Defensive Gun Uses: Three Shots, Three Seconds, Three Yards?

Previous Post
Next Post


I often read that most defensive shootings occur within three yards, with three shots or less, within three seconds. As far as I can tell, that is a complete myth.

The FBI’s Uniform Crime Reports lists the distance for some shootings of law enforcement officers. The stats average out to three yards or less. But not all…not even most reports list the distance of the shooting.

Moreover, the distance listed isn’t always the initial distance the officer was shot at, but the distance at which the officer was killed. Note: this data only exists for the officers, not non-law enforcement victims of shootings, which is the vast majority of victims.

As for the other two measures — three shots in three seconds — I can find absolutely nothing to validate those numbers, especially for non-law enforcement victims. At the same time, hundreds of online videos of shootings shows far more shots taken, often over a longer period of time.

I’ve asked a FBI agent, a sheriff, and the Texas Department of Public Safety if there’s any source for the distance at which non-law enforcement victims of shootings were shot at. They’ve all told me no.

If anyone can provide distance, time and shots fired information for defensive gun uses and post it here, I’d be most appreciative. Meanwhile, you might want to re-consider your EDC’s ammo capacity.



Previous Post
Next Post


  1. I said this yesterday. I’d like to know how that 3,3 &3 rule is calculated. I’ve been skeptical since this “stat” started being bandied about.

    This brings into question if it’s even a real calculation at all or just the playing about with extremely incomplete data for funzies. Incomplete data often leads to erroneous conclusions.

    Again something I said yesterday: If we assume that the average person doesn’t carry a derringer but rather carries at least a five shot snubbie then we have two pieces of conventional wisdom that are at odds with each other. One says 3, 3 & 3 the other says that “relatively untrained” people are likely to empty the gun. Statistically speaking I would guess that these two things can’t be true at the same time unless a hell of a lot of people use a derringer.

    • Not gonna lie I like to think this is in reference to my bringing up the rule of three yesterday and even though I’m not mentioned by pseudonym it makes my day

      • Yeah, my saying what I said was in reply to you IIRC.

        Personally, and I know this will tick some people off, but I see no basis for any of the numbers. There are some nice theories out there but very little hard data which means the theories are unsubstantiated bullshit that people tell themselves to feel better.

        Most of the theories seem to me to be people clinging to this rule to justify their carry of a small gun because that’s what they prefer (those in states with stupid mag limits excepted). It’s easy and this is a piece of conventional wisdom that justifies what they’re doing. If it turned out the numbers were actually 7 yards, 9 shots and 5 seconds a whole lot of people would have their Cheerios urinated in by reality but since this is basically impossible to calculate they can stick with this 3, 3 & 3 idea and tell themselves that they’re good to go.

        • And I would think that at three feet, even the average shooter using a semiauto pistol will fire three shots in the first second. Unless we are talking about Jerry Miculek, most double action will shoot a bit slower, but still faster that one shot per second. I think most of us can accomplish aimed fire of three shots at 7 to 15 yards in three seconds.

        • yea I did some googling and couldn’t find a reliable source for the rule of 3… bummer. I guess it just always made sense to me. Either way I love my snubby if only for reliability and ease of drawing and deep concealability . That being said I had considered something with more capacity after the shooting in Orlando right in my neck of the woods.

    • Yet this is one of the reasons I carry a revolver. At extreme close range a revolver won’t push the slide out of battery with a contact shot, yet even with my 3″ (exposed hammer) .357 I have no doubt about my ability to engage an enemy with lethal precision at 50 yards without a rest or 100 yards with a rest. What the revolver doesn’t have is the ability of laying down a high volume of cover fire. But we each must make our choices and take our chances…

      • Cover fire? <– Really? If 'a' warning shot (singular), is frowned upon, how would anyone justify "laying down a high volume of cover fire." during a self-defense use of a firearm? Someone please enlighten me.

        If a lawyer is attached to every transmitted bullet, I surely would not want to be the defender "laying down a high volume of cover fire."

