RELATIVE stopping power, one-shot-stop

Status
Not open for further replies.

Dr_2_B

Member
Joined
Jan 4, 2006
Messages
1,850
Location
midwest
I've seen many people debunk the Marshall and Sanow One-shot-stop statistics. But I think some people miss the salient point with their work. I agree that no handgun can be expected to have a 94% chance of stopping any assailant with one shot somewhere to the body. But I must say the Marshall and Sanow findings are extremely useful because they numerically compare effectiveness of different rounds in real-world encounters. (I realize there are valid questions about the statistics they used, that's for another discussion)

It should not be discounted that the relative effectiveness of the different calibers is entirely pertinent to the discussion of self defense. What I mean is this: Marshall and Sanow's findings show 90% effectiveness for one caliber and 40% effectiveness for another smaller caliber from real-world encounters. For me, it's simple. I want the 90% caliber. Does this mean I can expect 9 out of 10 assailants to fall with one shot? No. Does it mean this round is likely MUCH more effective than the smaller one? Yes.

In my psychotherapy practice, I know that certain techniques work well with, say, abuse victims. I don't care whether it's 90% effective or 80% effective. I'm going to choose said treatment over one that shows 50% effectiveness with such a client.

I believe Marshall and Sanow has great significance as long as it's used properly.

My $.02.
 
I'm with you matheath. I use to, a long time ago, put more emphasis on their statistics then realized there were flaws but still some validity if you apply the relatively effective principle to what they have compiled. For example, I do believe a 125 grain hollowpoint .357 is more effective than a 125 grain hollowpoint .38 and a good 9mm is more effective than a .380.
 
All that "research" can be wrapped up in two sentences.

1. A decent hit with a bigger faster bullet is probably going to stop a person faster than the same hit with a smaller slower bullet.

2. Sentence one isn't always correct.:)
 
There may be flaws, but those flaws exist equally for each caliber. Though the results may not be of great use in predicting the capabilities of any one load, the fact that they are comparing apples to apples makes them useful for comparing one caliber/load to another. If load "A" is listed as a 75% stopper and load "B" is listed as 50% we can not assume that the percentages are always going to hold up, but I think it can be reasoned that load "A" will be 50% better than load "B" regardless of the actual percentages.
 
Hey SixForSure, I'm with you, but I'm not sure I can even go so far as to say load A would be 50% better than load B. Your math is good, but I'm not sure I have that much confidence in the statistical gathering or, frankly, the veracity of the real world reports. I suppose I'd have to say load A is appreciably better, probably.

That said, give me load A, baby!
 
Understood. To say statistical probability is not my strong suit would be an understatement. I was pointing out to the nay sayers that the data is not useless as some claim. There are so many variables in any of this that nothing is assured. It needs to be considered in context with all factors. In considering any stoping data or ballistics data we need to remember the three most critical factors in stopping power:
1. shot placement
2. shot placement
3. the ever so critical shot placement
 
I've seen many people debunk the Marshall and Sanow One-shot-stop statistics. But I think some people miss the salient point with their work. I agree that no handgun can be expected to have a 94% chance of stopping any assailant with one shot somewhere to the body. But I must say the Marshall and Sanow findings are extremely useful because they numerically compare effectiveness of different rounds in real-world encounters. (I realize there are valid questions about the statistics they used, that's for another discussion)

It should not be discounted that the relative effectiveness of the different calibers is entirely pertinent to the discussion of self defense. What I mean is this: Marshall and Sanow's findings show 90% effectiveness for one caliber and 40% effectiveness for another smaller caliber from real-world encounters. For me, it's simple. I want the 90% caliber. Does this mean I can expect 9 out of 10 assailants to fall with one shot? No. Does it mean this round is likely MUCH more effective than the smaller one? Yes.

In my psychotherapy practice, I know that certain techniques work well with, say, abuse victims. I don't care whether it's 90% effective or 80% effective. I'm going to choose said treatment over one that shows 50% effectiveness with such a client.

I believe Marshall and Sanow has great significance as long as it's used properly.

My $.02.

