Warning: Most sane people will require context outside of this text to understand the discussion.
So, the big announcement at GenCon was that Wizards of the Coast (WoTC) said that Dungeons & Dragons 4th edition (4e) was going to start hitting shelves in May 2008. For those of you who aren't familiar with the D&D product cycle, let me explain:
Each new edition of the game starts off with three core books to purchase. After these come a procession of optional tomes, from which a gamer can pick and choose according to their interests in the game.
So, it comes that over the lifespan of each edition, one accumulates a store of revelant texts. However, these books become obsolete and fairly useless once the new edition is launched. Obviously, one can refuse to 'upgrade' to the next edition, but this curtails the ability to participate in the game fully, since official events (such as GenCon) are run using the most recent rules, that game stores will only support events using the most recent rules, that it's difficult to introduce people to the game using outdated rules, and that support for the old edition will cease - no new products, errata, etc.
Since the announcing of a new edition thus entails a loss for current gamers, it's generally met with scorn, as was the case this time around, on a variety of internet locales and at the announcement itself. However, it's also the case that gamers will stop buying product from the current edition (3.5e) once the new edition (4e) is announced.
Disclosure: I am fairly heavily invested in the current edition, and have received the announcement with less than good spirits.
So, WoTC faces a complex constrained optimization problem. They would like to roll out new editions frequently, but this tends to irritate people who play the game, hampering sales of the new edition. Similarly, they can't announce the new edition far in advance, else people will stop buying current product. But being unable to announce the new edition prevents all sorts of good things, such as openly consulting with people who play the game on what they think needs fixing.
Frankly, I would love to know if they have any sort of quantitative process for resolving these conflicting tensions. It would personally make me happy if they do. The process seems similar to that of deriving an optimal copyright term, and certainly bears many similarities to problems concerning dynamic time inconsistency and game theory.
Given the current economic rage over the economics of massively multiplayer games, maybe I'll try to figure out a model for the situation faced by WoTC.
P.S. From Robin Hanson:
...we have too few nerds, and all else equal we should want to subsidize nerds, to get more of them.