How to work up an accurate rifle load

Load development

The three editions of the Lyman Cast Bullet Handbook are helpful in finding loads for cast bullets, but in my opinion often show loads at higher velocity than gives best accuracy. They also include some information that is just not true.

Other reloading manuals include loads for cast bullets, often at those high velocities.

Equipment lists from Cast Bullet Association and American Single Shot Rifle Association matches show who used what loads and where they placed. I find these equipment lists valuable.

There are a multitude of sites on the internet where one can read about loads and cast bullets; these are also interesting, but sometimes incorrect information is presented.

Be careful, check other sources, be careful, and start low on the charge.

The only cast bullet shooting that I think I know about is the low speed end of smokeless powder rifle shooting. There's a high speed end, with velocities around and over two thousand feet per second. I've tried this, with some fair results, but the noise and recoil are bothersome and I get a little concerned about bore wear and throat erosion. But that's just me.

To get good accuracy with cast bullets in rifles:

1. Cast good bullets, bullets that are completely filled out, with no wrinkles or holes or obvious surface defects. Weigh rifle bullets after visual inspection and discard those that are more than half a grain below the average-these probably have a hole or void. Use any reasonable lubricant. Make sure that the gas check-if used-is on the bullet base firmly and square.

2. The bullet must be large enough. We are told that the bullet must fit the throat, and that is probably true. At least it is said a lot. My experience is that the bullet must be BIG. Thirty caliber bullets of .312" diameter on the base bands shoot well without any sign of leading in any thirty caliber rifle I've had. I believe that rifle bullets three to five thousandths of an inch over groove diameter work well most of the time. If the bullet is a bore-ride design, it is probably best if the bullet nose is a-little-tighter-than-slip-fit in the bore ahead of the chamber. Note that the bore dimension just ahead of the chamber IS NOT NECESSARILY the same as the bore dimension at the muzzle. It doesn't seem to take a lot of jacketed bullet shooting to open the bore up a thousandth of an inch or so in front of the chamber. 

3. The brass must be in good condition. Case mouths must be chamfered in and out and square, the brass should be of the same head-stamp, trimmed to the same length; it doesn't hurt and may help to uniform primer pockets, de-burr flash holes, turn necks and anneal the necks. It may also help to select a set of brass that weighs within plus or minus half a grain. Any individual step may not make much difference, but doing it all and having a set of forty or fifty cases that are as close to perfect as you can make them will help some and give you a feeling of confidence.

4. The rifle must be in good condition, with a clean bore and no copper fouling, bright rifling, a decent crown on the muzzle, proper headspace and in good working order.

5. Your sights and your age must be compatible. Any reasonable telescopic sight is suitable for shooters of any age. Target iron sights with peep rear and aperture front are suitable for shooters of any age as long as those who need them have the required glasses. I wear trifocals, and cannot see the bull at all with iron sights and without glasses. With my glasses I can see and shoot fine with irons. Open sights, particularly with a bead front sight, introduce a sighting error into shooting that adds to the accuracy error. An older person, or one who has less than perfect sight, using open sights, is not able to tell if the group size is caused by the load or the sights.

     So we need good bullets, of the right size, good brass and a good rifle with appropriate sights in order to get good groups. This isn't about a lot of money. Today we can buy a brand new 308 rifle and a telescopic sight for fairly short money.

Don't try to work up a load when there's a howling gale, or when it's 14 below zero or 105 in the shade. Don't shoot in uncomfortably bright sunlight, or when there's heaps of mirage, or when you don't want to be there shooting. Bud Welsh wrote that he only shot when the temperature was between 30 and 80; this isn't bad guidance.

One of the best times to shoot is late on a summer afternoon, two hours before sundown. Often the air is still, a lot of the heat is gone from the day and the sun isn't too bright. If you can catch a still day with a light drizzle and temperatures of 40 to 70, you've got the best conditions.

Keep in mind that the object is to test the rifle/load combination, not your ability to dope wind or mirage.

     The best way to work up a load is to load at the range. That way you haven't loaded a bunch of cartridges that don't shoot well, and you can load right there to shoot a lot of charges that do shoot well. Load some starting loads at home, and then load increased charges at the range. It requires hauling a lot of equipment around, but I think it’s worth it.

Some folks advocate the use of just one case while loading at the range. John Ardito loaded one case on the few times I watched him shoot. Using one case slows the pace of shooting and allows for barrel cooling. It also eliminates all case-related variables.

Start with the smallest powder charge listed. Increase the charge a half grain at a time. Shoot at least two five shot groups with each charge. When you have a charge that looks promising, shoot five / five shot groups with it.

