Although I think it was my post in another thread that started this, reading this thread I'll acknowledge (top of the hat to
@Lanefan especially) that there's one possible exception to my insistence that all traps have tells: the kinds of traps that allow decision-making
after they have been sprung. The room slowly filling with sand. The portcullis that prevents backtracking. The tripwire that rings that bell that alerts the guards. Etc.
Not that this means they can't have tells or shouldn't have tells...but I don't think it's a crime to have (some) tell-less traps like that.
In one sense this is a variation to the distinction I've made elsewhere between traps that are easy to find but hard to negate, and those that are hard to find by easy to negate. This type of trap is "easy to find" in that you just stepped on the pressure plate (dumb@$$!) but "hard to negate" in that you now have to figure out how to solve the predicament you just created.
As for the verisimilitude argument:
- It is totally realistic for some traps to have tells.
- It may not be realistic for all traps in the universe to have tells, but presumably the players are only seeing a tiny fraction of all the traps in the world. See point #1.
- In the absence of modern dentistry, another thing that would be highly verisimilitudinous would be rampant tooth decay among PCs. (Which is more fun, tell-less traps or tooth decay? I dunno; it's a toss-up.)
In any given dungeon, I'd rather have just one well-thought out trap that is subtly telegraphed and requires ten or fifteen minutes of player planning to discover and bypass...the kind that causes the players to all cheer because they have genuinely accomplished something when it is finally "solved"...than any number of trivial roll-a-die and keep going traps.
Also, all of the above applies equally to secret doors.