Why shoot groups?

[7br]

While searching for the beagling article, I came across Joe B's article on statistics and the shooter. My question is why not shoot a set numbers of shots, measure the right-left, up-down deviation from point of aim, figure the SD and let it go at that?

[Haywire Haywood]

A set number of shots is a group, no? I'm confused.

Ian

[danski26]

standard deviation

(Statistics) a measure of how spread out data values are around the mean, defined as the square root of the variance.

There are some problems with your proposed experiment. First how do you find your point of aim (mean)? You can't just sight in so you shoot a group "around" a point of aim and call in the "mean" value. I guess you could try with some type of laser aiming device but i would guess for the accuracy neaded the cost would be prohibitive.

Also taking the square of the deviance "group" around your mean value "POI" would tell you what? As shooters we are worried what the farthest outside of a "POI" a shot will ring. "SD" will not tell us that.

If you could set up the experiment what would you be hoping to glean from it?

If you are interested in some statistical applications to load development I can put you onto a few articles..........just have to find them in my messy bookshelfs for the web site address.

[mag_01]

You don't have to shoot groups ---If you so desire set a row of 1 in. dots and fire a shot at each say 5 or 10 dots . Then you will have an Idea of how the (gun) with that (ammo) is performing. ---- Mag

[7br]

standard deviation

(Statistics) a measure of how spread out data values are around the mean, defined as the square root of the variance.

There are some problems with your proposed experiment. First how do you find your point of aim (mean)? You can't just sight in so you shoot a group "around" a point of aim and call in the "mean" value. I guess you could try with some type of laser aiming device but i would guess for the accuracy neaded the cost would be prohibitive.

Also taking the square of the deviance "group" around your mean value "POI" would tell you what? As shooters we are worried what the farthest outside of a "POI" a shot will ring. "SD" will not tell us that.

If you could set up the experiment what would you be hoping to glean from it?

If you are interested in some statistical applications to load development I can put you onto a few articles..........just have to find them in my messy bookshelfs for the web site address.

Actually, point of aim really doesn't have much to do with it. What I am thinking about is the center of the group will be the mean. We should expect that 99.9 percent of our shoots should fall within 3 SD on either side of the mean. If we have a .3" SD, I would expect 99 out of 100 shots to fall within a 2" group. I need to go back and read Joe's article and see what number of shots a person would need to get good data.

[Hip's Ax]

Last couple of years with the advent of the 6.5/284 round in 1000 yard shooting guys have been doing "the ladder test" instead of shooting groups. As the barrels chambered in this caliber wash out in less than 1000 rounds every shot counts. What they do is start low, make each round 0.2 grains apart up to max. They shoot them in order at the target and due to the increasing muzzle velocity the shots "climb up" the target. Where the shots bunch up are the sweet spots, a little more shooting within the sweet spots and their ammo testing is done for way way fewer rounds than shooting groups would have taken.

[danski26]

Center of the group may not be the PIO there for not the "mean" value. Not too many approximates in statistics.

Again, what are you trying to learn by this experiment?

Your going to figure a SD "approximatley" to figure a avarage group size?????

Would you also take an average weight of an apple, then divide it by the total weight of a plate of apples, minus the avarage wieght of the plate to find out how many apples you have or would you just count five apples?

My point is what advantage does finding the SD give you if you want to find group size? Just measure the group size.

As mention above ladder testing is one way to fine tune a load with less rounds down range.

[danski26]

Ladder testing is not a group size test but a test of the relation of powder charge to the impact elevation on target. A useful tool none the less.

[Larry Gibson]

Ladder testing is not a group size test but a test of the relation of powder charge to the impact elevation on target. A useful tool none the less.

I tried the ladder test with several cartridges and found it only usefull at the range tested.

Larry Gibson

[danski26]

Most load testing is only good at the range tested. Just because you have a good grouping load at 100 yards does not mean you have a good grouping load at 600 yards. You would think at shorter ranges (100 to 200 maybe 300) the info may be usefull. Larry, you found that not so?

