In Gamasutra (free registration required, I believe) there’s a summary of Will Wright’s Game Developers Conference Keynote. Will Wright is the genius behind Sim City and The Sims, and is now working on this next magnum opus (how many can he have?), Spore.
The talk wove together astrobiology in talking about the required research, but what comes out is the tuning required to go from principle to finished game. Will’s told me that despite the simply awesome programming of Sim City, it was only 10% of the work, and 90% is tuning. Some quotes:
“How and when research should be done: even before preproduction begins”
“this initial research led to simulation prototypes, which were just simple apps showing some of the underlying science principles visually”
“many science prototypes needed to be built in order for some of these to become interesting”
and “research should be useful in four areas: innovation, risk, fun and deep messages”
but you’ll have to read the article to find out how research helps these. In short, there’s a lot of work doing the work to ensure that your design integrates a great experience and real outcomes.