Sizing effect on SD and ES

[sammorgan3]

I'm casting a lee TL452-230TC for 45ACP.
Over 5.0 gr of WW231 I'm averaging around 830fps with a standard deviation of 22 and extreme spread of 70fps. Seems a little high to me. 4.7grs gives SD of 22, ES of 70. 5.3 grs gives SD 26 and ES of 70. How much would diameter variations of boolits contribute to this? I'm getting around .002 deviation. I have a .452 lee Sizer I can run some through to try.

[hermans]

If you use the same make of brass and you can really accurately measure the load of 5.0 gr, then for sure would the 0.002 deviation in your boolit diameter contribute to your rather large SD. I do size my boolits, and if I take care to load them pretty accurately, I normally get around 6-10 SD.
It would be interesting to hear your results when you size them.

[Tatume]

On how many shots are you basing the s.d.? If fewer than 20 s.d. is probably meaningless.

This is easy to test. Fire 50 shots and record velocity for each. Put the data in Microsoft Excel. Use the randomization function to randomly select 5, and calculate s.d. Do this over and over, and watch how the s.d.s vary. Compare to the s.d. for 50, which you can consider the population s.d. for this experiment. Next, randomly select 20 shots and calculate s.d., and repeat. The s.d.s will be much more stable, and will be almost always closer to the s.d. for 50. This Monte Carlo simulation is very revealing.

There's lots of Internet information on s.d. directed at shooters. Much of it is wrong.

Take care, Tom

[sammorgan3]

Hermans, I used my new hornady auto dispenser. :D I will load another 20 and chrono these. Tatume, only 6 at each load level, initial workup. Mixed brass, seemed like deviations were still high. Will load 20 more @ 5.0 and sized boolits. Will report back.

[Whitespider]

An extreme spread (ES) of 70 FPS for just 6 rounds fired is huge. In handguns I like to see something less than 50 FPS for 25-30 rounds fired, and prefer something less than 40. I suspect something other than boolit diameter is the major culprit... it may be a contributor, but I wouldn't bet on it being the primary cause. I rarely pay much attention to Standard Deviation (SD), I find the ES to be more informative as to the load's uniformity... still, the SD should be at least something less than ½ the ES.

[Foto Joe]

You know there was a time in the not too distant past that I let Standard Deviation and Extreme Spread rule my world. Nowadays I'm more apt to take the numbers that you stated and say "That's pretty good for a handgun cast boolit" and move on to more important items. I'm going to assume that you're shooting these out of a 1911 pattern gun and I'm also going to assume that you are not shooting at triple digit yard targets. Given the distances normally shot and the short barrels of most handguns I think that you might be chasing an unattainable goal.

If you really want to chase the numbers down you'll need to load matched unfired brass, hand weigh and charge each case and match the weights and diameters of all boolits. Even then you might find that the limitations of the gun prohibit single digit SD.

[sammorgan3]

Yes, it's a Taurus stainless PT1911. I've read several times that SD and ES don't matter that much in handguns. Though just how large a spread is acceptable has been left pretty much up in the air. I'll still go through sizing the boolits, and this time at least match headstamps on the brass. 8% for ES just seemed rather large even for unmatched brass. Now that I think about it, I used some of my older primers to load these. That could also be a contributor. I don't even know exactly how old they are. :oops: (They were part of my grandfather's reloading gear that I inherited years ago. WLP in the white box.) I've never had a FTF with them however. I'll load up some more, and set the chrony up once all this rain moves out in any case. I imagine they'll do plenty well for informal target practice at <20yds.

[gray wolf]

If you have nothing to do and or, your the kind of person that just likes to know things fine, ( I am one of those people )
Then go full steam ahead, but in the end for a 1911 45 ACP you are not going to see a difference in how they shoot.

There are so many variables that would have to be eliminated before you could come to a logical conclusion for the experiment.
If you shoot at a target and get a flyer how could you tell if that round was outside
the acceptable parameters for the SD or the ES ? you can't test it, it's gone down range.

I try to take as much care in my loading as I can, I test my loads from a solid two bag sand bag rest.
I always work up my loads in .2 or .3 from a start load to a little below max. I shoot at least 6 rounds at the smallest size target i can from the rest at 15/20 yards with my 1911. I save 4 round out of each group of ten loaded to shoot off hand.
The load I pick is the load that will cluster the six rounds the tightest.

I surely do not trim 45 ACP brass, I never weigh it, I don't clean primer pockets and use Lee dies, ( no FCD ) and all mixed head stamps. I am a hard person to beat at reasonable shooting distances.

