War Play Read online

  The next major step in the military’s video game history came with the 1993 release of the blockbuster first-person shooter fantasy Doom. According to Timothy Lenoir and Henry Lowood, historians of science, Doom is solely responsible for changing practically every facet of PC-based gaming, including “graphics and networking technology, . . . styles of play, notions of authorship, and public scrutiny of game content.” (One of the game’s innovations was a new mode of play called “death match,” which, like Doom’s other innovations, is now a standard feature of many first-person shooter games.) Doom was an immediate sensation among gamers, with sales soon climbing into the millions.

  Around the same time that Doom was released, the Marine Corps Modeling and Simulation Office (MCMSO) received a mandate from the annual General Officers Symposium to begin looking for commercial video games that might prove useful for training. Because the Marine Corps budget is a great deal smaller than that of the other services, the corps has a long history of seeking cost-effective training solutions. General Charles Krulak, its commandant at the time, believed that PC-based war games held great potential for teaching Marines critical decision-making skills.

  Lieutenants Scott Barnett and Dan Snyder of MCMSO immediately began combing through dozens of military-related video games to see if any might be useful for training. They developed the online Personal Computer Based Wargames Catalog, on which they posted detailed reviews of the numerous games they investigated. Barnett and Snyder were looking for a fast-action first-person shooter—one that, crucially, allowed user modification. Of the many games they examined, only Doom (or rather its sequel, Doom II) fit the bill. As part of its marketing strategy, Doom’s developer, id Software, had released parts of the game as shareware and encouraged players to enact their own modifications.

  Throughout the spring and summer of 1995, Snyder transformed the game from an outer-space gothic fantasy into a military fire-team simulation. The Martian terrain and alien demons of the original Doom were replaced by a dun-colored landscape of pockmarked concrete bunkers and enemies who had been drawn from scans of GI Joe action figures. The cost of production? A mere $49.95—the price of one copy of Doom II.

  The point of the modified game, known as Marine Doom, was to teach Marines not how to fire their weapons but how to work together in teams and make split-second decisions in the midst of combat. “A real firefight is not a good time to explore new ideas,” Snyder explains. The game had another, equally significant, rationale. “Kids who join the Marines today grew up with TV, videogames, and computers,” Barnett reasons. “So we thought, how can we educate them, how can we engage them and make them want to learn?” Barnett and Snyder’s calculations were correct: their creation became a huge hit among Marines, though, like Army Battlezone, it was never actually used for training. According to Barnett, Marines would plead to be allowed into his base’s gaming lab even after it closed at night.

  Marine Doom was created at a time when the Pentagon had begun embracing simulation for a broad range of activities. As scholar Sharon Ghamari-Tabrizi relates, these activities included “part-task training; mission rehearsal; operational planning; strategic and tactical analyses; weapons systems modeling during research and development, testing and evaluation, and acquisitions; and long-range future studies.” Much of this emphasis on simulation was the result of post–Cold War military downsizing. With the collapse of the Soviet Union, the military’s budget had been reduced to a level commensurate with what Congress assumed was a greatly reduced geopolitical threat. The relative affordability of simulation technologies matched well with the military’s newly tightened budget.

  The Federal Acquisition Streamlining Act of 1994 also forced the military to change its procurement policies. No longer could it underwrite defense contractors’ R&D and acquisitions on an unlimited basis; instead, the Pentagon had to rely on what are known as “commercial off-the-shelf” (COTS) technologies—technologies that already exist and that have been developed by commercial industry. Take SIMNET: put together by military contractors, it required $140 million, ten years, and several hundred employees to build, even though it did use some commercial technologies. By contrast, Marine Doom, which relied exclusively on commercial technologies, was built by eight people in six months for $25,000. Military contractors now had to take on the stripped-down and flexible management practices of corporations—in effect, to become commercial businesses themselves. This had a deep and immediate impact on the defense sector, resulting in the merger or closure of a number of prominent companies.

  In order to maintain their livelihoods, defense contractors had to find other customers to whom they could peddle their high-tech gadgets. Yet even in this time of seeming crisis, the contractors ended up coming out ahead, as it quickly became apparent that another industry was hungry for their wares: the entertainment industry. The relationship born of this outcome was symbiotic: defense contractors would spin their technologies off into the commercial game industry, and the commercial game industry would spin its technologies right back. In an update of Eisenhower’s classic formulation, cyberpunk writer Bruce Sterling termed this win-win relationship the “military-entertainment complex”—the relentless exchange of technologies, personnel, and money that defines the bond between the military and the video game industry.