In rifle loads, do not use charges with any number to the right of the decimal point other than zero or five. Use charges of (for instance) 12.0 or 12.5 grains; don't use charges of 12.2 or 12.8 grains, this is nonsense.

With cast bullets at 1600 fps or under, my experience is that there are a few "cookbook" loads that will work in any gun-if anything will. The 32/40 is frequently shot with bullets from 165 to 210 grains, large pistol primers, and 14.5/IMR4227. I believe that most any cartridge based on the 32/40 case will shoot well with this powder and primer combination and any reasonable bullet. Then the 30/30, 32/40 and 38/55 will probably shoot well with Remington Large Pistol primers and around 14.5/IMR4227. Maybe a little more, maybe a little less. I've also found that 12.5/AA#9 works in at least the 30/30 and 32/40.

I recently got a Winchester Model 54 in 30 WCF or 30/30. Since my Martini bench rifle in 30/30 likes 14.5/IMR4227 or 12.5/AA#9, Remington Large Pistol primers and the 311299 sized .312" and NRA Alox formula lubed, that's what I loaded for the M54 as fixed ammunition with gas checked bullets. The first day, the second group was .625" for 5 shots at 100 yards. This rifle will average a bit over one inch for five, five-shot 100 yard groups.

The message is that if you start with a proven common good load, you should get reasonable groups out of the chute. 

Now the proven load may not be the best load, but it should shoot reasonably well. Reasonably well means that you should be able to average five shots in 2" or less at 100 yards with either a telescope or target iron sights with an aperture front-post front if you're under 40 years old.

If you can't do that, then making marginal changes to lubricant or powder charge or primer or seating depth or bullet alloy is not going to help. You're going to need another bullet. If a bullet that is properly fitted to the gun doesn't work, you're probably in trouble. The solution may be difficult to find but that is what makes shooting cast bullets the sport we like.

Here's another story. From summer of 2003 to summer of 2004 I tried to get a brand new Savage 12BVSS target rifle in 223 Remington to shoot cast bullets. This gun, with either the NEI 80 grain, or Eagan 65 grain bullets, should be capable of averaging under one inch five shot groups at 100 yards.

I shot it twice with jacketed bullets, 15 shots each time in three five shot groups. I lost one shot off the paper /one group with a bad sight adjustment. Groups were .825", .55", .50", .675" and .525"; an average of .615" for five shot hundred yard groups.

 I fired over 3000 cast bullet shots, used three different molds, hard to soft bullets, two primers, nine powders from Unique to IMR4350, two lubes and a wide variety of sizing die and OAL settings. The gun doesn't like cast bullets.

My hobby is buying rifles, making them shoot cast bullets and then selling them. In 40 years I can count on my fingers the rifles that wouldn't shoot cast, and I can think of none that would shoot jacketed and not cast-other than this Savage. 

The message is that if you start with a proven common load, you should get reasonable groups out of the chute. I expected this rifle to shoot averages of 1.25" early on, it didn't, and that should have told me that there was something wrong.

Once the rifle shoots reasonable groups, there are two things to vary to improve accuracy.

Vary the powder charge up and down a half a grain at a time. If your 30/30 shoots under two inches with 14.5/IMR4227, then try 13.5 and 14 and 15. Don't try 14.3 or 14.7, you're just kidding yourself. Don't change anything else, just the powder charge.

Once you've found the powder charge that your rifle likes best, vary the seating depth. Begin with the bullet almost as far out in the case as it will go, without de-bulleting if you have to take the cartridge out of the rifle. Set the cartridge length shorter and longer-risk de-bulleting-and shoot groups.

• Then change powders.

• Then change bullets.

• WRITE IT DOWN!!!

Never hurry while reloading or shooting. If you don't have enough time, then stop. There's nothing that makes me as nervous as a shooter in a hurry at the range.

Creighton Audette et al on Load Development

Bob Fitzgerald, an Old Colony shooter who knows his stuff, once told us that varying the powder charge up or down a little didn't change the elevation of the shots on the target. I tried it and found that Bob was right; half-grain and even full grain increases or decreases in the powder charge made little or no difference on the elevation of the bullet on the target. Sometimes.

Here's the story. With a load that shoots accurately in a given rifle, varying the powder charge up or down a bit does not vary the elevation of the bullets on the target. There are many people who know, or think that they know, why this is. And they're not shy about telling anybody who will listen. I don't know why it is, but I do know that it happens (sometimes) because I've done the experiments.

