On May 15, 2012, hundreds of thousands of people across the world loaded up the Battle.net Internet client and slammed the launch button for Diablo III, a game that the developers at Blizzard had been making for nearly ten years. Fans had waited patiently for this moment, counting down the days until they could again click-click-click their way through demons in a hell-ish hodgepodge of gothic fantasy. But at 12:00 a.m. Pacific time on May 15, when Diablo III went live, anyone who tried to load the game found themselves greeted with a vague, frustrating message:

The servers are busy at this time. Please try again later. (Error 37)

After a decade of turbulent development, Diablo III had finally gone live, but nobody could play it. Some people gave up and went to bed. Others kept trying. An hour later:

The servers are busy at this time. Please try again later. (Error 37)

“Error 37” turned into a meme, mushrooming across Internet forums as fans vented their frustration. Diablo players had already been skeptical about Blizzard’s decision to make Diablo III online only—a decision that cynics assumed was driven by fear of piracy—and these server issues nourished the belief that it had been a bad idea. It immediately occurred to fans that if they could play Diablo III offline, they would be fighting their way through New Tristram right now, not trying to figure out what Error 37 meant.

Over at Blizzard’s campus in Irvine, California, a group of engineers and live-ops producers sat in their self-proclaimed “war room,” freaking out. Diablo III had outsold their wildest expectations, but their servers couldn’t handle the flood of players trying to log into the game. Around 1:00 a.m. Pacific time Blizzard posted a brief message: “Please note that due to a high volume of traffic, login and character creation may be slower than normal. . . . We hope to resolve these issues as soon as possible and appreciate your patience.”

A few miles away, at the Irvine Spectrum outdoor mall, the rest of the Diablo III team had no idea that people couldn’t play their game. They were busy partying. Hundreds of hard-core fans, dressed in spiky armor and carrying giant foam battle-axes, had come out for the official Diablo III launch event. As Blizzard’s developers signed autographs and passed out swag to the crowd, they started to hear whispers about overloaded servers. Soon it became clear that this wasn’t a standard launch hiccup.

“It really caught everybody by surprise,” said Blizzard’s Josh Mosqueira. “It’s kind of funny to say. You have such an anticipated game—how can it catch anybody by surprise? But I remember being in the meetings leading up to that, people saying, ‘Are we really ready for this? OK, let’s double the predictions, let’s triple the predictions.’ And even those ended up being super conservative.”

Later that day, as fans tried again to load Diablo III, they found another vague message: Unable to connect to the service or the connection was interrupted. (Error 3003). Error 3003 didn’t grow as popular as its younger, catchier brother, although it did make one wonder how the other 2,966 errors had been averted. The next day, Error 37 reemerged, along with a host of other server issues that continued plaguing Diablo III players for days after the game launched. Blizzard’s war room was active 24/7 as tired engineers gathered around computers, sipping on coffee and trying to figure out how to bolster their network.

Within 48 hours they’d managed to stabilize the servers. Errors would still pop up sporadically, but for the most part, people could now play the game without interruption. On May 17, once things had settled, Blizzard sent out a statement of apology. “We’ve been humbled by your enthusiasm,” they wrote. “We sincerely regret that your crusade to bring down the Lord of Terror was thwarted not by mobs of demons, but by mortal infrastructure.”

Finally, the world could play Diablo III. Like its predecessors, the third Diablo would let you build up a character and hack your way through landscapes full of demons, collecting fistfuls of shiny loot along the way. You’d unlock abilities based on the class you’d selected (wizard, demon hunter, etc.), switching between a large array of spells and skills. And you’d power through dungeon after dungeon, all of which were procedurally generated so that no two playthroughs would be the same. It appeared, at first, to be the game that fans had been waiting for.

In the weeks to come, however, players would discover that Diablo III had some fundamental flaws. It was satisfying to rip through hordes of monsters, but the difficulty ramped up way too fast. Legendary items dropped too infrequently. The end-game was too challenging. And, perhaps most frustrating of all, the loot system seemed to revolve around the in-game auction house, where Diablo III players could use real-life money to buy and sell powerful equipment. This controversial system made Diablo III feel like a dreaded “pay-to-win” game, in which the best way to beef up your character wasn’t to play the game and make fun decisions, but to type your credit card number into a form on Blizzard’s website.

Since Blizzard’s founding in 1991, the studio had developed a reputation for making fantastic games, including the cultural touchstones Warcraftand StarCraft. When you saw the jagged blue Blizzard logo attached to a game, you knew you were getting something unparalleled. With Diablo II in 2000, Blizzard had developed the definitive action-RPG, a game that inspired countless all-nighters and LAN sessions as millions of teenagers gathered to battle disfigured demons and hunt for elusive Stones of Jordan. Diablo II was widely considered one of the best games ever made. Now, in May 2012, the rocky launch of Diablo III had associated the Blizzard logo with something that the company had never experienced: public failure. And even after Error 37, the problems were just getting started.

