Is it Possible to do a MMORG Right?

My previous post raised the possibility that massively multiplayer online role playing games, or something similar, might play an important role in the future, due to the fact that they, unlike movies, can be technologically protected from copying. That raises an obvious question: How good a story can a MMORG be?

Current versions face at least two intractable problems. The first is that, if the viewer is a character in the story, what he does should matter, should affect what happens. But the authors cannot write a different version for every player, so in practice there is, at least in World of Warcraft (I'm not as familiar with other games), a single plot line which the character is being walked through—meaning that he himself is impotent.

Many years ago, I  had the experience of participating in a role playing game (Empire of the Petal Throne, for any familiar with it) being run by the man who created it. He was a Tolkien-class world builder but not a Tolkien-class story teller. The game felt, not like a story or adventure, but like a guided tour. And that was a game where there was a real human being in a position to modify what happened according to what the players did—not the case for a MMORG.

That suggests one possible solution to the problem, a fusion of the ordinary role playing game, with a human dungeon master, and the MMORG. The online game provides the world and backstory, while providing a talented DM the tools to take the plot line in any of a wide variety of directions, depending on what the players do. 

One disadvantage of that model is that it, like D&D and its imitators, provides a world with a lot fewer real people in it than World of Warcraft. I am not sure if it would be possible to design something along those lines in which each human DM had control over a different part of the world and plot, making it possible for the actions of each player and group of players to determine one strand of a much larger, but internally consistent, story. Carrying my speculation farther forward, it might be possible to do the same thing using artificial intelligence to replace the human DM. I can imagine a primitive version of that at present, where a game has an elaborate network of potential plot lines, with the particular strand followed by a player depending on his choices. That is already doable, and done, by some single player games. It would be much more difficult with a massively multiplayer game, but perhaps not impossible.

This brings me to the other and related problem with the current version. Your player is told that there is some dire peril threatening the world, or at least large parts of it, and he is the only one who can deal with it. He succeeds in his quest, kills the monster. Half an hour later he observes another player, having been told the same story, fighting the monster he just killed. It is hard to tell a consistent story that way, or successfully suspend one's disbelief in a story that inconsistent. Over and over again.

To solve that problem, I think you need two changes, one or both of which may have been implemented in games I haven't played. One is to abandon the "save the world" point of view and give your characters objectives at a more plausible level. There is no reason why your character cannot save a child who is being attacked by a pack of wolves today, and observe another character saving another child from another pack of wolves tomorrow. That is only a partial solution if the two incidents are obviously identical. So the other part is to design the challenges with a lot of random variation built in, enough to make each run believably different from the last.

As should be obvious, these speculations are coming out of a very limited experience with the world of designing games. I will be interested in comments by those who know more about it, and in particular in examples of existing games that solve one or more of the problems I describe.