Quality of Nosler 2nds

[hardpan]

JD338,
is the +/- 0.2 gr a guarantee or a target? Reason is, I bought 3 bags of 160 7mm ABs, and weighed every bullet in one of the bags... Something new for me.

My overall variance mainly within +/- 0.2gr, but I have several bullets outside of that. I got a large number at 160.0 and a large number at 159.8. several at 159.9 and 160.1. I got 4 or so at 160.2, and 159.7, and 1 each at 160.3 and 160.4. and 2 at 159.6...

I am not going to return this at all. And I am going to load the heaviest and the lightest and shoot them all in the same group. I am very interested to see how much that group varies from my normal shooting with that rifle. I am expected little difference.

BTW, I measured about 10 180 gr barnes TSX. in that batch, the weights ranged from 179.5 to 180.6... I was pretty surprised...

My takeaway: I will do the experiment and then decide what impact weighing and sorting bullets has on my own shooting in my own guns.

OT3 got the engineer in me curious again!
Hardpan

 

[Oldtrader3]

Hardpan, You are finding the same characteristic bullet weight differences that I have found in normal production Nosler bullets, about 0.5 to 0.6 gr variation in weight. I don't know where Nosler gets their figures? Maybe off one machine? Are they mixing lines? I don't know but I see larger variances then claimed and some are bimodal, indicating either lack of control or multiple sources of bullets. I also got fairly large std. dev. which indicates 0.8 to 1.0 grs of weight variation in a larger sampling size. I got it on both 160 gr, 7mm bullets and 210 gr, .338 bullets, all Partitions

I don't know how many limit samples you can make from the outlyer bullets and cases that you have on hand? I would round up the outlyer high/low weight bullets and outlyer capacity (high/low H2O) cases and make limit samples of the four conditional limits that you have on bullet weight and case capacity. If you have enough to give you five shot groups of each, great if not improvise. For this exercise suspend degrees of freedom and confidence intervals.

When you have the samples made, shoot them into 5-shot groups and compare that group to a couple normal distribution 5-shot groups of measured samples at mean. Establish a mean and look at each shot in the limit group as to velocity variance from mean, Standard deviation. Look at the group physically and measure the distance of each bullet hole from mean, calculate standard deviation of the group. Do a sum of the means, ANOVA. Sum of the squares, whatever characterizes the group best and can be analysed against a mean control group. I threw away all my statististic books and am doing this from memory so it may need tuning.

Intuitively there should be differences. If there isn't you have a pretty well controlled process that is falling within whatever standard deviation is from the mean, control group. Then don't worry about the differences. If you find something write it down and try to replicate it later.

Degrees of freedom may be an issue but so what. Just try to find what is different and what you can expect in velocity and physical variance of the mean groups versus groups from the limit samples. Nothing complicated.
:roll:

 

[hardpan]

Hardpan, You are finding the same characteristic bullet weight differences that I have found in normal production Nosler bullets, about 0.5 to 0.6 gr variation in weight. I don't know where Nosler gets their figures? Maybe off one machine? Are they mixing lines? I don't know but I see larger variances then claimed and some are bimodal, indicating either lack of control or multiple sources of bullets. I also got fairly large std. dev. which indicates 0.8 to 1.0 grs of weight variation in a larger sampling size. I got it on both 160 gr, 7mm bullets and 210 gr, .338 bullets, all Partitions

I don't know how many limit samples you can make from the outlyer bullets and cases that you have on hand? I would round up the outlyer high/low weight bullets and outlyer capacity (high/low H2O) cases and make limit samples of the four conditional limits that you have on bullet weight and case capacity. If you have enough to give you five shot groups of each, great if not improvise. For this exercise suspend degrees of freedom and confidence intervals.

When you have the samples made, shoot them into 5-shot groups and compare that group to a couple normal distribution 5-shot groups of measured samples at mean. Establish a mean and look at each shot in the limit group as to velocity variance from mean, Standard deviation. Look at the group physically and measure the distance of each bullet hole from mean, calculate standard deviation of the group. Do a sum of the means, ANOVA. Sum of the squares, whatever characterizes the group best and can be analysed against a mean control group. I threw away all my statististic books and am doing this from memory so it may need tuning.

Intuitively there should be differences. If there isn't you have a pretty well controlled process that is falling within whatever standard deviation is from the mean, control group. Then don't worry about the differences. If you find something write it down and try to replicate it later.

Degrees of freedom may be an issue but so what. Just try to find what is different and what you can expect in velocity and physical variance of the mean groups versus groups from the limit samples. Nothing complicated.
:roll:

Hi OT3,
Pretty much, that is the process that I will follow with a couple of exceptions. I want to test the impact of weight on grouping. So, each group that I shoot will contain at least 1 bullet in the center of the weight band. 2 on the high end and 2 on the low end. The weather here is HOT, so it will be a while before I can get enough loaded and get to the range. Also, before I even started any of this kind of testing, my gun shot 1 inch 3 shot groups, and usually 1 inch 5 shot groups. Sometimes, though, the shooter (me), makes a mistake to put a shot out of the group. I will publish the findings here. But, doves were yesterday and Sat, and hunting supersedes range testing!
Hardpan,
BTW testing will include chronograph results. Cases will not be crimped.

 

[SJB358]

Sounds like a great experiment Hardpan! Always kinda neat to see how these tests work out. I think the shooter, the gun, and the barrel probably make more difference than the Partitions most of the time. I have had such good luck with PT's 2nd's, in all of my rifles, I really have never thought of them being any less accurate than any other bullet. Matter of fact, I have been slowly developing all of my heavy lifting loads around them! Scotty

 

[Oldtrader3]

Of course hunting trumps shooting. It is your dime. I am just being helpful, not burden you with work. I don't do that anymore, my wife does that now!

However, I would appreciate some insight into your findings as far as relevency and variance. That is when you have the inclination and time, thanks