THEORY OF ADVENTURE DESIGN

I've said time and again that the Arcade Ruleset is intended to provide fast and fun gameplay. There's an emphasis in being GM friendly - as his is always the worst part - providing a set of rules that are very easy to learn and general enough to convey most situations.
Yet, this is a game, and is governed by a set of rules. Those rules will dictate what can be done, and what don't. And that limit should be taken into account when designing an adventure; as this is not a precise simulator of real life, there are many things that you can imagine that won't be playable with these rules. So, before making an adventure, try to concentrate on the rules for a while, because they are the "language" in which the adventure must be written.

Basically, in any roleplaying game there are two types of encounters: combat encounters, and non-combat ones. With my ruleset the latest can be subdivided into skill test encounters, and powers encounters. A combat encounter is what it sounds: a fight on any kind, that will risk the lives or wellbeing of its participants. A skill test encounter means that a character has found some kind of obstacle, and to be able to do what he wants must proof himself valid. A powers encounters (in lieu of a better denomination) is similar to a skill test one, but with a biggest emphasis on the protagonism of a character: his will be the spotlight, showing that it is thanks to his special uniqueness that obstacles can be overcome and the plot advanced.
And for an adventure to be fun it should be as varied as possible. Which means that all three kind of encounters should be used in any adventure. Don't have the PCs always fighting for their lives, but threat them every now and then. Similarly, show them there are many difficulties to face in the world, but that many of them can be overcome. And this is very important: make sure every player have fun by giving their PCs their fifteen minutes of fame. One of the good points of being a GM is that when you write an adventure you already know the PCs; having that in mind, always write at least one encounter per PC in which he may shine; if possible, make them two. It's nice for a player to feel his character has made an essential contribution to the adventure.

But how should an adventure be structured? In my opinion the best way to do so is using a cinematographic approach, which means thinking the adventure in terms of scenes. Write the adventure as if it were the script of a movie; that means structuring it with the typical introduction-plot-conclusion scheme, too. The introduction serves two goals: getting the PCs know each other (unless we're talking about an ongoing campaign) and involve them in the adventure. Don't be afraid of breaking the old cliche of starting with all the PCs together and take for granted they already know each other; introducing them in successive scenes will give much more realism to the story.
But, of course, a rpg is not the same as a movie, because you must take into account a key element: those perky players, and their weird decisions. That said, the introduction and the conclusion scenes can be pretty much rail-roaded without it affecting the game's playability; it's what the players will expect, after all.
The plot will be a whole different matter. It is there where there's never any single way to do things; even if you think the "right" decisions for the PCs are very obvious, they are not. More, there's no such a thing as "right decisions", because during the development of the plot the game no longer belongs to the GM, but to the PCs. After all, they are the protagonists. What I'm trying to say is that, even if it's good to present ways to advance in a very obvious way, you can't seriously expect that the PCs will follow a linear plot (unless you force them, a bad habit that takes most of the fun from a game).
And all that means you'll have to write extra scenes, quite a good bunch of them. And they can be of any of these three kinds: alternate paths, dead end paths, or floating scenes.
Alternate paths, as the name suggests, are different ways to reach the same place. They can make things more dangerous or more safe; quicker (less scenes than the "original" plot) or slower (more scenes). With alternate paths is important to tie their benefits or perils with good or bad playing on the PCs part; good alternatives should be the result of players using well their characters, and vice versa. And there's always the same old alternative: left or right corridor? You, the GM, know they are the same, but it will give the players the sense that their decisions can influence how the game goes. Infact, it would be good if you considered the "original plot line" just as another alternate path, and take on writing the adventures including many alternatives since the beginning.
Dead end paths are either side quests or false clues (these can be misinformation or misunderstandings too). They are a series of scenes that won't later connect with the main plot, so what's the point with them? Actually, they are pretty useful, and serve different purposes. The first purpose is slowing things when they're going too fast; the Arcade ruleset is intended to provide a fast gameplay, but sometimes the PCs are too good and the adventure is menaced to be ended within an hour; trying to make them follow a side quest will ensure a little more time. A second purpose is adding difficulty to the adventure; evil guys should not be stupid, and they sure will try to mislead would-be do-gooders. Or adding ethical questions; for example: the main plot may require the PCs to act quickly, but won't they spend a little time to save an innocent? Yet a very important use for dead end paths usually overlooked is giving realism to the adventure. Sure, the PCs are the protagonists, but that doesn't mean the world revolves around them; foiling them into leading what finally has nothing to do with them or the main plot will give them, and the adventure, a nice reality bath.
Finally we have the floating scenes. They are a more evolved equivalent of what in D&D tradition were the random encounters. They are very similar to dean end paths in that they don't really belong to the core of the plot, and the uses they can be given. The main difference is that floating scenes have no predefined starting point, and can be thrown at the players almost at any moment. In order for that to work properly they must be short and very generic. In fact, they can be so generic that you may recycle them from one adventure to another, and always keep a bunch of floating scenes made for a setting. For example: in Ajsalium, heroes may be attacked by a hungry griffon whenever they are in the wilds; in Necronomicon, the players can suffer a theft attempt whenever they are just going anywhere; or in Galaxtar, hyperspace jumps can always go wrong and put the adventurers in the middle of an asteroid field.

Has all this rung a bell to you? Likely it has, because it's has been done before: "choose your adventure books". Very popular for teenagers, they are a full genre based on the concepts I've just exposed. So if you want a good example just grab one of these books from your collection (or ask your friends for one, if you don't own any, what would be rare for a roleplayer, that is what I assume every one following this blog will be).