Constraining Interaction to Create Emergent Narrative Greg Costikyan

Constraining Interaction to Create Emergent Narrative Greg Costikyan CEO, Manifesto Games

Before 1973. . . l l People would have looked at you funny if you said something like “games are a story-telling medium. Chess? Monopoly? Candyland? Or even Afrika Korps?

In 1973, two things happened Colossal Cave: . . . and Dungeons & Dragons:

“Interactive Fiction” l l Colossal Cave was considered “Interactive Fiction” from the start. . . Though interaction is limited (few viable actions at each location) And as fiction, it’s not that interesting. Later games in the genre work better as fiction (e. g. , Tom Disch’s Amnesia)

Interactive Fiction l l l Text adventures no longer a viable commercial genre—but they live on as a hobby/literary movement (see www. ifcomp. org) Graphic adventures declining in popularity, but some still appear Leads also to “action/adventure hybrids” (e. g. , Psychonauts, Fahrenheit)

RPG l l l Boom in tabletop RPGs in the 70 s Direct inspiration for computer/console RPGs (e. g. , Richard Garriott’s Akalabeth, the precursor to the Ultima series, was based on his D&D campaign) Indirect inspiration for MMOs (via MUDs) Leads to LARPs In 21 st century, spawns the “indie RPG” movement of experimental RPG design Still commercially both in tabletop & digital games

Cultural Clash over Role of Story & Games From the Start l l In 1977, the Game Manufacturer’s Association (collection mainly of tabletop wargame & RPG publishers) adopts the name “adventure games” for its field (over the objections of wargame publishers who prefer “simulation game”) Every GDC (and before it, CGDC) conference has had talks debating the role of stories in games

. . . and Continues l l Today the biggest debate among game scholars is between “narratologists” (who view games as a form of narrative) and “ludologists” (who maintain they must be viewed as formal systems) No end in sight (despite by calls by some, e. g. , Janet Murray, for a truce)

Basic Problem: l l l There’s a central conflict between the demands of story and the demands of games Stories are linear. Though they can leap about temporally, they are experienced the same way every time. Games are non-linear. Though they are experienced over time, game sessions are different each time.

From Story to Game l You can put most games on a continuum from “story-with-minor game” to “game-withvestigial story attached”.

Cortazar’s Hopscotch l l l Two Paths. But really just a play with time (Proust/Remembrance of Things Past, Joyce/Ulysses, Vonnegut/Slaughterhouse Five) These are hat-tricks—not going to see a genre of Hopscotch novels But still interesting: This is the minimal branching narrative (one decision point) More game-like than a typical story, but still a long way from a game

Hypertext Fiction Robert Coover, Eastgate Systems, afternoon: a story (Michael Joyce) l Multiple choices at each node, netlike narrative l Generally not a predefined resolution, instead strives for the reader to have an epiphany after exploring enough of the narrative l But… not necessarily a good way to tell stories… l And… no goal, aimless browsing—not a good game l

Gamebooks l l l A/k/a “Choose your own ending” or “which-way” books Fighting Fantasy Branching narrative, sometimes rudimentary game system Lots of dead ends (but at least one ‘win state’) Basically the same as hypertext—follow a link to the next bit of text…

Solitaire Adventures & Paragraph-System Boardgames l l Solo Adventures are similar to gamebooks, but use the more complicated rules of a tabletop RPG, thus more potential outcomes Para-System: Boardgame, leading to occasional short gamebook style adventures with resolution. Tales of the Arabian Nights. – Considerably more replayable

Dragon’s Lair l l Arcade analog to gamebooks Two paths at each decision point, one leads to death. Popular when introduced (1984) because the first game with cinematic-quality visuals… But sequels failed, because this sucks as a gameplay concept.

