Data sets, plausibility, and defensive shooting: what you don’t know can waste your time, energy, and money.

Data sets, plausibility, and defensive shooting: what you don’t know can waste your time, energy, and money.

As I sat eating lunch last week I found myself perusing a gun forum with which I’m not all that familiar. On it I ran across a post from a fairly well known trainer, one that most shooters would not recognize but those familiar with the training world might. I’ve never met the guy, let alone trained with him, but his comments left me distinctly perturbed.

The statement was in reference to some particular techniques that he finds important to teach. In defending his approach, he wrote “I know, statistically, it is unlikely that you’ll ever need these skills. Of course, statistically, it is unlikely you will ever need a gun at all.”

I’m not at all sure that he understands the implications of what he said.

Let me start with some perspective. The American Cancer Society tells us that approximately 1.5 million cases of cancer will be diagnosed this year. With the U.S. population standing at a tad over 307 million as of the last census, that puts those patients at about .48% of the population. That’s right – less than one-half of one percent of the population of the U.S. can expect to be diagnosed with cancer, which one would have to say is a pretty small number. (As it happens, it’s still quite a bit smaller than the percentage of people that Kleck and Gertz tell us will use a gun in self defense the same year, by roughly half. Keep that in mind.)

Those numbers make it statistically unlikely that any one person will develop cancer in any given year; the total number of cases is small compared to the whole population. Even though cancer of all types is not terribly common, we all know that not all cancers (nor diagnoses) are equally likely, let alone have the same outcome. Some cancers are far less prevalent than others; salivary gland cancer, for instance, occurs in perhaps 6,000 people per year – compared to nearly a quarter-million who develop who develop prostate cancer. That’s a huge difference despite the fact that neither is likely to occur.

What medical science doesn’t do is to flail about and proclaim that since any cancer is “statistically unlikely” to begin with, they’ll throw the same treatment at all of them in hopes that something works. That’s not how science is done, and it’s not how lives are saved.

Within that small data set of cancer cases there is a huge range of probabilities and outcomes. It’s that very fact that enables medical science to classify each case and use the best treatment approach based on where it falls in the data matrix. Since not all are alike, all do not get the same treatment.

This extends to the research realm as well. We don’t spend as much time and money developing cures for salivary gland cancer as we do for prostate cancer. We put our research resources where they will do the most good, where they will save the most lives.

Am I saying that defensive shooting is the same as cancer? Of course not. What I am saying, though, is that just because an occurrence of an event is unlikely doesn’t mean that all such unlikely occurrences are the same. A small data set does not imply homogeneity; even in small data sets there are differing circumstances and results. To imply otherwise is ignorant (or manipulative.)

Of course it’s statistically unlikely that at any given time you’ll need to use your gun. This is not news. Needing to use a gun to defend yourself is about twice as likely as you developing cancer this year, mind you, but it’s still unlikely. Just because it’s unlikely, however, does not mean that all skill sets related to a defensive shooting are of equal value!

Just as some cancers are more common than others, some defensive scenarios are more likely than others. For instance, how often in private sector self defense incidents are people called on to make 100-yard hostage rescue headshots with a handgun? It may have happened somewhere or at some time in history, but I think you’d be hard pressed to find a single case – let alone any sort of trend. Is that of equal probability to dealing with a simple assault inside of a car’s length in a parking lot after dark? Of course not.

Should we train equally in the skills necessary to deal with those two disparate events simply because neither is “statistically likely”? I don’t think so.

When we look at defensive shooting threats and scenarios, there are some that are possible but have rarely (if ever) happened; there are some which happen occasionally but not often, making them at least plausible; and there are those which happen often enough that we can see some sort of likelihood, a certain probability of occurrence. Our problem as students is that none of us has the unlimited time or resources necessary to train for everything which is merely possible. We have to take into account the likelihood, the plausibility, of what can happen when we make training and technique decisions.

Using the “statistically unlikely you will ever need a gun at all” argument in relation to training is a smokescreen, a way to ignore the concept of plausibility. It’s an attempt to deflect the student’s attention, to get them to suspend their critical thinking so that they don’t question the actual value of the technique. Yes, it is unlikely that you’ll need to use your gun – but saying so doesn’t magically transform “possible” into “likely”, and doesn’t elevate a rarely needed skill into something which is vital to learn.

-=[ Grant ]=-

  • Posted by Grant Cunningham
  • On November 14, 2011