I agree with you and I think Evan Marshall agrees with you. That's how I see it, a comparative guide to the relative effectiveness of different loads. Now, you will notice most of the nay sayers are big bullet worshipers at the church of the 1911, pastors Jeff Cooper and Chuck Taylor. They so despise the unholy 9mm that they refuse to admit that it's even in the same percentage range as the holy caliber. Some even refuse the power of the .357, but many will concede the point since it's a revolver and nobody serious about self defense would carry anything, but a 1911 after all...:rolleyes:

Rather than believe statistics of actual shootings, the pastors of the holy church make up books for their bible containing their proven formulae of "stopping power", which of course are based on momentum being that is the strong suit, the only strength of heavy and slow since it is linear in relation to velocity. They present these truths as handed down by God, just in case you don't believe them, God's initials quite coincidently being "J C".

Well, I know there's more to terminal ballistics than any silly momentum equation given proper torso hits and seems to me, knowing what I do about statistics and statistical inference, that the real world is the best reflection of what works. You can say all you want about the holy implement being "proven in war for a century", but give me some statistics to PROVE it! After all, the primary weapon in war is the long arm. The side arm is only an officer's decoration to be actually used in either very special circumstances or in sheer desperation. I know I'd never personally want to go into actual combat armed only with a handgun. :rolleyes: And, also, I'm not restricted to hardball. I don't give a flip about Geneva or Hague.

There is plenty missing IMHO in the understanding of physicists as to what happens when bullet meets bone and flesh. There's lots of variables and no simple momentum formula explains anything. That's almost as bad as trying to build a climate/weather model with a single variable algebraic equation. :rolleyes: Until all is known about terminal ballistics, the best measure I can find for how one caliber relates to another is the stats.

Now, it really isn't too complicated if you don't have some sort of agenda to prove. It really is common sense. You got your major caliber guns and your minor ones. .38, 9, .40, .45 example of some of the more powerful defensive loads, make good energy and still controllable. Then, you got your .32s and your .22s and your .25s. I mean, like Kinky says, how hard can it be? The only fence sitter I see in the line up is .380. To me, it's a compromise, currently a necessary evil if I need to pocket carry. That's really all I need to know about caliber performance, carry a caliber that makes 70% or better preferably 80% or better in the M/S stats and go with it. .380 barely makes that cut. Don't expect it do kill 70% of the time on the first shot, so keep firing until it works. That's about the size of it, way I see it. If you like .45, fine, but you're only kidding yourself if you think you can take on a bad guy armed with a 9 and the sheer superiority of your bullet's momentum is going to win the day for you. If the bad guy gets off the first accurate shot, your widow is likely to be collecting her life insurance.
 
Understood. To say statistical probability is not my strong suit would be an understatement. I was pointing out to the nay sayers that the data is not useless as some claim. There are so many variables in any of this that nothing is assured. It needs to be considered in context with all factors. In considering any stoping data or ballistics data we need to remember the three most critical factors in stopping power:
1. shot placement
2. shot placement
3. the ever so critical shot placement

Couldn't agree more. This is why, even as a noob to the firearms scene, I'm always saying to my friends who shoot with me (to their growing dismay, I'm sure), "The most important thing in defensive shooting is also the most important thing in real estate: Location, Location, Location."
Caliber and power makes a difference, but a .32ACP projectile through the brainpan tends to have the same overall effect as a .44 Magnum projectile through the brainpan.

Mark(psycho)Phipps( HAHAHA! )
 
You're right that the Marshall/Sanow data isn't "worthless." It's less than worthless. It's worse than having absolutely no data whatsoever.

The only times when a pistol bullet will stop someone with one shot is if the bullet either strikes an extremely vital structure (brain, spine, aorta), or if the shootee is psychologically incapacitated by pain, fear, etc.

Even if we assume the data is accurate, and hasn't been manipulated (there is strong evidence that the data has been manipulated and/or fabricated), there is still going to be next to no correlation between OSS% and actual "effectiveness," in terms of tissue damage, as it is.

Look at what's at the top of the OSS charts. "High-performance" hollowpoints. Look at the bottom. Cheap FMJs of minor calibers. If you were to poll everyone in the country that carries a concealed handgun, on what caliber and ammo they use, and how often they practice, you know what trend you'd find?

In general, people who carry premium hollowpoints in decent calibers practice more often than those who carry FMJs in minor calibers. This shouldn't be a world-shattering revalation. People who spend money on expensive ammo of are better shots than people who just buy whatever's cheap, because in general, they practice more. People who get a large caliber handgun are usually better shots than people who use .32s and .25s, not only because of amount of practice, but ease of aiming the weapon too. People who are better shots get solid hits to the upper torso more often. Solid hits to the upper torso incapacitate in one shot more often than shots that hit wherever.

