Making Choice Theory Mesh With Views/Controls/Constraints

September 14, 2007

So I’ve talking to Mendel Schmiedekamp about whether or not and how notions of choice and action can augment or benefit from his concepts of views, controls, and constraints (more info, [1], [2].)  I’m at the point now where I feel like I have reasonable notion of how such a theoretical framework might be constructed, but I’m not sure the two of us are quite on the same page.  I’m going to record how I’m thinking about it now here, and leave it subject to change later if something new pops up.  Nonetheless, I hope this explanation of my thinking will be useful as it stands.

Briefly, content is the actual occurrences of play.  Player A says X, Player B rolls some dice, the GM scratches his nose.  A view is a description of what content a player is paying attention to.  Controls are the internal forces that lead a player to take particular actions, such as change views, or contribute to the content in certain ways.  So, content is external to the player, and consists of observable phenomena.  Views and controls are internal to the player (which makes them partially observable, insofar as they are self-reportable.)

In a discussion of what the Fundamental Act of roleplaying is, I suggested “choice.”  (This is in contrast to two other options: “the process by which we agree on what happens in the fiction” and “socializing.”  It’s not really crucial to the subsequent theory whether or not choice is the most fundamental.)  Thinking down the choice path lead to these two definitions:

 Choice – The result of the individual, internal process of reducing a potentially infinite field of possibilities to a single possibility
Action – Making a chosen possibility into an actuality

So, can we use these ideas together to describe the process by which players contribute to a game?  I think we can.  Let’s start with content.  A player perceives content through their view, which reduces the sum total of contributions to the game down to those that a player is paying attention to.  This perceived information activates controls, which serve to eliminate possible actions based on a player’s goals for the situation.  Once the player settles on one possibility, they take action, contributing back to the content of the game.  Patterns will arise in the content, which Mendel calls constraints.  (It’s important to note that you can’t really talk about constraints in general, it’s necessary to discuss an actual instance of play.)

 So, what does all this let us do?

  • Talk about play on the level of a single player.
  • Talk about what kinds of information will be satisfying to perceive through a view (Mendel’s theory is information of intermediate complexity, which seems intuitively reasonable.)
  • Talk about what sorts of choices are available to a player.
  • Talk about what sorts of choices are satisfying to a player.
  • Talk about factors involved in individual decision making (i.e. controls.)

I think there’s more territory than that to explore, if, for example, we can connect this to goals, sockets, and payoffs.  It seems plausible that such a connection can be made, and I’m toying around with it a bit.

GNS as Resolution Criteria

August 28, 2007

While GNS “creative agendas” are often used in the context of what players want from a game, or what type of play a game system supports, I think of them instead as modes of resolution.  When deciding what events will transpire, the GM could take any of the GNS modes into account.  In gamist resolution, the primary criteria for deciding how the story procedes is “What sets up interesting challenges for the party?”  In narrativist resolution, the primary criteria is “What makes a good/interesting/exciting/fun story?”  In simulationist resolution, the primary criteria is “What is most likely to happen within the game world in response to these events?”

This way of looking at GNS promotes a slightly different usage of the terms “gamist,” “narrativist,” and “simulationist.”  When viewed as a creative agenda, one might say “This player is gamist.”  But when viewed as criteria for choosing what events come next, the GM is the sole deciding party.  We might say “This player prefers gamist resolution,” but only the GM can actually be gamist.  Another interesting outcome of this shift is that the 3 modes are no longer mutually contradictory.  In theory, a particular choice of what happens next can satisfy what would really happen, what sets up interesting challenges, and what makes a good story.  Of course sometimes these priorities will clash, and that will necessitate that the three modes be prioritized according to what is most important to the group, but they can all coexist in the same turn of events.

The three resolution mode questions, when prioritized by the group, make an excellent piece of technology for the GM to improve a game.  By asking themself those three questions when deciding an outcome, the GM can more closely match the types of resolution that the players prefer, and frequently satisfy players with different preferences at the same time.

  • What is most likely to happen within the game world in response to these events?
  • What sets up interesting challenges for the party?
  • What makes a good/interesting/exciting/fun story?

Choice: The foundation

August 24, 2007

The essential attribute which differentiates roleplaying from reading a book, watching a movie, or listening to a story is that in roleplaying, there is no audience distinct from the creators or performers.  All participants are active in deciding the outcome of the story.  The action they take is choice.  There are two main questions that need to be answered about choice:

  • Who has the authority to make which choices? 
  • What information can be taken into account when making a choice?

Who has the authority to make which choices?  The traditional arrangement is to divide the choices between the players and a GM.  The players make choices related to the actions of their characters, while the GM makes the choices about everything external to the characters (the world.)  Other games run GMless, with greater player control over the game world.  Beyond that, what “stances” are acceptable?  Traditionally, each player controls one character which they identify with (like an actor) and whose actions they control.  Is it OK to transition from “actor stance” to “director stance” to make statements about how the world responds to your character actions?

What information can be taken into account when making a choice?  Is it ok to make a choice based on what the player knows, but the character doesn’t?  This is often called “meta-gaming,” and frowned upon by many groups.  (For the record, I call it “active postmodernism.”  It breaks the walls of the reality.  I want to reserve the term “meta-gaming” for something else, specifically what I’m doing right now: thinking/writing/talking about gaming.)  Is it ok to make a choice based on what the player wants, as opposed to the character?  This is a less touched upon issue, but I think generally more contentious.  Choosing to stay with the party as opposed to going off by yourself is sometimes justified on these grounds.  “It makes the game more fun, so you should stay with them.”  “But my character wouldn’t do that!”