.380 ACP the Red Haired Stepchild?

[JohnKSa]

Firstly, I don't think you'll find any trace of me ever suggesting that all 10mm loads are better than all 9mm loads.

Dodging the point.

1. I didn't say that.

2. By your own admission it was a best vs. best comparison.

I have however made generalization based on the capabilities of a cartridge from the perspective of MAP, case volume, and projectile diameter.

So the two calibers at their best can't be compared because that's an anomaly, but we can rely on generalizations to tell us the truth? This only makes sense if you really, really want it to. It's not the logical conclusion.

Many .40S&W cartridges also come in ahead of the 10mm cartridges in the data sets, even at similar grain weights.

Which is exactly what we would expect if there are contributors that have nothing to do with terminal effects due to caliber difference that are dominating the outcome. It's telling that this is showing up in the results even after the data has been culled to try to emphasize any existing differences due to terminal effects.

If the effect is significant, it should be easy to prove it's there--because that's exactly what significant means.

Go back and read Urey Patrick's 1987 paper. Even while trying to argue that caliber difference was important, he admits that the effect is so small it will take very large numbers of shootings to make it show up. That was decades ago. In all the time since that paper (and we've certainly had very large numbers of shootings to work with) no one has come up with something that consistently shows a significant effect in real world shootings that fits the energy, momentum or bigger projectile theories.

I'm not trying to prove or disprove anything. I just want someone to show me some data that makes sense and supports the theory they believe in. If no one can do that (and no one has to date) then why should I assume that there's some significant effect I need to be concerned about?

 

[WrongHanded]

So the best vs. the best is an anomaly, but we can rely on generalizations to tell us the truth? This only makes sense if you really, really want it to. It's not the logical conclusion.Which is exactly what we would expect if there are contributors that have nothing to do with terminal effects due to caliber difference that are dominating the outcome. It's telling that this is showing up in the results even after the data has been culled to try to emphasize any existing differences due to terminal effects.

If the effect is significant, it should be easy to prove it's there--because that's exactly what significant means.

It was a best vs best of the data available that fit the criteria in LE shootings. Not the best 10mm ammo that existed vs the best 9mm ammo that existed. I have no idea what development did or did not go into those cartridges, and infact I do not have data of their velocities. Those would be important variables.

But among the cartridges running at 35,000 psi, such as .357 Magnum and .40S&W, we can see some correlations in the data sets. Lighter, and therefore presumably faster bullets dominant the top of the charts, and heavier and therefore presumably slower bullets occupy more of the bottom. This is also true for 9mm, though the +p+ loads tend to be at the top with lighter and middle weight bullets, and standard pressure loads and heavier bullets at the bottom. This seems true also for the 10mm and the limited data sets for .357SIG. It is also true to a slightly lesser degree for the .38spl, where +p and +p+ loads are at the top, standard pressure heavier loads towards the bottom.

Does that tell us anything? Maybe it's just anomalies, right.

 

[JohnKSa]

You do see the problem with drawing conclusions from a data set that has been culled to try to emphasize the particular claim it is designed to support (i.e. eliminate data that doesn't support the claim), right?

The fact that it still shows significant anomalies even after being massaged is even worse, in a twisted sort of way.

I went through all that. Courtney, M&S, Fackler, gelatin, statistics, energy, momentum, made up formulas like TKO, etc. What it finally convinced me of was that none of it was able to demonstrate that terminal performance differences within a general handgun performance class had a significant effect on the outcome of real world shootings. Some of it was contrary to physics, some of it just didn't make sense, some of it couldn't prove what it was supposed to prove, some of it depended on having to throw out a lot of real world data before it "worked".

I finally realized that I had been going about it all wrong. I had assumed the presence of a significant effect and was trying to find a way to prove it existed. Things got a lot simpler when I accepted the idea that there might not be a significant effect and stopped assuming it existed.

 

[WrongHanded]

You do see the problem with drawing conclusions from a data set that has been culled to try to emphasize the particular claim it is designed to support (i.e. eliminate data that doesn't support the claim), right?

That's a poor way to describe removing uncontrollable variables. If one decided to do live animal testing, one would not shoot some animals in the data set multiple times and others once. Or shoot some animals in the head and others elsewhere. If you accidentally shot an animal more than once, or shot an animal outside of the predefined target zone, you would be wise to discount those results. With real world data, the best one can do is begin with a set of carefully thought out and well reasoned criteria and hold to it.

So let me ask you, what criteria makes more sense when trying to determine the differences (if any) between terminal performance of various handgun loads? Please describe a superior set of criteria for attempting to evaluate real world data.