Correlation is not causation. The Roman gladius has killed more people than any other weapon in history. Does that mean the Roman gladius is the ultimate weapon? No, it just means that Roman military tactics were far better than their contemporaries'.

The only way you'd be able to get any kind of correlation between OSS% and bullet effectiveness is to very carefully catalogue where each bullet hit, and the exact reaction, then compare bullets which hit in precisely the same location, on people of about the same stature, with the same mental condition at the time of shooting. There are so many variables that such a task would be impossible, because there simply aren't enough shootings to get a meaningful comparison.
 
Jim PHL asked, "So, umm, am I OK with my .38 or not?"

Ah, there's the question! I'm sure the therapists here on THR will be willing to help you answer that question, weekly hourly sessions at $150 per week. They will help you understand that being "OK" is a function of your attitude, not just toward your .38, but your Mother, Father, Aunt Matilda and the dog you failed to brush every week when you were eight years old.

Geoff
Who is OK with his .38 and his .45 and his .44 Magnum and his .308 and his 20 gauge.... :D
 
The OSS numbers are garbage.

Relative? Significant if used properly? Salient point?

I shouldn't have to quantify a "scientific" study like that. The data is questionable. The scientific method is lacking. The statistics are impossible to prove, and worst of all, possibly fabricated.

And I don't worship at the altar of any caliber. I've carried everything from 9x18 to .45 and felt just fine. I don't need to justify jack squat, and I don't need made up statistics to make me feel better about whatever I've got on my belt.

Here is an example of why the OSS is garbage (discounting the outright fabrications of shootings that never happened, the discounting of multiple torso hits, and all of the other things we've heard before) I've used this analogy before, so here goes again:

Subject 1: 105 lbs. Weakling. Listens to Michael Bolton. Voted for Ralph Nader. Attacked while watering his ferns. Shot with .25 that lodges in his shirt. Suspect faints at the loud noise. 100% stop.

Subject 2: 290 lbs. Prison hardened muscle. Named Mad Dog (by his mother). Has spider web tattooed across his face. Eats puppys. While high on elephant tranqualizer, is shot by 10mm that lodges in his left lung. Mad Dog kills the shooter (with a shiv made from his own hardened feces) runs 200 yards, car jacks a bus load of nuns and makes his get away. Later has friend who flunked out of vet school remove the bullet with a pair of needle nose pliers. Goes to 50 cent concert that night. 0% stop.

OSS doesn't take into account any of the huge multitude of factors that can change the outcome of a shooting. When a US soldier gets shot and finishes his mission, we give him a medal. When a bad guy gets shot and finishes his mission, we freak out about our caliber.

Seriously? How in the world can OSS prove a damn thing? It isn't exactly a laboratory environment. If I pop a guy in the heart and he is DRT, but I hit him twice more on the way down, that shooting wouldn't count. I teach my students that if they get shot, then they need to keep fighting. We're taught over and over that it is all about training and mindset. Well guess what? There are plenty of bad guys with a fighting mindset too.
 
Statistics work given enough sample size to establish a mean with a decent confidence limit. You really can't draw conclusions on any given load with a sample size of less than 1000-1500, say, shootings per the norm: however, given enough sample size, the variables even themselves out.

As to bias, I have no reason to suspect Evan Marshal of fabricating anything. Why? I do suspect those who would slander him if they want the big and slow to be the best. That's the agenda I've always seen in the caliber wars. They can't honestly say, "don't carry anything, but a .45" if it's shown the .45 isn't the end all of man stoppers. I don't see too many people with a .22 agenda or 9mm agenda or .38. Just the .45 know-it-alls. Why is that?

Despite the stats and despite the Facklers and despite the Strasbourg's and all the Jello in TV land, fact remains in the real world, we can't all be carrying a 1911 all the time even if that's all we were allowed to carry. There are times you need a pocket gun. You might not want 40 ounces pullin' your jeans down. In the real world, shot placement means more than a few percentage points, a few "Taylor power factors", a few foot lbs. Just live with it, but it's still fun to read the stats. :p

There are those here that claim "carry the biggest you can shoot." Well, heck, I can shoot a .45-70 handgun and have fired the .50AE effectively, but I ain't stuffin' no Contender or Desert Eagle in my waste band! There are other factors besides any sort of power factors to choosing a concealment weapon. Little and light makes carrying right. Of course, you gotta be able to hit with it, it needs to be 100% reliable, and it needs to be of adequate caliber. It is in the statement "adequate caliber" where the arguments seem to be and why things like the M/S stats are kept and published to try to shed some real world light on the problem. But, others with agendas continue to want their "side" to win. Whatever. I carry what I carry and shoot it well. That's all that matters to me. I'm pretty competent with my PDWs.
 