Josh Mosqueira had always hated winters in Montreal. A Mexican-Canadian with a thick blended accent who had served as a Black Watch infantryman in the Canadian army, Mosqueira spent his early career years writing role-playing games for the publisher White Wolf while trying to break into the video game industry. After working on a few games and spending a seven-year stint at Relic Entertainment in Vancouver, Mosqueira moved across Canada to work on Far Cry 3 at Ubisoft’s massive office in Montreal, where winter temperatures tended to drop a few degrees lower than they should in any human-inhabited city.

On one particularly snowy day in February 2011, more than a year before Error 37, Mosqueira got a call from Jay Wilson, an old friend from his Relic days. Wilson was now working at Blizzard Entertainment in Irvine, California, and they were looking for a new lead designer on Diablo III, the game he was directing. Someone from Ubisoft had applied, so Wilson wanted to know what the culture was like over there. Would this prospective new designer fit in? The two friends got to talking, and then Wilson offered up another option: What if Mosqueira took the job?

Mosqueira said he’d have to think about it. He looked out his window, watching the snow fall, and realized there wasn’t much to think about. “Fast forward two and a half months, and I find myself walking into these halls as the lead designer for the console version of [Diablo III],” Mosqueira said. His job was to direct a very small team—three, at first, including him—that would adapt Diablo III for the Xbox and PlayStation. This was a surprising initiative for Blizzard, which for years had resisted releasing games on consoles, instead choosing to put out massive hits like World of Warcraft and StarCraft II only on PC and Mac. With Diablo III, Blizzard’s brain trust finally saw an opportunity to explore the giant world of console gaming.

Mosqueira and his team settled into a section of the office and started playing around with prototypes, trying to figure out how to get Diablo III feeling good on a controller. Blizzard had given Mosqueira’s team liberty to overhaul everything for the console version, and they took advantage of that freedom, changing the skills of every class to account for their new control scheme. “A lot of the timing of skills on console felt off, because instead of focusing on your cursor, your eye, you’re focusing on your character,” Mosqueira said. “So we essentially went in and tweaked every skill in the game.”

Toward the end of 2011, as the Diablo III PC team started crunching for the spring release, Mosqueira and his colleagues put the console project on pause so they could help finish the game. “The three of us—and at that time we were about eight—eight of us were all new to Blizzard, so we sort of felt like we have to,” Mosqueira said. “We want to be part of this. It’s going to be really exciting. It’s going to be a big moment in Blizzard history, and we’re just happy to be part.”

Then came Diablo III’s launch, and Error 37, and manic days at Blizzard in May 2012 as they tried to stabilize the servers. While Mosqueira and his crew went back to work on the console version, Diablo III’s other designers began trying to solve the game’s deeper problems. Players clearly weren’t happy with the loot system, for example, but what precisely was wrong with it? How could Blizzard make the endgame as addictive as it had been in Diablo II, where players spent hours and hours fighting through demons and hunting for gear even after they’d already finished the story?

The biggest problem, the developers realized, was the game’s difficulty. Blizzard’s designers had originally built Diablo III’s difficulty system to mirror Diablo II. You’d play through the full game once on Normal mode, then play it a second time on the challenging Nightmare mode, and crank through a third time on Hell mode. Diablo III repeated that structure and introduced a fourth difficulty option, Inferno. Designed for players who had already hit the level cap, Inferno was blisteringly tough, to the point where you couldn’t beat it without the game’s best gear. But Diablo III’s best gear dropped only when you played on Inferno mode, creating a nasty, demonic version of the chicken-and-egg dilemma. How could you get Inferno gear if your gear wasn’t good enough to get through Inferno in the first place?

There was one option: the auction house. If you didn’t want to bash your head against the wall in Inferno mode, you could dish out real money for better gear, which was the exact opposite of what most players wanted to do. As a result, some crafty players found ways to abuse the system. Thanks to Diablo III’s random number generator, the chances of getting loot from a powerful enemy weren’t much better than the chances of getting loot from smashing a stationary pot. Once players realized this, they’d spend marathon sessions doing nothing but breaking pottery. It wasn’t particularly fun, but hey, it beat giving away real money.

What became apparent to Blizzard in the coming months was that people were more interested in gaming Diablo III than they were in playing it, a problem that would take serious investment to fix. From May 15 through the end of August, the Diablo III team released roughly eighteen free patches and hotfixes that fixed bugs, tweaked character skills, and addressed various player complaints. The largest of these patches, on August 21, 2012, added a system called Paragon levels that would let players get stronger once they’d hit the level cap (60). It also made Inferno mode less difficult and added a bunch of unique effects to legendary gear, so getting a sleek new weapon would make you feel like a devastating war machine.

