In this year's highly anticipated game design keynote, The Sims creator and designer extraordinaire Will Wright offered insight into the research process followed in creating Maxis' upcoming title Spore. It was a highly entertaining session, where Wright hopped from subjects like astrobiology, Drake's equation on the probability of life in outer space, the Russian space program, Maxis' recruiting policies and the design of coal-mining cranes.
From a session structure standpoint, Wright decided to follow two parallel trains of thought, one (the “official” one) dealing with game design research methods, while the other (the one he actually wished he had talked about) focused on astrobiology research. This made the speech quite humorous, as Wright swapped subjects every few slides, advancing his two parallel lines until they fused in the conclusion. Obviously, the real goal was to discuss research methods, while using the astrobiology excuse to delve into the actual research done for Spore, and offer examples.
Wright began by discussing how and when research should be done: even before preproduction begins and, in his case, relying heavily on books about the game's subject. That's how he got into astrobiology, discovering the timeline of the universe, from a Big Bang some fourteen billion years ago, to the appearance of life on Earth around four billion years ago, and possibly on other planets as well. For Spore, this initial research led to simulation prototypes, which were just simple apps showing some of the underlying science principles visually. Wright showed as an example, a galaxy simulation software, where he could actually “plant” seeds of life and advance time to see how life forms expanded through space with time. Interestingly, he commented, many science prototypes needed to be built in order for some of these to become interesting in terms of gameplay. Only these were then refined to experiment with control and gameplay mechanics for the game. Obviously, this frequently led to more research, so the basic research-science prototype-gameplay prototype loop was closed.
By using these techniques, Wright could examine the different astrobiology theories and decide which ones would be used in the game, from Stanley Miller's pre-biotic soup from the fifties to more modern theories like panspermia (the pollination of life in space by means of spores traveling, both unintentionally and intentionally).
Wright explained how, at the end of the day, research should be useful in four areas: innovation, risk, fun and deep messages. Innovation was characterized as a clear sales driver, helping differentiate new games from simple clones on store shelves. In this respect, it helps widen the signal to noise ratio, where the signal would be the novel proposal versus the background noise of planar clones. In Spore, innovation drives procedural game content, making the player capable of creating new, unique creatures and universes. Innovation is also used to encourage player creativity, by allowing them to share their data. Finally, it also drives the need for an unbound game world within the game.
The second area where innovation can become useful is risk. Wright stated that risks can actually come from technology, design, production, marketing and even political (intra-company, that is) sources. In his current project, they identified sources of risk early on, to ensure the right steps were taken to control them, and overcome them in the end. For example, one of the main sources of risk with procedural content is the need for procedural animation techniques. These were tackled by Chris Hecker and Ken Perlin, pioneers in this area for over a decade. Similarly, measures were taken to assess the risk in editor usability for the game, the design scope of the product, the content production pipeline and the ways to use pollinated content (that is, user-generated data).
It was interesting how Wright explained that previous user editors actually produced a hyperbolic curve of quality vs. quantity: lots of poor quality content, and a smaller and smaller ratio of higher-value content. One of his goals with Spore's editor and pollination mechanisms was to increase this proportion, to ensure more people could create great stuff using the game's toolset.
In the end, Wright reflected on how games can be used to convey deep messages, to tell users something, and even maybe make them learn along the way. This was probably the best part of the talk, when Wright stressed how his game can actually become a vehicle for information sharing and education. In the case of Spore, his goal was to let players know that life in the universe is diverse, and comes in all sizes and colors. A second theme for the game would be that anyone, given the right tools, can be creative. Third, he wanted his audience to learn, by playing Spore, that the universe is not flat as we often see it portrayed in science books: it is a complex, three-dimensional space where planets, galaxies and solar systems can be aligned and oriented in a variety of ways. And last, but not least, he wanted his audience to get a real feeling of the size of a galaxy. We often forget that a regular spiral galaxy like our own Milky Way is “just” one hundred billion stars, each one maybe with its planets.
In the end, Will Wright's GDC 2006 keynote was surprising in terms of theme, but not in terms of the value of the information presented. Instead of a lecture on future game design methods we got a one-hour session on how to do proper research for next-gen games from one of our industry's brightest stars, and an insider's look at the research process for Spore. While no actual media from the game was shown (except for character sketches and some simple science demos), attendees were left with the clear impression that Will Wright's next-game may very well be the “Sim-everything” he wants it to be.