Turn this around, and we get another approach to load development. If it is possible to vary powder charges around accurate loads without varying elevation; then increasing loads from recommended starting loads should show a range of charges where elevation doesn't vary, and in there is where the accurate load is.

Creighton Audette is credited with devising this approach to load development, it is sometimes called the "Ladder Test".

For a purely hypothetical example, let's say we've got a rifle in 297 Blooper and the recommended starting load of LR2468 powder is 10 grains with the 165 grain cast bullet, and maximum is 15 grains. We would load some cartridges with 10, then 10.5, then 11, then ... to maybe 14 grains. We would then shoot some of each at the same target, watching as the bullets struck higher and higher on the target. At some point the elevation would stop increasing, maybe at 13 and 13.5 and 14 grains. With charges over 14 grains the elevation increases-the shots are higher on the paper.

Then we would know that somewhere in here, between 13 and 14 grains, was the most accurate charge, and we would have found this with a minimum amount of shooting and loading.

     From time to time, we hear remarks from new cast bullet shooters, concerning a wish that there was more information about how to work up an accurate cast bullet load. In as much as I have some specifics in which I differ from a lot of the conventional wisdom, I offer the following article as a variation on Joe Brennan’s works, as presented in Cast Bullets For Beginners And Expert. (Published by The Cast Bullet Assn., see castbulletassoc.org) Lets get a few preliminaries on record first. I agree with Mr. Brennan in several respects, to note a few:

His comments about range/shooting conditions.

     Don’t rush.

A point where changes of charge have little effect on impact point.

Selecting a bullet

This is to me the place to begin, and that is with a chamber cast. If you have on hand a Brownell’s catalog, ( Also available on line at admin@brownells.com, or by mail at Brownells 200 South Front St., Montezuma, IA 50171; phone 800-741-0015), look up “Cerrosafe”. The catalog discusses it, and CerroSafe is quite easy to use, and will give accurate measurements of the chamber and bore sizes. Except that it is difficult to get an accurate measure of the land diameter from the chamber cast. The chamber cast will tell you the best bullet sizes. I say sizes, plural, because most CBA bullets are of two dimensions; rear driving bands, and front body. I believe that these two dimensions need to be within .0005 Inch of a bore match, throat and lands, to shoot well. Ideally, the rear driving bands should be a “kiss fit” to the freebore section of the throat, i.e. That section ahead of the case neck mouth, before the start of the lands.

Size the bullet and gas check to this size, up to 0.0005 inch larger This dimension and the chamber cast will tell you how thick to turn the case necks. Again no more than half a thousandth clearance. The front body diameter should also be a “kiss fit” to the lands, that is close enough that a chambered bullet will show some engraving by the lands of the bore, just behind the start of the ogive. Finally, when chambered, it would be nice if the transition from front dimension to the driving bands of the bullet were to match the taper at the start of the lands. However, I do not believe this to be as important if the other dimensions listed above really fit. Using commercially produced moulds (Available from various supply houses). But it may be necessary to custom order a mould for the ideal fit.

It is not always possible to get a mould that will produce the above described fit. If the shooter has a bullet mould, proceed with casting up some bullets

For gas check bullets, the metal should be between a Brinell Hardness of 18 and 22. Linotype will give the harder range, with half and half Linotype and Wheel Weights the lower range. Later, when the shooter has worked out some of the details of getting a rifle to shoot accurately, he (she) can experiment with variations of the bullet metal.

Selecting a primer

If you have been hand loading for a period of time, you probably have a favorite brand. Go ahead and use those primers. Or switch to large or small in that brand if necessary for the proper size. Many experiments and experts say there is very little difference between primers; they are all very good.

Selecting a powder

Generally, compared to jacketed loads in the same caliber, a cast load will be a lesser charge of a faster burning powder. A powder may be selected from loading manuals (Lyman has published has published several editions to their “Cast Bullet Loading Manual”). that cater to cast bullets. Another good source is to study the CBA Match reports; if you can find the caliber of gun that you are working with. Observe the different powders that are being used, especially the most common “numbers” (powder designation) reported. One or more of these should serve well. No point in trying to “reinvent the wheel”. Make note of the three most used “numbers”.

Can not find any data

If you have Richard Lee’s book: “Modern Reloading (Richard Lee’s “Modern Reloading Second Edition”. Inland Press. Such books can generally be found at suppliers that cater to hand-loader’s); check out his Chapter 8, and especially the part that starts on page 91. Or: Locate a common cartridge (same caliber ), with similar case volume, one that you can find data for. Or, locate a cartridge with a significantly different case volume. Then by ratio of case volumes, calculate the probable minimum and maximum loads. Do not attempt to work with any powder that is slower than #4895 in burning rate. Having selected one or more possible powders, proceed to the next step, as described below.

