How to push the limits of how much you work

A little over a week ago, I heard about the death of Monty Oum, the brilliant mind behind the animated show, RWBY. One of the things I’ve always admired about him is the drive and passion he showed towards animation, and he became famous for pulling 18 hour days of work. Roosterteeth encouraged people to dedicate themselves to creating something in remembrance, which made me resolve to attempt one of these 18 hour days myself.

The rules were simple: I had to work for 18 hours. This included no more than 30 cumulative minutes of breaks for cooking (read: microwaving) food, which I would eat while mashing at a keyboard. I had to work on development – coding which would push the project further to its completion. No marketing, no blog post writing, and no setting up any of the miscellaneous business stuff I have to do every now and again. I always aspired to applying myself to game development as much as possible, and pushing past an 8 hour day seemed to be a good first step.

To spoil the ending, I learnt a lot. I highly recommend it to anyone to try.

My expectations going into it was that this would teach me to be able to work for longer, and focus more, but that wasn’t the main takeaway here. I thought that by the end, every line of code would be a labour to write, but this was the easiest part. In fact, the hard parts were after I completed a feature or fixed a bug. If I could convince myself to push past the need to take a break in these in-between moments, I found within 5 minutes of starting to solve the next problem, I didn’t feel I needed a break at all, even after 17 hours of programming! This made me examine myself: What really was holding me back?

Well, I’ve always been aware of momentum, and the necessity to maintain it. Ostensibly, this is why it’s harder to start fixing a new problem than it is to continue working on the one you’re currently knee deep in. After a feature is finished, you cathartically take in your first vision of the feature as it might be in the final game. It’s a moment to catch your breath. But you lose momentum, and I previously would have recommended that you avoid focusing on this catharsis, so that you lose as little momentum as possible. But in my experience, it’s difficult not to revel in the glory of a well-done chunk of work, and as you rightfully should – it sets up an excellent and natural reward system; and I wouldn’t be the first to say that a reward system may help you out more than you may guess. So the problem isn’t momentum, or not exactly. The problem is urgency.

In an 18 hour day, there is no time for messing around. The moment I woke up at 9:10 am (which may as well be the crack of dawn for me), I knew that even if I started work right away I could only hope to finish at 3:00 am the next morning. So I got to it. There was never a question of what I’d do next, the answer was always: “Exactly as much as I need to do before I get back to making games again.” The situation demanded my attention because it was urgent, as if I didn’t give it my full attention, I would have to work for even longer.

This sense of urgency made fixing this problem of momentum so simple. Whenever I had that moment after finishing a block of work, there wasn’t anything wrong with playtesting and feeling good about myself. But I needed to get back to work – and to do so quickly. In a normal, not 18 hour day, I would recommend taking a ten minute break here, but too often in my normal routine would I get coaxed into taking unnecessarily long breaks instead, ranging 30 minutes to 2 hours, and just working for longer hours to compensate. Each 10-minute video seemed like it wouldn’t cut into development that much. But of course, it turns into a slippery slope. Things needed to be urgent for me to get past that – perhaps things are the same for you.

When I finished that day, I felt sore and tired, but I couldn’t believe how easy it was for me to do when I put my mind to it. It wasn’t climbing Mount Everest, and many people will scoff at my suggestion it was a trial in the first place. And I wouldn’t dream of doing it regularly – there’s just no time for sleep. But if you’re reading this right now thinking, “I couldn’t possibly do that,” try it. It doesn’t have to be every day of your life from here on out, but at least you’ll be able to say you tried. And you might even be able to say you succeeded.

Comments (2)

  1. Catreece

    The infamous “all weeker” from college; I once did a 160 hour work week (no, that’s not a typo) which involved snagging tiny little 30 minute naps on the floor every now and then, with food delivered to me. Unlike the 18 hour day, I don’t recommend it. =P

    Removing distractions is the single, most difficult part of working long hours as you’ve found out. The next biggest is forcing yourself to get started on stuff. I’ve missed several meals at a time before without even noticing the time going by since I was working meticulously on modeling in 3D; it’s a lot harder to do that now that I work on writing since there are more natural stopping points. Rigging models was the same way; there would be clear and obvious points where I could stop and I wouldn’t forget what I’d been doing, and that’s where the temptation to take a break seeps in.

    However… while one can work like that for a few days in a row, I’ve done so for weeks at a stretch, and I can warn you in advance that it doesn’t work that well long term. It’s great for a short term boost to work out a particular section which has to get done, but sooner or later the sleep debt will catch up to you and the quality of your work will begin to suffer. Little errors will begin popping up, things will take longer than normal to accomplish. Sure you’ll be working 18 hours a day, but it’ll be taking you 18 hours to accomplish what used to take you 5, which means you’re literally getting less total work accomplished per day despite working much longer hours.

    This is why “crunch time” should remain a short term burst only. Never more than two weeks absolute maximum. If you have to push crunch time beyond two weeks, then at that point you’re sacrificing more productivity than you’re gaining. Your employees become miserable, the quality of their work suffers, and by three weeks of crunch, the quantity has suffered so much that you may as well not even bother.

    Unfortunately, the sleep debt is not so easy to fix, nor is the burnout; a simple “here, have a week off to nap the whole time” doesn’t magically fix things since we’re not really built to work that way.

