I think it'll take a varying and combination of approaches to solve this problem. And that might be a good thing, as it will add some variance to the the presentation of adventures.
I run D&D like a TV show. TV manages to tell a complete story in an hour, with commercial breaks. TV manages to demonstrate a Status Quo type campaign, where the end of the episode has the problem resolved and the crew all safely aboard the USS Enterprise, or more episodically, where the situation continues to evolve session to session. However, even the episodic style tend to tell a complete story within that hour, while advancing a larger story across multiple sessions.
the trick is pacing and content balance. We know D&D 3e combat takes about an hour. We know that my group can only meet for 4-6 hours.
Therefore, we only put in 3-4 combats in the adventure, so there's room for setup, dialogue and problem solving. Too many D&D adventures are designed to take forever, when the typical movie or TV director can tell the same story in a few hours. As a game, D&D rules resolution is going to take a bit longer than a TV scene. That's OK. But typicial D&D compounds that by packing in way more material than is needed to tell the story and challenge the group.
So the goal is, create satisfying content that can be completed within 1 session (4-6 hours in my group) that can handle a changing roster of PCs as players do/don't attend subsequent sessions.
I suggest that the first stage of any campaign should emulate Star Trek. The "group" goes out and deals with a problem in 5-7 scenes. Some of those scenes are combat enounters. Just like Star Trek, the away team consists of whoever showed up on set for filming. if Dr McCoy ain't available, he doesn't appear, or has a few minimal lines voiced by the GM to say "He's dead, Jim."
In D&D, the Enterprise is analogous to the PC's home village/town. It provides a backdrop of set pieces for in-town action, and external threats can be raised to get the PCs out in the woods or dungeons as well. Because it is a "big" place, PCs who aren't available can easily be explained away as busy doing in-town things. This is a bit smoother than ending the last session in the Dungeon of Disastrous Doom and need to explain why Chad's Barbarian isn't here anymore.
If you keep the scope of your adventures paced the same as a TV show, it will tend to fit that the players solve the problem and return to town by the end of the session, thus enabling you to start the next session, back in town, with a week or more passing that enables a larger variety of reasons why Chad's PC can't be here for the next adventure.
By starting all your campaigns in this Status Quo way, you also get to develop and test your team of players, before you overcommit on large PC-centric plots and such. Find out who's reliable and who's a crazy nutjob player. Plus, after doing 2-6 of these kind of adventures, the PCs will have made some friends and enemies in-game, that you can then use to change the format to episodic in a smooth fashion.
Once you have your group broken in to the new format, it'll be time to advance it and change things up. Once you know your group is reliable, try a cliff-hanger ending, where the problem can't be solved in a single session. This will be a good change of pace, and is safer to do when you know your group is going to make the next session.
You can also start changing up how things end. Just like Star Trek:TOS, it can be nice to save the day and get things back to normal, but after 1-3rd level, it's time to mature things up and have weightier consequences for the events of the session. This means that the party may return back home, after each session, but things have changed due to the game's events.
You should almost always try to end things back to a certain location (the ship, the town, etc), but the overarching situation doesn't stay Mayberry perfect.
In most TV shows, you'll see this in the form of 2 story arcs going on. In ST:TOS/TNG, the pattern was a problem for the ship, and personal problem for one of the cast.
In shows like Burn Notice, it'll be the "Client of the Week" and a few scenes making progress toward finding who's behind the burn notice.
In Buffy, it's the same pattern, Monster of the Week, and a few scenes dealing with the mystery larger threat behind the Monster of the Week.
If the GM follows the pacing patterns of such shows while building content for a session, they will be better able to meet the goal I outlined, which I believe meshes with what Rechan asked for.
Note, my advice tends to approach it from a story-oriented style of GMing, than a Sandbox style. By Sandbox, I mean, the players decide from a plethora of opportunities, on where to go, and what to do. I don't think they have to be mutually exclusive, but some of these ideas can be applied to Sandbox style, and some Sandbox style ideas can be applied to this.
Sandbox vs. Story is really like Buffet vs. Mom's Dinner. Buffet gives lots of choices, meaning more work for the cook. Mom's Dinner means we are eating what Mom prepared, and she probably put a lot of work into it.
As a GM, if you're going to limit the choice of meal, you've got to choose well. Make something that will appeal to the players, and not force them down a path they don't want to go. In my group, we have a general rule that the party will bite the plot hook. But the GM has a responsibility to not screw the party over for doing so. I've played in a published adventure with that premise that all of us players saw as "this is a total screw job" but the basic rule of trust was that we would bite the plot hook so we could get to the fun. Hence, the corollary that the GM must not violate that faith the players put in by ignoring their better judgement, for the purpose of expediting game play.
In a sandbox, I don't see why the GM can't make a multitude of opportunities that can each be resolved within the same timeframe. The core concept is about eliminating time wasting content, not choices or freedom.
