Blood, Sweat, and Pixels Read online

  Dedication

  For Amanda

  Contents

  Cover

  Title Page

  Dedication

  Introduction

  Reporting Note

  1: Pillars of Eternity

  2: Uncharted 4

  3: Stardew Valley

  4: Diablo III

  5: Halo Wars

  6: Dragon Age: Inquisition

  7: Shovel Knight

  8: Destiny

  9: The Witcher 3

  10: Star Wars 1313

  Epilogue

  Acknowledgments

  About the Author

  Praise for Blood, Sweat, and Pixels

  Copyright

  About the Publisher

  Introduction

  Say you want to make a video game. You’ve got this killer idea—it’s about a mustachioed plumber who has to rescue his princess girlfriend from a giant fire-breathing turtle—and you’ve convinced an investor to give you a few million dollars to make it happen. Now what?

  Well, first you need to figure out the exact number of people you can afford to hire. Then you need to call up some artists, some designers, some programmers. You’ll need a producer to keep things running smoothly, and a sound department to make sure the game has, you know, sounds. Can’t forget to hire some quality-assurance testers to check for bugs. And a marketing savant—how else will everyone know about your future best seller? Once you’re all staffed up, you’ll need to make a strict schedule that determines how much time your team will spend on each part of the game. If all goes well, you’ll develop a demo for E3 in six months, then be “feature complete” by the end of the year.

  After a few months, things seem to be going well. Your artists are drawing all sorts of cool enemies for your plumber to fight: ghosts, mushrooms, that sort of thing. The designers have sketched out some clever levels that will guide the player through raging volcanoes and fetid swamps. The programmers just figured out a fancy rendering trick that will make the dungeons look more realistic than anything you’ve seen before. Everyone is motivated, the game is making progress, and you’re handing out stock options like they’re free newspapers in the subway.

  One morning, you get a call from your producer. Turns out that rendering trick is useless, because it knocks your game’s frame rate down to ten frames per second.* The playtesters keep getting stuck on the volcano level, and your marketing guy is grumbling about how that might affect your Metacritic score. Your art director insists on micromanaging the animators, which is driving them crazy. Your E3 demo is due in two weeks, and you know there’s no way you can get it done in less than four. And suddenly the investors are asking if maybe you can slash that $10 million budget down to $8 million, even if you have to let go of a few people to make it happen.

  A week ago, you were fantasizing about the speech you’d make at The Game Awards after taking home Game of the Year. Now you’re wondering if you’ll ever even finish.

  I once had drinks with a developer who’d just shipped a new game. He looked exhausted. He and his team had been right near the goal line, he said, when they were hit with a revelation: one of the game’s biggest features wasn’t actually fun to play. The developer’s team had to spend the next few months “crunching,” working eighty- to one-hundred-hour weeks to scrap the feature and overhaul everything they had done to that point. Some of them slept in the office so they wouldn’t have to waste time commuting, because every hour spent in the car was an hour not spent fixing bugs. Up until the day they had to submit a final build, many of them doubted they’d be able to release the game at all.

  “Sounds like a miracle that this game was even made,” I said.

  “Oh, Jason,” he said. “It’s a miracle that any game is made.”

  In the years I’ve been reporting on the world of video games, that’s been a common theme. Developers everywhere, whether at tiny independent studios or publicly traded corporations, talk frequently about how difficult it is to design and produce games. Walk into any San Francisco bar during the annual Game Developers Conference (GDC) and you’re bound to find groups of weary designers trying to one-up each other with stories of coding binges and caffeine-fueled all-nighters. War metaphors are common—“tales from the trenches” is a ubiquitous expression—as are choruses of complaints that the outside world doesn’t get it. One surefire way to annoy a game developer is to ask, in response to discovering his or her chosen career path, what it’s like to spend all day playing video games.

  But even if you accept the premise that video game development is grueling work, it’s not easy for those of us on the outside to see why. People have been making games since the 1970s, haven’t they? With decades of lessons and experience to draw from, shouldn’t game development have grown more efficient? Maybe it made sense for developers to crunch in the late 1980s, when gaming was the domain of teens and twentysomethings who gorged on pizza and Diet Coke as they coded all night and slept all day, but decades later, video games fuel a $30 billion industry in the United States alone.* Why do game developers still have so many stories about staying at the office until 3:00 a.m.? Why is it still so difficult to make video games?

  To try to answer these questions, I went out and engaged in my favorite activity: bothering people who know way more than I do. I spoke to around one hundred game developers and executives, both on and off the record, asking them an endless number of nosy questions about their lives, their jobs, and why they sacrifice so much to make video games.

  There are ten chapters in this book, each telling the story behind how a different video game was made. One chapter visits Irvine, California, for a look at how the Kickstarter-funded Pillars of Eternity helped Obsidian Entertainment recover from its darkest days. Another chapter heads to Seattle, Washington, where the twentysomething Eric Barone shut himself alone in a room for nearly five years to create a serene farming game called Stardew Valley. Other chapters tell stories about Dragon Age: Inquisition’s technological nightmare, Uncharted 4’s brutal time crunch, and even the mystery behind LucasArts’ much-hyped Star Wars 1313.

