So how do you actually model and control that openness? Don't you get to a point where you're just creating more and more branches?

DB: That's not what we're doing. We're not looking at it as a branching story. We're looking at it as you're playing characters off against each other, and so that actually makes it harder. You can't just have, "Here's the roadmap".

So what we've spent a lot of time on is how the various things that are likely to happen: how they can be tested, what comes out of that, and that kind of thing. One of the first steps, which we did some time ago, was implement the story in a very simple interactive way. Not playing a game, almost more like a text adventure, where you could crawl through the story. We called it the Story Crawler, where you could say, "OK, I'll do that. I'll get that guy, and I'll do what he wants and see what happens then." And it's then "Urgh! OK. Then I'll get this guy, and I'll do that."

So you can see the interest, and in a way where you can play through hours of gameplay literally in a minute, just by clicking through. That gave us the confidence to verify the story, and it also allows us to see the extremely contrary player, like me, who will try and break it and do weird stuff.

So on top of that you're building, what, combat and stealth?

DB: Just. [Laughs.] Yes, that's right. That's the thing that drives all the high-level behaviors. But just testing and verifying that we can do it in a way that is going to work, at least to a degree that's interesting.

That's the nightmare, isn't it, where actually there's only one route through that's actually interesting. It's also trying to make sure that the player knows what's available. It's a world that you really do want to play in, rather than a world where you feel, "Oh, I don't care. Just give me the next thing to shoot."

The traditional way to handle open world games is, here's the world, and here are specific missions within that world -- is this the same thing?

DB: It can be as considered as that, yes. With anything, if you look at the cinematic parallels, whether it's 24 or whether it's some great battlefield thing, there will be things of interest -- explosions going off, an attack happening. There'll be some bad stuff that you'll know about -- hopefully multiple things.

The way The Outsider works, it's the choice of what to respond to, who to go to, who to listen to, who to work with, that tends to determine the flow. For example, if the Chinese agent comes up to you and says, "Hey, I can help you," you go, "Cool!" and he may actually give you some guys to help you go in and sort the problem out. And then you realize, 'Wait, what exactly are we doing here?' when they start breaking into the safe and nicking stuff!

Ian Bell has spoken about how, in making Elite, you accidentally made a very pro-Capitalist game. Have you thought the same thing?

DB: [Laughs.] Yes, it's interesting with Elite, for either of us, it was never intended as a pro-Capitalist game in any way. I think it might have captured a bit of the zeitgeist almost collaterally, but the way we looked at it is, I played around with some 3D before I met Ian -- a fighting game, where you get 10 score for the little red ones, maybe 20 for the little yellow ones. So you think, "I'll kill that one. 20. 40. 60, 100." Then you start to think, "What's the point of this?!" So we thought, "OK, what if the score instead is money, so you can buy stuff for your ship."

And then we thought, "What would you want to buy for your ship? Oh, bigger lasers. And let's give some of the baddies shields." And so you have this little arms race. Stuff on the trading and docking? That all mushroomed quite quickly from that. It wasn't that we sat down and thought, "Let's make a capitalist game -- no-one's done one of them." It's more that we wanted to make a great game that was evolving.

Essentially we were writing the game for ourselves. We were gamers, and luckily there were a lot of other gamers who fitted into that mould. I don't think Ian regrets it, but I do think the overtones -- that's not how it was intended. But in Britain, with Thatcher at the time, there were inevitably associations. She'd only just got into power around then. That certainly wasn't the intention -- it was much more the way of doing clever stuff with progression.