Blood, Sweat, and Pixels

by Jason Schreier

  One of the biggest fights centered on what Devine called the “rule of eight,” a guideline that the Halo Wars team had been following since back in the Phoenix days. What that meant was that the player could select only eight units at once, to ensure that the gameplay didn’t feel too clunky and complicated on a controller. Some of the designers began to question this rule, arguing that it was making Halo Wars too simple. Devine pushed back. “When you playtest every single day, you get very good at the controls,” Devine said. “And you forget that you had to learn them at one point. And so all of a sudden you think, well ten of these would be perfectly easy. Sixteen, even.”

  It’s a common dilemma in game development: when you’re working on the same game for years, you’ll inevitably get bored. It’s tempting to make changes just for the sake of making changes, to spice things up because you’re sick of using the same simple control scheme when you get into work every day. For Halo Wars, that became a constant point of tension. Devine and his team had intentionally made this game simpler than its RTS counterparts on the PC because the console controls would be so unfamiliar to people. And now some people were trying to make it more complicated? “We were trying to avoid layers of thinking,” said Chris Rippy. “One of the biggest problems with long game development is, when you playtest the game for too long, you invent problems and you add layers to the game that don’t need to be added.”

  Although they had impressed fans with that Devine-narrated demo at E3 2007, the Halo Wars team was in a sad state. They’d scripted the demo entirely by hand, with code that wouldn’t work in the final game. Sure, the graphics were running in real time, but the artificial intelligence wasn’t properly hooked up, and playing an actual match of Halo Wars wasn’t nearly as satisfying as the demo made it look.

  In Dave Pottinger’s eyes, Halo Wars needed to go in a totally new direction. “We didn’t have a game that was fun,” said Pottinger. “It didn’t work. The E3 demo was a pretty kick-ass demo but not something that was representative of the gameplay. It was just entirely built for E3.” In the months that followed, the Halo Wars team continued to argue over the game’s basic features. Were the units on the screen too small? Why was base building so complicated? Should they have “control groups” that would allow players to assign hotkeys to different squads of units, as in StarCraft and other RTS games?

  After all the squabbling, something had to give. That something turned out to be Graeme Devine. “When lead designers are fighting, it’s not good,” Devine said. “So we had one meeting where I said, ‘Dave, you take the lead design position. I’m having a lot of fun on the story. I’m just going to stick to the story.’” Devine remained the face of the game, showing it off at Microsoft marketing meetings and press events. And Dave Pottinger, who for years had resisted working on another RTS game, became the lead designer of Halo Wars.

  Right away, Pottinger made big changes to the core mechanics of Halo Wars. “We basically threw the design out and started over,” Pottinger said. He altered almost all the units, including the leaping Warthogs they’d shown at E3, which weren’t quite working out. (Warthogs would remain in Halo Wars, but they would no longer have the ability to jump gaps.) He ditched the sprawling base-building system, replacing it with “prefabricated” bases that each held a finite number of slots. He overhauled the resource system, reducing the amount of time players would have to spend managing their economy. And he added a “select all” button, which the team had resisted for a long time. With Halo Wars scheduled for release in just a few months, Pottinger threw out nearly all of Devine’s design work. “I took over the design,” Pottinger said, “and Graeme kept doing the story stuff which was a full-time job, and he did a great job at that.”

  At the same time, Ensemble’s other team was also dealing with turmoil. For years now, a large chunk of the studio had been working on their dream project, the Halo MMO. The MMO wasn’t exactly a secret—Microsoft knew that Ensemble wanted to make its own version of World of Warcraft—but Microsoft still hadn’t given it the green light, either. To some on the Xbox team, it was a surprise to find out that Ensemble had dedicated so many staff to the Halo MMO when the studio should have ostensibly been focusing on Halo Wars. “It was very common for people [from Microsoft] to show up and go, ‘What the fuck is this, why are you working on that?’” said Colt McAnlis. “I remember there were some key meetings I was in where people were like, ‘We didn’t know you guys were putting resources on this. We thought these resources were over there.’” (Others dispute this, saying that Microsoft’s management was well aware of how many people Ensemble had on the MMO.)

