A couple of weekends ago I had a nice long conversation with Adam over at Barking Alien and amongst the many, many things discussed was the question of how the world evolves and progresses throughout the campaign.
So let's give an example. In a superhero RPG campaign the heroes become aware that Doctor Demolisher has been assembling parts to create a weather-control device that he plans on using to blackmail the Eastern seaboard (too soon?). After a few sessions of wrangling with DD's henchmen as they gather the parts they need for the weather machine, the players suddenly decide one session that they are going to pursue a sideplot and infiltrate the notorious gangster Peppermint Ice's nightclub to foil his diamond-smuggling racket. Side note: I feel Peppermint Ice's moll sidekick should be Ice Ice Baby.
So, as a GM watching your players wander down the trundle path of a secondary plot, what do you do with Doctor Demolisher? Do you have him patiently wait around, hand poised on the level ready to throw, until the PC's finally show up and then he can do his "hands milking the giant cow" monologue? Or do you have him win?
Adam said that early on he learned a game-mastering lesson from a mentor that basically said this: at first the villains don't know about the heroes and don't care about the heroes--the campaign is about how that changes. But if the players don't interject themselves into the way of the villains, then the villain's plans progress accordingly.
That means that as a GM, in Adam's playbook, you have to be willing to let the villains win, and I think that's a bold step for a lot of people. We are so caught up with "RPG as story" that we can't envision a tragic ending, so we do everything we can to ensure that we get the ending we want, or feel like is appropriate. Maybe I'm wrong here, but even in pre-generated adventures I've seen where the final boss fight is a challenging one, the text often says something along the lines of "this encounter is extremely challenging for the players, and one or two player characters might die."
Right, because having a tragic death to highlight the momentous nature of the final conflict is acceptable, even expected. But I don't see a lot of "adventure paths" (whose name make me die a little inside) end with the notation "there's a good chance everyone is going to die here."
I think in many ways this is why certain universes play better than others. Most D&D games, for example, already have the villains having won. They're already set up in their subterranean lairs hanging out on their treasure piles and feasting on the local populace as they will. The heroes' actions move the dial from "all the way bad" to "it's a little better" with every tunnel they invade. Star Wars has a similar set-up: the Empire is running the galaxy. Note that this is one of my problems with The Force Awakens. I could never figure out if the First Order was some kind of lunatic fringe that just happened to have the manpower and resources and time to convert a planet into a space station that somehow could shoot a laser across solar systems and not have it take years for the beam to get there or if they were a re-skinned Empire still running most of the cosmos because taking out Palpatine didn't really upset the apple cart too much.
I would love to see a campaign where the players actually face the risk of failure and end up spending the next six months to a year of gaming time trying to fix their mistake. I just wonder how comfortable people would be with the change to the status quo.