- Be realistic!
Don't attempt to write a 400 pages campaign if you've never written a short adventure before. Start small and build from there. Starting small allows you to gradually build up your campaign, refine your themes and ideas, those things do not come overnight. Until someday you wake up with a 400 pages campaign! Crawl, walk then run. This is valid for both veteran and novice writers. Think small and expand. If you think too big, things sometimes get very blurry at a granular level.
I have found that the basic idea I started to write gets modified, replaces or even discarded as the creative process happens.
- Points of Light
The term "Points of Light" was coined by Wizards of the Coast as part of their enthousiasm-creation for D&D 4th Edition. Although the concept is not new, I really like the name. The theory behind Points of Lights is that an adventure contains just enough to run an adventure. Over the course of many adventures more and more come into the light.
Thus you start your PCs in the small village of Hommlet. There, they learn that bandit infest a nearby Moathouse. And then that those bandits are in league with the temple of Elemental Evil...By the end, the PCs have visited a number of previously unknown or mysterious locations all surrounding Hommlet. Yet each location was discovered gradually, thus not overwhelming the PCs with monsters and challenges they could not realistically face from the start...
- Cover and address the most likely courses of action / KISS
KISS stands for Keep It Simple Stupid (not just a great rock band) This principle should apply to everything. If something is COOL but does not really ADD to the adventure, then drop it. Copy it somewhere you will review later. A cool idea can be cool in a number of places, not necessarily your *current* writing project.
Do not try to tackle every possible outcome and every possible way the PCs can crash through the adventure. You need to cover the main 80% of possibilities and leave the other 20% up to the DM. Not only is that impossible, but it is also a waste of time. You want to give the DM the tools to ad lib in a way you would like the adventure to proceed, but that is about the extent of what you can do.
Your adventure brings the PCs to a fork in the road. The adventure and clues should lead them to the left path. You need to describe what they encounter there (obviously). Some PCs may head down the right path for a number of reason: missed clues, just to be annoying, etc. The other option is simply not to detail the fork (see KISS, later). That PCs head down the left or right comprise the 80% possibility. There should be little need to detail what happens for those parties who decide to go straight, or go back where they came.
In the above example, mentionning that the forest where the fork is has a lot of hungry wolves who attack people leaving the road. This gives a good idea to the DM what the PCs might encounter if they leave the road without any more details (and your 80% target just went up to 95%!)
- Many shorter adventures are better than a massive one
This one I want to drive the point to you the reader... It is better to write many shorter adventure than one big, never-ending story. Whenever possible, try to break your super-world-shattering adventure into a series of smaller adventures. Not only is a climax easier to build through a series of adventures, but also it allows for good break points, for a DM to place some transition adventure(s) of his own design in-between. In short, it allows for breathing room.
Good Gaming everyone...