One of the things we had to bolt down early was the RLD. We knew that a major factor in the RLD would be the level geometry. The essence of the game was the hide and seek [inverse-stealth] element. A related critical factor would be the player’s navigation and spatial awareness. With both of those in mind I broke it down working backwards. I actually worked within Unity itself to make these studies. It was convenient due to the ability to snap movement and also work towards eventually establishing a template for building module size.
00 – The image below shows step 00. That is what we would have if all buildings were the same and all street sizes were the same. The end result of this, as you can also read in the image’s text, is that positive and negative spaces were monotonous. Because no hierarchy can be established visually, the player does not have any spatial landmarking. Only buildings of different heights can alleviate this, and even so only to a small extent. Finally, it is also visually boring.
01 – We wanted to keep those buildings as modular as possible, so I started looking at the streets. By simply providing 3 different street sizes, a hierarchy of spaces was established. You can see the ratio of street widths to each other. This does not only begin to alleviate the visual monotony, but also provides a means for creating metric values for the RLD. However, it was still a long way away from not looking boring and landmarking is at best still poor.
02 – Image 02 above begins to show when a second ingredient is thrown in: formal variety to compliment the spatial variety. By combining the positive building volumes, first hierarchy is created between the positive volumes and second, negative spaces gain more definition. By “composing” the city, spatial landmarking can now be established.
03 – We now have the beginnings for a spatial puzzle where potential camera and roadblock player placement can be hinted via level geometry. Geometry is re-organized to hint to the player how to breadcrumb contestants into open plazas.
04 – Finally, by considering each intersection TYPE as a node we can assign a value and determine the level progression. This is something I will elaborate more on in a subsequent update.