Kleine IBT inzichten: what types of factors might influence shooting errors beyond the usual shoot/don’t-shoot format?

In dit onderzoek keken de onderzoekers naar wat voor soort factoren invloed kunnen hebben op ‘schiet fouten’ (fouten in vuurwapen gebruik) die verder gaan dan het gebruikelijke wel versus niet schieten criterium. En wat dan betekent voor vuurwapen training. Nadat ik de studie gelezen had zocht ik contact met de auteur met als vraag de studie te verhelderen. Als frontlinie instructeurs zijn we het aan onze doelgroepen verplicht op de hoogte te blijven van onder meer onderzoek. Het liefst zoals Brian Willis zegt met een ‘open mind and a strong filter’. Het eerste spreekt voor zich, het laatste betekent ook filteren wat je wel en niet mee wilt nemen. Hier komen de vragen over het mooie onderzoek.

What were the research questions and hypotheses?

-We wanted to explore what types of factors might influence shooting errors beyond the usual shoot/don’t-shoot format. So, we looked at competition-style shooting because threat identification was always the same; that is, participants knew the targets and non-targets in advance. This information eliminated the issue of threat identification as a primary source of shooting errors and enabled our exploration of other issues. In particular, we wanted to see how expertise and unknown environments influenced outcomes since our primary hypothesis was that cognitive failures could influence performance beyond threat identification failures.

What are the main results and what can first responder firearms trainers best do with this?

-Several main results were interesting. Foremost, more time to plan (Dark House stage in Experiment 1 and Visible stage in Experiment 2) produced more no-shoot targets hit, where participants fired upon (and hit) targets they should not have. That represented the most counter-intuitive finding since you would think more time to plan should reduce those types of errors, and while we can’t say with certainty why they happen, there are several implications to include planning itself being a source of cognitive errors in marksmanship (e.g., you think you can make a shot and try to plan out an assault, but mess up parts of the plan and produce errors as a consequence). Additionally, we saw an inverse speed-accuracy trade-off for experts, where they were more accurate as they went faster. This latter point is especially not captured well in standard, points-based marksmanship tasks (we have several forthcoming articles describing why points-based systems are insufficient to assess skill). Specifically, many marksmanship tests hold speed constant and count accuracy, such as giving all shooters a 5s or 10s time limit on a drill. There are no more points for firing 2 bulls-eyes in 2 seconds than there are for 2 bulls-eyes in 9 seconds. That simply does not reflect the realities involved in use-of-force training. Hit factor does a better job than points-based systems as it accounts for speed and accuracy, but we prefer small arms combat modeling since it also represents risk and performance variability in the outcomes. There is a lot of nuance to that point, but bottom line, expertise in firearms use involves a speed/accuracy trade-off where faster performance can also be more accurate performance. If marksmanship tests do not reflect this issue in the assessment (e.g., providing time limits and counting bullet holes on paper), they may not adequately or accurately capture skill.

What is the relationship found between planning and shooting performance? What are “planning errors”? What are cognitive errors in firearm use and why is threat identification not enough to investigate?

-Answers here clump together, so I will try to answer as one. Foremost, firearm use has many errors from negligent discharge to threat identification failures. Sensorimotor errors are also explored, and Paul Taylor has done some good work on related factors such as muzzle position. Planning represents an understudied type of cognitive failure where things such as working memory load can be a source of error. Take the Static stage of Experiment 1 for example. Shooters never moved from the starting spot, knew the position of the target, knew the identities of each target, and could plan out their every motion in advance. Despite these advantages, they still hit more no-shoot targets than when they were running through a shoot house with a flashlight. Granted, target placement, number, and type become major issues, which is why Experiment 2 was so important. Those conditions equated all those factors, yet we observed a similar outcome. In short, the ability to plan out movements represents a major tactical advantage, yet it can also represent another potential source for cognitive errors in marksmanship.

The why, however, becomes speculative. We can see differences between these conditions, but do not have enough information yet to make a causal inference. Different possibilities could be overconfidence in the shooter thinking they can make a shot and deciding in advance to take it, or people forgetting the plan and trying to recreate a plan after bumbling the steps. Memory load could also be an issue if they are trying to remember what they are supposed to do next rather than responding to the situation. There are many possibilities, and we cannot say for certain. All we can claim is that providing them the opportunity to plan induced more no-shoot hits than without that opportunity.

What is the relationship between good shooting performance and (wrong) decisions?

Strangely enough, not as direct as you might think. We have another paper that looked at marksmanship performance in different types of simulators (e.g., video game style versus military-grade; Blacker, K. J., Pettijohn, K. A., Roush, G., & Biggs, A. T. (2021). Measuring lethal force performance in the lab: The effects of simulator realism and participant experience. Human factors, 63(7), 1141-1155.). Marksmanship had no correlation with false alarms (i.e., hitting no-shoot or unintended targets) in a shooting simulator. Our conclusion was that marksmanship and cognitive performance can be more orthogonal than people imagine, where good shooters do not always make good decisions (actually, a good shooter making a bad threat identification decision is likely to produce an unintended casualty). That finding needs more work, but the related outcome in this study demonstrates how marksmanship abilities and hitting no-shoot targets are not as dirctly related as people might think. Cognitive performance is a different type of failure.

What are the recommendations for firearms training?

At a simple level, competition-style stages can provide more nuance and fidelity in marksmanship tests. Shooters must move more between targets, engage multiple targets, and conduct other behaviors such as reloading within a course. Typical marksmanship tests isolate particular skills in particular drills, which works for isolated skill development, but does not represent a real-world lethal force encounter. Competition-style stages move much closer to this type of engagement than firing upon a known target at a known distance in a known course of fire on a firing range with no shooter movement. The scoring is also important. Incorporating speed and accuracy into the assessment gives you better insight into their performance. Variability is great to identify too, but that moves beyond hit factor and into small arms combat modeling. Small steps at first, but most places should be able to implement hit factor without much trouble.

Thank you again as always for your interest, and hopefully you’ll be interested in the small arms combat modeling stuff we have coming out soon!


Adam T. Biggs, Joseph A. Hamilton, Andrew G. Thompson, Andrew Jensen, Joel Suss, Karen Kelly, Rachel R. Markwald, Not according to plan: Cognitive failures in marksmanship due to effects of expertise, unknown environments, and the likelihood of shooting unintended targets, Applied Ergonomics,
Volume 112, 2023,