Almost a year ago I started work on a horror game. I suck at horror games. I can play for about half an hour of the first Dead Space game in one go before I need a dread break. I played 1.5 hours of Amnesia over the course of an entire year. I’ve had Alien Isolation since Christmas and I’ve got about 2 hours in because every time the alien kills me I rage-quit and have to go and calm down a bit. So you get the picture. Because I suck at them I don’t tend to play them very often, but weirdly I love the idea of playing them and am fascinated by them.
Our game (Late Night Shop) hinges largely on jump scares and tension at the moment. The gameplay relies on a simple mechanic: spooky mannequins that move when you can’t see them. This is actively terrifying for me as a developer because I have to test it. Every day. Even after a year it still scares the hell out of me, which is either a testament to how good the game is or further proof that I’m a massive wuss (maybe both?). Practically what this means is that I’ll be writing some code to make lights flicker, or to make some audio cue fire off at the right time, and I’ll have to psych myself up… to debug my game.
Ridiculous isn’t it?
I literally spend about 5-10 mins doing busy work before hitting the play button in my editor window sometimes. I mean I’m doing it right now. And that’s just for debugging. It’s a whole other story when I’m actually trying to complete a level. Palms get clammy, adrenaline starts pumping. Then one of those creeps jumps out at me from nowhere and I visibly jump, swear and quit/tear off the VR headset. It’s all very embarrassing.
So why am I making this game? Mostly because I find the challenge of scaring people an interesting one. Because I’m a huge wuss I happen to know all the right buttons to push for a lot of other wusses. My core drive behind the AI design is to make them seem to behave unpredictably, with the aim to have them corner you like a pack of velociraptors. The uncertainty of exactly how many are stalking you or where they are gets me almost every time. Because this happens organically, but to a point predictability, it feels like you’ve been outsmarted when they get you, and vice versa when you make it through. For instance, if you follow the same path at the same speed, the same thing will happen. The enemies will be in the same position at the same time. All of this means that if you try lot’s of different approaches (or are just lucky) you can thread the needle through the maze and to the exit without being seen. So for me, finding this critical path is the aim so as to avoid as many jump scares as possible, which is kind of how we want people to play the game.
The current AI patrols around a fixed area, tracking the player down if they hear or see you. One of my future tasks is to produce more intelligent AI that actively hunts the player through a given level, creepy Elmur Fudd style. Since deciding upon this task I’ve realised that I’m going to probably have to play some more horror games for inspiration… Wish me luck. (N.B. I’m planning to write a blog on my AI design soon for anyone who’s interested 🙂 .
Saying all this, developing this game is still a lot of fun. Watching it scare the hell out of other people is great, especially in VR. It’s somewhat reassuring to see people have similar reactions to me when playing it. Can’t shake the feeling that I should be over it after a year but hey, maybe by the time the game is done I’ll be some kind of zen horror master that feels no emotion and is immune to surprise.
I’ll leave you with some vines of other people who work on the game also being scared by it. Kinda glad I’m not the only one.