I WROTE A LARGE DUNGEON: How I Will Do It Differently Next Time

Be Your Own Goblin King
(Save For Kidnapping Babies)

"THE WINE DUNGEON": You can read about it here. A play report here. Quick pitch: You are dogsbodies hired to retrieve wine from a cellar gone "rancid" and morphed into a dungeon for a wealthy sorcerer whose hosting parties for Salon Season. But funny that no one has seen his three sons of late...

I am currently expanding out LEVEL 2, bringing the total room count to 100, because I feel the story of this dungeon is not finished at LEVEL 1. However, I am also coming to the realization that combining the strongest rooms of both levels could yield a much better dungeon. This fills me with great dread because I really want to push this out before 2025.

So here's what I'd do differently next time I write a dungeon.

DEFINING ROOMS: I would write down notable monsters, treasures, and rooms that define the dungeon. These are the areas that anchor the dungeon and further refine the random stocking

RANDOM ROOMS: I'd also start writing out 1d6 Monsters, Traps, Treasure that define the space that I am thinking about. This helps keep you on an aesthetic theme in the dungeon when it comes to random stocking. This can also be expanded to 1d10 or whatever if you need it.

RANDOM STOCKING: Here is the random stocking table I like:

1-2 Monsters (3:6 treasure)
3 Traps (2:6 treasure)
4-5 Empty (1:6 treasure, hidden)
6 Special*

* Special can just be a room, shrine, or object that can't be moved from the dungeon but can function as a boon/bane; a shrine to Orcus would be a good example

DESIGN DOC: Then combine all the above into a single sheet as a sorta design document along with notes to myself about central "deal" of the dungeon and faction notes. This document would be added too as I think of things. At a high-level it might look like:

  • Brief overview of the dungeon
  • Hooks (with some factions)
  • 1-2 Factions (or NPCs) with wants, gives, goals
  • 1d6 Monsters
  • 1d6 Traps
  • 1d6 Treasures
  • Random Stocking table as above
  • Inspirational pictures

INTENTIONAL MAP DESIGN: Next, instead of randomly generating the whole dungeon like I did with the "Wine Dungeon", I think I would be more intentional with the map design which makes it easier to create an aesthetic theme and drive the feel of the dungeon.

I would take each of the defining rooms and build ~5 additional rooms that fit its theme. This also conveniently fits the random stocking tables. In other words, each defining room would have a room with:

  • A monster
  • A monster + treasure
  • A trap (30% treasure)
  • Empty
  • Empty + treasure*
  • Special
* Given everyone's shorten play time these days, I don't mind adding more treasure than less in a dungeon, especially if hidden and besides John Romero thinks all maps should at least have 4 secrets.

Do this for 5 defining rooms and you already have a ~30 room dungeon which is often as big or bigger than a lot of things out there.

"FILLER": I don't think adding rooms or hallways which link these key spaces together is "wasted". For instance here I talk about how "empty" rooms can be actionable in a dungeon. Hallways also can be more than counting 10ft squares. I also think random stocking can make for combinations you might not have thought of too. Also restocking the dungeon requires some extra rooms to take on a different purpose.

PLAYTEST: This is about the smartest thing you can do for a dungeon is to run it for 5-10 sessions to see how folks feel about it. My players still ask about "The Wine Dungeon" so I know that it has legs and is a pretty novel take.

So that's it! Other good posts about dungeon design:
  • The Two-Week Megadungeon: https://www.paperspencils.com/two-week-megadungeon/
  • Mastering the Megadungeon: https://inplacesdeep.blogspot.com/2017/06/mastering-megadungeon.html
  • So You Want To Build A Dungeon: https://alldeadgenerations.blogspot.com/2021/03/so-you-want-to-build-dungeon.html
  • Dungeon Checklist: https://goblinpunch.blogspot.com/2016/01/dungeon-checklist.html

4 comments:

  1. This is very useful! My Dungeon23 stalled last year, but there are some bits in there that have potential, and maybe I could rejig it into the sort of format you describe here.

    ReplyDelete
    Replies
    1. Yeah, I went into it with a randomly generated map first and I think that was what created friction between the rooms in my head and the shape on the map. That friction slowed things down.

      Not that random generation is not good for some "Friday Night D&D", but if writing a module I think its better to prioritize that sorta coherance.

      Delete
  2. I've found that identifying key themes and writing down ideas that articulate those themes and approach them from different ways (in the case of dungeons monsters, traps, treasures, etc.) goes a long way in making something big and multifaceted still feel cohesive. Solid advice for creating anything really, not just dungeons.

    ReplyDelete
    Replies
    1. Thank you for the read! Agree. I had the ur-concept down but I under estimated how not quite having specifics in place created a sorta friction. Then I started second guessing different room placements and such. Then the dread thought "man, I should just re-write this"

      Delete