Is there any data or experiences regarding failure variances for different types of lamps? I would guess some of the Chinese manufacturers have higher pre-mature failure rates. For example, if I took two incandescent lamps from the same box and put them in the same fixture, and one failed at 418 hours, and the other burned for 1732 hours, that would be a high failure variance. Possibly due to an imperfection in the filament of the first lamp. LEDs (not talking about retrofit lamps, but the LED diodes themselves) have a very consistent failure rate. They gradually get dimmer over time.

Possibly due to an imperfection in the filament of the first lamp.

It could actually be the second lamp, who has a failure in the filament: It may have been stretched and consequently run colder than designed, so yielding less than designed efficacy and light output.

Definitely, too large spread is the most common consequence of poor process control, mainly when the lifetime is limited by the operating principle (filament evaporation in incandescent, alumina penetration by liquid halide salts in CMH's, arctube blackening and consequent thermal instability in MHs generally,...), where the resulting life time is traded against other performance criteria (efficacy, color stability,...).
With quality products all these effects should be tight under control, so should yield the same results for all individual pieces (operated at the same conditions).

It could actually be the second lamp, who has a failure in the filament: It may have been stretched and consequently run colder than designed, so yielding less than designed efficacy and light output.

Definitely, too large spread is the most common consequence of poor process control, mainly when the lifetime is limited by the operating principle (filament evaporation in incandescent, alumina penetration by liquid halide salts in CMH's, arctube blackening and consequent thermal instability in MHs generally,...), where the resulting life time is traded against other performance criteria (efficacy, color stability,...).
With quality products all these effects should be tight under control, so should yield the same results for all individual pieces (operated at the same conditions).

Good point. Smaller spread is also good for group relamping. I'm sure we've all had to get the ladder out to replace a lamp, and then the one right next to it goes out like the next day. Because they were installed at the same time.
"alumina penetration by liquid halide salts in CMH's". I'm not sure how that works, but it sounds sophisticated, lol.

"alumina penetration by liquid halide salts in CMH's". I'm not sure how that works, but it sounds sophisticated, lol.

It is not by far:
The molten halide salts dissolve the alumina - the material of the arctube. It forms a gradually growing pit at the place where the salt pool reside during operation and once the pit goes through the wall thickness, the arctube leaks and so the lamp fails.
It is one of the major limiting factor of the CMH lifetime, but the seemingly easiest countermeasure (thicker walls) absorbs extra light, so reduces the efficacy. So if the arctube wall is thick, it takes longer for the salts to make it through, but it blocks more light.
Of course, there are design tricks to get better optimum, but these need either more complex arctube shape (the wall is made thicker only at the places, where the salt pool sits to make it last longer and thin elsewhere to not block the light), restrict the operating position range (to make only dedicated spot for the pool, so the majority of the burner is made thin), some salt composition tricks (suffice with less amount of salts, so smaller residual pool,...), in any case with any chosen design concept if all the pieces are equal and operating conditions the same, there is no reason why the lifetime should differ. So if it does, it means something is varying. And if that involves the manufacturing, it means the manufacturing is not well controlled. And that means a quality issue. And with such trade offs to be made, these issues means generally shorter lifetime and worse performance, compare to a better quality counterpart.