Edited to add:

Look at it like this. If I gave you 20 different 9mm handgun loads to test (as many of each cartridge as you wanted), and as many 180lb pigs as you needed, what would your test parameters be?

 

[jar]

For me, the major questions in choosing my self defense selections are:

1. Do I have the ammo already?
2. Do I have a handgun for that ammo already?
3. Do I have a holster that will hold that handgun securely and be comfortable to wear all day?
   
4. Can I shoot that handgun with that ammo accurately, repeatedly and quickly?
5. Am I physically able on that day to reliably operate that selection?

There may well be better choices available but any such better choices will need to wait until the first two items are available. The actual best selection varies from day to day based on weather, my physical shape on that particular day and what my plans are. A "NO" response to any one of the questions just means making a different selection; but still from what is already available.

Some days a full size 1911 will be the choice while on some other days I will choose my Ruger LCP II Lite Rack 22LR. But regardless, the selection will be what I have available and am confident of being physically able on that day to reliably operate that selection.

While I have quite a few very small .380 handguns, in almost every situation I end up selecting a larger .380 handgun over a small choice.My smaller .380s; a S&W Bodyguard, Remington RM380, Sig P230 are far harder to shoot accurately and repeatedly and so if shooting them I tend to shoot considerably slower than if I am shooting a S&W 380EZ or Ruger Security 380 or Walther CCP M2+ 380.

The reality is I will carry the ammo and handgun I own rather than that absolute best choice I don't own.

The reality is I will carry the ammo and handgun I own in a holster I own and that will be comfortable all day rather than something that I find uncomfortable or that I cannot reliable use that particular day.

 

[JohnKSa]

Look at it like this. If I gave you 20 different 9mm handgun loads to test (as many of each cartridge as you wanted), and as many 180lb pigs as you needed, what would your test parameters be?

I wouldn't do the test because I can find out all I need to from gelatin testing and from looking at real world shooting data. I would eliminate any of the loads that don't penetrate adequately and consistently or that don't consistently and adequately expand. My criteria for selecting from that subset of the ammo would be based on how accurate and reliable they are in my carry/self-defense gun, availability, muzzle flash, etc.

That's a poor way to describe removing uncontrollable variables.

If the goal is to see how different loads perform in real life, then you don't throw away data that shows how different loads perform in real life.

So let me ask you, what criteria makes more sense when trying to determine the differences (if any) between terminal performance of various handgun loads? Please describe a superior set of criteria for attempting to evaluate real world data.

Just look at the real world data. All of it. That's what you really want to know about. Do you really want to know only what happens to attackers who are shot only shot once, or only about attackers who are shot only in a particular part of their body? No, of course not.

Stop rationalizing throwing away real world data. Then accept what the data is telling you even if it's not what you expect or want. The story it's telling is very clear.

 

[WrongHanded]

If the goal is to see how different loads perform in real life, then you don't throw away data that shows how different loads perform in real life.

Bullets go down range, and either hit or miss, one at a time. Not in pairs, not 5 at a time. Shooting fast is possible, but to compare multiple hits is absolutely impractical. When the goal is to determine cartridge efficacy, it must be done per shot. Which means data including two or more hits must be excluded. This type of exclusion happens in science all the time. If you can't fathom that, then I understand why you have come to the conclusions you have.

The reason I suggested thinking about how you would do your own testing, is with the hope that you would come to the conclusion that multiple hits on a single target cause huge problems with defining results from data. Of all the variables in real world shooting that you are bringing up, multiple hits would be one of the hardest to negotiate. It really is that simple. So on the one hand, you complain about variables, on the other you want to introduce more variables that muddy the waters far worse.

 

[trackskippy]

Except maybe for crunching numbers, fast multiple hits, and preferably, good hits, "is" the answer.

I dont know how you intend to shoot, but rounds do, or should go down range in multiples.

And I know, I know, Im screwing up the argument by introducing reality, so Ill shut up. Carry on.

 

[WrongHanded]

Except maybe for crunching numbers, fast multiple hits, and preferably, good hits, "is" the answer.

I dont know how you intend to shoot, but rounds do, or should go down range in multiples.

And I know, I know, Im screwing up the argument by introducing reality, so Ill shut up. Carry on.

That's not the point. The point is you can't judge one cartridge against another when the first one has 5 hits and the second one only has 2.

 

[trackskippy]

The point is, it doesnt matter, as you cant judge anything on something that isnt repeatable, comparable, and verifiable. Shooting people or critters, isnt any of that, as each and every instance is its own critter.

Shooting something like gel kinda sorta is, as its something that is repeatable and can comparably show you what one vs the other does, in it. It has nothing to do with shooting people or critters.

