Thank You.

And since we’re all expressing opinions here , I guess those who are suggesting shooting max pressure charges over a long testing regime have never seen the results of a HP rifle blow up. I don’t hear any of you folks volunteering to pull the trigger for that . Seriously ?

Would you mind telling us how many rounds were fired for each group? The SD of MV values you obtained are very interesting to me. 7 fps is very low and indicative of great control, especially given that A) they’re .30 cal rifle rounds, and B) you managed to do it twice.

Specifically,
“This test shows that MV can be effected by twist rate, and how much.”
– Your data show a trend in twist rate only; in other words the graph is the best piece of information. There is no statistical significance with which to draw any conclusion. While necessary to analyze the data, reproducing the MV/SD table only confuses the issue to those not familiar with statistical analysis. In other words, your test shows the need for further tests in order to draw any sort of conclusion.
– How much is entirely unanswerable. R^2 should be 2 or 3 standard deviations if you want to answer that question. You simply don’t have statistically significant results with which to draw a conclusion.

“You’re somewhat right. The scatter in the data and the R squared value indicate that only about 1/2 the variation in MV is due to twist rate (Correlation Coefficient is 0.55) which means that random noise has as much effect as twist rate.”
– The scatter in the data and the R^2 value indicate a Design of Experiments or similar test need to be conducted to find the significant contributors to the poor fit of the data to the regression line. Once identified, these factors, hopefully can be controlled for. Costs will increase with increased precision in your measurements whether due to more expensive equipment, a larger number of iterations, etc.
– The scatter in the data at this point only indicates that there is too much variability to draw any conclusion. You’re suggesting that the hypothesis you are testing concerning velocity vs. twist rate is showing a correlation in your results. I and others, above, are telling you that the fit of the data here is insufficient to draw that conclusion although it remains a probability.

“The point in presenting these results is to show that the effect of twist rate on MV is VERY minor, and can almost be said to be statistically “in the noise”.”
– There is too much variability in the data with which to draw any conclusion except the need for further data collection.
– The point of presenting the data is to further everyone’s understanding, correct? I commend you. Thank you. Your interpretation of the results in incorrect.

“The point of deliberate scientific testing is to isolate variables of interest to determine true causality.”
– Agreed. You’ve a responsibility to identify, isolate, and control for those variables prior to drawing any conclusions. R^2 of .5XX isn’t demonstrating a significant data fit with which to draw any conslusions.
– Ideally, the true scientific method would include not only your future statistically significant data set, but also those from two other independent sources corroborating your results.

I take issue with how you’ve presented your findings and mislead others. Why would I purchase this book after you’ve either misrepresented or misinterpreted your results?

It’s always important to hear what everybody’s wondering about & then hear from some guys in-the-know regarding whatever the topic of discussion might be, so thanks for taking the time to respond.

I think you’d be stunned at what some guys have done in terms of decisions (myself included) when building rifles based on some of the long-held beliefs that get perpetuated somehow. Usually it’s by folks that have not taken the time or effort to flush out these ideas to see if they are actually true or just old wive’s tales. Sometimes it’s a case of a faulty understanding of the real science that’s actually at play that leads to an incorrect interpretation of what someone “saw” a few times during a flawed ‘experiment’, a day at the range, or during a hunt.

Bryan, sure wish I’d had your first book fifteen years ago – it has been pivotal in helping to finally explain some events witnessed during a few long-range groundhog hunts that had otherwise defied conventional wisdom and left us mystified for many, many years. It has finally made sense of a lot of supposed ‘explanations’ of things that just never seemed to add up. It also effectively silenced some very heated arguments about b.c. and it’s effects. Not to mention that it would have saved me a bunch of money by getting barrel selection right the first time! Written and explained in a way that I can actually understand it (albeit after reading it a few times so it could sink-in to my thick head), the book has quickly become a prized and frequently referenced tool.
(guess this officially makes me a “fan-boy” now – oh well, so be it). Gents, you really need to read these books. I can’t emphasize that enough.

So, I’ll stop gushing about Bryan’s books and say I think we can pretty much chalk up another historical ‘hot topic’ of debate that’s been fairly well laid to rest finally by some smart guys with ACTUAL first-hand knowledge and experience on the subject. Though, I have to admit, I’d still really love to see some actual piezo pressure trace data for my own curiosity.
Thanks again.

I feel like a true evaluation is best conducted on the extremes, if in fact we want to draw a conclusion from twist rate alone. Lets run the same test with a 18″ 17 twist and a 18″ 7.5 twist, pressure sensors attached, load work-up to pressures within +- .5% using 3 different powders and bullets. With the large separation of twist values we should negate a .0002″ bore or groove discrepancy, obviously cut rifled barrels using the same lot of steel and rifling hooks. If we see less than a 150 fps velocity change we can put this to bed as a myth, as it would amount to as little as less than 15 fps per 1″ of twist.