Alexis Kennedy on Cultist Simulator: The Way through the Wood, Part 1

On 25th May 2016, the day I left Failbetter, I made my last ever post on the company blog:

“[…] Failbetter is in an excellent place, and delights in an excellent team. Everything will be the same except that it will also be different. All of you are going to have a great time. Me, I’ll be walking the earth like Jules in Pulp Fiction walking the earth like Caine in Kung Fu. Oh all right I mean I’ll hang out my shingle as a freelancer. I have the entrepreneur’s bug and I’ll probably found another studio eventually, but I want to learn as much as I can from as many different kinds of project as I can, first.”

I liked to think that conveyed a suitably devil-may-care sense of romantic adventure, but honestly, I was terrified. I knew how hard life was for freelance writers. I’d spent plenty of time, latterly, on the other side of the desk, seeing the look of quivering wariness in the eyes of potential hires when they proposed a day rate. And I had a rep, but I didn’t know how much of a rep. I was very used to going to conferences and talking to mid-ranking exec types who’d never have heard of Failbetter, Sunless Sea or me. Most writers I met at conferences had heard of Failbetter — to the point where my feelings were slightly hurt if one hadn’t — but writers don’t usually sign the cheques.. I’d been my own boss for seven years, for good or ill, and suddenly I was dependent on the patronage of others. I had enough money to live on for a good while, but by no means enough to give up work — even if I’d wanted to.

It turned out okay. I was, indeed, a niche taste, but there were enough people out there who liked my work that I picked up offers fairly quickly. Paradox Interactive, whose strategy offerings included Stellaris — a game that I’d loved — asked me to write a guest story pack. And BioWare, the Dragon Age studio, wanted an extended engagement with me so I could write one of the factions for the new Dragon Age game. The idea of working inside BioWare as their first ever guest writer — the studio whose work had first inspired me to make narrative games, the Edmonton mothership itself — was a little dizzying.

But big studios move slowly. BioWare contacted me the day I announced I was leaving, but it took time to get everything in place, and the contract wouldn’t begin for another six months. Paradox moved a little quicker, but the wouldn’t want me for three months. I decided to take a month off, relax and catch up with my reading. I’d barely taken any holiday in years. It was about time.

I totally failed to take a month off. I just wasn’t built for not working any more — especially with the slow, chilling upwelling of the fear of possible irrelevance. I made it about a week before I cracked and stopped working again. I started doodling a project which I called, rather frivolously, Cultist Simulator.

It was the first time in perhaps five years that I’d written real code (with the daft exception of a Python script I’d put together to make it easier to create dividend certificates for shareholders). I was worried I might have forgotten how, and I was halfway right. It took me weeks to stop making the kind of silly mistakes I’d originally spent years learning not to make.

But once I’d got going, I understood all over again why so many indie game designers do their own coding. It’s exhilarating to have an idea in the morning and build it into the prototype yourself by afternoon. Even when it’s a bad idea. Especially when it’s a bad idea, because then you can take it out again the following morning, without needing to mention it to anyone.

Cultist Simulator was the culmination of four separate things I’d wanted to do at Failbetter, and never got a chance to.

The first was loose choices instead of tight choices. Think about an online multiple choice questionnaire. You wait at each question until you’ve made your choice, and then you move on. The questionnaire might be very sophisticated — it might change based on your earlier answers — but you only have one choice to make at a time. These are tight choices. That’s how most choice-based interactive stories work. You focus on one thing, you see the results, you think about the next thing.

Now think about playing a game like chess, or badminton, or Starcraft. At every moment there are a whole variety of things you can do, and many of those things you can also do a moment later.. Move your pawn now, move it next turn, move a completely different piece. In some games, ‘don’t do anything at all, just wait’ is also a meaningful choice. In Sunless Sea, there had been both loose choices — at sea, you could decide to turn in any direction at any moment — and tight choices, in our traditional beige interactive-questionnaire format. I’d realised that where the choices were looser, everything in the game remained relevant and interesting for longer. I wanted to take advantage of that.

The second: I wanted to make a game with mechanics of experimentation. I’d been briefly enthralled by Doodle God, a little mobile game where you mixed things to make more things and then mixed those more things to make yet more things. So you’d start with earth, water, air and fire, and then you’d find that earth and water together would make mud; air and fire made light; light and mud together made life. What had caught my attention is that combining human and death got you a corpse, and human and vampire got you vampire again… but you still only had one vampire, because the game wasn’t about resource management, just puzzle combinations. I’d been thinking about a game where you combined things to garner scarce resources, like an alchemist in a laboratory or an immortal experimenting with a society (get hold of a grand prophecy and combine it with an orphan to get a hero you can send to the underworld to get a secret truth which will be part of the recipe for a prosperous kingdom). I was particularly excited by the idea of using metaphorical or abstract ingredients in the mix, like grand prophecy or secret truth, much as I’d used Tales of Terror and Whispered Secrets as quasi-tangible goods in Fallen London.

The third: I wanted to make a game where part of the game was working out how the game worked. I’m going to admit here and now that this was partly a fancy formal constraint, and partly a sneaky way of getting around one of my long-term weaknesses. I’d never been good at elegance, I’d never been good at entry ramps, I’d never been good at tutorials, and plenty of people had bounced off Fallen London or, especially, Sunless Sea. I reckoned that if I could make obscurity a feature, not a bug, it’d mesh better with my natural instincts. And it would add a spice of wariness to the experimental mixing of ingredients: literally or figuratively, it could blow up in your face.

