Let us imagine that you’ve decided you want to run a sandbox-style game, and you have chosen an appropriate system. Now, you have pads of graph paper and hex paper printed off, and you are beginning to imagine what you want your sandbox to be like. What now?
For me, the first step toward setting up anything is to brainstorm ideas. I’ll get some scrap paper, and just write down whatever ideas I come up with. What kind of cultures do I want in the campaign milieu? What kind of creatures? People? Situations? Even single encounters, if they occur to me, get written down. From these notes, I am attempting to create four things:
1. An initial base of operations,
2. An interesting outdoors area to explore,
3. An overview of the region that the initial sandbox area is part of, and
4. At least three major and six minor adventure sites.
We’ll be looking at each of these in order, but it is important to remember that they are interlinked. One of the reasons that I brainstorm first is so that, for example, I have a strong idea of what my adventure area and wilderness will consist of, and can include clues to, say, the Tower of Amoreth the Arcane in the background of the initial starting area and the wilderness. In addition, adventure sites can interlink, offering (a) pointers for the players, to help them determine where else they can explore, and (b) “Aha!” moments when things learned in two or more locations suddenly point to a larger whole.
Point (a) is important because it helps the players to make choices. If you go back to the “Choices, Context, and Consequences” blog posts, you will recall that the Game Master needs to supply information (context) for player choices.
In the initial area, ferreting out these connections shouldn’t be too difficult. In expansion areas, where the players are presumably growing accustomed to seeking out and putting together scraps of information, the clues can become harder to discover/piece together. The goal is to make information available, and to have enough information available that the players will gain access to a reasonable amount of it. OTOH, the information should be difficult enough to gain/use that doing so gives the players a feeling of accomplishment. Lots of information, hidden with varying degrees of difficulty, is the best way to accomplish this.
(In early TSR modules, treasure was used the same way. Lots of treasure, hidden with varying degrees of difficulty, ensured that the players would both find treasure, and feel a sense of accomplishment based on the most difficult-to-locate treasures they uncovered. If the GM didn’t tell the players what they missed, that sense of accomplishment – as well as the sense of the module taking place in a mysterious area – could endure. It is only when the GM felt the need to tell the players what they missed, or the players to read the module later, that this set-up became damaged. You can easily prevent this problem in work you write yourself!)
((Failure to understand this set-up is one of the reasons that certain analyses of early TSR modules, and especially comparisons between early TSR modules and WotC 3e and 4e modules, fall far short of the mark.))
The foregoing also explains point (b) rather well. “Aha!” moments are (among other things) a reward for good play. In an initial area, “good play” should have a fairly low bar, so as to encourage play, exploration, and decision-making.
This is not to say that a cakewalk is desired, because a cakewalk offers no sense of accomplishment. Rather, again, a sliding bar is desired, where any effort includes rewards at its fringes according to the effort put in. Everyone gets rewarded; better play is rewarded more.
Likewise, if your overview includes the idea “Ancient Aztec civilization was once in area now being raided by Vikings” as a “hook”, it is critically important that your initial area includes elements both of the current Viking raiders and of the ancient Aztecs who were once there. The overview exists to guide your initial work, allowing you to foreshadow a larger world.
Note that, if your campaign milieu is going to include a megadungeon, I recommend that this lies outside the initial set-up area. Characters and players should have a chance to get their feet wet in the milieu before entering such an area.
If you think of The Hobbit, Hobbiton, the trolls, and Rivendell all may be considered as part of the initial set-up area, before entering the much more complex goblin mines and Mirkwood. Likewise, much occurs in The Lord of the Rings before the Fellowship encounters Moria.
If you intend on including a megadungeon, though, you can certainly include links, hints, and rumours in the initial area. This is actually a good idea. In fact, hints of any expansion areas you are already envisioning should be included in the initial area. You want the players to consider a larger world almost from the beginning.
Next: Initial Base of Operations.