Maybe I'll Just Make A Dungeon: Crystallizing the Central Features

Step the Fourth: Crystallize Themes, Central Features, Inhabitants, Brief History

Going into this project, I had a couple inspirational themes/central features:

  • The dungeon was of dwarven origin. I'm not sure why, but dwarven styled dungeons seem underrepresented in the OSR - at least, I don't know of any prominent OSR adventures that take place in such.
  • I want to play around with a style of "architectural puzzle" ala classic Zelda dungeons. This means these puzzles are in good working condition, and fits in well with the dwarven architecture trope. So the dungeon probably hasn't seen much ruination since its construction. It's pretty rare we get to explore non-ruined locations in OSR adventures. I enjoy the Mythic Underworld concept, but I have difficulty suspending disbelief especially if I'm responsible for actually building that underworld. More on that later, perhaps.

I want this to be a versatile module, which led to a couple more features:

  • It could be the first level of a larger dungeon, so it needs a connection down. But it also needs to stand on its own, so maybe that connection is sealed. But it can still be opened - the GM just decides what exactly is behind that seal. A kind of Quantum Prize.
  • I really want to see more in-built scalability of difficulty, and hence suitability for a wider range of character levels. So the dungeon in its "default" state should be on the order of other good examples of 1st-2nd level dungeons, and have good support for increasing difficulty.

Two further central features were inspired by the generated map itself:

  • I noticed the biggest rooms were all the same size and shape. This is obviously a quirk of the algorithm, but it did give me the idea that it is intended, and perhaps they were all of a similar purpose. These turned into "monument rooms", a large statue of a dwarf in the center, holding a valuable item, the room trapped in some way to protect it. There were 7 of these, which is an odd number... maybe there's a secret 8th in a differently shaped room?
  • Importing the map to Mipui, some of the Donjon-generated detail is lost, including which doors were locked and trapped, but interestingly it did transfer portcullises. I think dungeons could use more "windows" - ways to see ahead which could give hints at the dungeon's layout and "objectives" without allowing easy passage. So I decided to make most of these unopenable, at least from the side you're likely to reach first.

Preview of the next map-focused installment

Extrapolating from these themes

So the fact that the dungeon hasn't seen ruination also likely means it hasn't been inhabited by intelligent individuals since... what? What happened to the dwarves? I guess they're dead. Undead? Perhaps protect what they sealed below with their un-life? Ok, so some light ruination.

What if the items the statues hold in these monument rooms need to be collected in order to unseal it? That's 8 items. Kind of a lot. So maybe they aren't super challenging to acquire, but have a good variety of types of challenge. And the dwarves aren't a constant danger.

But what else is in there to give the party a challenge? They must be the first ones to open up this dungeon since it was sealed off. So let's style the majority of the wandering and inhabiting monsters after unintelligent dungeon dwellers; things that would fit through a tight, undiscovered entrance and have gone about their merry lives in darkness for decades. Normal rats and bats, spiders, oozes. And I really want to throw in some monsters from BX I think are underutilized: Cave Locusts, and Giant Geckos. And... Cats!

But we really need some kind of intelligent faction to interact with. And that was a perfect match for scalable difficulty. I'll include a number of factions that could begin inhabiting the dungeon after the party arrives. Challengers to the party's quest, and a first wave of more general dungeon inhabitants moving in. The GM can pick and choose between these to add spice and challenge.

So a rough history and outline comes together:

  • A small dwarf clan, maybe its head their 8 children, strike out on their own to create this stronghold
  • They Dig Too Deep, decide to seal away What Lies Below via these Macguffins, become undead in the process
  • The hold lies fallow until its entrance was rediscovered, the party arrives, and then a variable amount of challengers hot on their heels

So in a way, this module might examine the birth of a more typical Dungeon. I'm not sure how hard I'll lean into that concept, but it's there in some respect.

In reality, a lot of this took place in the back of my mind while working on various other bits of the adventure. I was going to start looking at further map changes in this post, but I really want to give that a lot of attention, so I think I'll cut this one off here.

I encourage you to comment below, rather than elsewhere.

Long live the Blogosphere!

Maybe I'll Just Make A Dungeon: Getting Over Myself

Well, there was a time when I wanted to keep this style of writing off my blog, but it seemed to help me actually write instead of endlessly edit, so...

So I Want To Make A Dungeon? Just follow these [insert final number] simple steps!

Be Warned: This only superficially and sarcastically resembles a step by step procedure that I would recommend following. I may or may not revise it into one later. I've been working on this pretty much constantly since my earlier post, and so this is an artificially re-organized and dramatized version rather than a real-time log.

