What Is RELIABLE?

21-Jun-12 – 16:40 by ToddG

When someone says his pistol is reliable, what does that mean?

I’ve met people who defined “reliable” based on the amount of ammo they carry every day. Someone who carries a Glock 19 with a spare mag wants 31 rounds straight without a stoppage and he’s happy. Of course, that’s a mess from a statistical standpoint. The sample size is too small to draw meaningful conclusions. It might sound smart, but do you really want a gun that is up against its performance envelope shooting such a small number of rounds under ideal range conditions? Probably not.

Others will tell you that they’ve never had a stoppage of any kind! But again we have to ask whether that’s a perfect run of 100 rounds or 10,000 rounds. What did he consider a stoppage? How many times did something happen that he blamed on limp wristing, bad ammo, or some other gremlin other than the gun itself? Data matters.

When I think about reliability, I think about it in two ways:

  1. What it takes to “prove” a gun is reliable enough to carry.
  2. How reliable a gun is long term.

For me to be comfortable carrying a new gun, it needs to shoot at least 200 rounds of the same JHP defensive ammo that I’ll carry when CCW’ing without a single stoppage. Ordinarily, I’ll put at least 500 rounds of practice ammunition through the gun first… mostly so as not to waste the more expensive JHP ammo on a gun that might be having a problem. If the gun cannot make it through that 500rd of FMJ-style ammunition, I’ll examine it and attempt to fix anything either by working on it myself or returning the gun to the manufacturer. Then I’ll start the cycle over again, 500rd of FMJ followed by 200rd of JHP, minimum.

I’ve made exceptions in the past to the 500rd “warm up” before shooting the carry ammo, usually for backup guns like my S&W 442 j-frame or my Ruger LCP pocket gun. They still get 200 rounds of JHP, though. Yes, it’s expensive. But because I test the ammo during my normal shooting routine, it’s really just the difference in price between practice ammo and the JHP. Over the long run, I’d rather have confidence in my pistol’s reliability than a few extra shekels in my pocket.

Of course, if you shoot any gun enough you’ll eventually have a stoppage of some kind. While many shooters either ignore it or suffer from selective amnesia afterwards, if you really want to understand how well your gun runs you need to keep honest and objective records about any instance when the gun fails to go through its entire cycle of operation properly.

My personal threshold for long-term reliability is 2,000 Mean Rounds Between Stoppages.  Now, I’ve had very few serious pistols turn in reliability that poor. Even my endurance test Glock gen4 17 with its problem childhood and broken springs has turned in almost 71,000 rounds of service to date with a total of 18 stoppages (a MRBS of almost 4,000). The HK45 endurance test gun had a MRBS of 16,667. The P30 went 7,608 MRBS before it started to choke due to the cracked frame at a little past 91,000 rounds. The M&P9 scored an incredible 20,778 MRBS with two recorded stoppages and one spring breakage that prevented the gun from functioning normally.

Part of my ongoing reliability evaluation involves not just my practice ammunition but my carry ammo, as well. At a minimum, when my gun stays loaded from the car to the firing line I’ll make sure the first round I fire is the one that’s already chambered. (This also helps break the habit of repeatedly chambering the same round which can eventually lead to bullet setback and a dangerous overpressure situation) Frequently, I’ll shoot my cold drill (the F.A.S.T.) with carry ammo. And once in a while I’ll check zero or do some other accuracy work with the JHPs, too, to make sure I know where the bullets will be going if I need them.

Finally, I have one more personal rule when it comes to reliability: I won’t carry a gun if it’s had a stoppage within the last 200 rounds. So if I’m at the range and the gun fumbles, either I put at least 200 more rounds through it without a problem or I carry my backup gun on the way home. Because even if it’s a one in a million fluke, I want to be sure the gun hasn’t developed a breakage or other systemic problem that could go unnoticed.

Which raises the issue of back-up guns. I’m not talking about a small second gun that you carry. I’m talking about having an identical copy of your everyday primary gun that you can pull out of the range bag and shoot if something goes wrong with your main pistol. Is having a second gun expensive? Yes. But if I had to choose between having three different types of guns for fun or having a matching set of  carry guns, well, I can find other ways to have fun. Another popular alternative is to have a dedicated training gun and a separate, identical dedicated carry gun. That way  no matter how much you abuse your gun during practice you’ll always have a clean, top condition pistol in your holster when you’re off the range.

Again, all of these numbers are bare minimums. My personal guns get shot a lot and if something goes wrong, I tend to put a lot of rounds downrange before trusting one again.

