Just googled myself to this thread. I'm working on a new cmapaign under... let's say stressful conditions. We have 2,5h sessions and use this campaign as a gap-filler when we are one player short of our normal group. So a session is run every now and then and I have to find easy to remember conclusions at the end of each sessions so we don't have to spend half an hour telling what has happened 2 months ago.
Mustrum_Ridcully told me about narrow-wide-narrow. I have started to use that concept in a micro version. I start every session with a definite starting point, give some room for decisions and come back to a definite end. The end needs not to be the only one possible, just a definite one. The setting is Dark Sun, so it's quite clear. But I don't plan much on an overall arc, I just think maximum two sessions ahead and have some kind of rough idea what might happen in background and in the future, which gets more and more clear once we come closer. And every further inspiration I get from the actions of the players. Just let it flow
For such a type of campaign that is not used in regular sessions preparing and thinking too much ahead is just a waste of time to me. I'm prepared to jump in for the next session if one player has no time, and a second session is maybe also roughly laid out and that's it. After the session I think on what might happen next. With the creature catalog creating some battle is fast and easy. Just put some skill roles in between, save some time for role player the characters and you're done.
This is not my usual type of play, but since the discussion was on narrow-wide-narrow it might be a good addition. Usually I'm the sandboxy type of GM.
1. I make up my own setting, ruthlessly thieving parts from other settings and putting them together to a rough idea. It's mostly brainstorming. I also have some rough ideas about possible larger background plots.
2. I work out the surroundings of the starting area of the PCs. Those will be more detailed, but just the part that they have access to in the beginning. The further away, the more vague it gets. In one campaign I even have a "border" around the starting area, a circle of almost impassable terrain and nobody knows anything about the outside. I just did that to keep things simple for me and to have the players focus on solving the problems inside the circle first (if not people will die, so they have to).
3. I work out interest groups inside the starting area. What do they do now? What are they goals? What are their opponents? Is their armed conflict? What territory do they occupy? If there is a chance I will have those groups in combat with the pcs involved (as opponents or allies), I generate some sample members of the group with full stats. Like the standard member, team leader, single important persons and so on. If there is no chance for combat, the important npcs get maybe a class and level or a few keywords with whom I may generate some stats on the fly.
4. I work out some quests or plots for the pcs to follow. Maybe some initial goal, some side quests and such things. Mostly based on the results of 2 and 3. Maybe I'll twist in parts of the background plots I have from 1. Those background plots usually have become more clear by now because they interact with what I have fixed already.
5. That's it with initial prep.
During the sessions I see where the pcs go and where they get involved. I check how that works with my idea for the interest groups and see what changes. Who knows what and will do what. Who want to have something from the pcs? Does he use diplomacy or force? Which groups become stronger rivals and might engage in war? Which groups rise due to their resources of the influence from the pcs? So the pcs are the may force behind the world. Their actions often are the impulse to change. From what the cps do and what the ideas of the players are I put up new quests and change the world. And depending on that I'll also start to develop the world further and tie in the background plots more strongly. Sooner or later the pcs will stumble upon the background plot and then they might see the red string and follow it. If not, next idea. Just listening to your players during the sessions, especially when they discuss and you have not much to do gives you tons or ideas.