Standard deviation and groups size/pattern.

[DenisA]

Hi all,

Recently in the 30 cal thread, I was educated (nicely) on the importance of SD in LR load testing.
I learnt that although a load can shoot one hole at 100, without good SD it will likely have excessive vert at the longs.

I've been going over previous chrono results for different loads. They're only 5 shot groups and I realise more shots are preffered for SD.
I have some groups that are terrible but have great ES. In one case, the ES was 3fps which will bear great SD.

So from this, I know that a LR load must group tightly and have low SD, but..... low SD does not mean good groups. Do you agree with that?

Or, does that mean that human error or poor technique has opened the bad groups with low SD?

Thanks,

[Norm]

Or, does that mean that human error or poor technique has opened the bad groups with low SD?

Could be..... Always look at the weakest link.

Or it could be that the exit velocity of the projectile results in the bullet exiting the barrel at a particularly bad point in its frequency. A slight change in powder charge or seating depth could help.

[DenisA]

Hi Norm,

Im pretty confident with my technique. I record myself load testing and always watch back over each session for flinch and/or other technique inconsistancies.

Thats not to say that I dont have bad shots, but to shoot tight groups and then spray groups is a little too inconsistant.

......point being, I do check my own variables first, but I could be missing something.

I suspect its barrel vibration being inconsistant at the muzzle as the bullet exits. I've never dwelled on it in the past, I've always just gone with the tightest group, but now that I'm checking the SD of old groups, I'm seeing some bad groups have good SD.

I'm curious to know if anyone else has found the same thing?

At the first powder charge stage of load testing, should we be choosing the group with the lowest SD or the least amount of spread for further development?

[AlanF]

Hi all,

Recently in the 30 cal thread, I was educated (nicely) on the importance of SD in LR load testing.
I learnt that although a load can shoot one hole at 100, without good SD it will likely have excessive vert at the longs.

I've been going over previous chrono results for different loads. They're only 5 shot groups and I realise more shots are preffered for SD.
I have some groups that are terrible but have great ES. In one case, the ES was 3fps which will bear great SD.

So from this, I know that a LR load must group tightly and have low SD, but..... low SD does not mean good groups. Do you agree with that?

Or, does that mean that human error or poor technique has opened the bad groups with low SD?

Thanks,

Denis,

Back to your original post, to get good accuracy at the longs, you must have reasonable grouping at 100, I'd suggest 0.5MOA or better for any F-Class. Velocity spread, unless it is quite bad, is unlikely to have much effect on 100yd accuracy, so I'd advise looking for reasons for poor grouping before worrying too much about velocity spread. Poor grouping has many possible causes, either human or equipment - ask the short range BR people. When you have good groups, then for longer range accuracy, low velocity SD is desirable. But its not actually essential if you have a barrel and load that are well tuned (where changes in velocity cause less than the expected changes in elevation). But minimising velocity SD is very good insurance regardless of your tuning abilities, particularly at the longest ranges we shoot (say > 800m). One thing that can be overlooked in the quest for long range accuracy is projectile quality, in particular uniformity of BC. You might have tiny groups and negligible velocity spread at 100yds, but if your bullets have widely differing speeds by the time they reach 1000yds, you will get bad vertical.

So in summary, good 100yd groups, quality projectiles, and either low velocity SD or careful load tuning (preferably both).

Alan

[Norm]

Denis,

I was not saying that you had bad shooting ability. But you stated that you were looking back at the results of "old groups".

Pretty hard to look back at old results that may have only included one or two groups. It only takes one bad shot to ruin a group and anyone can do that.

Maybe you are just looking at things too hard. I'm with Alan with the SD and good long range grouping.

[DenisA]

Denis,

I was not saying that you had bad shooting ability. But you stated that you were looking back at the results of "old groups".

Pretty hard to look back at old results that may have only included one or two groups. It only takes one bad shot to ruin a group and anyone can do that.

Maybe you are just looking at things too hard. I'm with Alan with the SD and good long range grouping.

Hahahhaha, LMAO. I didn't take your comment negatively, as me having bad shooting ability. I understood what you were saying. The old groups in question I'm looking at with good SD and poor grouping are not just 1 or 2 shot flyers. In a couple of cases, None of the five shots have overlapped.

I agree thats its hard to look back at old groups and analyse them, there are so many conditions and variables that cant be remembered or accounted for so long after the event.

I think your right about the bad point in frequency.

I agree with Alans summary also.

Thanks Alan and Norm.

[bruce moulds]

denis,
it is not hard to get loads that group 1 hole clusters at 100.
at 1000, the 100 yd cluster with the lowest s.d. will have the highest score in terms of more shots closer to the centre in verticle.
hope this helps,
bruce.