[7br]

My thoughts are that it might be a better indication of accuracy especially for load testing. I shoot smallbore pistol silhouettes and not groups. If all 40 rounds of a match were shot at 100 yards, it could very well be that a person with a smaller average group size could lose to a person with a smaller SD. For instance, I could have 7 groups of 1/2" and the last group of 5". IE, just over 1" average for all 40 shots. However, since the last five were at the edges, I missed two high and three low. The shooter beside me shoots 8 groups each at 3inches and never misses.

Maybe I am thinking about it wrong.

[felix]

Shot placement is most important in all shooting sports except in BR where the smallest group counts no matter where the placement. In other words, the guy who knocks down all the animals is the winner, no matter what the group size, or for that matter, he can care less about the placement as long as the animal goes down. Nobody cares what the velocity is either: good, bad or indifferent. ... felix

[44man]

I made one boolit mold for my BPCR that shoots as small as 1" groups at 200 yd's with extremely low SD and ES. Beyond 200, it starts to go crazy and I can barely hit any turkeys or rams. The SD's dont tell you a thing and has nothing to do with grouping ability. Shooting the smallest groups at the ranges you shoot is the only way to tell if everything is right. The best way to learn a gun or load is to leave the chronograph and math equations home.

[MT Gianni]

Groups of known loads let me know how I am shooting. Tighter groups help mr to eliminate fliers and improve mental concentration. We watched a military match at the Knob Cr. shoot last week, You would see shooters hit the first 3 target plates then miss 4 not because there was anything different than mental conditions. You start to think I am going to clean this station and you forget the front sight isnt on target or the trigger pull goes aray. Groups allow me to call my fliers and concentrate on the same target sighting. Gianni

[joeb33050]

While searching for the beagling article, I came across Joe B's article on statistics and the shooter. My question is why not shoot a set numbers of shots, measure the right-left, up-down deviation from point of aim, figure the SD and let it go at that?

This is an alternative method of measuring accuracy, the measure is called the "mean radius". Ed Wosika was/is very interested and knowledgasble about this.
The disadvantages are that it is difficult to measure a set of shots this way, (although there is software available that lets you put the target in a scanner and get the mean radius),and that the result isn't easily understood because we don't use this measure much. The advantage is that every shot counts, where in a 5 shot group only 2 shots count-this is pretty much an imaginary advantage.
I believe that Joroen Hogema has pretty well defined that relationship between mean radius and group size, and demonstrated that group size is a perfectly acceptable measure.
The old timers measured accuracy with "string measure", distance from the center o0f the target to the center of the bullet hole summed. Then a string measure of 10 inches for a 5 shot group means that on average, each shot was 2 inches from the center. I've always believed that this is maybe the "best" measure of accuracy, but with today's shooting it is impossible to use because the groups are so small thay measurement is impossible.
joe brennan

[joeb33050]

Ladder testing is not a group size test but a test of the relation of powder charge to the impact elevation on target. A useful tool none the less.

Creighton Audette invented ladder testing which is used to find the "best" load within a fairly narrow range of variable powder quantities. It gets you close, quick.
THEN you must select from a few, maybe 2 or 3, powder quantities; to find the "best" load.
We talk about this testing in the book; also see Jesse Miller's article about load development that describes an alternative to the ladder test.
joe brennan

[joeb33050]

My thoughts are that it might be a better indication of accuracy especially for load testing. I shoot smallbore pistol silhouettes and not groups. If all 40 rounds of a match were shot at 100 yards, it could very well be that a person with a smaller average group size could lose to a person with a smaller SD. For instance, I could have 7 groups of 1/2" and the last group of 5". IE, just over 1" average for all 40 shots. However, since the last five were at the edges, I missed two high and three low. The shooter beside me shoots 8 groups each at 3inches and never misses.

Maybe I am thinking about it wrong.

