Looking deeper into the findings….

A few days ago I saw Unc had posted a link to this article on “Self Defense Findings”.

What struck me about the article was that it came from Claude Werner. I got to train with Claude a year ago, and as I wrote in my AAR:

After we finished FoF, we went inside for a presentation by Claude. Claude maintains a database of some over 3000 incidents of “gunfights” in America. This database of incidents has provided him with a great deal of information and perspective. Furthermore, he’s read the law enforcement records, such as the annual FBI reports. All of this has enabled Claude to really understand what “gunfights” are truly like, at least here in the USA. I refuse to spoil it by talking too much about it here, you’ll have to attend a class or conference to hear it. This alone was worth the tuition.

So this new article was interesting because, after having seen how Claude collects and reviews data, I thought there could be some good tidbits to take home.

You have to go read the article to understand what follows here. But just in the off-chance the referenced website goes away (the Int3rw3bz has a habit of that), I thought the information was worth archiving in a PDF.

My Thoughts

As soon as I saw this article, I shared it with my fellow KR Training instructors because we’ve all trained with Claude and know where he’s coming from. I started to formulate this article, but Karl beat me to a response by commenting at the original blog (I’ve been busy). Still, I wanted to write my thoughts down.

Given Claude’s nature and hobby for collecting and analyzing data, I certainly read the article with interest. But upon reading it, the conclusions felt in stark contrast to other data I’ve seen on “private citizen” self-defense incidents.


Claude’s data shows 52% of incidents happening at home, 32% in a business. That’s the #1 and #2 locations.

US Department of Justice Robbery locations in 2007 have the street at 43.8%, residence at 15.2%. This comes from the April 2012 Rangemaster Newsletter

So, you are almost three times as likely to be robbed on the street than at home, and in the home only accounts for 1 robbery in 6. Similar patterns exist for rape, aggravated assaults, etc. In fact, good locks, an alarm system, and proper lighting can reduce your risk of violent crime at home to very low levels.

This also jives with data on encounters with plain-clothes FBI and DEA agents. Furthermore, looking at Tom Givens’ own student incidents, almost none of them happened in the home. I have papers with Tom’s data, but am having a hard time finding them online since Rangemaster recently redid their website: good redo, but many old links broken and getting at all the newsletter PDF’s is tough. 😦  If someone wants it that bad, I’ll go find my papers and print them here.

The upshot is Claude’s data makes it look like most violent encounters happen in the home. But is that really the case? His data set is the only one I’ve seen that draws that conclusion. Thus, it makes me want to look at the data set. Claude’s data is coming from the NRA’s American Rifleman magazine’s “Armed Citizen” column. It covers 5 years from 1997 to 2001 looking at 482 reports. That leaves out the last 10 years, and that’s rather significant when you consider the changing landscape of “Gun Culture 2.0″… concealed carry has expanded rapidly in the past 10 years. Could that mean because there are more guns in public, we’ll have more “gun incidents in public”? perhaps. But more consider the NRA. The NRA is about gun rights, not pepper spray rights or karate chop rights, so is the “Armed Citizen” column going to cover stories where someone fended off a criminal with a right-cross to the jaw? Nope. So we can’t say what’s reported is necessarily indicative of crime in general: only what the NRA chooses to report, and then only from the submissions they receive or discover. And that’s the key to mind: this isn’t a cross-section of all incidents, just the incidents printed. I’m sure the NRA gets a LOT more submissions than they can print, just due to practical issues like physical magazine space in which to print the stories. Thus there’s going to be some editorial selection, that this story will be printed but not that one. Would the NRA print a story that makes gun owners look bad? How about a story where some “Tactical Tommy” fended off the bad guy with his AR-15? Or would it be better to pick a story where “the old white man used his old Colt revolver”? (sorry to stereotype). Does it look better to the public to have stories of home defense because that’s a bit less political, than it is stories of people toting guns in public and using them in that venue? And again consider, this is 10-15 years ago and what the landscape was like then vs. now.

I’m not saying the NRA is being biased as I have no idea what sort of editorial choices they make, but they have to be making them because only so many stories can be printed every month. To me, that taints the data set and limits the conclusions one can draw from it.


Claude’s data concluded that most, if not all, incidents happened about at arm’s length. To an extent this holds with other data, whether it’s the “3 shots, 3 yards, 3 seconds” mantra, or the 0-5 yards, or the “within a car length”. It’s generally because the attacker needs to be close to you either to injure you or because if they’re mugging you because they need to be able to talk to you. But if the majority of incidents happen in the home, according to Claude’s data, why are the homeowners even allowing someone to get that close? One advantage of firearms over other self-defense tools is their ability to overcome distance and be effective at distant targets. Why aren’t people taking up a fortified position in their house and shooting from a distance? Especially if, again as Claude’s data reports, people do not have the gun on them and have to go to another room to get the gun first? Why are they then coming back to be so close to the attacker?

