First, any subset of the possibilities inherent in a character or setting can feel artificial if you notice things are bounded and limited. Thus where real life has effectively unlimited possibilities, life in a good story has limited possibilities. And noticing that the characters live conveniently falls within the bounds of good story can feel inherently artificial because real life is not so tidy.
There's a subset of writers who can't change the setting and characters to harmonize with a good story, however: the writers of substantially accurate historical fiction. They have to fill in the blanks in a historically likely manner, and they have to present all the events, known, probable, and invented, in a manner that will engage the reader. Hollywood's notorious for not giving a tinker's dam whether they get anything even slightly right as long as it's an exciting story and most people won't notice or care about the inaccuracies; but the aim of at least some historical fiction is to try to give the reader a sense of what things might really have been like, as far as the writer can manage to figure it out.
That is similar in certain ways to what I end up doing. The first rule I have is never, ever to contradict a model that has yielded a definite result. Then I usually fill in places where I don't know what "really" happened with the most likely possibility. Occasionally when one plausible choice is particularly good I favor it, and if another is particulary bad I avoid it.
Second, the way authors generally harmonize story with character and setting is that during the process of writing and revision, they change the characters and setting to harmonize them with the needs of a good story. This is not a luxury that a GM or players usually have in a role-playing game. I have played revised scenes before, usually to correct a misunderstanding but at least once to correct a game-destroying sequence of events from happening.
The most common reason for me to revise something is realizing that I was wrong -- I had to make a choice I didn't manage to channel and later I realize it didn't work the way I said it did. Mostly I can just note the change and amend the log, but after the first several sessions of my current campaign I realized there was something crucial about the setting I hadn't understood, and we replayed some scenes that had a significant effect on the characters' interactions.
No. I mean exactly what I asked, in the broad sense that I asked it. Can the natural actions of the character produce a story (in the most general sense) that's not good? If there can be good stories and bad stories, it follows that some sequences of events will make a good story and others will make a bad story. Is it possible that the natural actions of a character can produce a story that's just not good by the standards of, say, good fiction? I suspect, for example, that you don't even try to play certain sorts of characters because they will inevitably be boring. Is there a point at which a character seems like they should produce an interesting story but, when played out with fidelity to setting and character, are also boring?
...
Not all conflicts and not all resolutions are good stories, nor are they even satisfactory stories, by many standards. For example, the way a character resolves a conflict may simply be boring. Can't you imagine, for example, that some of the ways your character might cope with being undead might simply be boring as a story? Or maybe during the first few scenes, the flow of events and being true to character leads to a character that destroys himself or rolls up into a catatonic ball that he refuses to leave. OK. The character has resolved their situation but is it a good story?
There's no question that characters can behave in ways that's actively detrimental to doing anything interesting -- there really are campaign-wrecking responses and I've yanked a character out of play before he even got into it to prevent one once. It's just that being flexible about what story we're going to end up executing, and which characters will end up taking which roles, cuts down on the number of responses that I can't work with.
As far as boredom goes -- I've never been bored playing an immersible character, but there are things the character might do that are of little interest to anyone else. If I'm GMing I try to pay enough attention to when we've hit that point to start summarizing there, although of course one of the hazards of immersion is inattention to OOC perspectives. One of the sessions in my present campaign was wrong that way -- it was interesting to me and highly informative, in that I came to understand some of the character's previously inexplicable actions, but there was a point where it was time, not to change the character's actions, but to start summarizing rather than playing in detail. That I didn't was simple faulty technique on my part.
Now about the undead character -- I could use either permanent catatonia or self-destruction as the primary plot of a short story, or as a subplot in a novel or a campaign.
If I get stuck on the idea that the story I'm going to tell starts at reanimation and moves chronologically forward from there, then I may think the character wrecked my story if he does either of those things. The story I had in mind is wrecked, sure. All possible stories, no. The two possibilities that stop a forward-moving story dead leave a backward-moving one wide open: "How did he cope? -- He failed. And why did he fail? -- This is who he once was, and this is what happened to him." Self-destruction is stronger, being more decisive and active, but with sufficiently descriptive writing, either of these could be wrenching.
In fact, this character is providing a background subplot in my current campaign. I was expecting a more-or-less catatonic reaction, with a slow awakening that proceeded over months as his magical state stabilized; obviously, that's the sort of thing I'd portray intermittently and mostly in summary until he did wake up enough to do something interesting that the PCs could find out about.
There's a natural plot this character could have lent himself to, that could have worked if the character had awakened the way I expected him to: a mystery plot about who and what has come to be attuned to the magical wellspring in the enchanted forest he crashed in. I wasn't planning to run this anyway because the player knows too much about the character from other contexts, and that doesn't make for a good mystery; but if it weren't for that, I might well have run that -- except that the PC forestalled it.
The PC saw the undead character's glider going down and surprised me by deciding to rescue the pilot, despite the fact that the undertaking involved definite risk and there were very bad odds that he survived the crash. But she was willing to undertake it, and kept on being willing even after she discovered what she was rescuing. So he ended up reaching intermittent semiconsciousness in a much more psychologically favorable circumstance than I expected -- friendly human contact -- which mitigated some of the reactions I would have guessed at otherwise. (I can't actually be sure how he'd have reacted without playing it at least in my own mind.)