Soooo..... the only way to make "any" of them work, is to shoot whatever it is, to the ground, and continue to do that, until they are done, and that will take "exactly" as many rounds of any kind of bullet you might have, as it takes, and the faster you do that, most likely the better for you. (ie, shoot first, shoot fast, shoot often.)

All the arguing over numbers and guesses is just useless and wasted time. No matter what it is you chose, it will still have to be employed as I described above. If youre REALLY lucky, that "might" take one round, if youre not, it "might" take everything you brought along.

If you run out of ammo, you "might" be able to whip out your data, and bore them to death with your numbers and reasons why they "should" be dead. I wouldnt suggest that though, as youll probably lose the civil case for the use of extremely cruel and inhuman punishment. I know right now, after 10 pages, I want to sue a couple of you for it.

 

[jar]

The point is, it doesnt matter, as you cant judge anything on something that isnt repeatable, comparable, and verifiable. Shooting people or critters, isnt any of that, as each and every instance is its own critter.

Shooting something like gel kinda sorta is, as its something that is repeatable and can comparably show you what one vs the other does, in it. It has nothing to do with shooting people or critters.

Soooo..... the only way to make "any" of them work, is to shoot whatever it is, to the ground, and continue to do that, until they are done, and that will take "exactly" as many rounds of any kind of bullet you might have, as it takes, and the faster you do that, most likely the better for you. (ie, shoot first, shoot fast, shoot often.)

All the arguing over numbers and guesses is just useless and wasted time. No matter what it is you chose, it will still have to be employed as I described above. If youre REALLY lucky, that "might" take one round, if youre not, it "might" take everything you brought along.

If you run out of ammo, you "might" be able to whip out your data, and bore them to death with your numbers and reasons why they "should" be dead. I wouldnt suggest that though, as youll probably lose the civil case for the use of extremely cruel and inhuman punishment. I know right now, after 10 pages, I want to sue a couple of you for it.

Went hunting.

Took two shots at a rabbit.

One missed two feet behind, the other two feet ahead.

Rabbit did the math and fell down dead.

 

[trackskippy]

LOL. I shot a rabid, or what appeared to be rabid anyway, it looked like hell and was wandering around in the middle of the day and acting all shaky. First 45acp Hydra Shok round from my Commander just bowled it over and it came right back up on its feet, hissed a little bit, and continued waddling along. A second round to the head, which I was trying to do in the first place, killed it. What can I say, the adrenaline was spiking and I flubbed the shot!

Now, Im sure there is more than one person around here whos read all the data, and will tell you that a 45 Hydra Shok will drop a full grown man with a one shot stop, somewhere in the high percentile, yet it didnt even phase a couple of pound raccoon.

The only guarantees here are, there are no guarantees. No matter what someones statistics might say, EVERY time is a whole new thing, and might as well be the first time. If you're lucky, the stars align, the gun works, the round works, your aim is perfect, and the target is impressed enough to fall over, grave yard dead.

Or maybe it just waddles off, or worse.

 

[WheelGunMan]

This thread has gone totally off the rails.

 

[Leon tP]

After 10 pages of argument, I'm still waiting for someone to answer the basic question ask,
".380 ACP the Red Haired Stepchild?"

Hell, I was thinking maybe I'll just start talking about a Zombie Apocalypse and get this thread locked.

 

[JohnKSa]

When the goal is to determine cartridge efficacy, it must be done per shot.

I understand EXACTLY what you are saying. Trust me. I get it. What I'm saying is that in order to look at only cases where ONE shot was involved one must discard a LOT of real world data.

Now, if one wants to know what happens in cases where a person is shot only once, and only in the torso, then that might start to make some sense. IF we...

• Ignore that the torso encompasses a fairly large area with a fairly large range of organs which have a very significant range of immediate necessity to the person who owns the torso.
• Ignore that hitting a torso from different angles dramatically affects the chances of hitting multiple organs or maybe missing everything important.
• Ignore that different people respond differently to pain.
• Ignore that chemical impairment, mental attitude, and physical conditioning can dramatically alter the physical and mental response to injury.
• Ignore that the issue of whether bone is hit is nearly random and can certainly affect the outcome.
• Ignore the fact that even when we restrict things only to torso shots, we still don't exclude all of the CNS (the spine is in the torso) which tends to react the same to a hit regardless of caliber.
• Ignore that psychological stops are by far the most common method of "incapacitation" and outnumber all other kinds of stops by 4 to 1 and have virtually nothing to do with where the target is hit at all.

But unfortunately all of those things are true. So now it turns out that we didn't really isolate the variable we cared about at all. In fact, we didn't even come close to the goal. All we did was throw away a lot of real world data.

