Agree.
My goal as GM is to simulate a world in a neutral and fair way. The players then interact with that world in the same way we interact with our real world. To the degree, I can provide that sort of experience in an imaginary fantasy world, I consider myself a success.
The big problem for me is what I've bolded in the snippet above---namely, it's almost impossible for me as a GM, even with the absolute best intentions, to remain fully neutral/impartial/fair within all of the parameters available. Whether it be scene framing, adjudicating action, prefabrication of world elements, challenge and combat encounter creation, etc., I always find that inevitably some sort of bias creeps into my decisions.
Most of the time, that bias is in favor of the players, but sometimes it's not. Sometimes it's stuff that I just think, "Man, I really
really really want the players to see or experience X, because that would be sooooo cool!"
And suddenly that desire to have the player experience X becomes this hidden seed that pushes the action.
For example if you say "What do I see", the GM isn't responding based on their prior conception of the fiction. That isn't how we conceive of play at all. It is not this unfolding fiction that is happening that gets built up in binary exchanges of players say X, GM decides. There is that component of the GM making his decision. But you are ignoring things like players can make a case outside character for things, and the GM will often be considering their words. It isn't as simple as "I decide". My answer needs to make sense too. And most GMs I have played with, will allow back and forth, where players often explain hwy they think something ought to be present. The players don't have direct power, but they have the tools of persuasion (expected to be used in good faith, not to advance their character's interest) to help smooth out this process. In a typical sandbox the GM is making his decision not based on the prior fiction, but based on the world, the ongoing situations in that world, and what has just previously occurred (I think this is a much better term than the fiction, because the fiction seems to sidestep or minimize the role of the world).
The bolded part here is the problematic component. You keep talking about "the world" as if it operates in some wholly independent sphere of the shared fiction, as if it is possessed of some immutable, objective properties of existence separate from other components of the imagined fiction.
This is the whole crux of the argument around how and what player-facing game mechanics are designed to address---that there is no "world," there are only conceptions of the fiction in question. Saying that it's all part of the grand, overarching "natural, simulated world" doesn't give those conceptions any additional weight or gravitas.
Honestly, this was one of, if not
the biggest mental hurdle for me to get over in regards to knowing how to approach player facing mechanics. Because the "game world" simply
had to be this independent construction, operating under its own parameters. How else could anyone know anything about anything if there wasn't an assumed, "fully realized" game world?
Until it finally clicked that there is no "world," there are only
conceptions of the fiction. Any given conception exists in one of two states---1) something that is already established as true within the fiction state, and 2) things that are proposed to be true, but not yet known to be true (and potentially may end up being false).
By default, D&D assumes that a GM's notes / prefabrications / headcanon are conceptions that fall into Category 1 --- "Something already established as true within the fiction," until/unless the GM deems otherwise. The fact that the players don't know about the overwhelming majority of prefabricated "truths" is irrelevant, they're still considered "truth" for the fiction.
Category 2 conceptions are generally propositions from the players---"I kill the orc." This isn't known to be true until the game plays out, and the fiction state resolves. It may end up being true---and may end up being false, if the player's dice perform badly, or some other interposition happens first (e.g., the orc successfully runs away or the character trips and falls down).
Category 2 conceptions/propositions can be negated. For example, a player can say something as simple as, "Bob the Fighter walks across the room to head toward NotBob the Vile's private dining area." But this can be rendered untrue in any number of ways, e.g.:
Player 2: Joe the Wizard grabs Bob the Fighter's arm as soon as he stands up. [in character] 'I don't think you want to mess with NotBob right now, friend. He'll probably kill you.'
Or,
GM: You go to walk across the room, but the barmaid slips, crashes a tray of empty flagons to the floor, and falls into your arms in disheveled confusion.
In both cases, Bob the Fighter has not, in fact, walked across the room to NotBob's private dining area. At least not until the interposed propositions are either accepted or rejected as truth.
RPG gameplay is really nothing more than Category 2 conceptions/propositions steadily moving to Category 1---it was unknown if the conception is true, and now it is known to be true or not.
To take another example, if the players go to the head of phoenix moon gang and ask for her help finding the disappeared daughter of a local magistrate, the GM is going to respond, not decide, but respond, based on what the players say, what the leader's motivations are, weighing any rolls they might make, who the player characters are, etc. What the players say here could be very important. Then he might declare what the leader says or does, and even then he isn't often simply deciding.
The bolded portion of your quote cannot logically follow from the sentences that precede it. A
response is necessarily a
decision.