  Soon enough, the ax came down. Microsoft made it clear to Ensemble’s leadership that it had no interest in spending tens of millions on an MMO, and the project was unceremoniously canceled. “We made trade-offs,” said Dave Pottinger. “Part of why we did Halo Wars was to give us a chance to do the MMO. . . . I think unfortunately the MMO was going to be a huge thing, and Microsoft decided that it wasn’t going to be something they wanted us to do.” Some members of the MMO team split off and worked on new prototypes for a while—including a potential Age of Empires 4—while others moved over to Halo Wars.

  All of this drama was undoubtedly raising eyebrows at Microsoft. Ever since the original Age of Empires in 1997, there had been bouts of tension between Ensemble and its publisher. As long as Ensemble was shipping games—and generating profits—those tensions would ultimately fizzle. But Ensemble hadn’t released a new game since Age of Empires III over two years earlier, in 2005. All this talk about an MMO had frustrated Microsoft’s executives, who were scaling back on PC gaming in favor of the red-hot Xbox 360. (The Halo MMO, unlike Halo Wars, was designed for PCs.) Keeping Ensemble’s hundred-plus staff on the payroll was starting to seem like a bad investment.

  “Contrary to popular belief, headcount’s not completely free at Microsoft,” said Shane Kim, who was a vice president in the Xbox division. “The fact of the matter was it was sort of zero-sum, a lot of headcount. They’re very expensive. And we decided we needed to deploy that in other areas.”

  Unlike most game developers, Colt McAnlis enjoyed getting to work at 6:00 a.m. His coworkers wouldn’t typically arrive for another four or five hours, and he felt most productive when he was coding by himself, with nobody around to distract him. McAnlis was responsible for some of the most complicated engineering work on Halo Wars: art tools, light shaders, multicore threading processes. Solitude helped.

  One Monday in September 2008, McAnlis arrived at the office at his usual time. It didn’t take long for him to realize that something was off. Instead of empty desks, he found dozens of his coworkers mulling around. “I [thought], ‘Wait, what the hell was going on?’” he said. “What are you all doing here?” Nobody would give him a straight answer until a few hours later, when a coworker mentioned that a bunch of Microsoft bigwigs were coming to visit.

  Word got around that everyone at Ensemble needed to go to the auditorium for an all-hands meeting. As they filed in, they started seeing scores of Microsoft employees: HR people, vice presidents, executives. When the whole company had gotten settled, Ensemble’s CEO, Tony Goodman, stood up front and dropped a nuclear bomb.

  “Tony gets onstage and says, ‘We have some news,’” said Graeme Devine. “‘Ensemble is going to be closing after Halo Wars.’”

  After fourteen years and a dozen games, Ensemble’s time was coming to an end. The good news, Goodman said, was that Microsoft wanted them to finish Halo Wars. Everyone at Ensemble would keep his or her job for another four months or so, which would give people a fair amount of time to look for new gigs. “Microsoft wanted to talk to us about keeping us all on through the end of Halo Wars,” said Devine, “and [Tony] hands it over to Microsoft to explain exactly how they were going to do that, because we’d all just gotten the news that we have no jobs after we ship this game.”

  For two hours, Microsoft’s representatives, including VP Shane Kim, stood
in front of the room and answered questions from hurt Ensemble employees. “They’re going, ‘Why us?’” said Kim. “‘You’ve got Rare, you’ve got Lionhead [two other game studios]. . . . Surely there must be other places you can cut instead of us.’ Unfortunately, it’s one of the darker sides of the job.” Kim tried to be diplomatic, adding that Microsoft would be giving solid severance packages to those Ensemble staff who stayed until Halo Wars was out the door.

  Some people started crying. Others got angry. “I flipped the fuck out,” said Colt McAnlis. “My wife and I had just gotten pregnant for the very first time. And our baby was due at the end of January, which was supposed to be the exact date the company was supposed to close. We had a newborn that was supposed to be delivered the day I was no longer supposed to have a job.” Two people—the brothers Paul and David Bettner—immediately quit Ensemble and started their own studio, Newtoy, which went on to make the popular Scrabble clone Words with Friends. One might say their move worked out nicely: in 2010, Zynga purchased Newtoy for $53.3 million.

