Rather than re-state all the stuff I've said in three other threads on this subject, I'm just going to take all the text I've produced on these issues so far and drop them into this post. Sorry the post is long. But it seems silly for me to try and re-state stuff I have just (or two years ago) said better.
I apologize, in advance, for the fact that all of these responses, while obviously explaining how I design worlds, were written in response to other issues or questions. Still, I think you can piece together pretty clearly what I'm on about and why it would be a hassle to re-type it. That said, I think I've just decided to use all these found quotes as the nucleus for an essay.
How I make worlds
In order to explain how story comes into being in my games, I sort of have to explain what the player-GM relationship is in my games. My role is essentially as a puzzle designer. The worlds I make, in many respects, are experienced like mystery novels: the reader gradually figures out what is going on; sometimes ahead of him, sometimes behind him, so does the detective. The reader, equipped with a slightly different data set than the detective enjoys the process of solving the mystery, semi-independently of the protagonist. When my games work well, this is how they work. Essentially the entire world, campaign, etc. is a puzzle that both the players and characters ultimately decipher and solve, rectifying whatever flaw or evil act I have built into the system.
As a result, at the macro-level, my campaigns are formulaic. Every game involves the players and their characters discovering the true nature of their universe, discovering the thing disrupting it and then thwarting whatever that thing is. But below that top-level formula, there is near-limitless free will; yet, certain things inevitably happen because of the structure of the game world itself. In one sense, the players have complete control over what events take place but in another, they can only control the terms on which predestined events unfold.
One of the other fairly unique things about how my games interact with this thing called story is that they often can be read referentially and symbollically as well as at a more literal level. My worlds tend to be situated on alternate earths and often signify to the players by invoking certain common tropes and cross-cultural myths because these things are actually part of the structure of the world itself. For instance, in my current campaign, the characters will inevitably progres through seven cities which correspond to the seven forts of the original grail quest Spoils of Annwfn in the Book of Taliesin, the seven levels of sufic consciousness, the Seven Cities of Cibola and the Seven Cities of Antilla. This resonance is a cue to the players as to what is taking place in the game that the characters can never comprehend in quite the same way. The characters can do whatever they want and go wherever they want but, unless the campaign ends inconclusively, they will end up at the City of Glass. In what way will the City of Glass be the City of Glass? Where will it be? What will happen there? That's up for grabs. But it's certain that that is the end point. There are various other ways that the game is operating on this half-thematic, half-cryptographic level by tapping into allied symbol systems and cross-cultural myths that are just as predictive and less deterministic. (I'll save an example for any requests for clarification.)
The advantage of running games with a navigable and discoverable metatext is that there is an alternative predictive model available to the players in addition to the rules. This allows the players to guess at future events and, if they can arrive at in-game justifications for doing so, thwart or reshape events that they see looming in their future.
Are there blank areas on the map?
If I create a world based on certain principles and a player unilaterally creates a piece of it that violates those principles, then the central plot, theme, etc. of the world can easily fall apart. Because I don't tell my PCs what the big universe-building equation behind my world is, we have to work a little differently. The player has to ask "If I wanted to play someone from over there, what would they be like?" Now, plenty of things about "over there" are up for grabs but some portion of them has been predefined because they inhere in the structure of the world itself.
Let's say you had a bunch of ancient Hellenistic players gaming in a particular world and you were working on character background with them for a campaign set in this world. "Well my character is from some Brahmin-like culture," one might say, "but he's a Platonist (or Stoic or Pythagorean or whatever) and worships Poseidon." Because of the way people in the Classical world understood the world, they assumed that philosophy inhered and gods inhered in the very structure of the world and that no matter where you went, there would be Pythagoreans and temples to Poseidon everywhere. You can see that there is a basic conflict between two worldviews about what is local and particular and what is universal and entailed by the structure of the world itself. In this case, the inherent structure of the world you have built (this one) says that culture, philosophy and gods are local and particular and cannot be deduced from non-local data. Conversely, your player might say, "Well, my character will be very surprised by how travel works around here given that where he's from, water usually flows up hill and the sea is often burning." Here, he believes that the properties of water can vary at the local level based on regional particularities, whereas you know that the properties of water are universal in your world.
