Read Chris Crawford on Interactive Storytelling Online
Authors: Chris Crawford
GAME DESIGNERS HAVE NEVER
given storytelling a high priority. This is not to say they refuse to make some efforts in a storytelling direction. After all, they like to wrap themselves in the glamor and legitimacy that Hollywood enjoys, so they happily claim as much storytelling territory as they can convince others to accept. But their design decisions reveal their true priorities. Game designers put all their money on action, violence, and cosmetic extravagance. They see story as an embellishment they can tack onto their basic design, but certainly not a fundamental component of their products.
Therefore, after the game design is fairly complete, they hire “storytelling talent”—preferably someone with a Hollywood pedigree—to “storify” their games. The Hollywood talent is presented with the basic scenario and game structure and instructed to weave a story into the existing design. The storytellers do the best job they can under impossible circumstances, but their contribution is an embellishment, not a central component of the design. The story goes along for the ride.
An example of the second-class status given storytelling in game design is the classic Deus Ex. Game designers hold this game in high regard because it has such a strong storytelling element. Storytellers look at the same game and ask, “What story?” Sure, there’s something within the game that matches the academic definition of the term “story,” but Deus Ex sports a thin shadow of the real thing—and this is the game that game designers hail as a triumph of storytelling in games! There’s no question that vast talent and energy were devoted to the game side, but the storytelling side comes up short by any normal standards.
Nevertheless, game designers have slowly bumbled toward a conceptual approach to interactive storytelling that merits consideration. I call their idea the
environmental approach
. The basic concept is to create a large three-dimensional world broken up into buildings, rooms, caves, and other regions; equip that world with a variety of weapons, tools, and resources; and populate it with a variety of weakly interactive agents. These agents are seldom given any character to speak of. In general, they’re dispensers of information or providers of resources. Players move through this three-dimensional world searching for goodies in pursuit of some simple objective (kill the boss monster, save the princess, acquire the jewel, or some such). Along the way, players must deal with a variety of obstacles and make use of a variety of resources. Agents are placed and equipped to provide these obstacles or resources. Thus, players get the chance to bring all their videogame skills to bear: running, jumping, climbing, shooting, searching, acquiring, and so forth. Occasionally they might need an agent’s assistance, which is normally done through some mechanical process, such as presenting the correct gift, giving the right answer to a question, or carrying out a required task. So far, this environmental approach has accomplished nothing significant in the way of storytelling.
Emergence
, a hot buzzword these days, refers to the notion that complex systems can produce behavior that surprises us with its organization. One of the earliest examples of emergence came from the old computer program The Game of Life. This program introduced the concept of
cellular automata
, a map or array of intelligent cells, each of which followed some simple set of rules for its behavior. Cellular automata inspired a generation of research work and led to the creation of that great game SimCity. In the original Game of Life, a simple set of rules for cells produced a system capable of building complex phenomena such as “gliders,” moving patterns of activated cells, and “glider factories,” stable systems of activated cells that produced gliders at regular intervals. As the theory of cellular automata matured and interacted with ideas from complexity theory, the concept of
emergent phenomena
arose: the idea that sufficiently complex systems can generate even more complex phenomena that the system’s original creators never expected.
Most of this research was carried out with computer programs that created artificial systems and then generated their behavior. Although the academics who advance this field of research have developed and refined it considerably, a bastard form of the concept has trickled down to the programmer community as a fervent belief that big complicated systems can be made to produce almost any desired result, if only programmers make them big enough and complicated enough and give them enough time. From this belief the concept of
emergent story
has arisen, the hopeful fantasy that somehow, if programmers diddle around with complicated systems long enough, they’ll eventually get a story to emerge.
This isn’t the first time the concept of emergence has been eagerly seized upon. Starting about a thousand years ago, a concept that could be called “emergent chemistry” achieved some popularity among Western Europe’s intelligentsia. They figured that if only they could mix enough smelly chemicals in complicated enough ways, they’d eventually figure out how to cook up gold. For hundreds of years these people messed around, but they never did get any gold. (Ironically, the same chemical properties of gold that made it so valuable were what foiled alchemists.) Some alchemists gave up on the search for gold and instead searched for something that seemed easier: life. They didn’t get far, either. But their spirit lives on; some people simply refuse to learn the basics. You can’t just throw together a bunch of bits and pieces and expect wonderful things to happen.
Just as alchemists had to understand the basics of chemistry before they could start cooking up interesting chemicals, storybuilders are just going to have to understand the basics of drama before interactive storytelling works.
Lesson #18
Emergence is
not
the same thing as magic.
