Celia Pearce on gamestudies.org:
CP: When you were first working on SimCity, what was going on in the game world at that time? Were you responding to games that were out there, were you wanting something different? Were there things that influenced you at all in the game world or were you just totally in a different mindset?
WW: There were things that influenced me—not many though. There was a very old game called Pinball Construction Set by Bill Budge which was great. He was kind of playing around with the first pre-Mac Lisa interface, which was icon-based. He actually put this in the game, even though it was an Apple 2 game. He kind of emulated what would later become the Mac interface. But it was very easy to use, and you would create pinball sets with it which you could then play with. I thought that was very cool.
Also early modeling things, like the very first flight simulator by Bruce Artwick which had this little micro-world in the computer with its own rules, kind of near reality to some degree, but at a very low resolution. But yet it was this little self-consistent world that you could go fly around in and interact with, in sort of limited ways.
CP: What kinds of responses did it give you when you did stuff?
WW: It was very open-ended and I could do whatever I wanted to in it. The first thing I did was I went in and started exploring the behavior space. Trying all the different things with the airplane. What happens if I go straight up? How far can I go? What happens if I crash? What happens if I do this that and the other? So I could carry out experiments in this world. And in running those experiments I could get a more accurate view of what the internal model was. So it’s kind of a scientific process. It’s kind of a “hypothesize, experiment, change your hypothesis” type cycle that was going on.
I read this piece a while back, but re-reading it in the context of “learning programming” has been very enlightening.
(via Bret Victor)