One possible reason: “Defenders frequently communicate with their attackers before shooting.” And so, that suggests a few things to me. They feel in order to be heard they have to be near… but you know what? find your inner drill-sergeant and yell. Or if they can’t hear you, fine! Take up your fortified position and yell, because if they do eventually get close enough to hear you yelling, that’s far enough for them to come and get the message they’ll get shot if they come any closer. And that’s probably the other aspect of this. Most likely the communication before shooting is a lot of “don’t come closer, or I’ll shoot you” or “stay back” or other such things. But still, you can communicate this over a distance. Don’t go back to where you know there’s trouble, unless the trouble could be worse if you didn’t (e.g. spouse or child in there, etc.).

But what really got me about distance was Claude’s conclusion:

The perceived need for massive quantities of ammo, reloading, and precision shooting at distance is largely a figbar of people’s imaginations. There is simply no evidence to support the contention that any of those conditions occur during armed confrontation

That might be the case from the limited data set which Claude examined, and perhaps he’s meaning it within that context. But the presentation is such that it implies such conditions NEVER happen PERIOD in any sort of private citizen self-defense encounter. Thus you should never worry about these matters, because there’s “simply no evidence” to support they ever happen. And that’s wrong, and Mr. Werner might want to check with his friend Tom Givens for some data on this topic.

Which takes me to another reason for why people might draw closer to their attacker: they can’t shoot them from far away. If all you ever do is blaze away at a cardboard target that’s 3 yards in front of you, you’ll probably be pretty good at that distance. If you never shoot your pistol at targets 25 yards away, how much confidence do you think you’ll have if now suddenly you have to make that shot? Under stress you’ll default to what you can already do, and given a drive for success you’ll work to put things more in your favor, so if that means getting closer then you will. If more of these citizens had adequate training, training that pushed them to do things like shoot at 15 or 25 yards on a regular basis, how might this data be different? And again, with the sharp rise in Gun Culture 2.0 and more private citizens seeking formal training, what would data from the last 10 years show by contrast?


Throughout Claude’s examination of the stories, he finds that you just don’t need much ammo. He said:

If the defender fires any shots, most likely it will be 2 rounds.

And then the above comment that you don’t need massive quantities of ammo, and saying that a snub revolver (typically holding 5 rounds) is all you’d need.

When Karl commented on the blog posting, he mentioned how Givens’ students shot from 1 to 11 rounds. The blog owner replied taking Karl’s statement to task, and he was right in doing so because Karl left out one important part of Tom’s data: the average was 3.4 rounds. The FBI/DEA data holds about the same too.

So no, it’s not THAT much different from Claude’s data, but it’s still different.

But remember what average is, statistically. There were enough incidents that required more rounds, including at least 1 incident that needed more than 2 snubs worth of ammo, more than what a 1911 traditionally holds.

What’s hard to read about Claude’s findings is again the way it is presented, that 2 shots is all you need, you’ll never need more than 5 to take care of anything. This is simply not the case. Sure that might be the average, but boy… if you opt to train to just the averages, how do you think you’ll feel when you get to be the one statistical anomaly? Look at the edge cases in Claude’s own data and piece them together. The largest group had 7 VCA’s, and you’re going to need more than a 5-shot snub to deal with that many attackers. Again, this is about playing to statistics, about assuming you’ll be alright because the averages say. Does anyone say “gee, I wish I had LESS ammo”?


Claude writes:

At this distances, even .22s and .25s are highly immediately lethal.

A revolver, even J-frame, is perfectly capable of dealing with almost all of the incidents. The ones which were beyond the capabilities of a five shot revolver would be best deal with by a shotgun, anyway.

For those who do not practice, a revolver is far preferable to the autoloader because of the revolver’s simpler manual of arms. Eighty per cent of gunshot wounds are self-inflicted. Guns are handled many times more than they are shot and so safe gunhandling qualities are much more important characteristics than its ability to be shot accurately and reloaded quickly. Revolvers are much less likely than autoloaders to AD in the hands of novices.

Yes, I don’t want to get shot by a .22. I know a .22 could kill me. It’s not my first choice, but it’s better than no choice.

A shotgun would be better than a J-frame for sure. But I can’t carry a shotgun. Oh wait… this data shows that most incidents never happen out on the street where one might need to carry a gun. Hrm. Someone reading this data could draw the conclusion that we need personal protection in the home, but that we don’t really need it outside the home because most incidents happen inside and rarely outside. Thus why carry. Not sure that’s a good conclusion to allow people to draw.

