Just input some of my load numbers into the Pointblank software for future reference. I have heard people on the forum talk about standard deviation in loads and I have had enough statistics classes to know that zero deviation is best, but in reality not possible (at least in firearms). Anyway my question is this.

For a consistent hunting load what is the maximum amount of standard deviation you would be willing to accept with all other things being equal?

Also, what size of sample would you use to determine this unit of measure?

Well, when it comes right down to it, standard deviation, extreme spread and all those other magic and mystical formulae don't mean diddly!

What counts are the holes in the target. And the closer they are, the better...(the holes, not the targets)

I have had loads with extreme spreads of 10 FPS and a standard deviation of 4, but the 5-shot groups at 200 yards have been 2-1/2 inches.

Whereas, I have had loads with extreme spreads and standard deviations that would make the obsessive-compulsive loader turn the gun 180 degrees on the next firing, but the 5 shot 200 yard groups have been less than 1/2 inch.

In most cases, the standard deviation of your muzzle velocity isn't enough to have much effect on group size.

The commercial ammo I've tested seems to run about 35 fps standard deviation.

SD is a slippery critter to tie down. It generally takes a few dozen samples to get a decent approximation.

S.D. is not a speed number, but a percent. If I remember right the no. is the percent of shots that should fall outside the extreem spread of that string of shots. I bust my hump trying to get my cases all the same, and generally end up with a S.D. of 8 to 11. They seem to shoot well, but I have seen shot strings with a smaller S.D. no. Bob

300winnie ----- I like ricciardelli and Hot Core's responses to your post. I have been shooting over chronographs for six years now, and have come to the conclusion that thay are like women, just about the time you think you have them figured out, Whappo-they do something else. ----- I have 3 charts with a SD of 0 that I keep, just to show it can happen. Two of those were one hole groups (3-shots), and with the same rifle. The other was another rifle and while a very good near one hole, not quite. As odd as it sounds these were with rather large magnum rifles, .358 STA's. I have had some SD's that were very high that were also one hole groups with those rifles. ---- I will say this, when I have a one hole group with a very low SD, I have that warm fuzzy feeling all shooters want, everytime they shoot. This is my experience, for what it is worth. Good shooting.

SD is not a percent. It is a measure of dispersion... how spread out your data is. It comes out in whatever units you are making your measurement in. If you're measuring feet per second, SD is in feet per second.

If you want to assume normality, which is usually a good assumption for muzzle velocities, 68% of your shots will be within 1 standard deviation of the mean, 95% within 2, and 99.7% within 3.

If I am wrong I apologize, I have always heard the explanation of S.D. expressed in percent. I will research this and get back to you. Bob

Bob, please accept my apology. I was a little abrupt, and should not have been. After all, we're all just friends trying to help each other out.

Trust me on the definition of SD, though. I do a lot of stats.

No problem that way at all, I am just curious as to how I could have thought wrong for so long. I have understood the percent figure was what it was for years, and discussed it with other guys. In allthis time noone has ever strieghtened me out. Bob

denton

I have spent the last couple of hours going through google and jeeves to find out about standard deviation. A lot of information, but to be honest with you, most of it is too much of a brain strain for me. You are very right in what you said, but I found out why my idea of percent is stuck in my mind. Quite frequently S.D. is quoted in a percent, but as you say in a percent of something. In the case we are interested in it is a percent of bullets. Actually the percent that will fall inside of one standard deviation. One standard deviation is defined as within 99.7% of the mean, two is within 95%, and three standard deviations is within 68%. In my particular case approx. eleven bullets out of one hundred would be outside the 99.7% envelope. This is the way I interprate what I read, if you know any different, let me know. I am sure that I learned something today, I just don't know what. Again I am sorry for making a statement that I THOUGHT was right. Bob.

I just reread your post, and I thought the percents were the other way around. Within 99.7% was one S.D. Back to Google.

I am going to get tired of this, but again I'm sorry on the one two and three deviations. I told you it was too much for my poor brain. I've decided that this one time I am going to have to take someones word for something without figuring it out myself. In my own defense, it can be confusing, but the alphabet is confusing to me a lot of the time. Thanks for the information. I really do appreciate being set strieght on this. Bob

It does help if you are dyslexic.

Here's an easier way to think of it: What shooters call "maximum spread", statisticians call "range". For every sample size, there is a conversion from range to standard deviation. As range grows, so also does standard deviation.

For a group of 5, you divide range by 2.33 to get standard deviation.

BBeyer-

SD is the average distance of the data, from the average OF the data. Read that again slow, it's right

Units are in whatever units your data is in, measuring speed will put them in fps. All it says is that if you take the average of the speeds, each shot was an average distance from that average. That is the SD.

Well, sort of.

It is a special type of average deviation from the mean. You square all the deviations, then you add them up. That's the "sum of the squares". Then you divide by n-1, rather than n. Then you take the square root.

It's still a type of average deviation, but it's a little bent. That's a very good way to think of it. SCW has expressed it well.

The dispersion of velocity (within any reasonable limits)has been shown time and again NOT to correlate with short-range (100 yd, 200 yd) accuracy by bench rest shooters, most of whom don't even weigh each load.

At long ranges, like 1000yd, velocity variation can be very important, as working an exterior ballistics program will quickly show.

Denton, the way you are referring (can't remember the names of the two methods, divide by n or by n-1, off the top of my head) is definately more popular, I guess that was the way I should have described it.

All in all, I can't see the use of it in shooting. I don't even have a chronograph, but I still get pretty good groups out of my swede. I'm afraid if I got the chronograph I would quit seeing the forest for the trees, so to speak.

The way I was told to interpret what my chronograph was telling me about SD was to look at it as a number that would predict how much on either side of average any given shot would be. IE, if you have an average of 3,000 fps with a SD of 10, you could predict that with some degree of certainty that all shots would be within 10 fps of 3,000. The more shots in a string, the more certain you could be of being from 2990 to 3010.

My first chronograph was a Schmidt-Weston. It simply read out the fps for each shot. In order to get SD I had to imput a rather long complicated algerbraic formula into a calculator. I then bought a calculator that had a button for figuring SD, you just input the velocities and hit one button, viola, there was the SD. All that work for something I barely understood!

Now-a-days I simply write down what the pact chrono says. It also states something called AD,(I think it stands for average deviation). It's always a smaller number than AD. So what does AD do for us?

I don't see getting all tied up in your socks over standard deviation, either. To me, it's a relatively useless feature on a chronograph.

Standard deviation is something nobody would even mention if we had to do it longhand.
If it were not a one button function of the chrono, it would not be worth knowing.

Velocity spread is more useful for most purposes.

That being said, lower SD is better than higher SD.
But even a super low SD is absolutely NOT a measure of accuracy.

The holes in the target are what matters.

Travis F.