How many rounds to carry?

That question came up on a Facebook group I’m a member of recently. In response, I referenced my Armed Citizen database. The question was asked about my methodology, which is a fair question. I’ll address it my forthcoming eBook about the Armed Citizen but I want to first post the Introduction, which addresses the journey I have made about the Armed Citizen and my analyses thereof.



This book is the result of the overlap of several very widely different topics and experiences. As is often the case, as more information comes to light over time, perceptions can change.

During my time in the Army, I held several different intelligence (S2) positions. These largely involved information collection and analysis duties, not ‘spyguy’ stuff. The purpose of Intelligence in the military and government is always to facilitate decision-making. Having to provide and defend a cogent analysis of not only the information collected but the conclusions I drew from it was a formative experience for me. Information collection was only the beginning. From there, it had to be processed and turned into a usable product that decisions could be based on.


As I wound down my military career and entered the civilian world, I got into the commercial real estate business. As a Research Director for several different real estate firms, my S2 training and manuals were very useful to me. At the same time, the transition from mini-computer (Wang) to PCs in the business world was beginning. My boss was an extremely astute businessman and recognized the value of databasing information early on. Being able to construct my own databases allowed me to do several projects that were particularly influential in the way I looked at information.

One of the projects was to database the contacts that the brokers in our office used to develop business. Our firm’s business model was territorial with each broker having an assigned property type and area. To see how well this worked, my boss had me collect each broker’s contacts by Zip Code and create a map of where the contacts were in relation to the broker’s chosen territory. This process was very similar to the Intelligence Preparation of the Battlefield (now Battlespace) products I had prepared in the Army. The results were surprising to everyone involved. In almost no case were the majority of the broker’s contacts in his or her territory. Some were nearby, which was understandable, but in many cases, they were widely scattered and even far away. The brokers themselves couldn’t believe it until I showed them the actual maps.

What this showed me was how inaccurate conclusions based on data that isn’t properly disaggregated can be. Their information was written down in their Rolodexes with every contact date annotated. That system told them very well what the level of their contact activity was. What it didn’t provide was much information about how well they were following their business plan. Aggregating the data and then disaggregating it by location instead of contact name and date told a much different story.

Another database I had to create was of proposed and completed deals. Creating this database gave me a much better insight into the numerous factors that make up a transaction. Proposed rental rate, length of term, size of the space, etc. were all captured when the brokers proposed a transaction. Eventually we would enter whether the deal closed or died. That database gave our company a firm understanding of what the market was actually doing across the city and in the various submarkets. Instead of speculation about what actual rental rates and terms were, we had a very clear picture.

Training I took impacted my thoughts also. I took Massad Ayoob’s Lethal Force Institute I in 1991. Having a measured and structured component to training was an eye-opening experience. Similarly, when I started training with John Farnam of Defense Training International, I got a lot of good information, both formal and informal. John was kind enough to give me a copy of W. French Anderson’s book about the FBI Miami Massacre. The book provided a superlative example of an in depth analysis of an armed conflict.

The next leg of my experiences developed when I started shooting IDPA in 1998 and then started an IDPA club. A number of Match Directors and I were discussing how to develop stages every month for our matches. Stage development is a constant pressure for any Match Director to keep the matches fresh and interesting. Someone suggested that The Armed Citizen column of NRA’s American Rifleman magazine might be a good place to start. I had been tearing the columns out of the magazine for years but never paid close attention to them. So I dug them out and looked through them in greater detail. My response to the other MDs was that almost all of the incidents were less than five shots and a lot were only one or two. Many of them had no shooting in them at all. The general consensus was the round count wasn’t high enough and the situations weren’t complicated enough to make interesting scenario stages.

My conclusion was different though, so I started designing what I called Armed Citizen Scenarios for my matches. There were several ways to adapt the incidents into stages. One way was to put multiple strings into a stage. For instance, if a Citizen was wounded in the arm in an attack, I would have one string shot with both hands and a second string shot with the Dominant Hand Only. Or, when only one shot was fired at one criminal in the actual incident, I would specify a failure drill (two shots to the body and one to the head) on all the targets.


The Armed Citizen topic interested me enough to create a database all 482 of the incidents from the column for the period 1997-2001. The incidents were remarkably devoid of ‘ninjas coming from the ceiling’ and ‘face eating meth-heads.’ As I had done with the deal database, I broke out as many different characteristics (at home, in a business, number of shots fired, etc.) as I could. With the database populated, I ran a series of pivot tables and produced a short study of what the characteristics and outcomes of the incidents were. Although there were methodological issues with it, fifteen years later, it remains the only study of its type I am aware of. Like a vampire that won’t die, it continues to be widely referenced and reproduced on the Internet.

TAC 5 year w tables

One of the criticisms of my 1997-2001 study was that the NRA ‘cherry-picks’ the incidents to portray the actions of Armed Citizens in the most favorable light. Although the nature of what the Citizens might have done wrong was never really specified, I accept that as a valid critique. Only Positive Outcomes are reported in the Armed Citizen.

Flash forward more than a decade to the 2014 Rangemaster Tactical Conference, where I am an annual presenter. My colleague Craig Douglas threw down a challenge to me. “You should do a presentation on ‘Bad Shootings’ next year.” It was a virgin topic and gave me an opportunity to counteract the ‘cherry-picking’ aspect of the Armed Citizen. I accepted the challenge and casually started gathering information.

