Play Nice: The Rise, Fall, and Future Of Blizzard Entertainment

by Jason Schreier

Since the team was so small, they were all able to sit and work together—a callback to the older days of Blizzard, even as the company had grown into the thousands. “It was this miraculous moment,” said Jay Baxter, a programmer. “Blizzard at the height of its power decided to start up a garage company and let them do what they wanted.”

Hearthstone was completely free, only charging players for additional content such as card packs and expansions. “When people asked how successful we’d be, I said, ‘I guarantee we’ll make dozens of dollars,’” said Dodds. The results blew everyone away. By the end of the month, the game had ten million registered users. Within the next few years it would reach one hundred million—more players than any game Blizzard had ever made—and would eventually generate hundreds of millions of dollars in revenue per year. This was Blizzard’s little skunkworks project, the company’s lowest priority, a game that had come close to being canceled several times. Now it was one of the company’s top money makers.

In the coming months, Blizzard doubled and then tripled the size of Team 5 so its developers could continue making new cards and expansions for Hearthstone. It was a resounding success, but to some members of the original team, the magic was lost. “I think Hearthstone’s story is a microcosm of Blizzard’s story,” said Jay Baxter. “We went from creating to churning.”

Blizzard had turned into an empire by taking other people’s video game ideas, polishing them, and turning them into phenomena like Hearthstone. But in the years after World of Warcraft, as the company grew rich and bloated, Blizzard would also experience something it had never faced before: someone else taking from one of their games and turning it into a phenomenon.

flew him to Irvine for a meeting. But Icefrog, a reclusive designer who avoided using his real name, arrived with a list of requests that Blizzard’s executives considered unreasonable. “He wanted full creative control,” said Richard Khoo, a designer on Team 1. Plus, World of Warcraft demanded everybody’s attention, so DOTA seemed like a trend they could safely ignore. “My recollection is that Blizzard wasn’t in a place to start building the game at the time,” said Icefrog in a statement through his agent, “so discussions never really got far at all.”

By the end of the year, League of Legends had more than 100,000 concurrent players, and two years later Riot Games would sell a majority stake to the Chinese conglomerate Tencent for a whopping $400 million. Riot became a titan of the video game industry, with a sprawling campus in Santa Monica and an army of recruiters who began calling Blizzard employees and offering to double their salaries if they came to work on League of Legends

The timing was interesting. Shortly before BlizzCon, the Seattle-based video game company Valve had announced its own game, DOTA 2, that served as a standalone successor to the Warcraft III mod. In fact, the real name for Blizzard DOTA was Blizzard All-Stars—the company had tacked on “DOTA” at the last minute to prepare for a future trademark battle.

Many of them began to feel confident—perhaps overconfident—that they had something special on their hands. “The rallying cry across the whole company was, ‘We’re going to do to League of Legends what World of Warcraft did to EverQuest,’” said one designer. “Which was dumb, because League of Legends already was World of Warcraft

By the end of 2013, League of Legends had twenty-seven million active daily users and Valve’s DOTA 2 was the most played game on Steam, while companies like Warner Bros. and Electronic Arts had taken their own stabs at the MOBA genre and failed spectacularly. Blizzard was entering a market so crowded it was a fire hazard.

But many of the younger gamers who were into MOBAs hadn’t grown up with Blizzard games—there had been no new ones released between 2004 and 2010—and they had no interest in the company’s version of Smash Bros. Matt Schembari, a programmer on Heroes of the Storm, started to suspect that something might be amiss when he couldn’t convince friends or family to accept free early codes for the game. He recalled growing nervous when his younger brother, a teenager, told him that none of his friends wanted it either.

“I’d been working on Heroes for four years at that point,” he said. “You’re staring down an endless content treadmill.” But there was also something bigger—more existential. For Klinchuch, cracks were beginning to form in the company’s once-polished facade. “I had a real sense that this is a very different Blizzard than the Blizzard I initially got hired at,” he said, “and the longer I stay, the higher the chances of bad memories.” There were little things, like perks that seemed to be disappearing. In the past, each employee would receive multiple free tickets to BlizzCon—now they were lucky to get one. Where the company had once taken employees to Vegas to celebrate the release of each individual game, now they were combining multiple launches into a single wrap party. “I definitely had a real sense of decline happening around me,” Klinchuch said.

