Often when writing, particularly when writing a "creative" answerline (description acceptable, two answers required, answers that need a lot of prompts, an answer that is off the beaten path), editors will undersell the difficulty of playing the question and managing to construct the correct answer. This may not be an issue in playtesting, but think about it statistically - multiply that hazy feeling over the >100 teams playing the set, then by the fact that teams will be much more tired and strained after hearing all the previous questions, then by the fact that many teams are inexperienced and don't know what to do with more complex answerlines. This is the same reason that instruments like safety checklists or guardrails are used - it's certainly true in many cases that, even without these features, nothing bad would happen, but they're engineered to prevent the rare errors that do occur in their absence.
That last consideration means that for tournaments like ACF Fall and HS regs, it's doubly important to make your answerlines as fault-tolerant as possible. For harder sets (NSC, anything harder than two dots), it is perfectly reasonable to assume that people know the intricacies of the game. In other words, if you play ACF nationals, you know what you're in for. I'm also not suggesting there is a bright line to when complex answerlines should be allowed or when they shouldn't, or even which kind of answerlines should be allowed where, only suggesting you be cognizant of the audience for your tournament when you use them.
One can also consider the moderator's experience in this situation. Answerlines with long prompt chains, or with many alternate answers, are more likely to be screwed up by even the best moderators. Multiply this by the fact that, especially at the high school level, the quality of moderators fluctuates wildly, and it can be a recipe for disaster.
Some examples of fault-susceptible answerlines and how to make them fault-tolerant:
- Answers that require multiple pieces of information. Both "two answers required" and answers that require both work(s) and their creator (Chopin Etudes, Keats Odes) fall in this category. One can simply ask about one of the two required answers, or for the creator using only clues about that set of works.
- Answers about a nebulous concept or for which there isn't a reified, specific term. An example would be something like the tossup on "lower-class" in linguistics from this past ACF Fall, which is an excellent idea but empirically didn't play well. It may have worked better as a question on "class" ('this variable') or, as someone suggested, "English".
- Description acceptable. There's often a way to transform these into something much more linear. The one that first comes to mind is a biology tossup on "nuclear import" (it might have just been on transport, but the clue I buzzed on was on nuclear import). One could have simply written a question on the nucleus. It appears less creative, but it tests the same knowledge.
- Excessive prompting. Similar to "description acceptable", there's probably a way to change the answerline so that fewer prompts are needed. If you're writing about a small subset of something, just write about the better-known superset and use clues about the subset.
- Asking for a detail of a thing, rather than the thing itself. There was a tossup at an old NSC on "ice" clued from One Hundred Years of Solitude content. Were it being written for an easier tournament, I would probably suggest just writing it on One Hundred Years of Solitude.