A reloading experiment

C

cdoc42

New member
I want to share the results of an experiment I did today. I have two Remington Model 700 rifles in.270 caliber and I decided to see if neck sizing would provide any advantage in accuracy over standard full length resizing. In addition I looked for any difference between using standard vs magnum primers.
Recognizing there is no universal acceptance of a 3-shot group to prove anything, I used that because I don’t ever expect to shoot even that many at one time at a deer. All loads had the primer pockets cleaned.
Each rifle prefers a different bullet. Rifle #1 is 31 years old and has a 22-inch barrel; it prefers Hornady 150gr spires with 56.0 gr of H-4831. Rifle #2 is 20 years old, has a 24 inch barrel and prefers Hornady 140gr SST with 56.0 gr of H-4831. These loads had been previously developed and are accurate enough for hunting.
Rifle #1had:
1) R-P cases full-length resized, measured varied lengths but all close to maximum of 2.540 inches and charged with CCI Large Rifle Magnum primers. The 3 shot group was 1.098 inches with shots landing 1.25” to 1.50” above the point of aim.
2) R-P cases neck sized and trimmed to minimum length of 2.530 inches, charged with CCI magnum primers The 3-shot group was 1.973,” with shots landing 0.75 to 2.25 inches above the aim point.

3) R-P cases neck sized, trimmed to minimum 2.530 inches but charged with CCI large rifle primers (not magnum). The 3-shot group was 0.723 inches, with all 3 shots landing 1.50 inches directly above the point of aim, with 2 holes touching.

Now I need to repeat this load but with cases that range between 2.530-2.540 inches in length, neck resized. But clearly this load should not use magnum primers.

Rifle #2 had:

The same charge of H-4831, large rifle primers (not magnum), all cases varied between 2.530 and 2.540 inches. The only difference was neck resize vs full length resize.

4) Neck sized: the first shot out of a clean barrel went 5 inches high, directly above the point of aim. The next 3 shots grouped 0.723 inches, above and slightly to the left, 2.5 inches high.

5) Full length resized: 3 shots grouped 1.098 inches with 2 touching and breaking paper between them, landing in the same spot – slightly left, 2.5 inches high.

What did I learn? Both of these rifles are more accurate with neck sized cases and non-magnum primers. It remains to be seen if varied case length matters in Rifle #1.

For those handloaders who are early in your hobby, this is an example of how much fun it can be!
 
Recognizing there is no universal acceptance of a 3-shot group to prove anything, I used that because I don’t ever expect to shoot even that many at one time at a deer.
Click to expand...
You were not shooting "at a deer", you were attempting to see if there was a difference between neck-sizing and full length-sizing in handloads. In that scenario, five-shot groups would have been more meaningful.
I wrote a computer program once that would simulate five random shots within a given boundary (one inch). Often times three of the five shots would form a group, indicating that one can too often be mislead by mere chance. Three shots as a way higher potential for misleading results (inadequate sample size), than does five shots.
Now I suppose someone is going to post that ten shots would be better. :rolleyes:
 
If you shoot 5 to 6 three-shot groups with a given cartridge and rifle, you're probably gonna see a 4X spread in their sizes. 10 shot groups are better. I prefer 20.

How much does our shooting skills effect the group sizes?

I give credit to tool and die maker Ferris Pindell of Sierra Bullets in the 1950's for convincing them (and top ranked benchrest shooters) to full length size their fired rimless bottle neck cases setting shoulders back a couple thousandths. Case necks end up better centered on case shoulders, the part of the case that centers the neck and bullet in the barrel when fired. Mr. Pindell is one of the "P's" in the PPC family of cartridges.
 
Last edited:
Ha! Ha! Ten shots would probably prove there is no winner among the competing components of the experiment......

But I get your point. If I do the whole thing over tomorrow I would probably get different results. But it was fun.

However, if the plan was to actually find the best load for hunting, it might have made more sense just to fire one shot from each of the 5 loads at individual targets and call it a day.
 
Here are computer generated groups for the standard bivariate normal distribution with equal radial standard deviation in both the horizontal and vertical axes, as you would find in zero wind conditions for a gun that produces round groups. The upper left is 1000 consecutively generated shots. The yellow cross is the center of all 1000 shots corresponding to ideal point of aim at the center of the target.

The remaining three targets are the first, second, and third consecutive sets of three shots (1—3, 4—6, 7—9). The green X shows the center of each group of three. Ask yourself, how much will any one of those groups tell you about how to expect the total of 1000 shots will behave?

Imagine you used those last three targets to adjust your sights. After adjusting the sights to String 1, String 2 would make you move the adjustment up 5/8 MOA and left 1/8 MOA, then when you fired the third group, you would have to come back down the 5/8 MOA, and thinking you must have done something wrong or your scope was having a problem shifting between two aiming points 5/8 MOA apart, vertically. And you'd be wrong in all that cogitation. If you fired all nine, the average would still be about 3/8 MOA below the yellow cross (not as close as Sting two randomly happened to get you, but you have no way to know that from the groups). But where going by the individual targets only gave you a 1-in-3 chance of picking the closest one, the nine gives you a compromise that at least is not terrible.