  As you read, many of these stories may strike you as anomalous—games that were hampered by drastic technology changes, directorial shifts, or other wild factors beyond the developers’ control. It’s tempting to think, while reading these stories, that these games were all made under abnormal sets of circumstances. That these people were just unlucky. That the developers of these games might have averted hardship if they had followed industry standards and avoided common pitfalls, or if they had made smarter decisions from the get-go.

  Here’s an alternative theory: every single video game is made under abnormal circumstances. Video games straddle the border between art and technology in a way that was barely possible just a few decades ago. Combine technological shifts with the fact that a video game can be just about anything, from a two-dimensional iPhone puzzler to a massive open-world RPG with über-realistic graphics, and it shouldn’t be too shocking to discover that there are no uniform standards for how games are made. Lots of video games look the same, but no two video games are created the same way, which is a pattern you’ll see throughout this book.

  But why is it so hard to make them? If, like me, you have never tried to develop a commercial video game before, you might find it helpful to review a few possible theories.

  1. They’re interactive. Video games don’t move in a single linear direction. Unlike, say, a computer-rendered Pixar movie, games run on “real-time” graphics, in which new images are generated by the computer every millisecond. Video games, unlike Toy Story, need to react to the player’s actions. As you play a video game, your PC or console (or phone, or calculator) renders cha
racters and scenes on the fly based on your decisions. If you choose to walk into a room, the game needs to load up all the furniture. If you choose to save and quit, the game needs to store your data. If you choose to murder the helpful robot, the game needs to identify (a) whether it’s possible to kill the robot, (b) whether you’re powerful enough to kill the robot, and (c) what kind of awful sounds the robot will make as you spill its metallic guts. Then the game might have to remember your actions, so other characters know that you’re a heartless murderer and can say things like, “Hey, you’re that heartless murderer!”

  2. Technology is constantly changing. As computers evolve (which happens, without fail, every year), graphic processing gets more powerful. And as graphic processing gets more powerful, we expect prettier games. As Feargus Urquhart, the CEO of Obsidian, told me, “We are on the absolute edge of technology. We are always pushing everything all the time.” Urquhart pointed out that making games is sort of like shooting movies, if you had to build an entirely new camera every time you started. That’s a common analogy. Another is that making a game is like constructing a building during an earthquake. Or trying to drive a train while someone else runs in front of you, laying down track as you go.

  3. The tools are always different. To make games, artists and designers need to work with all sorts of software, ranging from common programs (like Photoshop and Maya) to proprietary apps that vary from studio to studio. Like technology, these tools are constantly evolving based on developers’ needs and ambitions. If a tool runs too slowly, is full of bugs, or is missing pivotal features, making games can be excruciating. “While most people seem to think that game development is about ‘having great ideas,’ it’s really more about the skill of taking great ideas from paper to product,” a developer once told me. “You need a good engine and toolset to do this.”

  4. Scheduling is impossible. “The unpredictability is what makes it challenging,” said Chris Rippy, a veteran producer who worked on Halo Wars.* In traditional software development, Rippy explained, you can set up a reliable schedule based on how long tasks have taken in the past. “But with games,” Rippy said, “you’re talking about: Where is it fun? How long does fun take? Did you achieve that? Did you achieve enough fun? You’re literally talking about a piece of art for the artist. When is that piece of art done? If he spends another day on it, would that have made all the difference in the world to the game? Where do you stop? That’s the trickiest part. Eventually you do get into the production-y side of things: you’ve proven the fun, you’ve proven the look of the game, and now it becomes more predictable. But it’s a real journey in the dark up until that point.” Which leads us to . . .

  5. It’s impossible to know how “fun” a game will be until you’ve played it. You can take educated guesses, sure, but until you’ve got your hands on a controller, there’s no way to tell whether it feels good to move, jump, and bash your robot pal’s brains out with a sledgehammer. “Even for very, very experienced game designers, it’s really scary,” said Emilia Schatz, a designer at Naughty Dog.* “All of us throw out so much work because we create a bunch of stuff and it plays terribly. You make these intricate plans in your head about how well things are going to work, and then when it actually comes and you try to play it, it’s terrible.”

  In all the stories in this book, you’ll see several common themes. Every game is delayed at least once. Every game developer must make tough compromises. Every company must sweat over which hardware and technology to use. Every studio must build its schedules around big trade shows like E3, where developers will draw motivation (and even feedback) from throngs of excited fans. And, most controversially, everyone who makes video games has to crunch, sacrificing personal lives and family time for a job that seems to never end.

  Yet many of the people who make video games say they can’t imagine doing anything else. When they describe that feeling of being on the cutting edge of technology, of building interactive entertainment that’s unlike any other medium, of working with a team of dozens or even hundreds to create something that millions of people might play, it all adds up to an unshakeable belief that for all the turbulence, for all the crunch, for all the unadulterated bullshit that developers often have to go through, making video games is worth it.