Train hard & stay safe! ToddG

  1. 31 Responses to “What Is RELIABLE?”

  2. I really never considered what to do after a malfunction to reestablish trust. I’m pretty unforgiving when it comes to failures. 200-rounds after a failure to regain trust is a new rule that I may have to borrow from you. Nice write up!

    By gtmtnbiker98 on Jun 21, 2012

  3. Great article, provides a blueprint for others to follow.

    By Jesus Banda on Jun 21, 2012

  4. Well said and another good example why it is wise to acquire two or three identical pistols. In addition, it is further proof why the M&P is better than both Glock and HK…

    By DocGKR on Jun 21, 2012

  5. Okay Doc, the statement of M&P being better than HK is a bit of a stretch!!

    By gtmtnbiker98 on Jun 21, 2012

  6. I’d say that an identical gun isn’t necessary. Similar, yes……identical, no. If someone has a P30 and a P2000sk, or G19 and G26, M&P and M&Pc, P99 and P99c, ect, I think there’s enough overlap with the platform to maintain competency while also providing a gun that may serve in a different function than your primary as opposed to sitting in a safe/go-bag.

    Obviously if you’re on a tactical team or a competition that would present some severe disadvantages if your primary goes down….though I think it’s a good note for our civilian concealed carry peeps.

    By Tyler on Jun 21, 2012

  7. When my guys come to me and want advice on what pistol to buy I make a couple of recommendations and then tell them to pick up a case (1K) of ammunition for the piece. I tell them to lube the dickens out of the gun and to shoot the first case of ammo without cleaning. My reasoning is that any “break in ” that the gun needs will be accomplished, and they will gain some experience with manipulaitons, and some trigger time prior to showing up to qualify. All of my duty weapons (pistols, rifle, and shotgun)have had this procedure. Of my officers the ones who listen generally have very few problems, while those that don’t may struggle.

    We had an XD on the line several years ago (after Springfield bought them, but before they became popular) that the officer, during a qual, drew and got a “click” with a live round in the chamber. He had bought the gun, but couldn’t see the importance of practicing with it prior to a qual. There was a shortage of amrorers, and the gun had to go back to Springfield for repairs. Kinda soured me on the pistol.

    By Pat on Jun 22, 2012

  8. Todd,

    We think alike in this area.

    People think I am weird for shooting my carry ammo, sometimes just because, like at an IDPA match.
    Carry ammo is expensive dontchaknow?

    I happen to think a guy needs three handguns that are identical twins;
    One to carry after proving reliability.
    One as a spare of #1 that has been proven reliable and basically sits on stand-by.
    3rd gun to shoot the hell out of at the range.

    If you carry you might use your gun, if you shoot your guns you will eventually wear one out or break it.
    Either way you’ll be out a gun for an unknown period of time.
    Best to have a spare in the safe that you can grab and holster up instead of wondering what to do next.

    By Chuck on Jun 22, 2012

  9. I’m curious – how does cleaning factor into this? If you have to clean the gun every 500 rounds, but with that cleaning it makes your 2000 rounds between failures standard – does that make it reliable but picky and annoying, or unreliable?

    By ford.304 on Jun 22, 2012

  10. DocGKR:

    M&P’s are great guns–if you have all the equipment necessary to detail strip them, which as far as I can figure includes not only a sight-pusher but probably a hydraulic press also (though, who knows, maybe some C-4 could get the extractor pin out). I think I’d take 4000 rounds between stoppages instead (though I’d really like to have the M&P’s safety).

    By SteveJ on Jun 22, 2012

  11. I might be pointing out the obvious but… Just by buying two of the same make and model of pistol doesn’t mean they both are going to be equally as reliable. It might be possible to statistically infer they’re both as reliable, but sometimes life will hand you a lemon. So when mantaining two identically setup pistols its important to shoot both of them.

    By Dropkick on Jun 22, 2012

  12. I guess inferences = assumptions.

    Methinks the inference would be that the “carry” gun would have to pass whatever reliability test the user chooses before it is “carried”. Then, if one chooses, said pistol could be shot little, carried much. All this while the “training pistol” takes the abuse as the full-time work-horse.

    And while I do not know the good Dr. Roberts, I imagine at least a small smile when he typed –”In addition, it is further proof why the M&P is better than both Glock and HK…”

    By Bill Lance on Jun 22, 2012

  13. Reliable is being able to clear a malfunction and continue through with what you were trying to do in the first place. Everything will fail; how you deal with it is what matters.