  The military’s reduced budgets in the 1990s also led to a greater dependence on reservist troops, which only increased the use of distributed interactive simulation systems for training. These systems enabled reservists to participate in large-scale training exercises and maneuvers no matter where they were based. As a further cost-cutting method, the immediate post–Cold War period saw the military’s emphasis shift to all-encompassing joint operations, as opposed to individual service missions. Two new declarations of military doctrine—Joint Vision 2010 (1996) and Joint Vision 2020 (2000)—codified this focus. In the effort to develop new simulation platforms that would meet the requirement for a jointly linked system, the four services of the armed forces were directed to overcome their traditional rivalries. The ultimate result of this attempt at cooperation was the Joint Simulation System(JSIMS), a single, integrated virtual battlefield—in technical terms, a mission rehearsal and command simulation environment—in which participants from all four services could operate regardless of location.

  There was one more reason for the military’s turn to simulation: modern high-tech warfare was increasingly fought through electronic and digital interfaces resembling video games. Early on, this rapid growth in the electronic mediation of warfare caused confusion even among military professionals. An oft-repeated anecdote involves the war game Operation Internal Look, undertaken by the U.S. military in July 1990, during the run-up to the first Gulf War. General Norman Schwarzkopf relates the tale in his memoirs: “As [Internal Look] got under way, the movements of Iraq’s real-world ground and air forces eerily paralleled the imaginary scenario of the game . . . As the war game began, the message center also passed along routine intelligence bulletins about the real Middle East. Those concerning Iraq were so similar to the game dispatches that the message center ended up having to stamp the fictional reports with a prominent disclaimer: ‘Exercise Only.’”

  Linking Entertainment and Defense

  In the 1990s, no less an entertainment icon than Mickey Mouse presided over the tightening of the military–video game industry bond. At a mid-1990s meeting of the Army Science Board, that service’s senior scientific advisory body, four-star general Paul Kern met Bran Ferren, an entertainment industry futurist with a friendly, expansive manner and a wild red beard. Ferren was the influential head of creative technology at Walt Disney Imagineering, the design and development arm of the Walt Disney Company based in Glendale, California. (Since its founding in 1952, Walt Disney Imagineering has developed dozens of innovations in the areas of special effects, interactive entertainment, fiber optics, robotics, and film techniques.)

  General Kern’s first thought upon meet
ing Ferren, with his tan explorer’s jacket and his untamed facial hair, was, “What’s this crazy liberal doing here in the middle of our organization?” As soon as he heard Ferren speak, however, Kern found him to be an inspiring, intellectually challenging figure who crystallized many of the nascent doubts Kern had been harboring about the static state of military simulation. Listening to Ferren describe the cutting-edge virtual-reality development efforts he was leading at Walt Disney Imagineering, Kern came to the abrupt realization that the entertainment industry had leaped far ahead of the military in regard to high technology—and, equally important, in the cost of that technology.

  Ferren pointed out to the assembled army officials that when he and his entertainment industry associates looked at the military’s modeling and simulation offerings, the offerings were, frankly, “unaffordable and pretty crappy.” The software was “lousy” and the hardware was “clunky and inflexible.” Ferren posed a series of questions to prod the army officials’ thinking. “How much texture memory can we have in the graphics process unit?” he asked as an example. He received a number of blank stares. “Texture memory?” someone responded. “What’s that?”

  Kern’s education was in mechanical engineering, and over the previous two decades he had gained a great deal of experience with computing and simulation, including the SIMNET program. He found Ferren such a font of valuable information and advice that he began meeting with him regularly, sometimes at Disney Imagineering headquarters, other times in his own office in the Pentagon. Their conversations were technical and wide-ranging, but Ferren took pains to drive home to Kern a simple message: “You gotta be where the action is.” If the military wanted to be part of the emerging technology base in Hollywood, which was linked to Silicon Valley, then it had to establish a presence there.

  Kern was so impressed by his meetings with Ferren that he charged his subordinates with making the military more Disney-like. The military had been at the forefront of technological development for decades, he told them. Why couldn’t it now develop its capabilities to match those of the entertainment industry?

  As it turns out, he wasn’t the only one asking that question. In 1996, professor Michael Zyda of the Naval Postgraduate School in Monterey, California, had chaired a National Research Council study titled “Modeling and Simulation: Linking Entertainment and Defense.” As tightly connected as the military and the game industry had always been, Zyda’s report argued that the two entities nonetheless had a great deal more to offer each other. When General Kern sent his subordinates scrambling to find someone who could better meld the entertainment industry’s technological know-how with the military’s training needs, Zyda’s name was at the top of the list.

  An intriguing combination of laid-back Southern California surfer and overcaffeinated Silicon Valley entrepreneur, Zyda frequently uses words like “awesome” and “totally,” yet those words are delivered with rapid-fire intensity and a buzzing physical energy. Compact and sturdy, with a gray mustache and thinning gray hair, Zyda is one of the people most responsible for the partnership between the military and the entertainment industry. While that link is hardly new—think of the numerous propaganda films produced by Hollywood during World War II—the mid-1990s witnessed the start of an unprecedented level of collaboration between the two groups. More than any other single person, Mike Zyda played a seminal role in this process.