Can the environmental approach provide a foundation on which genuine interactive storytelling can be built? To answer this question, consider the conceivable extensions to the design foundation.
The obvious approach is to make the world bigger. Add more buildings, rooms, tunnels, caves, doorways, alleys, and bridges. Give the player more space to explore and more places to discover.
Although this approach is popular, it doesn’t accomplish anything because the size of the universe has no bearing on the dramatic substance of the experience. Consider, for example, the list of stages in the original
Star Wars
movie:
That’s it: just 33 stages for the entire movie. I exclude several settings, such as fighter cockpits, because they can hardly be referred to as stages; they’re more like backgrounds against which a single actor performs. I also exclude a number of conjoined stages, such as the corridors of the Death Star, the alleys on Mos Eisley, and so forth. Even if you throw in all these secondary stages, you still come up with fewer than 50 stages total. Most exploratory games have more than 50 locations already; indeed, a single level of the old game Doom could boast that many distinct locations. Therefore, games already have all the stage locations you need.
The 3D games are chock-f of interesting objects: weapons, ammunition, medical supplies, tools, money, sources of energy, keys, books, and so forth. You might think that perhaps you could find a story lurking in an adequately large set of props. If only you add enough items to the object list, perhaps you could end up with something dramatically interesting.
This is wishful thinking with no foundation whatever. Stories aren’t about
things
; they are about
people
—that was one of the lessons of
Chapter 1
, “Story.” Adding more things to a game accomplishes nothing in the way of improving its storytelling potential.
Agents in many games are notoriously stupid; perhaps their stupidity could be overcome by building a large cast of one-trick pony agents who, taken as a group, can offer players some interesting dramatic interaction. Dramatic interaction cannot be subdivided into tiny fragments, however. You don’t provide romance by offering one girl who flirts, a second girl who goes out on a date, a third who kisses, and a fourth who falls in love. Storytelling requires characters with depth, not cardboard cutouts—and adding more cardboard cutouts doesn’t help.
So why not make agents who are smart enough to handle more interaction? This idea sounds more reasonable, and it could work. Character interaction is certainly central to any kind of storytelling, but that’s another strategy, one
discussed in detail in
Chapter 7
, “Simple Strategies That Don’t Work.” You don’t need an environmental shell to provide good character interaction. The character-driven approach is independent of the environmental approaches popular in so many 3D games.
Chapter 11
, “Personality Models,” also discussed some mechanics of the process of character interaction
.
Some designers have taken this route, hoping that more puzzles will make for a more interesting game. Indeed, more and better puzzles might make a product more entertaining—but as a puzzle product, not a storytelling product. Puzzles aren’t central to storytelling; that was another lesson from
Chapter 1
. Many stories do have something like puzzles—problems that the player must solve or obstacles that must be overcome—but a story is considerably more than a series of puzzles. A goodly number of products over the years have provided a series of puzzles, such as The Seventh Guest (1992) and Myst (1994). Both games were commercially successful, but nobody would claim they point the way to interactive storytelling. They succeeded as puzzles, not stories.
Some designers have speculated that the classic “journey structure” of storytelling can be built on top of an environmental base. For example, you could build something analogous to Mark Twain’s classic
Huckleberry Finn
. The player, as Huck, starts the story in his home town and can wander about there as much as he wishes, but the only way to push the story forward is to get on the raft with old Jim and head down the Mississippi. As the player travels down the river, he’s free to wander around the stages placed along its banks. Each stage can offer an interesting puzzle or spectacle, but the only way to advance the plot is to move farther down the river. Eventually the player, having experienced a variety of adventures, reaches the end of the journey, and the story reaches its conclusion.
This structure has much appeal. Players have plenty of freedom to explore each stage along the river as deeply or as perfunctorily as they like, but the river’s gentle flow always pushes them forward. This approach doesn’t seem as heavy-handed as some other methods, yet keeps players moving along the intended plotline.
The minor objection to this approach is that the dramatic universe has only so many rivers; you can’t solve the problem of interactive storytelling by putting every protagonist on a raft. Sure, you can come up with variations on the theme: wind-powered craft that carry the protagonist, like Odysseus, to a variety of locations on the journey home; train trips that offer stops at intermediate destinations; road trips down Route 66 on the way to LA. Ultimately, however, this approach is an artifice, not a solution; it can be applied a few times, but loses its charm when it’s overused.
The killer objection to this approach is that it does nothing, in and of itself, to provide drama. A journey is merely a thread that ties together separate components of a story; the story’s success depends on the quality of the components. Therefore, the journey structure begs the question by shifting the problem to the components while doing nothing about the components themselves.