Photo credit: ibropalic via / CC BY

The problem with gunfights is that almost no one has actually been in one.  This causes many people to believe anything that someone says, who has.  This is a dangerous proposition.  They may be right on the money, or they may have just been lucky, because, hey there are 5 guys who win in Russian roulette right? How do we fix this problem?

Well, we can do the next best thing and apply a rigorous analysis on videos of actual gunfights, in order to see what similarities exists in these fights, better preparing ourselves, should we find ourselves in this situation.  Keep reading to see the surprising results of that analysis.

Before we get into the nitty gritty, I’d like to thank the folks over at Active Self Protection.  They regularly feature self defense and gunfighting videos on their Youtube channel, which made my research way easier.


Let’s talk about how exactly I did this.  I’ve had this idea for a while that most folks aren’t really training correctly for a gunfight, but I didn’t really have any data to prove it.  As a former Intel Officer in the Marine Corps, I have a hard time just taking things on faith, I like to see the data for myself.

This led me towards looking for videos of actual gunfights, which led me to Active Self Protection.  They have amassed hundreds of videos, covering a huge variety of self defense scenarios.

I then selected a set of 30 videos and analyzed them for 14 traits.  Check them out below.

 Range in Feet Between Good Guy and Bad Guy

Number of Assailants 

Obscured Draw (hiding draw from assailant)

Delayed Draw (i.e not draw immediately on recognizing threat)

Number of Hands on Gun (two hand vs one handed shooting good guys)

Grappling (physical contact during gunfight)

Shooting and Moving 

Event Time in Sec 

Assailant Direction of Attack (clock face method) 

Use of Cover 

Police Officer (Civilian or Police Officer) 

Location Type (store, gas, station etc)

Good Guy Injured (any significant injuries)

Overall Success (good guy injured or killed = failure)


The next step was to watch all the videos, and record the data.  Once this was accomplished we could then move on to the important part of this process, the analysis.

If you like this kind of evidence based article, then drop your email in on the right, and join the team.  You’ll get the latest articles sent directly to you.  If you don’t, I’ll assume you date your sister.


After you watch just a few of these video’s, you can quickly start to see patterns once you clearly define the variables you are looking for, as we did above.


This was estimated visually using my Mk1, Mod 1, calibrated eyeball.  Across all 30 scenarios, I estimated that the average engagement distance was 9 feet, with some much further, and some at contact distance.

Number of Assailants 

This was the most surprising to me.  In 11 scenarios there was only one assailant.  There were also two scenarios with 3 and 4 assailants respectively. However, incredibly, there were 17 scenarios with 2 assailants.  

Prior to this analysis I would have thought that the vast majority of assaults and robberies, which is what these scenarios generally are, would have been one on one.  This is simply not true!

Obscured Draw 

This was a key predictor of success in any engagement.  In 73% of the engagements the good guy, hid his draw from the bad guy(s).  This also flies in the face of conventional wisdom, where we think of the good guy racing to draw his gun, and engage the bad guys in clear view.

Here’s a great example of obscuring the draw.  This guy is cold as ice.

Delayed Draw 

This was closely associated with the obscured draw, and likewise, was present in nearly all successful engagements, with the exception of uniformed police officers.

There were 53% of scenarios where the good guy saw the threat, waited, then delayed his draw.  Often times they delayed their draw, and also obscured it as well.

In every successful scenario the good guys, delayed, and obscured their draw.

Number of Hands on Gun 

This is a variable that I’ve seen a lot before, and had assumed that two handed shooting was a little rarer than the movies would have you believe.  There were only 57% of scenarios where the good guy used two hands to fire their weapon.

I also noticed that those with formal firearms training, such as police, and security guards, tended to revert to two handed shooting when they had enough space to do so.


This was somewhat surprising to me as well, as only 37% of these scenarios involved physical contact during the gunfight.  This was most often light pushing, or attempting to control the attacker’s limbs.  There were none where strikes preceded shooting.

Shooting and Moving 

This was another factor that was common in nearly all fights.  There were 73% of scenarios where the good guy was either moving and shooting together, or moving, then shooting from multiple positions.  This is a great example of a very dynamic fight.


Event Time 

The vast majority of engagements were roughly 20 seconds or less, but a few longer scenarios made the average engagement 26 seconds from first shot fired to last.

Assailant Direction of Attack 

This was a key finding.  There were 10 scenarios where the assailant approached head on from 12 o’clock.  There were 6 scenarios from 9 o’clock, and 4 scenarios from 3 o’clock.  It’s clear that attackers often times approach head on, but the most dangerous approach was from 6 o’clock, which often lead to injuries for the good guys.


Use of Cover 

Only 27% of these scenarios found the good guy using cover.  Often times there was no cover available, and the engagement was too quick, or violent for this to be a viable option.

