As development on LittleBigPlanet finally comes to a close, we sat down with Media Molecule co-founder Alex Evans to discuss how building a game that lets its players build their own games has changed the way he works, what happens to everything once it goes live, and why he wants the company’s first game to be unrecognisable at next year’s GDC.

LittleBigPlanet seemed robust even when we played it a while ago. Can you throw whatever you like at it without it falling over?

Not always, but everything in our game is user-generated, so even the developer content actually goes through the user-generated content pipeline, which means that literally everything you do is live. There’s no compile step, there’s no running it through an exporter – basically, the goal was that you’re editing with both hands on the controller. Our tools are completely different from a traditional game. Accidentally, by nature of being user-generated, we went a totally different route and it paid off by being this iterative thing in which I can literally load a texture in Photoshop and it updates live in the game while I’m drawing on the texture.

With that system in place, can you imagine making a game any other way now?

No. And the problem is that whenever it breaks it’s a bit like when the internet goes down and everyone goes: “What do I do now? I don’t understand!” And it’s the same thing with our game – a while ago we got a new SDK update from Sony and it broke part of our live update system. There was this guy going: “I can’t edit my level any more because it takes two minutes between making a change and being able to play with it – I can’t cope, I need it to be [clicks fingers] live”. That was really nice; it was a design we thought we were doing for the users, but it’s turned out to be great for us.

In terms of additional content via DLC, will you be able to ship new mechanical elements or will it be just new visual assets?

We can do both. This isn’t a promise that we will, but we can actually embed logic into objects after the fact. The example I always give is that if people really want fruit machines working, we can do that. I always thought we’d end up with more intelligent objects in the game than we have: fruit machines, guitars, that kind of thing. It’s turned out that our level designers prefer the basics and build the fruit machine out of switches and levers, which is mental to me. But if the community wants fruit machines we can put new gameplay logic into the game on a rolling basis. So that might be new gameplay modes or something, maybe something that changes the scoring. Texture packs are easy to do and it’d be criminal not to do them, but it’s things like… There is some good stuff coming, but I’ve got to bite my tongue.

The first thing you’ll want to do after launch, presumably, is turn away from the game and have a break – but you can’t.

No, you can’t. Luckily, I’ve made that mistake already. The biggest mistake I made in the industry was after I worked on Black & White and I said there was no way I would work on Black & White 2 for at least a year – and I ended up not working on it at all. What I should have said was: “We can do so much with Black & White – let’s do a PS2 version now”, but I was a young kid then, and I could have been bolshy and pushed things through. It’s getting to that mindset that it’s not a constraint to make DLC or anything. Quite the opposite: LBP is about the content people create, and I think that people will be surprised by what we can do and how far we can take it. I really don’t think it’ll just be stickers. Stickers are fine, and they will be there, of course, but I want to shock myself. I want to be at GDC next year when someone does a LittleBigPlanet reveal and no one knows it’s LittleBigPlanet.