      • Cover fire? <– Really? If 'a' warning shot (singular), is frowned upon, how would anyone justify "laying down a high volume of cover fire." during a self-defense use of a firearm? Someone please enlighten me.

        If a lawyer is attached to every transmitted bullet, I surely would not want to be the defender or associate "laying down a high volume of cover fire."

    • “relatively untrained” people are likely to empty the gun

      Just the opposite. Relatively untrained people only fire once or twice. Trained people fire many more rounds.

      • I read a couple of years ago about a study that found that cops (the only people who supposedly have enough training to be trusted with guns) frequently do full mag dumps and when questioned later honestly believe they shot 3 or 4 rounds. In fairness, I don’t think us unwashed heathens are immune to this phenomena either, but when someone is trying to kill you, counting your rounds seems rather trivial. As Detective Callahan used to say, ‘in fact, in all the confusion I lost count myself.’

  2. I imagine the “3 yards” detail is probably pretty reliable.

    Why? The motive for almost every attack would fall under one of these categories:
    — robbery
    — assault (including rape)

    That’s about it. And all of those crimes happen at contact distances. Thus, you would almost always find that you need to draw and shoot when your attacker is uncomfortably close.

    Sure, a small number of defenders would be shooting at an attacker that was trying to snipe from a distance. Those attacks are few and far between.

    As for the “3 shots in 3 seconds” points, I am having a much harder time seeing how that is reliable.

    • We still have to take into account on the stats would be skewed by cold blooded murder. You don’t need a lot of numbers in the 10-20 yard range to bring the average above three and the number of incidents that get to Trayvon Martin distance is likely low due to the difficulty of deploying a firearm while engaged in a physical struggle.

      Drive by shootings come to mind as having the possibility to seriously open up the distance. Ditto someone running away while shooting at you.

      I don’t trust any of the numbers until I see the data and how these numbers were calculated.

    • Yep. We see stats that say that rifles and shotguns are rarely used in crimes. Rifles especially. The bad guys use pistols or other weapons, knives, blunt objects, etc.

      A legit civilian dgu is going to be up close. I primarily rely on a j frame snubbie as my bad guy behave weapon. It’s short enough to make a bad guy grabbing it less likely.

      In a pocket holster few guns can be brought out as quickly as a hammerless j frame.

      And since I live in a mag limit state the round count issue ain’t really a thing.

    • Using the same reasoning, the 3-second thing is probably somewhat reliable, then. If most attacks are robbery or assault, those kinds of attackers are opportunistic predators, and probably not eager to die for their crime. Once the lead starts flying, they’re very likely to immediately break off the attack and haul ass (assuming none of that lead finds its mark and they are physically able to). I don’t know if that equates to just three seconds, but I’m also not sure the actual number is all that important. The key point is that the shit happens very quickly, and if you’re the unfortunate one having to suddenly deal with the aforementioned shit, the difference between three seconds and twenty is pretty academic.

    • i don’t know that I’d agree that most shootings are robbery or rape. Anecdotally, it seems, most gangbanger shootings are some sort turf war, revenge, or disrespect. I was on a jury in federal court for three gangbangers up for three murders, and two of the three were related to revenge for an ass-whuppin’ one of their fellow Traveling Vice Lords incurred. The other one was related to drug sales misdealing.
      One of the three they shot up a house and killed the guy inside. The other two were up close, one they shot a guy in a car, the other right outside of his car. As the recorded prison call said, “guess who got found stankin’?”

  3. I think it is a good question to raise.

    Though I suspect, that the 3-3-3 generalization of things happening quickly in close quarters will be supported.

  4. The scenarios for civilians, on the defensive, largely involve close, rapid, quick encounters. Unlike police we do not seek out trouble. It finds us. The smiling lad who asks “Sir, do you know the time” as he walks up and pulls a knife, is too close before the fight starts. If you are lucky you be able to push him off, clear the holster, and fire seversl shots while looking into his eyes and breathing his stink. Why so close? Because he isn’t an “imminent” threat at 21 feet. Civilian defensive shootings are always unexpected, closer, and faster than police. I tell my students to practice drawing every night and go to the firing range at least monthly.

