Thread: Let's talk about velocity Extreme Spread (ES)

This is mostly a discussion about which statistical measure is the most useful for the purposes of assessing what we want to assess - one tool the pros use (think election polling and the like here) is the calculated probable margin of error in the sample size. This is another useful way to assess if the info generated from a sample size is going to be reliable over the long term.

For our purposes on the topic of reloading and shooting, I feel that MV SD and ES are two different measurements and both give us potentially useful information but answering a different question. MV SD (or specifically a low figure) is an indication that if the group size is acceptable, the likelihood of a reliable and predictable impact point each time you pull the trigger is greater. A high ES figure indicates that the potential for a higher vertical spread of impact points on target is possible, everything else being equal. Ideally, you want as low an ES as possible, as low a MV SD as possible, and a load that performs with a small group size as possible. While using the pill design that you want, at the velocity you want, with a reliable supply of components, and not eating the insides of things.

Not much of an ask.

Originally Posted by No.3

This is mostly a discussion about which statistical measure is the most useful for the purposes of assessing what we want to assess - one tool the pros use (think election polling and the like here) is the calculated probable margin of error in the sample size. This is another useful way to assess if the info generated from a sample size is going to be reliable over the long term.

For our purposes on the topic of reloading and shooting, I feel that MV SD and ES are two different measurements and both give us potentially useful information but answering a different question. MV SD (or specifically a low figure) is an indication that if the group size is acceptable, the likelihood of a reliable and predictable impact point each time you pull the trigger is greater. A high ES figure indicates that the potential for a higher vertical spread of impact points on target is possible, everything else being equal. Ideally, you want as low an ES as possible, as low a MV SD as possible, and a load that performs with a small group size as possible. While using the pill design that you want, at the velocity you want, with a reliable supply of components, and not eating the insides of things.

Not much of an ask.

No. ES and SD are different ways of measuring the same thing. SD is better at it. ES tells you nothing that SD does not, and tells you less. Use SD if you care about measuring your velocity consistency. It is easy to calculate and all modern chronographs do it for you. Get a sample of at least 10 shots. End.

@McNotty

Here is an example of where poor velocity consistency does start to make a difference. SD of 27 at 650 metres. Load is very precise at 100m (10rd into well under 1moa) but at 650m - hit probability on smaller targets is poor. 14 shot group. Pretty well centred - but you'd get in trouble making corrections off a small sample.



The vertical on target is visible on the chrono as shot. Low velocity turns into low impact.

The actual result on target matches predicted from the AB WEZ pretty well.

Said group would still be fine at?? 3-400??

Yes, the effect of poor velocity consistency grows further out. An SD of 27, while resulting in lower hit probability on smaller targets at 650m - has basically no practical effect at 400m.

This 14rd group was 34cm tall - about 1.8moa. So if well centred (as it was) - still a high hit probability on a 2moa target even at that range.

27fps SD is very much on the high end.


As I show in the first post - hunters really don't need to know about velocity consistency at conventional ranges

Originally Posted by gimp

No. ES and SD are different ways of measuring the same thing. SD is better at it. ES tells you nothing that SD does not, and tells you less. Use SD if you care about measuring your velocity consistency. It is easy to calculate and all modern chronographs do it for you. Get a sample of at least 10 shots. End.

ES with a good consistency will give a low SD. If you have a load that the ES is generally OK but with the odd round that is an extreme outlier in terms of velocity the load can be crap for grouping but with what looks like a reasonable standard deviation figure (SD smooths outlier figures)... They are telling you different things about the same sample that provided the information. Of the two, ES is the 'more raw' measurement and probably the more useful...

Nope ES is a junk metric. If you think you have, say, the odd bit of bad brass, or maybe junk loading technique and get occasional apparent velocity outliers, you need to know whether they fall within the expected range of velocities or not, to know whether you actually have a problem. A properly measured (i.e. enough samples) SD tells you this. ES tells you nothing except what are the 2 farthest apart velocities you have measured.

Precision at short ranges is uncorrelated to velocity consistency.

Here's an example. I have a load with an average velocity of 2697 fps and an ES of 63 measured over 32 shots. If I get a shot that measures 2750 fps, do I have a problem? Show your working

I have another load that has an average V of 2705 and an ES of 40 measured over 10 shots. What kind of vertical dispersion should I expect all shots to fall within at 600 metres?

Perhaps I'm wrong - please explain exactly how you propose to use an ES number functionally - measured how and for what purpose. Maybe a real example with some real numbers

I think that we will get better information (more data sets) now that the likes of the garmin chronograph are here. i see a lot of them getting plonked down on the range everytime the owner shoots. In our club situation that can be the recording of 60 rounds per rifle (or more) for the day.

Its how people interpret that information, that will be the telling point, for me the calculations on ES and SD and average velocity are lost in the noise i am actually more concerned with what happens at the target. I leave the maths up to the devices and concentrate on my job of making a good shot.
If I shoot 60.10 and am told my ES and SD were double figures I couldnt care less( I would be surprised) but if the target says i shot 1/2 MOA for 10 shots I am happy.

Yeah the purpose of the thread was to educate on velocity consistency and provide a consistent way of talking about and measuring it, and who should actually care - because there are a lot of people pointlessly worrying about it for short range hunting loads, or trying to use invalid measures with small samples to make statements.

If you have the facility to shoot some decent samples at longer ranges and see what the actual vertical spread looks like, it's a direct measurement of the only thing velocity consistency matters for so that's clearly a good approach

Originally Posted by gimp

Precision at short ranges is uncorrelated to velocity consistency.

Agreed, a few times I have had the crony out when testing/practicing/running in various benchrest rifles

I can tell you its best just to get an idea on velocity (68 grain @ 32-3500 fps), only for interests sake, then turn it off......the ES is hideous, 40-60-80 & in good conditions the groups are tiny at 100 & 200

Yeah, could never figure out why there was so much emphasis on ES when SD is a much more useful indicator. And a normal hunting distances it’s not an issue regardless. I use velocity more as an indicator of approaching max loads & pressure

If it made a significant difference up short people with good barrels wouldn't be posting up 10 round ladders shooting sub moa. My old 6.5x47 was the first rifle I ever reloaded. I tried the satterlie method with 10 rounds and the entire group was about 0.6. I then as a novice thinking speed was the most important metric loaded another 5 going further up and picked the load on the bleeding edge of pressure. There was 5 grains variance from lowest charge to highest in the 10 round ladder and another 2.5 in the next. Poi did not change appreciably. The I sold the gun as didn't have a reloading g setup at my flat and when I started again realised that those results are not to be expected and were the sum of a inherently accurate catridge, a moderately heavy low recoiling rifle, good bedding, and high quality components that worked with that barrel. Id have had a sub moa 15 round group with a velocity spread of 200-250fps.

ES does in theory give the worst case scenario but it only does so for that session. Hornady testing showed that for ES type measurements ie ES or group size that the accuracy of the number tends to take sample sizes that are unreasonable to stabalise (ie in 100s) as they only consider 2 data points so to be accurate they pretty much require you to have shot the worst 2 shots possible or very near it which by pure chance takes luck. Sd type measurements such as SD or mean radius stabalise much faster around the 30 shot mark. They can be used to fairly accurately predict the likelihood of a shot falling away from the average. As a predictor its much more useful. It's provides a metric to actually make a judgement call of the likelihood of a shot falling outside a desired/required rrange. And ES tells you what could be the worst case whereas SD gives the likelihood that its happens. One can be used to make a judgement call the other i could tell anyone there load "could" shoot above or below there average by 100fps by pure unadulterated chance same as you could win the lottery.