Be careful of what you wish for. The broad array of what I came to call Negative Outcomes really surprised me. The categories I broke them out into are:

  • Brandishing/threatening
  • Chasing after the end of a confrontation
  • Downrange failures (shot an innocent while shooting at a threat)
  • Intervention (Proverbs 26:17)
  • Lost/stolen guns
  • Mistaken identity shootings
  • Negligent discharges, including self-inflicted gunshot wounds and Unintentional shootings
  • Police Involvement, e.g., getting needlessly arrested
  • Poor judgement
  • Unauthorized access (generally by small children)
  • Unjustifiable shootings, including warning shots

The categories are far from being the lurid list of ‘gunfights lost’ that those who objected to the 1997-2001 study probably expected. Rather than being tactical failures, most are simply the result of poor gunhandling, lack of familiarity with the law, or out and out carelessness and negligence. My list of such incidents is shockingly long. The only really noticeable category of tactical failures was what my colleague Tom Givens calls ‘forfeits,’ i.e., not having your gun when you need it.

Chapter takeaways

  1. There is a process to data collection and analysis.
  2. Information that isn’t written down and then analyzed in written form is prone to error. The human mind has a remarkable capacity for memory but that capacity can be disorderly and easily misinterpreted.
  3. Defensive Gun Uses by Armed Citizens tend to be uncomplicated affairs.
  4. Defensive Gun Uses have discrete characteristics that can be broken out for broad analysis.
  5. Negative Outcomes rarely consist of ‘gunfights lost’ but more often are negligence related Unintentional Shootings and Unjustifiable Use of Weapons. The exception to that rule being not having a gun when it’s needed.


14 responses

  1. […] Source: How many rounds to carry? […]

  2. I remember when all the firearms people for Federal agencies were given that FBI shooting book–I tore through it as an 18yr old (I don’t think it came out until 1988 or so…at least I didn’t see it until then). I can’t remember if it was them or the LA robbers who ate raw coffee grinds (found in their stomachs during the autopsy), but I remember all the autopsy photos. I also bought a .40 USP compact for my first pistol thanks to that book.

    As for “how many rounds,” I know Tom Givens has done research on non-police CCW shootings, and it wasn’t that many shots fired–like 3-5 maybe. Still, he carries a Glock 35 for his CCW. I wonder what his students who’ve been in firefights used and how many shots were fired.

  3. Reblogged this on RealDefense and commented:
    Claude Werner’s work is top notch.

  4. Forthcoming eBook? When will it happen? All you write is golden.

  5. On the positive, I just love the negative outcome work you are doing.
    Having been in the military and a LEO I find that lessons learned from incidents taken over time tend to allow us students to avoid a large breadth of possible negative outcomes in our life.

    I wonder if you could expand your thoughts on the the negative aspects of the data set? Not only do I agree with the criticism that the data set from the magazine is self limiting to the outcomes but I believe that the idea of positive and negative (binary) outcomes misses the value of a sliding scale of outcomes. Not using a sliding scale is often what messes up people’s decisions in reference to training, practice, tactics and equipment. It often becomes a cascading set of errors.

    The positive things that happen in the magazine do nothing to tell us what we should do in reference to equipment, tactics, etc. Unfortunately people tend to use those stats to make positive or proactive decisions. Those positive decisions do not fully apply to us as the negative lessons learned do.

    Us humans are horrible at using statistics and even worse at interpreting them. We often use average when we should use median. We often make decisions with median when we should analyze the data around that median point for the usual low and high factor of the data set (bell curve or power law) and develop our training around those data points points. Doing that can cause us then to use the same training and tactics to expand that curve to cover incidents that are farther away from the median.

    The framing of why a person didn’t shoot (zero rounds) and why a person shot twenty rounds can frame the discussion of how many rounds are needed for us as students. If I misinterpreted the data and use the median of 2 then a five shot revolver is just fine for a two person gunfight. That of course is ignoring much of the actual data and thus ignoring the very positive, “Why” analysis.

    Note that I’m not saying a person should not use a 5 shot. But understanding the, “why” of the data is important in that my needed ability, tactics, etc are different if I am shooting a 5 shot with two bad guys and a 15 shot with two bad guys. Joe above mentioned the Givens’ data being similar but he carries a high cap pistol. The, “why” of that is more important than if we decide to carry a 5 shot or a high cap pistol.

    1. That of course is ignoring much of the actual data and thus ignoring the very positive, “Why” analysis.

      What actual data is that?

      1. In this case it would be the data that you used to compose the table in the article.

  6. Grant W. Hiesterman

    Love your articles, research, and logic.

    I think there may be a typo near the bottom of this most recent communication:

    4. Defensive Gun Uses have discrete characteristics than can be broken out for broad analysis.

    You may have meant “that” rather than “than.”

    Warm regards, Grant

    G. W. Hiesterman (612) 386-8359

    “A short pencil is better than a long memory.”


  7. Reblogged this on Stuff From Hsoi and commented:
    Data is good, and I always appreciate Claude’s collection and analysis of it.

  8. I carry every waking moment and when out of the house I have 46 rounds available to me. This is not because I think it takes so many rounds to resolve a defensive situation, but because we are at war. We never know when that war will be brought to our store or mall or church or street. That war may involve several trained individuals with rifles and if I cannot escape, the gun battle may go on for a while.

    I don’t expect every man to carry according to what is appropriate for wartime. Only the responsible ones.

    1. Personally, my favorite weapon for wartime is the fragmentation grenade. Great attention getter when ‘Up Yours!’ isn’t strong enough language. I can think of several incidents where a frag would have solved the situation nicely. They’re kind of a nuisance to carry in civilian clothes, though.

  9. How many rounds to carry? Tom Givens has it right when he says “I have never met someone, who has been in a gunfight, tell me they wish they’d had a lot less ammunition”.

  10. Reblogged this on disturbeddeputy and commented:
    When in doubt, grab another mag. Too much ammo, food, or water are self-correcting problems.

%d bloggers like this: