by Mike Shea on 15 July 2013
In one of the most popular D&D adventures of all time, Tomb of Horrors, Gary Gygax described summoned demons and imps who would travel through the dungeon and periodically reset all the traps for the poor unwitting adventurers who would soon travel within. It was an easy way to explain how a falling floor pit trap would just so happen to be ready for your own PCs when they stumbled on it. Tomb of Horrors worked because, as crazy as it was, it made sense. It was a living place even when the PCs weren't there.
Today we're going to discuss the fine art of designing living dungeons — dungeons that make sense and operate outside the context of the PCs. How can a massive series of tunnels stay dry below sea-level? How can totally different ecosystems co-exist within a chamber buried for ten thousand years? That's up to us to figure out.
We often begin by looking at dungeons as a simple vehicle for adventure. In many cases this is perfectly fine. The primary intent of a dungeon is to bring adventure to PCs. It can quickly fall apart, however, if people start to scratch beneath the surface and say "this doesn't make any sense". How can a bunch of brigands hide behind a door with an ancient and powerful arcane lock? Have they been sitting there for ten thousand years waiting for us? In video games, this became known as a monster closet — a cubby containing a monster whose sole purpose was to leap out at our protagonist and be summarily killed. It's a poor life to live.
We want to do better. We want our dungeons to feel like real places. We want them to make sense in the context of our game's world.
We start by asking about the origin of our dungeon. Is it a wizard's inverted tower sunken into a mountain? Is it an extra-planar prison? Is it a series of caves taken over by kobolds? Once we know what the dungeon actually is, we can build it out and run it as it makes sense for that particularly story.
Even megadungeons like Greyhawk or Undermountain can still follow this logical model. Instead of a single purpose for the whole dungeon, we simply segment the mega-dungeon out into separate areas connected through tunnels and chambers. Each section has its own purpose.
Here's an idea to consider. What if every dungeon you introduce in your game is actually just a piece of a mega-dungeon? What if there is no end to the chambers and tunnels that lie below? Following the concepts of running faster than the other guy, what if we always ensured there were always unexplored chambers and deeper levels than those found by the PCs? We might not need to reveal it all at once. We might not even give them direct access to it. It would be cool, however, if the PCs always knew there was just one more unexplored set of chambers lurking just below.
Building any realistic element to your game comes from adding layers. A living dungeon is no different. We began by asking what it's used for now and we build it around that idea. For an added layer, though, we can ask "what did it used to be?". Maybe these prison chambers below the old castle were once the laboratory for a dark wizard long dead. The walls still contain the etches of his mad scrawlings and some say, when you put your ears to the stone, you can still hear his whispers.
Adding a former purpose for the dungeon gives players a real sense of the depth of something. Now it's a hideout for brigands. Before, it was a secret ritual chamber to a god long dead. Of course, the hard part is that now both origins need to make sense for the layout of the dungeon.
A living dungeon continually moves and evolves. Things are going on in there, like some sort of ancient factory. It's not always enough to simply say that "magic" keeps the dungeon operating throughout the millennia. The dungeon needs to work. Heat, wind, and water are great sources of energy to power a dungeon. A dungeon may have a series of pipes that lead down into a boiling spring well below the chambers. The boiling and cooling water flows through the pipes and pushes doors into place, resets pit traps, and builds up pressure in deadly boiling water trap rooms. Water flowing through an underground river could power massive waterwheels that pump water out of lower chambers and move massive stone doors throughout the dungeon. Wind power could also generate all sorts of movement throughout a dungeon, powering massive electrical generators.
You don't need to know how all this stuff works perfectly, but giving hits to your players that these sorts of things power the inner workings of a living dungeon gives them the idea that it can work, which is just what we need.
In a description of the giant Rube Goldberg machine in the Ok Go video for This Too Shall Pass, Adam Sadowsky, the engineer of the machine, described how he they could get such a complicated apparatus to work. "Small stuff stinks", he said. It's not hard to get a bowling ball (or a car) to go where you want it. The bigger the objects in motion, the less precise they had to be. There's no reason we can't steal this for our living dungeons.
How do stone pumps use volcanic steam to pump water out of a subterranean dungeon for ten thousand years? Make the pumps really really big. What if the pumps are stone cylinders over a thousand tons big powered by massive underwater pressure boiling from fissures in the bottom of the water?
When you're building a dungeon, think of the dungeon as a villain itself. When running White Plume Mountain, what if the dungeon itself is the new embodiment of the wizard Keraptus? What if the whole dungeon is his phylactery? Think of a living dungeon as a massive apparatus through which PCs venture; an apparatus that wishes to digest them and fuel its own growth. It may be figurative or it might be literal, that's up to you.