Hi, InvisibleMimsy. The question you asked is an important one, and one that we as a forum have been mulling over since the turn of the millennium.
There are a number of things that contribute to a puzzle's longevity. A lot of them are systematic and nothing to do with the parsing of the puzzle - the fact that we're a small forum, that even the easier puzzles tend to have 1-3 members contributing after the initial round of testing assumptions and establishing details (and these solvers may be relatively inactive) etc.
Note also that having puzzles last longer than you expected isn't the sign of failed or poor setting by the host - a lot of people find the more difficult ones interesting, but may not be able to come up with the right questions. This is fine. It's probably because the puzzle is, you know, difficult.
However, there are a few things that contribute to longevity.
- As Earnest alluded to, puzzles that involve setting up context or establishing a series of events tend to take longer than those hinging on a certain lateral leap. Consider albatross soup or the doctor sending his arm in the post against the dwarf in the elevator or the cowboy facing down the bartender's gun.
- Puzzles based on testing assumptions also tend to be solved quicker than those involving creative thinking - even though the basic question is 'how', look at the Two Americans who somehow had the same son versus the party punch with the poisoned ice cubes. Look also at the fact that forum members say 'H/A/M?' almost as a reflex.
- The importance of context is also a factor - the man who lies dead next to a stick and a pile of wood shavings only makes sense in light of his own life story (which must be established through questioning), whereas the man who finds his wife dead and a stranger in the car could happen to more or less any adult male.
But the main one, in my opinion, is the amount of information you're willing to give away from the outset. Consider:
'The music stopped. He died. Explain'vs.
'A man is driving quickly through the suburbs. He turns on the car radio, but hears only silence. He gasps in horror, and then, like, kills himself. Why?'The first will take longer to solve, but will have a range of satisfying alternate answers (you were probably thinking of the tightrope walker until you read the second framing), whereas the second cuts out a lot of lines of questioning, but will lead the solver to an answer that is likely the best fit for the puzzle question.
And this is a relative mild example. A lot of the puzzles seen on this forum over the last two decades have been exceptionally vague - both in terms of the scenario and what they want the solver to do with it. We'd see puzzles that simply said 'It was square'. That's it. Not only did it present a scenario that I had no trouble believing, I had no idea what to do with this information. I would basically have to ask the host what the puzzle was, which for me defeated the purpose of the exercise, and gave me no incentive whatsoever to start off. And the answer was invariably several paragraphs of text, illustrating a lavish story, most of the details of which were irrelevant, but there were enough important steps along the way to reveal what 'it' was, and lend enough credence to the suggestion that its being square was in some way unusual or noteworthy. This is all very similar to how I assess the underlying legal interest in a long-winded Caveat Against Dealings lodged against a certificate of title by a practitioner. I don't need to deal with this from you guys.
One final contributing factor is whether or not the key to the solution is establishing its logic, or working out specific details. The 20-questions style puzzles in which the spoiler will only be posted once we've established that the murder weapon was a dark green rotary telephone will necessarily take longer than the one in which we established someone was strangled to death with a flexible cord.
Among the puzzles you've posted, my favourites include:
'Face the music' - A nice amount of detail presenting an odd scenario and demanding an explanation, leading to a good set of assumptions-testing around your role in the concert and the physics of playing the note, followed by establishing the nature of the item that played it
'Sign of the times 2' - A very clear scenario demanding explanation was presented. The nature of what needed to be worked out was clear from the outset, and the answer was simple and satisfying.
I actually quite liked 'Songbirds of pain' - the scenario presented in the answer was great, and coming up with questions worthy of such an answer is extremely difficult. The two sentences you posted demanded a causal link, and the one you provided in the spoiler was a satisfying one. The puzzle lasted longer due to the solution ticking off pretty much all the above boxes - The answer required context (she was addicted to pain and intrusive surgery) a series of events (the various operations, the husband not being able to go with her and therefore needing to meet her at the airport, the big 'reveal' etc), applied creative thinking (there are many ways and reasons to break legs or leave spouses - which ones apply here) and vagueness (who are they? why are they breaking her legs? What's the timeframe here? Why does she leave her husband? What does he have to do with it, anyway?). Therefore, it was long, but satisfying.
Gimmicks like posting the puzzle in an accent, in poetic metre, or using LOLspeak can be a way to throw in a red herring, I guess. I personally find them off-putting, but others disagree.
Also, it should be noted that my approach to puzzles is famously conservative and this forum basically thrives off the efforts of members who aren't quite so precious
