Design Essays

I set to work on this concept, but progress was slow. I was attempting something completely new and different; almost every aspect of the game had to be invented from scratch. I groped about, attacking problems willy-nilly. To help sort out my thoughts, I decided to write daily design essays. Now, I have long known that an effective way to develop your thinking is to write it down. The arduous task of articulating your thoughts forces you to confront minor ambiguities, reject sloppy ideas, and tighten everything up. For example, I wrote The Art of Computer Game Design as an exercise to organize my own thoughts on game design. Now I decided to make the concept part of my routine. Every day I would write a single-page essay presenting my thoughts on some design problem. Taken together, these essays do not comprise a design document. They do not form any continuous narrative, nor are they consistent; ideas tackled in one essay are often modified or rejected in a later essay. I made a point of not reading the previous day's essay before writing one; I feared that it would exert too great an influence on my current thinking. The essays were not primarily historical documents for later consultation; they were the by-product of an intellectual exercise. It was the exercise in disciplined thinking that was important, not the actual written result.

I did consult my essays at regular intervals, but only when I wanted to step back and trace the progress of an idea. I would read an entire week's essays, or even a month's, to remind myself of insights forgotten or mistakes not to be repeated. Sometimes, an offhand remark in an essay would trigger a new idea to explore.

Here's one of the essays:

Thoughts on the Word Set

I'm having some problems deciding what types of words should be included in my word set. These words will decide the nature of the interactions between the characters, and I am having difficulty resisting the obvious choices. For example, consider groups of words associated with locations. I could have verbs such as go, stay, and associated pronouns such as here, there, and away, and nouns such as tree, camp, trail, and whatever other specific locations I wish to include in the set. The problem is, what would such words do for the game? Here I come back to the central design problem of the day: How do I get interpersonal interaction instead of just resource management?

It would be easy to make the game just a big puzzle. The group must explore the world to find the machine gun lying in the cave that will save them from a monster. But how can I get some interaction among these characters? I don't want it to be just a matter of finding the machine gun and mowing down the dinosaur. I want them to spend their time worrying about each other.

For example, if there is a machine gun, I would rather have it accidentally go off in Gardbore's hands, and nearly hit Chris, and then let Chris yell at Gardbore. That's interaction; mowing down helpless dinosaurs isn't.

Or how about this: They find a water hole. Kendra wants to drink, but Siboot warns that the water is bad. Kendra and Siboot argue.

What if I just prepare a very small sample scene and see if I can make it work? But how would I handle artificial personality? Maybe I could play act that part out, playing the roles of all characters in the group. My test is to see whether the scene and the language are adequate to make a game.

Yes, that's it. To work!

LESSON 60

Write design essays to clarify your thoughts.

I wrote nearly 200 such essays.

The essays traced a zig-zagging course as I struggled with the many difficult problems I encountered in designing this unconventional game. Here are some of the issues I struggled with and my final resolutions for them.

The Cast

I had initially intended that the game would have just two players: the human player and the Siboot. The human would learn to cooperate with the Siboot despite their difficulties in communicating. I later decided that this was too small a cast; I needed more characters to achieve rich dramatic interaction. I therefore added several new characters. This permitted me to bring conflict within the group; with only two characters, much of the conflict would take place with external actors. By moving conflict into the group, I permitted a wider range of dramatic conflict.

The Journey

My initial concept had relied upon the journey from the crash site to safety. A standard literary device, the journey provides many opportunities for a variety of adventures. I eventually learned, however, that the kind of adventures made possible by the journey format were too disparate in nature. I could find no common basis on which to build standard algorithms for the adventures provided in the journey format. I would have to write each and every adventure as a unique experience. This created a problem for repeat playings.

Suppose that I created a hundred unique adventures. If the player experienced all one hundred adventures in a single playing of the game, then a second playing would confront him with the same one hundred adventures, and they would have no surprise value and little entertainment value. Now, I could have reduced the number of adventures that a player would experience in a single game. Thus, if the player experienced only 50 adventures per game, then he could play two unique games still not enough to be sustainable. By reducing it to just 10 adventures per game, I could extend the replayability to ten games but then the player would not experience much in any single game.

This is a common problem in many games, and I had long before recognized the fundamental nature of the problem as well as its solution, but many people still have problems grasping it. Games people, in general, seem to have an intuitive appreciation of the problem, but their understanding seems to lack the depth to realize its full implications. Multimedia people, on the other hand, don't seem to have begun to get a grip on the problem. I've been shouting this concept from the rooftops for at least fifteen years now, but my rooftops seem to be too far from the population centers.