066 – Training for untrained performance.
If students spend most of their time mimicking the actions of poorly trained combatants who performed poorly in gunfights, then we shouldn’t really expect anything different.
There is a mainstream concept being applied in training design for combat shooting, especially for law enforcement and self-defense, that goes something like this:
Critical incidents and/or self-defense situations are stressful and you never know what you are going to do. We do, however, have some data from the FBI’s Uniform Crime Report, LEOKA, and various studies of shootings.
This data tells us that most shootings happen at close range, most shootings happen quickly, most defenders / officers will be behind the reactionary gap, and most people will shoot as fast as they can under stress.
Based on this data, we will train mostly at close range and will spend most of our time trying to shoot as fast as we can “at the speed of the fight” with “combat accuracy.”
This is of course an overly broad brush but that ought to sound generally familiar to anyone who has been around the training side of the industry for a while.
Our question here is: does this actually make sense?
When reviewing any study or data, it is always worthwhile to look past the surface level and try to understand where the data comes from before trying to determine what it means. This doesn’t mean data is bad. It certainly is not. It just means that we should understand the limitations of the questions and measurement methods that produced the data before we interpret it.
One of the things that is conspicuously absent from the majority (if not all) of the publicly available scientific literature and data sets regarding gunfight performance is an accompanying analysis of how each individual involved was trained. To be fair, this analysis would be enormously complex and likely almost impossible to produce. Still, it is worth noting that these factors are absent.
Most research we have seen that attempts to associate performance with training and/or pre-existing skill have been done using each officer’s experience as the data point. This is typically defined by years on the job, or with participation in a specialized tactical unit such as SWAT.
Unfortunately, neither of these metrics actually mean much with respect to baseline skill performance, especially in broad data sets spanning multiple agencies. This is particularly true for the “years on the job.” In most cases this has no bearing whatsoever on an individual officer’s hard skills. What we do know, with certainty, is that the historical average level of firearms and tactical skill in law enforcement is minimal.
There is also a second item worth noting with respect to the most-used large data sets (the FBI’s Uniform Crime Report (UCR) and the Law Enforcement Officers Killed and Assaulted (LEOKA) data). By nature of reporting requirements and collection methods, the data contained in these reports is biased towards engagements where the officer either lost the fight or was substantially injured in the process.
This does not make the data useless. However, it is certainly worth nothing. If we ignore in analysis that the bulk of data contained in the commonly cited large-scale gunfight performance data sets is taken from gunfights that were lost—or at least that nobody would choose to replicate—then how we interpret and apply the data may not produce the outcomes we are looking for.
More pointedly, allow us to suggest that this should give us all some pause with respect to whether or not this data is indicative of performance that we should be seeking to replicate.
This brings the discussion to where the rubber meets the road—impact on training design and implementation.
At Building Shooters we are not primary researchers. Our focus is combining research with practical experiences in both training and real-world environments to impact training design and delivery.
Vocational training models assume that a student’s eventual mastery of the subject will be acquired during on-the-job performance. Unfortunately, this assumption catastrophically fails for firearms skills.
Firearms simply are not used enough in the real-world for these skills to be mastered, much less acquired, on the job. At a systemic level, all ability with respect to firearms must be developed solely through the training process.
Looking through this lens, it seems difficult to justify the design of training around replicating real-world performance that mostly failed.
Allow us to briefly switch subject matter—so as to avoid any pre-conceived biases and assumptions.
Imagine a study about American football that was conducted on offensive line performance. The purpose of this hypothetical study is to analyze quarterback sacks. Therefore, a fundamental requirement for data about a play to be included in this study is that the quarterback was sacked.
Data collected in this imaginary study includes how the offensive line was arranged, how the defensive line was arranged, and what play was run by the offense.
Let’s pretend that this study produces the following results, showing patterns of plays and offensive/defensive line-ups. Together, these patterns account for 90% of all sacks. These patterns are as follows:
There are three primary defensive line-ups. A, B, and C.
There are three primary offensive line-ups. 1, 2, and 3.
And there are two primary plays run by the offense, P1, P2.
Defensive Line-Up | Offensive Line-Up | Offensive Play | Percentage of Occurrence |
A | 1 | P1 | 25% |
B | 3 | P1 | 50% |
C | 2 | P2 | 15% |
From plays used in the data set, 50% of the time an offense is in line up 3 against defensive line up B, running play P1 and another 25% of the time, an offense is in line up 1 against defensive line up A, also running play P1.
