Standard Deviation help

[Catpop]

Well I finally got to the standard deviation issue. I found a web page, "calculator.net", that will figure it for you if you plug in the numbers. I tried it and it works, but -----
Question: Do I want "population sd" or "sample sd". The numbers are a little different and it probably doesn't matter, but I would like to know what the forum uses and suggests.
Question: I think a lower sd (standard deviation) is better. Am I correct?
Question: Is a sd of 10 good?
Question: Are 5 - 5shot groups better for figuring sd than 1 25 shot group
Thanks men, Catpop

 

[berettaprofessor]

Question: Do I want "population sd" or "sample sd". The numbers are a little different and it probably doesn't matter, but I would like to know what the forum uses and suggests.
Question: I think a lower sd (standard deviation) is better. Am I correct?

Lower SD is better. You want to use the sample SD

Question: Is a sd of 10 good?

Depends entirely on what you're measuring and what the mean is. If your SD is 10 mm in group size shooting at 100 meters and you're averaging a 15 mm grouping, than that's going to put your worse groups at 25mm...about an inch or 1 MOA. If your SD is 10 cm, then that's bad.

Question: Are 5 - 5shot groups better for figuring sd than 1 25 shot group

By definition, to get a standard deviation you have to shoot more than one group...if you're looking for a group SD. The number of shots/group has more to do with the power of your conclusions.

 

[Bwana John]

Depends entirely on what you're measuring and what the mean is. If your SD is 10 mm in group size shooting at 100 meters and you're averaging a 15 mm grouping, than that's going to put your worse groups at 25mm...about an inch or 1 MOA. If your SD is 10 cm, then that's bad.

Is he measuring group size or velocity?

 

[Vol46]

Is he measuring group size or velocity?

It is usually used for velocity. Assuming that is the case, yes, a 10 FPS SD is pretty good - better for a 3000 FPS rifle load, but pretty good even for much slower handgun rounds.

 

[kelbro]

Question: Do I want "population sd" or "sample sd". The numbers are a little different and it probably doesn't matter, but I would like to know what the forum uses and suggests.

Even though you are measuring the entire population, sample is better for predictive.

Question: I think a lower sd (standard deviation) is better. Am I correct?

Yes, lower is better.

Question: Is a sd of 10 good?

10 is good, single digits are desired.

Question: Are 5 - 5shot groups better for figuring sd than 1 25 shot group

Not really. Larger samples are almost always better but the difference between your two options is negligible.

 

[Andrew Leigh]

In shooting std deviation is generally used to measure the speed of the bullets, I assume this is your goal? It is generally believed that std deviation is a measure of consistency in the reloading process as it measures the degree to which you deviate about the mean. The better your process the lower the deviation. To put this into perspective I aim for 14fps as my outside limit, I have shot 0.6MOA groups with this standard deviation, I have been as low as 5fps but that was exceptional for me. If I can consistently get a std deviation of 8fps I am really pleased.

In your case the number of results (population) is very small, the more the results the better the output, but given 5x5 or 25 these are both tiny and little difference will be seen. The population method is where you have every result recorded and you calculate the std deviation based on this. The sample group would be if you take just some of these results and do the calculations. So in your case 25 would be the population and 5 would be the sample.

If you are using std deviation to find your average group size and deviation then 5x5 groups are obviously better.

The average shooter will shoot a better 3 shot group than 5 shot group than 25 shot group. The risk of having flyers / poor shots increases with the amount of shots fired. There is a nice paper done by the "Long Family" the originator of the Optimum Barrel Time theory which explains this.

http://www.the-long-family.com/group_size_analysis.htm

Cheers

 

[Catpop]

Sorry again for being unclear,
I'm using velocity data from 1911 .45acp at 25 yard 5 shot groups.
Thanks for helping guide me thru this new area of fun, Catpop

 

[HighExpert]

If you are getting consistent 10fps deviations you are doing better than half the ammo available commercially. Congrats.

 

[Owen]

Standard Deviation, assuming a normal distribution (bell curve), will give a number about 1/6 of the total expected range. Inother range of velocities is goingtobe +/- 3 standard deviations. That is, if you are shooting a mean velocity of 500fps, and getting a standard deviation of 10, then 99.6% of the velocities your ammo produces will be between 470 and and 530 fps. Not bad, but not great either.

If you are shooting a rifle at 3000 fps, with an SD of 10, then you are looking at 2970fps to 3030 fps. The "goodness" of a standard deviation is related to the mean velocity.