    • If the thug gets within talking distance of you before you are aware of the threat you have already lost. He is ready and you are not.

    • Not to mention that we subject under MUCH higher degree of legal scrutiny than cops. A cop kills a man running at them with a knife from 21 feet and he’s legally fine. Charges won’t even be filed. A non-LEO does the same thing? Charged with first degree murder. Of course, a man running at you with a knife from 21 feet away means you are, at most, 1 second from being eviscerated, so clearly both of these theoretical individuals are well within their rights to open fire. But the cop can fire sooner and with less doubt, and with a far lower chance of legal retaliation.

  5. JWT, I assume you’ve seen the SOP 9 (s.69) NYPD report, old but interesting, summarized well here:

    I have distilled this from the studies: If I want to survive a handgun shootout, then I should allocate most practice to what W.D.M.Bell called “dry fire snap shooting.” In his day, in Africa, one could do this on a casual walk, like Thomas Jefferson was wont to do. Today I think it means use of sights, a laser sight, perhaps a red dot…and just picking a target and getting your EMPTY gun on it accurately as fast as you can, and pulling the trigger.

    I wish I’d downloaded it, but in the Law Enforcement Journal the FBI published an interesting study a few years back. The key conclusion in studying LEO/Thug shootings was this: Whoever gets their hand on their gun first, essentially gets their gun up first, wins. They also found the thugs had more shooting experience than they had assumed. The element of surprise was key.

    The article made such an impression on me that when I walk or drive the mean streets (rare, I live in a very safe township) I tend to pocket-carry a G36 pistol: I can hang my thumbs out of the pockets, four fingers in the pockets, Lord Mountbatten style, having three fingers on the grip. Meanwhile my actual carry gun is at my waist in a good holster, but often under a suit coat and an overcoat! The system works for me. The Glck 36 is very light. I put a 3″ by 4″ cardboard blocker in the pocket that holds the 36, which complete prevents printing. My pocket holster is probably unique. It is the core of a Blade-Tech Stingray. I flick it off with the shooting hand thumb in draw. This whole bit actual has some examplars in the carry items of the surprisingly well-dressed sheriffs of the old south-west.

    So the pretty guns are in the safe. A Glock 30S or 19 Gen 4 MOS/RMR is holstered for CC, and a 36 is often in my overcoat pocket. 3 rounds 3 seconds 3 meters? That seems a reasonable guesstimate. 10 rounds 5 seconds 7 meters seems like adequate preparation.

    The NYPD study would tell us that alertness, getting to cover, and a fast draw/first shot are more important than range scores, which did not correlate with success in actual shootings.

    I hope you’ll post any interesting supporting info you find.

    • A hammerless j frame is the quickest gun I’ve found to come out of a pocket with. Haven’t tried a compact Glock.

      I’ve heard of that cardboard trick elsewhere. If it works, it’s good.

      • There are pocket holsters (e.g. Recluse, Uncle George, deSantis SuperFly) that cover the outside of the gun with a flat sheet to avoid printing.

        To avoid accidents, especially with pistols that have light triggers, you should always carry in a holster that covers or blocks the trigger. Never carry loose in a pocket.

        • True. I found the cloth and leather pocket holsters slow. I like having the ‘blocker’ and the holster separate. The “core” of a Blade Tech STR holster is just that part of the holster, the kydex, that surrounds the pistol. In other words, the holster with the entire belt-attaching bit removed. A little sanding can reduce the lumpy side where the screws attached the belt bit.

          What is left is strong, covers the entire trigger, yet can very easily be drawn from. I’m as used to the pocket set in my overcoats as I am to to the Crossbreed Supertuck and clones for my main carry item.

          It is convenient that the STR core (and the CBST…) are nearly perfect fits for both of my most usual carry options, the 19 and the 30S. Small adjustments are made. I doubt I’ll change any of my handgun habits at this point, believing with Fairbairn that long-term practice draw of the same items in the same place in the same holster (or sheath) is one of the keys to effective reaction under stress.

