Single digit E.S. dont mean squat.

[heat_and_fam]

Generally, that should be true, but with ES, you could get into a case of chasing your tail. The tails of the distribution are always what we in statistics call long. This means that they extend way off into eternity in both directions. Just by random chance you could get a speed somewhere in this "eternity" area which would lead you to think that this is a poor load when in reality it was just a fluke of random chance. The SD will tell you a different story (assuming a reasonable sample size) in this case. If you had shot say a 30-50 shot string and calculated your SD in this case that SD may be perfectly acceptable but that one shot will put your ES out of acceptance.

In that scenario I just described, the ES and SD are telling you different stories and I would argue that the SD is telling a story closer to the truth.

All of this is also assuming that the distribution of shots is unimodal or single humped. If you have two (or more) modes, two (or more) humps, then that is a different story altogether. If your reloading process is under control you should have a unimodal distribution of speeds. If not, then something happened in your reloading process that you'll have to figure out.

I'm just trying to help you not spend time you shouldn't by getting sidetracked by one random fluke should that happen. If the SD and ES are telling you the same story, then great. If not, I would trust the SD more and I would inspect my distribution to make sure that there wasn't some other influence on my reloading process (so ES isn't entirely useless, it's just nowhere near as useful as the SD).

You seem to be intent on lecturing me…to order a point of some sort. Yet have to caveat your statements with a bunch of assumptions. Not sure why you think i care? Not be be rude, but I didn't come here asking for help with my handloads or for someone's dissertation on SD…I simply said ES matters, as does an accuracy/harmonic node.

 

[DNADave]

You seem to be intent on lecturing me…to order a point of some sort. Yet have to caveat your statements with a bunch of assumptions.

I'm not here to argue. I gave an expert opinion and stand by that opinion. ES tells me nothing about how I should expect a load to behave and SD does.

We disagree and let's just leave it at that. Sorry you feel attacked.

 

[heat_and_fam]

I'm not here to argue. I gave an expert opinion and stand by that opinion. ES tells me nothing about how I should expect a load to behave and SD does.

We disagree and let's just leave it at that. Sorry you feel attacked.

An expert opinion…as a statistician…On hand loading? We'll disagree on that too. Again, not feeling attacked, just not sure why you decide to tag me on your post. Your argument that you value SD more on the process means nothing to the next guy. Some don't value ES or Sd, only the target. I guess, I'm glad you found something to hang your hat on. Thanks for sharing.

 

[Culpeper]

Learn Something New!

There's the water. Drink it or don't, I really don't care. If you don't know the difference between descriptive and inferential statistics you have more than a little bit of catching up to do.

The 5x ES standard is a shorthand way of telling people they don't have enough data to know if the distribution is normal or not.

Lol we already determined that on page one.

 

[Culpeper]

An expert opinion…as a statistician…On hand loading? We'll disagree on that too. Again, not feeling attacked, just not sure why you decide to tag me on your post. Your argument that you value SD more on the process means nothing to the next guy. Some don't value ES or Sd, only the target. I guess, I'm glad you found something to hang your hat on. Thanks for sharing.

I've never seen so many people enrolled in the NRA youth program. Who is your instructor?

 

[DoneNOut]

I'm not here to argue. I gave an expert opinion and stand by that opinion. ES tells me nothing about how I should expect a load to behave and SD does.

We disagree and let's just leave it at that. Sorry you feel attacked.

I'll buy that SD thing. I have routinely used the delete button on my chronograph to throw out goofy shots from the rest and consider it a chrono lighting error or something! LOL. So in a 10 shot string I get 2750-2760fps and a couple 2620s or 2800s I'm like, "Nope, you aren't like the rest, delete!" I feel confident to set my ballistic data at 2755fps and go shoot stuff.

