I've pulled out a few items to focus upon.
My response to this is slightly round about.
My own experience is that play becomes more engaged, and visceral, when the stakes reflect player buy in, rather than a GM-established McGuffin. A very large number of modules involve McGuffins ("fetch quests" are the paradigm; just today I saw a post which suggested that it is good GMing to require a PC to go on a quest to get ingredients for the magic item that s/he wants for his/her PC).
I don't think that the GM establishing what the game is about
has to involve McGuffins. But I think it can.
If the GM buys into the players' stuff and embellishes it and works with it, I find the engagement and visceral nature of play increases. That's a mixture of aesthetic preference ( @
Lanefan upthread said he doesn't want pressure when playing; for me it's pretty vital, and McGuffins are the enemy of pressure in the relevant sense as the pressure is purely tactical/operational, not gut-wrenching) and generalization from experience. Both are prone to idiosyncrasy!
I won't say that I don't use McGuffin type elements from time to time in my game. They tend not to be as basic as finding ingredients or the like. I usually tie such quests more tightly to in game elements or events. Given that we're playing D&D in a fantasy world, I don't want to discard such item based quests, which are a big part of the genre expectations.
However, I do think such quests function differently in a game as opposed to fiction. So I try to take that into consideration.
I agree about player engagement, though. It's why much of my "GM backstory" actually draws upon a lot of elements created or introduced by my players.
Why am I using a rule that goes contrary to my preferences and that has irritated one of my players? Two reasons: I want to play the game more-or-less as written, to get the "Traveller experience"; and the rule is there to make getting psionics fairly hard, and I'm happy for that part of the game experience to be delayed a bit because it will change the nature of the game once this player's PC does develop psionics.
Why not just ban psionics, then? See the first of my two reasons. But then why, given that reason, am I not using sector-mappig? Because (i) I contain multitudes etc, and (ii) that would have such a ubiquitous blocking effect that it would make the game effectively unplayable for me, and so on that point Traveller has had to yield.
To compare this with the finding of the map example (removing the arbitrary choice of it being in the breadbin, see my comments below), what if the map is meant to be hard to find? Obviously, the breadbin makes no sense. But let's change it a bit....what if it was in the orc chief's treasure chest? That makes sense in the fiction, and makes the finding of the map more of a challenge for the PCs.
Does this still violate your player attempting to determine the map in the first room of the complex that they enter?
Or is the map just the maguffin itself? Is it just the impetus introduced to get the players to explore the complex? The more they explore, the greater the chance for some compelling aspect of gameplay to emerge.
Doesn't finding the map immediately undo that?
This all assumes importance being placed on the possession of the map (it's needed for some greater purpose, or it can lead to further adventure, or finding it is the current goal, etc.).
I posted a lot about this upthread. The difference I see is that in your orc example the player knows the fictional positioning - the GM has framed something, and the player has to deal with it. (If the player declared an action to sneak within dagger distance of the orc, and the GM fiated failure, that's a further matter, but I hope you're happy for me to assume that the player finds his/her PC at sub-optimal distance from the orc either as the result of a failed check, or in other circumstances where the GM was at liberty to frame the PC, and thereby the player, into adversity.)
In the case of the hidden document, the player doesn't know the fictional positioning - it's secret fictional positioning, secret backstory that leads to failure.
Upthread we also discussed invisible opponents, or NPCs in social encounters with hidden motivations or quirks. My view about these is that they're fair game if (i) the hidden stuff is knowable by the players within the current framing, and (ii) the hidden stuff in some sense is salient (because if not salient then, in practice, not knowable even if knowable in principle), and (iii) the failure to find the hidden stuff won't be a "rocks fall" moment.
Obviously factors (ii) and (iii) in particular are highly contextual - I would take more liberties playing with friends than with strangers.
My view is that the hidden document - which in this thread has served as placeholder for the generic "clue", or the generic thing that is central to the unfolding of play - violates (iii), and may well violate (ii) if the GM has decided that it's hidden in some largely arbitrary or unlikely place (my example upthread was the breadbin in the kitchen).
Okay, I will admit to thinking the mention of the map being in the bread bin was arbitrary in the context of an example, not that you meant if the location of the map is arbitrary, then why not simply allow it to be where the player hoped.
In the case of an arbitrary choice such as that, I would not in any way feel beholden to having the map be in the kitchen rather than the study. It's unimportant, and it may as well be in one place as another. I suppose that I'd question if the determination of such unimportant game elements really qualifies as player agency, though.
If the map or its location were not arbitrary, but were instead planned as part of the framing of the challenge, would that be different in your eyes?
This obviously isn't exact science, but what is motivating my comments here is that the practical result of the map being hidden in the breadbin is that quite a bit of the actual episode of play, at the table, will be the players declaring moves for their PCs that trigger the GM to narrate stuff about the rooms of the house being searched by the PCs until eventually they think to search the breadbins and the GM tells them they find the map. Because of issue (iii) the play couldn't continue without that moment taking place; because of issue (ii) it is an extended period of play; and thus a lot of time is spent on something where the players exercise little agency and the game doesn't really move forward.
Contrast: there are two scroll cases in the study, one with the rune of Ioun and one with the rune of Vecna, and one of the PCs is an invoker who is affiliated with both these (mutually opposed) deities, and finding the map in one or the other would count as a big reveal. We now have (i) and (ii) both satisfied, so no risk of a type (iii) misfire because the hidden thing is going to be revealed. Personally I would be quite comfortable with this sort of framing.
Between the two examples - of breadbins, and of two scroll cases on the desk - lie a range of other possibilities which differ as far as (i), (ii) and (iiii) are concerned. It's not an exact science. But I've tried to explain why I incline to one end of the spectrum, and the method I use to try and satisfy myself that that's where I am.
That's all fine. I can understand your preference even if I don't share it in the same way. I suppose that part of the disconnect was that I was viewing the finding of the map as a challenge, rather than some arbitrary element, and so, to me it seems odd to have a player be able to introduce a solution to the problem through the mechanism of a simple Search/Perception check.