Seating depth prs rifle

[Markg308]

Which load would you pick for a prs rifle? Curious to see the different opinions.

 

[dminn1]

Run your two bugholes against each other at distance and a winner will emerge.

 

[JFrank]

12 & 13

 

[Cfshooter]

12 & 13

Agree!

 

[Varminterror]

It’s a PRS rifle, I wouldn’t overthink it.

 

[RegionRat]

I can't see your target for jump 16, but it appears you should be looking at the shorter jump values since that is where they seem to knot up.
I see you randomized the jump target, but if you just plotted the results in order, you would see there isn't a real question here. Your smaller jumps gave you significantly tighter groups. Your velocity stats are not worth much as they have no statistical weight.

The advice above to explore the best results at distance as your next step was good. In my opinion you will want to play with smaller jump values and also tune the charge there while testing shotfall at long range. YMMV

 

[Markg308]

I can't see your target for jump 16, but it appears you should be looking at the shorter jump values since that is where they seem to knot up.
I see you randomized the jump target, but if you just plotted the results in order, you would see there isn't a real question here. Your smaller jumps gave you significantly tighter groups. Your velocity stats are not worth much as they have no statistical weight.

The advice above to explore the best results at distance as your next step was good. In my opinion you will want to play with smaller jump values and also tune the charge there while testing shotfall at long range. YMMV

I think you are mistaken where I wrote the group size on each target for the jump distance

 

[RegionRat]

I think you are mistaken where I wrote the group size on each target for the jump distance

Sorry, didn't have my glasses.... Then for sure 13 is worth exploring to see how wide that zone is.
The only challenge I see (now that I have my glasses on...) is that the harmonic isn't even but the one at step 13 looks like a better candidate.
I would take the charge and seating depth testing to distance around this step and see if you get some forgiveness for climate changes.

I have a minute to plot this so you can see what I mean. Your velocity is good and steady. The group stat is on the right and the speed is on the left.

 

[Eagle Six]

If it were me I would shoot #13 again 3, 3 round groups BTO 1.825, 1.826, and 1.827. And because I don't load at the range I would also take 3, 3 round groups of BTO 1.826 with 30.7, 30.9, and 31.1 grains. From that I would stretch it out as far as practical in good conditions.

 

[Rob01]

Way too much testing and wasting components for a PRS rifle. I load mine .020" off and rarely move them as long as I can fit them in the mag. I would have loaded them every .010" off to see where they work as you will be burning your throat out and that distance will be changing anyways. Don't overthink it.

Also as mentioned go shoot them at longer range and see if it holds tight. You will almost never shoot at 100 in matches. Shoot them between 300-800 yards.

 

[Varminterror]

What do our error stats and confidence interval look like for these mean radii of these groups? The plot above might make us want to chase #13, but if it’ll we repeat the test, does that valley really repeat, or are we finding confidence in coincidence? Maybe the error bars on this would support a trend, maybe replication would corroborate… or maybe we see something completely within the noise of the experimental data set, and the small group here happened only because ONE group had to be the smallest…

 

[RegionRat]

Your point is well taken, however....

The value of the above test isn't to have the final answer, but to select those places that are worth more investment, and eliminate ones that are not.

With a 6:1 ratio in group ES between step 8 and step 13, I would bet on spending more effort searching around step 13 at distance. YMMV

 

[Varminterror]

With a 6:1 ratio in group ES between step 8 and step 13

What's the SD of all of the group sizes pictured?

Is it only coincidental in your mind, if we just qualitatively observe that ~.35 were the average group ES for all of these groups, that the SINGLE minimum group size and the SINGLE worst group size are roughly the same value away from mean?

Just playing devils advocate, evaluating whether a trend actually falls outside of a Normal Distribution is a quick and easy check to see if we really have differentiated results: Reminding of course, that Sample Range (ES) for a Normal Distribution is ~6x SD... so a ~.35 average with ~.1 SD should produce a group at 0.05 and a group at 0.65... Also observing that 12 of the 21 groups fell within +/-0.1 of that average (57%), again, assuming a roughly 0.1SD, which SHOULD be ~68% of the groups for a Normal Distribution, and 18 of the 21 (86%) fell within ~+/-0.2 of that rough average, which SHOULD be ~95% for a Normal Distribution... Based on a quick and dirty check, it looks like non-differentiated noise...

Without corroborating data to confirm the trend, it really looks like observation bias to pick the low as a "good" and the high group size as a "bad" from this data set. We can chase coincidence by BELIEVING that 13 is better than 8, but the data really doesn't look like it's true. Repeat the test, the complete test, and confirm the trend, and there might be something there. But this doesn't appear to bear the scrutiny of statistical differentiation for me to believe chasing 13 is better than chasing 8, other than the common observer bias to derive confidence from coincidence.

The value of the above test isn't to have the final answer, but to select those places that are worth more investment, and eliminate ones that are not.

