Who’s In the Driver’s Seat?

If there is one thing every writer knows, it’s that good stories are about characters. And every game designer knows that if games are going to be taken seriously – or better yet, praised as “cinematic” – they must have good stories. So, it’s a simple matter of deduction: Games should tell stories about characters. And the market seems to get it, because every year games lay on heavier stories with an increasing emphasis on character development.

But in fact, no matter how airtight the logic or lucrative the rewards, the decision to hitch games to character-driven stories is a misguided one that seldom produces respectable stories and even less frequently yields the best thing games can offer: fulfilling and enjoyable gameplay.

The problem is game writers refer to the “player character” as a single entity, when it’s in fact a compound creature, like Jekyll and Hyde. And, as the poor doctor discovered, only one can be in control. Either the character is driving the game’s narrative or the player is.

Writer’s Block
There are inherent obstacles to telling character-driven stories in videogames. Some are technical (like the extraordinary difficulty in animating characters, particularly their faces), some are budgetary (like the cost of good voice acting). Stories about characters rely on subtlety that is very difficult to convey with stiff models and bombastic voice actors. There’s also no career path drawing talented writers into the industry, meaning a lot of the writers who do find their way into games have come because they haven’t found more traditional success. But that’s just the tip of the iceberg.

Writing better stories for games is not, ultimately, a matter of putting Shakespearean words in Mario’s mouth, because Shakespeare was writing fixed stories for passive audiences. Hamlet wouldn’t work as a game, because the player could never be expected to act – or rather, not act – as Hamlet does, and so the player would be forced to behave differently than his personal character.

For a game’s story to complement, rather than obstruct, the gameplay, it can’t be fixed in advance but must be responsive to the player. While it’s relatively easy to make things like exploding barrels or bloodthirsty orcs respond to the player’s free choices, it’s difficult to make complex psychological characters respond plausibly. To make the characters work in the story, the designer must take away the player’s ability to affect them as he likes. Most obviously, characters necessary to the story can’t be killed, and this immortality leads to a host of artificial gameplay constraints or narrative leaps.

But in some ways, it’s easier to excise characters from the story than it is to make them respond in a rich and enjoyable fashion to all the things the player might want to do. The fact the rather clumsy interaction and wooden interlocutors in Façade have been noted for their responsiveness and human qualities underscores the problem. Moreover, good characters have arcs, and it’s tricky to craft compelling character arcs outside a linear story.

The problem is particularly acute with a specific character present in every game: the player character. In a good adventure story the protagonist should drive the action, and in any self-respecting game, the player character should be the protagonist. But the more the player’s “character” becomes integral to the story, the less the player is able to define the PC because the game’s design will fix the character’s role in the story (regardless of dialogue trees and alternate endings).

Turning the player into a passive audience dissociates him from the PC in a way that cleaves the game into two parts: playing a game and watching a story. This loses the best thing games can offer: the chance to be a hero (or villain). Attempting to infer a personality for the character, on the other hand, can force the designer to put fairly bland dialogue in the PC’s mouth.

So what’s to be done?

Down in Back
Although it’s worthwhile to try to find a way to make character-driven stories dynamic and responsive, there’s an obvious alternative: Move storytelling away from the character-based model by pushing any fixed, character-driven aspects into the game’s back story.

In almost every genre where storytelling forms more than a superficial part of the experience, there are games with compelling settings and interesting narratives that don’t rely on characters to drive the action. Consider Myst, Daggerfall or X-Com. While these games might feature characters, the characters were typically part of the back story, and the player’s avatar was the one pushing the action forward.

In each of these games, the designers fitted a story to the game and to the engine’s limitations. Myth‘s story, for example, is one of war, and it compares favorably to the stories in RTSes that tried, uncomfortably, to tell stories about characters using an engine suited for depicting impersonal hordes of soldiers.

All of these games share the feature of a rich back story. Indeed, the player’s goal in Myst and X-Com is in large part a matter of uncovering that back story. But in each game, there is an ongoing narrative defined by the player’s actions as well, simply one that’s not told through dialogue or character interactions. The player in X-Com experiences an extraterrestrial onslaught, the steady improvement of Earth’s technology by cannibalizing alien devices and so forth, through the gameplay itself. In this regard, the player and the character are almost perfectly attuned. Similarly, the player and the character come to understand Myst‘s strange worlds together by working through the mechanics and physics of puzzles. Daggerfall, for its part, had no illusion of interesting characters – most are embarrassingly flat – but instead used background lore told through books or artifacts to create a living, complex world.

Next Generation
Daggerfall and X-Com also relied heavily on procedural content, content that creates itself based on how the game unfolds. And it is precisely this sort of content that might make a truly player-driven game work.

The unapproachable but genius independent game Dwarf Fortress relies on a procedurally generated world and dwarves to allow the player incredible freedom to choose the course of his action. As flat as the “characters” in the game may be, the narrative created by their interaction is exciting and complex, driven by action rather than by complex psychological dilemmas. At the same time, however, this narrative is so closely linked to the gameplay experience it becomes transparent. The player plays the story; he doesn’t watch it.

Procedural methods can generate richer characters than those found in Dwarf Fortress, as The Sims reveals. But it can’t generate the dialogue that character-driven stories seem to require. We can become attached to a sim or a dwarf, just as we can become attached to an X-Com squaddie, but they’re no Hamlet. It’s as impossible to imagine Planescape: Torment being generated procedurally as it is to break away from the rigid course The Nameless One must follow in that game.

Combining procedural characters and dynamically managed scenarios with the rich back stories in Myst or Daggerfall, however, would yield a world more “alive” than those found in most games today. The technical and creative obstacles to designing such a game are, in fact, much more modest than those inherent in character-driven stories. The main impediment, then, is the misguided notion that it’s worth sacrificing a player-driven game to achieve a character-driven story.

It may be uncomfortable for designers to let go of the wheel and let the player drive. After all, it makes for a bumpy ride. But in the end, that just makes the trip all the more exciting.