String loads

My load testing technique. Assuming you have a cast bullet that is a good fit to your bore; load a string of each of the most common loads, in half grain steps. Starting at the lowest charge, if working from a manual, up to the top listed amount. Or if you are using data from the match reports, cover the whole range and an extra half grain on each end, i.e. half grain under and over. Fire each string on a separate target keeping a record of where each round hits, and it’s charge. Virtually all the regularly used cast bullet loads are not “gee whiz” high pressure loads. So there is no need to worry about safety in modern guns. Exception: a DOUBLE charge can be disastrous. Len Wilson, a former CBA competitor, used to talk of “Instantaneous Disassembly”. Understand that the release of varying amounts of energy in your “mechanism” will create variations in vibration and harmonics of your gun.

To demonstrate; refer to figure #1. This target was fired at 100 yards, from a rifle cal. .308 Norma Mag., 180 grain Hornady bullet, CCI #250 primer, and H450 powder. The numbers on the target represent the sequence of firing, and the charge in grains of each round. Note that the results are strung out over a height of almost 6 inches; but that rounds numbers 3 & 4 printed very close to one another. I take this to represent a ‘Harmonic’ charge. Over the years, this rifle has shot extremely well using thrown charges in that weight range. That charge range is well within the accuracy potential of many powder measures.

Incidentally, this powder was originally selected with no data. That was 1963 or 1964, and Hodgdon had not yet published data for H450. All I knew was that it was of approximately the same burning rate as H4831. This kind of response is what to look for; and notice it only took 5 rounds to get there; although there had been some previous indications that this charge range would be a good place to begin testing.

As another example; I prevailed upon another CBA competitor, Gerald (Jerry) Jahn, to tested this technique in his Heavy Class CBA competition rifle, cal. 30BR. He found that a string from 27 gr. to 30 gr., of VN133 shot into about a two inch area at 100 yards. By carefully monitoring each hit, it appeared as if the steps between 28 & 29 gr. were in the middle of the group. Some of the hits were so close, that without checking each shot impact, we would not have known which was which. Using 28.5 gr. Jerry fired a one-hole group at 100 yards using thrown charges from his RCBS powder measure. The bullet used in this process had the characteristics as described above. Using this load, he shot the following month’s club matches and nearly won the shoot. If this technique does not produce a base from which to refine your load; proceed to the next section.

TIC – TAC – TOE chart

How do you decide between 3 primers, 3 powders and 3 different charges? Yeh how do you do it? Well my approach is what I call Tic-Tac-Toe.

If you were to shoot 10 rounds each, that would be 3x3x3x10 = 270 rounds! Not impossible, but certainly time consuming.

If you shoot 5 round groups, using only 2 primers with each powder and 3 charges with each powder; mixing as follows:

This assumes that the powders are all of about the same burning rate, that is close together on burning rate charts. Note that each component has been used 3 times, and yet we have only 9 groups, or 45 rounds. Record the group size in each space matching the load. Then by cross referencing, we can see which combos, gave the better groups.

     Now it is time to select the best combos. I don’t think primers will make as much difference as the powder charge, so we could have skimped on primers and shot more charge variations. Select the three best combos. Repeat the process using 10 shot groups ( 30 rounds), (Most statisticians will tell you that you need more than one or two groups to indicate which load is best.) Again, do the Tic-Tac-Toe chart to further refine your load choice.

It only took us 75 rounds (not 270) to get to get where you are ready to refine the powder charge with variations of 0.5 grain or even 0.1 to 0.2 grains.

Now having arrived at a reasonably accurate load, it is time to refine it with several 10 shot groups. Now shoot several groups with the above minor variations. Keep records. Use group sizes as just as important as SD. The more groups you shoot, the better will be your choice of the load. At this point, a Chronograph is useful in refining a load, if available. This provides information on Velocity, and ranges of Extreme Spread and Standard Deviation. Wishing you loads of X’s.

Chronographs And Load Development

I've had a Pact chronograph for 15 or more years. It has been most valuable when preparing for a long range match with a bullet with an unknown ballistic coefficient (BC).

For Example:

We got the chance to shoot at 600 yards on the range at Camp Edwards on Cape Cod. My C. Sharps 1875 in 45/70 with a Wolf No-Grease-Groove 446 grain bullet, Rem LP primers and 21.5 SR4759 with a Dacron wad was shooting well at 200 yards. I had no idea of sight settings required for 600 yards. I went to the range with the chronograph and the gun and cartridges.

