I want to talk a little bit about establishing setting today. Also as a side note; writing this from Munich. Word.
Setting is important. If the game is going to have any immersiveness, the setting has to allow it. It describes the world that allows the plot to happen. It's the clues and hints for what is to come. I don't think any random concepts can do that (and more), however. To me there are a couple keys to a good setting that let it do its job.
Internal inconsistency always bugs me. When the idea behind a story is contradicted by the action in the story, all illusory disbelief is immediately broken down. This contradiction is often a simple oversight on the part of the writers, but the story is broken nonetheless. Often I bet that the logical inconsitency is discovered late in the writing process, and it is realized that the problem negates the original plot entirely. Wow. Sucks. The key to me is to determine the basis for the setting coherently. If an inconsitency is found, it must be reasonably explained; ideally prior to the viewer/player discovering it.
Of course, avoiding that entirely is better.
With fantasy, there's always an escape. The traditional 'A wizard did it'. Meh. That's lame. It may explain it but it does nothing for the audience. It's forced suspension of disbelief.
I'd rather see fantasy itself be coherent. Now, that doesn't mean coherent with reality. That's science fiction, in theory. Reality has all sorts of obnoxious physical laws that make things all normal. Who wants normal! Thus fantasy introduces magic as a mechanism for coherence. Magic lets the world be interesting. But is magic internally consistent?
I think it ought to be. But most of the time it isn't. Harry Potter is of course a gross example. I think the setting of the series would be stronger given a more defined magic, but clearly the world's populace does not think it was necessary. Mind I enjoy the books a great deal myself, despite my whining. I enjoy Jim Butcher's books more than otherwise because of the mindfulness to consistency. Now, it could just be that I don't read in depth enough to note the inconsitency, but I'll argue that this means that he did good enough to fool me. Butcher's magic breaks physical laws, but never without cost, and the cost is always proportional to the reward. I never feel like asking, 'Why didn't he just...?', but more often go 'He should do...', then he does. Or he comes up with some wacky (but simple) idea, that totally should work, and totally does, like arrows made of salt.
Back to me. :)
When developing Elementus, a pen and paper RP system, I built up a concept behind elemental magics, divided into nice symmetries and methods, each concept in its place in a nicely formed grid. The system was massively broken from a gameplay point of view, but that's neither here nor there, yet. I loved the apparent rules behind a concept that should be beyond rules. Eventually I tried to take the concept behind magic and explain other phenomena in the world. What I ended up with was, well, pretty out there. What I found, however, is that, to the players, none of this matters. At all! They really couldn't care less.
But it provided me with ideas, and ideas that came from the details tended to make coherent sense within the established rules. It's kind of like basic math. You start with basic axioms. Decided upon truths. And from those you derive the details. The details must then be based on the axioms, unless you're wrong. That happens.
So when something happens to my players (be me the GM or writer), I can explain why. When the players do something unexpected, I have a wealth of concepts to draw from to try and resolve it. When they say 'I think I should be able to do this', I can say 'Yes you can, and it is awesome'. Then they do it, and they feel awesome for having thought of it.
I want to reiterate that now, since I feel it's really the message of this entire post. If the setting is consistent, the player will become immersed in the story; while immersed the player can make reasonable guesses; the player can do something awesome. If the setting is not consistent, the player will be wrong more often than right. The player's options are reduced to scanning through all available options and hoping one works better than the others. Derivation is not allowed. No salt arrows. No creativity. It means that the game does not encourage thought, nor exploration, nor experimentation, and that the game cannot reward the player for engagement. What becomes of this is a toy, more than a world.
I've had some time to ponder the setting. I have some ideas that I like, and I feel that they are internally coherent, despite being a bit.. different. It's time to get them down in writing, and see if they hold up to the first order of analysis.