You can see how reacting to individual small groups can have you chasing the correct sight setting. You also can see why board member Statshooter, a professor who teaches statistics, uses 30-shot groups to evaluate shot placement and group size.

attachment.php
 
If you are using a "skinny barrel" rifle and or are concerned with POI shifting as the barrel heats up you can do several smaller shot groups on several targets without tweaking the scope setting do a composite using a program such as OnTarget TDS

https://ontargetshooting.com/

great program and Jeff provides excellent customer service
 
Any barrel profile will walk shots away from group center as it heats up when it's fit against an out of square receiver face.

32 inch skinny light weight 308 Win barrels correctly fit won't walk 20 shots fired 15 seconds apart.
 
Neck sizing vs FL sizing has nothing to do with accuracy. Neither does "standard vs magnum primers".
2.530" in length is too short.
 
I once did a computer experiment very similar to Uncle Nick's and came up with virtually identical results. If you have a two-minute rifle and shoot five-shot groups, you will almost always have three or more shots making a very tight group (often even touching) and at least one shot that stands noticeably away from the group.

This is why most internet shooters ignore the stand-away shots, calling them "flyers" and look only at the three shots touching. That way they can call their two-minute rifles half-minute rifles.

Tim
 
T. O'Heir said:
Neck sizing vs FL sizing has nothing to do with accuracy.
Neither does "standard vs magnum primers"
Click to expand...
Easily proved otherwise with good quality rifles, tools, components, reloading techniques and marksmanship skills.

If one cannot tell the difference, at least one of the above isn't up to snuff.
 
Last edited:
TimRB,

Lots of folks also shoot all day long until they randomly get one exceptional group and then declare "this is what the rifle can do when I do my part". It is human nature to try to find cause and effect explanations for random events and coincidences. It is why there are so many conspiracy theories.
 
My takeaway from this is any pursuit of a MOA rifle is best confined to the activities of bench rest and long range shooters. Those of us who handload for recreational routine target shooting or hunting can recognize from Unclenick's 1000 shot presentation that 99.98% of the shots fired were 2 MOA or less (as best I can discern from the picture). As pointed out, any day one gets 3 shots in 1 MOA or less (with flyers that don't exceed 2 MOA) you should be happy about that.
 
Cdoc42,

MOA and better is easily attained by hunters and recreational shooters using and doing all the right stuff.

Service rifle competitors shooting 22 and 30 caliber semiautomatic rifles in 3 positions at 200 through 1000 yards with handload or reload ammo that tests half MOA at 600 yards. That level of rifle and ammo performance is needed to win the big matches and set records putting 80 to 90+% of their shots inside the 1 MOA X-ring. People shooting 24 to 30 caliber bolt action box magazine match rifles in the same events often reload or handload ammo that equals the accuracy of the benchrest discipline's ammo.

One of my first service rifle competition shooting Instructors said one's marksmanship quality is best revealed by letting him shoot a very accurate rifle and ammo.

Are you familiar with the "ball and dummy" exercise? I had to quit flicking my finger off the trigger when the round fired. Hold the trigger back until the rifle quit moving from recoil.
 
Last edited:
Cdoc42,

Keep in mind I dictated the group SD to produce a 2 MOA 1000 shot group by giving it a 0.75 MOA standard deviation. The one MOA ring is about ten holes in diameter, so 30 cal at 300 yards would make holes roughly proportional. I mainly wanted to show groups that make it look like the gun and load are under 1 moa most of the time (3-shot groups) can actually grow to two moa when you give outliers enough opportunities to show up (a thousand opportunities altogether). But in this instance, 50% of the population (the red 50% CEP circle you can barely make out in the 1000 shot target) will be at about 0.63 MOA.
 
Problem with Nick's and Bart deal, it doesn't take in real world shooting conditions.

The 800yd,900yd,1000yd target is 72" square with is six ft square with means you can have a shot any where is that square and get a point. You have most accurate rifle in practice but most can duplicate group in match.

F-Class is score shoot you can shoot 20 shot 5ft group, max 6ft.
 
The 800yd,900yd,1000yd target is 72" square with is six ft square with means you can have a shot any where is that square and get a point. You have most accurate rifle in practice but most can duplicate group in match.

F-Class is score shoot you can shoot 20 shot 5ft group, max 6ft.
Click to expand...

I wish it was that easy
 
old roper said:
Problem with Nick's and Bart deal, it doesn't take in real world shooting conditions.
Click to expand...

It depends on how you define that. The whole value of the Gaussian distribution used to generate the holes is it reflects an enormous number of natural phenomena very accurately. You can add wind and other factors into such a program if you choose to, but in the no-wind conditions I specified and with a gun tuned to small round groups, you'll find that switching between the groups it generates and those made firing actual bullets, it is indistinguishable. Bryan Litz has a hit probability app you can use that dials other factors like wind in with it and, IIRC, you can choose the target.
 
You must log in or register to reply here.
Back
Top