When looked at through the lens of Vector Theory; the simple, branched and interlinked paths are just some of the possible ways a game scenario can be designed. But in a traditional game design context, these are basically the only acknowledged forms of scenario design.
Even in current indie-design groups, it seems that a game is either structured according to one of these methodologies, or it is an amorphous entity that only resolves into some form of structure in retrospect.
I'm going to propose a couple of other options. Nothing radical, and once you see them diagrammed out (or explained), you'll probably see that a few of these game structures have already made a presence in games you've experienced.
The interlinked path with multiple endings is another fairly common design format. It's actually something that a lot of computer games do now; a few years ago it was considered revolutionary to have multiple endings based on game-play decisions in a computer game. Plenty of game modules and pre-written scenarios follow this type of set up, it's fairly common in roleplaying conventions as well. One ending might be victory (with a specific text-blurb to read out), another ending might be failure (with a separate text blurb).
Another common twist on this theme is to have a strict structure, but leave a spectrum of possible endings for the characters to encounter depending on the objectives they resolved during the course of the story. Nothing revolutionary here, so let's move along.
The difference between the fully and partially interlinked paths is subtle but important. In the example shown, the central path through the story is the most expected one, but deviating to the left or right are possible. The key is that once you have moved to the left at one level, the rightmost path with become unavailable in the next level.
Let's look at a specific example.
In the introduction, the characters meet an NPC, they can fight him (left path), use his information and move on (middle path), or they can befriend him (right path).
The impetus has the characters meeting the NPC's enemy. The leftmost path describes a scene where the enemy is friendly to them, the rightmost path describes a scene where the enemy is actively hostile to them, while the central path has the enemy simply sizing them up.
If the characters had befriended the first NPC, the second wouldn't befriend them. You simply can't go from the right side to the left without passing some kind of middle ground, the story wouldn't make sense. But there are still choices available, the enemy could be sizing them up (middle path) or he could be outright hostile(left path). The choices are reversed if the characters fought the first NPC.
If the characters followed the central path, the most options are available to them (but the story loses some of it's dramatic twists and turns).
This has still been pretty rudimentary stuff to anyone who has GM'd more than a couple of sessions. Now, let's pull in a few more Vector Theory specific concepts.
Mirrors were mentioned in some of the early Vector Theory posts (here and here). They could be deliberately included as a part of a scenario layout, or they could be added on-the-fly if a group of players starts to deviate too far from the story a GM is trying to narrate. The diagram provided above is fairly typical of many story situations; you might have a couple of choices to follow, but if you choose something that the GM hadn't planned for, then a gentle nudge will be required to put things back on track. Some of the better pre-written scenarios I've read have included contingency plans to get things back on track ("Have the players encounter an avalanche to prevent them heading out of town", "an oracle gives them a warning", "the character's conscience tells them not to follow this course of action", etc.).
Using Mirrors is always dangerous.
If they are used clumsily, a group will accurately perceive the event as a form of railroading. They will feel that their decisions don't matter, and they they had better simply follow the GMs story to avoid problems.
If they are used well, a group of players won't even realise that they've encountered a mirror, they'll simply think that they've chosen an unexpected path that has opened up a new opportunity for them. But they might expect that such lateral thinking will always provide them with new possibilities, and the GM who has used a mirror well once might be expected to use it just as expertly on future occasions.
On the other hand, not using mirrors can be just as dangerous. Simply telling a player "Sorry you can't do that", just because it deviates from your storyline is a really immature way to play (IMHO). Traditional players who like to be led through a story from start to finish won't look for lateral solutions to problems so they probably won't notice mirrors when they hit them, they might just think to themselves "Cool. I didn't see that coming", or they might just say "Yeah...but what dice do I roll now?". Indie players looking for the nodes and decision points in a story will probably think to themselves "I didn't think I'd pull that off" or "S%*t, I didn't manage to break the GMs game, I'll have to try harder"...
A few more Vector Theory specific ideas for module design are coming soon.