Every world has a structure that implies that certain things can vary locally and individually and certain things are universal. Many people who play RPGs seem to like putting new things in the the "local and individual" column but rarely take anything out of that column and slot it into the "universal" category. I find the most vibrant and interesting worlds are ones that switch a few (not too many) between the two columns.
Examples of navigating on the metatextual level
A metatextual reading of a novel or a a campaign world is about decoding the symbol system the author/GM uses in order to hypothesize about what is going on.
Example #1:
When I began reading George R R Martin's Game of Thrones, I was struck by the emerging civil war between the Lannisters and the Starks and its obvious echoes of War of the Roses. Given that I knew Martin was referencing the War of the Roses, obviously the character of Tyrion was some kind of figure of Richard III. I then wondered what aspects of Richard III, aside from physical appearance Martin would invest in him. Obviously, the character wouldn't be an exact correspondence because Richard III was a York not a Lancaster. So, when he ended up in charge of King's Landing and his two nephews' care, I was quite delighted, especially the way Martin ended up inverting the Richard III myth and having Tyrion end up imprisoned in the tower.
Of course, all this got more complicated when we met the Dornishmen later in the series and realized that what Martin was actually doing was overlaying the Seven Kingdoms of the Reconquista overtop of the Seven Kingdoms of the Heptarchy so that Cornwall and Granada were synthesized into Dorne.
Example #2:
This is my favourite long gaming anecdote that I have tried to reduce so that it doesn't take 30 minutes to tell. This is the experience which sold me on metatextual gaming.
I played in a game in which the characters lived in a world called Midgaard. It was one of the nine worlds which were, in this specific order, Vanaheim, Elfheim, Midgaard, Asgaard, Jotunheim, Svartelfheim, Niflheim, Utgaard and Hel. There is no way I can replicate the incredible richness and genius of this campaign-- it is the second-best campaign I have ever been in. There were many stunning realizations that I cannot do justice to. But eventually, we the players figured out that the nine worlds actually corresponded to the outer nine planets of our system; once we removed Mercury from the model, suddenly, we saw the alphabetical correspondence. Venus, Earth, Mars, Asteroid Belt, etc. The only thing that didn't fit was Hel. This profoundly informed our reasoning as players-- the home of the gods was not a planet. Had it been destroyed in Raagnarok as our myths implied?
We hypothesized like mad about this, as players, because our characters lived in a magic-rich medieval-style society on Midgaard. Our characters could not experience the realization that the adventure was taking place sometime in the past or future of our solar system.
Eventually, our characters were captured and enslaved by Dark Elves and taken to a special building they had discovered that they needed our magical affinity to understand. The building was millions of years old; the back room was occupied by some kind of enormous magic engine that seemed to affect time in some way. The front room was occupied by a series of high seats looking out at the sky through an enormous bay window and spread out infront of the seats was a huge computer system which seemed to glow in all colours of the spectrum that we couldn't figure out either.
We went away puzzled from the session, feeling like we were on the verge of figuring out what was happening. In the shower, 5 days later, it occurred to me that even though the building was rooted to the ground, the engines in back indicated that it was not a building but a ship. And what is that part of the ship faces out into space? The bridge. And what colour is the computer? ...Obviously, we had located the Rainbow Bridge -- the building's purpose, therefore, was to take us to Asgaard!
Sure enough, armed with this understanding of the significance of where our characters were, we focused all our energy on getting the computer to run because we knew it was the Bifrost Bridge. (We later discovered the Dark Elves' code name for their archaeological dig was Project Heimdall.) Sure enough, the building transported us back in time to when (based on the Russian Phaeton hypothesis) the Asteroid Belt was a planet; and, living there were the creatures our characters called the gods. At every stage, our characters' motivations were justified in terms of their own world's events but, as players, our choices were based, to varying degrees, on our understanding of the world's metatext.
