Save vs Neckbeard: Building a Megadungeon Part 1

I knew that sooner or later I was going to build a megadungeon. It’s the fantasy RPG equivalent of a Hell Week obstacle course. Many try, most fail, and some people lose their minds in the process. But there’s something irresistible about the idea of the megadungeon – a magical and wondrous meatgrinder that can occupy a party for multiple campaigns. And the best ones like Stonehell, Barrowmaze, and Rappan Athuk have developed a well-earned notoriety for straddling the line between meaningful exploration and grinding struggle. Others reek of madness, sometimes delusional (The World’s Largest Dungeon) and sometimes inspired (Anomalous Subsurface Environment).

So, here we are. Lots has been written about this subject by people smarter than me. Check out what they have to say. This is my take.

The principles I’ll be following:

The dungeon will exist in the world of 0e D&D but is not tied to any particular setting – I figure a setting will evolve naturally. Plus, Greyhawk is for losers.

For this exercise I will be primarily using the tables, monsters and treasure in Delving Deeper, an 0e retroclone because it’s:

  • 1) Easier to navigate than the original rules.
  • 2) Free (unless you want a print copy).
  • 3) Has neato gonzo elements like androids and shit.
  • 4) The lightest version of D&D which should make it easier to adapt to other systems.

If I need other tables, I’ll make them up because Old School.

The dungeon will have 1d10 levels. Each level must fit on an 8.5 X 11 piece of paper. Each level will have 1d6 connections with other levels. Levels may be connected horizontally, and each level will have 1d100 rooms (I’m not fucking around, we’re building a megadungeon here).  The contents of those rooms will be determined using the following table, adapted from Delving Deeper:

  • 2 Treasure guarded by a trap
  • 3 Trick or trap
  • 4 Treasure guarded by monster
  • 5 Monster
  • 6-8 Atmosphere*
  • 9 Monster
  • 10 Treasure guarded by monster
  • 11-12 Unguarded treasure

*for the purposes of this exercise, no room will be considered “empty” – all rooms and passages will have at least a hint about what’s in an adjacent room or passage. Other options may be considered as long as they are consistent with the behaviour and needs of the level’s inhabitants (nests, litters, barricades, etc).

I’ll roll 2d4 for the number of eras that have passed since the dungeon was created.

The first era’s inhabitants will be random humanoids. Every other era will add two more primary inhabitants

If they need them, each inhabitant must have had a source of food and water. Those sources may still exist, and they may be located in the dungeon or adjacent to it. In addition, all humanoid inhabitants must have had a specific motivation for occupying the dungeon, determined by the following table:

  • 1-2 Refuge
  • 3-4 Religion
  • 5-6 Military
  • 7-8 Settlement
  • 9-10 Research/Exploration

Each era’s inhabitants won’t need to have occupied the entire dungeon and may have only lived on certain levels. If there’s a reason why the previous inhabitants remained on the other levels (undead, living statues etc.) they can stay. Otherwise, the previous inhabitants have:

  • 1 Left the dungeon
  • 2 Died off
  • 3 Moved to a different level in the dungeon
  • 4 Adapted to the environment, changing as a species in some important way. Rolls joint.

The entire dungeon must be drawn by hand on graph paper.

The dungeon’s history, inhabitants and current state must form a coherent story – no matter how gonzo.

I fucking pray to Orcus that this doesn’t turn into a total disaster.

Next post: I establish the eras, levels, number of rooms, and inhabitants. And then try to figure out what the hell has been going on in there.

2 thoughts on “Save vs Neckbeard: Building a Megadungeon Part 1

  1. Exciting! I love how you are leaning so heavily into the randomisation.

    It makes me wonder if it might be possible to create an entertaining game just out of the “building the dungeon” part a megadungeon campaign. It always seems a bit melancholy to see people (myself included) keying huge maps, knowing that most of it will probably never be explored by players. But what if the creation process was designed to be fun in itself?


Comments are closed.