But Blizzard knew these patches were just bandages—temporary solutions to get players to do more than smash pots. There was still a leaking wound in Diablo III’s side. And it would take a great deal of time to stitch it up.

In the center of Blizzard’s sprawling Irvine campus is a giant statue of a Warcraft orc. Surrounding that statue is a ring of plaques, each with a different message that’s meant to be a mantra for Blizzard employees. Some of them seem like they’ve been ripped from parody motivational posters—“Think Globally;” “Commit to Quality”—but one resonated strongly with the Diablo III team throughout 2012: “Every Voice Matters.” Players were frustrated, and Blizzard’s developers felt compelled to listen to them. Diablo III’s producers and designers took to as many Internet hangouts as possible, from Reddit to the Blizzard forums, to collect and collate feedback on how to make the game better. In blog posts and letters throughout the summer and fall, Blizzard promised players that they had a long-term plan to fix the auction house, improve the loot system, and make Diablo III’s endgame more fun.

This sort of commitment was atypical. Usually, a developer would release a game and then move on, maybe leaving behind a skeleton team to fix any lingering critical bugs before the studio dove into its next project. But Blizzard had built its reputation as a premier developer by sticking to games for the long haul. Blizzard would update all its games with free patches for years and years after they launched, believing that the support would lead to goodwill from fans, which in turn would lead to better sales.

By the end of July 2012, Diablo III had sold an astounding ten million copies. Blizzard’s developers felt like they’d made a fun game, but they also knew it could be so much better. “It was a diamond in the rough,” said Wyatt Cheng, a senior technical designer. “We knew we needed to polish it a little bit more. It just needed that extra little bit.” It helped that Blizzard’s CEO, Mike Morhaime, had told the Diablo III team to keep working on updates and releasing free patches for the indefinite future. “There are few other companies where (a) we could sell millions of copies and still feel like we could’ve done better,” said Cheng, “and (b) given some of the initial launch problems, [we’d be] given this long runway to work on the game and make it better.”

That was one way of looking at the situation. The other was that the people who worked on Diablo III—some of whom had been on the game for nearly a decade—wouldn’t get a break. Anyone who’s spent a great deal of time on a single project knows how relieving it feels to finish—and how when it’s done, you never want to look at it again. “I was listening to a podcast,” said Cheng, who had started on Diablo III in the early days. “There’s a person who’s been making the tours promoting her book”—the psychologist Angela Duckworth—“and she used to write about grit. She said grit is this quality that a lot of successful people have. And it’s this persistence, to push forward with something. Anything worth doing, it’s not necessarily day- to-day fun sometimes. Sometimes it is. Great when it is. But grit usually means that somebody sees the long-term goal and they see the long-term vision and they push through any obstacles that they have on a day-to-day basis, with the end in mind.”

The “end,” or at least the next big milestone, was Diablo III’s expansion. Traditionally, Blizzard would produce meaty expansion packs for every game the company released, and Diablo III’s developers knew theirs would be the best opportunity to overhaul the game. Toward the end of 2012, they started putting together a giant Google document full of problems that they needed to fix and features they wanted to add, like an item revamp and a new set of goals for the endgame.

But they needed a new leader. Diablo III’s longtime director, Jay Wilson, had told the team that he planned to step down, citing burnout after nearly a decade spent working on the same game. (Oddly, Blizzard would not give Jay Wilson permission to be interviewed for this book.) Blizzard needed a new director, not just to lead development on the expansion, but also to shape the future of Diablo III.

And there was one newcomer who might make the perfect fit.

When he first saw the opening on Blizzard’s internal website, Josh Mosqueira wasn’t going to apply. He’d been enjoying the challenges of porting Diablo III to consoles, and he liked overseeing a small team. Although his squad had expanded from three to around twenty-five, it was still a drastic departure from his days at Ubisoft, where he’d had to help coordinate a team of four hundred–something people. Even when Wilson and other leaders at Blizzard encouraged him to apply for the director position, Mosqueira was reluctant. “I was just really happy being a lead and having a console project to look after,” he said. “Just getting my hands dirty actually working on the game and not just PowerPoints.”

But Mosqueira also loved the culture of Diablo III’s development team, and soon enough he was persuaded to file an application. After a gantlet of interviews, not only with Blizzard’s management but also with all his coworkers, Mosqueira was called into the office of Blizzard’s cofounder Frank Pearce for the news. He’d gotten the job. Mosqueira hadn’t been on the Diablo III team for very long, but people respected him as a designer and leader, and Blizzard wanted him to drive the future of the game. “When they told me, it was a pretty amazing moment,” Mosqueira said. “Quickly followed by a lot of panic, when you realize that Diablo’s one of the big franchises, not just at Blizzard but in the industry, and to be given that responsibility was intense.”