[KHGS]

[

Back to your original post, to get good accuracy at the longs, you must have reasonable grouping at 100, I'd suggest 0.5MOA or better for any F-Class. Velocity spread, unless it is quite bad, is unlikely to have much effect on 100yd accuracy, so I'd advise looking for reasons for poor grouping before worrying too much about velocity spread. Poor grouping has many possible causes, either human or equipment - ask the short range BR people. When you have good groups, then for longer range accuracy, low velocity SD is desirable. But its not actually essential if you have a barrel and load that are well tuned (where changes in velocity cause less than the expected changes in elevation). But minimising velocity SD is very good insurance regardless of your tuning abilities, particularly at the longest ranges we shoot (say > 800m). One thing that can be overlooked in the quest for long range accuracy is projectile quality, in particular uniformity of BC. You might have tiny groups and negligible velocity spread at 100yds, but if your bullets have widely differing speeds by the time they reach 1000yds, you will get bad vertical.

So in summary, good 100yd groups, quality projectiles, and either low velocity SD or careful load tuning (preferably both).

Alan

Alan is spot on with what he says, one of the best long range accuracy destroying factors is inconsistent base to ogive measurements! Keith H.

[aaronraad]

Denis

Stability can cause a short range groups to appear over-sized depending how frequently a projectile is oscillating around it's centre line of flight. Here is an article on LRH from BTG Research and USAF Academy:

http://www.longrangehunting.com/articles/bullet-pitch-yaw-1.php

A poorly made projectile will not group at short or long range...unless you understand and control the error. L.E. Wilson was interviewed for an article in American Rifleman that was published in April 1962, titled Bullet Jacket Accuracy. It's one of those articles I wish I'd read 10 years ago. I think the implications from the article were too difficult to comprehend at the time. If you can get hold of a back issue it's worth a read. If I can find my copy I'll PM you a pdf.

[Ben F]

I will give you a short example of load development with a chrony.

In testing h4350 in the US (I wont test with anything less than eight and prefer ten shot strings), the shorter the string the better the ES will look which is not preferable when you are shooting long strings.

I got my 284 to shoot 12 ES and 4 SD but this group was .8 inch at 100yds. Two tenths up it shot 15 ES and 5 SD and grouped .3 inch, this is the load I used in Raton.

Load test for Extreme Spread and group not SD, you can get a gun to shoot 3-5 SD easily but if the velocity changes as the barrel fouls and heats up your ES over ten shots can be more than 30.

Hope this is of some benefit.

Ben Ferrara

[KHGS]

Ben has also hit the nail on the head! ES & group size is what counts, SD is a prediction, ES is real world. So ES, group size, good base to ogive specs & good BC = good long range groups.
Best, Keith H.

[AlanF]

In my opinion for our purposes you can use either SD or ES, as long as you use plenty of shots. Some say up to 30 shots are worthwhile in terms of getting accurate data. Particularly with many shots, SD is better because it gives more information. All that ES will give you is the difference between the fastest and slowest shot. All it says about the others, regardless of how many there are, is that they're somewhere (anywhere) in between. On the other hand, SD is directly influenced by the difference of EVERY shot from the average of the group. As an example, let's say you fire 9 shots with a spread of 5fps, and one flier that increases the ES to say 15fps. In this case, the ES (of the group) will simply be 15fps, giving no credit for the extremely good 9 shots. On the other hand SD will still indicate a very good result, and so it should. When shooting for score, such as in F-Class, getting a flier is not as disastrous as it is when shooting for group in BR. Even Stuart Elliot, one of the top exponents of short range BR in Australia, says he ignores fliers when testing, which implies he would have more respect for SD than ES.

Alan

[DenisA]

Theres some very good information and personal experience. Thanks Ben, Keith and Alan.

Having been reading up on ES and SD, the way I understand it is that they are 2 different tools. Due to the accounting of each shot in SD, SD is a tool that will always demonstrate consistancy of an average while ES, being an overall spread, doesn't always reflect consistancy unless the ES is very low.
By my way of thinking, SD puts a number to consistancy and the lower it is the more consistant the MV is.

Ben, You've answered my question though. You've acknowledged that a larger group can have lower SD and in your case by increasing the SD by 1, you've closed the group by .5moa.

So it looks like a more consistant MV does not neccessarily mean more consistant grouping.

So basically, in developing loads, group size is still King and SD's the wench that keeps him happy.

[Brad Y]

Or go to 900m or 1000yds, shoot some 5 shot groups and repeat that a few times. You will soon see what has the best vert!

[KHGS]

Quite correct Alan, SD is important, I believe ES is just as important & they are linked, but neither are worth anything for long range grouping if there are flyers. SD & ES may not cause or prevent flyers at long range. Both do help groups to be symmetrical though.
Flyers occurring with good ES & SD would indicate a bullet problem to me. This bullet problem may be simply that the barrel in use just may not "like" that particular bullet. But one thing is certain whatever the reason is for flyers they do wreck long range scores just as much as groups. Especially when other shooters shoot a perfect score! One cannot afford to ignore flyers.
Shooting groups at 1000yds can be inconclusive due to the weather affects unless we have access to a 1000yd range whenever we need to so we can choose the weather conditions to test in. For most of us short range groups with low ES (& SD) with bullets suitable for long range performance will have to do. Many shots need to be fired though to get meaningful figures.
My thoughts only, I do not have the answers unfortunately just theories.
Best, Keith H.