Now if I were shooting bulls eye matches in competition I might change that some, and for all my rifle loads I go the distance as far as being particular ( can you say Anal )

My revolver rounds are trimmed for an even consistent crimp, I will use the same head stamp when I can.
I will say I am very careful about my powder charges.


I used my new hornady auto dispenser.
I would take 20 charges from that dispenser and weigh them on your good balance beam ( you do have one, right ? )
And see just how accurate it is. It may surprise you. I had an RCBS dispenser for a while when I was rich and famous.
every charge came off the machine and was weighed on a balance beam, don't ask how many went back in the hopper.

So I guess the bottom line is something like this: As long as what we do makes us happy, and we keep out of trouble,
Then no learning experience is ever wasted, time perhaps, but not the exercise.

[Tatume]

If you shoot at a target and get a flyer how could you tell if that round was outside
the acceptable parameters for the SD or the ES ? you can't test it, it's gone down range.

There are tests to determine outliers. However, as I caution my students, never discard an outlier unless you know why it fell on the outside. Even then, you need good reasons for discarding the datum.

I once analyzed a set of water temperatures that spanned the end of WWII until year 2000, at six minute intervals. I found some February temperatures that were about 80F, and some August temperatures that were about 45F. It turned out that the technician who cleaned the temperature probe housing sometimes didn't place the probe in a bucket of seawater like he was instructed, but set it on the hood of his truck instead. It continued to record while he was working. That explained the 80F temps in February, and they were true outliers. The cold August temperatures were found to coincide with thunderstorms, which were dumping ice-cold water and hail. The August temperatures were not outliers.

Take care, Tom

[Larry Gibson]

On how many shots are you basing the s.d.? If fewer than 20 s.d. is probably meaningless......................There's lots of Internet information on s.d. directed at shooters. Much of it is wrong.

Take care, Tom

First part is 100% correct and second part is absolutely correct.

I've extensively tested numerous lots of U.S. Ball and commercial ammo and an ES with 10 and 20 shot test strings of 70+ fps is not uncommon. BTW; I consider a load to be good if the ED of a 10 shot test string is less than 50 fps and the SD falls between 15 and 40% of that. I pay more attention to the ES than the SD though, especially with handgun of rifle loads intend for long range shooting (100+ yards with the Handgun and 300+ yards with a rifle).

Larry Gibson

[oldfart1956]

Sam, I'm certain running the boolits thru a sizer can't hurt anything. I do it for the sake of uniformity. But you mentioned this is a TL boolit and I'm thinking if they're out of round by .002 you might just wipe out the grooves. Be sure and lube, then size, then re-lube and it may help preserve the shallow TL grooves. Also I think S.D. is a neat number to look at, and worthless especially for handguns. Probably for rifles at under 100yds. as well. I received Gun Tests Magazine for many years (excellent publication) and kept a log of their ballistic tests for about a year. One thing stood out, S.D. had nothing to do with accuracy. Go figger? More than half of the time the load with the worst S.D. turned out to be as accurate or more accurate than the one with the smallest S.D. ! Who knew? A few years back I bought a chronograph and did some testing on my own and got the same results. The most accurate loads were seldom (if ever) the ones with the lowest S.D. number. This was with revolvers however. On self-loaders (your 1911) if may have more effect simply because it might make a difference in how the gun resets each time. (chambering the next round) It would probably take a mechanical rest to really see a difference. Just my opinion but ignore that stoopid number. Go with the most accurate that feeds 100% and shoot it. Audie...the Oldfart. P.S. enjoying the new PT 1911 as well. :)

[Whitespider]

One thing stood out, S.D. had nothing to do with accuracy. Go figger?

That's not surprising at all.
SD doesn't indicate anything about load uniformity... Standard Deviation is a way to predict the the odds of any one shot falling between various high/low parameters within the Extreme Spread. Standard Deviation calculations also assume "normal" distribution of the data points... which is not the case with erratic loads.
A Standard Deviation of, say 20 FPS, simply theorizes that 68% of all shots will fall within 20 FPS above or below the average of all shots (1 SD), and 95% will fall within 40 FPS above or below (2 SD). But what if most shots actually fall closer to the upper and lower of the Extreme Spread (say 100 FPS) because of powder position in the cartridge?? See where I'm going?? Even though the SD calculation says 20 FPS, virtually no shots will fall within 1 SD... most won't even fall within 2 SD.
(In reality... the SD would be somewhat higher than 20 for the above example.)

