How many shots do you reckon it takes to identify your best load?
“Mind blow” I mean it’s all made up self deception!!!??
- Thread starter Utectok
- Start date
Good example, ammo for canadian champs. Select bullet - different than previously used, had acquired data known to work in a range of barrels with available powder. Tested around 200 before loading for shipping and competing. A mix of getting known muzzle velocities to understand downrange performance, and different seating depths to reduce group size. You know well before then if a load is going to perform- electronic targets give instant feedback. 1/2 moa at short range is good enough sling, no scope. No point chasing benchrest groups when its supported in the shoulder.
And there you have it.
Even with previously acquired data (which probably took hundreds, if not thousands, of rounds), you still shot 200 to get what you needed. That's a far cry from the 15-20 most people shoot during 'load development'.
This is my whole point: most people are not shooting the numbers they need to actually be able to make informed decisions about the variables they are manipulating, so the load they end up with is essentially chosen at random.
Here's another example. Suppose someone shot the following targets during load development (3 shots at each charge level, systematically increasing the charge by 0.2 gr each time). Which should they pick as the 'best' load:
I expect most people would pick the third one to the right in the middle row.
However, I generated these by first creating a pool of 1000 simulated shots at each charge level. I assigned a mean group size for each charge level, as follows:
2MOA, 1.5MOA, 1 MOA, 1.5MOA, 0.75MOA, 1MOA, 0.5MOA, 1.75MOA, 2MOA (reading left to right, row by row). I then randomly chose 3 shots from each pool.
This simulates real underlying differences in accuracy, and attempting to find those using 3 shot groups. Most people would choose load 6 (a 1MOA load), when in fact the most accurate load is load 7 (the 0.5MOA load).
So - we do it again, but this time shoot 5 at each charge level:
Still struggle to tell the difference between the 1MOA and <1MOA loads (here, you might choose the centre of the middle row, which is actually the 0.75MOA load)...
I expect most people would pick the third one to the right in the middle row.
However, I generated these by first creating a pool of 1000 simulated shots at each charge level. I assigned a mean group size for each charge level, as follows:
2MOA, 1.5MOA, 1 MOA, 1.5MOA, 0.75MOA, 1MOA, 0.5MOA, 1.75MOA, 2MOA (reading left to right, row by row). I then randomly chose 3 shots from each pool.
This simulates real underlying differences in accuracy, and attempting to find those using 3 shot groups. Most people would choose load 6 (a 1MOA load), when in fact the most accurate load is load 7 (the 0.5MOA load).
So - we do it again, but this time shoot 5 at each charge level:
Still struggle to tell the difference between the 1MOA and <1MOA loads (here, you might choose the centre of the middle row, which is actually the 0.75MOA load)...
Absolutely. No argument there!
Wizardry…
So... this is what a genuine 0.5MOA rifle would look like. 100 shots, simulated from a distribution approximating a 0.5MOA gun. If you can't put 95 shots inside the central circle, it's not a 0.5 MOA gun...
Good job I have angling to fall back on! Rethinking my desire for a better rifle. Best just to go straight for “pretty” - little point in investing in 0.001 improvements from rifle or ammo components made of unobtainium, in the hope of better performance. It’s a good lesson in learning better to shoot what you have.
Angling is even worse for this sh*t - well, at least fly fishing is (I don’t know anything about other disciplines).
People invest untold amounts of time, money and resources refining their gear and flies, with no idea at all whether it makes a difference. At least with rifle shooting you at least have a hole in a bit of paper to tell you something. With fishing, you really have no clue at all. The number of possible explanations for why you didn’t catch is close to infinite…
Which is why I spend 95% of my time fishing a 5wt with a floating line and a size 12 black nymph! And do well enough that I go home happy.
its all wizardry to me, my second rifle a got in the 80s a ruger no1 came with a lee hand loader and some powder,cant remember what , filled the yellow scoop up and bashed the 85gr sierra bullet in with a hammer , that powder ran out and started to do the same with vit 140 never read a loading manual and long before tinternet and didnt know any reloaders to ask, shooting a tennis ball size at 200 was impressive to the folks i did know. turns out later that the scoop wasnt far off 35gr and been using that combination for for years with a press, scales and trimmer now even on the reds in scotland. i couldnt tell u what the fps , the wap facter is or how long it would to get to the moon but they cover your thumb nail at 100yrds and any thing i hit doesnt like it
Alternatively, you spend an awful lot less time obsessing about minutiae and far kite time actually shooting/fishing/whatever…The problem with being a statistician is that all the fun gets sucked out of your life
I’d certainly far rather be on the hill than at the range or a reloading bench!
What a fascinating pod cast and thread. To my mind it underlines what empiricism indicates, safety in lots of results pointing the same way.
I found this excellent article early on in my reloading career and it has proved most instructive if a trifle deflating...subsequently I have always based load development and testing by looking for the worst group not the best...
As a stalker rather than target shooter I am concerned with my worst group rather than either my best or my average...I want to be confident that the worst case scenario is that at 100 yards my bullet is not going to be more than half the worst group size away from the intended bullet placement...the group size that I can achieve in those conditions not the one from the bench. So all shots are counted and nothing is discounted as a flier.
Although I am never going to accumulate the sheer quantity of data a target shooter like 308tikka, I have logged and kept the target or every one of the 3248 rounds fired through my 308W from when it was new. No target record of those I put thorough deer obviously, but they are included in the tally. I have put through lots of 5 shot groups at the range with all of my hunting bullets. All of which have given me and my rifle and average of just over 0.8" with worst around 1.5" and best around the 0.4" Much the same as the Target Master factory ammo from HPS. I use the HPS ammo as practice, source of Lapua cases, and as a control if something odd happens I can eliminate the ammo and look at the shooter, the loose mod, scope mounts and etc.
