What is standard deviation?
- Thread starter jambau
- Start date
How to Calculate Standard Deviation: 12 Steps (with Pictures)
Standard deviation tells you how spread out the numbers are in a sample. Once you know what numbers and equations to use, calculating standard deviation is simple! Look at your data set. This is a crucial step in any type of statistical...
www.wikihow.com
Standard Deviation Calculator
This free standard deviation calculator computes the standard deviation, variance, mean, sum, and error margin of a given data set.
www.calculator.net
The important thing to keep in mind is, that the standard deviation is just part of the equation; Since it's an expression of the variance you can expect from round to round, it does NOT mean that you can expect every shot to be within +/- 1 standard deviation. Std Dev is used to calculate the expected velocity range for your entire 'population' of shots, which is +/-3* Std Dev, so, if your Std Dev =12, you can expect +/- 36 FPS of your mean (average) muzzle velocity. A lot of folks chase StdDev with a great deal of passion, BUT, it isn't always a direct correlation to downrange accuracy. I've had single digit StdDev shoot like hammered shit, while stuff with 24 FPS Std Dev put 10 in one, .5 MOA hole @100M, during the same range session with the same rifle & bullet, just different powders; go figure....
Depends on the use.
For stuff I’m maybe going to shoot 500 ish yards I don’t care about it at all as long as it’s not atrocious.
For my rifles I regularly shoot well past 1000 yards I’d want it in the single digits and in the 5 or better range is preferable.
Good information. Thanks. My string included a clean cold bore shot that ran faster than all the others. In fact, my notes indicate that velocity ran much closer after the first 4 shots. I'm going to fire another string with the fouled barrel and see if SD improves.
So my SD's didn't change much. I'm going to try a couple different primers. I'm beginning to shoot at longer ranges (out to 1K). I'm making hits but as my skill level progresses I want to make sure I'm not being handicapped by my equipment. Any other suggestions?
A 1 moa rifle with an SD of about 10 will do quite well at 1000 yards.
At 1000 with my 260 and summer DA I’m at 7.77 mils
a shot a 20 fps drop is 7.91
Not much.
Last edited:
One thing to note, standard deviation is a statistical measurement that requires a statistically significant dataset. A lot of SDs that are touted on the internet are based on 3 or 5 shots that make guys feel good. Kinda like snapping a picture of that 3 shot clover that happened once and claiming a quarter minute rifle.
I like tracking SD mainly because I like running numbers. But to me, trigger time is far more valuable than chasing SDs at the bench.
I like tracking SD mainly because I like running numbers. But to me, trigger time is far more valuable than chasing SDs at the bench.
Based on a Gaussian distribution, +/-1 Standard Deviation statistically represents 64% of typical outcomes around the mean. +/- 2 standard deviations is 95% , +/- 3 standard deviations is 99.7%. I think what Secant said is important on sample size.
In shooting it's basically used to determine consistency of velocity which will translate to consistency of vertical dispersion at distance. I guess I haven't noticed much correlation between dispersion at 100 yards and SD, but at 1k it can be apparent.
Kind of the rules of thumb I've gone by is under ~10 is what you're going for on a 10 shot string and there will not be a ton of noticeable difference between a SD of 10 and an SD of 4. Only the amount of money it takes to get there
If you're at 12 now I don't think you'll have any problems at distance.Thanks everybody. I needed to have this put into perspective. The rifle in question consistently shoots 1/2 and better MOA out to 300 with this load and an SD of 12. That's as far as my local range goes. I only get to shoot to 1K about 4-6 times per year. So, given what Steelhead said, I'll give different primers a try. If it works good but if not I think I should leave it be unless I find a place to shoot past 1K. This was and expensive rifle for me and I don't want to burn up a barrel chasing numbers when I can be enjoying shooting the damn thing. This shit can drive you nuts!
I suggest you end your load development efforts immediately.
1/2moa and 12sd is more than acceptable for 4-6 range trips to 1k... Some people just love chasing better sd/es numbers and can go nuts doing so
So I was at the range today. Had my pet load for the CTR but loaded with different primers. Shot them at 200yds. One of the primers, CCI BR-2, gave me an avg of 15 fps more and lowered my SD to 10 over 20 shots. My 4 group averaged 1.15 inch. Very pleased with both my set ups now. Time to stop playing and start shooting.
Or build another one?
Or build another one?
10-15 or better.
Read this article because it has a lot of info and will answer your question regarding what level of precision SD plays in hit probability.
That's funny! BR2's lowered my velocity by 10-15, and narrowed my SD by 3-4 fps in my pet load compared to 210M primers.
Exactly. I was chasing the SD and velocity dragon for a long time. I have since stopped doing that bullshit because realized I'm not a benchrest shooter. As long as things are sub 3/4 MOA and under 15 SD for a 10 shot string, I'm loading a shit ton and focusing on my shooting skills.
I agree and a tight sd is nice but guys get so focused on a SD of 3 that they never run the numbers between a 50fps velocity swing at 1500 yards... That one or two outlier may not be enough to blow up your sd for 10 shots but a .6 mil drop change from shot to shot is a no go chief. I focus more on my es being under 35 fps consistently at a given da and temp.
Extreme spread is technically the difference between the highest and lowest values in a data set.
What really matters in shooting is the velocity change from shot 1 to shot 2, from shot 2 to shot 3, etc, etc.
What really matters in shooting is the velocity change from shot 1 to shot 2, from shot 2 to shot 3, etc, etc.
Read The Perverse Nature of Standard Deviation by Denton Bramwell. It's a short paper on Standard deviation that every reloader should read.
So, I just googled that paper, and it really fell apart in the first page were he says that small data sets can reliably convert ES to SD based on simply multiplying a coefficient. He uses dataset examples consisting of 2 to 7 data points.
I’m only an engineer, but performing any sort of statistical analysis on two data points is just a really bad recommendation. His credibility took a nose dive for me anyways.
I personally don't "savor" anything over an SD of 8. Won't go into the POI deviations between the different numbers, but below 8 provides a decent calculated hit probability at 1 mile for 338LM. I consider success at SD 8 or less across 10 shots.
It depends on what your intentions are though too. If you only have a 1K range then just run it.
Similar threads