And dealing with a snub? Folks, you have to know who Claude Werner is. He’s one of the masters of the snub revolver. He shoots IDPA matches and wins them with his snub. He was chief instructor at the Rogers Shooting School for a number of years. I’d say Claude’s abilities with a snub are far superior to the average citizen. Snubs are hard to shoot, and Claude knows that (again, see my AAR of his snub class). Yes, one reason I carry a snub as a back-up gun is because of the manual of arms: it is simpler, and if I had to give my BUG to someone else because the fur was flying that badly then I know at least I can expect them to “point and click” without having to worry about levers and gizmos and malfunctions and such: just keep pointing and clicking. But they are still very difficult to manage, very difficult to shoot well, all having very long and heavy triggers. Revolvers are not my first choice nor recommendation for anyone.

If we’re worried about AD’s? My suggestion? Get training. Understand and abide by the rules. And never think you’re above having an ND happen to you.


I don’t take odds with the data Claude collected, in terms of what he did and the summary picture that came out of the data set. I think it’s all reasonable collection and analysis of what was there. Furthermore, it does paint an interesting picture that’s worth knowing.

What I find problem with is some of the conclusions and suggestions in here, like that a snub is sufficient, that long distance shooting never happens, that reloads don’t happen. My fear is that someone could look at this data and use it as justification for developing a training program, or to justify they don’t need any training at all. “Why should I train how to reload? they never happen, and besides it’s a pain to reload my snub because it’s so small.”  I believe this is Karl’s fear as well, tho it wasn’t perhaps expressed well enough in his comment (given the response from the blog owner). Claude may have stated at the beginning: “You decide what suits your needs best to solve this type of problem.” but to the untrained and unknowing, they’re going to look at Claude’s data, conclusions, and suggestions as authoritative and will likely use his data, conclusions, and suggestions in formulating what suits their needs best — because they’re a n00b and don’t know what their needs are and how to satisfy them! That’s the problem.

There’s something to be said for understanding all the data, and how that generates some averages and yes how that can and should influence our training both in terms of what to train and what not to train. If most self-defense incidents end up fitting that “3 shots, 3 yards, 3 seconds”, it would stand to reason that’s something to first ensure you can do (e.g. can you clean the “3 Seconds or Less Drill” consistently, constantly, and on demand). That doesn’t mean your training should only encompass that sort of work, but if you can’t do that stuff it’s best to master it before you go on to things like group shooting at 25 yards, if self-defense is your shooting goal. But you should eventually move on to being able to shoot groups at 25 yards and not be satisfied with what the data says, what the averages are, because while certain data sets may not support a need for it, who knows… you may get to be the lucky one that establishes a new data set.

16 thoughts on “Looking deeper into the findings….

  1. Great post.

    Statistics often lull us into a false sense of security. Even knowing that in a given data set the most shots fired was 11 doesn’t mean that you won’t need 12. We can derive where to focus most of our attention, but we need to make sure we train for the unlikely scenarios as well.

    • That’s my main point. I just don’t want to see people look at the data Claude presented and take it as gospel or justification for how they should train. Given the way the data was presented, I can see potential for people to base their training upon it or use it as justification for not needing to train either parts or completely.

      Just don’t want to see that happen because no, you don’t know how the flag will fly for you.

  2. As with so many things, perspective is everything, and thanks for giving us your perspective so I could post it at The Thinking Gunfighter blog. My perspective is not that Claude was saying anyone should train a certain way, or that everyone should just carry around a snub. It was more along the lines that lots of the things we worry about the most probably aren’t that important most of the time. Sure, there will be outliers and exceptions but we need to remember they are outliers and exceptions and train or plan accordingly. And the more information we get the better we can conduct thta training and planning.

    • Indeed.

      Claude did make some definitive statements, and from those statements one could draw conclusions that you don’t NEED to train for shooting at long distances, or to reload, or whatever. That’s the problem and key source of contention here, by both myself and Karl (KR, that commented on your blog): that risk of drawing conclusions and justifications about one’s training based upon this data, without looking deeper into things.

      I think in many respects it’s matter of how the data was written/presented, not the data itself.

      And while data and statistics should shape our training, it should not define it. And that’s what we don’t want to see: we don’t want to see people defining their training by averages. Take things into account, keep perspective, but master all you can.

  3. Great Blog. I commented on “The Thinking Gunfighter”, we had a recent case in San Diego CA where the CCW holder came home to a burglary/home invasion and fired his weapon and missed. I believe in the garage. The CCW was pistol whipped with his gun by the over-powering suspect (Large Parolee). The CCW was an older man (60 +).

