by Tom Bissell
The panel moderator, Chris Kohler, from Wired magazine, introduced himself next. His goal was to walk the audience through the evolution of the video-game character, from the australopithecine attempts (Pong’s roving rectangle, Tank’s tank) to the always-interesting Pac-Man, who, in Kohler’s words, was “an abstraction between a human and symbol.” Pac-Man, Kohler explained, “had a life. He had a wife. He had children.” Pac-Man’s titular Namco game also boasted some of the medium’s first cut scenes, which by the time of the game’s sequel, Ms. Pac-Man, had become more elaborate by inches, showing, among other things, how Mr. and Ms. Pac-Man met. “It was not a narrative,” Kohler pointed out, “but it was giving life to these characters.” Then came Nintendo’s Donkey Kong. While there was no character development to speak of in Donkey Kong (“It’s not Mario’s journey of personal discovery”), it became a prototype of the modern video-game narrative. In short, someone wanted something, he would go through a lot to get it, and his attempts would take place within chapters or levels. By taking that conceit and bottlenecking it with the complications of “story,” the modern video-game narrative was born.
How exactly this happened, in Kohler’s admitted simplification, concerns the split between Japanese and American gaming in the 1980s. American gaming went to the personal computer, while Japanese gaming retreated largely to the console. Suddenly there were all sorts of games: platformers, flight simulators, text-based adventures, role-playing games. The last two were supreme early examples of games that, as Kohler put it, have “human drama in which a character goes through experiences and comes out different in the end.” The Japanese made story a focus in their growingly elaborate RPGs by expanding the length and moment of the in-game cut scene. American games used story more literarily, particularly in what became known as “point-and-click” games, such as Sierra Entertainment’s King’s Quest and Leisure Suit Larry, which are “played” by moving the cursor to various points around the screen and clicking to the result of story-furthering text. These were separate attempts to provide games with a narrative foundation, and because narratives do not work without characters, a hitherto incidental focus of the video game gradually became a primary focus. With Square’s RPG–cum–soap opera Final Fantasy VII in 1997, the American and Japanese styles began to converge. A smash in both countries, Final Fantasy VII awoke American gaming to the possibilities of narrative dynamism and the importance of relatively developed characters—no small inspiration to take from a series whose beautifully androgynous male characters often appear to be some kind of heterosexual stress test.
With that, Kohler introduced the panel’s “creative visionaries”: Henry LaBounta, the director of art for Electronic Arts; Michael Boon, the lead artist of Infinity Ward, creators of the Call of Duty games; Patrick Murphy, lead character artist for Sony Computer Entertainment, creators of the God of War series; and Steve Preeg, an artist at Digital Domain, a Hollywood computer animation studio. The game industry is still popularly imagined as a People’s Republic of Nerds, but these men were visual representations of its diversity. LaBounta could have been (and probably was) a suburban dad. The T-shirted Boon could have passed as the bassist for Fall Out Boy. Murphy had the horn-rimmed, ineradicably disgruntled presence of a graduate student in comparative literature. As for the interloping Preeg, he would look more incandescent four nights later while accepting an Academy Award for his work on the reverse-aging drama The Curious Case of Benjamin Button.
LaBounta immediately admitted that “realistic humans” are “one of the most difficult things” for game designers to create. “A real challenge,” he said, “is hair.” Aside from convincing coifs, two things video-game characters generally need are what he called “model fidelity” (do they resemble real people?) and “motion fidelity” (do they move like real people?). Neither, he said, necessarily corresponded to straight realism. Sesame Street’s Bert and Ernie, for example, had relatively poor model fidelity but highly convincing motion fidelity. As for the Uncanny Valley Problem, LaBounta said, “just adding polygons makes it worse….The Holy Grail in video games is having a character move like an actual actor would move. We’re not quite there yet.” Getting there would be a matter of “putting a brain in the character of some intelligence.” I was about to stand up and applaud—until he went on. One thing that routinely frustrated him, LaBounta said, was when a video-game character walks into a wall and persists, stupidly, in walking. Allowing the character to react to the wall would be the result of a “recognition mechanic,” whereby the character is able to sense his surroundings with no input from the player. Of course, this would not be intelligence but awareness. The overall lack of video-game character awareness does lead to some singularly odd moments, such as when your character stands unfazed before the flaming remains of the jeep into which he has just launched a grenade. What that has to do with character, I was not sure. If “personality is an unbroken series of successful gestures,” as Nick Carraway says in The Great Gatsby, the whole question of believable characters may be beyond the capacity of what most video games can or ever will be able to do—just as it was for James Gatz.
