So when did you realize that it was anything but a niche game?
WW: I think our first big break was about three months after release in, I think, February 1989. At some point, some writer at Time magazine got hold of it, and wrote a full-page review in Time. It was the first time Time had ever really reviewed a computer game, and this guy just loved it. After that, a lot of other people started buying it. Our orders shot up.
Then, later that year, we came out with the IBM version, and the IBM was just hitting the market at that point. It was interesting, because SimCity came out, and typically with the game market, you release the game and most of your sales are in the first six months, like 80 percent of sales, and then they would dwindle off, and you would come out with another game.
SimCity was a totally different profile. In the first year it did very well, but the next year we actually sold quite a bit more than the first year, and in the third year we sold even more. It was actually increasing year-on-year. It didn't have this really short lifespan, like most games. So SimCity paid for a lot of mistakes, which was great, because we made a lot of mistakes.
Did any video games influence SimCity? It seems like it was influenced by urban dynamics and Life rather than other games.
WW: Not really. I think it had more influence actually from the Macintosh design. Probably the biggest inspiration for SimCity was MacPaint. When you think about it, there's this pallet of tools, you have a canvas out here, and you grab the tools and you draw with them. We always thought that MacPaint was really the underlying architecture for it. And even before MacPaint, there were things like Pinball Construction Set, which had a Mac-like interface.
What is it about John Conway's work that caught your imagination?
WW: It's so extraordinary, because the rules behind it are so simple. It's like the game Go. A lot of people I know lost major chunks of their life from both of these endeavors. From an outside person's point of view, it might seem totally pointless, but there's some underlying aspect I think they capture, both Go and Conway's Life or cellular automata in general, of reality and complex systems.
That they can arise from fairly simple rules and interactions, and that became a major design approach for all the games: "How can I put together a simple little thing that's going to interact and give rise to this great and unexpected complex behavior?" So that was a huge inspiration for me.
That approach is quite a departure from more linear games in that you, as the designer, do not know where it will end up.
WW: Yes. One of our first models for interactivity was like the Choose Your Own Adventure stories, which is a branching tree. Or even chess, which in some senses is a branching tree -- you don't have that many branches on a given move in chess. When you're dealing with a simulation, you can view it as a branching tree, but it's so dense that it's better to be viewed as this possibility space. It is a much more open-ended world, the possibility set is much larger in that it's meant to be sculpting the possibility set so that it has meaning to the player.
So a game like chess, there's just one goal structure. In a game like SimCity, because we're not formally defining success, there's a lot of psychology involved for the players. The first thing they have to decide when they play SimCity is, "What kind of city do I want? For me what is success? Is it a big city? Is it a city with low crime? Or low traffic? High-land value?"
Players can have different balances of those factors that they consider success, which puts more meaning into that possibility space. Different players can have different goal stakes within that space, and so the diversity of space will be much higher.
Did you have hopes for what people would learn from playing SimCity?
WW: I realized early on, because of chaos theory and a lot of other things, that it's kind of hopeless to approach simulations like that -- as predictive endeavors. But we've kind of caricatured our systems. SimCity was always meant to be a caricature of the way a city works, not a realistic model of the way a city works.
So we're kind of enhancing dynamics, emphasizing them in a way where you can start seeing ideas, these kinds of truisms in city planning that we manage to capture. Things like the idea that roads don't relieve traffic, they breed it. You think we need to build more roads, but when you put more roads in, it actually encourages more volume, and the traffic gets worse. Those are kind of counterintuitive dynamics that you can capture in a model like SimCity, even if it's not accurately modeling traffic elements in a real city.
J.C. Herz's book Joystick Nation says the road-building model of SimCity represents a bias towards public transport. Is it biased?
WW: Well any simulation is a set of assumptions. So there is bias in any simulation, depending on how you look at it. A lot of people thought when they played SimCity we were really biased to mass transit. It's interesting. One of the fun things is that a model like that gives you something to reflect against. In fact, when people start arguing with the model, that's when I think it's been successful.
First of all you have to clarify your internal model -- how does a city really work? Most people, they'll kind of roughly describe it, but they've never really thought in detail what the linkages are between different things. But when they're playing a game like SimCity, which is one set of assumptions, it clarifies their own internal assumptions. And then, when they can start arguing with it, it's crystallized in their internal model to the point where they can now argue against that model, so in some senses I think that's the point of it.
Is it the same case with nuclear power? I found a 1989 New York Times article where Jeff Braun says SimCity's approach to nuclear power is to tease people into thinking it's a good idea, doesn't pollute, generates three times as much power as a coal plant, but it will blow up and that's inevitable.
WW: The nuclear power plant, that's a good example, because a lot of times we will simulate things not the way they actually work, but the way people think they work, for entertainment reasons. So in the original SimCity if a nuclear power plant caught fire, it would blow up at some point -- which they don't do. I mean, they just don't blow up like that. But it's because people were expecting that.
It's like if you show a gun in the first act, you know somebody's going to shoot it later. Or like in Hollywood movies whenever a car crashes, it blows up. Cars don't blow up when they crash usually, but in Hollywood they always do. So a lot of the times, we will simulate things on purpose inaccurately just for entertainment value.