But it was never quite clear to developers on the team if the game was generating enough revenue to support a team that had now grown to nearly two hundred people. “They were never transparent about what our successes and failures were,” said Phill Gonzalez, “even to the point where they didn’t want to tell us what skins were popular.”

Still, Blizzard’s executives suspected that a WoW killer would be on the horizon sooner or later—so they figured they should be the ones to make it. The natural next step seemed like a World of StarCraft, but they struggled to envision StarCraft lore fitting into an MMORPG, so instead they decided to develop a new fictional universe.

Other video games, like Half-Life and Fallout, painted a grim, dystopian picture—in contrast, Titan would feel bright and optimistic, perhaps attracting people who wouldn’t normally play Blizzard games. “We were really excited about appealing to a broad audience,” Thavirat said. “Gamers, non-gamers, young, old, men, women, everything in between.”

Titan would be set on an alternate version of Earth in the 2070s. The hook was that it was essentially two games in one, with players taking control of superhero-like characters who lived normal lives during the day and secretly battled against evil forces at night. An early presentation showed the player, as a professional chef, popping a dish into the oven before going off on a secret mission. When they returned, the dish was perfectly cooked and ready to serve.

Chris Metzen and Rob Pardo had never seen eye to eye. Metzen was a boisterous, emotional artist and a fountain of story ideas, while Pardo was quieter and more analytical, with a head for game design and an aloof demeanor. The two had first worked together as creative leads on StarCraft’s expansion, Brood War, then paired up to help steer Warcraft III and World of Warcraft. Both grew increasingly powerful at the company, becoming vice presidents when Adham retired in 2004.

Blizzard staff who worked with them, it felt like parents about to get a divorce. “We’d just sit there for a full hour and watch Metzen and Pardo argue,” recalled one developer on StarCraft II. “Metzen would say it should be pink. Pardo would instantly say it should be blue, and only an idiot would make it pink. It was painful to watch.”

People who worked with the pair quickly learned that they had divergent visions for what Titan should become. Metzen wanted to create a superhero universe, like the Marvel and DC comics he loved, with godlike figures duking it out in the skies and streets, while Pardo wanted characters to be more like secret agents—spies with superpowers who worked in the shadows to protect the world from insidious threats. As a result of these battles, Metzen left and returned to the project multiple times, which created more chaos and confusion.

As the Titan team expanded, bringing in veteran developers from across the video game industry, they built an endless number of prototypes for the civilian portion of the game: fishing, farming, photography, gardening, hacking. This non-combat section, which they called Titan Town, was essentially Blizzard’s take on Animal Crossing or The Sims—players would be able to deck out a house with furniture, run a business, and go on quests in their neighborhoods.

During playtests, the Titan team found that some of these game mechanics were fun in isolation, but nobody could envision what game developers called the “core loop,” or the sequence of actions that players would spend the bulk of their time doing. “It always felt like it was right around the corner,” said Thavirat. “Right around this milestone, this is where things will come together.”

Riot Games was using all that League of Legends cash to poach Blizzard’s staff—including, at one point, nearly the entire animation team. And the underlying technology behind Titan—all designed from scratch for this project—was hampering their progress,

An internal repository called TitanArt grew so bloated, with thousands of images, that artists would sometimes draw characters or cities only to later learn that someone else had already done the same years earlier. “The amount of art we did was enough for five games,” said artist Vadim Bakhlychev.

“There was a feeling that Blizzard had essentially written a blank check to fund this game and that bred a sense of complacency within the team,” the artist said. “We were not working with any kind of urgency.” Making games was always hard. Making new franchises was even harder. And making a new franchise at a company full of perfectionists, with the pressure of surpassing World of Warcraft, with a team that was growing larger than any project in Blizzard’s history—well, that was proving to be impossible.

“He at times seemed like an absentee game director,” said one developer. To mitigate this problem, Pardo had brought in two lead designers: Jeff Kaplan, who had designed quests on World of Warcraft, and Matt Brown, who had worked on The Sims and SimCity at EA’s Maxis. But Pardo remained director of the project and would occasionally jump in with feedback, forcing the team to change course and potentially throw out months of work. “I think when you want to lead something, you have a responsibility to lead,” said Connie Griffith, who worked as an assistant to Pardo. “And if you are not able to give it your full attention, you need to relinquish control.”