The ballistic program, "Balistic" version 4.0 by W. R. Frenchu gave me an estimate of BC of .617 at standard temperature and pressure.

Plugging that BC into the ballistic program gave me a ballistics table and a trajectory table showing that to be on at 600 yards the bullet would strike 84 inches high at 100 yards.

I sighted the gun in at 100 yards, the elevation was "3" or .150"(Rear sight divisions are .050" with a vernier adjustable to .010" divisions.) Distance between sights is 34.5", so .010" at the rear sight is 1.040" at 100 yards (100 yd = 3600"/34.5").

I needed 84" more elevation at 100 yards, 84"/1.04" = 80.8 X .010" on the sight = .808" more elevation was required. The .150" at 100 yard zero + .808" additional elevation required = .958" total sight elevation to be on at 600 yards.

My first 600 yard record shot was on the paper. If I understand my notes, after zeroing in at 600 yards the elevation was .940" rather than the calculated .958, or about 20 inches lower than calculated.

So the chronograph and some calculation and the ballistic program will serve to get you on the paper at long ranges when you've never shot that gun/load at long range. My experience is that it is not unusual for new shooters to have trouble getting on the paper at long range.

This process hasn't failed me yet.

If there are any things I know about my chronograph and shooting, they are: (normal range of velocities)

Variations in velocity have little to do with group size.

Individual velocities have little to do with where the bullet hits.

Higher velocity than average bullets can hit higher or lower or east or west of the center of the group.

Standard deviations of velocity with n = 5 are statistically meaningless. n must >/= 30, like it or not.

I've chronographed thousands of bullets, written down the velocities, done some statistics to these sets of velocities, and learned very little about loads and accuracy. 

I queried several internet sites about the use of chronographs, and got a number of responses from long range cast bullet shooters stating that variation (Standard deviation or extreme spread-each has it's champions, which flabbergasts me.)in muzzle velocity for long range loads must be kept low for those loads to be accurate. Others chimed in suggesting that short range accuracy isn't affected as much by variation in muzzle velocity as is long range accuracy. I was almost sold, until I went to "Ballistic" by Frenchu.

Here's what I got (BC = .5):

It seems pretty clear that variation in velocity, at least at the lower velocities, are as "important" at 200 yards as at 1000 yards.

I'm left with some folks who agree with me that in their experience small SD’s don't make for small groups, and who wonder, as I do, why not.

     ("Shooting Times", May 2007, pg. 27, "Velocity Versus Accuracy.- Is there A Definitive Relationship?" by Lane Pearce agrees with me that accuracy and small SD’s are not related.)

Other shooters believe that the chronograph is a valuable tool in developing a load, and that small SDs do indicate better accuracy.

"I only acquired my first Chrony a couple years ago. It gave me a window into what was really happening with my handloads. The various handbooks list x grains of y powder and a z grain bullet for w fps - but one never knows if one is actually GETTING what the books say. Then too, if one wants to use a non-standard powder (H4831 in 222 Rem with a 60 grain cast bullet) to achieve a gentler smoother acceleration, the Chrony lets one know how well the stuff is working (2375 fps, a small SD, and 2 MOA).

One can use a chronograph judiciously to develop a load that is close to max with a powder for which there is minimal data, such as the surplus stuff available from GIBRASS and others. Or, one can verify that a moderate load with such powders really IS moderate. I love chronographs." John Bischoff

Gas checks are hazardous to your chronographs

I set up my Chrony at around 12 feet or so from the rifle muzzle.

When shooting 22, 24, 30, or other caliber gas checked cast bullets I find that the Lyman brand gas checks come off at the muzzle or thereabouts and will damage the Chrony. Those gas checks have chipped the plastic sky screens, cut off the 1/8 inch wood dowels I use when it is cloudy and I don't need the plastic sky screens, dinged and dented the various vertical faces of the Chrony including inside where it folds, damaged the readout liquid crystal cells, and generally caused considerable damage and dismay.

The best fix for this is to go to a glass shop and ask for a piece of scrap Lexan at least 0.25 thick. Cut it big enough that it will protect the whole Chrony box, and use a couple pieces of coat hanger wire to hold it in front of the Chrony (use the inner two sky screen holes)." John Bischoff

"Like Joe, I don't use chronograph data for accuracy loading. After I read Harold R. Vaughn's "Rifle Accuracy Facts" I knew that just measuring velocity was useless for determining accuracy. There are just too many things happening in the barrel vibration modes to think that just velocity is important.