  So about your plumber game (Super Plumber Adventure). It turns out there are solutions to all your problems, though you might not like them. You can cut your budget by outsourcing some of the animation work to a studio in New Jersey. It might not look as good, but it’ll be half the price. You can ask the level designers to add some extra platforms to the volcano level so it’s a little less punishing. When they object, remind them that, hey, not everybody loves Dark Souls. And you can tell the art director that the programmers have their own jobs to do and really don’t need to hear his opinions on chiaroscuro in video games.

  Hitting that E3 deadline might be a little tougher, but what if you ask your employees to stay late for a couple of weeks? No more than two weeks, of course. To make up for it, you’ll buy them dinner, and maybe even offer them nice bonuses if the game gets a 90 on Metacritic.

  You’ll also have to cut out a few features. Sorry, sorry. I know—they were awesome. But it’s not like your plumber really needs to transform into a raccoon. You can save that for the sequel.

  Reporting Note

  The stories in this book are based on interviews I conducted with roughly one hundred game developers and other industry figures between 2015 and 2017. Most of those people spoke on the record. Others spoke on background, requesting anonymity because they were not authorized to participate in this book and did not want to risk damaging their careers. You’ll probably notice that most of the people who speak in this book are male, which is a depressing (and unintentional) reflection of an industry that has, for decades now, been dominated by men.

  Except when otherwise noted, anything you see in quotation marks was said to me directly. This book does not contain re-created dialogue. All the anecdotes and details in this book come directly from the accounts of sources, and all were corroborated by at least two people whenever possible.

  Some of this reporting was based on trips to studios and homes in Los Angeles, Irvine, Seattle, Edmonton, and Warsaw. I paid for my own travel and did not accept accommodations from any companies or developers, although I did say yes to a few lunches, which seemed kosher. Well, the lunches weren’t kosher. What I mean is that it seemed ethically accep—You know what, let’s just get on with the book.

  1

  Pillars of Eternity

  The most important question in video game development has nothing to do with making video games. It’s a simple question that has stymied artists for centuries and put an end to countless creative endeavors: How are we gonna pay for this thing?

  In early 2012, Feargus Urquhart, the CEO of Obsidian Entertainment, found himself unable to come up with an answer. Obsidian, a relatively small game development studio based in Irvine, California, had spent the past year working on a fantasy role-playing game (RPG) called Stormlands. They’d never made a game quite like this. It was weird, ambitious, and, most important, funded by Microsoft, whose game producers had decided they wanted a massive, exclusive RPG to launch alongside their next console, the Xbox One. Around fifty of Obsidian’s ~115 employees were working on the game, which meant it cost a lot of money. Which was fine—as long as Microsoft kept sending checks.

  Urquhart had grown accustomed to financial pressure. He’d been working in games since 1991, first as a tester at the publisher Interplay, then as a designer, then as the boss of the powerhouse developer Black Isle Studios, which a suddenly cash-strapped Interplay dismantled in 2003. That same year, Urquhart founded Obsidian with a few other Black Isle veterans, and they quickly learned that running an independent studio was like juggling dangerous objects. If they didn’t have a new contract ready to go as soon as the last one had finished, they were in trouble.

  Urquhart has thin, light bro
wn hair and a stocky build. He speaks quickly, with the authoritative tone of someone who’s spent decades pitching games. Over the years he’s played a part in creating some of the most beloved RPGs in gaming, like Fallout and Baldur’s Gate. While speaking on panels and to the press, he’s always candid about the difficulties of running an indie studio. “The life of an independent developer is, every morning you wake up wondering if your publishers are going to call and cancel your games,” Urquhart said. “I wish I was more of a sociopath or psychopath and I could just completely ignore the fact that there’s this guillotine pointed at my neck all the time. But I can’t. And I don’t think a lot of developers can, and unfortunately there’s that constant threat, and that threat is used a lot. It’s used all the time.”

  On the morning of March 12, 2012, Urquhart’s cell lit up. It was the Stormlands producer at Microsoft, texting to ask if Urquhart could hop on the phone. Right away he knew what was about to happen. “It’s the text from the girlfriend who’s going to break up with you,” Urquhart said. “I just got on the call knowing what it was going to be.”

  Microsoft’s representative was frank: they were canceling Stormlands. Effective immediately, fifty of Urquhart’s employees no longer had work.

  The producer didn’t say why Microsoft was axing the game, but it had become clear to Obsidian’s top staff that development was not going smoothly. There was an inordinate amount of pressure, not just to make a good RPG, but to make a good RPG that could sell Xbox Ones. Stormlands’ ideas had felt disjointed to Obsidian’s staff, and, at least from Obsidian’s point of view, Microsoft’s expectations were impractical.

  As some former Stormlands developers described it, the game was full of ambitious “high-level” ideas, many driven by Microsoft’s vision of what should be in an Xbox One launch game. The Xbox One was all about the Kinect, a motion-sensing camera that could recognize full-body gestures, so what if Stormlands let you use Kinect to haggle with shopkeepers? The Xbox One would support “cloud computing,” allowing each player’s console to interact with Microsoft-owned computer servers, so what if Stormlands had big, massively multiplayer raids that would let you meet up with other players on the fly? These ideas all sounded interesting on paper, but it was unclear whether any of them would work in a game.