After becoming director, one of Mosqueira’s first moves was to sit down with the rest of the staff on Diablo III, who had all once been his colleagues but now reported to him. He asked how they were feeling. What did they like about the game? Where did they see Diablo III going in the years to come? Usually, video game expansions were additive—they’d provide new content, new areas, new loot—but for Diablo III, Blizzard wanted to make something transformative. “It quickly became apparent that they really wanted to use the expansion to not just adjust and pivot from the launch, but really create this platform that will take Diablo into the future,” Mosqueira said. “That’s the kind of pressure that the team put on themselves. And they were thinking really big.”

What had also become clear to Mosqueira, and what he tried to convey to the rest of his team, was that they didn’t really know what Diablo III was yet. Mosqueira liked to point out that when gamers got wistful about Diablo II, they weren’t remembering the game’s original incarnation—they were thinking about what it became in 2001, after the developers reacted to fans’ feedback and released the expansion Lord of Destruction. That was the game people remembered. It was the result of millions of players giving Blizzard feedback, and Blizzard reacting directly to that feedback.

“What makes it really hard is you can build a game, you can test a game, and you can think you know the game—until you release it.” -Josh Mosqueira

“What makes it really hard is you can build a game, you can test a game, and you can think you know the game—until you release it,” said Mosqueira. “Within the first day, probably more people have spent more hours playing the game collectively than [in] the entire development up to that point. So you’re going to see things you never intended. You’re going to see players reacting to things. . . . They’re playing with it, they’re engaging with it, they’re interacting with it. And really, what makes it hard is learning the discipline and the rigor to know how to react to that.”

As the Diablo III team looked ahead to the expansion, which they were calling Reaper of Souls, they saw it as a chance to make amends—not just for Error 37, but for all of Diablo III’s early faults. This was the team’s chance to make their own Lord of Destruction and reach the heights that Diablo II had set all those years ago. “We took it as, this is our one opportunity to win back the fans,” said Rob Foote, a senior producer. “Let’s do it.”

It might have seemed odd, to an outside observer, that a game ten years in the making would launch with so many problems. Mosqueira’s theory was this: Diablo III’s issues were a direct result of a development team haunted by the beloved Diablo II. As he said in a 2015 talk: “The specter of Diablo II loomed large over the team. The pressure of trying to live up to the legacy of this incredible game weighed heavily on the team and impacted so many of the decisions.”

Where BioWare’s Dragon Age: Inquisition team was bedeviled by negative reactions to Dragon Age 2 (see chapter 6), Blizzard had the opposite problem: Diablo III had to surpass Diablo II’s massive success. The designers of Diablo III had been willing to make big innovations in some areas, like the flexible skill system, generally considered a highlight of the game. But, in Mosqueira’s view, they were too rigid about other series traditions.

As a newcomer, Mosqueira was willing to challenge everyone’s conceptions of what made Diablo feel like Diablo, even if it meant picking fights with some of Blizzard’s veterans. For the console version, which was still in development alongside Reaper of Souls, Mosqueira had fought hard for an “evade” feature that would allow players to use a joystick to roll around on the ground, dodging enemies’ attacks. This was controversial. “Evade was extremely contentious on the team,” Mosqueira said. “Extremely, extremely contentious. I would get into very heated discussions with some of the other designers about why we needed it on console.”

Mosqueira argued that players would get bored walking around for hours at a time without having some method of switching up their movement, like the beloved jump button in World of Warcraft. Other designers pointed out that offering an evade feature would diminish the impact of items that boosted players’ movement speed—a concept from Diablo II—therefore making the game less rewarding in the long term. “Both are really strong arguments,” said Mosqueira. “Both are right arguments. At the end of the day, you have to say, ‘OK, I’m willing to sacrifice some of that long-term reward for the short-term visceral feeling.’ . . . I understand I’m giving away the power reward, but for this to feel like a console game, my thumbs need to do something at regular intervals, and it just feels good. It’s a console thing to do.” (Mosqueira eventually won this battle, and the evade feature made it into the game.)

By experimenting on consoles, where there was less pressure to adhere to the formula of Diablo II, Mosqueira and his team could take steps that seemed radical to the rest of the team. “I think that was one of the most liberating things,” he said. “The PC team, for all the right intentions, because of all the pressure, the expectation, some of their initial design was very conservative. But on console, it was a bit of the Wild West. And in some ways, looking back at it... there’s a level of being very naive. We’ve been mucking around with this game for about six months, not knowing all the history behind all the decisions leading up to this moment, just walking in like kids and pushing buttons.”