Standard Deviation just ain't gonna' tell ya' squat about load uniformity unless data distribution is "normal"... which it rarely will be if Extreme Spreads are high. It's the Extreme Spread (ES) that indicates uniformity; Standard Deviation just gives you something to feel all warm and fuzzy about... but it's meaningless unless the Extreme Spread is first reduced to a meaningful level. My personal "meaningful" ES number is something less than 50 in typical handgun loads, less than 30 in longer range handgun loads, and less than 20 in any rifle load. Do I always achieve those goals?? Admittedly, sometimes a guy just has'ta compromise.


I've read several times that SD and ES don't matter that much in handguns. Though just how large a spread is acceptable has been left pretty much up in the air.

That’s because it depends… it depends on the size of the target, the range you’re shooting at, even the type of handgun and how you hold it.

Shooting soup cans with a 1911, at <20-yards… a 70 FPS ES likely ain’t noticeable.
Shooting golf balls with a hog-leg single action, at >20 yards… a 70 FPS ES is a really big deal.

Looking at ballistic tables the difference in handgun boolit trajectory ain’t much for 70 FPS out to 75-yards or more… but it ain’t about external ballistics, it’s about internal ballistics and barrel time. The laws of physics state that for every action there is an equal and opposite reaction. In firearms this means as soon as the boolit begins moving the gun is recoiling… in handguns this means muzzle-rise as the boolit is still in the barrel. A slower boolit spends a bit longer in the barrel… meaning the muzzle rises just a bit more before the boolit exits. A slower and/or heavier boolit typically raises point-of-impact at typical handgun ranges.

The more relaxed you hold your handgun the greater the change in point-of-impact by velocity variation. I shoot more single-action revolvers than any other type; I use a relaxed grip and allow the handle to roll and rotate in my hand… a 70-100 FPS ES will cause measurable vertical stringing even at 15-yards, at 35-yards it’s enough to cause occasional misses at beer cans, and at 75-yards a milk jug has the occasional chance for survival. If I can keep my extreme spreads less than 40 FPS the vertical stringing is near eliminated out to 75-yards or so, and is only minimal at 100-yards.

Because of my single-action “style” I tend to hold any sort of handgun with a relaxed grip, and velocity variations make a difference… even a 70 FPS ES. Also, because of my “style” of grip, most handguns sighted for someone else tend to shoot high for me… my handguns have the rear sight cranked down tight, or near tight to the frame. With fixed sight handguns I usually need to go to a lighter/faster boolit to hit where I’m lookin’… for example I shoot a 150 grain in my .40 because the 170 grain boolits hit way high.

When I was making the annual trip west to shoot prairie dogs I always took my .32-20 Blackhawk along. I have a (hot) load developed that sends a 120 grain boolit down range at ‘round 1500 FPS, with an extreme spread less than 30 FPS for 50 rounds… I shot dogs at ranges well past 100-yards with that load. A 70 FPS extreme spread would make hits like those nothing but pure chance… at best.

[sammorgan3]

Hm. I got a little discussion started didn't I? Perhaps I was worrying about the details too much for a handgun. I think I'll just stick with the 5.0 behind the sized TL boolits for my shooting. I'm definately not trying to nail milk jugs at 75yds! :lol: Maybe beer cans at 15-20yds 90% of the time. :redneck:

I will check that auto dispenser against both of my balance scales (RCBS and Lee).
Loads will be churned out on Pro1000. Not really going for super acurate, don't know why I got so worked up over the numbers. Probably new chronograph syndrome. :lol:

[Tatume]

Standard Deviation calculations also assume "normal" distribution of the data points ... which is not the case with erratic loads.

I have constructed frequency distributions and normal quantile plots for several large velocity data sets, and have not observed a significant departure from normality. Small sample sizes make determination impossible, but given sufficient data small arms velocity distributions are far from uniform and do not show much skewness. Velocities are normal.

[Larry Gibson]

I consider SD to be about as meaningful to load uniformity as "average group" size is to the accuracy capability.

Larry Gibson

[Tatume]

I consider SD to be about as meaningful to load uniformity as "average group" size is to the accuracy capability.

My opinion on average group is tempered with thoughtful consideration also. Accuracy capability (actually "precision") can be better determined by other means, which most people are unwilling to undertake because they won't like the outcome.

I will add that I have made no claims as to the usefulness of s.d. for the average shooter, and have in fact made a case of just the opposite. On the other hand, I don't like to allow statements to stand when I know them to be flawed.

