|Image from Wikipedia|
What makes a hazard in a tabletop RPG fair? I believe that clues are what make hazards fair. They can be subtle clues, or even indirect clues, but if there are no clues you are playing a game of chance rather than a game of skill. Especially if hazards can kill a character dead with only a saving throw as potential saving grace, clues are critical (the issue is slightly blunted, though not removed, of the lethality rules are more forgiving).
This is relatively straightforward with physical traps, such as darts, daggers, and collapsing ceilings. Just think about the mechanism required for such traps and add description. Scorch marks, dead NPCs, holes in the masonry, dust on the floor in a particular pattern, watermarks on the walls, etc.
But what about illusions? Theoretically, an illusionary floor could cover a pit full of spikes covered in save or die poison. Sure, you would still get a saving throw, but having to roll a saving throw should come after making a mistake. So what is the clue for an illusion?
There has been some discussion of illusions by Courtney over at Hack & Slash. For example, this example of detecting illusionary pit traps suggests using magic saving throws after physical interaction is attempted (such as throwing a coin into an illusionary pit). And in a post about another illusion-powered trap, Courtney writes:
There should be at least a single word in the description of the object to indicate it’s chimerical nature.
For example, describing a rolling boulder as preternaturally silent. This approach is not entirely satisfactory to me, especially the saving throw method, because there are no clues at all prior to interaction. Moving to the second example, this is better (a clue is presented at the outset), but it also seems somehow artificial, and perhaps difficult to apply to non-visual illusions. I don’t think the “one word” rule is bad, exactly, but I think clarifying exactly what kind of clues illusions might create would be useful, especially when moving beyond examples of visual illusions that don’t make noise.
Consider this quite from The Deed of Paksenarrion by Elizabeth Moon (book 2, Divided Allegiances, chapter 13):
Young sir, if you think it is easy to produce even illusory fire, I suggest you try. My old master, who is well-known in the arts, always said that a fine, convincing illusion was far more difficult—because reality carries its own conviction, and saves its own appearances. If you make a flame, it is a real flame, and you don’t have to worry, once you’ve got it. But an illusory flame can go wrong in many subtle ways—even such a thing as forgetting which way the wind is blowing, so that it flickers the wrong direction.
Perhaps this is a potential answer. No illusion is perfect enough to not have any logical inconsistencies. So here is a simple rule of thumb similar to Courtney’s “one word” dictum. Every illusion has at least one inconsistency or “deja vu” moment (like the cat from The Matrix). Maybe something plays in a loop or doesn’t quite react correctly to the environment. I believe this could easily handle illusions involving any of the senses, because the heuristic is something incongruent, rather than something chimerical.
Potential inconsistencies may also, of course, be caused by something else that is not yet understood. For example, a flame flickering the wrong direction might also be an indicator of a draft coming from a secret door. So inconsistencies should not necessarily be seen as a dead giveaway of an illusion, but they are something that needs to be understood before a party can proceed without potentially subjecting themselves to the mercy of the saving throw dice.