Using the same logic commonly applied to data from police shootings, together offensive line-ups 1 and 3 running play P1 are 75% of our recorded plays. Therefore, we should spend most of our time practicing play P1, from line-ups 1 and 3 against a defense in A and B configuration. Right?
If we ignore the fact that the study’s data only incorporates plays where the offensive line failed, this logic makes a lot of sense. And, indeed, it is likely that this training approach would produce some degree of improvement and a reduction in the number of sacks when these same configurations are encountered in the future.
However, does anyone expect a team that trained and subsequently played this way—only practicing and running plays that historically don’t work—would be successful?
2025 Schedule Now Available
“Having the written material plus seeing the methodology and technology on the range along with hearing Dustin present the nuances of the material really helps paint the picture of how we should all be presenting firearms training for those who have signed up to protect their communities.”
-John Baggett, Rangemaster, Central Alabama Sheriff’s Office with 200+ personnel
“Despite being only sixteen hours long, [Understanding the Terrain] provided me with more valuable information and knowledge than the majority of forty-hour gunfighting classes I have taken. In particular, it has allowed me to enhance my understanding of adult learning and neuroscience concepts, which I can now integrate into the curriculum I am developing.”
-Aaron K., Career LEO and State Curriculum Developer, June 2024
Should we train primarily to do things that the people who didn’t have the benefit of this training did—things that didn’t work?
If not, then why should we blindly take this same approach to training for armed combat—where the consequences are a lot more severe and second chances far less frequent?
There are several areas to look at here, with volumes to be written on each subject (and much room for legitimate debate). For the purposes of this article, we’ll simply provide concept summaries.
First, context matters.
Our brains have systems dedicated to relational memory. Most of the scientific literature on the subject is focused on what is termed episodic memory. By definition, this is consciously accessed; however, it is also generally understood and acknowledged that our unconscious and pre-cognitive processing incorporates similar relational elements of memory.
When we are presented with challenges to solve, our brains, both consciously and subconsciously, look for prior experiences that match contextually with the situation we are experiencing. Our brains then use these prior experiences and the remembered outcomes to help generate our actions.
It’s worth noting that these actions don’t have to make cognitive sense. (Picking up shell casings while being shot at, for example.) Our subconscious brain just needs to recognize a relationship (hence the term relational memory) between the action, the context of the experience, and a generally desired outcome. This can be a big deal in training if we create a large volume of skill performance with inappropriate or unrealistic operational context.
Going back to our hypothetical football example, a combination of defensive line up B, offensive line up 3 and offensive play P1 produced 50% of all sacks. If that configuration accounted for most of what a team practiced—especially if the practice was structured like what we do in tactical firearms and reality-based training, where the offense never loses—then the following potential outcome is created, by the training process itself:
A quarterback under high stress during a game loses the ability to do high-level cognitive reasoning. He then observes the defense line up in formation B. Seeing formation B, the quarterback calls an audible to shift to offensive formation 3 and to run play P1—even though this action historically provides the greatest possibility of failure.
Why does the quarterback do this? Because whenever he sees defensive formation B in practice, he is in formation 3 and runs play P1. He also spends most of his training time doing this.
It feels right and—in training—he always finds a way to win. His subconscious brain associates this context and associated action with a positive outcome.
The quarterback in this case has many options available to him. Unfortunately, his training-related experience has informed his subconscious that he needs to “fight like he trains.” Unfortunately, due to poor training design, that may mean lining up in a formation and then running a play that, together, have the least possible potential for producing a successful result.
In other words, one of the outcomes produced by the training in this hypothetical is contextually guiding the quarterback into deciding to put the offense into a tactically disadvantaged position.
The offense doesn’t end up at a disadvantage because of a bad situation outside of their control, nor are they proceeding with a suboptimal situation because it was tactically necessary for some tertiary reason or objective. In this example the quarterback is deciding to do the dumbest thing possible—because it matches with the familiar context experienced frequently (with successful outcomes) in training.
Please Note: There are, in fact, critically important reasons for students who are involved in reality-based training to “win.” Please do not misinterpret (or misquote) this article as suggesting otherwise.
