It's good to see this thread reopened. I have some catching up to do. First, 44 AMP.
This is where we run into problems with statistical models. Because statistics where a longer, heavier trigger pull prevented an ND simply do not, and cannot exist. We may get a collection of individual anecdotes, but not useful data of any kind, because NO ONE reports an ND that did not happen.
I agree that the kind of direct empirical data that would be ideal to have do not exist, nor will ever exist until technology advances to the point where all guns have the equivalent of a black box embedded in them. I trust I won't live to see that day.
In the absence of perfect data are we powerless to conduct a meaningful analysis? Of course not! Are there any data we can start with? I'm unaware of any civilian data that would not be horribly corrupted precisely because of the fatal flaw you point out -- most citizen NDs are likely not reported, and those that are are not compiled anywhere that I know of.
Police are required to report their NDs, and I assume most do, but I'm sure some are not captured. So a given police department should have data for that department, and that could be a starting point for extrapolating to civilian handgun use.
As far as I know few if any police departments report their NDs to a central authority that compiles and integrates all the data by gun model and trigger pull weight. Someone recently posted a link to a 4-year study of shootings by federal law enforcement agencies, including the number of people shot during enforcement activities (presumably good shootings) and the number shot during non-enforcement activities (from NDs during training or otherwise on the job). My impression was that the fraction of ND-related shootings was awfully high compared to the total, and most of those occurred during training. If I understand what these data represent they are flawed because NDs that don't hit anyone are not included.
As an illustration, lets say you are doing the bad thing, finger on the trigger, and you fall, stumble, get startled, knocked off balance, whatever, and you clench your hand on the gun and on the trigger.
With a DA gun, your accidental pull might only bring the hammer back part way, not firing the gun. When that kind of thing happens the usual response is "wow, got lucky that time" and a resolve to be more careful in the future. And then, it's forgotten. SO, no data for statistics to play with.
The same amount of pull could fire an SA (safety off) and an AD/ND will NOT be forgotten. Not all of these get turned into data, either, but some do, and that is the data you have to work with. It's ..."flawed".
The German study provides an estimate of how often pressure is unintentially applied by a finger resting on the trigger, while performing relevant policing exercises with other limbs, in excess of 5 lb (enough to activate a SA trigger) and in excess of 12 (enough to activate a DA trigger). Not surprisingly, the 5-lb limit was exceeded more often than the 12-lb limit, and with these two data points we can extrapolate to other trigger weights of interest. From what I can tell, the study almost certainly overestimates the rate of the 12-lb trigger pull being exceeded, because only trigger finger pressure was measured, not length of pull. The stock long pull on most DA/SA guns would be expected to reduce the frequency of unintentional pulls based on pressure alone.
I earned a living conducting risk assessments on uncertain systems far more complex than a handgun. Unfortunately, I do not have ready access to my simulation sofware. However, allow me to propose an initial simple model for estimating the ND rate for a generic defensive gun use for a simple SAO handgun:
N = T x S x P
where:
- N is the ND rate (a value between 0 and 1, where a perfectly safe gun, something we all know doesn't exist, would be 0);
- T is the likelihood of one's trigger finger straying, unintentionally and unknowingly, onto the trigger (a value between 0 and 1, and which one of 2damnold's studies gives us an estimate that is closer to 0 than 1, but is clearly greater than 0);
- S is the likelihood of the thumb safety being deactivated (a value between 0 and 1 for which I have no knowledge of any empirical data, but we can agree on a reasonable way to represent this variable); and,
- P is the likelihood of enough pressure being applied to the gun's trigger to unintentially fire the gun (a value between 0 and 1, a couple of data points for which are provided in the German study).
All models are wrong, but some are useful. I propose that this very simple model can be used as a staring point to usefully determine the relative safety (in terms of ND potential) of various handguns.
The model can be exercised deterministically, that is representing each variable with a point estimate and getting a point estimate as a result. But, each variable is obviously highly uncertain, so a probabilistic assessment would be far better, representing each variable by a probability distribution reflecting our state of knowledge, and accounting for any correlations among the variables.
Of course, the model can be made more complex and realistic, if needed.
reading about adopting the shrouded revolver being a way to reduce city liabilities comes from an Ayoob tale. My best guess is NYC, but am not sure as to location.
