What do you call a game that’s good for you, good for its genre, good for the gaming industry, and good for gaming technology, but isn’t a good game? Easy: Façade.

The new “one-act interactive drama” by Michael Mateas and Andrew Stern is brilliant in every conceivable way except as something to play. It feels churlish to criticize a work of such commendable ambition and scope for faults as mundane as being frustrating to operate and only marginally more entertaining than an average session of Westfront PC. But achievement does not exist in a vacuum, and it’s Façade’s technical strengths that that prove most responsible for the faults that make this potentially landmark title into little more than an unwieldy footnote of our current computer gaming climate.

Trip and you "enjoying" a drink

Mateas and Stern have in effect created a developmental prototype for the most realistically interactive computer game in history. Theirs is a world very much like ours, in which every word, action, and even hesitation has myriad unforeseeable consequences that affect all that follows. This means that, though specifics may vary from session to session (thus making a traditional walkthrough useless), the overall shape of the experience is always similar; the player, then, cannot simply go through the motions, but must always work to effect a resolution of the central conflict.

That conflict concerns Grace and Trip, a married couple in meltdown mode. They’ve invited the player’s character, an old friend, to their posh apartment, but can barely stop arguing long enough to open the door. Though they initially put on a brave show of connubial bliss, it’s not long before each injects bickering and insinuations into the conversation; it soon becomes clear that Grace and Trip have been lying to themselves and to each other, and that without the help of the clear-eyed (and ostensibly objective) player, one or both will soon choose to sever their relationship entirely.

The player’s toolkit is limited entirely to conversation. He or she may minimally manipulate objects or hug or comfort Grace or Trip, but there’s no real action to speak of. In all you say, you must be cognizant of not only how your words affect the Grace-Trip pairing, but also the emotional well-being of the individual character you’re speaking to or not speaking to. It’s a brilliant idea, one of the most detailed and considered representations of social intercourse yet seen for the personal computer. And as a narrative concept, it’s not inherently unworkable.

But programming a simulation of this sort about characters of this sort is not the same thing as telling a story about them. That requires a continuous sense of directed, forward motion, as well as writing that can fulfill the promise of the setup, with deep characterizations and subtext being requirements instead of mere good ideas. It’s in this area that Mateas and Stern most crucially falter: In trying to anticipate every possible dialogue variant, deviation, and combination, they’ve neglected to make the story's foundational conversation compelling and colorful enough to sustain interest from beginning to end, across multiple playings.

Instead, there’s a series of lightly connected scenarios. Your responses to them determine Grace's and Trip's attitudes toward you and each other. Trip offers to make drinks, and Grace doesn’t want them—how do you react, and how will Trip interpret your answer? Grace asks you about an abstract painting hanging over the sofa that typifies her interior decorating style, but is saying you like it the right thing or the wrong thing? What irreparable damage will be inflicted on their union merely by your examining and commenting on a statuette of a bull to be found on an armoire's shelf?

In terms of style of interaction, I am instantly reminded of Emily Short’s Galatea (2000), which attempted a similarly complex interweaving of emotional states and conversational topics, but succeeded to a considerably greater degree. Short’s story, about an art critic's investigation of a statue that’s inexplicably come to life, could branch off in any number of directions, depending on the predominant tone of the conversation between the statue and the player over a lengthy period of time: discussions of art, life, love, or countless other topics could result in some two dozen endings, some suggesting a more successful exploration of the story than others, but all of which, without exception, satisfied.

There’s only one satisfying (and successful) ending here: Grace and Trip come to terms with their faults, take responsibility for their actions, and stay together. But unlike in Galatea, in which contextual clues surreptitiously build on top of each other to gently guide you along the way, one can only reach Façade’s sole “correct” ending through exhaustive, exasperating trial and error. If you miss a clue—something that can happen all too easily—and get herded into a section of the story you’re not prepared for, you might find yourself stuck in what seems like an infinite loop: “No, let’s talk about us, about our relationship,” Grace might repeat three, four, or five times, offering little additional clue as to what she wants from you, let alone from Trip. Façade quickly becomes about learning exactly what each wants to hear, saying it within the specified time frame (usually only a few seconds), and then praying that it doesn’t upset the precarious emotional balance you've cultivated with the other character.

Trying to elicit information from Grace

This isn’t the behavior of either a game or an interactive story—it’s the behavior of a children’s toy, a guessing game with surprise time limits that must work within a set of tight parameters in order to operate at all. Once this realization sinks in, as it did for me after a dozen or so playings, Façade rapidly loses what little entertainment value it had in the first place. Establishing and maintaining relationships is hard enough in real life, and a computer game that replicates that process—without the levity, whimsy, or humor so vital in the real world—is something that can never be “fun” in the traditional sense.

That’s not to say that fun should have been the authors’ goal. They wanted to challenge and further our perceptions of interaction, and they’ve done so in a way that suggests we might be approaching a new crossroads in computer gaming history. But without a spoonful of sugar to help their medicine go down, Mateas and Stern sabotage their own work and their own message. As impressive a simulation as Façade is, it would impress even more if it offered entertainment to match its innovation. Edward Albee's 1962 play Who’s Afraid of Virginia Woolf? is rightly considered a landmark in American theatre for its frank, often violent depiction of an ever-sparring married couple, but you’ll find fewer hard-hitting dramas as screamingly funny.

Interestingly, Mateas and Stern even acknowledge some theatrical roots for Façade, providing a red-drapery-and-voiceover opening, a slow fadeout at the end, and an option to generate all the dialogue from any completed session in script form. Façade, like too many plays produced these days, is functional but little more, an outline for drama that never actually occurs. One can’t help but feel, though, that when the curtain comes down on interactive fiction, it will be the doors Façade opened that will be seen as more important and worthwhile than anything about the game itself.