"Reliability"

[Fon1961]

Reliability gets referenced a lot in a wide range of forums/threads. I'd like to see some objective data. Do gun manufacturers typically provide reliability data for their guns? If so, are there industry standards for determining and comparing reliability? If manufacturers don't provide this are there any independent entities that do?

[Lon]

Other than the results from a large .mil or large .gov testing process, I have never seen a manufacturer provide reliability data. Here's the best place I've seen for unbiased (read: not done by the manufacturer of the pistol) reliability testing.

http://pistol-forum.com/showthread.php?9-2-000-Round-Challenge

[GardoneVT]

Reliability gets referenced a lot in a wide range of forums/threads. I'd like to see some objective data. Do gun manufacturers typically provide reliability data for their guns? If so, are there industry standards for determining and comparing reliability? If manufacturers don't provide this are there any independent entities that do?

The closest entity that I can think of would be European proof houses, and they only test firearms with a limited number of overpressure rounds. Better then nothing, but not the same as a long term test.

Way I see it, setting up a "Consumer Reports" type test has several major logistical issues.One, who has the motivation to set it up? Guns and Ammo wont do it, because theyre an advertising vehicle for the industry.

Manufacturers wont do it on a mass scale, for obvious reasons. The only way this works is if a wealthy individual writes off millions of dollars . Testing one 9mm pistol for 50K rounds would cost $15,000 , and that doesnt include range costs and labor compensation for the time of the shooter-because guys like TLG are hard to find, and Cleetus aint qualified for this kind of work.

Thats 15K+ per GUN; and HKs variants alone would run up the bill fast. Multiply $15,000 times every modern 9mm produced today, and that's one heckuva ammo bill. Talk about lead poisoning.....

Then follows the challenge of updating the list due to parts changes, so that people dont accidentally buy a turd because last year it topped the ranking . As TLG observed years back, gun companies change parts all the time without public notice-and any one of them can turn a reliable machine into a bucket of Taurus.

[Fon1961]

Other than the results from a large .mil or large .gov testing process, I have never seen a manufacturer provide reliability data. Here's the best place I've seen for unbiased (read: not done by the manufacturer of the pistol) reliability testing.

http://pistol-forum.com/showthread.php?9-2-000-Round-Challenge

This thread is awesome, thx for posting it....skimmed the entire thing....need to go back and read the details now lol

[Fon1961]

Manufacturers wont do it on a mass scale, for obvious reasons. ...

I agree with you in general, however, don't you think every new weapon design would have to go through some type of a formal qualification/validation which would include both non operational and operational testing. Not just for performance assessment but also for liability reasons....? In addition can a mfg really afford to take a chance mass producing parts that may not function together? Maybe a lot of it's done by modeling and simulation?

The main concern I hear about pistols is reliability, I'd just like to see data.

[GardoneVT]

I agree with you in general, however, don't you think every new weapon design would have to go through some type of a formal qualification/validation which would include both non operational and operational testing. Not just for performance assessment but also for liability reasons....? In addition can a mfg really afford to take a chance mass producing parts that may not function together? Maybe a lot of it's done by modeling and simulation?

The main concern I hear about pistols is reliability, I'd just like to see data.

Tanfoglio has youtube videos showing their QC process, but not every manufacturer is that forthcoming .Some firms operate on a business model which says if only 2 people in 10 actually shoot their guns enough to see a problem,it makes more sense financially to deal with the rare warranty claims then to invest in QC which drains profits and goes unnoticed by their customers.

[Fon1961]

I could buy that some companies would take a gamble from a warrantee perspective...but not one of the major manufacturers. I also think that liability is a huge factor. I would also like to believe that this data could be used to objectively demonstrate the superiority of their product or highlight what changes that were needed to make it so. I've been trying internet searches to learn more but mostly just get posts that make claims with very limited or unscientific data/rationale.

[RJ]

Maybe it would be possible to look at the problem from the other direction, i.e. What are the requirements?

For example, for the original Glock pistol, what were the reliability figures it had to meet?

I realize wiki is not authoritative, but a quick search shows this entry:

http://en.m.wikipedia.org/wiki/Glock

"No more than 20 malfunctions are permitted during the first 10,000 rounds fired, not even minor jams that can be cleared without the use of any tools.

After firing 15,000 rounds of standard ammunition, the pistol will be inspected for wear. The pistol will then be used to fire an overpressure test cartridge generating 5,000 bar (500 MPa; 73,000 psi). (The normal maximum operating pressure Pmax for the 9 mm NATO is rated at 2,520 bar (252 MPa; 36,500 psi).)[11] The critical components must continue to function properly and be up to specifications, otherwise the pistol will be disqualified."

