Alexis Kennedy on Cultist Simulator: the Way through the Wood, Part 6
When Lottie and I launched Cultist Simulator, it was a success. Not a giant success — not a headline-grabbing rags-to-riches success — but it sold about 5x as many copies as we expected; it got good reviews and some awards; and it meant we could afford to put a deposit on a house. For a studio of two people, this was an excellent result.
But I (and Lottie) have talked about all that before, at length. We’ve talked about what happened in the following year, and the year after that. We’ve talked about how we made Cultist, how we stayed within budget, how we tried to make an experimental but commercial game, how we tried to handle marketing. This Medium series— and the book they’ll go into — are intended as a window into independent videogame development, for people thinking about maybe doing that sort of thing, and people who are just curious. So in this post, I want to give you an idea of what it was actually like, making Cultist Simulator. And then I want to end with a warning.
Here are the things I spent my working hours doing between September 2017, when the Kickstarter ended, and May 2018, when we released the game:
25% sitting in front of a keyboard writing code (game code or the JSON config stuff used for Cultist content)
15% sitting in front of a keyboard writing actual narrative in actual English words
10% sitting in front of a keyboard planning stuff in a spreadsheet
5% planning on a whiteboard
2% pacing back and forth between the whiteboard and the keyboard or going out for a walk to think things over
2% reverting to Twitter because I couldn’t concentrate, either to talk to people or to post in-progress work or both
1% writing post-it notes with an idea I’m convinced is too important to be forgotten, and attaching it to monitor, desk, or whiteboard, where it will usually curl and fade until I sheepishly throw it away six months later
5% talking to Lottie
5% Talking to other collaborators (four different freelance artists, UI dev, composer, video editor…)
5% Biz stuff, tax admin, cash-flow projections (Lottie handles the bulk of this but I still have to get involved in a business this small)
5% Some sort of IT or devops issue, like a font inexplicably can’t be uninstalled or I don’t know what’s wrong with my Steam upload script
10% Community relations: support tickets, forum stuff, AMAs, press interviews, events and conferences (again, Lottie picks up most of this stuff but I can’t not be involved)
3% answering support tickets to beta players and messages from Kickstarter backers
2% writing a blog post that was meant to take half an hour and ends up taking most of a day
5% laundry, cooking, housework looking after a small child because she’s off school for some reason, cleaning out the gerbils, trying to convince Chi the ragdoll to shut the hell up
Let’s take the last thing first. I can’t emphasise enough how incredibly domestic the whole thing is. This is a picture of my desktop workstation, the one I use when I’m actually coding. It’s in the living room, but it’s got a river view and a fancy stand-up desk (with optional bar stool) and so this is the place I generally share pictures of.
However… I usually do the 75% of the work that isn’t coding in the spare bedroom Lottie uses as her office. Prepare to have any illusions about the glamour of indie development shattered forever:
So the rhythm of my days for three years has been something along the lines of ‘write something allusively melancholy, fix fiddly content bug, hang up washing.’ On the up side, this helps keep us both grounded. But on the other hand, there is a constant disorienting gulf between the moment-to-moment reality of working in a quiet flat with one person and two cats, and the awareness that at any given moment there are usually quite a lot of people playing our game.
This gulf was nowhere more disorienting than on the day after launch. I know I’m hammering this point home but it really does mess with your head. We went from quiet, if frenzied, work in our flat on something that lived on our PCs to seeing, in the first 24 hours, twenty thousand people download — and play — and talk about — and email us about — the game. That’s twice the population of my hometown, or probably about the number of people I’ve met in my whole life.
And as I said at the beginning, Cultist was a success, but not a giant success. It wasn’t a tenth as successful as FTL or Stardew Valley or Papers Please or an Undertale or Stanley Parable, let alone Minecraft— all of which were made by core teams* of only one or two people. But even at our level, something that started as a Javascript doodle when I couldn’t sit still has ended up letting us put a deposit on a house.
And this, of course, is why a noticeable number of successful indie developers start to believe their own legend; or become determined recluses; or flat-out lose their minds. Most indie developers were, basically, working on a laptop in an apartment when they made their first game. Jason Schreier’s Blood, Sweat, and Pixels has an endearing anecdote about meeting Eric Barone, the solo developer behind Stardew Valley:
When we’d finished packing, Barone opened the front passenger’s door and climbed to the driver’s seat. The door had been broken for months, he said. The car, a hand-me-down, had belonged to his family for nearly twenty years. I asked if he planned to get the door fixed. He said he hadn’t really thought about it […] On that Thursday in Seattle, as he climbed across the front passenger’s seat, Stardew Valley had already sold 1.5 million copies. Since he’d launched the game [six months previously], it had grossed close to $21 million. Eric Barone, who was twenty-eight years old and couldn’t open the front door of his car, had over $12 million in his bank account. And he still drove around town in a broken Toyota Camry.
Imagine yourself as you are right now, in your life, and imagine that tomorrow someone dumped twelve million dollars into your bank account and told you that a million people now know your name. It’s a nice problem to have, obviously. But how sure are you that you’d come out of that the same person you were when you went in? Especially if, like most indie developers, you’re a bit of an introvert?**
It’s natural to focus on the twelve million dollars. But the ‘million people now know your name’ can also be weird as hell. I’m not any kind of actual famous, but between Fallen London, Sunless Sea, Cultist Simulator, Horizon Signal and another dozen much less well-known projects, easily a couple of million people have read my work. A number of those people know my name, a number of them have had a personal or an emotional reaction to my work, a number of them have written kind or unkind things about me.
