What I wanted -- and what several other people I play "tabletop" RPGs with wanted -- was the feeling of "our group" being the "stars of the show" -- the ones fated to save the day, etc. while engaging in group problem-solving, clever banter and other socialization. But we wanted to be able to save the world at our own pace, when we could all agree to get together online, not log in and find out it had been saved without us, or that someone else was currently on hogging all the action.

This led me to think about a possible hybrid. I was thinking in terms of text-based MU* systems, so I called it a MESS: Multiplayer Episodic Storytelling System. The idea was that a group of players could get together and form a team, and a new instance of the world would be spawned just for them; when they logged off, that instance would be packed up and stay waiting for them until that team logged in again. To conserve server resources, as well as to make story construction easier to manage, I thought of breaking up the world into individual "episodes" and "scenes," which would be unpacked and presented to the players as needed. I actually started modifying PerlMUD into such a system. Then my first life got knocked sideways by Second Life. So much for programming MU* systems.

Now I'm back to thinking about the episodic system, though, because here we are talking about LARPs and RPGs, and we have the same sorts of problems that MMORPGs have. Many games still manage to be enjoyable despite this (I plan to spend time in Numbakulla this week), but I think the general idea still has merit.
Building sets and props is a lot of work. One wants to be able to get as much use out of all that work as possible, i.e. leave everything up long enough for many people to visit. But SL restricts the amount of land and prims you can have at any given time, and having big builds tends to be a problem because if more than a handful of people visit at once, the server goes to its knees. What I propose is a "set packer/unpacker," which would allow someone to build a "set" or "scenario," with props, scenery, and (preferably) scripted NPCs, then pack it all into an object that would unpack on command, on a "stage" owned by the game provider or even by the group playing. Several setpacks would be created, for different scenes. Ideally, they would be able to be linked using commands so that exiting from the boundaries of one would cause it to pack itself and unpack the adjacent set (and put the characters in the appropriate locations of the new set).
Think of it as "holodeck programs," if you like.

To go with this, I'd like to see a simple toolkit for setting objectives for each scene. This gets at the "episodic" nature of the system. Examples of objectives would be things like "make Enchantress laugh" where the Enchantress is an NPC on the scene, with a set of simple "emotion" parameters triggered by specified objects and actions, with "laughing" as an outcome of certain of those emotions. (The emotions would operate on scales similar to those used for physical attributes in more traditional RPGs, but would be more easily modified.) "Solving" a scene would result in one or more scene outcomes, e.g. changing to a different location, characters gaining new abilities, NPC providing necessary information, etc. Scenes would be strung together in story arcs. The action wouldn't be quite as free-flowing as a SL/LARP, but would have a strong sense of "story." This would be a way of developing multiplayer (but not massively multiplayer) interactive fiction in SL.
As an extra feature, people could hire out as NPCs or NPC teams for a higher level of participatory theatre, bring their own setpacks, provide the experience for the customer, then pack their stuff and leave.

Maybe it's overkill. Maybe with the population density of SL, setting the scenes up at distances from each other with some kind of teleport (possibly using the URL launch that I hear the HTML support will allow) would eliminate the need for the pack/unpack functionality, as no two groups would be likely to be there at the same time -- the sets would just have to "remember" who had been there before and act appropriately. But if a game got popular, it would be nice to know that your team could play the part of the game you're currently working on without having to worry about someone else being there at the same time, striving for the same goals or cross goals. And being able to eliminate the land problem would be nice, too-- someone with a limited amount of land could still build great setpacks and sell them to people to host on their own land.
Most of all, it would be important to make these development tools so easy to use, that anyone with good storytelling ideas (and an SL account) could use them to create these kinds of games.
What do the rest of you think? Is any of this making any sense? Is anyone else interested in these concepts, even some of them? Would you play games like this? Would you build them, if the tools were easy enough to use? I do have more details written up, if anyone is interested. I know I don't have time to develop this, but I'd like to work with someone else, if this strikes someone's fancy.
neko