Would it make sense that this is an objective measure of 'reliability'?

[Malamute]

I could buy that some companies would take a gamble from a warrantee perspective...but not one of the major manufacturers. I also think that liability is a huge factor. I would also like to believe that this data could be used to objectively demonstrate the superiority of their product or highlight what changes that were needed to make it so. I've been trying internet searches to learn more but mostly just get posts that make claims with very limited or unscientific data/rationale.

The problems that Glock has had with a couple model variants over time would indicate that they were willing to live with sending guns out that weren't up to their usual reliability parameters. ISP's guns being one of those problems. I don't think they ever admitted that there was actually a problem, even though they took a fair sized lot of guns back that they were unable to get to run right. Just an example of a generally accepted high standard of reliability that doesn't always extend across the board on all their products.

I think establishing your acceptable baseline is going to be necessary to make judgments on reliability. One persons "outstanding!" or "flawless" level of shooting and reliability acceptance may be anothers weekend practice level, and barely register on their radar.

[Al T.]

Test of the firearm, sure, but also the ammo. :eek:

[Fon1961]

Establishing or acquiring performance requirements would be a great and you could argue that they are objective and comparable...but without performance data I'm not sure how useful it would be.

The impact of ammo would need to be considered if there was a malfunction, failure, ect ... agreed

I'm looking for requirements and reliability vs those requirements ...I guess that would take all the fun out of debating it tho lol...

I did find an interesting (based on a quick look) comparison site ... compares several parameters between guns ...reliability not being one of them ... http://gun-tests.com

[Malamute]

Not a specific performance requirement, but interesting reading,

http://pistol-forum.com/showthread.php?9-2-000-Round-Challenge

[Fon1961]

Not a specific performance requirement, but interesting reading,

http://pistol-forum.com/showthread.php?9-2-000-Round-Challenge

It's a great thread and very interesting IMO

[TCinVA]

Manufacturers do their own internal testing on various products they offer, but that is usually trade secret stuff they don't release publicly. About the best you will get is the testing requirements for military contracts and the like, but even that has limits.

The BATFE contract for .40 Glocks was a pretty rigorous test, and yet one of the reasons the FBI is moving to 9mm is that the same weapon that won that contract has proven to be problematic in mass issue.

[Alpha Sierra]

One could suffer paralysis by analysis

Or one could be practical, purchase a pistol from a reputable manufacturer and simply go shoot. If reliability problems arise, get them fixed or sell the pistol and buy something else.

[Hambo]

One could suffer paralysis by analysis

Or one could be practical, purchase a pistol from a reputable manufacturer and simply go shoot. If reliability problems arise, get them fixed or sell the pistol and buy something else.

Roger that. I wouldn't hop on the wagon of a brand new design, but pistols such as the 9mm Glocks and Beretta 92F have shown over decades how reliable they are. Even utterly reliable pistols suffer parts breakage with use and time. Pick a proven design, buy some spares of whatever has been shown to break, and drive on.

[TCinVA]

That's about the best you can do.

[justintime]

working as a range master and seeing thousands of rounds go down range everyday, I see every manufacture fall to failures of all kinds. Almost everyday I see a multitude of different malfunctions and I fill up a coffee can a week of bad ammo from reputable but mostly garbage ammo that has failed. All I can say is practice clearing malfunctions!

[ReverendMeat]

I think instead of trying to test one pistol for, say, 10,000 rounds it would be a more informative to test 10 randomly selected guns of the same model for a thousand rounds apiece. For designs that are known to be reliable, testing more samples would be more likely to turn up more manufacturing defects/QC issues, which tend to manifest in the first couple hundred rounds.

[GardoneVT]

I think instead of trying to test one pistol for, say, 10,000 rounds it would be a more informative to test 10 randomly selected guns of the same model for a thousand rounds apiece. For designs that are known to be reliable, testing more samples would be more likely to turn up more manufacturing defects/QC issues, which tend to manifest in the first couple hundred rounds.

For a statistically valid selection, one would need at minimum 30 pistols, and preferably more then that of each model.

Otherwise we run the risk of sample bias-ten guns out of tens of thousands won't reflect the likely reliability of the production line.

We also cannot assume that a brand is reliable enough to merit not testing, as once again were introducing personal bias into the testing process.