A second common result of designing the bulk of training repetitions to mirror unsuccessful (or at least undesirable) real-world outcomes is that it may produce poor fundamental skill development. This then ultimately results in tactical skill performance limitations that most armed professionals will never overcome.
We’ve written at length about this specific subject matter in past articles and books, both with respect to visual skills and with respect to the impacts of expectations on skill set development. Here we will be brief.
If someone trains mostly inside 5 yards to hit somewhere in the silhouette of a static piece of paper, then this person spends most of his or her time performing a task that can be accomplished without developing shooting skills.
What we mean by this is that this task can be accomplished by using the same gross motor skills that all developmentally healthy humans acquire simply through interacting with the world around them. No gun-specific skills are needed. The gun is operated with the same collection of brain circuitry the person uses to operate, say a beer can or a lawn mower.
Does this work? Certainly it does. In fact, there are entire training programs built around these general principles (see Shooting to Live by Fairbairn and Sykes).
The problem with this type of training is not that it doesn’t work. It is that it doesn’t work well—or more accurately that it ends up putting significant, and usually unnecessary, limitations onto a student’s skillset. For example, the inability to even hit a person under any level of stress once distances increase or any level of precision becomes necessary.
The basic reason these limitations are formed is that the development of shooting skills that can be applied by the procedural (unconscious competence) memory system requires constructing entirely new brain circuitry that is specific to the gun and the tasking.
This isn’t an insurmountable task if you know how to do it. However, it does require some effort.
At a fundamental (e.g., physics and chemistry) level inside the brain this means that a non-trivial level of energy and resources are needed. After all, the requirement is literally to build new and novel connections and structures.
Our brain functions are governed by physics. Therefore, the brain works according to the principle of energy conservation. We don’t like doing the unnecessary—and most people won’t—because doing so is a waste of the brain’s energy.
Wasting resources makes us less likely to survive at a basic systems level (i.e., we will run out of power and our brain will shut down sooner if we use more power than is necessary). As a result, we tend to avoid doing so as a default setting.
If we spend most of our training time doing tasks that do not require the use of specific brain circuitry to be successful, then our brains simply won’t ever build that circuitry.
Why would they build it? Our basic operating system considers doing so a waste of perfectly good resources.
As a police or military trainer, if you have one day of training and a grand total of 36 rounds to expend annually per trainee (again, see Shooting to Live by Fairbairn and Sykes), then getting your students to operate a gun with their “beer can circuitry” may in fact be a legitimate (and possibly the only) option for your training design.
However, if your training options and resource levels include weeks or months of access to your trainees (not all firearms training needs to be done on the range—a lot of it may even be better done off the range), then we would suggest that, at best, you’re doing both your students and the general public a serious disservice if you give them unnecessary limitations on purpose.
This is especially true if those limitations provide additional tactical context that will carry through to the real world (see above). For example, developing a shooting technique that doesn’t work past about 5 yards.
This effectively traps the student in tactically disadvantaged positions because those aren’t just the only contexts where the student has trained, those are the only contexts where their combat skills can even work.
In conclusion, as an industry we should really take a step back and look at the data we’re using and what it means. We should also ask ourselves what brain circuitry we are developing in our students and why when we run training.
The concepts behind effective training design are relatively straightforward and can be thought of as a sort of reverse engineering. Start with what you WANT the student to do. Not in training—for real.
The art here is that this needs to be realistic—and it needs to work. Training people to do things that don’t work is more than possible. It happens frequently; sometimes intentionally, sometimes by accident.
Once you know what you want people to do though, then it’s simply a matter of building the brain circuitry to make that performance happen. Real world performance, especially under high stress, is more or less the result of the function of brain circuits. What exists will be used. What does not exist cannot happen.
If we want things to happen in gunfights, then we need to build the brain circuits that will make them happen with our training processes. Anything else is literally a waste of time and resources.
With respect to the hypothetical football example, it is certainly fair to point out that tactical situations, for police officers in particular, are different than for football teams (they clearly are substantially different).
It’s also fair to point out that officers may have much less control over some of the contexts in which they find themselves than the quarterback in this hypothetical (also true). Police especially need to get close to people and that forces a lot of close-range interactions and uses of force.
None of this, however, changes the fundamental point of this article: If we spend most of the available training time asking students to mimic context and skill performance that historically was performed by poorly trained individuals who lost in real gunfights, then we really shouldn’t expect to produce anything different.
