Monday, March 04, 2013

Not Sim City

So, I was thinking about how to make a game like SimCity but without totally dropping the ball. The idea is that it has to be simpler and faster than the old SimCities, but not as simple or fast or always-onlined as the new one.

Here's my thought: what if it was a city where space could be zoned in large chunks or small chunks depending on the need? Cast aside the thought of road planning as a major career. Instead think of the zoning.

You start at a city center. A square kilometer or so of space starts off at a medium chunk size. You lay down a civic center, a police station, a bunch of chunks of commercial and light industrial, and so on. You weigh whether you want to build a bridge across the river, or flatten some scenic cliffs.

At the edge of your square kilometer of city land, the chunk size is in square kilometers. So you lay down some suburban residences, some farmlands, some woods. There. Now you not only have your city, but also the area surrounding it. Pretty and useful and fun.

As your city density rises and you run out of that initial square kilometer, you begin to "absorb" the nearby areas, breaking them down into chunks and extending into them exactly like the first square kilometer. So the dense suburbs near the city center are converted into more city, bit by bit. Hell, you might even absorb noncontiguous areas, creating two dense city centers with some chunky, largely undeveloped suburban space between them.

There's no reason you can't extend this zoning system further, either. Break down the city-block-sized chunks into much denser areas, and then reshape the buildings, parking lots, and parks within. Craft to your heart's content. Or use very large regions and paint them as being in the same country as you, different countries, different border types...

I like the idea of starting big. Like any painter, you start with a very rough sketch, jotting down the important elements. A forest here, some suburbs here. Then you start to break it down. Do the gritty work as you need to. And maybe you never do - maybe that part of the picture is fine as a rough sketch.

Start big, then start to break down into smaller pieces as complexity goes up. The smaller pieces allow you to handle growing complexity.

For example, imagine you have a commercial zone and surrounding it are six residential zones. Individually, each residential zone benefits from neighboring a commercial zone, and visa-versa. The commercial zone gives them jobs, the residential zone gives the commercial zone workers and a market.

But six residential zones is too many for one unpolished commercial zone. The tremendous number of people flooding into the commercial zone is too high for it to handle automatically, and even though it tweaks itself to find a good balance, each of the residential zones has slightly different parameters and therefore expects slightly different things of this one commercial zone. Efficiency plummets. Crime probably rises.

So you click on the commercial zone and switch it over to medium-level detail. This will allow you to build the zone like a machine, so it can handle the requirements on it.

Maybe the residential area to the north is wealthiest. So you put a swanky mall full of Starbucks on the northern end. Much of the rest of the commercial zone is dominated by a much less swanky set of shopping areas and Dunkin' Donuts. This means that the swanky crowd from the north quickly find a spot that works well with their specific needs, while the rest of the crowds are efficiently funneled into the area that works well with their specific needs. Of course, you could reach either one at any time, but it's a matter of lowering the friction. The better and more effectively you match the input/output needs, the higher the efficiency and lower the crime rate.

But every interaction has side effects, and malls aren't the only things in a commercial zone. Malls have requirements as well - for example, a massive amount of transport and warehousing and parking lots. So you have to build these secondary structures to "catch" and "buffer" the side effects and requirements of your primary buildings.

If you leave it at just malls, the employment levels in the commercial district would be terrible - no jobs higher than minimum wage is a great way to insure your nearby residential zones wither into economic dog days. So you also need to create office buildings.

These have a different set of secondary requirements and side effects that can be mitigated with different secondary structures such as police stations, subways, eateries, pubs...

So the initial challenge is this: set up your commercial district to filter inputs and outputs into various primary structures, and build the support structures needed to keep them running with a minimum of crime, traffic, and annoyance. If your filtering is bad, friction rises as people end up defaulting to unsuitable locations. If your support layout is bad, you waste lots of space or end up with a high crime rate or a low efficiency rate.

Then the challenge evolves into keeping your commercial district suitable as the populations and economic conditions of the surrounding zones change. It always comes down to that in construction games: choking on your own creation. It isn't the difficulty of the game that kills you in the end. It's the layout you created. And this seems like it would highlight that.

No comments: