Designing Dungeons Part 3
This is the third and last piece that I’m going to write about dungeons (at least for now). They’re on my mind, because I’m designing a megadungeon right at this moment.
Maps
When you think about dungeons, you probably quickly get an image in your mind of some twisty, almost maze-like rooms and corridors. And that’s great. Maps are a fun part of dungeon design and play. When people started making dungeons in the early days, they tried to make mapping the dungeon a challenge for players. The idea was that it might be fun for the players to get lost because a spinning room changed their orientation, or a gradual descent in a long hallway put them deeper underground than they thought.
I’m going to argue, however, that attempting to create challenge and fun by disorienting players doesn’t actually work in today’s games, if it ever did. It rarely results in the PCs being lost in a dangerous place and instead results in frustration and arguments at the table. The art of communicating the dungeon’s description to the player tasked with sketching a map for the players to reference is a difficult one already, without putting in ways to make it even harder. That’s why I hate mazes in dungeons, and why, when I put a maze in my current project, I give the GM a way to handle it narratively with some die rolls to see how quickly the characters get to various points. That way, there’s no, “okay, and then the right-hand path goes for 10 feet and turns to the left, and the right-hand path goes 10 feet and then comes to a T intersection. No, the right passage goes to the left. Wait, no…”
And of course, if the group is using some kind of VTT, dungeon terrain like Dwarven Forge, or similar visual aids, the designer’s tricks aren’t really going to work anyway.
But sneaky tricks aside, what makes a good dungeon map?
Lots of designers talk about “Jaquays-ing the Dungeon,” referring to the work of my friend Janell Jaquays, who is sadly no longer with us. I love that term. Simplified and taken to its broadest applicability, you might say this means, “Ensure there are multiple ways to reach most places in the dungeon. Rather than just one linear path to get from the entrance to the end, connect the rooms or encounters with multiple passages or connections.” Your map will look like a series of looping paths rather than just a line from room 1 to room 2 to room 3, etc. I’d take this even a bit further, actually, and discard the idea of a dungeon entrance and a dungeon end. Put in multiple entrances, with some only discovered after some exploration of the dungeon, like a secret passage discovered from inside rather than from the outside of the dungeon, or a locked door where the key can only be obtained after exploring within.
There’s a certain kind of art to drawing a dungeon map. I don’t mean that it takes drawing talent—anyone with a sheet of graph paper can do it. I mean the layout of the corridors and rooms, the shape of the rooms, and the way it all fills the page can be very visually pleasing. Look at other published dungeon maps and study them simply from a composition point of view. Find the ones that look the best to you and learn from them. A pleasing dungeon map has a compulsion built into it. It makes the GM want to run that adventure and makes a player want to explore.
Sure, you’ll hear people complain that the traditional winding dungeon map isn’t realistic, and I could come up with a bunch of geology-based reasons why it’s not as unrealistic as you might think, but who cares? It’s appealing, interesting, and cool, and when you’re making products for other people, those are important considerations.
The whole graph-paper-dungeon concept has one real drawback. And that is because you’re drawing on a flat piece of paper, your dungeon will be flat. But one of the best things about a subterranean locale is that it can easily exist in three dimensions.
Give the dungeon lots of verticality. Stairs are great for this, obviously, but ledges, drop-offs, balconies, galleries, pits, sinkholes, chimneys, and ramps, all add interesting texture to the dungeon. Think about the ceilings as much as the floors and walls. A hatch (or a secret hatch) on the ceiling is always intriguing.
Presentation of Information
Look, I think a lot of us have come to the realization that the “wall of text” is not GM-friendly, despite how guilty I and the next designer who’s been doing this for 38 years or so are of writing such blocks of text. (What’s that, there’s hardly any designers still doing this?… I’d rather not dwell on that.) It’s hard to quickly find what you need, and at the table, easy reference is everything. I’ll even take the next step. Maps with a numbered key are not very GM-friendly either. What is GM-friendly is putting the text the GM needs right on the map itself. Then there’s no page flipping between the map and its relevant text. If it sounds like there wouldn’t be enough room for the designer to say what they want in or next to the map, it might be a good exercise to see what’s really necessary for a room or area description.
The GM needs to know:
Is there light?
Is there a dominant smell or feeling in the room? For example, is it damp, cold, or dirty?
Is there something alive or moving in the room? What’s it doing?
What furnishings or features are in the room? (Start your description with the most visible or obvious thing.)
If there are locked doors, traps, hidden things, or interesting contents inside a feature or furnishing?
The way I’m solving the space issue in what I’m working on now is that, in addition to a big overall map (or multiple such maps), I’m zooming in on a small section of related areas. I can fit about six to ten areas and the text describing them on a two-page spread.
In this project, each cluster of rooms will get a page of background, lore, and notes about what happens to the area should the PCs change things and come back. This is where I can go on at greater length about tactics and whatnot too. This page is meant to be read ahead of time to understand the detailed areas, but it’s likely unnecessary to refer to it during play. Then there’s also a page of game stats for creatures and items.
Take a look at this unedited rough sample.
Non-Combat Challenges
I’m going to close this series of articles by simply pitching that while we often associate dungeons with “hack-n-slash”-style gameplay, I prefer to design dungeons where that’s not the assumption. I like weird things to interact with, strange locations to explore, and mysterious objects to investigate. Maybe some fighting too, although preferably with some interesting tactics or opportunities for the PCs to use non-standard means of bypassing or overcoming a challenge.
I remember after I wrote Return to the Temple of Elemental Evil, players came to me and said that it took forever to get through all those fights. My response was usually, “You fought everything?” Even back then (and that was 20-some years ago—I’ve learned a lot since then), I tried to make it possible to bluff/charm/sneak your way past most of the combats if you wanted. Still, the fault is likely mine for not making that even clearer to both players and GMs. That’s the lesson to learn. Give the players options to take more than one approach, and don’t obfuscate those options. Put multiple factions of NPCs in the dungeon so the PCs can ally themselves with one to bypass the others, or pit them against each other. Develop a hierarchy among the dungeon inhabitants so that taking out someone important (or impersonating them) puts the others in disarray.
And even better, don’t rely on NPCs and combats too heavily in your dungeons in the first place. Crossing a narrow ledge, piloting a raft down an underground stream, figuring out the weird mechanism controlling a big gate, overcoming the defensive measures protecting the magic gem in the glass case, learning that the statue can be moved to cause a secret door to open, trying to convince the ghost to tell you the password, dealing with an unstable ceiling or a floor that can’t hold an adventurer’s weight—these are all potentially exciting or intriguing dungeon situations, too. Dungeons are a great way to please all kinds of players because all different kinds of challenges can be presented in rapid succession, all in the same mysterious, dark, and frightening environment.


I love reading these lessons.
I hope you keep it coming, you’re going to help me write my adventures for my system!