For a game that had been in development for so many years, a fresh perspective could be useful, especially when Blizzard began reexamining core parts of Diablo III, like the loot system. In the PC version of Diablo III, enemies would spawn fountains of loot when they died, offering that familiar dopamine rush for players as they picked up cool new weapons and accessories. But without a mouse and keyboard, sorting through all those glittering rings and amulets could be a chore. As the Diablo III console team playtested the game, Mosqueira found that this loot overload was impeding people’s progress, forcing them to stop every few seconds just to organize their inventories.

That’s when they tweaked the formula. “We said, ‘OK, every time a gray or a white item’s going to drop, seventy percent of the time it’s going to just drop gold,’” said Mosqueira. The change might have seemed drastic to Diablo II devotees, but it wound up forming the foundation for what the team called Loot 2.0, a system that would improve Diablo III on both PC and consoles. “We started realizing that maybe we can be dropping less,” said Mosqueira, “and if you’re dropping less, well, we need to drop better.”

With Loot 2.0, Mosqueira and his team hoped to address every criticism players had about gear in Diablo III. Fans complained that it took way too long to get the top-tier “legendary” items, so Loot 2.0 guaranteed that each major boss would drop legendary gear. Fans pointed out that, when they did finally get legendary items, the game would generate those items’ stats randomly, making it likely that a player would spot a shiny orange weapon and get excited, only to find that the weapon was useless for his class. So Loot 2.0 introduced a weighting system, skewing the random number generator to increase the chances that when a player picked up a legendary item, it would be something she needed.

As Blizzard’s developers met throughout 2013 to talk about what they wanted to do with Reaper of Souls, “randomness” became a driving conversation topic. After all, random numbers had always been the pulsing heart of Diablo. Since the first game in 1996, which sent players battling through procedurally generated dungeons under the decrepit city of Tristram, Diablo games had relied on a random number generator for just about everything. Dungeon layouts were random. Treasure chests were random. Most magical items were random, too; the game would piece them together from a large table of prefixes and suffixes, with each item’s attributes tied to its name. (A “Lucky” belt would boost the amount of gold you received from monsters. A sword “of the Leech” would offer health every time you attacked.)

This randomness was what gave Diablo its mass appeal. Playing a Diablo game was sort of like campaigning in Dungeons & Dragons: every time you played, you’d have a different experience. There was something naturally addicting about finding a new item and hitting “identify,” knowing you could wind up with pretty much anything. Diablo appealed to that same instinct that makes us want to feed all our cash into slot machines and lottery tickets. It would have fit in nicely next to the craps tables at a sparkly Vegas casino.

It took a long time before the designers realized that their obsession with random numbers was hurting Diablo III. “I started to worship at the throne of randomness,” said Kevin Martens, a lead designer. “And when I could make something more random, I would, missing the point that randomness is a tool to get replayability. . . . When people ask me, ‘What’s the major difference between Reaper of Souls and Diablo III?’ my shortest possible answer is: We shaved the rough edges off randomness. We made randomness work for the player instead of against them.”

That was where Diablo III diverged from Las Vegas—Blizzard didn’t want the house to always win. Josh Mosqueira and his team realized that the way to keep players happy was to give them the edge. “When Diablo III shipped, whether you got a legendary [item] or not was just a whole bunch of die rolls,” said Mosqueira. “Sometimes you’d get lucky; sometimes you wouldn’t get lucky. . . . In Reaper, we just said, OK, we don’t want to cheat. We don’t want the player to feel that we’re making it easier for them or anything like that, but we just need to raise the floor, so it doesn’t take me 104 hours to find a legendary.”

They also needed to do something about the difficulty. Back when they’d developed the original version of Diablo III, Blizzard’s designers had believed that players wanted a game more challenging than any Diablo before it. “We had this video, ‘Diablo’s going to brutalize you,’” said Martens. “We had people on our team talking about how hard it was and how even though they’re experienced developers, they still get murdered. The truth is—now we’re on Hindsight Mountain looking back—that some people want it really hard and some people want a little bit easier. And everything in between.”

It wasn’t just that Inferno mode was too tough. Players had lost their appetite for playing through the same campaign multiple times, with nothing changing but monsters’ strength. The structure that had felt rewarding in 2001 was, for several reasons, a slog in 2012. Video game design had made drastic leaps in the past decade. Dozens of Diablo clones had emerged over the years, and some of them had even improved Diablo II’s structure (though none were quite as successful). When Diablo III came out, people expected a less repetitive rhythm.

Reaper of Souls was the opportunity to solve these problems. Blizzard’s short-term solution for Diablo III had been to make Inferno mode less challenging via postgame patches, but with Reaper, they could take things further. “It was probably late November of 2012 when I started to think maybe we should completely redo the difficulty system,” said Martens. That was a tough prospect, though. “The whole game is built for these four difficulties. Every monster has their numbers tuned for each of these four difficulties.”

