Feedback Loop: Many choices, or none, make a game.

Is just one choice all it takes to turn a novel into a video game? Before you say yes, consider when a game is created out of many choices and when we are left with none.

Richard Eisenbeis looks at Katawa Shoujo in an April 24 Kotaku article. Eisenbeis holds up the dating sim/visual novel as proof that one choice is all it takes to turn a novel into a game. It is a shallow analysis and the implication that one can just stick a choice in a novel and have a game is wrong.

If we step away from the screen with only Eisenbeis’s assertion, we lose out on understanding what developers have to do to take a story and turn it interactive.

Creating a good game means understanding the times when a million choices create an interactive work and the instances where no choices are required.

The Little Choices

A story is a sequence of events, so there’s a reason to ask the question: if a game cannot force its player into that sequence, can it tell a story? Kill Screen’s Darshana Jayemanne challenges the pat answer that linearity is inseparable from direction by challenging our assumption of what is linear.

The mistake is assuming linearity somehow eliminates choice. It’s clear this couldn’t be more wrong. Some games, like Max Payne, may set a plot in stone but still provide the player with tools to author their own story.

During its heyday, I played a lot of Unreal Tournament 2003. As I played, a complexity emerged from the game that enriched my experience. Even through the game has a bare-bones single player (a long tutorial really) I felt more about its characters, weapons and settings than some more thought-out game stories. Yet there is no way I could ever alter the plot, there are no in-game choices that will stop me from working my way up the ladder, fighting against certain rivals, and finally winning the fantasy tournament. Without a plot choice to be seen, UT2003 is still a game. So are Half Life, Diablo, Freelancer and all the other linear games out there.

These are all narratives that, from the perspective of a plot-writing developer, have nothing distinguishing them from what Eisenbeis might consider a novel version of Katawa Shoujo, one without a branching choice. What makes them games?

These games are made great by the plot created outside of the directed narrative, they benefit from a narrative of play.

This narrative of play emerges out of three primary elements. Story is the a sequence of events, the order of our actions in a game. Plot is the causality that chains events together, the motivation behind our actions. Narrative is a phrase that covers the method by which stories are told, from the grossly mechanical format to the techniques of delivery.

In fact a game has two overarching narratives. There is the plot and how it is told to us through cut scenes, voice-overs and tricks of the camera, the directed narrative of the game, scripted out by developers. Then there’s this other thing, a narrative of play, the millions of little choices we make throughout a game and how we interpret them. Where does this come from?

In their 2004 GDC presentation Randy and Harvey Smith talked about how to create emergent gameplay at even the simplest levels of gaming.

This looks like a chart of the last round of Team Fortress, but it's also a story.

The Smiths define emergence in games as the ‘combination of first order mechanics to create a second order dynamic.’ Which means that emergent gameplay occurs when the basic way that things work in the game combine in a logical but unexpected way.

Here’s a simple example from UT2003: Mechanic 1: You receive and can use the Bio-Rifle, which has primary fire and an alternate fire that launches globs of biosludge that sit around like short-lived mines. Mechanic 2: Maps have multiple levels that sometimes need to be traversed quickly. Mechanic 3: The biosludge globs explode when touched or left by themselves too long. Mechanic 4: Explosions throw your character. Dynamic: Jumping above a biosludge glob as it explodes pushes your jump much farther, allowing you faster and easier travel around the map.

This emergent mechanic is, in itself a story, but it’s also a plot. Doubts? Try this instead, re-read the above paragraph, but replace each step with an act from Freytag’s pyramid. Now Mechanic 1 is our Introduction, a brief explanation of how the weapon works so we can properly understand the item. Mechanic 2 is the Rising Action, our basic conflict, an obstacle standing in the way of our goal (domination of the map). Mechanic 3 is the Climax, a turn towards victory and a moment of drama, where we get the sense of how things will resolve, the majority of the action will be experimentation with this mechanic. Mechanic 4 becomes the Falling Action, with all the mechanics now understood for this weapon, conflict begins to unravel and we gain mastery. The Dynamic becomes the Resolution, mastery of the weapon gained through understanding allows us to overcome our obstacle.

