In many of the discussions that come up here, such as those about the comparative effectiveness of various handgun loads, ammunition capacity, and legal issues along the lines of of "will it hurt me in court", someone asks for actual data, and sometimes, someone attempts to draw a conclusion from the lack of such data. From time to time, people even attempt to prov a negative.
That constitutes wasted time and bandwidth.
First, it is fundamental that the number of data points be sufficient to illustrate cause -and-effect relationships. The necessary numbers will depend upon the number of relevant variables,
That's true in any kind of statistical analysis.
When it comes to defensive use of force events, (1) there are few actual events in the first place; (2) the factors that affect the results are numerous; (3) most of the important data are never measured, much less recorded; and (4) the data that are captured reside in thousands of court transcripts and police reports, and are not compiled into a national database
When it comes to trying to use actual event data to analyze defensive shootings, the number of variables involved is usually so large that would take an astronomical number of data points to draw any reasonable conclusions, and what it come to the use of deadly force, (1) the number of relevant incidents is far too small to supply the necessary data, and (2), most of the relevant data are never recorded.
To illustrate the latter point, let's consider handgun stops of assailants. The question of interest may be one of whether a 9mm or a .40 made a difference , or of how many shots were really required. No matter. The data are not available.
It is not a matter of how often the attacker was fatally injured—that’s not important to us. Whether the defender was injured doesn’t tell us anything. How many shots were fired might be indicative if we knew why. What is important is how each each event unfolded.
Consider the subject of wounding effectiveness. The human body is not a water jug, or a gelatin block, or a steel plate. It is an extremely complex mechanism, and some handgun hits may have little or no immediate effect, and others, much more.
Averages will not help us.
To evaluate the effectiveness of shooting (in terms of timely stopping, and not lethality) by looking at actual event data, one would need to know which round struck where on the body and damaged what inside, not only in what order but at what precise moment in time.
Why? Well, a bullet that strikes a body in a particular place can have a markedly different effect than one that strikes very close to the same place, Not only that a bullet may have an immediate internal effect that differs very greatly from that of a bullet entering, at the same point, a target that is at a different angle in or more axes. The important targets are small things, concealed within a large, moving three-dimensiopsl envelope.
In the real word, the relevant data cannot be recorded. But if they could, we would also have to know how each round actually affected the person shot, and how much ability the aagressor possessed after each shot.
We cannot get those data. They are never recorded. They cannot be.
But if they were, there would still be far too many variables--the physical condition and the state of mind of the attacker, and so on.
Do we need to discuss why a couple of simple stats would not be conclusive?
Well. the numbers of shots fired in different encounters might be somewhat indicative, but they would not be very useful. The results will depend on a myriad of other factors in each incident.
Same thing when it comes to caliber.
And again, the relevant variables far, far outnumber the events.
That's just the way it is.
No, actual event data cannot really help us. We need something else.
That's why penetration and expansion testing, voluminous range shooting results including measurement of speed of fire and precision, expert medical judgement must all be brought to bear.
For civilians accustomed to firing at stationary targets at the range, realistic FoF simulation can be extremely enlightening.
The FBI recommendations for agents and for their law enforcement partners are based on medical judgments, range shooting results, and ammunition testing.
Except for things pertaining to shooting through plate glass, there are no reasons why civilians would have different requirements.
We have been speaking of handgun effectiveness--terminal ballistics, number of rounds, rapidity of fire, and so on.
How about court cases?
"If its a good shoot.... Show me a case...."
Well, for the same reasons discussed above, we cannot use actual conviction and acquittal data to come to any meaningful conclusions. There are simply too many variables in a murder case, and too few data, and each case hinges not on one variable, but on the totality of the evidence, for each juror.
Want to get an idea whether that pistol grip shotgun might hurt you in court? Don't look for actual data.
So, what do we do? How about jury simulation? Dr. Glenn E. Meyer has done a lot of that, and he has written about it.