I lay significance to the the first shot cold bore POI but over a few hundred groups can be reasonably sure that my POI is going to be within half an inch of POA at 100 yards if off a good high seat rest. Quad sticks introduce a little more vertical variation to bring it up to 3/4" from POA but windage is unaffected.
I found this excellent article early on in my reloading career and it has proved most instructive if a trifle deflating...subsequently I have always based load development and testing by looking for the worst group not the best...
Small-bore Shooting - Group diagnosis - by W.H. Fuller
Historic Arms Resource Centre facility for British and Commonwealth Enfield and other Miniature calibre training and small-bore TARGET RIFLES (BSA- Birmingham Small Arms Co.), conversion units and accessories with representation of other nations' training rifles, plus Postal League competitions...
www.rifleman.org.uk
As a stalker rather than target shooter I am concerned with my worst group rather than either my best or my average...I want to be confident that the worst case scenario is that at 100 yards my bullet is not going to be more than half the worst group size away from the intended bullet placement...the group size that I can achieve in those conditions not the one from the bench. So all shots are counted and nothing is discounted as a flier.
Although I am never going to accumulate the sheer quantity of data a target shooter like 308tikka, I have logged and kept the target or every one of the 3248 rounds fired through my 308W from when it was new. No target record of those I put thorough deer obviously, but they are included in the tally. I have put through lots of 5 shot groups at the range with all of my hunting bullets. All of which have given me and my rifle and average of just over 0.8" with worst around 1.5" and best around the 0.4" Much the same as the Target Master factory ammo from HPS. I use the HPS ammo as practice, source of Lapua cases, and as a control if something odd happens I can eliminate the ammo and look at the shooter, the loose mod, scope mounts and etc.
I lay significance to the the first shot cold bore POI but over a few hundred groups can be reasonably sure that my POI is going to be within half an inch of POA at 100 yards if off a good high seat rest. Quad sticks introduce a little more vertical variation to bring it up to 3/4" from POA but windage is unaffected.
Post on this by Brian Litz
“One thing you learn when you shoot a lot of groups is the wide range of group sizes that are normal.
Many shooters seem to expect a gun to shoot basically the same group size all the time, but that's simply not how it works.
What we've seen over 100's of 5-shot groups from .223 thru .375 is that the standard deviation of 5-shot groups is about 30% of the average. Statistics tells us that 67% of groups will be between +/-1 SD of the average, and 95% of groups will be within +/-2 SD's of the average.
So for example, if your long term group average is 0.5 MOA, then 67% of your groups will be between 0.35 and 0.65 MOA. Likewise, 95% (19/20) groups will be between 0.2 and 0.8 MOA. Anything in that range is completely normal for a 0.5 MOA average.
This is the nature of dispersion.
For this reason, it actually takes a lot of 5-shot groups to accurately characterize the true average precision. We consider 5-shot groups decent, but sometimes it takes ten 5-shot groups to resolve a genuine precision difference.
Remember this next time you're doing load development and call one load 'better' than another because it shot a 0.4 vs. 0.5 MOA group. Those single samples are more likely to be the same than different.
Attached picture; each row is 10 groups of the same rifle/ammo combo. “
“One thing you learn when you shoot a lot of groups is the wide range of group sizes that are normal.
Many shooters seem to expect a gun to shoot basically the same group size all the time, but that's simply not how it works.
What we've seen over 100's of 5-shot groups from .223 thru .375 is that the standard deviation of 5-shot groups is about 30% of the average. Statistics tells us that 67% of groups will be between +/-1 SD of the average, and 95% of groups will be within +/-2 SD's of the average.
So for example, if your long term group average is 0.5 MOA, then 67% of your groups will be between 0.35 and 0.65 MOA. Likewise, 95% (19/20) groups will be between 0.2 and 0.8 MOA. Anything in that range is completely normal for a 0.5 MOA average.
This is the nature of dispersion.
For this reason, it actually takes a lot of 5-shot groups to accurately characterize the true average precision. We consider 5-shot groups decent, but sometimes it takes ten 5-shot groups to resolve a genuine precision difference.
Remember this next time you're doing load development and call one load 'better' than another because it shot a 0.4 vs. 0.5 MOA group. Those single samples are more likely to be the same than different.
Attached picture; each row is 10 groups of the same rifle/ammo combo. “
Attachments
Or you could do what I did
Shoot three and call it “good enough”….
www.thestalkingdirectory.co.uk
Shoot three and call it “good enough”….
Challenges of loading non lead ammunition
Finally got round to loading up some 203gr Peregrine VLR4s in my 300 Norma Magnum Treated myself to some fresh Lapua Brass. Quickload is a bit ambitious on overall length putting it at 93.47mm In practice with bullet with driving rings that are in relief it is better to keep the rings in the...
www.thestalkingdirectory.co.uk
Attachments
That’s what I do now!Or you could do what I did
Shoot three and call it “good enough”….
Challenges of loading non lead ammunition
Finally got round to loading up some 203gr Peregrine VLR4s in my 300 Norma Magnum Treated myself to some fresh Lapua Brass. Quickload is a bit ambitious on overall length putting it at 93.47mm In practice with bullet with driving rings that are in relief it is better to keep the rings in the...
Similar threads