Kevin Martens zoomed out. What if, instead of treating difficulty levels as stages, the Blizzard team overhauled Diablo III’s structure entirely, making it so monsters would scale with the player’s power over the course of the game? And then what if they added a new modifier system, so that anyone who wanted more of a challenge could switch to Hard or Expert modes to boost enemies’ health and damage? If you wanted to make things a little easier, you could simply flip back to Normal. To solve Inferno mode’s chicken and egg problem, Blizzard would kill both the chickens and the eggs.

To an outside observer this might have seemed like an obvious method—most other games use difficulty modes this way—but for a Diablo game it was revolutionary. “Initially it seemed like this impossible mountain to climb,” Martens said. “You knew you needed to change this major thing, [but] you never thought about the game in that terms, in automatic difficulty before. We’d never considered it in that manner.”

They’d never considered it because of Diablo II. It had never occurred to the team that they should change the difficulty structure because that was just always how Diablo games had done things. Playing through the game on Normal, then again on Nightmare, and then a third time on Hell was part of what made Diablo Diablo, wasn’t it? Blizzard had taken flack during Diablo III’s early development just for the minor act of making health orbs drop from enemies, which some fans saw as breaking series tradition, so it had been tough to even think about a move as drastic as upending Diablo’s entire structure. But what if they did? And what if they found a better replacement for it?

In the months after launch, several Diablo III players had complained about not being able to teleport back and forth between the game’s four acts, and Blizzard had been looking to find a way to address that feedback. “We worked with the engineers and they said, ‘Oh yeah, we can figure out a way to do that,” said Rob Foote. “And actually, I think it was an engineer who said, ‘But couldn’t we do something better?’”

Again, they all went into brainstorming mode. What if, instead of solely letting players teleport between areas, Diablo III gave them a whole new mode that changed everything? And what if that mode became the focal point for Diablo III’s new endgame?

They would call this new feature Adventure Mode. Once you’d finished Reaper of Souls, you could open up this new Adventure Mode and jump into any area of the game, from the deserts of Caldeum to the frozen peaks of Arreat. Each of the game’s five acts would give you a series of randomized bounties like “kill a boss” or “clear a dungeon,” and the more bounties you completed, the more loot you’d snag. Adventure Mode would also add special events and what the game called “Nephalem Rifts,” multitiered dungeons that shuffled areas and monsters from all throughout Diablo III like some sort of gothic mixtape. As Blizzard envisioned it, Adventure Mode would entertain players for hours and hours once they’d finished the game. It sure beat smashing pottery.

In August 2013, at the Gamescom trade show in Germany, Blizzard prepared to announce Reaper of Souls to a packed room full of reporters and fans. This expansion would center on the demonic archangel Malthael. It would come with a new class, the crusader. And it would introduce all sorts of features, starting with Loot 2.0, in a free patch that Blizzard hoped would signal that the developers of Diablo III were listening to the complaints.

“Right before we were going to do the announcement, the energy in the room was tense,” said Josh Mosqueira. “I could feel that everybody was [thinking], ‘Mmm, this better be good.’ You could feel that they almost were expecting to be disappointed.” Then Blizzard put on the announcement video: a four-minute opening cinematic for Reaper of Souls, introducing the world to Malthael. Holding a nasty scythe in each hand, the archangel sliced his way through a group of Horadrim mages and attacked his former brother, the angel Tyrael. “The Nephalem will stop you,” Tyrael said. Responded Malthael: “No one can stop death.”

Gamescom’s audience erupted in applause. “It’s almost like this wave of excitement,” said Mosqueira. “You could feel it. I said, ‘OK, I think people are willing to give us another chance. Let’s not fuck it up.’”

“OK, I think people are willing to give us another chance. Let’s not fuck it up.” -Josh Mosqueira

Originally, Blizzard had planned to put out Reaper of Souls later in 2013, but the Diablo III team realized they needed more time, delaying it to the first quarter of 2014. This was a surprise to just about nobody. Blizzard had a reputation for taking its sweet time with games—Diablo III had taken ten years, after all—and you’d be hard-pressed to find a Blizzard game that hadn’t missed at least one deadline.

One quote, delivered by the director of StarCraft II, Dustin Browder, has always stuck out as a telling description of how Blizzard makes video games. In June 2012, over a year after Blizzard had hoped to ship StarCraft II’s first expansion, Heart of the Swarm, Browder spoke to me about the game’s progress. “We are ninety-nine percent done,” he said, “but that last one percent’s a bitch.” Heart of the Swarm wouldn’t come out until March 2013. That last one percent took nearly a year.