So if the goal is this, as stated, it doesn't appear there are any answers to guide selection towards places worth more investment, nor to eliminate ones that are not.

 

[CharlieNC]

Your point is well taken, however....

The value of the above test isn't to have the final answer, but to select those places that are worth more investment, and eliminate ones that are not.

With a 6:1 ratio in group ES between step 8 and step 13, I would bet on spending more effort searching around step 13 at distance. YMMV

And if you do a moving average it will highlight that group 13 and both of it's adjacent neighbors are small vs the remaining results. Further evidence that that it's not an isolated, chance occurrence.

 

[Coyotefurharvester]

I used a three group run(average) to find the "best" node
19,20,21 average 0.246 slightly better than other 3 group targets.

 

[Varminterror]

And if you do a moving average it will highlight that group 13 and both of it's adjacent neighbors are small vs the remaining results. Further evidence that that it's not an isolated, chance occurrence.

Yes, the moving average will ALWAYS display a trend - that's what it does - it moves with the sample set, because it's made to do so. We use rolling averages to determine stability in sample data when the bounds of the noise are known, to understand if we are trending towards a boundary or if an errant data point is just noise... But as I asked in my first, "do we really know what the mean radius and standard deviation of group sizes among this entire sample set really should be?" we really don't know - so we don't know if we have a ~0.35" rifle sending a non-differentiated Normal Distribution of group sizes with ~0.1" SD, or if we actually have a differentiable trend.

If we shoot 21x 3 shot groups of the same load, we'll have ONE group which is smallest, ONE which is largest. That's not a trend, and it's not differentiation, that's coincidence. We see in this plot 13 direction changes for group size trend - out of sample points. That's noisy data, no matter how we cut it. MAYBE a true trend will evolve, but it's very easy to apply simple, rudimentary statistical analysis to REALLY question whether there is anything meaningful out of this plot. We have no understanding of the SNR, because there isn't an establishment of what a standard mean radius is expected to be, and there's not replication to prove that 8 would remain a large group in a subsequent replication, or that 13 would remain a small group in subsequent replications.

The stats are pretty simple - measure mean radius from centroid for the groups, or establish a common standard for the rifle using ONE load, OR, just shoot the test again... If 8 squirts out twice and 13 squeezes in twice, and those 13 peaks all repeat, then we can say the SNR is acceptably high to know the group sizes are meaningful. Otherwise, eh, it's chasing smoke.

 

[CharlieNC]

What's the SD of all of the group sizes pictured?

Is it only coincidental in your mind, if we just qualitatively observe that ~.35 were the average group ES for all of these groups, that the SINGLE minimum group size and the SINGLE worst group size are roughly the same value away from mean?

Just playing devils advocate, evaluating whether a trend actually falls outside of a Normal Distribution is a quick and easy check to see if we really have differentiated results: Reminding of course, that Sample Range (ES) for a Normal Distribution is ~6x SD... so a ~.35 average with ~.1 SD should produce a group at 0.05 and a group at 0.65... Also observing that 12 of the 21 groups fell within +/-0.1 of that average (57%), again, assuming a roughly 0.1SD, which SHOULD be ~68% of the groups for a Normal Distribution, and 18 of the 21 (86%) fell within ~+/-0.2 of that rough average, which SHOULD be ~95% for a Normal Distribution... Based on a quick and dirty check, it looks like non-differentiated noise...

Without corroborating data to confirm the trend, it really looks like observation bias to pick the low as a "good" and the high group size as a "bad" from this data set. We can chase coincidence by BELIEVING that 13 is better than 8, but the data really doesn't look like it's true. Repeat the test, the complete test, and confirm the trend, and there might be something there. But this doesn't appear to bear the scrutiny of statistical differentiation for me to believe chasing 13 is better than chasing 8, other than the common observer bias to derive confidence from coincidence.



So if the goal is this, as stated, it doesn't appear there are any answers to guide selection towards places worth more investment, nor to eliminate ones that are not.

That's like saying if you look at velocity across 20 incremental charge weights, and calculate the SD across those 20 charges, then those velocities are all within +/-3SD . Inferring charge weight did not affect velocity. That's an incorrect basis to use.

 

[JFrank]

If there’s nothing meaningful in the statistics then one might as well just follow the target and have conviction, it’s certainly as good as just loading at .020 off just because Joe reloader did so. In this test .019 off is spitting rounds as well as the groups on each side.

 

[Rob01]

So how many of you guys shoot PRS?

 

[CharlieNC]

If there’s nothing meaningful in the statistics then one might as well just follow the target and have conviction, it’s certainly as good as just loading at .020 off just because Joe reloader did so. In this test .019 off is spitting rounds as well as the groups on each side.

I don't think it has been concluded that there is no statistical significance to these results, or that there are not strong indications for followup. There simply has not been an analysis of the data, and generally speaking not many are interested in a such an analysis. I personally spend as much time analyzing my target data as it takes to load the bullets for the testing, but I'm a nerd!