I use it mostly for safely loading old cartridges that have little or no data available, i.e. 25 Remington, 455 Mark I, older wildcats. If I am using the same or similar powder and approach the listed velocity, I know I am closing in on the same pressure level. There is not room for even one round of over-pressure with these older guns.

By varying the velocity and bullet weight, you can change point of impact for fixed sighted revolvers. Once you find what weight bullet and what speed "sights in" for your revolver, you can change powders and just load to the right velocity, and it will hit the same point of impact.

Lastly, I am just curious about how fast my best cast bullet match loads are going. Each rifle seems to like a little different speed with the same bullet, but I have never found any logical reason. They just are like women, no reason for something, it just is". Ric Bowman 

"I do see some good things about using a chronograph when working up loads with new and unknown powders. By knowing your velocity with a proven good load, this velocity gives you a good starting point when switching powders. Of course different powders might not shoot their best groups at the same velocities but it does give you a starting point to work from.

A good example of the chrono being a big help was back when Scot sold their powder line to Accurate and Accurate started selling 3032. Those of us using Scot 3032 simply started using the new AA 3032 with the same amount of powder. Without a chrono, we would have never known why the AA powder wasn't shooting as well as the Scot. The reason was that the AA powder was quite a bit faster than the Scot and we were getting 200-400 more fps. Once the velocity was reduced to the proven good point, the new lot of powder turned out to be every bit as good as the original, maybe even better if you were using close to a full case of powder. Sure wish we could still get the 3032 powder!

I agree that a good SD doesn't guarantee a good group.

I only dig out the chrono when I am working up new loads." Dan Hudson

"I have had an Oehler 35P for some time now. My best 1000 yard loads are tested on it after load work up at that range, they always yield a very low ES and SD they OFTEN DO NOT SHOOT THE BEST Groups at 200 yards. I work up loads with low SD and ES then test them at 600 and 1000 and usually find what works the best there.

I am often asked to test various lots of powder by a major powder company and they want to know the MV and the ES and SD for various types of powder, I assume this is to adjust burn rates and such.

I love mine, as it's taught me a lot about load workup." Kenny Wasserburger

"As one example, I was shooting reduced rifle loads, cast, and tilting the bbl up to position the powder. Two of the shots were way outside of the velocity of the others ... I had forgotten to tilt. It was easy to see the error on the chrono, which in turn explained why two shots were out of the group. Without the chrono data, I might never have known what had happened." mike in co on Cast Boolits

"I would have to say that a chronograph is an essential tool when one is working with a (chose one) odd caliber, bullet and/or powder not found in a manual. If velocities are very erratic or way out of the ball park then something isn't right and needs be corrected. It also helps to build confidence in a load if one Knows exactly what it is doing versus what one Thinks it is doing." The Nyack Kid on Cast Boolits

"I value my Pact I because it gives me a better "picture" of what my loads are doing. With load data in the manuals varying so much, this can be a MAJOR help.

Working with my .35 Whelen Ackley, for which no data is directly available, I feel a lot more comfortable knowing my velocity, since that CAN be a good indicator of if and when I may be loading too much of a given powder. Could be useful in working with small charges of the fast powders, too. Just another piece of equipment to keep me out'a trouble ... or at least TRY to do so.

Everything a man can learn about his loads can be valuable, if only to perplex him with the results. Still, it's really nice to know the velocity when we can't know the pressure. It's not the same thing, but it's better than NOT knowing the velocity, at least.

Loads with very small ES's and SD's point to that load being very consistent, which should mean that SOMETHING's going right with that load - a very nice sign when you get it.

It's just another tool to help us know what our loads are really doing, or at least a bit more about them. I'm keepin' mine.

I DID learn to buy the ones with the remote sky screens, though. Shot my initial Chrony, and went with the Pact I because of the remote sky screens. I make SURE there's plenty of clearance NOW, and friends don't shoot over it casually until I look over their setup - not a problem, since I always answer "Sure," when asked to use it, "As long as you'll get me another if you shoot it." Blackwater on Cast Boolits

"I have a Chrony set up in my tunnel so all it takes to measure the velocity is to turn it on which I almost always do. I have data on thousands of loads. Most of them were part of my effort to perfect a competition load for the 223. The best loads have horrendous variation in muzzle velocity (SD = 20 to 70 fps) and ranges for five shot groups up to over 200 fps. - although most are under 100).