    As a retired LEO (30 yrs), the numbers don’t surprise or shouldn’t surprise. CCW is rare in CA so most gun owners would only have legal access to their weapons on their property. The “potential” to be a victim on the “street” is clearly higher than the home. The study shows that Americans are clearly either not arming themselves for protection or are not being allowed to arm themselves. Less than 500 cases of armed defense out of the thousands of robberies/assaults every year is pretty astonishing. Just flip around the “good guy” with the ”
    bad guy” and see how the stats change regarding gun use and crime.

    I couldn’t agree more with you regarding home defense. The lights in my house are going on, my drill instructor voice will be demanding the intruder to leave before I deploy force. My clear warning will help in any lawsuit that may occur due to force being used. I will give the intruder the opportunity and CHOICE, to leave before force is used. I have my plan and it’s purpose is not only to help my family survive, but to survive the civil suit if required. My weapon is in an area where I would either have to “retreat” to get it or I’m stuck and can’t get to the front or back door to escape. My 870 is basic and is not “tactical”. I don’t need the inside of my house or bedroom looking like the inside of the local police supply store.

    If armed with a firearm, we will be bringing the gun to the fight. That responsibility alone should give pause and not be taken lightly. Training between the ears and rounds going down range do not have to be inclusive to themselves, both can be accomplished at the same time.

    Thanks for allowing me to comment and writing on such an important topic. Cheers, Dennis

    • Hi Dennis. Thank you for taking the time to comment.

      I think it’s a combination of things. People aren’t allowed to arm themselves, but this (in terms of legal impediments) is decreasing. Then when people can arm themselves, they aren’t. The number of license holders is still a low percentage of most state’s populations. And then even when people can do it, many times they don’t, e.g. the have a license but only carry when they feel they need to, as if they have some ability to foresee when the flag will fly. But again, these things are changing. More people are getting licenses, more people are carrying, more people are carrying more often, more people are seeking formal training, so it’s all getting there, slow but sure. I also think the general population is starting to “get it” more, which helps. I mean, look how much more we’re seeing guns and gun-related things in mainstream culture — and presented in a good and positive way.

      You’re right tho… what matters more is training “between the ears”, such as having a plan. This will take you further because yes, being armed brings great responsibility.

  4. I think they were both good articles. I am of the group that has a carry permit but chooses not to do so. My reasons are simple, I live in an area of the country that it just doesn’t make sense for me to do so. I’m 55 years old and although this isn’t a really good reason not to carry, I don’t because I’ve never felt the need in all these years. I’ve never been afraid in my surroundings. I know this is stereotyping at it’s worst but I’ve been what I call a ‘biker’ (riding Harley’s and hanging with ‘biker gangs’ (gag)) almost my entire adult life. A rough crowd as considered by some but if I was ever in a position to feel threatened it was by any group of college age football players not the seemingly ‘lowlife’ element. Anyway, good article.

  5. Awesome analysis. The primary thing which hit me – and which you detailed far better – was that the “Armed Citizen” columns is in no way a database of crime or shooting incidents. As such, it’s a useless assemblage of anecdotes signifying “not much of anything” aside from which cases came to the editors attention, and which of those were thought worthy of publishing. What can be more accurately determined is a cross-section of what kinds of cases the NRA editors LIKE.

    • Well, I still think there’s something useful to take from the analysis, but we have to keep in mind the data set and everything about it and how that then affects the conclusions drawn.

      Basically, take it for what it is in whole… don’t try to read any further into it than what it is, but still consider it an interesting data point and analysis.

      • Exactly. Just because the data doesn’t lend itself to analysis on some things doesn’t mean it can’t be used for analysis of anything. Pretty much ANY type of data collection has some flaws in it but the trick, as you say, is take it for what it is.

  6. Glad to be introduced to your blog by this topic! I think you’ve done a good job of explaining some valid objections.

    The one thing I wanted to touch on is the notion of ammunition capacity in reported incidents, as much is made of the 11-round shooting that Tom chronicles. Even in the DVD he did for the NRA Personal Defense Series, his description of the incident suggests that the defender probably didn’t need the 11 rounds that were fired. The assailant had fallen against an open car door during the exchange and was wedged in an upright position; the defender continued to shoot to slide lock because the perp hadn’t dropped to the ground, which was interpreted (correctly, I believe) as still being a threat. It’s entirely possible that the incident was in fact resolved in the first couple of rounds, and it was only in the lull precipitated by slide lock that the defender actually figured out what had happened and that he didn’t need to shoot any longer. If he’d had only five rounds in his gun, the only change might have been that he found this out sooner!