I run D&D like a TV show. TV manages to tell a complete story in an hour, with commercial breaks. TV manages to demonstrate a Status Quo type campaign, where the end of the episode has the problem resolved and the crew all safely aboard the USS Enterprise, or more episodically, where the situation continues to evolve session to session. However, even the episodic style tend to tell a complete story within that hour, while advancing a larger story across multiple sessions.
the trick is pacing and content balance. We know D&D 3e combat takes about an hour. We know that my group can only meet for 4-6 hours.
Therefore, we only put in 3-4 combats in the adventure, so there's room for setup, dialogue and problem solving. Too many D&D adventures are designed to take forever, when the typical movie or TV director can tell the same story in a few hours. As a game, D&D rules resolution is going to take a bit longer than a TV scene. That's OK. But typicial D&D compounds that by packing in way more material than is needed to tell the story and challenge the group.
So the goal is, create satisfying content that can be completed within 1 session (4-6 hours in my group) that can handle a changing roster of PCs as players do/don't attend subsequent sessions.
I suggest that the first stage of any campaign should emulate Star Trek. The "group" goes out and deals with a problem in 5-7 scenes. Some of those scenes are combat enounters. Just like Star Trek, the away team consists of whoever showed up on set for filming. if Dr McCoy ain't available, he doesn't appear, or has a few minimal lines voiced by the GM to say "He's dead, Jim."
In D&D, the Enterprise is analogous to the PC's home village/town. It provides a backdrop of set pieces for in-town action, and external threats can be raised to get the PCs out in the woods or dungeons as well. Because it is a "big" place, PCs who aren't available can easily be explained away as busy doing in-town things. This is a bit smoother than ending the last session in the Dungeon of Disastrous Doom and need to explain why Chad's Barbarian isn't here anymore.
If you keep the scope of your adventures paced the same as a TV show, it will tend to fit that the players solve the problem and return to town by the end of the session, thus enabling you to start the next session, back in town, with a week or more passing that enables a larger variety of reasons why Chad's PC can't be here for the next adventure.
By starting all your campaigns in this Status Quo way, you also get to develop and test your team of players, before you overcommit on large PC-centric plots and such. Find out who's reliable and who's a crazy nutjob player. Plus, after doing 2-6 of these kind of adventures, the PCs will have made some friends and enemies in-game, that you can then use to change the format to episodic in a smooth fashion.
Once you have your group broken in to the new format, it'll be time to advance it and change things up. Once you know your group is reliable, try a cliff-hanger ending, where the problem can't be solved in a single session. This will be a good change of pace, and is safer to do when you know your group is going to make the next session.
You can also start changing up how things end. Just like Star Trek:TOS, it can be nice to save the day and get things back to normal, but after 1-3rd level, it's time to mature things up and have weightier consequences for the events of the session. This means that the party may return back home, after each session, but things have changed due to the game's events.
You should almost always try to end things back to a certain location (the ship, the town, etc), but the overarching situation doesn't stay Mayberry perfect.
In most TV shows, you'll see this in the form of 2 story arcs going on. In ST:TOS/TNG, the pattern was a problem for the ship, and personal problem for one of the cast.
In shows like Burn Notice, it'll be the "Client of the Week" and a few scenes making progress toward finding who's behind the burn notice.
In Buffy, it's the same pattern, Monster of the Week, and a few scenes dealing with the mystery larger threat behind the Monster of the Week.
If the GM follows the pacing patterns of such shows while building content for a session, they will be better able to meet the goal I outlined, which I believe meshes with what Rechan asked for.
Note, my advice tends to approach it from a story-oriented style of GMing, than a Sandbox style. By Sandbox, I mean, the players decide from a plethora of opportunities, on where to go, and what to do. I don't think they have to be mutually exclusive, but some of these ideas can be applied to Sandbox style, and some Sandbox style ideas can be applied to this.
Sandbox vs. Story is really like Buffet vs. Mom's Dinner. Buffet gives lots of choices, meaning more work for the cook. Mom's Dinner means we are eating what Mom prepared, and she probably put a lot of work into it.
As a GM, if you're going to limit the choice of meal, you've got to choose well. Make something that will appeal to the players, and not force them down a path they don't want to go. In my group, we have a general rule that the party will bite the plot hook. But the GM has a responsibility to not screw the party over for doing so. I've played in a published adventure with that premise that all of us players saw as "this is a total screw job" but the basic rule of trust was that we would bite the plot hook so we could get to the fun. Hence, the corollary that the GM must not violate that faith the players put in by ignoring their better judgement, for the purpose of expediting game play.
In a sandbox, I don't see why the GM can't make a multitude of opportunities that can each be resolved within the same timeframe. The core concept is about eliminating time wasting content, not choices or freedom.