Drawing Board IV: Lots and Structures

This article originally appeared on The Next Generation blog at Simtropolis.com
************************
 
One of the gripes that has come up with Cities XL is that having fixed lot sizes makes it a major headache to expand a road;  To accomodate the extra space it is often necessary to demolish the lots lining the road on at least one side.  Typically the retort is that "well, it's a headache in real life too."  True enough, but it's not necessary in real life to tear down the entire lot to shave a few extra yards off it (and then leave it vacant because it doesn't "fit" anymore!)  While part of this is a problem with how roads are laid, it also indicates a deeper problem with the zoning model in both Sim City 4 and Cities XL.

I'd like to look at an alternative way of approaching lots and buildings; But first, as usual, the Disclaimer:
 
***************
IMPORTANT DISCLAIMER
 
This series of posts are not to be read as "official" ideas for the Metropolis Project, they are simply my personal thoughts about what some cool game features might be.  The final feature set for the game will be decided by the community and the development team contracted to build the game, and may not include any of these things.
***************
 
Like any of these simulation elements, the best practice is usually to look to real life and see how we can best approximate it.  Why can you expand a road in a real city without tearing down all the houses on the block?  Obviously, because the size of the lot is not immutably connected to structures on it.  That's the weakness of having a "lot" as a connected unit of both structure and surroundings.  Ironically, it was less noticable in Sim City 4 because both lots and roads were tied to the grid, so you could always fill up space with one-square lots; (Or to put it another way, the only road expansions possible were dramatic ones, from road to highway, so some demolition made more sense.)  In Cities XL, they introduced greater fluidity for width and angles of roads, but did not complement that improvement with equally fluid lots (in fact, the fixed size made them even less flexible.)  If Metropolis is going to move off the grid (which I firmly believe it should) it will have to encompass the separation that I mentioned above: making lots distinct from structures.
 
The mechanics of this would be relatively simple, and although it would mean greater demand on the system I think median computing power is up to the task (and will be even more so by the time these ideas can actually be put into practice).  Lots would be defined as vector shapes;  These could be either placed several ways: manually, with an Illustrator-style "pen" tool (for fine-grained control); In a number of predetermined shapes (most lots are actually rectangular); Or with an "auto-fill" function, which would fill a space between roads with an appropriate number of lots for the area and the zone type (or a user specified number).  When a lot became occupied, the simulation would use an algorithm to determine what kind of structure would appear on the lot, and it's position;  on a suburban lot, for example, a house would probably not fill the entire lot;  instead, a house model of appropriate size for the initial property value would be selected and then placed somewhere relative to the street.  A key point here is that the structure would not constitute the lot, it would simply be a subsidiary feature contributing to it's overall value (more on property values later).
 
The advantages of this system are numerous;  To take my example from the beginning, it would be easy to expand a road by simply shaving off a meter or two of the lot;  But (unless the structure was bulldozed as a result of the construction) doing this would not prompt the simulation to re-evaluate the lot's occupancy or structures;  Except for the footprint being smaller, the lot would be unchanged, and the footprint would only play a part in later calculations for new construction.
 
But beyond this, there are many other benefits: oddly-shaped lots would not become a nightmare for building designers, as any building (of any shape) could fit on any shape of lot (provided there was sufficient area in there somewhere).  It would also make incremental development possible:  again, looking from a real life perspective, a homeowner who is making more money doesn't often tear his house down and build a better one;  But they might decide to add a garage!  Secondary structures could fit easily into this model.  Lastly, lot areas would conform easily to curved roads, or other boundaries (water, for instance.)
 
While there's clearly a lot more complexity to this model than using a grid, like SC4, or using square lot-buildings as CXL does, I think separating lots and buildings is necessary in order to create uneven lot shapes, which in turn is a crucial complement to non-grid roads.  After all, if you're going to break free of the grid, there's no point in doing it halfway.