    By Steve B on Jun 22, 2012

  14. Steve B,

    So if a pistol gets a double-feed or stovepipe once every mag then it’s still reliable because it’s easily fixed with remedial action?

    According to the definition of reliability, that simply can’t be true. Reliability is a person or systems’ capability of performing without error or fault. In the case of a firearm, that means functioning without the firing cycle being interrupted.

    By Tyler on Jun 22, 2012

  15. you can buy your 3 m&p !!! i will stick with my 1 GLOCK

    By scc on Jun 22, 2012

  16. @ford.304
    Your threashold is yours. You ulitimately have to make that call. I have not owned a gun that would not go 1000+ rounds without a stoppage, without cleaning, and with adequite luberication. I frequiently run my Kimber hard, shooting very cheap ammo (wolf, Privi, AE, all of which shoot very dirty). I will usually only clean when I know there is an inspection coming up, or on a six month rotation.

    I would have to check the book, but in 9K+ rounds I think I have had 4 magazine related failures (threw them away and the problem went away) and 3 failures due to out of spec rounds (the last lot of Federal AE that the department bought has had 3 rounds deformed to the point that they would not chamber… I use a chamber guage on my duty ammo but not practice ammo).

    Sooo, did I have 7 failures of the gun, and thus by these standards an unreliable gun? Or did I have magazines that I wore out that eventually caused problems, and bad ammo? I am very comfortable saying that My gun has gone 9K rounds without a weapon related malfunction. Springs are replaced at regular intervals and Extractor is about to be replaced as a precaution (at 10K). Detail stripping/cleaning with a more rigorous inspection every 6 months.

    By Pat K on Jun 23, 2012

  17. I have a jar of factory ammo that has had problems in several calibers. Primers missing/put in sideways/backwards, bullets sat too deep, rifle rounds inconcentric, case mouths peeled back, ect.

    Magazines are disposible. They are all numbered and at the first hint of a problem they get tossed.

    By Pat K on Jun 23, 2012

  18. One other reason for a backup pistol: IF you are involved in a DGU shooting, AND the police decide they need your piece for evidence. Depending on the jurisdiction, there is no telling when or if you might get it back. What do you carry in the meantime?

    By Mike Settles on Jun 23, 2012

  19. Tyler — Agreed that an identical gun might be demanding more than necessary. But the farther we get from that (changing triggers, sights, holsters, etc.) the less suitable it becomes. Carry a G19 with the same trigger and sight setup as my EDC G17 in the same holster would not be a concern on my end.

    ford.304 — The less maintenance a gun requires to function properly, the better as far as I’m concerned. But if your gun needs to be cleaned every 500 rounds and that’s ok with you, then cleaning it every 500 rounds should be part of the reliability proving.

    Dropkick — Yes, as others have said, the backup needs to be vetted just like the primary.

    Steve B — Cannot agree with your sentiment. Knowing how to clear a stoppage is important, but not having a stoppage in the first place is far better. The two aren’t mutually exclusive and the more reliable my pistol is, the less chance I’ll need to resort to fixing something broken.

    Pat K — Very good point re: what counts against the gun and what doesn’t. I record everything but make reasonable educated assessments as to whether something might be the gun’s fault. I tend to err on blaming the gun, though.

    By ToddG on Jun 23, 2012

  20. My Steyr S9 has had 1 failure to go into battery in 8 years (rd count unk.) In a class I borrowed the equivalent sized Glock, with which I was lucky to get through a magazine without a malf. (Good practice for clearing stoppages I guess but bad juju in a match or serious social work situation) The glock’s owner had no reliability problems with it. A lady friend bought a Sig 229 at the gun show, and it functioned perfectly for her. Again I could not shoot a magazine without a stoppage, and neither could the instructor at Insights when she took it to a class. In both cases I took care not to limp wrist.
    In today’s enlightened society we no longer believe in angels whizzing in the touch hole, but I am at a loss to figure this out.

    By revjen45 on Jun 23, 2012

  21. Anecdotes aside, I trust my life the the astronomical improbability of being struck by lightning while going through my business. (fingers crossed after I make this post…=) but I would never trust to exempt myself from mortal consequences with odds as great as 1 in 200 or even 1 in 1000.

    If your gun doesn’t shoot when you pull the trigger, you die. Anything short of a bang when the trigger’s back needs to be understood and resolved to the tone of 500+ rounds in one cleaning period with not a hiccup. (I don’t care what kind, as long as they match the profile of your defensive ammo. for me, comprable fmj with the same stats from the same company will do…)

    If that sounds like too many, you probably don’t shoot enough or train enough to justify carrying a weapon.