 

[Justice1327]

I know what you are trying to get at, but ES is not the measure that will get you there. ES is only the tails of the distribution, the extreme ends. No matter what distribution we look at (normal, Poisson, Negative Binomial, Gamma, etc.) the tails all have extremely low probability of occurring. Close to zero probability. What I care about is where are the bulk of my bullet speeds going to be. There are only two statistics that matter for this and those are the measure of central tendency and the measure of dispersion. Here, we'll use the mean and the standard deviation to measure each.

To make sure that my bullet speeds are under control, I want a small standard deviation around a mean that I find acceptable. Plus or minus 3x the SD will tell me the upper and lower bound where I would expect 99.7% of my bullet speeds to be. No need to go to 5x as that last 0.3% is really likely out of my control anyway. If I find that I have a SD of 10 for a load, that means that my total spread from minus 3 SD to plus 3SD is 60fps., I'm probably OK with that as long as the mean speed meets my expectations. 99.7% of my shots should be within that range with the occasional, very rare at 0.3%, shot outside that range.

I don't care about the ES, I care about the SD as that tells me where the bulk of my shots are going to be and I think that is what you're after in the end. From your last sentence I can tell we're after the same thing, but I want to tell you that SD is the measure that will tell you there, not ES.

Yep, you did a more detailed explanation than I did. ES is past data and SD is a predictor…SD is what is important.

 

[heat_and_fam]

I've never seen so many people enrolled in the NRA youth program. Who is your instructor?

The only entertaining part of your comment is your profile pic…a grown man who contributes to the NRA and thinks they're actually out to defend his 2nd amendment rights. You believe in Santa too? The tooth fairy? That's nice.

 

[dstancu]

Just wanted to share something with you guys. I bought another labradar ran them side by side this weekend with multiple guns. 30 nosler 2 normas and lapua improved. Both reading different. One read es 1 sd .8. Other es 30 over same loads side by side. Every gun being different but every other geoup or so it be that far apart. I believe es matters but only if its correct. So i can see why some people dont think it matters. I think it does but if its shooting accurately at distance then the es is closer then the chrono says.

You just touched on a very important point, sir!
The LabRadar is very accurate in what it measures, BUT, the readings are hugely influenced by its position in relation to the trajectory of the bullet (aiming at the same target) and distance from the barrel axis, quality/strength of the return signal to the radar's receiver, and, most importantly, very good repeatability of said position every single session (allways place it the same!)
In other words, I can almost guarantee that two LabRadars trying to measure the same gun's shooting will not produce identical sets of results (because ony one can be in on place at a given time ), but if you don't move them, the sets should show a somewhat stable delta (differential) over the course of measurements.
There is no magic sauce in long range shooting (paper or game animals), there is only physics and science in general. To have any success, one must understand how the world around us really works, be it man-made or God's creation.

 

[Culpeper]

The only entertaining part of your comment is your profile pic…a grown man who contributes to the NRA and thinks they're actually out to defend his 2nd amendment rights. You believe in Santa too? The tooth fairy? That's nice.

OMG, we will write a letter of reference for you to the last place you were banned or got run off.

 

[DoneNOut]

OK guys, start acting civil or I'll start handing out bans like spilled Tic-Tacs!

Something to lighten the mood.

 

[heat_and_fam]

OMG, we will write a letter of reference for you to the last place you were banned or got run off.

Just to lay this out…you replied to me (was not speaking to you in the first place) with some snarky comment…and now you wanna whine because I gave you some attitude back? Grow up peter pan.

 

[Cberndt]

I'm rather surprised at the childish behavior here. It seems some people don't want information if it's counter to their beliefs.

Blocking people? Are you even mature enough to own a firearm?

SD trumps ES. Period. If you don't understand that it doesn't mean you shouldn't try to.

100' ES for hunting at 100 yards will work just fine. It doesn't take long for that to start to suck though, depending on the SD.

Statistically a sample size of 30 is minimum to draw a conclusion.


I know I don't post here much, and my comments will not be taken well by most, so if you only take one thing from this, step back and check your childish behavior. I'm sure there's an off topic section here somewhere for that.

 

[Mike6158]

So this is what the ES SD argument looks like. I thought it was more civilized than that.