We found, using two methods, Monte Carlo simulation and analysis of the data and ranges, that sets of groups have standard deviations that are fairly consistent. Forgetting the arithmetic for a moment, if sets of five 5 shot groups are shot, then on average the smallest group of the five will be half as big as the largest. That's on the average, and we've got a lot of data to demonstrate this.
So in rough terms, if a rifle averages 1" five shot groups, then sets of five will have smallest group averages around .5" and largest group averages around 1.5".
This means two things to me, first that the problem described above isn'r likely to happen, and second that virtually any rifle can shoot a 1/4" group if you shoot enough groups.
joe brennan

[joeb33050]

Shot placement is most important in all shooting sports except in BR where the smallest group counts no matter where the placement. In other words, the guy who knocks down all the animals is the winner, no matter what the group size, or for that matter, he can care less about the placement as long as the animal goes down. Nobody cares what the velocity is either: good, bad or indifferent. ... felix

Not quite right, Felix. In BR shooting for group we sight the gun to NOT hit the aiming point, otherwise the aiming point would soon disappear. I have not had a problem with this. And, in BR shooting, both modern BR and CBA and ASSRA, there are SCORE matches where the object is to hit the aiming circle. In ASSRA 100 yard 10 shot matches, five shots are fired at each of two bulls.
joe brennan

[joeb33050]

I made one boolit mold for my BPCR that shoots as small as 1" groups at 200 yd's with extremely low SD and ES. Beyond 200, it starts to go crazy and I can barely hit any turkeys or rams. The SD's dont tell you a thing and has nothing to do with grouping ability. Shooting the smallest groups at the ranges you shoot is the only way to tell if everything is right. The best way to learn a gun or load is to leave the chronograph and math equations home.

44Man-you're talking about SD and ES of VELOCITY not being good predictors of accuracy-the SD being talked about here is the standard deviation of distance of shots from some center. My position is that MV and SD of MV are not, in my experience, good indicators of accuracy. In the book you'll see my opinion, then other opinions from folks who love that chronograph and use it to develop loads. A mystery to me. However, see Shooting Times, May 2007, pg. 22 for an article that lends support to your (and my) position.
joe brennan

[felix]

Not soon disappear, Joe, but immediately with a 16X or mo'betta' at a hunnert. ... felix

[S.R.Custom]

I tried the ladder test with several cartridges and found it only usefull at the range tested.

If this is a weapon that will get used at several different ranges, i.e. a hunting rifle, then shoot three ladders at say, 100, 200, and 300 yards. Your "nodes," as you point out, will vary some at the different ranges. But likely as not, you will find some commonality, and this willl be "the" load. All in all, a much more efficient means of load evaluation than just random group shooting, because you can actually watch the trend line being drawn on your target...

[Larry Gibson]

If this is a weapon that will get used at several different ranges, i.e. a hunting rifle, then shoot three ladders at say, 100, 200, and 300 yards. Your "nodes," as you point out, will vary some at the different ranges. But likely as not, you will find some commonality, and this willl be "the" load. All in all, a much more efficient means of load evaluation than just random group shooting, because you can actually watch the trend line being drawn on your target...

That may work, I've not tried it. But for a general load to use hunting, varmint shooting or for long range work what I found with the ladder test that the loads in the "node" group (if there was one - about half the cartridge rifles I tested did not develop a "node" group) was the very large extreme spread of velocity. While, due to barrel harmonics, those two loads may hit close together at a specific range at other ranges, specifically long range those two loads are far apart.

I prefer groups for testing and zeroing. Also I test for minimal extreme spread of velocity. After all you are shooting into a cone of fire (the bullets do not all go in the same hole now do they?) whether you shoot one shot, three shots or a string of 10 shots. When you hold off or adjust your sights you are adjusting for the center of the cone of fire (the center of the group) and not really adjusting for the shot. This is so even with an extremely accurate rifle.

Larry Gibson.

[trk]

Skip the math and statistics (they're ok for what they're worth - I've had 6 or so graduate courses in stat.).

I shoot 5 or 6 five shot groups. Some will be small some will be larger. When I get so I can shoot 5 or 6 five-shot groups that are right-much small I know this:

I can do it. It's repeatable. It's not a fluke.

If the groups are not small enough, I change something to improve the group size.

If I can shoot a 3" set of groups at 100 yards the woodchucks are going to suffer.

If I can shoot (standing) a 6" set of groups at 100 yards that's good enough for deer.

Other standards apply as well. Like hitting 10 pennies in a row at 100 yards. Then one moves to asperins glued to black construction paper.