"There is no replacement for shot placement." That's what I tell my wife. Should (God forbid) she have to shoot someone,she'll put #00 in the center of his chest. That's a "proven stopper" for ya.

Now,without rehashing everything that was just said,nothing is 100% or 80% or 50% guaranteed effective. A lot of it's the mindset of the person getting shot. If I was up to no good,I'd probably stop at the threat of a bullet. If I was determined to do what I was doing(I mean really determined) it might take more than one round to stop me. That said I'd rather take a .380 than a .44 magnum.
 
I agree. Marshall and Sannow undertook probably the greatest effort to conduct REAL research into street shootings and ballistics. Though there are limitations of the data, it is still good stuff so long as one does not make the fatal mistake in believing that there is such a thing as a "one shot stop".

Take probability stats for what they are, a historical probability, but with hundreds of uncontrollable variables.
 
The problems with the OSS data are well discussed elsewhere, but basically, it's not that the percentages are not guarantees, or that there are uncontrolled variables (inevitable in such a database), but in the statistical methods used to include or exclude each particular data point (shooting). Excluding all incidents involving more than one shot, for instance, may well disproportionately inflate the OSS percentage of less effective rounds due to the circumstance that when the first shot fails, another often follows. When that happens, neither shot is included in the database, when properly the first shot should probably have been recorded as a "failure." Conversely, when the first shot stops and no second shot is fired, it's recorded as a "stop." Do each of these ten times, (20 incidents involving 30 rounds), and your percentage winds up at 100%, when it probably should be 50%. The only way I'm aware of to record a "failure" in the OSS system is to fire one shot, hitting the vital area of the torso, having a failure to stop, and not screwing up the data by shooting the still-murderous felon again.

That said, you can still glean some pretty good info from their studies. Just don't take the percentages too seriously, especially when certain ones seem illogical, given the other data.
 
But it's the best data we have! I mean, come on!!! It's better than nothing. Who cares if it's mathematically equivalent to:

1+1=5
3+4=8
5+1=3
6-2=9
7x3=18​

It's all relative because the errors are spread equally among calibers! :)
 
The way I see it, if you are killed by a freight train does that make you any "deader" than if you were killed by a bus?

A pox on all one-shot-stop myth perpetrators. May they all be stopped by a single shot to the arse with a high velocity staple gun.:evil:

BTW, there was a recent school shooting in my area. A young man was shot four times in the abdomen with a 10mm pistol, he survived. According to M&S, shouldn't he be around 450% dead?:neener:

Clearly defined parameters + sound and clear methodology + independantly verified experimentation = a good study

JH
 
Subject 2: 290 lbs. Prison hardened muscle. Named Mad Dog (by his mother). Has spider web tattooed across his face. Eats puppys. While high on elephant tranqualizer, is shot by 10mm that lodges in his left lung. Mad Dog kills the shooter (with a shiv made from his own hardened feces) runs 200 yards, car jacks a bus load of nuns and makes his get away. Later has friend who flunked out of vet school remove the bullet with a pair of needle nose pliers. Goes to 50 cent concert that night. 0% stop.
ROFLMAO

Correia, you owe me a new keyboard.

JH
 
There have been so many attempts to quantify the expected effects of a given weapon/caliber in the outcome of a fight. Folks have been shooting livestock, ballistic jello and all sorts of materials for years. Marshall and Sanow took a relatively revolutionary approach in completing studies based on already available historical data and throwing some statistics around it.

Are their conclustions relative? Yes. Are their conclusions generally repeatable? Yes Are they specific enough to define every situation's outcome based on statistics/ballistics? Of course not. They have created a very good general "rule of thumb" theories.

In God We Trust..... everyone else please bring data.
 
RYANM wrote:

You're right that the Marshall/Sanow data isn't "worthless." It's less than worthless. It's worse than having absolutely no data whatsoever.

No it isn't. It's a relevant data set where the author's explain their methodology sufficiently enough to allow for the work to be interpreted and replicated.

RYANM wrote:

The only times when a pistol bullet will stop someone with one shot is if the bullet either strikes an extremely vital structure (brain, spine, aorta), or if the shootee is psychologically incapacitated by pain, fear, etc.

