Udo's Techblog

After alienating the rest of my audience (yes, all two of them) with politics, it's back-to-technology time!

BoingBoing's Cory Doctorow reports how the FarCry people are utilizing automated content generation to alleviate the load on content designers. Finally, this idea is getting some traction!

Take World of Warcraft, for example. There is a whole army of people who do nothing but design quests, buildings, plots, creatures, zones and everything else. The bigger the game gets, the more it costs to produce new content, because it's all interlinked and has to balanced out. The last expansion, Wrath of the Lich King, introduced literally thousands of new "get me three rat tails"-style quests. Apart from economics, a huge part of what makes MMORPGs like WoW less like actual worlds and more like treadmill social games is that this content has to be re-used over and over. Millions of people play the same storylines, and even when looking at individuals, there is a lot of repetition. You kill the same named mob over and over again until it finally drops something worthwhile. You keep clearing the same dungeon again and again with your raiding party. In a way, it's even worse than pre-canned single player RPGs, because there you can at least keep the illusion that a story is taking place for you only. And the good ones don't respawn unique mobs, too. But in a MMORPG, they have to, because of the way the whole experience is designed. The only way to simulate uniqueness and quench the creeping sense of repetition is to introduce even more carefully-designed static content, so players can at least have the illusion of meaningful choice.

Pit against this static concept a dynamic world, where the mobs are actually living entities with their own agendas, where the environment actually responds to actions of players and mobs alike. A world where creatures have their own dynamic ecosystem, where NPCs build their own societies, settlements, cities and dungeons. Every single one of them has a unique history. When they're killed, they stay dead - with all the consequences. If an important mob is killed, what will happen to its followers and minions? What really happens when war sweeps over the land? How does a landscape slowly change as an evil necromancer is setting up shop in the dungeon below? It's pretty hard and unrewarding to explicitly design these highly dynamic scenarios, to speak nothing of fiscal impossibility. Those questions can only be answered by a real simulation.

I postulate that the simulation, while undeniably complex, does not have to be especially pedantic and overwrought in order to deliver some pretty cool results. The simpler the rules of the game are, the more dynamic and surprising the interplay of all the entities can become. While it doesn't necessarily mean that balancing factors and design decisions will not be necessary, I suspect that a self-balancing world is achievable and fun to play in. At least, it would be for me, because WoW is definitely not the way to spend my time...