s) that you want to change. If you have unlimited resources and time, have at it. If you can find enough components that will meet the above criteria, then you would be better served doing some smaller load tests and saving the rest for competition loads. Otherwise choose a method recommended by whoever you trust the most.

Some methods will work for you and some won't. I don't use the same methods as the benchrest guys because I don't control as many variable as they do. I'm just saying what works for me. I'm a scientist, but I'm also willing to accept a certain level of risk to balance my time and resources ($$). My load for my F-TR rifle can and has shot clean scores at midrange, so I must be doing something right. Shooting is not all science, and sometimes things happen that are not normal (good and bad). That is why most shooting sports have national and/or international records, and why we still shoot competitions. Even the rail gun benchrest guys still go to competitions because you don't know which competitor is going to have the best day. You wouldn't go to a competition if you already knew what the results were going to be, right? I know I wouldn't shoot a group if I already know what is going to be the result. That would be an exercise of futility and insanity.

 

[Toby Bradshaw]

I'm pretty well done with this thread, but the notion that one has to use the same piece of brass, same barrel heat, etc. is incorrect. As long as the different loads are fired in random order, the variables that you seem to be worrying about are part of the environmental variance in any statistical analysis. They add noise to the data, and make the signal more difficult to detect, but don't preclude a standard statistical analysis.

When a drug trial is done, do you imagine that the same person must get the drug and the placebo for the trial to be valid?

 

[chkunz]

I'm pretty well done with this thread, but the notion that one has to use the same piece of brass, same barrel heat, etc. is incorrect. As long as the different loads are fired in random order, the variables that you seem to be worrying about are part of the environmental variance in any statistical analysis. They add noise to the data, and make the signal more difficult to detect, but don't preclude a standard statistical analysis.

When a drug trial is done, do you imagine that the same person must get the drug and the placebo for the trial to be valid?

Thanks for you response to my question. You and a couple of other folks actually tried to answer my question. It is a simple question that has a definitive answer and the answer can be instructive especially to folks that have thought about this subject but have never been exposed to the science behind the answer. The wisdom is in how you apply the answer. It is just one tool that is available in the quest for accuracy. Others seemed more interested in attacking me for asking the question and made wild accusations about my ability to understand how complicated the pursuit of accuracy is. Many of the responses were mostly about the individuals approach to load development and match strategy and some of it is excellent information and I am sure produces excellent results and wins in competition. Again thanks for you response and take care, Clyde.

 

[BoydAllen]

Yes, it seems to have been an waste of time, pointing out the reasons that coming up with a magic number to answer your question would have ignored many important factors. I have run into a fair number of people who seem to think that statisticians produce magic numbers, something that statisticians are more than willing to go along with, because they don't know what they don't know, and often confuse complexity with reality. After projections have been made not many take the time to dig them up, and check them against what actually happened.

 

[chkunz]

Yes, it seems to have been an waste of time, pointing out the reasons that coming up with a magic number to answer your question would have ignored many important factors. I have run into a fair number of people who seem to think that statisticians produce magic numbers, something that statisticians are more than willing to go along with, because they don't know what they don't know, and often confuse complexity with reality. After projections have been made not many take the time to dig them up, and check them against what actually happened.

BoydAllen Whose time are wasting?

 

[BoydAllen]

Your previous post seemed to indicate to me that you have a preconceived idea about the nature of a proper answer that I did not pick up from your OP. I and others have tried to explain why there is no magic number that grants 95% confidence, and it seems that you have rejected all of that. Certainly I do not resent the time that I have spent, but it seems to have been a wasted effort on all of our parts, because it did not achieve our goal, except that not everyone who reads a thread posts on it, and for that reason, perhaps saying that it was a waste of time, is not correct.

So many times I talk and correspond with shooters who grab onto some detail that they have read, and it becomes more central to their efforts to advance their shooting program than it really deserves, if it does at all. By presenting my reasons to take exception to the concept that you seem to me to be advocating, perhaps I may have influenced someone to think about the subject in greater detail, and to consider it with a broader perspective.

 

[chkunz]

Your previous post seemed to indicate to me that you have a preconceived idea about the nature of a proper answer that I did not pick up from your OP. I and others have tried to explain why there is no magic number that grants 95% confidence, and it seems that you have rejected all of that. Certainly I do not resent the time that I have spent, but it seems to have been a wasted effort on all of our parts, because it did not achieve our goal, except that not everyone who reads a thread posts on it, and for that reason, perhaps saying that it was a waste of time, is not correct.

So many times I talk and correspond with shooters who grab onto some detail that they have read, and it becomes more central to their efforts to advance their shooting program than it really deserves, if it does at all. By presenting my reasons to take exception to the concept that you seem to me to be advocating, perhaps I may have influenced someone to think about the subject in greater detail, and to consider it with a broader perspective.

Excellent response, thanks so much. I was concerned that I may have been seen as wasting the time of others. My concern is that I think I am seeing where folks are sometimes making decisions based on insufficient data or they are interpreting data incorrectly. In some cases it is no better than flipping a coin. But we do not normally have the time or resources to obtain as much data as we need. I fully understand the limitations of statistics but it can be a useful tool is applied correctly and with wisdom. Thanks again and take care.