That started me thinking about the theme of the game, which had been unusually fluid up until now, and I realised that experimentation and mystery would work beautifully if the player was a student of the occult, of the ‘invisible arts’.

And this in turn dovetailed into the fourth thing I wanted to do: create a new setting for new games. I’d spent seven years working on Fallen London, and I was heartily tired of that. But I hankered badly for the opportunity to build a new mythology, revisiting some of the themes that had obsessed me from different angles; setting up new trails for players to follow and labyrinths for them to get pleasantly lost in. The approach of setting multiple games in the same elaborate world had worked really well at my last studio, and I wanted to repeat that trick.

My initial prototype was ugly. UI design has never been one of my strengths anyway, and I was trying to pick up web coding again after a long gap. The UI I built was grey boxes in a grey box with monochrome text, but even making that all line up the right way strained my rusty skills. But it worked, and it was instantly compelling.

What made it compelling? I had a half-dozen verbs the player could use — talk, dream, renounce, and so on — and objects they could be used on. So far that was rather like the traditional TALK TO GHOST and DROP CENSER of traditional parser-based interactive fiction, but the key difference here was that each of these actions would start a brief timer, and you wouldn’t see the results until the timer completed. So you’d begin a Talk action that would take thirty seconds to complete, and while it was completing you’d experiment with other combinations. Just as you had begun another action — or got stuck — an earlier action would complete, opening new possibilities.

So straight away there was a game loop that attracted players — try Talk with a Secret on an acquaintance, what did that do? It tries to RECRUIT them! What if you tried Talk with Sanity? it consumes your Sanity but allows you to TELL FORTUNES and gain Shillings. That experimentation loop was what had worked so well with Doodle God, the previous game of combinations I had enjoyed, but here was an extra wrinkle: delayed gratification.

This was more significant because resources in the game were so limited. You needed money, and money took time to earn. You clicked Work and Health and got funds, but only after a long delay. So you cared about that timer loop completing, but while it was ticking away, you had other things to do. So the effect, overall, was that you were always waiting for something interesting to happen — and you always knew exactly how long that something was going to take before it arrived. It put me immediately in a state of constant pleasant anticipation.

So the timers meant you were always waiting for something to happen, but what factors made it something interesting? Three factors: and fortunately, the second was something I was better at than I was at coding.

The first was simple: because the things you got were often limited resources, you cared that you got them. Health and Sanity were both resources, and if you ran out of them, the game was over. So an event that gave you Health or Sanity was always something interesting. A step further: I added a mechanic where you needed shillings to keep buying food, or your Health dropped. Immediately, anything that gives you shillings is as interesting as health.

The second was the possibility of other, different, somethings, for either curiosity of profit. If an interaction gave you an Atlas of Dreams or a murmuring amulet, that immediately suggested other possible interactions — that you could get something by studying the Atlas or selling the amulet. At its best, the game was a delightful whirl of experiments which gave rise to further experiments. At its worst, it was a struggle through a maze of trial-and-error, and a series of blind alleys and frustrations. I could already tell that finding the sweet spot between trial-and-error, and mystery, was going to be one of the big design problems.

The third factor for making a something into a something interesting was also something that I hoped would help fix the trial and error: words. Story, narrative, lore. Whenever a player examined an object, or got a result, the prototype would give them a pithy or enigmatic line of text. Try fasting, and you’d read “When your mouth waters, when your vision swims, you see the red cup; and now and then it will speak to you.” Combine Dream with Sanity, and you’d read “Last night, I dreamt again of the Wood behind the world, and the dwelling-place at its heart.”.

So, each piece of text was — hopefully! — a tiny reward in itself — long enough to engage attention, but not long enough to feel like homework. More than that, though, the snippets were fragments of a larger story. They also provided clues towards the underlying system of the game — trails of breadcrumbs that the player could follow.

I put the prototype online, and wrote a carefully casual blog post about it: just a doodle, bit of an experiment, not much there yet, see if you think it’s worth doing more on… I was still desperately apprehensive of the response to my first solo piece since Fallen London.

The response was good. It didn’t go viral or anything, but the kind of people who read a mid-ranking game writer’s blog were very enthusiastic and wanted to see more. And that’s all I wanted — not a proof of likely commercial success, just a sense that I wasn’t writing something so incomprehensibly niche that only I would play it.

So I started building the actual game. Still at a slowish casual pace, but now I had a task list and a rough plan, and now I was learning Unity, the game engine we’d used for Sunless Sea, probably the most popular one for indie developers. And now I was thinking about a Kickstarter. It would be a further test of people’s enthusiasm for the project; it would raise a little cash; it would get a little more attention.

The only problem was, I had paying work to do. The story pack I had to write for Paradox was coming up in October, my work with Bioware would begin in January, and I wouldn’t get much done over Christmas. My quixotic personal project would only get a couple more months before I had to mothball it. All the minimal buzz I’d got would fade.

So, I resolved to get an alpha version in Unity — something with actual graphics, something more than greybox, something that people could download to whet their appetite more convincingly — done in two months. I found a freelance artist to draw some basic art, and a freelance UI developer to put together a user interface that didn’t rely on my dreadful skills. I was persuaded that the game should look like a digital card game or board game — something abstract, playful, potentially compelling.

Two months was a pretty short time to put together even an alpha version, especially with the design still so vague and diffuse and my technical skills so limited. It was a pretty daft plan. Arguably it wasn’t really a plan at all. But I’ll talk next time about how it worked out.

Stay up to date on the latest by connecting on Facebook, Twitter, and Youtube