Rob Pardo first discovered the joys of game design while playing Dungeons & Dragons with his friends, who always wanted him to be the Dungeon Master, allowing him to create his own rules and build his own campaigns. When he wasn’t crafting fictional universes, he was gravitating toward anything competitive: ice hockey, Risk, video games. He took pride in being the best at anything he played—a trait inherited from his father. “He’s probably competitive to a fault, and instilled that in me,” Pardo later said on a podcast. “He’d never, ever let me win. So if I beat him, it was legit, and if I started winning at some activity we’re doing, then mysteriously we would stop doing that activity.”

After bouncing around departments for a couple of years, he left Interplay for a startup, where he and his team pitched a new game to Allen Adham. Adham wasn’t interested in the game but was impressed by Pardo’s multiplayer skills and design chops, and soon the two started talking about Pardo coming to Blizzard.

Who was this kid sitting with the testers and telling them what to do? It was soon clear, as Pardo was promoted to become the top designer on StarCraft’s expansion, Brood War, that he was being positioned as Adham’s successor, and when Adham left the company in 2004, Pardo took his role as lead designer of World of Warcraft. After that, Pardo would oversee every game Blizzard made.

Yet the accolades were breathless. “Pardo didn’t invent this kind of game, he merely perfected it,” the magazine gushed. Although the Time article wasn’t Pardo’s doing, Blizzard staff groused about how he had been portrayed. This was a company that tried hard to avoid glamorizing their developers—that would stamp “Design by Blizzard Entertainment” atop the credits of every game as a way of showing that everyone at the company had contributed. Adham and Morhaime had always preached that Blizzard had no rock stars, yet Pardo was becoming an industry celebrity.

As Pardo rose through the company’s ranks, he became a polarizing figure at Blizzard. Employees griped that he would take credit for other people’s ideas, and in one interview he even claimed responsibility for Adham’s old donut theory, which had been one of Blizzard’s core principles since long before he had joined. He also cultivated a reputation for passing the buck when things went wrong.

As Pardo’s responsibilities grew and his time became stretched between projects, he became known for what many of his subordinates called seagull management: He would swoop down, poop on ideas, and then fly away. “He’s a brilliant designer; he was just a terrible manager,” said Greg Street, a lead on World of Warcraft. “Which was common at Blizzard.”

On previous games, not everyone had always agreed with Pardo’s decisions, but at least he had made them. On Titan, he appeared to be unwilling or unable to commit to a vision—perhaps because of the pressure.

One morning in the spring of 2013, as developers on the Titan team trickled into the office, they noticed that rows of folding chairs had been set up in the common area for an impromptu all-hands meeting. When it started, Pardo stood up and dropped the big news: Titan wasn’t working out, so he and the other Blizzard executives had decided to reboot it.

Later, Pardo stood up in front of the company, tears in his eyes, as he gave a presentation about why Titan had failed. He noted that he should have done more about the game’s technical and artistic problems, which Team 4 members saw as a backhanded way of blaming other people for the game’s issues—the last straw for some. The longtime Blizzard designer had stewarded a lot of great games and was Morhaime’s heir apparent, but now he had lost the trust of Blizzard’s other leaders. Several executives and directors, including Metzen, demanded that he be removed from a position of leadership. They didn’t want to report to him anymore; didn’t want him overseeing their new projects.