It’s no mistake that this also roughly maps to how tutorials are designed. Look to Portal:

• Introduction: A player is shown a new game mechanic.
• Rising action: Interaction with that mechanic becomes more complex.
• Climax: The player is asked to solve the most complex puzzle that can be created with the single mechanic.
• Falling action: The mechanic is then combined with previous mechanics in easy-to-resolve ways.
• Resolution: Now that the player has mastered the mechanic, it becomes a regular part of new and more complex puzzles.

Sequence and emergent gameplay combine to create hundreds of little plots. Combine all these small plots and you’ve created an enormous web of causality without ever needing to give the player a choice in the directed narrative’s plot. You’ve got a game.

In any game, interaction equals plot. In a good game it takes more than just that.

What about the hypothetical Katawa Shoujo that exists without any branching narratives? There’s no skinning the clicks with weapons or spells. What could possibly emerge out of just clicks that would make a game like Tetris or Cow Clicker engaging? What is the other half of the narrative of play?

Choosing Your Universe

In the Smiths’ 2004 GDC presentation they cite a framework for emergence proposed by Marc LeBlanc called MDA. Part 1 is Mechanics, part 2 is Dynamics. The first two are the core of the Smiths’ emergent gameplay. The third part is Aesthetics, defined as “human emotional responses.” According to LeBlanc, Aesthetics is where emergent gameplay transforms into emergent narratives.

To understand what an emergent narrative is we need to look at a form of literary criticism called “reader-response.”

This school of criticism studies the formation of the reader’s response to a text, the interaction between the reader and the book. It looks at that interaction as a narrative that emerges from both the reader’s identity and the design of the text.

Students of game studies may recognize this concept from more recent ludological studies, like Sherry Turkle’s popular 2011 ethnographic book Life on the Screen, in which she examines video game players who use interactions inside of the game, but outside of the directed narrative of the game, to explore their own identity.

If even simple clicks, or no clicks, can create a reader-response in a text, then no choice is needed within the directed narrative to allow us to play. We don’t need a storyline handed to us to become the heroic but lonely space pilot defending a planet in Astroids or to imagine ourselves making slow heavy strokes with Hiro Protagonist’s redneck katana from Snow Crash.

In contrast, Bulletstorm‘s thin characters don’t work because the designer failed to create the right environment to support them and the player. The person who designs the narrative has a role to play in eliciting a reader-response. No matter what, at some level our experience with a game is still mediated by the depth and breadth of the universe created for us.

At the site What Games Are, Tadhg Kelly describes the design of a game’s universe and how it interacts with the player as storysensing. Really, this is the emergent part of narrative design. When it is done right, players discover a universe designed to flow around them, but not with them. Though a narrative’s universe may not exist outside of our point of view, when designers are successful, we feel that our plot is impacted by stories existing elsewhere, outside of our agency and view.

When the narrative design is successful, we are freed from the burden of being the center of the universe. With that freedom we are able to explore a game, not just through its scripted options, but through reader-response.

Journey is one of the best examples. [Please excuse the following spoilers for Journey.] It is a game in which your actions don’t really amount to anything in the directed narrative. No matter how long your scarf gets, in the end it is destroyed entirely and rebuilt entirely as the directed narrative requires. But it doesn’t matter because the mysterious ruins of the game and all the items and characters that inhabit it provide a depth whose exploration creates the type of narrative of play that critics sit around writing articles about and that the rest of the players can’t wait to discuss with their friends.

Successful narrative design is not accidental. To find failures, look no further than Bulletstorm, Deus Ex: Invisible War or the Kayne & Lynch series. You cannot just take any story, stick a choice into it, and end up with a game, certainly not with a good game. We can judge successful narrative designers by their grip on the game, it must be tight enough to give us structure, but not so tight as to extinguish our imagination.

Story Told

So, what’s the better answer to Eisenbeis’s question, how many choices does it take to make a game? Both zero and a million. No branching directed narrative is required, no interactive tools are needed at all. Or a million tiny plots can add up to an enormous complex narrative of play. In some cases, both.

Not everything that is ‘playable’ is a game, and not every game is playable with a controller. Let’s keep this in mind as we critique games and seek the next masterpiece in narrative design. In the meantime, I’ll be dreaming of Unreal Tournament 2003.