I know the right load with less velocity variation should/might shoot better. The few I have tried so far didn't. I am still looking. I could probably go back and compare SD or range to five shot group size, or even best four out of five since I compulsively keep that too.

Since I'm not quite anal enough to keep data on individual shots, I can’t compare velocity to elevation but if there is a correlation it is hard to see. Time after time I have seen a very high velocity shot go low or a very low one go high. I'm not claiming there is an inverse correlation just that there is no obvious correlation with these loads and the rifle I am using. A more systematic look at where the individual shots hit compared to the muzzle velocity might show one.

These groups are at fifty yards and I don't doubt that velocity variation affects elevation at longer distances. The only evidence I have is my groups at 200 yards in last year's nationals. They had horizontal spreads that would have put me in the money but vertical spreads that were awful. Wind is supposed to be the problem at 200 for my tiny bullets, according to others. Right now vertical dispersion, which may or may not be because of velocity variation is more of a problem." John Alexander

Chronographs measure velocities, and calculate, among other things, the average and standard deviation of a string of shots. These calculations are precise, with one digit to the right of the decimal point at least. The chronograph manufacturers neglect to tell us three things, (maybe more):

• 1. Conditions during measurement must be the same as conditions of use.

• 2. The calculated average velocity is an estimate, with error a function of the number of shots and the calculated standard deviation.

• 3. The calculated standard deviation is an estimate, with error also a function of the number of shots and the calculated standard deviation.

Let's call the numbers calculated by the chronograph the STDEV, standard deviation and AVG, average. (We can also easily calculate the AVG and STDEV any set of numbers-like velocities-with EXCEL on the personal computer.)

Conditions

The AVG and STDEV of a set of, say, ten velocities, can be easily calculated. If there are ten guns involved, or one gun and five different loads, or two guns and four different bullets, or anything but one gun and one load; some folks might wonder what that calculated STDEV or AVG means. I'm one of those folks.

Any calculated statistic about a string of shots should be made with the same gun and the same load on the same day = same conditions of heat and humidity. We kind of know this intuitively. But what we don't intuit and what the chronograph manufacturers don't tell us is that for any calculated AVG or STDEV to mean anything, the shots whose velocities are measured must mirror the conditions under which actual shooting is done with that given gun and load.

My shooting/testing protocol at the Miami range is to shoot two sighters and ten shots for record in a fifteen minute relay. If I want to know something about the average or variation in velocities of a load shot under these conditions, then I have to shoot the ten record shots over the chronograph, and use their velocities in the calculation. And I need to shoot and record several sets of ten, to find the AVG and STDEV of the entire twenty or thirty or?? Shots.

     Varmint (and egg) shooters shoot a lot of single shots out of a cold clean barrel. To capture reasonable AVG and STDEV of these shots, the shooter needs to shoot and chronograph a lot of shots from a clean cold barrel. Shooting twenty shots in twenty minutes with that varmint (or egg) gun will allow calculation of an AVG and STDEV that is absolutely precise, and has not much to do with the AVG or STDEV of single shots from a cold clean gun.

So the first new thing we need to know about chronographs is: Chronographing conditions must match the anticipated shooting conditions for any calculation to have a connection with reality.

Averages

     The AVG of a string of shots, calculated by the chronograph, is an estimate of the population average, that we'll call MU.

If we want to know the average height of third-graders at St. Ignatius grade school, we can measure six of the ninety-four of them and calculate an average. We might not be confident in that average; which is like the AVG from the chronograph. We might want to measure all ninety-four third-graders, and calculate the average of the heights-that average is the true, real, underlying average and we call it MU. So in the chronographing world, AVG is an estimate (guess) of MU.

If we know AVG and STDEV of a string of shots, we can calculate the bounds of MU to any percent of sure. I use 95% sure. We find that Student's t distribution is very handy in this exercise, and do some figuring. The result is this table, that shows the +/- factors for various "n"s, where n is the number of shots fired in the string.

Let's say we fire 13 shots and get an AVG of 1200 fps and a STDEV of 20 fps. The table, next to n = 12, (number of shots -1) shows the +/- factor of .64. Multiply .64 times the 20 fps and get 12.8 fps. We can now say that we are 95% sure that MU is 1200 +/- 12.8 fps , or between (rounding) 1187 fps and 1213 fps.

So the second new thing we need to know about chronographs is: The AVG is an estimate of MU. The AVG and STDEV calculated by the chronograph coupled with the number of shots in the string can be manipulated to show the low and high bounds of MU.