That's an expensive enough enterprise, but we'd also have to test the magazines -as a Glock 17 with a standard capacity may not necessarily perform the same way as a Glock 17 loaded with a 10-round magazine, and folks in restrictive states need reliable data as well. 30+ magazines for each tested model of firearm will add up fast-and in the case of HK this project would require a Federal Reserve loan to fund.

[Fon1961]

There maybe more data out there than you think. I agree the best testing is most likely done by yourself at the range.

http://smartgunlaws.org/design-safety-standards-for-handguns-in-california/

[ReverendMeat]

For a statistically valid selection, one would need at minimum 30 pistols, and preferably more then that of each model.

Otherwise we run the risk of sample bias-ten guns out of tens of thousands won't reflect the likely reliability of the production line.

We also cannot assume that a brand is reliable enough to merit not testing, as once again were introducing personal bias into the testing process.

That's an expensive enough enterprise, but we'd also have to test the magazines -as a Glock 17 with a standard capacity may not necessarily perform the same way as a Glock 17 loaded with a 10-round magazine, and folks in restrictive states need reliable data as well. 30+ magazines for each tested model of firearm will add up fast-and in the case of HK this project would require a Federal Reserve loan to fund.

Well yeah, you're right of course. I still think 1000 rounds through 10 guns is still more significant than 10,000 through one gun. Nix what I said about designs that are known to be reliable, pick 10 Tauruses (Taurii?) and see how many fail in 1000 rounds. Or 30 guns for 500 rounds. Just use only the standard capacity mags that normally are included with the gun. State compliant mags, extended capacity mags, pro mags, ignore. While people who post on this forum certainly would be interested in much higher round counts, to the general gun-buying public 500 rounds is a lot. Obviously this rout is going to be a lot more expensive but in concept it'd still return more pertinent results (again, as far as the general gun-buying public is concerned) than anything with a sample size of 1.

[ToddG]

Statistically valid reliability testing is far, far more difficult and more expensive than one might think. Essentially, I'm not aware of any entity that does it. Not gun manufacturers, not agencies, no one. Some get as close as they can but if you've ever been involved in the simple shooting of, say, 20k rounds through each of five guns per model from, say, three manufacturers it gets ridiculous in terms of time/manpower consumption and, of course, ammo cost.

Then you need some way to factor in ammo-caused malfunctions, shooter-induced malfunctions (do you count that against the gun because people shoot guns and some guns are more prone to shooter-induced stoppages than others, or do you try to eliminate the human element for the sake of pure mechanical reliability alone?), secondary vendor issues (sights fell off because sight manufacturer machined them wrong... been there, seen that), etc., etc.

Regarding something discussed earlier in this thread, it is a mistake to assume that every part redesign/change is tested to original degrees of intensity. For example, a company might produce the Pistol Model 2014 and shoot ten of them to 10k rounds each. In 2016 when they decide to go from tool steel to stainless for the sear, they may not test it at all because an engineer somewhere tells them the test is unnecessary. And for all I know, that might be legitimate. I'm not an engineer. But having seen such changes made in multiple companies and having lived through the calamity that followed more than once, I really do believe you can never be sure that the Model 2014 you get today will be as good or as bad as the one you'll see two years from now.

[Alpha Sierra]

http://smartgunlaws.org/design-safety-standards-for-handguns-in-california/

This is your reference? You might want to re-think your sources.

[Alpha Sierra]

I really do believe you can never be sure that the Model 2014 you get today will be as good or as bad as the one you'll see two years from now.
Is there any consumer product that you can say the opposite about?

[GardoneVT]

Is there any consumer product that you can say the opposite about?

A large difference between firearms versus other goods is the product lifecycle. Handguns , unless they're very defective or poor sellers, stick around a long time.

IPhones, laptops, TVs, cars, etc usually are redesigned totally within 5 years if not sooner. By comparison, the Beretta PX4 turned 10 this year.

[Will_H]

A large difference between firearms versus other goods is the product lifecycle. Handguns , unless they're very defective or poor sellers, stick around a long time.

IPhones, laptops, TVs, cars, etc usually are redesigned totally within 5 years if not sooner. By comparison, the Beretta PX4 turned 10 this year.
Cars are generally a long cycle time, closer to a decade with a mid cycle refresh, that is limited to relatively minor visual changes, but that's really semantics.

It does make one wonder though. What other durable goods change at a rate as slow as firearms? Refrigerators?

[Fon1961]

This is your reference? You might want to re-think your sources.

No, it was just an example of data that's being collected, it's not "my reference"