This is nothing, of course, by the standards of even a modestly successful musician or actor. But musicians and actors spend a lot of their time performing in visibly grander contexts or audiences as they get more successful. My day-to-day life — again, like that of most indie developers — is very similar to what it was eleven years ago, before I started all this. I still spend most of my day typing. I’ve been recognised in the street a grand total of twice. I just have more Twitter followers, and use a slightly fancier computer, than I used to.
And I’m not complaining! I am enormously delighted to be able to spend my days doing the kind of work I most want to do with the collaborator I most want to work with. But if there’s one thing that stands about the experience of indie game dev, it’s this weird and constant disjunction. I don’t think I’ll ever get used to it.
Here are some other things about the experience of developing Cultist Simulator. Some of them will sound familiar to other indie devs,
I made important bits of the game in places even less glamorous than our spare bedroom. I put together the first prototype of the research mechanic on a train with my daughter sitting opposite and occasionally kicking me in the shins. I made a major experimental change to the code that decides which cards go in which slots on on an Easyjet flight with my knees jammed up into my chin. I wrote a lot of the Glover & Glover content, including the notorious Mr Alden***, while I was sitting in the corner of a supermarket in Wapping.
Right up until launch day, and beyond, I was constantly convinced there wasn’t anything like enough content in the game. Once a week, on average, Lottie had to talk me out of adding huge new chunks of mechanics or narrative to the plan, or letting the launch date slip. (We didn’t let the launch date slip.) I’ve been making software for twenty years, and I’m not usually a total idiot. I know about scope creep. But when you’ve been face-first in a game for months and you’re terrified about how it’ll be received, damn me if it isn’t useful to have a producer to talk sense into you.
Of course, the game didn’t need more content. It probably needed 20% less content and 20% better balance. But writers always think the game needs more writing, just like coders swear the technical debt will kill the project and artists know their work would be better if they could spend another day finishing it. All of us are right and none of us are. “Art is never finished, only abandoned”: that’s not actually true. It’s just that the hard bit is deciding what to call finished.
Related: not once, not twice, but three times, I had to make what seemed like impossibly hard decisions about what to cut. I say seemed because I can’t now really remember what it was that I so desperately wanted to keep. I know we had to cut some of the Legacies, because it took longer to do interesting ones than I’d expected, and that was a shame. But most of the rest of it was never remotely as important as it seemed. Always the way.
Back to the list at the top. I probably only spent about 40% of my game development time actually developing the game. That’s probably not unusual for a small studio (by the time I left Failbetter, I was certainly spending less than even that on actual gamedev). Of course ‘actually developing the game’ is a treacherous phrase, because all that other stuff that doesn’t look like ‘actual’ game development still needs doing. One of the illusions around all this stuff is that games are made by someone sitting down in front of the Unity or Unreal editor and typing and then getting up when it’s done. You can, in theory, make a game by doing that and only that. But it’s a lot less likely to get finished, or polished, or released, or purchased.
Even experienced game devs aren’t immune to that illusion. So week after week, I’d be too optimistic in my estimates, because I couldn’t believe I would really only spend 40% of my time writing and coding. And Lottie would wrestle me down from ‘too optimistic’ to ‘maybe a little too optimistic’. And still, when I hit Thursday of any given week, I’d realise the week was nearly done and I’d only ‘actually’ done a day’s work. Then I’d see the deadline coming up like a distant tsunami wave, and I’d start to panic. Pretty much every week.
But I’m really glad we stuck to the deadline we did. Here’s why.
Shigeru Miyamoto, arguably the most venerated game designer in the world, once said:
“A delayed game is eventually good, but a rushed game is bad forever.”
Or did he?
I’ve been hearing this quote for years, and only today did I think to go looking for the source and context. I can’t find any. It’s everywhere on the Internet, unsourced. A Wikipedia editor suggests Miyamoto said something subtly different, but they can’t find the original source either. It does have the suspicious pithiness of an apocryphal quote — especially suspicious because the elegantly chiasmic rhetoric does sound very much like the kind of thing English speakers might expect a venerated, philosophically inclined Japanese game designer to say. And Miyamoto has worked as a producer as well as a game designer. He knows what’s what.
But, of course, even if he did say exactly that, a principle that’s good for the star designer of the most beloved game developer in the world creating flagship titles for major console releases… it’s not necessarily a good principle for a microstudio taking a punt on paying the rent.
If we’d spent twice as long on Cultist Simulator, it might have been half again as good. It would certainly cost twice as much. And it probably wouldn’t have made twice as much money. Worse: it might not have made any money. The launch could have been a disaster. The changes I made in the extra time might have made the game worse. Everything could have been very different. When you’re a tiny company, there’s very little you can control, but one of the things you can control is the amount of risk you want to take. It’s very easy to spend as much money on a video game as you can lay your hands on. It’s not nearly as easy to be sure you’ll make all that money back. People don’t want to talk about the failures, or to hear about them. If you’re thinking of making games, if you want to emulate Miyamoto, you should probably make sure you have Nintendo at your back.
That’s it on Cultist Simulator for now. Next time, I’m going to go all the way back to the very beginning of my career, and probably the most outright insane thing I’ve ever done.
*’Core teams’ because even in microstudio indie development, very few games are made exclusively by just one person — there’s generally at least a composer or a graphic designer or an additional coder or another freelancer involved. 90% of the work on Cultist was done by me or Lottie, but there are ten other people in the credits, and each of them made an identifiable and important contribution.
**I’m not throwing any kind of shade on Barone, who comes across as sane and likeable in Schreier’s account.
***I’ve heard speculation that Mr Alden was inspired by someone specific. If that were the case, I don’t think I’d be allowed to say who. Probably not even in the book.
Stay up to date with the latest updates by visiting Weather Factory or get connected on Twitter, YouTube, and Soundcloud.
