I have been working since 9:30 on a bug that I inherited from a senior colleague, who we'll call Mike. He made this creature two years ago, but I am hoping he remembers how it works. Unfortunately, he's not in the office this morning. His chair is empty; his screens are dark. Nobody mentions it at the morning meeting.
After lunch, Mike trudges in.
He's worked here for a year longer than I have, and he's been in the industry five years longer than I've worked anywhere. This is my first Real Job, and I'm trying very hard to be the best Junior A.I. Designer the world has ever seen. The game actually launched a few months ago, but because this is 2008, I'm working on an MMO and the so-called "Live Team" anticipates updating the game for a decade to come.
Mike sits down with coffee in hand, waiting for his computer to boot up. I smile in greeting. He shrugs.
"Long night?" I ask.
"Oh, not too late. Like 2. But I had trouble sleeping."
Mike's lips jut forward in an unconscious pout, but they barely move when he talks, which is in a lower, more gravelled register than I remember last week.
"Sorry to hear it," I say.
"It's not my fault. Have you not seen my sprint? It's ridiculous—they want me to pull this feature out of my ass in two weeks, without any help."
"Did you tell them it's not possible?"
"No. Then I'd look lazy. I just have to do it. Besides, it is possible. It'd be easier if someone would help me, though."
I look away. I have plans tonight.
"It's not so bad, you know," he says, suddenly upbeat. "At least it's not like launch."
Feeling the Crunch
I've decided "crunch" refers to the feeling that descends after three weeks of bad sleep and staring at an LCD screen more than 12 hours a day—the feeling that everything outside your body is made of grey fog, and your real self has calcified into a crumpled wad shoved deep down into your abdomen.
In case you're not aware, science has found again and again that crunch is harmful. It's a lifestyle that is less likely to welcome people with dependents, disabilities, or diverse life passions. Over 45-hours of work for over 2 weeks certainly harms the quality of games. Besides, it's inefficient. Some studies point to 35 hours per week being the maximum one can work at peak productivity. One study found that eight 60-hour weeks produces as much as eight 40-hour weeks, yet game developers swap stories of 80-hour-per-week months.
So why do we do it?
The standard story of crunch is predatory, and not unique to games. An exploited worker, pushed to their limits, spirals into burnout or divorce or depression or custody loss. In the short term, the game is made, and in the long term, the worker leaves the industry. Then we all blame big companies for yet another victim.
It's a familiar story, and my sympathies are offered to every worker who has been squeezed for profit against their will. I wholeheartedly support labor protection efforts to make these kinds of practices illegal in more countries, for software development and other industries. But this story of victimization is not accurate for a large portion of the games industry, and pretending that it is the only cause is genuinely hurtful, because we ignore the real reason that many people voluntarily choose to crunch, beyond basic job security.
"He Needs the Sleep"
It's Friday night, 7pm. He's leaving for the weekend. They've been getting their game ready for console certification, and it's been a few hard months, but they submitted today. This is one of the first weekends in a while in which they are almost guaranteed to not be working on Saturday.
The office is lively. Most computers are still on. A colleague pours from her personal bottle of scotch into a few glasses. He declines, feeling like a spoilsport.
On the way out, he spots a colleague slumped at a desk. A box of takeout sushi ("overtime food") sits in front of him, the chopsticks primly stacked.
"No, don't wake him," someone says. "He has two toddlers at home. He needs the sleep."
The truth is that many game developers crunch ourselves. We choose the risks of overwork over the risks of idleness. We are conspirators in our own self-destruction, whether we work for big companies or we work for ourselves. Even as we say we disapprove, we also tend to say it is necessary, or unavoidable, or some other word that absolves us of our decisions.
Even as a regret, crunch is comforting.
It's comforting in at least the following 10 ways, none of which involve obedience to authority, public recognition, promotions, raises, or fear of unemployment. They are self-inflicted, as much as any cultural value can be, and I understand them because I have felt them too.
I share them with you now in the hopes that we can identify the causes of a serious problem and move towards accepting that even if overwork is natural in our culture, it is not inevitable.
It takes a tremendous amount of courage to admit that maybe, just maybe, you made a mistake that cost you hundreds or thousands of hours without any benefit. Instead, especially in the reeling emotional aftermath of a game launch, it's much easier to believe that there was value in destroying yourself and statistics must just not apply to you. I am not picking on McMillen here -- I have never met him, but he seems to be a smart, talented man. More importantly, his after-the-fact justification of self-destruction is not unusual among my many smart, talented colleagues.
Perfectionism: Software, much like any art form, is never finished. Every creator must decide when to stop creating. It's even more difficult to know when to stop when your creation is an abstract, intellectual experience that's different for every person who interacts with it. It's comforting to work on making your game just a little bit better. It might not be better in any measurable way, but it feels good. 8.
Camaraderie: When someone shares your values and endures adversity with you, over hundreds or thousands of hours, a strong bond can form. Some even believe loyalty during hardship is one of the only true tests of friendship. Small wonder, then, that when 10, 50, or even hundreds of people are all crunching together, it can become a source of genuine joy and companionship. If we can complain about our evil boss as a common enemy, so much the better. After crunch ends, it's common for people to mention missing those late nights together, with a tinge of bittersweetness. 9.
Learned Helplessness: After you've worked on a few games and perceived a pattern of "naturally occurring" crunch, regardless of what company you work at or what game genre it is, it's difficult to believe it can be any other way. You start to think that It just "always happens" and there's nothing to be done. 10.
Culture: At some point, learned helplessness evolves into a superiority complex, stating that crunch is simply how good games are made. After all, if Uncharted 4, Fez, and Fallout 4 used crunch in their process, who are we to question? Warren Spector is one of my heroes of game design, yet he has endorsed crunch as necessary.
The good news is that some game developers don't crunch. We might think they're lucky, or even think it's limited to bad games, or games with dispassionate creators. But that would be incorrect! Award-winning, profitable, beautiful games like Invisible Inc, Don't Starve, Canabalt, Hundreds, Regency Solitaire, Monster Loves You! and many others actually didn't crunch.
How do they do it? As the leader of Kitfox Games, I want my team to be at maximum efficiency and efficacy, so I want to minimize crunch. I asked my betters how to do it. Jamie Cheng, CEO of Klei Entertainment, told me that he changed his company's culture. After the difficult development of Mark of the Ninja, they looked for funding without deadlines, invested in experienced developers, and built for platforms that allow updates (tolerating an imperfect launch). The fourth step? "Believe you can do it."
Only a year later, Klei shipped Don't Starve with "almost no overtime, and certainly nothing sustained."
This year, the International Game Developer's Association announced intentions to reward companies that produce great games without crunch. Maybe soon, we can collectively admit that game development without crunch is not only possible, but admirable.