« Future Not Quite Perfect | Main | Thoughts on Mechanics »
April 17, 2006
Statistically Analyzing Story
This is a way overblown post title, but hopefully it's grabby. Anyway, this post arises from a conversation I had with Nick Wedig.
In high school English class, I learned that a good story relies as little as possible on coincidence -- the events of the plot ideally arise organically from the essential natures of the characters involved. As a result, the action plays out with internal logic, coherence, and a sense of inevitability to it.
RPGs seem like they would make a good venue to actually test the idea that a good story has an inevitable force to it, because we can replay scenarios to see what happens differently each time. For example, we could take Dogs in the Vineyard, and take a particular town writeup and a particular set of PCs, and then simply replay the scenario three times with the same players, to see if substantially the same story happens, or if the story varies wildly from run to run.
My suspicion is that the stories will turn out very differently. Two different productions of the same play can be very different, even though the scenes and the dialogue are the same. But: I don't know that this is the case!
EDIT: Hey, Nick has a weblog!
TrackBack
TrackBack URL for this entry:
http://www.typepad.com/services/trackback/6a00d8342026ca53ef00e550327aaf8834
Listed below are links to weblogs that reference Statistically Analyzing Story:
Comments
I think that Dogs is a very good choice of game for this, as some other games might play out more similarly each time. This is because of the amount of player choice Dogs gives: the players come in and judge a town and deal with the consequences. Judging or acting differently could have different results. Whereas the story level differences for a D&D dungeon crawl would be largely the same, because the environment is basically static. (The tactical level differences could be significant, because that is where D&D gives the most choice and the most options.)
"Stastically Analyzing Story" makes me think of roleplaying as algorthimically generating stories, which it is in a way.
Posted by: Nick Wedig at Apr 17, 2006 10:39:41 PM
I don't think an RPG would be a reliable experiment on the subject - simply because you've got a GM and a bunch of players who won't be interested in precisely the same game 3 times. They'll actively try and reach new places, simply because they're more interesting then replaying the same game they've already done.
A slightly better idea would be running the same game a few times with the same characters but DIFFERENT players, but the same char played by different people tends to be rather different.
In addition, don't get confused between "a lack of coincidence" and "predetermined". "Lack of Coincidence" means avoiding scenes like "Oh, by complete coincidence, we just HAPPENED to overhear the bad guy telling all his plans!", or "By complete coincidence, we happened to overhear precisely the correct snippet of conversation to make us THINK Snape's the bad guy!"
But avoiding such "well, things just HAPPENED to turn out really conviently for the writer" bits is not the same as being predetermined. Take a character who's, say, constantly debating between his loyalty to his state vs. his loyalty to his family, who comes up to a critical decision point during the story/game. He can decide either way. It's not inevitable - otherwise, there'd be no dilemma (or a very angsty one...). It's also not coincidence - each possible choice has sufficient justification.
--Standback
Posted by: Standback at Apr 18, 2006 4:00:22 AM
I actually did this. With Dogs, no less. 2 groups running through Clement's Patch, kind of in parallel. I learned 2 things:
1) The story diverges not merely through the preferences of the group - which I had been prepared for - but the way fictional content was introduced into the shared space. For example, first time through my introduction was quite dry: you ride through a town....you are greeted by the steward. Steward explains the situation, and Dogs jet off to resolve it. Steward is furniture. Second time round, I'm more into description, and draw out an idyllic scene, tilling, hoeing and a kid playing in the road. "We speak to the kid, and ask him about the goings on in town." This results in the players deciding that the town is not being well looked after, and the very first conflict involves shaming the steward in public, who becomes a major figure in the plot.
2) It's not clear to me how this can be done rigorously. As a GM, you have flexibility in how to play your NPCs - they have definite drives but they're not simple stimulus-response machines (and this should be true of characters in most RPGs). Last time round sister Althea got flustered under questioning and tried to shoo the dogs out. This time round...what? If I do what I do before to keep things consistent, then I may be steering things away from the path. If I choose to play her differently, then I'm steering away. The only way to be consistent would be to make it a blinded design - which would mean 2 different GMs, unless I can find a way to wipe my brain between groups. And 2 different GMs - well, that's a whole nother thing right there.
Also, it's deadly boring running a game as an experiment. Probably 2 different GMs is the best way to approach this.
Posted by: Alex Fradera at Apr 18, 2006 4:48:55 AM
Interesting line of thought...
However, I think the answer to the question can be seen simply by looking at the structure of RPGs in play. Let's even continue to examine Dogs in the Vineyard. Every time you roll the dice? That's because you see it as plausible for the story to go either way (or at least it should be).
Think about it, when a player declares that his dog hovers mystically above the ground a group (usually) either lets it happen or shoots it down entirely (depends on just how much the players are playing the magical aspects up). You don't (always) roll for this sort of thing because it isn't (always) a plausible addition to the story.
But each time you do roll, you're saying, "I think the story makes sense either way." Thing about your own play, how different would the story have been if you had succeeded (or failed) on that one critical roll. It still would have been a really cool story, but definitely not the same one ("For want of a nail...").
Thomas
Posted by: Thomas Robertson at Apr 18, 2006 7:48:19 AM
I would think this would be easy enough to check. Ask people who run demos at Cons how their experience has been - they usually run the same scenario with VASTLY different audiences enough times to potentially see trends. Data, waiting to be mined!
Posted by: Rob Donoghue at Apr 18, 2006 9:39:32 AM
No two games run by the same scenario will ever produce the same story. Thats the beauty of having players. It is rather like handing two different writers the same scenario, even with A, B and C the same you'll still get different stories. This is something that proves true in creative writing workshops across the country.
My personal data point has to do with running the same scnario twice during the same weekend at a con (something I've done multiple times). Even though the game was the same and it had me as a touchpoint both times were radically different stories given the people at the table.
Posted by: Jere at Apr 18, 2006 9:53:08 AM
Dogs is an interesting game to use because it doesn't really have any tactics and because of the sheer number of dice rolled there's less influence of chance.
I've run the same Dogs scenario four times now (Fort Yellow) and had fairly similar end points, in that the Dogs sort out the mess eventually but the trip had been rather different each time.
In these games I think I can put the differences down to the way the players approach the game, that is, whether they are more interested in exploring their character's role or the story as a whole or winning.
Posted by: Steve Dempsey at Apr 18, 2006 10:55:10 AM
I don't have much to say other than "try it! test it!" and agreement with Thomas that the dice are going to be an X factor over and above any human factors reshaping the story.
But I'm posting because I think some of the commenters above have missed a key part of Neel's post: he's talking about running the same scenario three times with the same players. Which is a neat idea.
I ran a session of Paranoia XP a while back with a kind of "Groundhog Day" premise. It wasn't deep, storywise, but it was a lot of fun seeing how information was gradually revealed and things went differently every time. Each iteration was in a way a comment or riff on the previous iteration.
Posted by: Rob at Apr 18, 2006 12:09:46 PM
Rob: Right on about using the same players. Also, can you say more about your Paranoia game?
Posted by: Neel Krishnaswami at Apr 18, 2006 3:00:49 PM
Did I never post about that game? I remember promising Bryant I would. OK, soon.
Posted by: Rob at Apr 19, 2006 1:38:39 PM