As SOON as you can figure out a way to REALLY isolate terminal effect due to caliber in shootings, I'm going to be right there looking at the data. But this "pseodo-isolation" doesn't begin to accomplish the goal. What it does is give the illusion of validity while actually obscuring reality by excluding actual data that can provide useful insight.

...hope that you would come to the conclusion that multiple hits on a single target cause huge problems with defining results from data.

Not if you just accept the data as it stands. If you just look at the data, what it's telling you is very clear. It's only if you refuse to accept what it says and feel like you need to look for something else that it gets complicated. Just look at the data. Don't throw out the data that you think is problematic. Don't only focus on the part that you feel might show the effect you want to find. Look at ALL of it and just go with the results.

So on the one hand, you complain about variables, on the other you want to introduce more variables that muddy the waters far worse.

That's an absolutely backwards way of looking at it. I didn't introduce any variables at all!

Those variables are all already there in the data and they are combining to produce the overall effect. I didn't put them there. You believe that it's possible to "unsum" the data and get rid of the variables that you don't want. That's simply not reality. In addition to the one or two variables the particular strategy for throwing away real data you choose to endorse might minimize slightly, there are many others that aren't minimized at all and can still have huge and essentially random effects on the outcome.

The water IS muddy, but I didn't make it muddy. The variables are already there--I didn't introduce them. What I'm telling you is that even though the water is muddy and all the variables are inextricably wrapped into the data, it's still possible to look at the results and draw very useful conclusions.

If I gave you 20 different 9mm handgun loads to test (as many of each cartridge as you wanted), and as many 180lb pigs as you needed, what would your test parameters be?

Ok, if you really wanted to do this, here's what it would take.

Just having them be the same weight would not be enough--shooting a bunch of pigs that are all the same weight would just be a waste of pigs and ammo.

They would all need to be:

• Very similar in weight.
• In very similar physical condition.
• The same sex and age.
  
• Very similar body fat percentage.
• In a similar state of arousal/calm at the moment of shooting.
• Xrayed or otherwise scanned to verify that their internal anatomy was sufficiently similar. Internal organs can vary in size and even in arrangement in different individuals. Ribs can vary in size and thickness, etc.

Each shot would need to be taken with the pigs precisely positioned and with a precise aim point and shot angle to keep things as close to identical as possible for each shot. As a starting point, each aim point/shot angle/loading would need to be repeated at least 10x, each time on a different pig to try to eliminate individual variability and slight differences due to aiming/angle errors. That might need to be increased as we get into the experiment, based on how the results start to play out, but I think 10x would be a good starting point.

One would want to use a number of aiming points, I'm thinking a minimum of 12 "frontal" shots (4 midline, going down the body, 4 on each side of the front going down the body) and at least 4 from each side.

For each aiming point, you would need at least 5 shot angles. Straight in, angled up/down/right/left at say, 30 degrees.

So for each loading, we would need to shoot (12+4+4) x 5 x 10 = 1000 pigs. We might have to increase the number of repetitions later if we see that we're still not isolating the variable enough and need to increase the number of shots per aim point/shot angle/loading.

For 20 loadings, we're talking about 20,000 nearly identical pigs shot under very tightly controlled conditions. I think that would have a chance of showing loading differences.

Now, when we finished, what would we have? Would we have something that told us how likely one round was to be effective on the street compared to another one? No, of course not--because all the variables we strove to eliminate in the careful testing still exist on the street. We would have managed to prove what we already knew to be true. That, if very carefully isolated, terminal performance differences can affect incapacitation times. But in the real world, chaos still prevails and it dominates the outcomes, not the puny terminal effect differences due to caliber selection that can only be seen under artificially controlled conditions.

How do we find out what happens in the real world? Look at the real world data. Which is convenient (and fortunate for the pigs) because we already have it. It will tell us that what really makes the difference in the real world doesn't come in a 20 round box or in the form of a number stamped on the side of a gun.

 

[Elmer]

It's disconcerting that people still quote "data", discredited as falsified years ago. But as I said, many seem to enjoy fiction.

 

[Elmer]

This thread has gone totally off the rails.

After 10 pages of argument, I'm still waiting for someone to answer the basic question ask,
".380 ACP the Red Haired Stepchild?"

Hell, I was thinking maybe I'll just start talking about a Zombie Apocalypse and get this thread locked.

Yes, the discussion of how cartridges and projectiles can be vetted for use in self defense, and the science, or lack of, behind making those choices, has absolutely nothing to do with the question.....

Wait.... "Ask anyone if they want to get shot with a .380"!!!