          Agree, there is nothing like a shrouded/hammerless J-frame for shooting while the pistol is still in the pocket, or from under the coat when the pocket is cut to allow access to your holster.

    • I have read the study, and yes, it is old, but still has some good insights, and your conclusions are all valid. But this study, like the UCR, is focused on law enforcement only and this particular one is limited to NYC Police.
      It is backed up by a more modern study completed by the TxDPS, which points to the same thing as you’ve noted. That is, key aspect of training to be focuses on is getting the weapon into the fight as quickly as possible. About 15 years ago TxDPS started changing their draw and retention training to reflect this.

  6. The Tactical Professor has an article analyzing the FBI report. It never said 3,3,3 although there area a lot of encounters where this is true. Many Police-criminal interactions occur at short range because they have to close with the suspect to make an arrest. A private citizen doesn’t have to do that.

    • I carry an LCP in a back pocket holster, spare mag front pocket. I am a jeans and t-shirt kind of guy. I can hit targets reliably out to 20 yards. I don’t feel under gunned.

  7. No one is going to reference the analysis of 5 years of “Armed Citizen” articles?
    I refer you to the web page of where they reprinted the entire thing
    It totaled 842 incidents of armed citizens, no law enforcement
    The average number of shots fired was 2, with only 3 incidents of reloads.
    One of the reloads was shooting at an escaped Lion with a pistol!
    Average shooting distance was “arms length” without actually touching
    Most shootings took place in the home, and victims often had time to retrieve their gun from another room
    So 3 yards and two shots sounds right.
    No analysis of total time of the incidents
    Read it for yourselves

    • Great resource, but using it as a data set to determine distance, time, and number of shots is fundamentally flawed. The Armed Citizen, although valuable publication, selects their articles for their readers. That’s one of the reasons we see so many of the instances are home invasions. It is self selecting. For instance, how many of the articles include robberies and murders where the good guy/armed citizen fails and dies? Good data points, but the selection bias is far too great to get a distance-shots-time trend.

  8. Look up Tom Givens,

    He’s been training armed citizens for decades in the Memphis area. His database includes 60+ citizen incidents reported to him first hand by students, cross checked with police incident reports. One of his recent newsletters (available free on the website) updated his findings based on the database.

    Certainly not a huge sample, and it may be geographically specific, but at least it’s something.

  9. I think there’s allot of jack ass internet know it alls who don’t know the flying Fuck they’re talking about when it comes to guns and well… Just about anything else. The “333” thing may be true, it may not be. If you have serious questions, talk to some LEOs and citizens who’ve had to do it and learn through others experience. Instead of some loser with his finger up his ass has to say online.

  10. As mentioned upthread, Tom Givens generally agrees with 3/3/3 rule (although he says “steps”, not yards), and that’s good for me.

    He also says that FBI UCR is not a good information source because these are not typical civilian SD encounters. Better sources include BJS Annual Crime report and Victim Survey, TBI state crime reports, FBI and DEA agent shootings, Rangemaster shootings.

    • “Better sources include BJS Annual Crime report and Victim Survey, TBI state crime reports, FBI and DEA agent shootings, Rangemaster shootings.”
      Unfortunately, those reports do not include data on distance-shots-time for non-law enforcement shootings. Nor does the NCVS.

  11. Thanks for the 20mm footage. Never saw one with a suppressor, now I’m thinking, hmmm, where did I see that government reclamation site with that 20mm brass. . .

    (IMHO) EDC / DGU distances don’t really matter as much as your reaction time in the moment. I believe everyone would experience the “time period” differently depending upon many physical and mental factors, and several environmental factors. I believe it is determinable that (has been determined that, in general) most (U.S.) people tend to drive faster in light to moderate fog. They drive faster on their way to and from work (places that they travel to very regularly). I believe you can reach / touch any part of your body with greater accuracy with your eyes closed, than anything attached to your body (i.e., your eye glasses or a holster). And your level of surprise (experienced) messes with your perception of the passage of time (you lose your metronome, and time comes in fractions of seconds in slower slices, or it happens so fast that you don’t even believe that you experienced it). Further, high levels of surprise tend to put your brain into a micro shock-like condition, and paired with actual trauma, your brain wads the whole experience up and chucks it, and if you survive you have a blank spot on your memory.

