I’ve thrown around a lot of theory, but how does that come together in practical terms? Here’s the outline it implies when trying to create a dungeon.

1. Decide the dungeon’s purpose in play.
2. Decide the dungeon’s identity within the game world.
3. Using real-world references, random generators, or your own drawing skill, get a map.
4. Stock the map with challenges, puzzles, rewards, and any other details you want to add.
5. If inspired to do so, repeat steps 3 and 4 (adding new space and content, respectively) until satisfied.

1. Dungeon design should reflect its purpose. If it’s just there to contain a single encounter or treasure, or just as a random cave (to shelter from the rain in, for example) or building, it should be small and simple; a room or a handful of rooms. If it’s a megadungeon intended to take the players through several levels of gameplay, it should be a multi-floor complex with a complicated, interconnected floor plan, secret doors, traps, and foes and obstacles throughout.

Venues intended for combat should have interesting architecture: large open rooms with furnishings and features that can be used creatively (balconies and tables, pillars and pits). Venues intended for puzzle-solving should be built around the puzzles (perhaps taking a cue from computer games, in terms of room design). Venues intended for exploration should have lots of secrets and interesting features.

2. Think for a moment about the dungeon’s location, origin, and use; these will all inform both its layout and features.

Where is the dungeon? The surrounding terrain will be reflected in its design. For example, caves or underground complexes will often be made of the local rock – or need supports if built in soil. A fortified bridge over a waterway will be split, and defensible from both ends. Even a building in a city will vary depending on its surroundings: built on a slope, it might be partially embedded in the ground, so that the first floor is also a basement. In a densely packed city, buildings will be narrower and more vertical than in a scattered town.

How did the dungeon come to be? I don’t just mean how the space was first created – natural cave or animal burrow, or its construction – but also how it got to its current state. Perhaps a natural cave was worked and reinforced by its later inhabitants. Perhaps the castle has fallen into disrepair, with blocked passages and crumbling towers. Perhaps the mansion has passed through the hands of a dozen families and generations, with each adding or renovating a little according to their tastes and needs.

Be careful here. I’m not saying that each dungeon needs a complicated, complete history before you can start designing it. This step need be no more than a general feeling or a couple sentences, but anything at all will help you design and stock it coherently.

3. Whether you generate it randomly as you go or hand-draw it lovingly in several colors long before play starts, a dungeon will generally need to be mapped by the time the players are done with it. There’s nothing wrong with looking for maps other people have made and repurposing them for your dungeon, even tweaking them to fit your needs more closely. I recommend, for starters, the maps of the inimitable Dyson Logos. Especially if you’re trying to scatter a number of dungeons around a large map, this can save you a lot of time.

Let’s say you’d prefer to draw your own by hand. As with writing, it’s next to impossible to tell you how to draw a good map; that’s partially a matter of your own goals and style. If you want to learn how to draw maps to your satisfaction, though, here are the standard tips for would-be writers:

A. Find references. If you want to write, read a lot; if you want to draw maps, look at a lot of maps. Don’t just look, of course. Think about what you like and don’t like about them, and why. Not the art style – although that’s something you may want to try to copy, it’s also not necessarily going to be shown to the players – but the construction. Borrow good ideas, and poke at bad ones to figure out why they’re bad.

B. Practice. If you want to be good at making maps, make a lot of maps and expect most of them to not satisfy you. There’s nothing wrong with that: since it’s helping you refine your skill, the time’s not wasted, so at worst you’re only out a trivial amount of scratch paper and ink or pencil lead. When you’ve got some maps that you like the look and feel of, try running the players through them and see how it works out. This is also part of your practice. If you don’t play-test maps, it’s hard to get a feel for what works and what produces a satisfying session. Don’t worry if you make some maps that fall flat; the important thing is that you’re creating and learning and improving, rather than just sitting there wondering how to draw a nice dungeon.

C. Use feedback. Using what you’ve learned by looking at maps, drawing or stocking your own, and using them in play, think about what works for you, and how. Think about what the players liked, and what they rushed past or seemed bored, confused, or frustrated by. If you’re feeling really brave, ask them directly for constructive criticism. Keep all this new data in mind next time you try to create – but as before, don’t let it paralyze you; don’t be afraid of making mistakes.

4. Stocking your map should be relatively easy because it should follow from the other steps: the atmosphere, the enemies, traps, puzzles, interesting odds and ends, and loot or other rewards will often suggest themselves based on what you’ve already made. If you’re feeling stuck, though, try making a list of things that you can put in. Brainstorm for a while. Add things that specifically don’t match the rest of the atmosphere and then reconcile the contradiction, to make things a little more interesting. Which brings us to…

5. Don’t be afraid to go beyond your original design ideas, even if it means tweaking or completely overhauling the dungeon’s seeds from step 1! Maybe in brainstorming, you thought of something interesting you want to add, but the dungeon needs to be reworked for it to fit. There’s nothing wrong with doing that. If you can’t bring yourself to, well, then you’ve got the seed for another dungeon already growing; that’s great!

Next time: some examples illustrating what I mean.