The reason I describe it as "GM-driven" and as driven by the GM's interests/concerns is because the world, the events in it, the reactions of the NPCs, the hooks, the possibilities, have all been authored by the GM (or taken by the GM from something someone else wrote).
Or, as I've stated before, taken from what the players have suggested, often off-hand, or it can be related to their backstories. A lot of what gets worked into the campaign is what the players have said in the past, they just don't usually realize it.
But the choices that are made ("prices are high", "everyone else holds firm on price", "rumours of war") are all made by the GM. And they do seem to steer the fiction in a certain direction - at least, they certainly steer it away easy access to silk.
And see, to me, they are made by the world. Just like AW states to "Turn it over to the NPCs" there is a brewing war in Calimshan, and as a logical result, the prices of Calishite goods is increasing. However, there are some merchants that are involved in smuggling, extortion, or other schemes, and locating one of them yields different results.
In addition, it could be, at least in part, the determination of a die roll.
In so far as the world looms large in play, and imposes constraints on and consequences for player action declarations for their PCs, it is the GM's vision of the world that seems to be paramount. The players, in the course of play, learn more about that. That is what I mean by "learning what is in the GM's notes". The player, by (say) having his/her PC looking around for the seller of silk at the the lowest price, is learning something about the GM's account/conception of the world. The shared fiction isn't being established in response to, and as part of the context of, the players declaring actions for their characters.
Some of it is in response to the actions of the players, some of it isn't. As I stated in other posts, like the one about illusionism, I have no problem changing something on the fly if that's more appropriate. While prepared material provides a foundation and framework for me to use, nothing is written in stone until it occurs in the campaign. Once the players have had that interaction, then there is something more that is known.
However, I approach it from a world-building and logical standpoint, not what might make an interesting story or scene in the moment. Because this is an ongoing campaign, with future ramifications for the player's other characters, I like to maintain an internal consistency.
Upthread of your post, I posted this about the role of the GM and player in BW (pp 268-69 (Revised); 551-52 (Gold)):
In Burning Wheel, it is the GM's job to interpret all of the varous intents of the players' actions and mesh them into a cohesive whole that fits within the context of the game. He's got to make sure that all the player wackiness abides by the rules. . . . Often this requires negotiating an action or intent until both player and GM are satisfied that it fits both the concept and the mood of the game. . . .
[T]he players . . . have duties . . . [to] offer hooks to their GM and the other players in the form of Beliefs, Instincts and Traits . . .
The player offers the hooks. The GM responds to them. If there is uncertainty we talk (as I posted upthread, I am not interested in GMing blind). That's part of the force of Luke Crane's comparison of GMing Moldvay Basic to a cross between telephone and pictionary. In Basic the GM
isn't talking to the players in that way, kibitzing with them, negotiating the framing with them. But in BW this is standard stuff. The GM isn't "guessing" or "interpreting" (I'm assuming we can put aside the philosophical questions of solipsism, other minds, etc in this context).
For me, at least, it's not about creating on the fly. It's about the context of and rationale for authoring the motivation. I prefer it to emerge from the play of the game - ie roughly, as an output; not an input.
And to me it's a bit different. Yes the players offer hooks, but I don't always incorporate those in the moment. I am, however, always reacting to the character's actions. Those actions have natural ramifications, based on the goals of the NPCs, or the many events outside of the character's control that are happening in the world. Why?
Because that's the way the world seems to work to me. If you're Elliot Ness going after Al Capone, you have your motivations, and Al has his. Yes, Al gains some new motivations because of your investigation, but in general his motivations are entirely independent from yours. And, at least early on, his motivations in relation to yours are probably more along the lines as orders to his goons to make you go away. They have motivations too, most of which are also unrelated to you. They are more closely tied to their selfish, greedy needs, their love of violence, their wish to move up in the organization, and/or not be "taken care of" by Capone or his men.
My "writing style" is to see where the story leads me, both within the campaign and offscreen. As soon as you create an NPC with a motivation and goals, it points to a certain direction. But then you place it in a world where there are other NPCs with motivations and goals, and things change. I don't spend a lot of time thinking about most of them, and in the majority of the cases, if it is a prewritten NPC, I probably don't consider what's happened until they do come into play. Now that I'm rereading the AW rules, it's not all that different from what's recommended there.
As a player, I think it's not that hard to tell when the GM is running the game based on his/her (pre-)conception of the fiction, rather than in response to the players' hooks.
My direct experience, since I use both, is that's it's not easy to tell the difference. At least to the players I've had. So I'm asking for more concrete evidence as to how you would be able to tell.
I've already posted
this several times. Here it is again:
One of the players is a gamemaster whose job it is to keep track of the backstory, frame scenes according to dramatic needs (that is, go where the action is)
By "going where the action is" I mean framing scenes according to dramatic need: ie picking up on the players' hooks.
But I do think it's good advice to GMs to (in general) avoid boring stuff. (Obviously what is boring is relative. As I posted upthread to [MENTION=29398]Lanefan[/MENTION], I find shopping for clothes boring in real life and boring in game also. I wouldn't suppose that's a universal view.)
All I'm saying is that you're the one who drew the inference from conflict to combat, and it's in your game, not mine that the players' response to an angry bear was to kill it.
Conflict doesn't have to equate to combat, and I also gave a number of examples of other options that could have occurred that were neither conflict nor combat.
I don't control the players, and they opted to fight, yes. But my point was, after following what is largely a default course of action for many D&D PCs, they learned something and it dramatically changed how they approached things going forward. It didn't have anything to do with the rest of the motivations or story arc at the time, it was just an encounter. But it did change the course of the story, because it changed the way the characters behaved in the future. It had a surprisingly large impact.