Take care, Tom

[sammorgan3]

I can see where ES would be more useful for shooting. With a large enough sample you will know that "all" velocities will fall within this range. If "all" shots count then there you go, you have to sample "all" of the spread. If most shots count then SD can matter. Ill not bother with SD anymore. :lol:

[Tatume]

When calculating ES only two shots are used. The rest are discarded.

The same is true of most group measurements. Almost all of the data are discarded.

[Whitespider]

When calculating ES only two shots are used. The rest are discarded.

L-O-L‼
Not exactly... although it is true that ES is calculated using only the most high and low variables, it is therefore known that all other data falls somewhere between those two points. The data ain't discarded, rather it remains between the the two points. If the data were discarded the the result of the calculation would not be called the "extreme spread" of all data recorded, it would be called the "variance" between two points.

sammorgan3 brings up a very good point, as does Larry Gibson.
When shooting at any target, every single shot counts... the group "average" or velocity "SD" is meaningless for that purpose because they only account for a percentage or, more correctly, a probability. When I settle the crosshairs on a tiny target way out yonder, it does me no good to know there's a 68% probability of hitting that target... just as I won't be happy knowing my handgun has a 68% probability of hitting the soup can at 50 yards. If I miss the target, I want to positively know it was me that screwed-up. If I positively know my .45 Colt will place all shots (every single shot) into an area that can be covered with a golf ball at 35 yards... then any miss at a golf ball out to 35 yards is on me.

A low SD may not equate to the smallest group... but any boolit/bullet I've ever tested sees the group tighten as the ES gets smaller. A low ES does not guarantee a tiny one-hole group... ya' still haf'ta be shooting an accurate gun using a boolit/bullet it likes for itty-bitty groups. But using the same boolit/bullet at the same velocity level, a lower ES will shrink group size every time.

My final verification test for any load, depending on the type, is to shoot 20-50 rounds over the chronograph into a single group... then I know (at least with a 99.67364% probability) exactly what the gun and load are capable of.
After that, it places any and all blame on me... no room for excuses.
Besides, confidence promotes better shooting... 100% confidence in the load will, in fact, improve your shooting.

[Tatume]

Yes, exactly, but not worth arguing about.

[sammorgan3]

Well I ran my new Lee 4-20 yesterday and cast about 200 of these bullets. Now diameter seems to be holding fairly constant, though reading between .451-.452. (Cheapo calipers usually used for OAL) I'll have to mic them and be sure. If that's the case they may be too small, I seem to remember that barrel slugging at .452 but being a dummy I didn't write it down.

[Foto Joe]

What alloy are you using. It does make a difference on what size boolits will actually drop from the mold.

[Harter66]

Mixed brass tells me everything I needed to know about the spreads.......

I used to not sweat the ''ACP'' class cartridges, then I shot about 200 9mm 1 morning to rotate the stock in loaded mags. The S&B had belly buttoned primers while 1 in 10 of the Win would soft eject (the ol' roll down your arm thing).

[sammorgan3]

Using wheel weights. Water dropped because back when I cast those I didn't know better. Ones from the other day are air cooled. Raining like hell today so I imagine I'll lube them up today.

[Foto Joe]

As an experiment which just might pay off, I would suggest a trip to the hardware store for some 97/3 or 95/5 solder (lead/antimony). Ad 2% tin to your wheel weight melt then cast some more and see what size they drop. You might find that your boolits are dropping a tad larger. The higher content of a tin alloy will diminish the shrinkage when cooling.

[Tatume]

I doubt you're going to find lead/antimony solder with 3 or 5% antimony in hardware stores. More common alloys use copper and silver.

[Foto Joe]

I've found that the Ace in my area carries 97/3. It's spendy but until you make a decision to order multiple pounds from someplace like RotoMetals it'll do the job.

[sammorgan3]

Hm. Tin lessens shrinkage. Got it, thanks for the tip!

[Foto Joe]

A wise person once told me "Using clip on wheel weights you can cast good boolits, ad a little tin to the wheel weights and you can cast great boolits".

Don't expect miracles from the tin, such as you're not going to get a .429 to drop .452 but it does help. Buying tin from the hardware store is expensive ($40 per lb) but experimentation purposes its a good way to find out if it'll work for you.

[sammorgan3]

As long as these 45acp are minute of coke can without much leading I'll be happy. Now when I start working on 44 magnum I'll be a little more picky, and If I can ever lay my hands on a 400 or 500 grain mold for my 460 I'll be chasing perfection. (15" bbl TC Encore with 4x scope on top. Probably add a brake if I start slinging 500gr of lead with it.)