Weather Factory

This is the side project I’ve been working on for the last few months. It’s really only an alpha: just a sliver of content, rough edges on the UI, some bugs. In an unusual commercial strategy, I’m going to halt development on it next week and not touch anything for some months. This is annoying for me: but I’m starting a hefty contract with BioWare in a week, and I wanted to get something out of the door to gather feedback from a wider audience before I enter the Veiled Temple of Edmonton.

Here’s some of my intentions and experiences with the project to date.

Polish and its Discontents: Cool, Gah, Pretty, Yeeesh

So far, the reaction to CS has been an equal mixture of ‘cool!’ vs ‘gah!’, with notes of ‘pretty!’ vs ‘eesh!’ I’ll talk about the cool<–>gah continuum in a bit, but a prefatory note about the pretty<–>eesh one: I think the game is a little too polished for the end to which I released it.

I released it to garner feedback on an experiment. But it’s got a splash screen, some fancy animations, and some very nice art. So although I warn people it’s an alpha, it doesn’t look very prototypey on first glance. Consequently, people go ‘pretty!’ but then when they run across a bug or a rough edge, they go ‘yeeesh!’ That’s my lesson learnt for the month: either leave it look unpolished, or go all in. Otherwise your alpha testers might get expectation whiplash.

What Walks on Three Legs in the Afternoon?

Cultist Simulator started as a Javascript prototype, which I got a lot of feedback on. I rebuilt it as a greybox Unity prototype that only a few people ever saw; and then I rebuilt it again in its current form (with the help of Martin Nerurkar and Catherine Unger, for UI work and for art respectively). The only bit that survives – almost unchanged – from the first prototype is the JSON notation for the content files.

So I’ve actually built this thing three times. Which sounds like a crazy waste of time, but has served a couple of really useful purposes.

First of all, I needed to thaw out my long-dormant coding skills – and learn Unity! I was a C# coder, but Unity brings a lot more to the table. The first iteration of anything I wrote was always going to need throwing away and rebuilding. I’ve still got plenty of blind spots and bad habits, but this is been a tremendously valuable re-education.

And secondly: I wanted to explore design ideas as much as I wanted to build a product. Anyone who’s built a game will know that even when you prototype and iterate, there are assumptions and decisions that survive iteration after iteration, like rats aboard a reconditioned and renamed ship, or unlucky builders accidentally bricked into a wall. Rebuilding meant I had to keep revisiting some of these decisions, and had the space to make them differently. Of course some decisions are just bricked right on in there anyway: that’s life.

Specifics, Kennedy! What are the things I wanted to experiment with?

The Magician and the Juggler

I wanted to work on a narrative, texty game that ran in real time. People try to do this every so often, and it doesn’t often work out well, for very clear reasons – text generally rewards time, patience, relaxed engagement.

But my last major project – Sunless Sea – used a mixture of static texty narrative where the game waited for input, and navigation where it occurred in (pausable) real time. It worked, but I was always conscious of the transition. I wanted to try to do something where the game allowed input, but didn’t wait for it.

So Cultist Simulator uses a resource-juggling mechanic to make the passage of time important. There’s always at least a couple of timers running – at a fairly leisurely pace, with a pause button, but there’s urgency. This meshed with a couple of other intentions.

Firstly, I’ve talked elsewhere about how I wanted to make a game based on looser choices. Making decisions about where to invest resources is a loose choice; making, or failing to make, a choice in a limited interval is also a loose choice.

Secondly, I have a long-term aesthetic interest in keeping writing as lean as possible. I don’t mean short – I mean no longer than is effective. Writing in games tends to expand to fill the space available. When you limit the time that a player has to read text, then there is a powerful compressing effect on what you put in those text boxes.

(It’s also immediately more difficult to cater to every player – some people read more slowly, often because they might not speak English as a first language, or they might have sight difficulties. A pause button helps with that, but of course if you encourage people to use the pause button all the time, you’re back where you started. I’m working on that.)

Thirdly, I wanted to provide a real sense that the player is balancing their occult studies with the needs of everyday life. A lot of famous occultists have died destitute. This is probably because occult powers don’t actually exist, but it’s appealing to imagine that their concerns are more rarefied, or they don’t manage to unlock the secrets of the universe before they run out of tea and soap. You can provide this balance in several kinds of traditional narrative game or interactive fiction, but it’s often necessary to foreground the consequences, or to leave them vague. Having those cards and timers show where you are at a glance makes the pressure obvious at all times.

(Assuming the game state is comprehensible at a glance! But that’s another issue.)

Life with Coffee Spoons

So some people, as I rather expected, can’t stand a game with that kind of constant pressure. A lot of my audience have come from Fallen London, after all, where time works the opposite way: wait longer, get more actions.

Other people really enjoy it. There is something about making half-understood decisions under time pressure that’s especially exciting. Games like Alpha Protocol and The Walking Dead have used this effectively in narrative; games like Crypt of the Necrodancer or FTL have used it effectively in strategy.

And other people need to be sold on it. The balance in the alpha is a typical early-build curve: it’s all a bit too mean in the beginning, and once you’ve worked it out it’s a bit too trivial (I haven’t implemented any of the content for menace states like Fascination and Despair and Unnatural Lusts yet – just Hunger and Sickness).

And, I’m glad to say, this is very encouraging. In 2017, the way forward for small indies – and I’m working solo with freelancers on this – is niche games. As Tom Bidaux says of Kickstarter: ‘It’s much better to have a game that some people love and lots of people hate, than it is to have a game that everyone likes’. If Sunless Sea hadn’t taught me that, Dark Souls would have.

Use Your Words

I haven’t talked at all about the actual setting, or the theme! What’s with that?

Well, first of all, at a talk recently I suggested that making Sunless Sea – a game where you take a small crew with limited resources out across dark waters, hoping to find something to sustain you on the other side – was influenced by running Failbetter, a studio that at the time was a tiny struggling venture with limited resources, embarking on a Kickstarter…

And now I’m a solo indie doing contract gigs from time to time, I’m making an experimental game about an individual on a quixotic quest for something they don’t understand, trying to balance their use of time carefully with satisfying their saner needs.

So there’s that.

And I wanted to build a new IP from the ground up, because that’s fun and it worked out very well last time. I have some thoughts about other projects that could use this stuff, and I wanted a setting full of mystery and implication and insinuation.

But also, I like games that balance horror and wonder. Cultist Simulator isn’t really Lovecraftian. I say that about all my games, though no-one ever believes me. It’s a game about tugging at the skin of the world to see what’s underneath, with no certainty about whether you’re going to find a thick stream of blood, or dawn-coloured light.

If that appeals, sign up below, and I’ll hit you up when I take the project further. Or there’s that link above, if you’re happy to put on your alpha trousers and see something really incomplete.

Sign up to hear more when it happens.