Suppose that you generalised this to the whole of play - having the players provide vectors, and responding to those via on-the-spot reasonable thinking.
My position, in this discussion about mysteries, is that this does not mean that the mystery lacks "reality"/"objectivity" - it doesn't mean that the players are just authoring their own solution.
To refer back to the Cthulhu Dark session I've mentioned upthread:
*A player decides to play a butler, Appleby;
*I ask the player why Appleby is in London, and his explanation is that the Earl for whom he works is missing;
*I frame Appleby into a scene involving another (NPC) butler, who is explaining a cleaning process for silverware - thus providing a clue (which the players missed) to the place of lycanthropes in the mystery;
*I introduce a document, discovered by Randal (the other player's PC), which he first reads as a topographic map (ie that is what I describe it as seeming to be), but which I know is actually a phrenological study of a hyena skull - only later is this document studied more closely (ie by way of action declaration) and the truth recognised;
*I introduce another document, also discovered by Randal, which is phrenological study of the Earl's skull in his doctor's files - it shows the same patterns as the hyena skull, a clue that the Earl is a were-hyena;
*Over the course of the session, in more low-key ways than what I have described above, I introduce clues linking the Earl and one of the other principal NPCs both to Central Europe and East Africa, which in turn is a clue to a link between (European) werewolves and (African) were-hyenas - at least one of the players picks up on these clues and recognises that link.
The decisions I make as GM are responses to the player's ideas - starting with Appleby' missing master; Randal's investigative journalism looking to criticise British imperialism, and then building on that and on the successes or failures of their action declarations.
In this thread,
@Crimson Longinus,
@Bedrockgames , and
@EzekielRaiden (and perhaps also
@Maxperson and
@FrogReaver?) have all asserted that the play I've just described does not involve a "real" or "objective" mystery, because it was not pre-written by the GM.
But none has explained why: there are clues presented; the players, both in the play of their PCs and in the more "meta" orientation towards the game, miss one (the silver) and draw inferences from others (eg the link between the East African were-hyenas and the Central European werewolves); and they finally work out what has happened to the Earl, although probably not as early as they might have done.
This is where the difference between
just making up whatever and
following principles in the creation of a shared fiction makes a difference.
If, at the start of the session I've just described, the two players had written down guesses as to what had happened to the Earl,
at that point in time there was no "true" solution to check against, because the game hadn't been played yet.
Upthread there's been some discussion of surveillance cameras. In the game where the GM didn't think of surveillance cameras when prepping, the same would be true: had the players written down guesses as to what the cameras would show,
at that point in time there was no "true" solution to check against, because the GM hadn't made it up yet.
The camera discussion has been full of accounts of how the GM can make a
principled decision about what the cameras reveal (the principles discussed have been mostly the sorts of principles that govern "living world" GMing).
The example of play I've described involves
everything being done via principle decision-making, although the principles are different from those that govern "living world" GMing. There was a mystery; bits of it were solved; the players did not make up their own answers, but arrived at solutions via inference (just as they would have if we were playing a traditional CoC module).