  Zyda’s interest in computers dates back to his undergraduate years at the University of California at San Diego, where, as a freshman math major in 1973, he got a job working in the lab of a physical chemistry professor named Kent Wilson. The job interview was brief, with Wilson asking Zyda only whether he was willing to learn three things: computer graphics, programming, and how to write grant proposals.

  Wilson’s lab offered the kind of freewheeling intellectual and creative environment that marked the high-tech world at the time. Zyda worked alongside seventeen other undergraduates, experimenting with computers, lasers, and chemicals. Among his coworkers were Bud Tribble and Bill Atkinson, both now legendary in the annals of computing. Atkinson, the eleventh employee of Apple Computers, is the creator of MacPaint, QuickDraw, and HyperCard, while Tribble managed the original Macintosh software development team and helped design the Mac OS and user interface.

  The “life-changing” experience of working for Wilson led Zyda to pursue a master’s degree in computer information sciences at the University of Massachusetts at Amherst, where his adviser was Victor Lesser, a major figure in the field of simulation. After receiving his doctorate from Washington University in St. Louis, Zyda entered a job market that was remarkably ripe for new PhDs in computer science.

  At the time of Zyda’s graduation, many universities around the country were just beginning to create computer science programs, but the lack of graduates with relevant experience meant that there was a dearth of qualified faculty. Without even sending out applications, Zyda was recruited by the Naval Postgraduate School (NPS) in Monterey, California.

  Zyda’s focus when he arrived at the school, in February 1984, was real-time graphics. The army’s SIMNET program had started the year before, and there was growing momentum in the military for visual simulators, which at the time cost between $10 million and $30 million each. In 1988 the army tasked Zyda with building a visual simulation system for the fiber-optic-guided missile. The FOG-M was an early version of a drone; it had a TV camera in the front and a thirty-kilometer fiber-optic cable spilling out the back. A soldier watching a video screen would guide the missile with a joystick and crash it into the intended target. Rather than relying on technology built by defense contractors, Zyda and his students built their simulation system to run on a $60,000 Silicon Graphics machine. They finished in six short weeks. When they presented their system to the army personnel at nearby Fort Hunter-Liggett, the response was immediate. “We’re cutting you a check for $100,000,” they told Zyda. “We want to take this system out into the field starting today.”

  Zyda and his students next built a simulator for the vehicle on which the FOG-M was mounted, after which they needed to network the two systems together. Luckily, Zyda had learned networking while on a three-week consulting trip in 1987 to Tokyo, where he had built a piece of code that would allow any number of workstations to be connected. He and his students now used that software to build a networked virtual environment that they called the “NPS Moving Platform Simulator.” Soon after, Zyda received a phone call from a man named George Lukes at the U.S. Army Topographic Engineering Center.

  “I just read a paper you wrote on your Moving Platform Simulator system,” Lukes told Zyda. “It looks like SIMNET. Can I come to Monterey and talk with you?”

  Zyda had never heard of SIMNET, because DARPA hadn’t written any papers or given any talks on it. “What’s SIMNET?” he asked, confused.

  After sending Zyda a paper describing SIMNET, Lukes went out to NPS for a demonstration of Moving Platform. After the demonstration, Lukes took Zyda aside and told him how impressive it was. He then offered a proposal.

  “Listen,” Lukes said, “the army is just about to take ownership of SIMNET from the defense contractors, but no one in the military knows how to read and write the networking packets. There’s also no one who knows how to read the terrain databases that the contractors have created. Will you do it?”

  This was just the kind of challenge that Zyda and the students in his graphics class enjoyed. Using money provided by Lukes, they taught themselves how to read SIMNET’s packets and databases. The source code they created led them to build the Naval Postgraduate School Net, or NPSNET, a collection of Silicon Graphics workstations attached to a local-area Ethernet. NPSNET was in essence a SIMNET-connected simulator that enabled officers both to observe and to participate in the virtual training of their soldiers.

  Zyda and his students’ efforts quickly attracted the attention of numerous Department of Defense offices, all of which were interested in the training possibilities of v
irtual technology. In 1995, Zyda was asked to take part in a National Research Council (NRC) study called “Virtual Reality: Scientific and Technological Challenges,” which advised the government on the kinds of virtual-reality research it should invest in. Though he was a relatively lowly member of the team, Zyda ended up writing about one-third of the final report.

  Because of these efforts, the next year Zyda received another call from the National Research Council. The NRC had just received funding from Anita Jones, the Pentagon’s director of defense research and engineering—who was responsible for overseeing the department’s science and technology program, research laboratories, and DARPA—to put together a conference and a report on areas of potential collaboration between the defense and entertainment industries. Would Zyda be willing to chair the committee?