I thought it was good, and lots of people raised good points... I also thought it very quickly sidelined from dungeon design as game design do dungeon design as world design. early in the post, the discussion was about how one navigates a dungeon as a space, later in the post it became about how one writes a dungeon as a vehicle for narrative and lore.
so, I wanted to talk a bit about dungeons as game design, because it's how I like them the best, and I feel like it is not talked about enough.
First though, I want to clarify some terminology, I'm going to use a few terms here.
Mini Dungeon: A single location in a tabletop RPG which has a self contained story, and can be fully explored, and all of its challenges overcome within less than five nights of gaming, often just in one night of gaming.
Dungeon: A large playable space which hold challenges that need many sessions of play to overcome, but is still in some way self contained
Mega Dungeon: Synonymous with Dungeon as used in this text.
Jacquaysing: A term describing the application of good game design to a Dungeon, used in honor of Janelle Jacquays.
With definitions out of the way, let's get on to the main text.
So, the post that inspired this article opens with a statement,
"My take on fantasy RPG dungeons is that if I pick two rooms at random on your map and there is only one path to get from the one to the other you shouldn't be allowed to call that thang a 'dungeon'."
the post later goes into "but how do I make that kind of layout make sense" which gets into the asthetic design thing... But I'm going to challenge that later question directly.
First off... If you've ever been in any building on a college campus, you can realize that it makes sense. Real buildings are designed as loops, and intersecting mazes, with multiple strange often inexplicable connections between floors, and often (in spaces designed to be public, or by organizations and individuals with large sums of wealth) to be visually appealing despite the requirements of their intended function.
I don't think much of the question is actually about it making sense though. I think a big problem a lot of people come up against when designing D&D dungeons isn't realism, or worldbuilding, it's context. I think that a space that is designed in a labrynthine way might technically not be necessarily unrealistic, but if that space is entirely without context, it may still feel unrealistic.
So, context. I like a two methods of providing context to a dungeon's unusual design.
a dungeon is not a singular designed space, it is multiple designed spaces in close proximity, and the interactions at the peripheries of those spaces are a large source of the location's complexity.
even if a dungeon was a single space "as originally designed" in lore, its modern inhabitants don't treat it as such.
I tend to like using those two methods together. A dungeon might once have been a castle's stores, connected to a large mine, that broke into a cave system, which grew around a crashed interdimensional spaceship, but now it's not any of those things anymore, the mines are empty, the ships original inhabitants are long dead, and the castle was abandoned by its builders, instead it's the source of a gold rush as every local warlord, cult leader, and bandit king seek the treasures of the lost dungeon.
But before layering the techniques, you have to... understand each technique individually, so lets run that down. I should note, as this is about game design, I'm going to be talking about the consequences to playable space that this lore decision would cause, not how to implement this kind of worldbuilding structure into the asthetics of the dungeon.
First off, multiple spaces. I generally think that if you want a mega-dungeon but you don't want it to be: A. Bland, or B. unbelievable, you should make the dungeon be different spaces that intersect in complex ways, instead of one unrealistically large, and unrealistically complex space. this should mean two things.
these spaces are designed in different ways. One might have lots of hallways and small rooms, another might have lots of closely connected large rooms. One might have many secret areas, one might have no secret areas. the structure of the layout should be different between the different spaces, both to aid navigability ("We're in the dwarf ruins, because the walls are smooth, and the tunnels all have sharp corners"), and to give the space a distinct feeling from another space it intersects, which will make the gameplay of exploring it more enjoyable.
each space should have multiple connections to multiple other spaces. If a space only has one entrance, to one other space, that isn't an interesting area, that's a large closet. Each space should connect to at least two others (so that it can both be used as a path between them, and also be discoverable from either one), and connect in at least two ways to at least one of those connected spaces.
These two principles would lead to a dungeon that held exploration and challenge regardless of the number, or number of types of inhabitants.
In regards to amount of inhabitants tho, again, I think you should make it more than one faction, and I think there are a couple good guidelines for doing so.
the factions don't like each other. the gameplay of an RPG is just as much social interaction as it is decisionmaking, and that should be reflected in the Dungeon (which will be the primary playable space of many campaings). If there is conflict, and story, it will push the players towards social roleplay (not just tactical roleplay) and give greater context and variety to their decisions.
the factions will partition the space like miniature nations.
