mitigatedchaos

How are you guys liking this sudden series of polygon-based urban planning posts?

argumate

wonder if we could rig up an actual simulation of this stuff

mitigatedchaos

I have no doubt that I could, but I can’t justify doing so without funding.  You know how it is.

argumate

fund it via KickStarter on the basis that it’s developing levels for an fps-

actually that’s needless deception, the indie games market has demonstrated that people are more than willing to spend on incredibly niche geeky shit like simulating perfect virtual cities

mitigatedchaos

Well, let’s see.  To justify the project as a game, I’d need about 2,000 people to pay about $10, or about 1,000 people to pay about $20. 

So let’s think about what that might look like.

As you may know, Skyrim compresses distance in a symbolic way by somewhere between a factor of 10 and 100.  It’s how you can walk ten meters, kill a pack of wolves, then walk another ten meters and kill another pack of wolves.  What often isn’t discussed is that most citybuilders do this as well, which is why they require such laughably overbuilt transit infrastructure.  Cities: Skylines gives you about 36km2 to play with.  Singapore is about 720km.  Manhattan is about 59 kilometers.

So presumably, this simulation game would work to aggressively limit computation so that it can do a lot more computation.  It would rely on a grid of 5m squares instead of allowing as much freedom in road-building.  Agents would all be represented in the sim, and their paths would be stored, but visual traffic would be represented as a statistical aggregate and not following individual cars from place of employment.  The visual style would be fairly simple to conserve on cycles - even people would be represented with 2m tall boxes (children with shorter boxes, obviously).  Traffic paths would primarily be updated opportunistically rather than continuously.

In exchange, each resident in the city would be simulated, and a wide array of more complicated municipal policies would be available, including state ownership of housing developments, the various components needed to try the One Thousand Villages, etc.  There would also be “markets” that bid somehow on jobs or properties, to decide what to do with a mixed-use zone for instance.  Just managing the fine details would end up requiring some minor automation and I bet there would be a “zone for subdivision” fill tool that caused a private property developer agent to insert a dynamically-generated slice of suburbia.

Possible elements could also include political costs/capital needed to push through bulldozing houses to install new light rail, disruptions from the global economy, etc.  Another element could be the cultures thing @lockrum brought up, but rather than real ethnicities (which could make people upset, and also are the largest source of real ethnic tension), they’d be represented as high school cliques or various subcultures (jocks, otakus, hipsters) and you could forcefully integrate them Lee Kwan Yew style, or they might self-segregate automatically, or they might be atomized by the distribution of jobs and housing prices.  (They would probably be depicted as different-coloured two meter rectangles instead of grey ones.)

Would people fund that?  I’m unsure.