In January, at Slamdance, Mateas and Stern met some investors who were excited about interactive drama. Many phone conversations later, they had a deal to raise $2 million for a commercial game. This was a crucial step for them. Stern, in particular, sees himself heading a commercial interactive-drama studio. Both he and Mateas believe that today’s video games occupy only a fraction of the potential market for interactive-video entertainment.
“Most people—your sort of regular Joe or Jane on the street who loves television and movies—don’t really get a whole lot out of games,” Stern said, when I asked who would buy interactive dramas.
“I think there’s a real market for more character-rich, story-centered interactive experiences,” Mateas added. “I think potentially it’s a market that dwarfs the entire current video-game market. There is a huge untapped market for experiences that are not about action adventures, quests, killing monsters, and solving puzzles.”
They have given their next game the working title “The Party.” It is still in the conceptual stage, but they expect that, where Façade had two computer-generated characters, The Party will have ten, a far more complicated proposition, but dramatically richer. It will require not just two programmers but, once it enters production, ten or more. The graphics will be more detailed and polished. The action will take place in a larger space. The game will last about forty minutes, rather than twenty. It will support more physical action, allowing the player to do things like rendezvous with characters in a private room, lock doors, carry things around, and fire a weapon. It will, they expect, understand the player better than Façade does, and support many more player moves.
And its aesthetic will be different. If Façade is a psychological drama, The Party will be a darkly comic social melodrama, along the lines of Desperate Housewives. In the prototype scripts, you find yourself cohosting a dinner party with your wife (or husband, if you play as a woman), who begs you to keep the conversation and liquor flowing smoothly. As guests arrive, the party fills with characters who have various designs on you and on each other. Your ex-girlfriend may try to break up your marriage; her angry husband may deck you; your neighbor may be snooping and your boss fishing for excuses to fire you. You can try to keep everyone happy, or you can hurl insults, or seduce your best friend’s wife, or announce that you’re gay, or refuse to admit guests (in which case your wife may let them in while shooting you angry looks), or lock your boss in the basement. You can try to mind your own business and be left alone. At every stage, however, the other characters—and behind them the drama manager—are conniving to draw you in. Madcap complications ensue.
There will be sex in the game, and there will be violence. There will be a gun, but only one bullet, so no shoot-outs. Here again, the designers invert the conventions of Video-Game Land, where shooting people is easy but talking to them is hard: in The Party, violence will be rare and dramatically meaningful, ricocheting through the game, as in life, with unforeseen consequences. Sex, likewise, will be dramatic rather than pornographic. It may disrupt a marriage or get someone killed. The sex will not be X-rated, but it will be realistic. “You may not literally see it, but the characters will be moaning,” Stern said.
Mateas and Stern expect work on The Party to take two and a half years, at least. They hope to make the game a paying franchise and use the proceeds to push on toward their real goal: a game that understands natural language and generates its own drama.
The Party, like Façade, will assemble bits of prerecorded dialogue and preauthored plot points; the drama manager, as if stringing beads, will sequence the bits as it monitors the action. In the end, the game can be no bigger than its supply of prefabricated dramatic possibilities. The door to a world of truly open-ended drama will unlock only when a computer learns to write its own dialogue and plot twists, using rules that teach it to emulate a human playwright or screenwriter.
I raised an eyebrow. Can it be done? A simple prototype, Mateas said, is “totally doable within twenty years.”
“We have every intention of doing those projects,” Stern added.