The Problem of Using Real Life Spaces in Games

with No Comments

When teaching level design (I’ve taught for over 3 years at Madison Media Institute when I was in Madison, Wisconsin), I like to give new students and assignment called “Chapel.” This isn’t an original assignment. This was given to me by Professor Jon Skinner at Southern Methodist University back in 2008.

The parameters of the assignment were simple: build a space using only BSP (BSP could be thought of as the lego blocks used by designers inside game engine software to build level geometry). You weren’t allowed to program any scripted events nor use any kind of art except for basic textures (so a brick wall could look like a brick wall, for instance, but you couldn’t have a nicely built love seat or lighting fixture).

The other part of the assignment was to use, as a template, some preexisting structure. While not required per se, using some sort of cathedral or chapel (hence the name) was suggested. In any case, you had to use some real life building.

Most of the exercise is about finding a comfort level building in 3D, as well as learning the software (in this case, Unreal, but the same applies to the other game engines, even ones without BSP such as Unity). However an important aspect of the exercise is learning a fundamental game design truth: real world spaces aren’t built for fun gameplay.

This should seem obvious, but understand what happens to the designer when building a space. You can easily get lost in the internal beauty and logic of the layout itself, without keeping track of the true utility of what is being built.

This is a much easier problem to get trapped into when creating 3D spaces, as opposed to 2D spaces, because designing a 2D space (such as a platformer game like Super Mario Bros.) is already abstracted from reality such that your brain is geared towards creating gameplay, while being less concerned with the reality of the space itself.

In 3D space, however, it is easy to get caught up in making reality. After all, it looks so real!

In the case of the Southern Methodist Guildhall “Chapel” assignment (and my own when I was the teacher), the gameplay was simple “deathmatch*” play, usually about four to eight players. What kind of gameplay problems occur when using a real world floor plan without adjusting it for the type of game the space is being used for? A few examples, taken from the Chapel assignment:

  1. Pews and seating causing obstruction to players.
  2. Irregular floor plans potentially causing navigation problems for players.
  3. Linear floor plans creating poor circulation of player movements; players get trapped in one corner of the map, unable to escape.
  4. Lack of interesting tactical options for players (if the floor plan doesn’t have these features built in naturally, such as a balcony to climb to, or basement to hide in).
  5. Real world dimensions that do not work in a virtual space. The huge culprit here being doorways; in games, it feels natural for doorways to be much larger than in real life!

*Deathmatch gameplay is a typical game mode in which each player tries to shoot another player, with no teams. Think of it like a free-for-all paintball match or laser tag.

So what should the lesson be for the aspiring designer? Abandon the study of real world architecture and design?

Absolutely not.

Instead, the designer should focus on how to adopt and adapt these spaces to properly accommodate the intended gameplay. For deathmatch gameplay, making sure large, circular routes around a map is a basic design pattern that can be designed into any real life floor plan. For a first person action game, creating fascinating secret paths and providing helpful cover objects to help shield the player from enemy attacks can also be accommodated.

Real world design features and gameplay design can be in harmony, and this is a small but important slice of what great level design is all about.

Leave a Reply