Just like with the "different spaces" game design process, have multiple connections, with multiple factions.
those decisions will result in situations where even a space that was once uniform, uninteresting, or maybe slightly nonsensical seeming, is contextualized. "why are there three hallways between these rooms when two would do?" is a question that is never asked, because the players are instead asking "which hallway is controlled by a friendly faction" or "what path can we take between these rooms that keeps us out of the battle-prone borders?" The space itself can be incredibly unrealistic, but if the contents within contextualize it such that the player puts other things first, their disbelief will be suspended by distraction.
another important note, is that two "factions" doesn't have to mean two nations. In certain levels of a dungeon, a single dragon with no followers or minions could be a faction on its own, or for that matter, so could a lone wizard, or owlbear, or a certain group of untintelligent monsters like boars, or bullette.
and now, for how I combine those principles, generally use a few tricks.
one faction in a region. This one is simple, it gives the players a good sense of what faction's territory they're in, and makes both the boundaries between factions, and the boundaries between territories make sense.
one dungeon exit per faction. Dungeons may feel self contained, but at the end of the day, everyone has to get out of the house sometime, and if your factions don't at least have a way out of their current level of the dungeon (In deep levels an exit from the dungeon itself might be unreasonable... or not, deep levels are also usually full of dangerous enough things that they could feature interdimensional portals and the like) they're going to lose both realism, and gameplay interest. If your players have an even chance of encountering any given dungeon faction, than the story at the table can be truly emergent, with the players experience of the dungeon being shaped as much by their own decisions as they are by yours.
3. Leave some space. if all of the factions are right up against each-other, the players will have an awful time navigating the dungeon, and it isn't "realistic" either. Unless one faction is actively besieging another, the space between factions' population centers is going to at least be a few rooms of no-mans land or wilderness.
4. story from every direction. For each entry into a faction, design ahead just a little with the setup of what the faction expects from that pathway, and how they've responded to it. A faction's main dungeon exit might be heavily trapped, but a border with a nearby faction only lightly patrolled. A border with an enemy will likely be guarded actively, or maybe even barricaded in some way. the shape of the players' encounters with a faction should be determined by their direction of approach, not just in social dynamics, but also in physical space.
Okay, that's everything I have on like "these are two easy ways to contextualize dungeons so that your players won't notice/care that they have unrealistic elements because those elements will exist in context and seem normal as a result."
but, I have some other general dungeon design tips. Most of these are "Jacquaysing"-esque tips, but they're worth mentioning anyways.
connect different levels lots of times. Level 1 should have more than one staircase to level 2, because then the player's path through the dungeon can branch and loop in a greater variety of ways.
simple labrynths are weak dungeon design. Generally "labrynths" are like... mazes on children's menus at restaraunts, they're a single "correct" path, with lots of branching "incorrect" paths. They lead to dead ends, and backtracking, and pissing off your players. Branches are fine in moderation, (every side room with only one door is technically a branching path), but actual complexity of a space will come from loops, not branches.
no choices without context! A dungeon should be full of choices, choices about rations, choices about allegiance, choices about who to stab next, but one of the main choices of a dungeon, is the branching path, a choice of where to go next. Branching paths may all eventually boil down to "left or right", but there should be enough information present in the choice that picking one over the other means something. It should never feel like a coin flip could make as good a choice as you can. Maybe the left hall is narrow, and the right path is wide, maybe a breeze comes from one path but not the other, maybe the paths slope confusingly, maybe one path smells really bad. Whatever it is, make sure that the players can see a discernable difference between paths, and that the difference present relates to something on the paths... And a cool breeze generated by "Glathdinar's disk of cooling blades" (electric fan) ten feet away should be used sparingly. fakeouts are only fun when they're actually unexpected, and they should usually be far enough from the decision itself that they're an actual joke or twist instead of a sight-gag.
Anyways... yeah, have fun with your dungeons, make good ones, and have a good day!