Text & Graphic Adventures l l l More free-form: Not predetermined paths, but limited game spaces until new ones are opened (beads on a string concept) Free combination of game objects within spaces Not that different from a gamebook, except that the ‘text’ can respond interactively to you—new paths opened/items available

Graphic Adventures l l Characters (but limited decision-tree interaction) Cut scenes (but when overused, kill gameplay—e. g. , Tex Avery: Overseer) At best, this is a happy compromise: Compelling story, entertaining gameplay (e. g. , Grim Fandango) All games are structures—but graphic adventures quite constrained—necessary to ensure excellence of story

PC/Console RPGs l l l Ultima, Final Fantasy, Zelda, etc. Intimately tied to story, but far more freeform on a moment-tomoment basis. Often multiple ways to overcome obstacles Some choice of spaces to enter Character growth But one (or a handful) of outcomes, story experience not much different from player to player.

PC/Console RPGs (con’t) l l PC/Console RPGs still highly dependent on story—but a greater degree of freedom— more “gamelike” Limited repeat playability because tied to an essentially linear story

MMORPGs l l Large-scale environment, thousands of players Sometimes a “story of the game, ” but players have no impact on outcome—linear story irrelevant to gameplay. Mini-stories in the form of quests. Since the game goes on forever, and it is hard to allow players to meaningfully impact the world, real story is impossible.

MMORPGs l l l To add story, you need to bring the game to a conclusion: A Tale in the Desert… Or allow real changes to the world (but hard to do in a multi-server environment) These are “story settings”—but have almost lost the connection to story in exchange for becoming good social environments as well as good games.

Tabletop RPG l l l Game system very similar (sometimes identical) to PC/Console --but vastly more freeform: since there is a GM, players can do anything he deems physically possible. While there are “adventures” (=pre-written stories), most GMs create their own stories for their friends.

Tabletop RPGs l True ‘roleplaying’ for the first time—showing off for friends. – l (“Roleplaying” in MMORPGs is bogus, because no possible impact on game outcomes… ) “Stories” are created through play, and for participants, can be if anything more powerful than the ones they receive through interactive media…

Tabletop RPGs l l …but are invariably dull as hell if told to nonparticipants (expedition write-ups suck). Many RPGers don’t give story a second thought: more interested in roleplaying, problem solving, or character advancement (the Blacow player types).

The Continuum l Thus, you can view the continuum between story-with-minimal interaction (Hopscotch) through the game-with-some-storyconnection (tabletop roleplaying) as an attempt to find compromises between the highly linear nature of story and the inherently non-linear nature of games

Constraining Gameplay l l l I used to think that was all there was—there was only one dimension along which “narrative games” could lie. . . But maybe a better way of thinking about it is that to tell a satisfying story, gameplay must be constrained to ensure that story does emerge. . And reducing gameplay to interaction within “beads on a string” is only one way. . .

“Embedded Stories” l l Multiple stories embedded in the game—each linear, but encountered by players in different orders, thus improving replay value. MMO quests. “Paragraph-system boardgames. ” True of some (not all) console/PC RPGs

Beads on a String— But Multiple Paths Within Each Bead l l l Asset development for digital games is expensive—hard to get away from “beads on a string”. . . But you can allow multiple ways to solve each problem—and multiple ways to shape a character (fighter, sneaker, hacker). . . And multiple outcomes (victories of different game factions).

Ending the MMO l l The “never-ending” MMO with multiple shards essentially cannot permit players to shape the overall arc of the story, if any. But if you end the game, you can. ATITD has two possible outcomes: the players accomplish the tasks necessary for Pharaoh to triumph over the Stranger, in 1 year of play —or not.

A Tale in the Desert (con’t) l l l And high degree of player freedom during that year. Commercially risky—you lose a big piece of the player base with each game end. But artistically worthwhile.

My Life with Master l l Narrative arc is explicitly fixed (the villagers will destroy Master). Game explicitly played in scenes with beginnings, middles, ends. No dierolls for individual actions; actions are unconstrained. But a die-roll is made to determine whether the player “succeeds” or “fails” in this scene —and he must roleplay the results.

