Q&A: Joar Jakobsson on Rain World

By now, if you've been reading my blog, you've seen my previous coverage of Rain World. The game is probably one of my most anticipated indies, besides SuperHOT, Distance, and Hyper Light Drifter. Joar Jakobsson - the man behind Rain World's concept, programming, art, and design - was kind enough to answer a few questions about the game and its development.
---

What inspired you to start a career as an indie developer?

I have always been fiddling with games on the side, but it has always been more of a hobby rather than anything else. I like the game medium because of the interactivity, it’s possible to build worlds that people can actually explore, driven by their own curiosity.

You’ve been developing Rain World for three long years. How does the current game compare to your original vision?

Hehe where do you put the line for “the original version”? I remember many different versions of rain world, both from when it was a few boxes moving around in a maze and later versions as well, that also looked quite different from what we have now, save from the main character. I think the Movement Prototype* I uploaded way back is still around on the internet, if you can find that you can try your hands at a very early version of Rain World. If it really was, though, the world of Rain World wasn’t really invented back then. Short answer - pretty much everything

In your TIGForum devlog, you’ve said that the art style was inspired by graffiti and old cartoons, but what were the inspirations for the world and gameplay?

Oh, that’s a hard one. The world was actually originally intended to have more of a resemblance to the sources of inspiration you mentioned, but drifted to something slightly more realistic. You can still see remnants of those styles though - the monochromatic palette of the levels comes from the old cartoon style, and many of the plants are supposed to borrow part of their shapes from graffiti ornaments. Gameplay has never really had those influences, inspiration for gameplay would rather be something like the harshness of nature.

As a fan of stealth games, I find the AI developed for the game quite fascinating. Was this focus on AI your intention from the beginning? Has it been difficult developing such a layered system?

I have always been interested in AI, so working that into the game has come naturally. Because the game always has had stealth elements, a somewhat decent AI was a necessity. I don’t really know if it was difficult - rather frustrating. Like, there is no mind-bending mathematics going on, just a lot of glitch fixing and endless iteration.

Given the reactive and adaptive nature of the AI, have the creatures ever behaved in an emergent way that surprised you?

Sometimes they do surprise me, but it's rarely mind-blowing because their actions are so limited. The lizards' choices in any given situation are basically all about where to move - the world doesn't allow for many more interactions than that as it's not a puzzle game with a lot of doors and levers and the like. I have tried to get a few interesting interactions in there though, such as lizards occasionally picking up objects and the like. I'd love to spend more time on AI interactions, but it might be a balance where you don't want to end up with the lizards doing so much other stuff that they don't have time to hunt you.

Currently Rain World seems to revolve around a simple gameplay loop: collect bats, evade lizards, return to shelter before the rain falls. Do you have plans to expand these mechanics, perhaps introduce new mechanics, in the full game?

Yeah, followers of the devlog will know about the pups. Basically the game will revolve around the cycle you mentioned, but when you find a few orphaned pups things will get mixed up a bit.

Reading through your devlog on TIGForums, it seems like the relationship between slugcat, lizard, and bat is core to the mechanics and game’s fundamental structure, and at this point, seems finely tuned and well balanced. How would the addition of new creatures and enemies affect and enhance this core system?

This trio of creatures will always be the core of Rain World, throwing in something else might wreck the balance. I do have ideas of other creatures, but they would not be as reoccuring as the lizards and flies. Rather they would be placed in special locations on the world map, and feature as occasional encounters, to spice up and bring excitement to exploring the world.

Also, was the focus on such a small number of creatures due to the limits of your current engine or just an extension of your vision for the game?

It’s sometimes difficult to keep apart limitations and vision, especially in game making where you can only get a good artistic result if you very consciously work with your technical limitations rather than against them. That said, there has never really been a technical limitation to the number of creatures, but it has rather been a question of development time. I early on decided that I wanted few, well made creatures rather than many species with less care given to each one of them.

---
You can support Rain World on Kickstarter, and follow its development on TIGForums. The game was recently Greenlit on Steam.

*Here is the download link for that early Movement Protoype (.rar)*
There are no enemies and you can't balance on poles, but you can climb around and see how the slugcat moves and animates. Player 1 controls are Arrows for movement, K to jump. Player 2 controls are ESDF for movement, Q to jump