    If we take this somewhat ambiguous incident out of the data, I think we’d find that the number of shots which were *physically needed* to stop the rest of the attackers in Tom’s data set was much closer to Claude’s report.

    This isn’t to say that everyone should carry nothing more than a J-frame, of course. I’ve said many times, in classes and in print, that snubby revolvers should be considered expert’s weapons and chosen only with a clear understanding of their limitations. I have been known to carry a Glock 19 and a spare full-capacity magazine myself, but intellectual honesty compels me to acknowledge that in the vast majority of reported private sector defensive shootings, the five rounds in a J-frame have proven to be more than enough. In fact, I’ve yet to find a reported incident in the private sector where a lack of capacity resulted in grave injury or death to a defender, and am truly interested in seeing any factual accounts that anyone may have along those lines. (My interest is in data, not dogma; if the facts contradict my opinions, then my opinions must logically need changing!)

    Keep up the good work, and thanks for your thoughtful analysis.

    • Thank you for stopping by and for the kind words.

      You are right about Tom’s 11-round student. This is the problem with averages — it only takes 1 to skew everything. Maybe we should get median and mode from Tom as well (I’m serious about that).

      But the thing is, we’ll never know. And certainly we can find incidents where “a bunch of rounds” were needed. Tom’s data set is awesome, but not the be-all-end-all.

      But you ultimately right — most reported private sector defensive shootings are well likely to be handled with just a 5-shot snub, or some other “small-capacity” gun (like common semi-autos — think Kahr — carried by people because they want a small gun for carry).

      I just want to be careful to set the bar too low. That is, if data shows it’s typically 3 rounds, could that create a bad standard? I mean, we’re human… we’re lazy for the most part. If people see they don’t need to fire many rounds, they don’t need to worry about reloads, or malfunctions because they don’t happen often in reported incidents well… could that become justification to train to a low standard? I would hope not! As well, could that become fodder for political means, be it training regulations held to a lower standard, or justification for bans and limits?

      So I won’t deny the realities of the data, I just want us to be mindful about how the data is used, y’know?

      • I tend to agree with you and don’t mean to imply that any one statistical analysis should be the absolute arbiter of all training. In fact, I cringe when I see “defensive” shooting courses where the number of rounds in a string have been predetermined by the instructor (or the curriculum.) Instead, I believe in teaching students to shoot a variable string of fire, ending when they’ve visualized their threat ceasing, in order to get them out of the habit of training to always shoot 2 or 3 rounds. (Because this might cause them to expend a large percentage of their gun’s capacity with each string of fire, they also get more experience in recognizing the stimulus of the empty gun and linking that recognition with the skill of reloading.)

        Analysis can, however, help drive our resource allocation: if we know that a certain large percentage of attacks happen under certain circumstances, that’s where logic suggests we first focus a large percentage of our limited training resources. Once those circumstances have been adequately addressed, we can then expand the range of circumstances in which we can utilize our learned skills. This isn’t to say, for instance, that we should never shoot at long ranges with a handgun (as some people mistakenly interpret), but rather that we understand it’s simply a less commonly needed skill and therefore probably shouldn’t be allowed to consume an inordinate amount of our training resources. As you said, it’s all in being mindful of how we use the data available to us.

        As to seeking median & mode for the data: given that our society doesn’t understand math all that well, I’m not sure it would make much of an impact; after all, most people believe that “average intelligence” means exactly half of the population is dumber than they are!

        • We’re very much on the same page.

          I do agree that we should use this data to drive how we train. It shouldn’t dictate, but it should provide guidance. I mean, if the typical gunfight is “3 shots, 3 yards, 3 seconds” and chances are high that’s what you’ll experience, then yes we should work to ensure people can deliver that sort of performance under that sort of pressure first… before skills like 25 yard groups or 4 second El Presidentes.

          I’ve been very fortunate to have Karl Rehn as my teacher, boss, and mentor in this area. He’s established a training curriculum that goes along these lines and is really well-formed towards getting people the right things at the right times, and progressing in a logical and effective manner.

          • Good comments, and that was in part the genesis for The Thinking Gunfighter blog, using data to drive the training. Excellent point that data shouldn’t dictate but should instead guide our training and practice. Is it really worth spending 25% of our time learning to do a fluid, rapid tactical reload, for instance, given that reloads are relatively rare and given that there are other reloading methods? Given the limited resources available to most of us it is probably better to devote time to what is probable instead of what is improbable. Once we have the probable down then we can look at outliers.

Join the discussion!

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.