In my style the world exists and the players play characters in that world.
In your style the world comes into existence or takes form as the players play their characters.
I personally wouldn't use your descriptions here, because they elide the difference between reality and fantasy. I would say that in your style, much of the fiction is authored in advance. (And if the GM does have to make stuff up on the fly, s/he does her best to make it
as if it had been authored in advance - so it should be "objective", neutral etc - rolls on tables are one popular way of doing this.)
Whereas in my style, more of the fiction is authored in the course of, and as part of, playing the game. (Of course, from the point of view of the PCs the world exists just the same as in your style.)
A lot of metagaming seems to go on in your style of play.
What have you got in mind?
The GM isn't "neutral" in coming up with content - that's metagaming. But the players are engaging the fiction from an in-character perspective - that's not metagaming. This came upon the other worldbuilding thread: another poster on that thread advocated for player "input" into the worldbuilding at the metagame level - talking to the GM about what they want. Whereas my approach emphasises action declaration and playing the game - "story
now" - rather than negotiating the fiction at the meta-level in advance.
I also wonder if your style even needs a DM. Why not just state what you see when it's your turn and the party can react to established fiction or create more of their own?
As to whether a GM is needed, I would have hoped that it's obvious what the GM's job is: frame scenes that put pressure on the players via their PCs, and hence force them to make choices.
I woud hope that the difference between cooperative storytelling and action declaration and resolution is obvious. If not, the Eero Tuovinen blog that has been talked about a bit has a good discussion of it.
It prioritizes a consistent world. I mean if as DM I have to make up what's behind a secret door on the fly all the time that world is going to lack verisimilitude.
I've got no evidence to provide here but testimony. For what it's worth, I think you're overly worried about this. There are a range of techniques that get used to manage content-introduction, the most important two of which are "say 'yes'" and "go where the action is".
Remember, if the game is going well, then the player already has some goal or agenda in mind in declaring that his/her PC searches for a secret door. (And if the game isn't going well and has collapsed into random action declaration, than verisimilitude issues resulting from secret doors are the least of your worries - because you've already lost the core verisimilitude of genuine characters with genuine motivations.)
So that goal/agenda is the anchor and guidepost (to mix metaphors) for whatever gets narratied next. If the player succeeds, then they get what the want - which they alreayd think is verisimilitudinous, because they asked for it! If they fail, then you turn their agenda back on them - and, again, by drawing on the material the player has already made part of the situation, you are assuring it is material the player will engage with rather than reject on verisimilitude grounds. (More on this below - because it's really about "fail forward".)
Modern - Make death harder, Fail forward, Metagame mechanics, Adventure paths.
Maybe I'm being a bit hard on modern gaming. Perhaps you can elaborate.
"Fail forward" is used in two different ways.
One is its original meaning, as used by designers like Ron Edwardsm Vincent Baker and Luke Crane both in discussion and in their games. In this usage, the
forward refers to pacing and narratie trajectory. It's an anti-railroading device. I'll explain how: in a traditional Call of Cthulhu or Dragonlance module, if the PCs (and thereby the players) don't find the clue, or the secret door, or whatever is the "ticket" to the next situation, then the game grinds to a halt.This gives the GM an incentive to railroad the players into the situation where they'll find the clue ("Are you
sure you don't want to check what's in the desk drawers?"), and/or an incentive to fudge Search checks. "Fail forward", as an indie-designer response to this, is: don't frame the PCs (and thereby the players) into situations where there has to be a definite outcome for the game to progress. Instead, just "say 'yes'" to the players' action declarations (they get where they want to go, they find the secret doors they want to find, etc)
until you come to a crunch point that actually speaks to the dramatic context of play. (Given the way these designers frame and present their games, that shouldn't take too long!)
At the crunch point, you call for a check. If it succeeds, the player gets what s/he wants (so it's like saying "yes"). If the check fails, the GM narrates some adverse consequence which means the player and PC
don't get what they wanted; but the adverse consequence drives the narrative on, by engaging with the dramatic context of play (whatever that happens to be in a particular game) and hence provokes more action declarations. Rinse and repeat.
The "fail forward" technique I just described is pretty fundamental to no myth play, or any other player-driven RPGing where the focus is on story and dramatic arcs.
However, the terms "fail forward" has now been co-opted by the very railroading designers it was meant to be an antidote to! So now "fail forward" gets used to mean something like the following: instead of fudging the Search check, even if they fail narrate a success but through in a tweak or a complication! And instead of railroading the players to have their PCs search the desk drawers, if they don't then have the clue arrive by carrier pigeon instead (or whatever other means takes the designer's fancy). I don't read that many contemporary modules, but the ones I've read seem to have quite a bit of this sort of thing. Perhaps the most common one is thta if the PCs kill the "big bad" early, then everything goes on just the same under the direction of a lieutenant.
Some RPGers, including some posters in this thread, say that this is not railroading, because the players are free to delcare whatever actions they like for their PCs, and the GM doesn't actually fudge any rolls. My own view is that it absolutely is railroading, because it means that nothing the players choose to do actually matters to hwo the fiction unfolds.
I find it amusing, but also a bit frustrating, that a temr that was coined to describe an anti-railroading technique is now most often identified with ultra-railroady AP RPGing.
(If it's not obvious already: AP-type RPGing is at the bottom of my list of preferences. Next is dungeon-crawling - and there I prefer "fun" stuff like White Plume Mountain or Castle Amber over something like Tom b Of Horros. Next is "sandboxing" with a sympathetic and usefully responsive GM - with a doctrinaire GM this is as railroad-y as an AP. At the top, obviously, is some form of "story now"."no myth".)