For this week’s theory/criticism post, I’ve invited over Calvin Ashmore, a close friend and Digital Media student at Georgia Tech. Calvin studies game adaptations, both in tabletop and electronic forms. He hopes one day to make a Pride and Prejudice game, which should tell you something about why we’re friends.
Calvin’s been working with a concept called “story worlds,” which has a lot of bearing on how games construct and relate to fiction. It’s a different perspective from what we get in tabletop theory, so I thought it would be interesting to share.
Stories are not static. They move, they change, and the things that happen in them happen for reasons. Story worlds describe narratives using rules. When events in a work of fiction unfold, they do so because it is appropriate for them to do so. A common rule to hear in fictional writing is that if a gun is introduced in the second act of a mystery story, then it needs to be fired in the fourth. This rule is not just a guideline or a convention, it is a mechanic. Rules such as these indicate how fiction works, to the degree that if a story does not follow them, then it feels broken or dysfunctional.
Essentially, story worlds treat stories as simulations. However, this does not mean that stories must be realistic, or simulate exactly what characters would do in a narrative situation. A simulation is an approximation of some other system, be it the rules of physics, the dynamics of characters, or the rules of drama. All of these are systems, they can be described by formal models, but compelling arguments can be made that we perform simulation in our heads during the very process of reading.1
In a world-centric perspective, plot is not static, it is mutable and volatile. There may be determined outcomes, but the essence of a story is in the path it takes to reach its destination. If we look at a story as having a world, then the story itself is just one of many possible valid paths.2 Story world theory pushes toward looking at narratives as fictions3, rather than as texts. The meaning from story worlds comes not from the plotted outcome of a story, but in the system of rules that map causes to effects. In this perspective, games and narratives have a great deal in common.
Applied to game design, story worlds are particularly useful for thinking about composing a design based on a source text or a setting. If it is possible to interpret a text to create a story world, then the world’s rules and mechanics can be applied toward the game design. This is useful for both adapting an existing IP into a single game (a huge number of video games are adaptations, after all), and also for adapting a setting (a huge number of tabletop roleplaying games are adaptations of settings). By using story worlds to inform design, it is possible to recapture some of the elements that make the story enjoyable and create a similar type of experience for players. Story worlds are useful for creating a sense of immersion, and also a sense of participation within fictional universes.
- Story worlds are abstractions of a text. The world models the rules of a story. Any model is a simplification.
- Furthermore, this abstraction is interpreted. If two game designers set out to make adaptations of one text, they’ll produce two very different games. That doesn’t mean that one is wrong and the other is right, but that there are many possible valid interpretations of worlds that can be made from a story.
- Story worlds are internally consistent. The world works in a way that makes sense to itself. The logic does not need to be real world logic, it can be dream logic.
- Most importantly, story worlds extend beyond the text. There is no point in developing a world that limits what players can do to be less than what happens in the text. It should be possible to see what other things exist within the textual universe, for characters to make different decisions than those that happen in the text, and to see what other outcomes are possible.
One of the most important issues from the perspective of game design is interpretation. What is the right level of abstraction for building a story world? How do we distinguish what is relevant in a story, and filter these into coherent rules? There is clearly no single right way to interpret a world from a text, but there are also interpretations that are better or worse than others. It is possible to identify five key criteria for judging story worlds:
- The interpretation identifies states (properties and information) and classes of events that are relevant in the story. This distinguishes the information that is relevant to the world from the information that is not. Details that have no consequence on mechanics are not part of the state.
- The Interpretation identifies mechanics that govern events and actions. What are the ways in which state can change, and what control do players have over it? How is player intention mapped to the actual effects of actions?
- The original narrative may be mapped onto the world. It should be possible for the world to describe the original story in its language. It is fairly silly to create a story world in which it is not possible to describe the source material using the world’s rules.
- The interpretation should allow reasonable departure from the outcome of the source text. Of course, what is reasonable may be a matter of contention, however, while it should be possible for the original narrative to be mapped onto the world, it should also be possible for the world to accommodate new narratives that are different. This point is the one that requires the most creative judgement, but it is also the most important.4
- The interpretation should be minimal. A converse to the point of reasonable departure is that story worlds should only include those states, events, and mechanics that are relevant. This point comes directly from game design. A story world which includes irrelevant features in the model invites players to use them, even if it goes against the theme of the world.
The first challenge is that games made from story worlds are not gamelike enough, namely that the rules for a story are not interesting for gameplay.5 My answer to this is “too fucking bad.” If the point of a project is adaptation, then forcing the game to be more gamelike by including gameplay that is irrelevant breaks the principle of minimality, sabotaging the adaptation.
The second challenge that often appears is that story worlds are not storylike enough, that they allow players to have freedom when the designer would much rather like the player to reach the interesting outcome of the story. My answer to this is “then write a fucking story.” If there is only one interesting path or outcome (a principle which is absurd if you know good players), then the design should be for a story, not a story world.
- See Keith Oatley: The Science of Fiction. ↩
- See Marie-Laure Ryan: Possible Worlds. Ryan argues for an approach to narratology that borrows elements from artificial intelligence, describing narratives using formalized rules, exploring how stories can exist within a network of possible worlds. ↩
- See Kendall Walton: Mimesis as Make Believe. Walton argues that fiction is part of a boarder class of representations, and in order to understand representation, we must understand play. ↩
- In a Pride and Prejudice story world, a reasonable thing to allow in the world is Elizabeth Bennett marrying Mr. Collins. However, allowing Elizabeth to run away and have a lesbian affair with Charlotte Lucas is less defensible. That would be much more in the vein of Wuthering Heights instead. ↩
- i.e., the pervasive notion that it’s not going to be a fun game if there’s no killing, applied to aforementioned Pride and Prejudice game. ↩