  Rich Geldreich, the graphics engineer, also left the studio shortly after hearing about the shutdown. “My brain had cracked in two,” he said. “I went crazy. I just couldn’t handle the fact that Ensemble imploded. It was such an awesome company. I had five years invested in that company, and all that technology and coding and everything and it just was gone. And everybody was cracking around me. We were all going crazy. It was ridiculous.” (Soon afterward, Geldreich took a job at Valve.)

  Those who remained at Ensemble found themselves in another uncomfortable situation. It wasn’t that they’d be out of jobs in a few months—it was that some of them wouldn’t be out of jobs. Tony Goodman informed everyone that he’d negotiated a contract with Microsoft as part of the closure. With this deal, he told the team, he’d be able to start an independent new studio, which he was calling Robot Entertainment. The Microsoft contract would allow them to make an online version of Age of Empires as they tried to get their business going.

  The problem was, Goodman had secured only enough money to hire less than half of Ensemble’s current staff. And only a few people knew exactly who had made the cut. “People all of a sudden were jockeying for new jobs,” said Colt McAnlis. “It was madness for a while because none of us really knew what was going on. There was always this watercooler chat: ‘Hey did they talk to you? No they didn’t talk to you? Well, if they offered you a job would you accept it?’”

  Those who were invited to join Robot found out quickly, but to those who weren’t getting spots, the following days were miserable. At one point, to clear the air, a manager tacked a piece of paper to the wall with a full list of Ensemble employees who would be going to Robot. Dozens of people came crowding around to see who’d made the cut. “It was like that scene in all the movies when the teacher comes out,” said McAnlis. “‘Here’s who’s been accepted into the cheerleading squad,’ and everybody rushes to the board.”

  To many at Ensemble, this new layer of politics was making a bad situation even worse. “In our view we saved half the jobs,” said Dave Pottinger. “[To] whoever wasn’t offered a spot at Robot, we screwed half the company over. . . . I think we tried to do our best. We made mistakes.” Other than Rich Geldreich and the Bettner brothers, everyone remained at Ensemble, despite the chaos. (The severance packages helped.) And over the next few months, Ensemble tried to put aside politics—and the fact that their studio would be shutting down—to finish Halo Wars.

  “The motivation was to go out on as good a note as we could,” said Dave Pottinger. Ensemble’s staff wound up crunching for months, sleeping at their desks up until the very last minute to finish Halo Wars by the end of January, when they knew they’d all have to leave. “The thing I’m most proud of at the end was, when we announced the closure, the next day [three people left],” said Pottinger. “Every single other person stayed to finish the game.”

  “[We were] dead men walking,” said Graeme Devine. “It was pretty crazy.” In early 2009, Microsoft flew Devine around the world on a press tour. Traveling alongside the polarizing British designer Peter Molyneux, who was about to release the RPG Fable 2, Devine conducted interviews and showed off demos of Halo Wars. When journalists asked him about Ensemble’s closure, he dodged their questions. “I remember pieces of paper being put in front of me by Microsoft [telling me] ‘This is what you can say,’” Devine said.

  For those surreal final months, Ensemble’s staff tried to put their internal politics aside, coming together in hopes of releasing something they could all take pride in having made. Even though they all knew the studio was closing, they kept coming into work every day. They kept crunching. And they kept doing everything they could to make Halo Wars as good as possible. “There were great days like nothing had ever happened and nothing was happening, and then there were depressing days,” said Chris Rippy. “But it was a prideful group and everybody wanted to do a great job, and everybody knew this was kind of the studio’s legacy, and wanted to do something to represent the studio properly.”

  On February 26, 2009, Halo Wars came out. It was received fairly well, getting solid reviews from gaming sites like Eurogamer and IGN. When they look back, those who were on the Halo Wars team say they’re proud to have released even a half-decent game after what they went through. Microsoft even hired a new developer—Creative Assembly—to make a sequel, Halo Wars 2, which came out in February 2017.