Step the First: Decide to really make something instead of constantly brainstorming on yet another amorphous project I get distracted away from and build up anxiety about not finishing

Return from an RPG info-sphere hiatus and come across a couple of genuinely inspiring posts on making dungeons at the equally indispensable Swamp of Monsters and Papers & Pencils. Remember Bryce's Black Maw experiment. Remember I saw a few neat dungeon-building "games" recently - Ex Umbra and Delve / Rise.

Think "Hey - maybe I'll just make a dungeon. I've never really done it before despite how much I've read about and thought about dungeon design. It might even be a nice, chill past-time exercise, like a crossword puzzle; just following procedures and progressively fill in some gaps. Some nice soothing distraction from our IRL hellscape."

Decide to keep it vanilla, trope-y, not get my head too high in the clouds with heady artpunk/OSR-envelope-pushing concepts. Focus on making something Bryce wouldn't pan. Use the lingua franca of BX and put my handy OSE books to good use. Keep it to one level (though we'll definitely want to have multiple floors/sub-levels for some good verticality. Every time I see a dungeon restricted to one lateral, level I nope out.)

Spend sometime revisiting  some old links and compiled dungeon design notes from around the blogosphere of the last decade (Philotomy's advice, Melan diagrams, Jaquaying the Dungeon, Goblin Punch's Checklist, Blackrazor's Stocking Method, a bunch of others), and honestly come up short on good fully "worked" examples of a well-rounded old school dungeon. (Fittingly, I saw only yesterday that Necropraxis agrees after reading the exact same posts that inspired this project. Though the recent Gygax 75 Challenge comes close. Note that I'm not sure I actually intend to fill this gap with this series - the dungeon itself takes precedence!) 

So less a simple, relaxing exercise, but between all the resources and my own design thoughts, I think we can work this out. 

Step the Second: Prevent Blank Page Syndrome AND Icarus Syndrome by just getting something concrete to start with

All of my backburner dungeon projects have started with ideas for central features and themes, or a rough shape of a map. Moon College started with the idea of a bunch of thematically related magic items. Its map was literally shaped around them.

And I've always assumed if I ever made a "whole package" dungeon that I'd be meticulously designing the map layout myself, (and I will almost certainly be doing that after this exercise). Starting from an "externally sourced" dungeon feels... pedestrian.

But partly inspired by the Papers & Pencils guide, and given the self-constraining nature of this project, I decided to turn things upside down and just start with a pre-made map. And not just any pre-made map. A random Donjon map.

I've always kind of written off Donjon as being too random and vanilla, and it really is, particularly the room contents. And you're also locked into its rigid generation rules.

However, after playing around with the advice from Papers & Pencils, I did find some sweet spots in the generator settings that spat out some layouts I didn't hate.


And! I discovered that Mipui, an online map editing tool we've been using in the Castle Xyntillan game I've been playing in, can import maps from Donjon.

Mipui doesn't have the most intuitive or... nuanced toolset, and it can be visually buggy and slow even with modestly sized maps, but it's definitely the most comfortable tool for editing old school dungeon maps I've found - at least online. And its image exporting is pretty good, if you don't need any fine detail beyond what the tools it supports. I definitely intend to reproduce the map in another tool, maybe Tiled, or actually hand-draw it, since that's another thing I haven't really done since childhood despite my latent, 1.6 decade-old (insert gasp and sigh of despair) Art & Design college major.

I also intend to produce two maps - one for the GM (with optimal iconography and notes for running) and one suitable for displaying to players in a VTT or cut up and assembled at the table while playing (no secrets). That's another pet peeve of mine - a lot of effort and artistry goes into creating some beautiful maps that only the GM gets to see! It doesn't take much to get those to a condition that they can be very simply used in Roll20, as long as you aren't using "S"s that stick out of walls, or crossing passages over each other.

So I generated a few Donjon maps and settled on one that spoke to me. It's far from great, but all we're really looking for here is a starting point for a dungeon that already has good looping, flow and multiple routes, and some good variety in layout.

The Donjon map imported to Mipui

Step the Third: Start tweaking, and then stop tweaking

Get rid of the most glaring absurdities. The passages that lead back into the same room, the ridiculously long linear corridors, the dead ends (should have used the "no dead ends" setting) - this isn't a funhouse dungeon.

Starting adding some flare and non-standard shapes. But don't get obsessive about the layout yet. I'll revisit how and why I refine certain aspects later.

Snip snip... ah, much better.

This step, at least, was nice and relaxing - in the same way as fixing up a wonky generated Minecraft village.

Many more steps to come - if I'm up for it. I've got an outline of everything I've done so far, but don't want to Icarus up this post before I Icarus up the dungeon itself.

I encourage you to comment below, rather than elsewhere.

Long live the Blogosphere!