In Boon’s view, when talking about believable characters, one had to specify the term. Were you talking about the character the gamer controls or other, nonplayable characters within the game? A great example of the former, Boon believed, was the bearded and bespectacled Gordon Freeman from Valve’s first-person shooter Half-Life. Part of the game’s genius, Boon said, is how Gordon is perceived. In the game’s opening chapters “everyone treats you as unreliable, and you feel unreliable yourself. By the end, people treat you differently, and you feel different.” Gordon’s journey thus becomes your own. (Also, throughout the game, Gordon does not say a word.) As for believable nonplayable characters, Boon brought up two of the most memorable: Andrew Ryan from BioShock and GLaDOS from Valve’s Portal. Both games are shooters, or neoshooters, in that BioShock has certain RPG elements and the “gun” one fires in Portal is not actually a weapon; both characters are villains. While the villains in most shooters exist only to serve as bullet magnets, Ryan (a sinister utopian dreamer) and GLaDOS (an evil computer) are of a different magnitude of invention. The gamer is denied the catharsis of shooting either; both characters, in fact, though in different ways, destroy themselves. For the vast majority of both games, Ryan is present only as a presiding force and GLaDOS only as a voice. These are characters that essentially control the world through which the gamer moves while raining down taunts upon him. In GLaDOS’s case, this is done with no small amount of wit. In her affectless, robotic voice, GLaDOS attempts, whenever possible, to destroy the gamer’s self-esteem and subvert all hope of survival. “GLaDOS is so entertaining,” Boon said, “I enjoy spending time with her—but I also want to kill her.” The death of Andrew Ryan, on the other hand, is one of the most shocking, unsettling moments in video-game history. It has such weird, dramatic richness not because of how well Andrew Ryan’s hair has been rendered (not very) but because of what he is saying while he dies, which manages to take the game’s themes of control and manipulation and throw them back into the gamer’s face. These two characters have something else in common, which Boon did not mention: They are written well. They are funny, strange, cruel, and alive. It is also surely significant that the controlled characters in BioShock and Portal are both nameless ciphers of whom almost nothing is learned. They are, instead, means of exploration.
Patrick Murphy jumped in here to say, “It’s not whether the character is realistic or stylized; it’s that he’s authentic.” In illustration he brought up Kratos of his company’s own God of War series. Kratos is a former Spartan captain who, after being slain in combat by Ares, manages to escape Hades and declare war on the gods. Among the most amoral and brutal video-game protagonists of all time, Kratos, in Murphy’s words, “doesn’t just stab someone; he tears him in half. That helps sell him. Veins bulge out when he grabs things. It gives him an animal feeling that’s really necessa
ry.” The narrative of the God of War games is set on what game designers refer to as “rails,” meaning that Kratos’s story is fixed and the narrative world is closed. The gamer fights through various levels, with occasional bursts of delivered narrative to indicate that the story has been furthered. It probably goes without saying that no one plays the God of War games to marvel at the subtlety of their storytelling, which is pitched no higher than that of a fantasy film. It is a game that one plays to feel oneself absorbed into a malignant cell of virtual savagery. Kratos’s believability is served by the design and effect of the gameplay rather than the story. In short, he has to look great, which provides a fizzy sort of believability. If Kratos does not look great in purely creaturely ways, the negligible story will be dumped into the emotional equivalent of a dead-letter office. This is one of the most suspect things about the game form: A game with an involving story and poor gameplay cannot be considered a successful game, whereas a game with superb gameplay and a laughable story can see its spine bend from the weight of many accolades—and those who praise the latter game will not be wrong.
Steve Preeg, by now wearing a slightly worried expression, opened by admitting that he was not a gamer and professed to know very little about games. But he knew a bit about believability and character. To explain the difficulty he had with animation in The Curious Case of Benjamin Button, he showed us “draft” shots from the digital process by which Brad Pitt’s character was aged. In the earliest attempts Pitt looked undead—utterly terrifying. Just shifting the width of his eyes a tiny bit, Preeg said, made the difference between “psycho killer” and “a little boy who just got home.” He showed us how he did this, and the difference was indeed apparent. Preeg then turned philosophical. In Hollywood, he said, “we have very clear goals.” He worked under a director, for instance, had a clear idea of the script, and knew whether sad or happy music would be playing under the scenes he was required to digitally augment. Every eye-widening and face-aging task he was given as an animator had a compelling dramatic context attached to it, which he used to guide his animation decisions. His art was always guided. “Your characters,” he said, turning to the panel, “have to be compelling in very different ways, depending on what the audience wants to do.” Preeg was silent for a moment. Then he said, “You guys are going to have a very, very difficult time.”
After the panel, I sought out the man who, during its EEDAR portion, turned to his colleague and said, “This can’t be a good thing.” His name was John Hight, and he was the director of product development for Sony Computer Entertainment, Santa Monica Studios. One of the projects he was currently overseeing was God of War III, a game whose budget was in the tens of millions of dollars. Yet he was no pontiff of the AAA title. Hight had also greenlit and helped fund thatgamecompany’s downloadable PlayStation 3 title Flower, a beautiful and innovative game—a stoner classic, really—in which the player assumes control of a windblown petal and floats around, touching other flowers and gathering their petals and eventually growing into a peaceful whirling versicolor maelstrom. (When faced with releasing the tranquilizingly mellow Flower, no one at Sony could think of an apt category under which to market it. Hight called it a “Zen” game, and that was how it was shipped. Only later did anyone realize that the category was Hight’s invention.) Hight, who was in his forties, had worked on dozens of games over the course of his career, from RPGs to shooters to flight simulators to action-adventure games. When I asked him why he had scoffed during the EEDAR presentation, he said, “The scary thing is that someone is going to enlist that data to find the ideal game that hits all the proper points, and they’re going to convince themselves—and a lot of bean counters—that this is a surefire way to make money.” When I asked if he could imagine any circumstances in which such data would be useful, he said, “I’m sure that our marketing people will at some point be interested, and if it helps them have the courage of my convictions, that’s okay.”