[/end examples]
Most fantasy worlds resist a metatextual reading because they are not designed to be figured out in this way. Most people who create fantasy worlds use the Jungian/HP Lovecraft method of reaching into the collective unconscious and pulling out whatever jumble of myth, history and invention they find. But I think the best writers and GMs create worlds that can be understood on both a textual and metatextual level. In my best games (ones where I'm not hampered by the D&D rules system), every puzzle I create is solveable on both a textual and metatextual level. The brilliance of the Midgaard campaign is that we still could have used the ship to take us to Asgaard without the realization of what it was.
More on fractals and what world equations look like
First off, I'll begin with my metaphor. Imagine that you're looking at an attractive abstract pattern on your computer screen. It may be that someone has carefully drawn this pattern pixel by pixel or it may be that the pattern is actually a Julia set-- an infinite abstract pattern generated by a single complex equation.
Most campaign worlds are constructed pixel by pixel -- some gigantic bitmap that someone has lovingly created; let's call that the imaginative campaign world. The kind of world I find most exciting is one which is, instead, generated by a single complex equation; let's call that the mytho-poetic campaign world. The difference between the two types of world is what happens if you zoom in or move off the screen. If you zoom in on a section of the imaginative campaign world, depending on its resolution, you will either find a carefully hand-drawn scene or blocky pixelated images. On the other hand, if you zoom in on a section of the mytho-poetic campaign world, you will initially see blocky pixelated images-- these will then resolve to images as detailed at the larger image. If you move off the screen of the imaginative campaign world, there will either be nothing or another, adjacent bitmap. If you move off the mytho-poetic campaign world, you will experience much the same thing as if you zoom in: blocky images resolving into smooth detail. Of course, the speed with which the image resolves is based on two things: the speed of the processor and the difficulty of the equation.
Sorry for the lengthy extended metaphor. So, what does a good campaign world equation look like? A good mytho-poetic world is constructed much the way you make a conspiracy theory: substitute correlation with causation. Examples of good world equations: What if the 7 angels of the 7 churches of Asia plus the Son of Man are the same people as the Eight Immortals of the Tao? What if Arthur's original quest for the Holy Grail described in Spoils of Annwfn took place in the Americas and the key grail artifacts were actually the key artifacts described in the Book of Mormon? What if the nine worlds of Norse myth were actually the outer nine planets of the solar system?
Start with one preposterous instance of a myth system, science, historical event or epic story appearing symetrical to some other myth system, science, historical event or epic story and you'll soon find other correspondences. In the Eight Immortals story, there was the text in the Book of Revelations that only someone whose name was written in the Lamb's Book of Life could enter the New Jerusalem taking on new meaning when it was noted that one of the ways the Monkey King of Chinese myth became thrice-immortal was to erase his name from the list of fates of all mortal beings. With the Holy Grail thing there were the different forms of the grail corresponding to different Mormon artifacts: the Lance of Longinus/Javelin of Teancum, Sword of Laban/Excalibur, the grail carved from the emerald which broke from Lucifer's crown when he was cast down from heaven/one of the Seer Stones of Zarahemla.
So, to write a mytho-poetic campaign, you come up with your conspiracy theory-like idea. Then you do heaps of highly selective research to find little facts which, taken out of context, make your theory appear credible. Then, you should be able to deduce roughly what each part of your world is like based on the interplay of these ideas. Of course, you still have to do work when characters do something unexpected and you need to do some episode prep but I find that I do 50% of my total campaign prep at the beginning while I'm fashioning my conspiracy theory and the other 50% preparing for individual episodes. But I find my total campaign prep time is about the same.
Can these games be played without the players buying into the metatext?
Yep -- the risk that everyone will choose to play the game on the mechanical and textual levels is a very real one. In fact, the first such game I was in, I was introduced late into play and the GM was amazed that I actually got the rainbow bridge reference. He had just put it there for himself.
Which brings me to the most important thing I have to say as this thread winds down: I can, in no way, take credit for the theory of mytho-poetic construction and metatextual play of game worlds. The person who turned me onto this was Philip Freeman, the most brilliant person I have ever met.
I used to fill reams of notebooks on world-building back in high school and college, and now I'm burned out on it.