Police Officer 

There were 40% of scenarios where the good guy was either an off duty police officer, or and on duty officer.  This is probably for several reasons, one of which would be the high likely hood of investigation for any officer involved shooting, even in other countries where many of these video’s came from.

Location Type 

There were 5 scenarios in each, gas stations, stores, and restaurants respectively.  The other scenarios were outside, or in other random locations like stairwells.  Only 1 scenario was a home invasion.

There is a big caveat to this one; however, because we were analyzing video we would only be able to see examples that had cameras accessible.  Random street crime, or crimes in rural areas aren’t likely to be recorded. So take this one with a grain of salt.

Good Guy Injured 

There were 5 scenarios of the 30 where there were significant injuries or death on the good guy’s part.  Often times, these were very violent ambush attacks, with little real chance of success.

Overall Success 

This one is somewhat subjective, but for the purposes of this analysis only scenarios where no good guys/ by standers were injured were counted as success.  There were a total of 83% of engagements that were successful, according to this criteria.

Another caveat here, folks.  Since these videos were not truly randomly picked, or uploaded to Youtube, we can’t say this is the overall success rate of gunfights.  Most folks don’t like watching the good guys being killed, so you’re less likely to see these types of video’s in general.


It definitely took many hours to review all the video, and analyze each scenario for 14 traits, but it was worth it.  I think we’ve all learned that a few of our preconceived notions about concealed carry, and gunfighting in general were wrong.

I was most surprised that the 2 assailants scenarios were the most common out of these 30 engagements.  If you actually watch the videos on Active Self Protection, you’ll start to notice that during robberies, there is often times a gun man, and a grab man who’s job it is to actually collect people’s valuables.  I don’t think many people think about that, and only expect to have to face one bad guy.

We’ve also learned the incredible importance of delaying, and obscuring our draw, which allows us to choose the optimal time to engage the threat.  Of course, there are times where the threat is coming right now, and there can be no delay, but if you can delay and obscure, you should.

You can also expect these events to be around 20 seconds, and they will probably require you to grapple, then shoot with one hand.  If there is more distance between you and threat, you are also very likely to have to shoot on the move.

Training Recommendations and Tips

Shooting on the move, one handed shooting, and gun grappling are all very advanced shooting techniques.  First, I would recommend you learn the basics of flat range shooting, and then move onto these more advance principles.

Without further adieu, here are the top 11 concealed carry tips based off of the analysis we have done above.

  1. Carry a high capacity weapon, ideally 8 rounds or above. 
  2. Be able to access your gun with one hand. 
  3. Keep your head on a swivel, checking your 6, the most dangerous area. 
  4. Delay your draw. 
  5. Obscure your draw. 
  6. Do not engage immediately unless you must. 
  7. Practice pivoting and engaging threats 
  8. Practice gun grappling and one handed engagement within 3 feet 
  9. Shoot from cover if available 
  10. Practice shooting and moving at man sized target at 3-5 yards 
  11. Validate all live training with force on force training

Ideally you would train these dynamic methods using force on force with training partners, or in a class.  I would also recommend getting your own airsoft pistol(Amazon Affiliate Link), ideally in the same model as your carry gun.

This will allow you to get comfortable in a dynamic environment, which is what you will actually face on the street.  Strangely enough, there were no engagements where the bad guy was standing in a well lit environment, and then turned 90 degrees to face you, while you blasted away.

Square range training has it’s place, but keep in mind what kind of fight you are actually training for!

Post your thoughts, comments, or questions below.




  1. Hi Jake, we’re working on analyzing all 1000 of them for better overall statistical validity. Honestly I don’t think #9 is quite right, as most people are using concealment not cover, and normally not until the first clash is over.

    Thanks for being part of the ASP community and for making some consideration of what actually happens in gunfights!

    • I appreciate the comment! I’d be really interested in your analysis of all 1000. It would definitely bring some more validity. I think you have a point about #9, but it could have been the video’s I selected were slightly different then the norm as compared to the rest of your 1000 video’s. I did my best to minimize selection bias, but I think your analysis would do better getting rid of it. Great work!

  2. Given the delayed draw and obscured draw observations and conclusions, were you able to get any insight into the most advantageous carry position. The officer in the restaurant was carrying behind the hip and that worked very well in that particular scenario. I don’t think appendix would have been as good. How about other scenarios?

    • Very good question. I had that same thought as I carry mine in the 5 o’clock position watching that video. I think the honest answer is that if you did carry in the appendix, you’d have to wait for a different opportunity to delay and obscure, or create your own. I think the real keys are to bide your time, and be able to access that gun with one hand. I saw no one do the classic scrunch your shirt weak hand, then draw.