. . . The Constrained Narrative RPG l l l In other words, the game specifically constrains the players to an explicit narrative. . . Which can, however, vary greatly in detail from playing to playing. And unlike traditional RPGs, the burden of storytelling is shared among players and PCs.

The “Narrativist” RPG l l l More generally, a new breed of experimental “Narrativist” RPGs work to share the way the story is shaped among players and GMs E. g. , Ron Edwards’s Sorcerer, in which all players have paranormal powers, which they can use only by unleashing their inner demons—always at a steep personal price Not so much “games as stories”—but “games as theater”

Gamist-Narrativist-Simulationist Theory l l Evolved by Ron Edwards and other participants at The Forge Attempts to few RPG gameplay as motivated by a desire for accomplishment (gamism—”I want more EP”), a desire for exploration and verisimilitude (simulationism—”that’s not realistic!”), or a desire to participate in a compelling story (narrativism).

Bartle & Yee Player Types l l l Interesting overlaps with the Bartle (achievers, explorers, socializers, killers) and Yee (relationships, immersion, grief play, accomplishment, leadership) player types. . . But the motivation behind GNS theory is mainly to try to understand how to design games to shape narrativist gameplay. . . And it all ultimately boils down to figuring out what set of constraints on gameplay allows for a high degree of player freedom, and forces the emergence of a coherent narrative.

Can This Be Done Digitially? l It’s hard to see how (most) GNS-inspired games can be modified for use in digital media. . . Since they depend (as all tabletop RPGs do) on the creativity and flexibility of a live gamemaster (and live players). . . But. . .

Constraining One Place is Okay if You Free Up Somewhere Else l From this, we can learn at least one important thing: You can impose strong constraints on gameplay (e. g. , determine in advance the outcomeof a scene) if you free up player action in other spheres (no die-rolls for success/failure of individual actions) thus giving players the sense that they still have freedom of action within the system

How Else Can We Constrain Gameplay to Force A Narrative to Emerge? l l Worth thinking about. We need to get away from “beads on a string”—I think we’ve basically rung the changes on what can be done with that approach.

Approaches to consider. . . l l l Breaking the narrative into discreet chunks that can be encountered in multiple orders Having more chunks than will be encountered in a single play-through, so there are still surprises with repeat play Imposing a defined arc on the narrative (beginning and ending fixed) but allowing high degrees of freedom in between.

In General. . . l l Conceive of gameplay and story as discrete entities, and look for non-traditional ways for them to interact with each other. Finding different ways to grant players “freedom of action” while working within a constrained narrative—or ways of constraining player freedom in one area while freeing it in another to produce an emergent narrative

References l l l l Colossal Cave: www. rickadams. org/adventure/ Interactive Fiction Competition: www. ifcomp. org Graphic Adventures: www. adventuregamers. com/ Dungeons & Dreamers, Brad King & John Borland, Mc. Graw Hill-Osborne Media, New York, 2003 The Forge: www. indie-rpgs. com/forum/ Game Manufacturer’s Association; www. gama. org Janet Murray’s Di. GRA 05 talk on narratology/ludology: www. lcc. gatech. edu/~murray/digra 05/

References (con’t) l l Hopscotch, Julio Cortazar, Pantheon Books, New York, 1987 (originally published in 1966 as La Rayuela) Robert Coover: www. brown. edu/Departments/Literary_Arts/coover. ht m Eastgate Systems: www. eastgate. com afternoon: a story: www. eastgate. com/catalog/Afternoon. html

References (con’t) l l l A Tale In the Desert: www. atitd. com My Life With Master: www. halfmeme. com/master. html Sorcerer: www. sorcerer-rpg. com GNS Theory: www. indie-rpgs. com/articles/1/ Bartle player types: www. mud. co. uk/richard/hcds. htm Yee player types: www. nickyee. com/facets/home. html