Hight’s first title was a video-game version of the old Milton Bradley tabletop classic Battleship for the Philips CDI (a doomed early attempt at an all-purpose home-entertainment center that was launched in 1991 and discontinued seven years later). As the game’s producer, coder, animator, and writer, Hight had no legal claim to develop Battleship when he began his work on it. When the time came, he simply crashed a toy expo, walked up to the Milton Bradley booth, and, after a short demonstration, strolled away with the rights. The entire game cost him $50,000. (“Very exciting,” he said.) He had also been around long enough to remember an argument he had in 1994 with a colleague about whether this new genre known as the shooter was “here to stay.” Hight told me, “Doom II had come out and done pretty well, but there really weren’t many companies doing first-person shooters. I think it was seen as a novelty.” Hight’s colleague had insisted to him, “What more can you do? You’re sort of just pointing a gun and shooting it.” Hight was able to convince the man otherwise and proceed with his shooter. It became Studio 3DO’s Killing Time, an innovative shooter for its day in terms of its gameplay (it was among the first shooters that allowed the player to crouch), setting (the 1930s), and relatively knowledgeable employment of an outside mythology (namely, Egyptology).
“When I first got into the industry,” Hight said, “there were a lot of really hardcore gamers, and we were basically making games for us. We weren’t making games for an audience. It was for us. And we got so specialized and so stuck in our thinking.” These men’s minds were typically scattered with the detritus of Tolkien, Star Wars, Dungeons & Dragons, Dune—and that was if they had any taste. Many of the first relatively developed video-game narratives were like something dreamed up by an imaginative child (a portal to Hell…on Mars! Hitler…as a cyborg!), with additions by an adult of more malign preoccupations. The writing in such games was an afterthought. For Killing Time, Hight told me, he “was literally writing the dialogue the day” of the recording session with the actors, one of whom approached Hight after the session was over and asked, “Do you guys ever just write a script and give it to the actors ahead of time?” A decade and a half later, Hight was still abashedly shaking his head. “Back in the day, most designers insisted writers really couldn’t understand how to develop good, interactive fiction. So there was this designer–writer divide that the game industry sort of started out with.”
When I asked how it could be that a panel putatively devoted to believable characters did not manage to discuss writing even once, Hight gently averred: “That panel was mostly composed of technical artists. The roots of games are in technical people. My background is computer science. I was a programmer for ten years. That’s kind of how we approached it. How can we make this thing run faster? How much more can we put into the game? How can we make the characters look better?”
With its origins in the low-ceilinged monasteries of computer programming, video-game design is, in many ways, an inherently conservative medium. The first game designers had to work with a medium whose limits were preset and virtually ineradicable. There were innumerable things games simply could not do. In this sense, it is little wonder that the people who were first drawn to computers were also drawn to science-fiction and fantasy literature. As Benjamin Nugent notes in his cultural history American Nerd, sci-fi and fantasy literature is almost always focused “on the mechanics of the situation. A large part of the fun of reading a sci-fi series is about inputting a particular set of variables (dragon-on-dragon without magic) into a model (the Napoleonic Wars) and seeing what output you get.” A video game is first and foremost a piece of software (which is why many magazines do not dignify games with italicized titles). The video-game critic Chris Dahlen, who by his own admission comes out of both a software and an “artsy-fartsy” background, argues that games “don’t pose arguments, they present systems with which to interact.” In this view, games are not and cannot be stories or narratives. Rather, some games choose to enable the narrative content of their system while others do not. A d
isproportionate number of game designers at work today come out of a systems, programming, or engineering background, which has in turn helped to shape their personalities and interests. One result of this is that it forces designers to imagine games from the outside in: What variables do I inject into the system to create an interesting effect?
For any artist who does not sail beneath the Jolly Roger of genre, this is an alien way to work. As someone who attempts to write what is politely known as literary fiction, I am confident in this assertion. For me, stories break the surface in the form of image or character or situation. I start with the variables, not the system. This is intended neither to ennoble my way of working nor denigrate that of the game designer; it is to acknowledge the very different formal constraints game designers have to struggle with. While I may wonder if a certain story idea will “work,” this would be a differently approached and much, much less subjective question if I were a game designer. A game that does not work will, literally, not function. (There is, it should be said, another side to the game-designer mind-set: No matter how famous or well known, most designers are happy to talk about how their games failed in certain areas, and they will even explain why. Not once in my life have I encountered a writer with a blood-alcohol content below .2 willing to make a similar admission.)