Raph Koster's Theory of Fun is that games are fun because they teach us things. I was recently thinking about a different angle of that: For me games are more fun when I still play them sub-optimally, when I haven't understood the game perfectly yet. Because for example back in the first days of World of Warcraft we still did crazy stuff like forming groups with random strangers to do quests, or going on raids without making a selection of who was allowed to come and who wasn't. And that was fun, even if it wasn't perfectly efficient. Over the years I learned a lot of things about World of Warcraft and MMORPGs in general, and certain sub-optimal modes of play were discarded as being not efficient. There are millions of ways to play a game sub-optimally, but only one way to play it optimally. The more I learned about the game and the more I approached optimum efficiency, the less options I had, and the less fun I had.

Unfortunately that loss of fun is portable. I gave up on the re-released Final Fantasy XIV very, very early in the beta process because I noticed that fighting with my spellcaster in FFXIV was the same as fighting with a spellcaster in every other MMORPG I had played for the last decade: Start out at maximum range and pelt the mob while it comes running towards you. If you get a mix of spells with different damage, range, and effect on enemy movement it is blindingly obvious for anybody who played a similar character before in a similar game on what the optimum sequence is. There is no fun of experimentation or learning things any more. And because there is only one optimal sequence, you end up pressing the same buttons over and over in every bloody combat until you are bored to death. (And for FFXIV beta I played the console version, where the button sequence was annoying.)

The hope for the future of MMORPGs is more dynamic worlds, because that means that you DON'T always know what will be happening, where exactly the monster and the treasure is, or how to best defeat them. At least, that would be the ideal of a dynamic virtual world. If any virtual place just goes back and forth between two possible states, players will quickly have learned both possibilities, and we're back to boring optimal behavior.

What I really would like is games in which rewards are also changing dynamically. Why should your hundredth orc give the same xp as the first one? Why should there be still treasure in the orc village after you killed every inhabitant five times over? Dynamic rewards have important consequences on how players behave: In Everquest players didn't move, because it was better to "break" a spawn and then camp it all day. In World of Warcraft players move because of quest rewards, but because everybody has the same quests, people all tread on the same paths. I'm currently playing the Borderlands 2 DLC Tiny Tina's Assault on Dragon Keep, and it has a parody of a scene in it where you need to make other "players" rage-quit in order to be able to get credit for killing a monster that respawns only every 10 minutes and is camped. That is funny in Borderlands 2 because it is so true: Other players in MMORPGs are frequently regarded as competitors for the same limited resources, and not as potential allies and new future friends. Moving away from static quests with static monster spawns and static quest targets towards a dynamic reward structure that doesn't punish players for grouping would go a long way to make these games more interesting, more social, and ultimately more fun again.