    By Johnkard on Jun 24, 2012

  22. I don’t follow the hydraulic press comment on the M&P. I’ve field stripped my own using a hammer and punch.

    By Rob M on Jun 24, 2012

  23. Err make that detail stripped

    By Rob M on Jun 24, 2012

  24. Reliable? One of my pistols has >5000 rounds through it with exactly one failure – a stovepipe – in the first 500 rounds, when my 12 year old daughter first shot it and limp wristed her grip.

    Will the next shot go bang? Likely.

    By mikee on Jun 25, 2012

  25. As a reliability professional, you have to understand both type 1 and type 2 errors to select a good measure. Type 1 error is when you pass your gun when it fails, type 2 error is when you fail your gun when it passes. There are several measures we use. Here is one that may be of use:

    Logistics reliability: Probability that your gun needs a repair measured in mean rounds between malfunction
    Statistically, we measure this with a simple chi squared function to an 80% confidence level, meaning you will incorrectly fail a good gun one time in 5.

    MTBF (80% lower confidence level) = 2R/(chiinv(0.2,2*n+2))
    with R being the number of rounds
    n being the number of fails used to calculate the degrees of freedom in your chi square formula or table.

    If you run your gun until it malfunctions, you use the same formula, but you don’t have the “+2″ in the degree of freedom.

    Logistics is always exponential, if you fix your gun after it malfunctions, either by cleaning it, or by replacing the malfunctioning part (which may be a bad magazine, a bad round).

    By DonM on Jun 25, 2012

  26. We had 2.5 ton trucks that the left tail light showed a high failure rate, but the right tail light showed a low fail rate. How could that be? The lights were identical, and both came from the same parts bin as they were installed on the truck….

    When ever the truck convoy stops, the good drivers get out and check their load, grabbing the tail gate, and putting their foot on-the-left-tail-light.

    Oh. We modified the truck to beef up the light installation to create a strong and convenient step.

    Moral: Different usage environment leads to different measured reliability. You get different numbers from range measurements compared to everyday carry. That difference can amount to 10% or more. That is why range measurement reliability events that demand 2000 rounds with no function won’t give you meaningfully greater confidence in reliability for everyday carry. Better would be to ‘carry’ to the range, and keep track of your first mag full of rounds several days in a row.

    By DonM on Jun 25, 2012

  27. “…no malfunction” that should be

    By DonM on Jun 25, 2012

  28. If you demand a very high probability of having a functioning firearm, say greater than 99 in 100, you should probably be carrying a second firearm. That way you have two chances to find a good one.

    Gun 1: 99 chances of success in 100 demands for operation
    Gun 2: 99 chances of success in 100 demands for operation
    Gun 1 OR Gun 2 (either one gives good service) 9999 chances in 10,000

    Compared to Gun A with 199 in 200: gives 9950 chances of success in 10,000

    By DonM on Jun 25, 2012

  29. “Gun Saves Your Life” is a system.

    Having the gun work is only part of it.
    Having the bullet(s) work on your assailant is part.
    Having the bullet(s) hit the target is part.
    Having sufficiently few assailants that you and gun can deal with it is part.
    Having the time to use your gun is part.

    “The Magic of the Drop” is one of Fredric Remington’s famous sketches. It doesn’t help much to have 3 digits (0.999) of reliability in your gun when you only have 1 digit (0.9) of reliability in any of the others. The different reliability numbers are merely multiplied together.

    By DonM on Jun 25, 2012

  30. I have done the training gun with a pristine proven reliable backup identical carry. My carry gun has 1,325 flawless round and has been fired only 200 times since I purchased my identical range gun a couple weeks ago which now has 714 rounds through it. Its very nice to shoot hundred of rounds down range and as soon as you’re finished have a clean reliable identical weapon to strap on.

    Both are USP45s

    By Greg P. on Jun 25, 2012

  31. Interesting. I have a revolver that has maybe a thousand through it thus far. No failures of any sort. Hard to see it failing any time soon.

    By Gamma on Jun 25, 2012

  32. I’ve never owned a revolver that wasn’t reliable. All problems were traced to bad ammo. I have seen service revolvers choke, but the guns were very worn and did not get the PM they should have had. I have seen probs w newer wheelguns, most due to added safety locks.

    A few yrs back the USMC was only asking for 300 MRBS for the 1911s they were looking to buy…

    IIRC, the FBI SWAT 1911s did about 10K MRBS in testing (7 or 8 in 80K/20Kx4).

    By MikeO on Jun 29, 2012

Sorry, comments for this entry are closed at this time.