  A few studios rose from the ashes of Ensemble. There was Robot Entertainment, which went on to make Age of Empires Online and then a popular tower-defense series called Orcs Must Die! A different faction of ex-Ensemble developers started a company called Bonfire, which was also purchased by Zynga, then shut down in 2013.* Those developers later founded yet another studio, Boss Fight Entertainment. Nearly a decade after Ensemble’s closure, many of its former employees are still working together. They were a family, after all. “We hated each other, but we loved each other too,” said Pottinger. “Ensemble lost very few people over its run compared to any other studio. . . . You don’t have a turnover rate of—it was less than four percent or something like that—you don’t have that if you don’t make people happy.”

  Back at E3 2008, just a few months before Ensemble closed its doors for good, the Halo Wars team put out a cinematic trailer that may have, in retrospect, been a subconscious cry for help. “Five years. Five long years,” the trailer’s narrator said, as a group of marines fought off the invading Covenant alien forces. “At first it was going well.”

  A group of Covenant forces descended from the sky, taking out marine after marine in what quickly became a bloody massacre. The narrator continued: “Then setback after setback, loss after loss, made what was going to be a quick and decisive win into five years of hell.”

  6

  Dragon Age: Inquisition

  The website Consumerist once held an annual poll for “Worst Company in America,” asking readers to select the country’s most reviled corporation through a March Madness–style elimination tournament. In 2008, as the US economy collapsed, the insurance salesmen at AIG took top honors. In 2011 the award went to BP, whose oil rigs had just spilled 210 million gallons of crude in the Gulf Coast. But in 2012 and 2013, a different type of company won the award, beating out the likes of Comcast and Bank of America, as over 250,000 voters flocked to declare that the United States’ worst company was in fact the video game publisher Electronic Arts (EA).

  There were many reasons for this ignominious victory, including the rise of optional “microtransaction”* payments in EA games and the spectacular disaster of the publisher’s online-only SimCity reboot.* What may have rankled gamers most, however, was what they believed EA had done to BioWare.

  BioWare, a development studio founded in 1995 by three medical doctors who thought making games might be a cool hobby, leaped into fame in 1998 with the Dungeons & Dragons–based role-playing game Baldur’s Gate (a game so influential that it would play a major role in the
stories of two other games in this book, Pillars of Eternity and The Witcher 3). In the following years, BioWare became renowned for a string of top-tier RPGs like Neverwinter Nights, Star Wars: Knights of the Old Republic, and a space opera called Mass Effect that appealed not just to people who liked to shoot aliens, but also to people who liked to smooch them.

  In 2007, Electronic Arts bought BioWare, and in recent years, it had seemed to fans that the studio was in a funk. BioWare’s two new flagship franchises, Mass Effect and the Tolkien-ish Dragon Age, were beloved but stalling. Dragon Age 2, released in 2011, was panned for feeling incomplete. And 2012’s Mass Effect 3, which put a bow on the spacefaring trilogy, angered hard-core fans with a controversial ending in which the player’s choices didn’t seem to matter.*

  Surely, fans thought, these missteps were EA’s fault. The problems started when EA purchased the studio, didn’t they? All you had to do was look at the long list of iconic studios that EA had acquired and then shut down—a list that included Bullfrog (Dungeon Keeper), Westwood (Command & Conquer), and Origin (Ultima)—to fear that BioWare might be next. Fans went to Consumerist to send a message, even if that meant declaring a video game publisher more heinous than a predatory mortgage lender.

  Amid these theatrics, BioWare’s studio head, Aaryn Flynn, was staring down a far more relevant challenge: Dragon Age: Inquisition, the third game in the fantasy series. Inquisition was going to be the most ambitious game that BioWare had ever made. It was a game with a lot to prove—that BioWare could return to form; that EA wasn’t crippling the studio; that BioWare could make an “open-world” RPG, in which the player could move freely through massive environments. But, as Flynn knew, Dragon Age: Inquisition was already behind schedule thanks to unfamiliar new technology. Their new game engine, Frostbite, was requiring more work than anyone at the studio had expected.