Standard Deviations

The STDEV of the velocities of a string of shots is a measure of the variation of the velocities of those shots. Shooters believe that "The name of the game is the same" for accurate shooting-that there should be little or no variation in any of the reloading or shooting activities. Then we think it "good" if the STDEV of velocities is small. Elsewhere we discuss the correlation (or lack thereof) between the STDEV of velocities and accuracy.

     There is a standard deviation of the population (not a sample) of shots that could be made with any given gun, load, conditions combination; the true or real or underlying or hidden standard deviation.

An example: If we want to know the standard deviation of the weights of the people in Bemidji, we might go there and weigh the first ten people we meet and calculate the standard deviation of their weights. Probably some of us would hold that calculation suspect. We know that there is a true or real standard deviation of the weights of the people in Bemidji, and we know that to find that number we could weigh every citizen at the same time and calculate the true, real, hidden, underlying standard deviation-that we'll call SIGMA.

The STDEV is an estimate (guess) of SIGMA.

Any gun, load, conditions combination has a SIGMA, a real standard deviation; and  STDEV, an estimate of SIGMA. What your chronograph tells you is STDEV, an estimate of SIGMA.

If we know the number of shots in the string, and the STDEV, then through the magic of statistics we can say some things about SIGMA.

First is how sure we are about what we're going to say. "Sure" comes in percentages, we can be 90% sure or 95% sure or 37% sure, or any percent sure. I don't think that being much less than 90% sure is very sure, and far prefer to be 95% sure.

Next is what SIGMA is. We never know what SIGMA is exactly, we have to say that SIGMA is between this number and that number.

An example from the Sharps notebook: On 10/17/90 I was measuring velocities at 200 yards. I shot thirteen shots, recorded nine, with an average of 997.6 fps and STDEV of 12.4 fps. Taking the facts that there were nine shots and the STDEV was 12.4 fps, we dust off and employ the ever-popular "chi squared" distribution to find that we are 95% sure that SIGMA lies between a low of 8 fps and a high of 22 fps. Doesn't that 12.4 fps STDEV look extremely precise? It is, and it is also vague. Precise, but vague. What comes out of the chronograph is STDEV, an estimate of SIGMA.

So the third new thing we need to know about chronographs is: the STDEV calculated by the chronograph coupled with the number of shots in the string can be manipulated to show the low and high bounds of SIGMA.

 Because the calculations and the chi-squared distribution itself have been found by some to be stultifyingly boring (what's wrong with those people?); I have prepared a table of UCL (high) and LCL (low) bounds of SIGMA for some values of STDEV and the number of shots in the string.

The LCL and UCL bounds of SIGMA, the true, real, underlying standard deviation; for some values of STDEV and n = the number of shots in the string. See the big bold entries. With n = 10 shots in the string that have been chronographed, and STDEV from the chronograph = 12 fps, SIGMA lies between a LCL of 8 fps and a UCL of 22 fps. More correctly, if we chronograph a series of 10 shots and calculate the STDEV to be 12 fps, then we are 95% sure that SIGMA lies between 8 fps and 22 fps.

Another way to look at the relationship between the STDEV and SIGMA, for various numbers of shots

in the string, is this graph. SIGMA lies between the lower and upper lines on the graph. The lines are percentages of the STDEV calculated from the string of shots by the chronograph. To go back to the example above of a string of 10 shots with a calculated STDEV of 12 fps. The vertical line at n = 10 shots in the string cuts the bottom graph line at about 70%. Then 70% of 12 fps = 8.4 fps, rounded to 8 fps is the LCL bound. The 10 shot vertical line cuts the top graph line at about 180%, and 180% of 12 fps = 21.6 fps rounded to 22 fps is the UCL bound. We're 95% sure that if a string of 10 shots has a calculated STDEV of 12 fps, then the true, real SIGMA is between 8 fps and 22 fps.

More information about standard deviation estimates and a handy calculator for finding LCL and UCL is on the EXCEL workbook "CONFIDENCE INTERVAL ESTIMATOR OF SIGMA SQUARED" in the appendix.

   Here's a table showing the muzzle velocities of sets of twenty-five shots fired with each of two loads. One fouler was fired, then five shots each with loads A and B, alternating, in each fifteen minute relay. A,B,A,B,

The muzzle velocities of the record shots were recorded in EXCEL, then the STDEVs of the "last five" shots and all the shots "so far". EX: See the bold 7.5, this is the STDEV of shots 6-10, the "last five". Then see the bold 18.3, this is the STDEV of shots 1-10, all the shots "so far".

This real life example shows how calculated STDEV can vary.