In the excellent DM's Advice thread, this particular bit of advice caught my eye:
Now, my initial reaction here is a big, heck no. Story, in my mind, is crucial in an RPG. But, stepping back a second, I wonder if there isn't a problem with how we each define what a story is.
Almost certainly. 'Tis the way of internet arguments.
To me, a story is simply character, setting and plot. That's why you can have 5 word stories, Drabbles (100 word stories) and various other very short fiction. To me, a story isn't a closed system. But, I think a lot of people define story in terms of a complete narrative. That you have a definite beginning, middle and ending and until you get to that last bit of punctuation, you don't actually have a story, but, once that final bit is in, it's locked in and THAT is the story.
I'm with you. I'd also say serialized fiction is a pretty big blow to the "a story has a single, specific, fixed beginning, middle, and end" definition. One of my favorite stories, for example, officially did not have a defined end when it was first started, despite being far less open-ended than all but the hardest of railroads in D&D. We have reached that end
now, but eight years before, that end was nowhere in sight yet.
Which, in my mind anyway, possibly explains why you see such different DMing advice. In another thread, we were talking about WotC's adventure paths and is an Adventure Anthology an actual campaign or not. Is it a campaign if you have a loose collection of episodes that are not really connected to each other, like Candlekeep Mysteries or Radiant Citadel? Is Dragon Queen a campaign because it has a definite beginning, middle and end point? Where does that leave things like Curse of Strahd, which is a pretty wide open sandbox until you get to the end and face Strahd. After all, if you complete Curse of Strahd, the final act of that campaign will be bearding Strahd in Castle Ravenloft, regardless of what order you do the rest of the module in. Is that more or less a story because it's open ended until the conclusion? Or, should we only strive for "emergent" stories?
I don't have any conclusions here. I'm just tossing this out to see what folks think.
We should have a spectrum of options because that's the best way to help the most people enjoy what they play. (There are deeper conclusions to draw from this statement, but they are too spicy to bring up here.)
Some "stories" should be Dragonlance-y, where there's a clear story and maybe even specific characters the players can play through it. That's a bit of an extreme, but it's clearly had a huge impact on how we play D&D (having almost totally displaced the old dungeon-heisting style, outside of the small but active OSR scene.) Toned down slightly, this gives us things like ENWorld's acclaimed APs like
War of the Burning Sky and
Zeitgeist, where there are major plot events which pretty much always occur, but adapt around the party rather than requiring specific characters or options.
Some "stories" should be akin to 2e Dark Sun, where there is a metaplot, whether or not the players are involved with it. The world grows, changes, and the players must respond--even if their response is to escape or defer. This approach depends critically on the execution, but I would say Paizo's well-received APs like
Wrath of the Righteous and
Kingmaker fall into it.
Some "stories" should be like your example of
Curse of Strahd, where there's a beginning and an ending, but the middle is free rein. This is a pretty popular style, I'd argue it's what people actually took from Dragonlance, the idea that there was a frame in which play would occur but some kind of "win condition" or at least a release point where the campaign could be considered definitively "done."
Some should be anthologies, as you mentioned earlier. Many small stories that can be combined in different ways. That's often better for groups that don't know if they'll be able to meet regularly, so bite-sized content is more compatible with their interests. This is also useful for DMs who want to do some stuff themselves, but want to have the fallback option of prewritten content if they're pressed for time or if plans change.
Some should be hexcrawls, or "pointcrawls," or other seed-like things. Something that defines a starting point only, leaving the party free to pursue whatever they like. This is also where I would put most "sandbox"-y or at least sandbox-like stories, as well as fully open-ended campaign settings, where there really isn't any "story" yet at all, other than the fact that the world is physically there and contains things with names and has a history. There's also room here for Dungeon World-style "Adventure Starters," where all you get is the initial presentation of The Adventure, and the story is played out moment by moment, "what will you do?" style rather than prewritten or assembled after the fact.
And then, of course, there will always be the pure homebrew, do-what-you want kind of thing. Where it's up to the group (whether "DM always decides," "collaborative effort," or any other preferred method) to determine what, if any, "story" applies to their game.