This essay will only interest people who want to know more about narrative-building games.
I ran a game this weekend called "Myths and Legends". It was half LARP and half card game, like all of these. The game was intended to reverse this game's test - seeing if a game could exist almost entirely on that "second tier" of immersion. The "subquest" level. I knew from experience that the gameplay in these games was minimal, so that didn't need adjusting. I simply got rid of the "overarching plot", so that there were only subquest-level hooks.
The result was pretty predictable. The game, intended to run in the background over a forty-eight hour stretch (gaming weekend) ran in about thirty. At that point, too many of the players were sleep-depped out of coherence, spelling the end of the game. As expected, honestly: there was what amounted to a tiny con running in the background the whole time.
The game progressed fairly well. It was built on the idea of constructing and completing sidequest-style events. I made the "terrain" complex by using multiple eras you could visit and using some gating methods to mix up which players could be in which events. I let them use their "powers" to resolve events with each other in various story-riffic ways, and the resolution of the events allowed them to make more cards - both powers and new sidequest hooks.
Even with that level of complexity, the game's mechanics were too shallow to really hold the player: only the sidequests really drove the game.
The interesting thing is that although the game was exceptionally good at starting sidequests and directing aimless players into sidequests of their choosing, at some point the players formed a third layer - an overarching plot. The game was actually a hindrance to completing overarching plots, which is probably the reason the game died a little early.
But... but... the players formed third layer plots!
That's really important. I've been whining for ages about how I can't get players to form long-term plots. Sure, one or two players out of ten will come up with some kind of long term goal... but about 90% of these players formed extensive, complex, and frankly clever, well-done, long-term plots.
These plots weren't really suitable to long-term play, unfortunately, because they were "this is how it will happen!" plots rather than a "this is what we want" plot which has some slack built in. I tried to introduce the latter kinds of plots, and it probably would have worked if I had introduced them earlier... but they would have been my plots, and I would never have seen the rather astonishing synthesis the players did.
Because the plots were of the "this is how it will happen!" variety, trying to get the player to "play through them" is basically just torture, and kills the game. So, I either need to have a method of allowing players to finalize the plot without playing, or I have to find a way to make them construct "this is what we want" plots.
If I go with the former, I'm going to have to use some kind of "rebirth" method. Once you've done a plot, you have to change your whole outlook. A new character, or the same character in a new stage of his life, or something. That way the game can be kept fresh. Keeping these fresh, unattached characters from being sucked into the whirlpool of a more advanced plot might be necessary... and any way I run it there will have to be some kind of carryover...
If I go with the latter method, I have to isolate more exactly what caused the players to generate plot in this game so much more frequently than in the earlier games I ran. Something about the terrain having more dimensions of complexity allowed the players to twist in fun ways until they bumped into something that clicked. I think I increased the cohesion of the chunks that the players could try to assemble...
Anyway, it's still an irritatingly-paced game, so I may have to change the pacing, which will throw all my data off.