...but that assumes that the probability of a frame failing at any moment is proportional to the number of the original batch still remaining at that moment. Radioactivity behaves this way (which is why the term "half-life" has crept into the popular vernacular. I suspect Diabloscott was slyly referring to this with his/her reference to "disintegrate" and "decaying".) So do some drugs (roughly, for a while) as the liver and kidneys metabolize and excrete them. Unless you know that the system obeys "first-order" kinetics, it is not meaningful to talk about a half-life.

A term that requires no assumptions about the system, just simple counting, is the "median" survival: the time by which the first half of the frames in the original batch will have failed. If quality control was really sloppy, but some frames were built very well, you could have half of them failing in the first year, but the ones that didn't fail might last nearly forever, i.e., you couldn't predict that half of the survivors would fail in the second year just because half of the originals failed in the first year. Conversely, if the factory produced mediocre frames of nearly constant quality, none might fail for, say, 5 years, then they would start failing all around the same time.