Your assertion is nothing more than an assumption, it is in no way a sound scientific conclusion. One thing I do know is that deer are not incapacitated by pain, fear, etc., and our research team has often seen deer incapacitated very rapidly by pistol bullets hitting only the lungs without hitting the CNS or supporting bone structure. If similarly sized mammals can be instantly incapacitated by handgun bullets without hitting the CNS or supporting bone structure, why not humans?

Even if we assume the data is accurate, and hasn't been manipulated (there is strong evidence that the data has been manipulated and/or fabricated),

There is relatively little evidence regarding fabrication of the data. However, there are high levels of correlation between the M&S data set and carefully controlled experiments in live animals which suggest the reliability of the data set.

there is still going to be next to no correlation between OSS% and actual "effectiveness," in terms of tissue damage, as it is.

Actually, the correlation coefficient between the M&S OSS data set and an model using the permanent and temporary cavity sizes is over 0.9. The correlation coefficient with a model using the permanent cavity and pressure wave magnitude is even higher.

Most statisticians would consider this level of correlation to be pretty good.

Correlation is not causation.

You are right. Correlation is not necessarily causation. But it can be the basis of further research which eventually does demonstrate causation. In other words, correlation does not _ALWAYS_ imply causation. However, the lack of correlation does make a pretty good case against causation. This being the case, when there is a high level of correlation, scientist do tend to keep their mind open and keep exploring the possibilaty of a causal relationship.

Michael Courtney
 
Dodson wrote:

But it's the best data we have! I mean, come on!!! It's better than nothing. Who cares if it's mathematically equivalent to:

1+1=5
3+4=8
5+1=3
6-2=9
7x3=18
It's all relative because the errors are spread equally among calibers!

But it's not mathematically equivalent to gibberish. I've published numerous statististical works in the peer reviewed journals and I teach college level statistics. The statistical analysis in the M&S books is compelling and sound and believably accurate within the expected margins of error as determined by standard statistical techniques. There are a small number of errors in the analysis and reporting, as often occurs when data sets of this size are published. However, these relatively minor errors do not detract greatly from the relevance of the work.

Michael Courtney
 
I've published numerous statististical works in the peer reviewed journals and I teach college level statistics.
I invite you to show us YOUR analysis of M&S methodology. Not your opinion, but your work please.
The statistical analysis in the M&S books is compelling and sound and believably accurate within the expected margins of error as determined by standard statistical techniques.
I assume you disagree with MacPherson's statistical analysis? Why?
There is relatively little evidence regarding fabrication of the data.
What glib argument are you going to use to attempt to explain away findings that show greater than 100% stopping power?
However, there are high levels of correlation between the M&S data set and carefully controlled experiments in live animals which suggest the reliability of the data set.
Please identify for us published controlled experiments with animals that, according to you, seem to validate M&S findings?
 
Dodson:
I invite you to show us YOUR analysis of M&S methodology. Not your opinion, but your work please.

In preparation for publication.

I assume you disagree with MacPherson's statistical analysis? Why?

If you summarize a specific statistical analysis in your own words (showing at least that you understand it) and give a reference, then I will be happy to comment on the validity of the analysis and the conclusions.

Our research group has considered all of the published criticisms of the M&S work of which we are aware, and none of them is sufficient to conclude that the work is fraudulent or flawed to the degree you assert.

Dodson:
What glib argument are you going to use to attempt to explain away findings that show greater than 100% stopping power?

Summarize the criticism you are referring to and give a reference. Also sate clearly what you believe the implications to be for the validity of the data set as a whole. This will enable me to comment on the specific criticism you are referring to and its impact on the general validity of the work. The answer is scientifically sound and reasonable, and in no way glib.

Dodson:
Please identify for us published controlled experiments with animals that, according to you, seem to validate M&S findings?

We've done some of our own experiments in live animals, these are in preparation for publication. But the Strasbourg tests also find excellent correlation with the M&S data set, and these have been published for some time. The pressure wave ideas put forward as a result of the Strasbourg tests have found an awful lot of support in the peer reviewed literature since that time and it is pretty much beyond doubt in the neurology field that pressure wave can cause incapacitation and neural injury without easily detectable wounding.

Michael Courtney
 
Michael, while your preparing your paper, would you mind providing some actual documentation that the Stausborg tests ever actually occured? :p
 
Status
Not open for further replies.
Back
Top