I put a lot of stock in it.
The advent of powerful search engines has led many people to believe that just about everything that one might want to know has been measured, recorded, and compiled.
That is just not the way it is
That constitutes wasted time and bandwidth.
First, it is fundamental that the number of data points be sufficient to illustrate cause -and-effect relationships. The necessary numbers will depend upon the number of relevant variables,
That's true in any kind of statistical analysis.
When it comes to defensive use of force events, (1) there are few actual events in the first place; (2) the factors that affect the results are numerous; (3) most of the important data are never measured, much less recorded; and (4) the data that are captured reside in thousands of court transcripts and police reports, and are not compiled into a national database
When it comes to trying to use actual event data to analyze defensive shootings, the number of variables involved is usually so large that would take an astronomical number of data points to draw any reasonable conclusions, and what it come to the use of deadly force, (1) the number of relevant incidents is far too small to supply the necessary data, and (2), most of the relevant data are never recorded.
To illustrate the latter point, let's consider handgun stops of assailants. The question of interest may be one of whether a 9mm or a .40 made a difference , or of how many shots were really required. No matter. The data are not available.
It is not a matter of how often the attacker was fatally injured—that’s not important to us. Whether the defender was injured doesn’t tell us anything. How many shots were fired might be indicative if we knew why. What is important is how each each event unfolded.
Consider the subject of wounding effectiveness. The human body is not a water jug, or a gelatin block, or a steel plate. It is an extremely complex mechanism, and some handgun hits may have little or no immediate effect, and others, much more.
Averages will not help us.
To evaluate the effectiveness of shooting (in terms of timely stopping, and not lethality) by looking at actual event data, one would need to know which round struck where on the body and damaged what inside, not only in what order but at what precise moment in time.
Why? Well, a bullet that strikes a body in a particular place can have a markedly different effect than one that strikes very close to the same place, Not only that a bullet may have an immediate internal effect that differs very greatly from that of a bullet entering, at the same point, a target that is at a different angle in or more axes. The important targets are small things, concealed within a large, moving three-dimensiopsl envelope.
In the real word, the relevant data cannot be recorded. But if they could, we would also have to know how each round actually affected the person shot, and how much ability the aagressor possessed after each shot.
We cannot get those data. They are never recorded. They cannot be.
But if they were, there would still be far too many variables--the physical condition and the state of mind of the attacker, and so on.
Do we need to discuss why a couple of simple stats would not be conclusive?
Well. the numbers of shots fired in different encounters might be somewhat indicative, but they would not be very useful. The results will depend on a myriad of other factors in each incident.
Same thing when it comes to caliber.
And again, the relevant variables far, far outnumber the events.
That's just the way it is.
No, actual event data cannot really help us. We need something else.
That's why penetration and expansion testing, voluminous range shooting results including measurement of speed of fire and precision, expert medical judgement must all be brought to bear.
For civilians accustomed to firing at stationary targets at the range, realistic FoF simulation can be extremely enlightening.
The FBI recommendations for agents and for their law enforcement partners are based on medical judgments, range shooting results, and ammunition testing.
Except for things pertaining to shooting through plate glass, there are no reasons why civilians would have different requirements.
We have been speaking of handgun effectiveness--terminal ballistics, number of rounds, rapidity of fire, and so on.
How about court cases?
"If its a good shoot.... Show me a case...."
Well, for the same reasons discussed above, we cannot use actual conviction and acquittal data to come to any meaningful conclusions. There are simply too many variables in a murder case, and too few data, and each case hinges not on one variable, but on the totality of the evidence, for each juror.
Want to get an idea whether that pistol grip shotgun might hurt you in court? Don't look for actual data.
So, what do we do? How about jury simulation? Dr. Glenn E. Meyer has done a lot of that, and he has written about it.
I put a lot of stock in it.
The advent of powerful search engines has led many people to believe that just about everything that one might want to know has been measured, recorded, and compiled.
That is just not the way it is
