Problem-Solving Characters

version 2 by Ron Newcomb

  • Home page
  • Beginning
  • Previous
  • Next



  • Section: Similarities to Narrative Scenes
    In an effort to inspire more narratively driven works that incorporate reactive characters, I'd like to share some interesting parallels between how narrative scenes are constructed and this extension in order to fire the imagination.
    * Kurt Vonnegut said every character should want something, even if just a glass of water. In this extension what a character wants is represented by the top-level action in that character's agenda, as a parser command: GET WATER. Or, TAKE WORLD. Or, FATHER, NO.
    * Attaining that desire isn't going to be easy, else why read about it? There will be obstacles and there will be struggle, and by the end of it the character will be irrevocably changed by the experience. This extension represents obstacles as check rules, especially ones custom-made by the author for the particular characters in the particular story. (Most of the standard rules in Inform's library merely enforce the basics of physical reality.)
    * When Hindered By rules are, effectively, character reactions.
    * A scene highlights a character and an attempt to get what is wanted at that moment, such as an item or a solution to a problem they're thinking about, or some kind of acknowledgement from someone. (The main character of a scene isn't necessarily the main character of the story.) The scene ends in one of three ways: the character gets what they want, or they don't, or it's postponed. These map directly onto the three major statuses of an attempt object: successful, failed, and hindered.
    * A scene has three parts, beginning, middle and end. The beginning introduces who the scene's main character is, where they're at, what they're trying to do, etc. Generally this part of a scene wouldn't be interactive, because it would be asking the player to make decisions before the author has oriented them. Instead, it's a fine place to mention things the player has done, to both "prove" to the player they can affect the story world, and hold their interest through a long non-interactive description by briefly focusing on them. "What the character is trying to do" and "what a character will do next" are phrases in this extension.
    * The mid-scene usually has the conflict, whether physical or verbal or internal or some combination. It has a to-and-fro sense to it, almost like a dance. To observe a conflict is to see multiple small obstacles and circumventions in rapid sequence. The mid-scene is where the parser gets to do a lot. The mid-scene can be choppy and full of dead-ends plot-wise, which can cause a player to focus attention on the game elements over the writing.
    * The end of a scene shows whether the character got what they wanted, and sometimes how they deal with it. In a CYOA this is a fine place for the three or four options on what happens next. In a parser work the commands might be at a slightly higher level of abstraction, like VISIT somewhere or someone, or INVESTIGATE some case or place, and so on, for which a new scene would delve into the details. The end can also be a good place for foreshadowing, since the player is no longer concentrated on in-the-moment game elements. To this extension, the end of a scene means an untried action in that character's agenda is about to change status to failed, succeeded, or hindered.
    * If the player-character is the main character of the story, and the walkthrough for the game were formatted to fit the AGENDA testing command, then the agenda would be the story's plot outline. If the player-character is NOT the main character of the story, they can be positioned as a third-party to the story's central conflict. This allows very interesting choices for the player on how to relate to the main conflict: choose one side over the other? Mediate between them? Escalate the conflict? Or perhaps a three-way dispute would be more interesting. If you ignore it to pursue your own sandbox goals, will the other characters allow you to do so?
    * If puzzles seem out of place but something 'crunchy' is still desired, one alternative could be the type of turn-based, multi-player gameplay found in ordinary board games, card games, and sports. These games differ in how well players can determine who's the current winner, and in how much players can interfere with each other's plans. Golf and bowling play out like several side-by-side games of solitaire, while Settlers of Catan is a board game of Screw Your Neighbor married to a card game of Screw The Winner, where every road ends in a fork. This extension allows the "how to play" knowledge that drives the NPCs to be broken into bite-sized pieces spread across When Hindered By rules. (More sophisticated what-if analysis would require the extension Hypothetical Questions, by Jesse McGrew.)
    Inform's standard actions allow characters to navigate and manipulate a mostly physical world. But by writing actions to convince, coerce, support, offer, blackmail, protect, teach, conceal, investigate, chase, negotiate, lie, entrap, broadcast, steal, inform, pressure, manipulate, fool, bring to justice, love, and all the other things that people don't do to medium-sized dry goods, the above framework becomes a little more interesting.
    [some conversational actions in a verbal conflict:
    "you cannot say that to me because" (hypocrit?)
    "that is incorrect because" (reason?)
    "your view of the same facts is distorted because" (ad hominem?)
    "I believe(d) this because" (confirming or excusing)
    "I had to / was forced to, because" (defending)
    "What (or why) was..." (learning)]