I'll cover the city in a moment, but first, a short primer on hammer shrouds.
To reduce the risk of snagging the hammer spur in clothing during the draw you have two options, and both have been used. One way is to remove the hammer spur, the other way is to cover it. Cutting the spur off a hammer (and doing nothing else) CAN lead to unreliable ignition. Works, usually, but for reliability "retuning" with different spring tension might be needed.
Interesting. I've never thought about bobbing a hammer resulting in a decrease in ignition reliability, but it makes sense. I assume a heavier hammer spring may be required, at the cost of degrading the gun's precision.
Covering the hammer takes two forms, those that completely cover the hammer (and this includes "hammerless" revolvers) and those that leave a slot so you can still reach the hammer to cock it for SA shooting. "Shroud" generally means it is an after factory add on. Factory guns with "protected" hammer spurs are usually described as "integral hammer shroud".
S&W has made several models of guns with both covered and accessible hammers. One of the models with an "accessible" hammer is the Bodyguard Model 38. Look that one up, you'll see what I am talking about. A good picture is worth a ton of words.
I recall seeing a photo of a slotted shroud. How does one access the hammer to cock it?
Shrouded and "hammerless" revolvers predate liability concerns by generations. If Mas Ayoob was talking about liability, he was talking about what exists today (or then), not about what brought them into being.
Upon further reflection, I seem to recall Ayoob's anecdote being about a shopkeeper getting into trouble by cocking his DA revolver, the NDing it into a guy who was approaching him threateningly. My impression was the shopkeeper's biggest mistake was exclaiming that he didn't intend to shoot. Moving on, I think Ayoob generalized about similar things happening to police and some police departments, as a result, moving to shrouded or DAO revolvers. That said, I wouldn't bet the baby shoe money on my memory being fully accurate, although it comes from reading I've done in the past 3.5 years.
Out of curiosity, which PD is it? How was the modification done? Was the reason to reduce liability exposure?
(I'm going from memory here, so I might get minor details wrong, bear with me, you can look it up in detail at your convenience)
The city was Miami. I think it was the late 60s, or maybe early 70s, when DA revolvers were still standard issue.
The situation resulted from a lawsuit where an officer held a suspect at gunpoint in a video arcade. Gun fired, suspect killed. Lawsuit brought. One side claimed officer cocked his gun, and then it accidently fired. One witness said he saw the gun cocked (SA mode). Officer said he did not, he fired DA, when the suspect "reached for a weapon".
There was an on going issue at the time with officers and SA fire. It was "common knowledge" that street punks were bluffing the cops, not complying with cops orders, even at gunpoint. They would only believe that the cops would shoot them if the cop "proved" he was serious by cocking his revolver. At this point, you should be able to see the problem with that, cocked revolver (SA pull) cop with finger on the trigger ready to fire...
Part of the result of the suit was that the city had the SA notch on the hammer removed, rendering the guns DAO. So an officer could NOT cock the gun and the "risk" of AD/ND from the short, light SA pull was eliminated.
Trying to come to grips with the idea of a cop pointing an uncocked revolver at me not being serious. But, I'd be very compliant long before he reached for his piece.
To be clear, you are referring to a long and heavy DA trigger. There are short and light DA triggers, ala Glock.
Yes, I am referring to the DA trigger as found on typical DA revolvers and semi autos such as the Sig, Beretta, and others.
Glock is not a true DA trigger, and Glock does not refer to it as such. They call it a "safe action trigger" and are talking about more than just the little tab on the front.
The Glock is "partially cocked", and the trigger pull brings it to full cock, and fires it. A "true" DA action starts with the hammer/striker uncocked, brings it to full cock and fires it. A technical difference, more than a practical one, but since GLock says their gun is not a DA trigger, we don't refer to them as such.
I understand that Glock calls their guns something else, but classifying a trigger by what it does seems to be applicable to any gun. Thus, it seems reasonable to regard a Glock as a DAO, and I believe there are SAO striker-fired guns (Springfield XDs?). I consider the use of "Safe Action" to be smart advertizing propaganda; after all, the gun community has bought it. It took brass to call a trigger action that is inherently less safe as a Safe Action, but people have bought more unreasonable claims.
