Publication
Kantor, M. A., Bleetman, A., Tenbrink, J., & Garg, H. (2025). Close-quarter concealed knife attacks: How much time do you have? The Police Journal, 0(0). https://doi.org/10.1177/0032258X251340171
What Was the Issue?
Law enforcement officers face real and immediate danger from close-quarter knife attacks, yet most of the research—and a lot of training—focuses on threats from 21 feet or more. What’s missing is an understanding of how quickly a concealed knife attack can unfold from conversational (interview) distance (around 8 feet), and how attacker characteristics like age, sex, and movement type impact the speed of that attack.
How Did They Look at It?
The researchers ran a study with 74 adults (ages 18–65) who performed three different knife attacks—thrust, overhead (ice-pick), and horizontal slash— in a random order starting from 8 feet away. Each participant chose their own concealment spot for a rubber knife and executed the attack on a stationary dummy. The attacks were timed from first movement to contact. Researchers then looked at the differences based on age (younger ≤35 vs older ≥36), sex, and knife motion using statistical testing.
What Did They Find?
Average attack times:
Thrust: 1.43 seconds
Horizontal: 1.55 seconds
Overhead: 1.60 seconds
Fastest time overall: 1.04 seconds (thrust). The slowest was 2.75 seconds (overhead).
Males were faster than females in all attacks
Younger participants were generally quicker, but the age difference wasn’t statistically significant
Thrust attacks were the quickest motion across the board
Knives were most often concealed at the hips (left, right, or back)
Weapons typically became visible when just 2–3 feet from the target
So What?
Officers might have less than 1.1 seconds to recognize and respond to a concealed knife attack from 8 feet. That’s less than the time needed for an officer to draw and shoot the suspect. Reaction time is about 0.15 to start moving. If there is a startle response that takes 0.4–0.8 seconds, and times to draw and fire a weapon range from about 0.87–2.05 seconds.
Training needs to focus on rapid movement, tactical decision-making, and defensive/offensive actions that actually fit these compressed timeframes—especially against thrusts, which were consistently fastest.
Males, regardless of age, represent a higher time-based threat, but anyone—male or female, young or old—can deliver a lethal strike before a standard firearm response can be executed.
These findings support a shift in tactical thinking beyond the traditional 21-foot rule. Officers need reps against realistic, close-in, spontaneous attacks, because that’s what they will have to deal with in the field.
My $.02
I want to focus on the thrust attacks in this article. They were the fastest and are also likely to cause serious internal damage. The mean attack time for these was 1.43 seconds. 95% of these attacks would be expected to take between .94 and 1.98 seconds. Based on the reaction time, startle, and draw times discussed above, we would expect officers to be able to use their firearms in anywhere from 1.02 seconds to 3.00 seconds. This means that most officers would end up being stabbed before they could draw their firearm and shoot. Admittedly, officers could sidestep or backpedal to gain time, but the attacker could also use their off hand to interfere with the officer's draw or aim. In short, a standard draw and fire approach is not likely to work.
The natural question then is what should we show them? I'm not sure because I haven't seen any solid empirical data on possible solutions. Maybe drawing one handed while attempting to block/capture the suspect’s knife hand with the off hand or having the officer fall to his back and keep his feet between him and the suspect while drawing and firing might work. Or maybe some completely different approach is needed.
The nice thing is that if I adopt the EcoD approach, I don't have to know the "right" solution to this problem. If you really think about all the variables involved (suspect capabilities, officer capabilities, variations in distance, angle, and direction of movement) there can't be a single "right" thing.
What I need is to understand that this is a problem—and to carefully study how these attacks unfold in reality. I need to know things like—at this distance, drawing a weapon typically takes longer than delivering a stab. Then I need to create activities where trainees can explore this problem and develop solutions. I might give simple instructions like “Don’t get stabbed” or “Stop the attacker.” If they get stuck, I can nudge them toward different options—asking, for example, “What happens if you try...?” I am watching carefully to see what sorts of things seem to be working better and nudging students toward those. We adjust the scenario as soon as students start to look comfortable. As instructors, we share observations to identify what's working and guide students toward more effective responses. I am watching videos of situations like this so that I can make what the students are doing as close to the real thing as safely possible.
Being a trainer in the EcoD realm is not about knowing all the right techniques—it's about designing problems that reflect reality. It is about allowing the trainee to explore those problems, so that they become adaptable problem solvers. The trainee is a wayfinder. You are a guide.
The speed that comes from rote repetition is seductive, but this study and others show that speed alone can't save you. Many attacks happen too fast. What works is awareness, positioning, and adaptability. You only get that from reps that look like an actual fight.
Pete - great work and raises a ton of questions about the validity of how we train for quick, violent, almost instant encounters in which the first indicator that the officer is in a knife fight, might be that they are already cut. It reminds me of how we were taught and subsequently practiced firing a weapon one handed with my off hand. My instructor mentor taught me to start every one of those sequences with the weapon on the ground in front of me. He said, "People get hit in the hand, not because people focus on and shoot at the hands. But because when you are in a gunfight, your weapon (and hands) are between you and the target (hopefully) so they sometimes get hit when the bullets are flying toward you. And, Damian, where do you think your weapon is likely to end up if a bullet tears through your hand? That's right, on the ground." So for the last 20 years, every time I start a one handed/off handed drill, I start it with the pistol on the ground in front of me and then complete the sequence of the drill to include follow-through and scanning rather than just swapping hands with my pistol.
There is a great book research in the 1990's and published in 2001 called, "The Logic of Steel" by James LaFond (link below). He researched and did interviews related to over 250 blade encounters covering events related to everything from street crime to prison attacks. It is eye opening as some of the statistics he published were related to "brandished blades" vs. "concealed blades". It has been a couple of decades since I read it but I remember the percentage of knife attacks after a blade is shown or brandished is very low compared to the number where the blades were concealed up to the point of the attack. I feel that 21ft rule is much more related to the latter than the former and it underscores and relates to your points of having way less time to draw and fire a weapon than it takes to do so.
Great work! Keep it up!
Debo sends.
https://www.jameslafond.com/book/logic_of_steel#faq
Great article