Or, "I know a guy that knows a guy that got shot 11 times with a .380, and he finished his Taco bell meal and then drove himself to the hospital"!!

Better?

 

[amd6547]

Some of us like redheads.

 

[JohnKSa]

It's disconcerting that people still quote "data", discredited as falsified years ago.

I take a somewhat different view. I don't think M&S falsified their data at all. I think that they made mistakes in terms of throwing away data that they should have kept and in how they analyzed the data because they didn't understand the problem well enough to take on the project, and I think their basic starting premise was problematic, but, at least from my perspective, I don't see anything wrong with their motives or ethics.

 

[dogtown tom]

It's disconcerting that people still quote "data", discredited as falsified years ago. But as I said, many seem to enjoy fiction.

But, but, but...........they used LIVE goats in the Strasbourg Goat Tests. Those goats died. How can that be falsified?

 

[Elmer]

But, but, but...........they used LIVE goats in the Strasbourg Goat Tests. Those goats died. How can that be falsified?

Everyone knows it's that ballistic pressure wave that does all the damage......

 

[trackskippy]

Well, the people in the movies just all fall down dead at one shot. And there are more people who have learned about getting shot and what to expect from that than the goats.

 

[Elmer]

I take a somewhat different view. I don't think M&S falsified their data at all. I think that they made mistakes in terms of throwing away data that they should have kept and in how they analyzed the data because they didn't understand the problem well enough to take on the project, and I think their basic starting premise was problematic, but, at least from my perspective, I don't think see anything wrong with their motives or ethics.

Well, lacking knowledge of current copyrights, I'm reluctant to post any of my copies of IWBA or AFTE journals, but you might look and see if any websites still link to any of these for further study.


Eugene Wolberg, Gary Roberts. "Book Review, Handgun Stopping Power: The Definitive Study." AFTE Journal 1992

Duncan MacPherson. "Sanow Strikes (Out) Again." IWBA Journal 1997.

Martin Fackler, MD. "Book Review, Street Stoppers: The Latest Handgun Stopping Power Street Results." IWBA Journal 1997.

Duncan MacPherson. "The Marshall & Sanow 'Data' - Statistical Analysis Tells the Ugly Story." IWBA Journal 1999.

Maarten van Maanen, "Discrepancies in the Marshall & Sanow 'Data Base': An Evaluation Over Time." IWBA Journal 1999.

Martin Fackler, MD. "Undeniable Evidence" IWBA Journal 1999.

 

[JohnKSa]

Yup. I was aware of those during the timeframe they were actually being published. IMO, M&S were never trying to mislead anyone, they just got in way over their heads and at the same time poked a hornet's nest. Their lack of experience and background in statistics and data collection and analysis made them easy pickings for folks who knew what they were doing and took the time to carefully look at their data. It's very difficult to cull a data set and be sure that you don't really screw up your statistics and they just didn't know what they were doing so they didn't realize how badly they messed things up. That and it appears that they got sloppy with their data management and let some really stupid errors creep into their data.

If they HAD known what they were doing and intended to falsify their results, they could have done a much better job. For one thing, they would have hired someone with a statistical background to make things look right.

Here are some links that have some of the articles you mention.

http://thinlineweapons.com/IWBA/1999-Vol4No2.pdf
http://thinlineweapons.com/IWBA/1997-Vol3No1.pdf

If you read through that stuff, and know what's what, it's pretty obvious that the authors are bent on crucifying M&S by whatever means they can. For example, in one place, MacPherson berates a curve fit formula someone came up with based on the M&S data set based on the fact that it produced a negative number for input values that "logically" should have produced a very high value. But he knows very well that when you do a curve fit that the fit is only useful in the range of the data that is fit. Outside that range, a curve fit formula often produces absolutely ridiculous values. That's how curve fits work.

Now, I'm not defending the formula or the M&S data, because both are problematic. What I'm saying is that MacPherson's criticism was not really above board. Even if the M&S data had been 100% valid and the curve fit formula perfectly represented it, there would still be every reason to expect that outside of the fit range the formula would provide useless, ridiculous, totally meaningless results. MacPherson knew this was true (either that or he's a massive fraud) and yet he went ahead and published the criticism as if it meant something.

 

[WrongHanded]

Well @JohnKSa I don't entirely agree with you. But I think I understand where you're coming from.

Personally, I see some things in the M&S data sets that make perfect sense. I also see some correlations that I do not believe could be easily thrown off by the uncontrollable real world variables you have mentioned. I don't take away any sharp sense that "this round is superior all others". But I certainly see value in the data as it stands, knowing that it is imperfect.