    Either way [again, IMHO] if you have a gun, the first time you hear it “go off” should not be your DC use.

  12. Whatever. I don’t train for 300 yard rifle duels. I train for real world defense, which includes mostly, but not exclusively, defense in my home and defending myself on the street.

    Who cares if it’s 3-3-3 or 4-4-4 or 5-5-5? The correct metrics might be nice to know, but they don’t really matter.

  13. I hear all the time, and mostly from guys who carry revolvers, that six shots will cover the average DGU. Well, when you factor in all the DGUs where no shots were fired, it’s probably a lower average. But why base capacity on an average? For little or no more effort, you can carry a gun with the capacity to cover 99.99% of all shootings.

  14. To Jon Wayne Taylor.

    You might consult with Tom Givens who runs Rangemaster.

    Don’t know who he is?
    You should.
    He keeps stats of the actual shooting instances his students.

    I have heard that the Course of Fire Shooting requirements of the FBI have changed to basically the distance of a car.

    I don’t know actually what they train for.

    But I teach my friends to shoot at distances within their house.

    That is 0 to 5 meters at most.

  15. This is very interesting and gives one pause to think about how you cannot really over train or be over prepared for a potential SD situation. I do not believe there is any logic or statistic we can apply to predict exactly what that encounter will be like. I offer an example from a good friend, who shall remain nameless, who I had the privilege of practicing with at my local private range for several years.
    He is a career military gunnery sergeant, two tours in Iraq, with a short period in between as LEO here in the U.S. During his short time as LEO he had two SD situations, both of which resulted in him having to shoot the perpetrator(s).
    The first involved undercover work to arrest a drug dealer. During the deal, which took place inside a house, the dealer looked at him and said, “You’re a cop!”, grabbed a machete and came at him. He drew his weapon from his rear waist as he was backing up, and fell as the dealer raised the machete. His first shot, at about three feet, missed!
    The second took the dealer under the chin, and surprisingly, he lived. Mind you, this is a well trained shooter with thousands of hours behind the trigger and combat experience (so he knew what it was like to shoot under extreme stress) who could hit an 18” steel plate at 100 yards offhand, with any handgun we had at the range, with 8/10 regularity.
    The second was totally different. He was in his patrol car, early Sunday morning, pulled up to a quickie mart/gas station to get a cup of coffe. Was just stepping out when two guys who had robbed the place came running out, saw him and both opened fire while running toward him. He took cover behind the car, and fired back with his .40 caliber. He hit the lead shooter who went down from a well placed torso shot. The second shooter jumped over his partner and kept coming and firing. My friend hit him in the cheek, which seemed to make him mad, and he turned and ran away while firing back over his shoulder. The last shot my friend fired clipped him neatly behind the ear and he fell dead. The first guy tried to get up three times, finally bled out, and expired. When the scene was secured, it was found that each robber had emptied a fully loaded 17 round 9mm at my friend, 34 shots. Not one hit him. He THOUGHT he had fired 6 shots, it was actually 7, and for that one minor miscount he was put on a desk and investigated for murder. He was absolved when the coroner could find no powder burns on the one shot behind the ear.
    Each situation is different and all we can do is be prepared.
    My friend decided to sign up for another tour in Iraq because he felt safer there.

  16. I have been carrying for well over 10 years. Avid shooter, have yearly memberships in three ranges and belong to a club. I have tried to monitor all the shooting in the large city I live in. I do not recall any cases other than gang related that any one shot more than three times. And never where anyone had to use a extra magazine. Maybe the Police etc.
    So I ask how many civilians posting on this blog have had to do so? I doubt there are any. I doubt there are very few who even have been in a gun fight. I would bet, that more worry about having a high round count than their own diet and will not of heart failure before they get into a gun fight. I worry more about a Text Message driver than needing more than on magazine.

Comments are closed.