“The thing that makes scheduling challenging is iteration,” said Rob Foote. You have to allow for iteration if you want to make a great product.” Iteration time was the last one percent. Blizzard’s producers tried to leave blank slates at the end of their schedules so that their development teams could push, tug, and polish every aspect of their game until they felt like they had something perfect. “And it’s challenging too,” said Foote, “because people say, ‘What’s in that, it’s a lot of time, what are they actually doing?’ They’re iterating. We don’t know what they’re going to do, but they’re going to be doing something.”

Even with the extra time for Reaper of Souls, Josh Mosqueira and his crew had to cut some features. In conjunction with Adventure Mode, the Diablo III team had devised a system called Devil’s Hand that would place fifty-two high-powered enemies throughout the game’s world. Players would be able to kill them all for collectible items in hopes of eventually getting all fifty- two. The Diablo III team didn’t have enough time to get Devil’s Hand’s collection system in shape, though, so Mosqueira decided to cut it. “We figured, we have extra time, but we can’t get both of these right,” he said. “And the really important one is Adventure Mode, because that really changes the way people play the game. So we had to put Devil’s Hand off to the side.”

As the months went by, everyone at Blizzard felt ecstatic about the progress they were making. Since Error 37, they’d changed Diablo III’s formula, overhauled the loot system, and thought they could win back millions of players with Reaper of Souls. But Mosqueira still felt like the game had a critical flaw that they had yet to address, something that seemed at odds with how they wanted people to play the game: the auction house.

When Blizzard first announced Diablo III’s real-money auction house, cynics saw it as a cash grab. After all, Blizzard got to take a healthy cut every time a player sold an item for cash. Blizzard’s developers argued that they had more noble motives, insisting that they’d built the auction house to improve the experience of trading items for players. Back in 2002, Diablo II: Lord of Destruction had become infested with third-party gray markets, in which people would trade real money for powerful items on sketchy, insecure websites. Blizzard’s goal was, as Kevin Martens put it, to provide “a world-class experience” for those players, one that was safe and secure.

Not long after Diablo III’s launch, however, it became apparent to Blizzard that the auction house was hurting the game. Some players enjoyed trading, sure—especially the ones who were farming loot and selling it for a healthy profit—but for many, the auction house made Diablo III a significantly worse experience. It reduced the thrill of hunting for gear. What was the fun in getting a great roll on a cool new piece of armor if you could just hop on the market and buy a better one?

One group of players, who called themselves the Ironborn (after Game of Thrones’ House Greyjoy), made a point of refusing to use the auction house. They even sent petitions to Blizzard asking if the developers would consider adding an Ironborn mode to Diablo III. “This was a community of players who were saying, ‘Hey guys, Diablo III is the exact same game, but I’m having this whole new experience choosing to play it without the auction house,’” said Wyatt Cheng. “You can look at Diablo through that lens and say, you know what, we’ve got this really amazing game, but the auction house is having this distorting effect on how some people might perceive it.”

One day in September 2013, as Reaper of Souls was in full production, Josh Mosqueira sat in a meeting, doodling in a notebook. It was one of Blizzard’s routine monthly strategy meetings, at which CEO Mike Morhaime would get together with company executives and project leads to talk about the business, and much of the technical finance talk was bouncing off Mosqueira. Then the conversation turned to Diablo III, and suddenly they were talking about the auction house.

“[Mike] said, ‘Well what do you think?’” said Mosqueira. “If I was anywhere else, I probably would’ve said, ‘You know what, I think we still need to figure things out,’ or ‘I’m not quite sure.’ But looking at those guys, and knowing how important having our players have faith in us was, I said, ‘You know what, guys? Maybe we should just kill it.’”

After some brief discussions about how the logistics might work—How would they communicate it to players? What would they do with current auctions? How long would they need to wait?—the decision crystallized. It was time for Diablo III’s auction house to die. “I was [thinking], ‘Wow, this is happening,’” said Mosqueira. “I think to Mike’s credit, Mike’s a huge gamer. He loves the games. He loves players more than anything else. And he’s just willing to make those decisions, and say, ‘You know, this is going to hurt. But it’s the right thing to do.’”

On September 17, 2013, Blizzard announced that it would shut down the auction house in March 2014. Fans were, for the most part, thrilled by the news. Now you’d be able to go hunt for loot in Diablo III without the nagging feeling that you could just pay money for something better. Wrote one Kotaku commenter: “Good job Blizzard, you’ve actually restored some faith in me. I may actually play this again.”

“Diablo is best experienced when, as a result of slaying monsters, you get better items that allow you to make your character more powerful,” said Wyatt Cheng. “And if the activity that I do to make my character more powerful does not include killing monsters . . . then that’s not a good place to be.”

