An Audience with Soren Johnson

1 Comments

Soren Johnson is best known for his work at Firaxis, where he's been deeply involved in the Civilization series since its third incarnation. He was Project Lead on the definitive Civ IV, while still coding all the game's artificial intelligence. Now he's moved to Maxis to join Will Wright working on the cutting edge of PC games design. And he's knows a lot about post-punk records. Clearly, it was high time we had a chat with him.

Which landmark games inspired you?

Soren Johnson: The one which stood out for me was Seven Cities of Gold. And I know there's a number of other people who say the same thing. At that point, we were so used to the repetitive stuff you saw in arcades, and here was this game which would create a random new world for you every time... when you were creating a continent, it actually took four or five minutes.

It took so long your imagination was thinking... Wow! What's going on inside my computer? And then it spits out this huge world for you to explore. The idea that there could be so much inside your computer, which could be so different and it wasn't just randomly scattered. It looked intelligent... That really opened my eyes.

Fantastic. What else?

Johnson: Sid [Meier] had a string of three games, which were hugely important in terms of making me think about what topics for games and how dynamic they could be. They were Pirates!, which he says was his response to Seven Cities Of Gold.

Then Railroad Tycoon, which showed how fun it was to run a business... That game was all about planning, these overlapping goals. And Civ just took it to the next level, with probably the best topic I could imagine for a game. That is, all of the world's history. And obviously SimCity was a huge one. I played a ton of that on my Mac.

Recently I've been thinking about how history could have been different if certain things never happened. Imagine a world without Civ...

Johnson: I remember when it came out. The thing that was funny to me is that there was no sense of surprise. It was obvious that someone should make a game like this. Railroad Tycoon had come out. SimCity had come out.

Populous had come out. Someone was going to make a great game about the history of the world. If Sid hadn't made it, it would have gotten made. Sid did it brilliantly of course. [But] he wasn't the only person working on it. Chris Crawford was working on his own game - Guns and Butter, I think - and Dan Bunten, the guy who did MULE and Seven Cities, was working on his own idea.

And these are great game designers... but Sid's the one who managed to work out how it worked. To me, it's interesting that while it's clear that game was going to happen, a lot of the specifics of how the game played out depended on what Sid decided to do... which means there must be other ways to do a history of the world.

Why make games at all?

Johnson: It's a field where you're writing the rules right now. Some day 100 years from now, they're going to writing about the stuff we do now, because this is the crucial moment for games. Beyond that... well, 100 years ago, if I'd been born, I think I might be making board games. It's not just games for me - I come from a real board game, strategy game backdrop.

This is what I'm about. I feel that games are such a broad category. You can do so much with games. People put it up and compare it to... well, are games like music or movies or books? I see games not like a new medium, but a new way of communicating - a new language, so much broader than a specific artistic medium.

It's so fascinating to work on. Your imagination keeps on rolling when you're dealing with games. In all the other media, you feel as if you're eventually going to some kind of limitation, but with games there's no idea that's so far off the wall that you don't think "Hmm, I guess we could make that work some way or another."

Games kind of hark back to the days before the schism of art and science, in that technological progress can also be artistic progress...

Johnson: I know if I was around 200 years ago... How cool it'd be to have these great scientists who are really into music and whatever, and you could actually have most of modern knowledge in your brain at one time. 200 years ago, that was theoretically possible and is a neat idea.

I find a lot of game designers just have a ravenous appetite for stuff. Will Wright is the classic example. There's nothing which doesn't interest him in some way.

And this is another thing I really like about writing about games, especially games which aren't about some made up fantasy world... when I was working on Civ, there's literally nothing I can do or experience or learn which doesn't relate somehow to my job.

You majored in History. That ties in with Civilization too.

Johnson: It's just very interesting to me. Here's history... and here's this new language of interactivity. Can this be combined in an interesting way? Is this the way to jump ahead or to the side of this giant long tradition of history and prose? I found that very interesting. I used to find it a lot more interesting than I do now.

The more you get into designing games, the more you find the medium and language has huge possibilities, but also has specific limitations. The entire idea of player agency means certain topics aren't going to be appropriate.

For instance, in world history, one of the most important books of the last 10 or 15 years is Jared Diamond's Guns, Germs and Steel... He's saying that all history is determined by geography. I read that before working on Civ III, and was all pumped up - there's all these great concepts which you could put to work in a macro world-history game.

The thing is, if you make a Civ game based off the ideas in Guns, Germs and Steel, it'd suck. The whole point is that there aren't choices that determine whether a civilisation does well or poorly - it's whether you have the right crops. Do you have the right animals? Are you in the right place? In early versions of Civ IV, we tried.

Horses will always be on one continent and not another. We'll show people how this stuff works. But it didn't work in terms of gameplay. It felt unfair.

You've done a lot of work on AI in your games. I think it's a topic which gamers tend to misunderstand, myself included.

Johnson: I think when talking about games in general, it's important to differentiate between when you're talking about an abstraction game or a real world game. Because the two get mixed up a lot. When you're talking about a real world game where you're moving around some kind of environment, you have an expectation that things will function.

Straight human behaviour stuff. You can get a long way with convincing animations and having the guys say smart things at the right time, but for the most part, these are all deeply asymmetrical games (ie, games where the AI isn't bound by the same rules as you - Ed).

This is why no one is interested in playing symmetrical shooters that aren't multiplayer. But when you're talking about these real world games, if it's an asymmetrical situation... the issue with AI isn't one of challenge. It's one of: do they act like people? Do they run away if they're scared? Do they get stuck on things?

But it's different for a game like Civ, yes?

Johnson: If you leave the real stuff behind and go into abstraction, then it's a totally different ball game. What I've worked on has always been these symmetrical strategy games. To me, the interesting thing there isn't so much the AI questions, but the game design questions.

When you're making a multiplayer game, when you make a design choice, all you have to worry about is whether it's a feature humans will have fun with... But a game like Civ is multi- and singleplayer. You ask whether the feature is fun, but also whether it's something the AI is ever going to understand. Because if it's hopeless at it, it's pretty useless.

Finally, what attracted you to Maxis and Spore?

Johnson: Spore is a hugely important project. It's really at the forefront at a lot of progressive ideas about game. Just being smart about connectivity - just because a game isn't MP, doesn't mean that you can't do cool things to connect people in the game.

Letting computers do what they're good at - that is, procedural generation. This is something that, in many ways, the games industry is heading down the wrong path and trying to do things with brute force.

Almost any game I work on, I want to have a significant dynamic element where we're using simple rules to create a great deal of variety. Spore obviously is pushing that a great deal. And, of course, it's really cool to work with Will Wright and the team.

Kieron Gillen