For 100% in-world reasons, I realized that given who he was, what he was doing, and the resources he had available, there had to be an enchanted equivalent of a flight recorder in the wreckage, and the PC recovered it. She already had a fair idea of who he was and why he was up there; she got to watch the dramatically appalling results of failing to survive the use of a particularly dangerous form of magic the player is interested in the PC's possibly needing to employ in the future. So the undead character and the events around him are suitable for setting up dramatic action in the PC's main plotline later.
If I had been intending to run the mystery plot and got too set on my original plans, maybe I'd be going around saying that the immersively-played PC wrecked my plot. Or if I were willing to interfere in the natural course of events more than I am, maybe I'd have tried to come up with some artificial obstacle to keep her from finding the revenant. But with the approach I take in fact, I don't see that I have a problem. Not being able to run one scenario forces me to set up another for the next leg of the campaign, but the worst ill effect I see coming from this is that I might have to take a break to figure out how to set up something else.
This character is turning out to be a recurring NPC, but as matters stand there isn't a plot I want to run involving him that primarily addresses the desires of the players and the PCs. (There's a possible plot with those characteristics that I don't want to run because it's much likely to turn out badly for him than not, and I like the character.) So he's in the background, serving to round out the setting and demonstrate that there's something going on in the world beside whatever the PCs' major attention is on at the moment.
There's another thing I'd like to get at, about RPG plots, constraints, and medium differences.
In writing groups that are likely to include roleplayers, invariably there are people who try to translate their campaigns into novels. Invariably also there are people who proclaim that no one should ever attempt this because it can't be done right, in spite of commercially successful counterexamples by authors who have done just that. The people who proclaim that it can't be done are probably throwing out the data from the successful translations because they don't recognize them for what they are: the marks of a naive attempt simply to write a campaign structure unchanged in a novel are painfully obvious; a skilled translation into a new medium with a very different expected audience doesn't end up leaving "This is my D&D campaign" bootmarks all over the text.
An experience I found most informative was reading an attempt at translation by someone who wrote serviceable prose, but who had not in any way recognized the structural differences between a campaign and a novel, which threw those differences into high relief for me. So I found myself reading a work that had six characters given nearly equal weight in a single-stranded plotline involving periodic fights. Aside from some pretty game-specific D&D tropes, it was like having neon signs reading "Keep the party together!" and "Balance spotlight time!" -- and I suspect it worked well enough as a campaign. It didn't work as a novel, because the extra strictures a lot of people put on RPG sessions have nothing to do with telling a good story as such, and their naive importation into novel format gave it a fair amount of dramatic deadwood in the form of distractingly redundant characters.
A novel can have several main characters, but they don't share the same dramatic function. You don't see novels with six characters differentiated primarily by their tactical abilities all having absolutely equal weight and attacking precisely the same problem from the same dramatic angle. Rather, important novel characters fulfill different dramatic functions and at least some of them will be participating in different subplots. You just don't see this idea that everyone must move in lockstep doing the same thing. It's possible to get away with that in an RPG because in the RPG format the characters are also serving as viewpoint characters, playing pieces, or both: every player who sees the story through the eyes of a character sees through a different pair. Which, of course, is not a method of differentiation that works in writing.
I'm not surprised that by the time someone adds the idea that to run a good session the characters must stay together and address the same problem to the idea that a good story is the one planned in advance, they've got a recipe for running a campaign that feels as free and natural as a straitjacket.
I never accepted some common ideas about roleplaying, in part because the first thing I did that much resembled roleplaying wasn't D&D. I joined a collaborative storytelling forum, where the owner made up a fantasy tavern and the rest of us introduced characters. It looked rather like freeform message-based roleplaying in some respects, but it had no particular connection either to freeform RP subculture or to tabletop subculture, with the result that we simply didn't import either set of ideas about The Right Way to Play. We ended up with a set of procedures that worked out to:
- The setting is fantasy.
- You control your own character.
- Get together with the writers of the characters you're involved with to work things out. (We'd often get together in chat if a scene had a lot of dialogue, to speed up the exchanges.)
It didn't produce great literature, but we had fun with it. And as a result I know for certain, by experience and counterexample, that some of the ideas floating around as to the only right way to handle the group construction of fiction are unnecessary restrictions.
- The story will not necessarily turn into a preposterous display of self-centered egotism if the guy who owns the setting also plays an important character as his own.
- You do not necessarily need game rules to keep from degenerating into "Bang! You're dead!" "No, I'm not!"
- It will not necessarily wreck the story if some of the characters find separate things to do.
- It is a sign of vanity and immaturity to write any story but one about a character who starts weak and slowly struggles to accumulate power.
The thing that caused me to move to the RPG format is that I wanted to do some tactical simulation (but I never succeeded in doing that satisfactorily online in chat until recently); and I wanted more consistency in the setting, and better integration between characters and setting, than you see in the freeform format, where no one is playing Editor in Chief of the World. But when I started playing RPGs, I didn't take up all the common ideas above, and my baseline idea of what you can do to tell a collaborative story is less restrictive because of it.
There's only one more of these things coming.