But Morhaime may have still been haunted by his decision to remove Mike O’Brien as lead of Warcraft III back in 1999, which had led to the departure of three top Blizzard programmers and the creation of ArenaNet, a significant competitor. In retrospect, Morhaime would tell confidants, he felt like he should have taken a more diplomatic approach back then. Now he didn’t want to repeat the same mistake and drive away one of the people who he believed was most integral to Blizzard’s success. Problem was, by doing nothing, Morhaime might drive away a group of other people who were integral to Blizzard’s success,

Pardo took a sabbatical as Morhaime and his team tried to determine his future at Blizzard. At first they talked about potentially shifting his responsibilities, but after a few months, Morhaime met with Pardo and told him it would be best if he resigned. In the summer of 2014, Pardo announced that he was leaving Blizzard,

On July 9, 2013, Kotick won the staring contest, and Vivendi accepted the terms as originally proposed. Vivendi agreed to sell at $13.60,

More importantly for the future of the company, the deal meant that Activision was now free to operate without Vivendi’s influence. On paper, the voting power of Kotick and Kelly was capped at around 25 percent—later reduced to around 20 percent by a legal settlement—but in practice they were now in control.

Cataclysm, which would overhaul the entire world of Azeroth through a series of apocalyptic events, Blizzard announced that the game had reached 12 million subscribers.

first to 11 million subscribers, then to 10 million. Players were upset with some of the changes in Cataclysm—the world felt too empty; the dungeons were too hard—and loudly voiced their discontent. Suddenly, Kotick and other executives were showing up at Team 2 meetings to talk about retention and engagement, while Morhaime promised on investor calls that they would work to produce content more quickly.

But developing an expansion for World of Warcraft was a slow, inefficient process hampered by clunky tools, content bottlenecks, and a chaotic culture inherited from the old days. “In my head I thought it’d be a really organized structure, but it wasn’t,” said designer Chris Kaleiki. “It was completely based on tribal knowledge: if you want to know how to do something, go to someone and ask them.”

World of Warcraft still drove the bulk of Blizzard’s business, giving the other teams extended timelines to work on Diablo and StarCraft. During company meals and parties, members of Team 2 would end arguments with colleagues on other teams by sniping, We’re paying for your game

revealed that Blizzard would be embarking upon its own ambitious moonshot plan to release a new World of Warcraft expansion every year. Their expansions typically took about two years to make, so they hoped to double the size of the World of Warcraft team and divide it into two units, each alternating so they could stagger new expansions every year. But as the old saying goes, nine women can’t make a baby in a month.

production director J. Allen Brack did the revealing

“It wasn’t a bad strategy; it was an ambitious strategy,” said Street. “But it totally failed. We never were able to hire enough leaders to get the second team up and running.”

“Wall Street predictability was something Bobby was looking for,” said Dusty Welch, who worked for Nestlé and Dole before joining Activision.

It was a stark contrast to Blizzard, whose leaders saw schedules as malleable and were always willing to delay games, although Activision executives liked to joke that Blizzard actually shared that desire for predictability: if they agreed to a deadline, you could rest assured they’d miss it.

Bobby wanted his games to be good, but he also wanted them to be predictable—to fit neatly into Activision Blizzard’s fiscal calendar so the company could show growth from year to year.

Kotick was once asked by a journalist what made Activision so successful. He responded by describing his company as “ruthless prioritizers of opportunity”—a mantra that concisely summed up how Activision had done business for the previous twenty-five years. Kotick had spent decades betting on video games that had, in his words, the “potential to be exploited every year across every platform,” and the floors of his company were littered with the corpses of franchises that had been sucked completely dry.

Back in 1999, Activision had found unexpected success with Tony Hawk’s Pro Skater, a skateboarding game that let players ollie and kickflip to a killer pop-punk soundtrack. Kotick acquired the game’s developer, Neversoft, and they began producing new sequels every fall. Sales tapered off after a few years, and by the late 2000s fans were wondering why Tony Hawk’s name was attached to so many duds.

Mick West, a cofounder of Neversoft, said the company never questioned or even regretted the decision to go annual—until much later, after the fifth or sixth games, when he departed. “There were concerns that it was getting harder to be more creative in a limited time frame,” West said.

Dusty Welch, who was head of publishing, recalled one memorable planning meeting in which Kotick stopped him in the middle of a discussion about yearly releases. “He said, ‘Dusty, I like this, but this is not aggressive enough,’” Welch recalled. “‘I want to get to a launch every quarter, and then I want to get to a Guitar Hero launch every month.’

Paul Reiche, whose studio created the lucrative toy-based game series Skylanders, watched his franchise find massive success and then die off a few years later. “I think annualization made money,” Reiche said. “I think annualization sort of hurt the creative capabilities.”