    I’ve known two individuals who have worked off of 2 hours sleep per day for years at a time. They don’t feel tired, and they still seem productive since it’s possible to train the body to pack all of the needed rest into a much smaller area, but the moment either of them had a full night’s sleep, it became obvious they’d only been working at a fraction of their actual potential. The extra hours per day really just didn’t do them justice in the long run.

    So what can we bring away from this in practice?

    If you’re doing simple, monotonous work, then it’s not that big of a deal to push an 18 hour day, or even longer. Simple tasks can be done even while unfocused and this can be a good thing. We all know that part in programming where you’ve solved all the problems, and now the key is to just copy over hundreds of repetitions of a concept but each one’s juuuuust different enough that a copy/paste won’t work, so you have to re-add large sections manually. It’s too obvious to you that it’s boring, but too complex for the computer to figure out, that little irritating niche. That’s the perfect point with which to employ a much longer day to just get it over with, especially if there’s no obvious stopping point until it’s done.

    When you’re trying to work on complex problems though, such as solving the programming issues in the first place? Don’t take the longer day. 14 hours absolute max for most people is the limitation there, any more and the capacity to focus and commit to complex problem solving will dissolve into almost nothing. You’ll still FEEL like you’re fine, that you can solve the same problems, but it really just isn’t so. I’ve worked on problems for 5+ hours while tired, but didn’t feel tired, and just ended up spinning my wheels pointlessly. Go to bed, wake up and bam, done in 20 minutes in a vastly more elegant manner.

    The point is, even if you don’t feel tired, it does still take a huge toll on your capacity for problem solving and intuitive thought processes. Things that should be simple become very difficult to connect the dots on. Some jobs can get away with it easier than others, and some will have almost no room for overly extended work hours without sacrificing more quality than it’s worth.

    In the end, there are a lot of lessons to take away as stated in the original article that can make your work more efficient even when not crunching.

    This is why most game companies provide free lunches at your desk; it’s cheaper to give you a sandwich and a coffee than it is for you to buy your own, leave your desk for an hour, then spend half an hour trying to get back into the groove of things again. Just keep in mind that longer hours do not necessarily equate to greater productivity, and the more complex your task is, the less efficient those extra hours will become. There is always an inevitable cutoff point where you’re wasting productivity by trying to work longer, than you would’ve gotten by just taking a break. Where that point is exactly is different for everyone, but it’s always there and it’s best to learn where it is so you don’t step over it. The fun part about that is it shifts about constantly even for the same individual based on a multitude of other aspects.

    Anyway, goodish article and I agree that it’s a good idea to push your limits to discover where they are, and how to be more efficient, just be careful not to get into the mindset that working more hours = more productivity in a linear fashion, because it’s really a severely diminishing returns formula, though it suffers a lower penalty to quality for the first few days.

    And again, to stress this very clearly to any management who blunders into this topic: overly long crunch time, as in anything past two weeks, will harm your team’s productivity vastly more than it will help. Don’t become one of those low-quality companies that tries to push all-year crunches, it not only makes you a lousy employer for what amounts to abuse to your employees, it also makes you a lousy employer for not understanding how biological efficiency works and wasting everyone’s time and resources.

    Reply
    1. Dylan (Post author)

      I completely agree – crunch has to have an end goal for the motivation to be there. What classifies as crunch will vary slightly between disciplines and people, of course; I’ve done some art asset production on projects in game jams before, and its much easier to push past those points. But programming is intellectual – it’s not like chopping wood. Contemplation is a massive part of a lot of aspects of game development, and programming is one of its greatest influences.

      I’ve seen a fair amount of talk recently which shifts focus from time management to energy management. I’ll likely make a post about my full thoughts on the subject later, but working longer isn’t the only way to be more productive – you should also be working smarter. And working long hours can mean you aren’t working particularly smart.

      To speak anecdotally, a lot of the proof is in the pudding. I know a fair few people who boast only requiring 4 hours of sleep, never spending a waking moment not working, but I don’t see them actually producing a significantly more impressive amount of work – and often quality will suffer alongside it. Contrary to that, some people will slowly work 1 hour a day for 3 years, but the end result of the project is brilliant because they were never rushing themselves.

      As you mentioned, forcing someone to work for longer never. Ever. Works. Every time I’ve seen someone work productively for very long hours, it’s because they have the passion and determination available to pour into the project. Even in that case, people get easily burnt out relatively quickly. It’s almost inviting someone to not foster that passion for a project when you bully them into crunching long hours. It shows that they don’t care about what you think, and is a symptom of a very brute-force professional relationship which raises a lot of red flags for someone who doesn’t have their job hanging over a flaming pit. If you’re lucky enough to have an employee who is willing to work those long hours and be productive during them, then guess what: politely asking probably would have gotten you further.

      This article wasn’t meant to encourage people to work 18 hour days – it was chronicling my experiences in something which inspired me to work for longer. I find the more I think of the work I do as work that I’m doing overtime for, the harder it is to push through with it. I’m still experimenting with the setup that’s right for me, and when I find it, I’ll make sure to share that as well.

      Reply

Leave a Reply to Dylan Cancel reply

Your email address will not be published. Required fields are marked *