D&D has always solved this problem, at least as far as physical combat is concerned, via hit points. I like a system that is able to extend a solution into other parts of the game. And for my personal purposes "the GM decides" is not a solution.
You speak of Fiat like its an approach taken in all situations; I see it as more of being when needed to improve or augment the game. As I said before, let's say the big bad monster is on its last legs with 76 HP and the rogue scores a critical hit for 67 HP of damage; I'm going to "fudge" that and offer a kill - and not tell the players that the monster "really" had 76 HP, because in my view it didn't "really" have 76 HP - it had 67.
I don't understand why you say that the GM has absolute power in a practical sense. The rules constrain the GM's power the same as they do anyone else's - for instance, the rules can state that if a player rolls a hit and does damage to the GM's monster, the GM has to knock off those hit points.
Yes, of course. But what I mean by "in a practical sense" is that the GM can always find his way around something if he really wants. Thus my Tarrasque example. A GM, in my view, can also create a reason for something happening or not happening and not have to explain to the players via the rules - citing page numbers and such. That's the GM's prerogative. I would add...as long as he has a good reason to do what he chooses to do, a GM can do
anything. Of course this opens the door for GM abuse, but if the GM is a reasonably mature human being that shouldn't be a problem.
The rules of an RPG might say that the GM can feel free whether or not to have regard to the players' dice rolls in adjudicating the consequences of PC actions but they don't have to, and I very strongly prefer a game that does not say that.
As far as I can tell,
all editions of D&D have said that, that "Rule Zero" is in effect (DM Fiat).
What you describe, to me, places the GM as more of another player, the player who plays "everyone else," and less as the story-teller.
Of course the GM can just ignore the rules and cheat, but then so can the players. That doesn't tell us about what limits on authority it is practical for the rules of an RPG to impose.
I think this is the heart of the matter, of our disagreement, and where (I think) you veer from traditional D&D - in every edition. The GM cannot "cheat." Or at least, they can emply Rule Zero to alter a situation as they deem necessary.
This doesn't mean that a GM should do whatever they damn well please whenever they damn well want to, especially when a player is being annoying. It is their job to remain impartial, to be a fair judge (referee), and to be willing to engage in discussion with players, even disagreements, about rulings. A good GM, in my view, is willing to change their mind - but
not because "the rules say so" but because the player presents a good argument as to why the ruling should be changed that out-weighs whatever reason the GM had for making the ruling in the first place.
So I don't think a GM can cheat as much as break the "contract of trust," so to speak, but being biased or unfair.
So I think the difference here as to do with both the nature of GM power, and whether rules are scene as absolute laws or guidelines - or to what degree. I've never played in a game of D&D in which the rules were absolute laws, or the DM didn't employ Rule Zero (Fiat) to some extent (the trick, though, is doing so without the player's realizing it).
On the bigger question, as I posted earlier, the only RPG I am interested in playing in accordance with the concpetion of GM that you describe is CoC.
Your description of the GM's role also tends, in my mind, to reinforce my sense that by "imaginative experience of the PC" you mean something like "forming a mental image of the events in the game fiction that are occurring to and surrounding my PC" as opposed to "emotionally inhabiting my PC and working his/her will on the world via the resources I have at my disposal." The latter approach to RPGing - which again is my strong preference, CoC excepted - more-or-less requires that the system give the players some resources whereby they can produce outcomes on the fiction that are binding on all other participants, including the GM.
First of all, I don't see how the two definitions you offer are contradictory. But the latter part is perhaps where we disagree, and where I say that even if these resources lead to "binding" outcomes, the GM can still find a way around that if he really wants to, thus has absolute power in a practical sense.
This is why I'm putting emphasis not on rules-as-absolute, but on what could be called the social covenant, the "contract of trust." The players are deciding to trust the judgment and fairness of the GM, and the GM is in turn pledging to be fair and just.
4e is a particularly rules heavy way to do this, of course. And there are things to be said - both from the "objective" viewpoint of design and the "subjective" viewpoint of personal taste - about the merits of rules heavy vs rules light systems. But I think there is no direct pathway from these things to a particular conception of either imagination in RPGs, or the role of the GM.
I'm open either way, but I still think there is
some relationship between "rules weight" and imagination, at least in some facets of the game.
Balesir's remark certainly captrues, for me, some of the reasons I don't like "unlimited authority" GMing. As a player, if I want to read a story or watch a film I'll do that. As a GM if I want to write a story or a screenplay I'll do that. When I'm RPGing I want to do what is distinctively pleasurable, to me, about RPGing - namely, frame scenes in which my players' PCs are engaged and then see what the players do. In order for me to genuinely "see what the players do" they have to be able to do things independently of my will. This is what player resources are for. Even p 42 involves player resources - skill bonuses, healing surges etc. Once I tell a player the DC and the outcome on success or failure, they get to roll the die and I'm as bound by that outcome as the players are.
None of which is antithetical to what I was saying above...the difference being that I reserve the right (as GM) to use Fiat if I deem it necessary to the improvement of the game, and the overall enjoyment of the players.
As for "4e is a pre-furnished apartment" I don't really have a strong grasp on the analogy: "furnishings" seem like world-building, but (i) 4e has a very light emphasis on world-building compared to traditional D&D, and (ii) worldbuilding may be creative for the GM but is nothing but a source of limitation for the players.
To the extent that I have some grasp on your analogy - that what is "pre-furnished" is the range of options that players can take for their PCs - then I don't think I agree. Here's one example that @
Hussar often gives: how often in classic D&D did a battle captain PC lead an attack while making a rousing cry to his companions, brining it about that not only does the battle captain get to attack the enemy, but so do all his/her companions? This is an utter staple from the more romantic end of the fantasy genre (Excalibur again) but is possible only in 4e.
I disagree. Its just that it requires the players to think imaginatively, the GM to think on the fly, to provide some kind of target number for the player to roll against.
You can still do this in 4e, but the problem is that page 42 is in the background, and behind the "density" of the AEDU system.
Multiple times you have suggested that I am misconstruing the significance of your reference to Excalibur, but I think you may not have fully appreciated the point I was making: that there is a fundamental contradiction between (i) the claim that 4e is in some way limited or "pre-furnished" compared to classic D&D, and (ii) the fact that for a whole range of utterly staple genre elements of fantasy literature, from the battle captain to the heroic comeback to the warrior who imposes his/her will on the mooks as they swarm him or her, only 4e can deliver them - and furthermore does so primarily at the behest of the players without needing to detour via the will of the GM.
(i) Is not entirely true in that I'm not saying it is as much "limited" as it employs a system of options that effectively lead to a de-emphasis on improvisation.
(ii) As I said above, I disagree with this. PCs can do anything in every edition of D&D
. A good DM/GM will allow a PC to try anything they can imagine. Just like a good teacher will never say, "Bad question," a good GM will never say "You can't do that." They might say, "That will be very, very difficult, but go ahead and try..."
For me, that is a game that is serving, not subordinating, the end of imagination.
It may boil down to what degree you and I prefer to have resources defined. I prefer less definition, which I feel allows for (or encourages) more improvisation.