The Campaign Builder's Guild

[blockquote=beejazz]This isn't literature... this is something you're going to be doing once a week for (hopefully) a very long time. I think it's best to mix it up whenever possible.[/blockquote]

Interesting.  I take the exact other viewpoint normally, in that I try to use the 'literature' approach for campaign design and plot design.  I consider my setting to be a book we are writing, and I try to even look at it in chapter format and even sequel format.  I'm not saying that a good book shouldn't have twists and turns and surprises, merely that I do look at it as literature, and when the game resembles a good book, I consider it succesful.

@Vereg: I'd compare my stuff more to a television show, if that's any help. It's maybe a little tighter than a television show would play out (alot of the animes I watch would devolve into PvP rather quickly... rival protagonists being fairly staple) but certainly not as tight as a book would be (unless you're counting the truly wild things, like American Desert).

@snakefing: I think my idea is that one shouldn't *start* with unity. It's important to distinguish "from many, one" from, say "from one... more of the same." A central core is important (the world is going to end in 86 days when the asteroid strikes) but one should not forget the periphery (panic ensues... you might be able to stave off the end if only people stopped looting for five minutes... by the way, there's this lone kid who can't find her family in the panic, etc. etc. etc.) A skilled GM knows when to diverge and when to converge subplots, and when to begin in the beginning and end in the end or whether it's a better idea to throw in some en media res and cliffhangers.