Rule No. 2
Every police officer knows (or should know) the four rules of firearms safety by heart. My versions of them have been:

1. Treat all guns as if they were loaded until redundantly proven otherwise.
2. Never let the muzzle cover anything you are not justified in shooting at that moment (I now think this rule needs to be modified, as I explain below).
3. Keep your finger off the trigger until the moment you want the gun to discharge.
4. Be sure of where your bullet is likely to end up.

Being law enforcement officers doesn't give us license to flout these rules. If we could, they wouldn't be so heavily stressed at the academy and make so much objective sense.

I know of no firearms instructor or gun-competent person anywhere who does not preach these four rules strongly and repeatedly. If Rule No. 2 means what it says, and we allow our officers to point guns at people they aren't justified in shooting at that moment, then we are contradicting ourselves. So either we stop preaching Rule No. 2 or we adhere to it in practice. Pick one. You can't have it both ways.

This is a serious concern. If we train our officers to cover with the muzzle those people we aren’t shooting in simulations or targets on the range, or tacitly allow them to do so in practice, and one of them has an unintentional discharge and wrongly shoots someone who he or she was covering with the muzzle, then all a prosecuting attorney has to do is get a photograph of the safety rules posted at our range or a copy of a department manual that contains them, and we have a serious problem. We can't say something to officers and then say it's OK for them to ignore it.

In fact, I believe that Rule No. 2 can't be applied 100 percent of the time, as I explain below, so we probably do in fact need to modify it, for liability reasons if nothing else. However, I see it disregarded in too many circumstances in which it should apply.

There is a too-clever-by-a-half argument that proposes that by keeping our muzzles up and therefore muzzling everyone we come upon during a search or challenge we are actually not violating Rule No. 2. It goes like this: "I'm not violating Rule No. 2. If I find someone during a search then I do intend to shoot them until I have determined they are not a threat." Obviously this is a semantic twisting of Rule No. 2, and in any case it ignores the risk-management part of our jobs.

Intimidation
You might give up a tiny bit of "intimidation" with the muzzle slightly depressed as you challenge a suspect, but any cop who can't compensate for that with intimidation from his or her presence and verbal commands probably lacks a vital skill necessary for the profession. Also, consider that Col. Jeff Cooper said covering a man with muzzle depressed (actually, not just slightly lowered, but at a 45-degree downward angle) might be more intimidating and deterring than covering him with the muzzle "on target," as it communicates a trained professional, confident that he is in charge of the situation. I'm not sure I agree, but the colonel's opinion is always worth considering.

Do you give up time?
How much time do you give up by depressing the muzzle to the suspect's feet? That is really at the heart of any objection to the muzzle-depression suggestion. A drill to measure this was done with hundreds (if not thousands) of law enforcement officers at the Smith & Wesson Academy during the 1990s. It was run at five to seven yards, with the arms straight and the muzzle depressed to something like a traditional low ready. The average time difference to get the first shot off on a Smith & Wesson Bobber target compared with starting with the muzzle on the target (fingers off the trigger in both cases) was usually 0.14 seconds. When I tried the same experiment on the range, I got times of 0.33 to 0.35 seconds both ways at five yards - that is, there was no difference in time. Running the same experiment at 12 yards, I found a difference of between 0.05 to 0.15 seconds (and a better shooter would be faster). I'll use my results here because they are from an experiment designed expressly to measure the time difference between the specific postures I'm talking about in this article.