Now it felt like they’d found the perfect formula for Reaper of Souls. In addition to the new area (Westmarch), and boss (Malthael), the expansion would come with Loot 2.0 (free to all players via patch), Adventure Mode, and an overhauled difficulty system. The week before Reaper of Souls, Blizzard would remove the auction house. As they finalized the development of the expansion and prepared for launch, Mosqueira and his team felt like this was the big moment. They were going to win people back.

When Reaper of Souls launched on March 25, 2014, there was no Error 37. Blizzard had bolstered its infrastructure this time around. Plus, the company had decided to improve the error messaging system so that even if something went awry, the warning wouldn’t be so vague. “I think one of the other lessons that we learned is if you were there, anxious, and you logged in, and you got Error 37, you [thought], ‘What is Error 37? I have no idea what this is,’” said Josh Mosqueira. “Now, all the error messages are more descriptive. They say, ‘This is the problem we’re having. Here’s a time frame where you can expect this problem to be fixed.’”

As they watched the reactions pour in, Blizzard’s staff collectively breathed a sigh of relief. So did Diablo players. “Diablo III has finally rediscovered the moment-to-moment gameplay that made the series great,” wrote a reviewer for Ars Technica, “and fixed—or removed—almost everything that got in the way of that greatness. Reaper of Souls is the redemption of Diablo III.”

Two years after launch, people were finally falling in love with Diablo III. “When you went to the forums, or when you started getting feedback directly from fans, the issues they were complaining about were more specific and less broad,” said Kevin Martens. “That was when I really thought, ‘OK we’re finally getting there.’” As Martens and other designers browsed through Reddit or Battle.net, they were encouraged to see players complaining about underpowered items or asking Blizzard to buff specific builds. No longer were people offering the three words of feedback that serve as a death warrant for any game: “This isn’t fun.”

What was most gratifying for Josh Mosqueira was that people especially loved the console version of Diablo III, which launched on PS3 and Xbox 360 in September 2013 and on the newer consoles (PS4 and Xbox One) in August 2014. After decades of clicking, many gamers almost felt sacrilegious saying so, but playing Diablo III was more fun with a PlayStation 4 controller than it ever had been with a mouse and keyboard.

In the coming months and years, Blizzard would go on to release more patches and features for Diablo III. Some were free, like a dungeon called Greyhollow Island and a revamped version of the cathedral from the first Diablo game. Others cost money, like the necromancer character class. Although fans lamented the absence of another big expansion pack (which, as of early 2017, hadn’t happened), it was clear that Blizzard had committed to supporting Diablo III for years after its release. Other developers might not have bothered—especially after the launch day catastrophe. “Mike Morhaime, the president of the company, said to us, ‘We want to win and earn the love and trust of our players,’” said Wyatt Cheng. “We had all this work we had done in this game. We believed in the game. We knew it was great, and it would’ve been so tragic, I think, if we were at a company that said, ‘Oh, Error 37, pull the plug.’”

“It would’ve been so tragic, I think, if we were at a company that said, ‘Oh, Error 37, pull the plug.’” -Wyatt Cheng

Josh Mosqueira, on the other hand, was done with Diablo III. In the summer of 2016, Mosqueira left Blizzard, joining Rob Pardo, a veteran Blizzard executive and the lead designer of World of Warcraft, to form a new studio called Bonfire. “Leaving this team and this company was the hardest non–life threatening decision I’ve ever had to make,” Mosqueira said. “I felt that I wanted to take a chance and try to do something totally different.”

At least he’d left Blizzard with something great. Diablo III was one of the best-selling video games in history, having sold thirty million copies as of August 2015. It also proved a point that would influence numerous game developers in the years to come, including but not limited to the makers of The Division and Destiny (whom we’ll meet in chapter 8): every game can be fixed.

Often, the developers of a video game hit their stride toward the end of a project, when they learn what their game really feels like to play. For Diablo III and games like it, launching was just the beginning of the development process. “Even with a game that has a really strong vision, a really strong identity like Diablo,” said Mosqueira, “I think one of the challenges is that at the beginning of a project . . . before a game comes out, everybody has a slightly different version of the game in their head. One of the hardest things to do is get it out. But once it’s out, there’s less discussion about that because you can now [see] what it is. Game development’s really hard, but it’s a different type of hard before it’s released. It’s more existential before it’s released.”

Diablo III was proof that, even for one of the most accomplished and talented game studios in the world, with near-limitless resources to make a game, years can pass before that game properly coalesces. That even for the third game in a franchise, there are still an impossible number of variables that can throw everyone off. That even a game that launches with crippling issues can, with enough time, commitment, and money, turn into something great. In 2012, when Error 37 spread across the Internet, gamers had thought Diablo III was doomed. And then it wasn’t.

If you liked this book excerpt, don’t miss BLOOD, SWEAT, AND PIXELS, out on September 5 in print, e-book, and audiobook.