World Models vs Narrative Choices

Broadly speaking, there are two main structural forms of interactive fiction:

  • In parser games, the player has a variety of verbs to use as world modifiying or introspective actions and usually the possibility of moving between spaces. As a consequence, they tend to be less novelistic and more puzzle oriented.
  • Choice fiction (pick-a-path, choose-your-own-adventure, hypertext, etc) uses pre-defined or dynamically generated links between sections that describe a specific action or movement in the context of the story—the player can only act or move given the choices that are available.

Leaving aside the experiential/affective considerations (which I’ll hopefully discuss another time), the core distinction here is between the story unfolding through actions modifying a world model and the story developing through predefined narrative branches.

Parser games based on a world model tend to be best suited to telling a story through puzzles and exploration, whereas narrative choices fit better with works based around a dramatic structure.

It’s interesting to consider whether these approaches are mutually exclusive, and if not, how they could be combined.

The reason why I decided to go with narrative choices in Muturangi was that the linear narrative constraint was better suited to the manuscript I already had. I had to resist the temptation to mire myself in some kind of dynamic conversational multi-agent system which would be larger and more ambitious than what I’m already doing. Something I’d probably never be able to finish.