I recently wrote a post in ARGN about the question of interactivity in mmoe-type games. Basically, I say that interactivity with respect to altering the story is overrated.
Warren Spector, the designer of Thief and Deus Ex, both groundbreaking games when it comes to player interactivity, said in Edge that he wanted Deus Ex 2 to exhibit emergent behaviour with its non-player character AI. No more scripted events, he said, we want the game environment set up so that interesting and varied things just happen. And this makes perfect sense to me, to have increased player freedom within levels.
But once the level is over, the interactivity stops. The game doesn’t ask the player, ‘So, where should our hero go now?’ and I think this is so for a few reasons. Firstly, it’s extremely difficult to design in this sort of interactivity beyond the level stage. Secondly, I don’t think it makes for a good story. It makes for a choose-your-own-adventure, and not withstanding the fact that I don’t think many people like that sort of thing (people still go to the non-interactive movies, after all), I stand firm on my belief that pre-written stories tend to be best. Yes, of course you can have a certain amount of flexibility within the story if you’re doing it in real time and trying to respond to players (e.g. RPGs, mmoes) but the overall story should be planned out right from the start.
I was re-reading Microserfs by Douglas Coupland the other day, and it was set during 1993/4. In one section, the group of programmers go to a conference about the different types of games that are being developed. They mention interactive stories, and say something interesting – people want to be entertained, they don’t want to have to choose what happens next in the story. Every time some new genre gets developed, the whole argument about story interactivity is rehashed.
mssv 
I think part of the reason this question has never been satisfactorily resolved is that some people just want to be passively “entertained”, while others genuinely DO want real interactivity. People who fall into the former group are more comfortable as entertainment consumers than as producers; the people in the latter group are budding artists and producers themselves. I see no inherent contradiction or undue difficulty in creating MMOEs that address the interests of both groups.
Clearly that’s the case, from the success of the Beast. However I still do think there is enormous room for improvement, and I’m not even sure that the designers of mmoes are actually aware of this distinction between ‘lean-back’ and ‘lean-forward’ players.