Several months ago I got the idea for this post when a person on the Swords & Wizardry G+ group asked for advice on making a megadungeon. I have two - Nightwick Abbey and Uz's Undercity - and I have run them quite a lot. By my (probably flawed) math, I have run Nightwick alone for more sessions than James ran his original Dwimmermount campaign and online playtests combined. My online group often asserts that megadungeons are my bread and butter. Presumably I've learned something about making them and running them in all that time, and this is my attempt to try to organize that learning.
A megadungeon needs a theme. The standard theme is, of course, "a wizard built it for weird wizard reasons." That is a fine one and allows for a wide variety encounters, tricks, and traps, but for reasons I will reveal later I don't think its necessarily the best idea. I think you need something stronger - more particular to the dungeon. Make sure the theme is something that speaks to you. It needs to come from some kind of media you don't mind revisiting to draw from the well when you're not in the mood or stuck for ideas. For me that meant horror movies.
It is also important for levels and sublevels to have themes that, while tied to the dungeon as a whole, make them unique. One of the bigger problems with the first version of Nightwick Abbey is the first level showed all of its cards too early. The "new"* version instead has much more themed levels - a cloister, a garden, torture chambers, catacombs, etc. - that help the DM and the players keep from getting bored. It also means there's a greater sense of discovery because either the thing you discovered is new and interesting (a new level with a different theme) or it hints at something about the level currently being explored.
The Player's Map of First Level of Nightwick Abbey. Letter Designations were Assigned by Players.
I use geomorphs for Nightwick and the Pettigrew Papers for Uz, and both of these sources allow me to have micro-themes within the bigger themes of the level and the dungeon. In the case of Nightwick Abbey each geomorph has a broad description of what it is before I start stocking it. The geomorphs have since become fairly obvious to the online group - who keep track of their divisions, but I'm not so much bothered by that. Geomorphs are a very easy way to Jaquays your dungeon.
Stocking algorithms are incredibly important to the way I design dungeons and run games in general. The main reason I use them is the help keep the voices in my head quiet, but I think its worth commenting on how they affect my games. Once I have assigned a geomorph/pettigrew complex a theme I divide the number of rooms in it by three (always rounding up if I have to). That gives me the number of rooms with monsters in them. Then I use
Courtney's Treasure Tables to generate an equal number of small caches (1d3 treasure parcels each). Half of these will go in rooms with monsters, a quarter of the ones without monsters will be trapped, and a quarter will just be free treasure. If the theme of the geomorph/complex necessitates a boss monster then I will increase their treasure parcels to the 4 + 1d4 - 2 one. I give each geomorph or complex a special if I can think of one. If I can't think of one then in the case of Nightwick Abbey I don't sweat it that much because Nightwick's entire operating mode is a special.**
On the level map I posted there are only maybe two truly empty rooms. Only 1/3 of the rooms have a monster encounter, but the rooms without them often have elaborate decorations or clues as to the nature of the dungeon. These are usually based on the geomorph, level, or dungeon theme. This is why it is important to have a very personalized theme: when you're stocking an 80 room level, eventually you will run out of ideas; however, if the themes you've picked are resonant enough with your brain you should be able to fill in the gaps with something. It also important to remember that something is better than nothing. All D&D is hackwork and a half-assed idea that gets your game on the table is better than a perfect one that takes months.
Imma Stock all the Rooms!
Back in the dim antiquity of 2009 when I first started thinking about the dungeon that would become Nightwick Abbey, there were a lot of hot takes saying that megadungeons needed to be huge. At the time I felt that Nightwick was too small but was unwilling to enlarge it due to laziness. Then I ran level one. For five years. ~77 rooms got me about 5 years of play without my needing to make a second level (though I kept promising I would). Experience with the Uz Undercity - which is a little less traditionally designed - has convinced me that 60 - 80 rooms a level is plenty mega for players to get lost and have plenty of options to explore. I wouldn't advise trying to get by with just one level, but the current version of Nightwick has 2 60 - 80 room levels and two ~30 room sublevels. This has been enough prep that I haven't touched it in two years and it seems like I may not have to for a long while yet.
One thing to remember is to restock your content. A simple version I use is that a room restocks on a 1-2 on the dice. The first week after the room has been explored you roll a d20, the next week a d12, then a d8, then a d6, then a d4 and you roll that d4 for each additional week until it restocks. This has worked very well for me, when I remember to do it.
I'll end with some pictures of my "Nightwick Abbey Prototype" - the graph composition notebook I keep my dungeon notes in.
*It is some years old.
**It's a living dungeon that shifts when the PCs do things it doesn't like.
.png)
.jpg)
