What Makes A Good Dungeon, and a Good Dungeon Campaign/Adventure?

The challenge. I'm not talking about hack'n'slash, or how difficult that combat encounter was. I'm talking about challenging the PCs in more than just fighting. There have been some great answers here about not just fighting. Likewise, Dausuul is spot on by stating a dungeon is a framing device. If you treat a dungeon as a microcosm of an adventure/campaign then if you're world is fun, there's no reason the dungeon can't be also.
Throw some stupid encounters in there: the 'bad-ass' kobold who's "gonna mess you up!" Screw with the party by splitting them up and have one of the 'found' players play a doppelganger out to kill the party.

My favorite dungeon: Arenaworld!
A DM canceled on us at the last second, so my group asked me to create a one-session adventure. Not having anything particularly ready, I cheated and literally created the dungeon straight out of the DMG. I placed the group in a dungeon with a town in it where no one had memories of their past (incl. the PCs). Rather than stay in the boring town they went adventuring. Up they went to the level 1 dungeon. The end result was that the DM never showed up again and I continued the 'cheat', level by level.

My favorite dungeon episode:
A DM got tired of us ignoring his mazes by cutting through walls, so he made the dungeon invulnerable. We succeeded in taking a door of it's hinges and carting it around as an invulnerable tower shield. Came in quite handy.
 

log in or register to remove this ad


I've been having lots of fun over the last year reading Monte Cook's Dungeon a Day. His dungeon is now 11 levels deep and growing!

Things that have made it interesting: each level has a theme, a history, and a meaning. Players can tease these bits of background out. There are ongoing layers of history (original builders, later residents, invaders, etc...).

These different layers of history and meaning have created different possible reasons for adventuring there - NPCs who want things or knowledge hidden there; PCs can learn of things deeper in that will fulfill goals they've set themselves; one level relates distinctly back to the nearby town and its politics; one level ties in to a nearby tower ruin; another level ties back to the castle ruins above the dungeon.

The dungeon could either be run as a straight "place to explore and clear out" (difficult but not impossible) or as a place the PCs find themselves drawn back to again and again for "missions" with specific goals.

There's a good variety between fighting, talking, exploring, and puzzle solving.

Anyway, if you want to see a master dungeon design in the process of growing, I'd recommend highly that you take a look at the site.
 

What makes a great dungeon?

I don't have time to answer the question fully. If someone could look up my 1st enworld post, it might explain my position more fully.

A great dungeon has many elements. Just off the top of my head:

1) Creative. When in doubt, to the novice, and to within a close approximation, creative means assymetrical. It contains surprises and unexpected features. In most cases, its enough to tell a bad dungeon from glancing at the map and seeing if there are rows of similarly sized rooms. This sort of feature is logical (real buildings are often made that way), but is usually the mark of an inexperienced dungeon designer.
2) Believable. The dungeon exists for an understandable purpose and its possible for the players to see why it might have been built. Traps exist only where such traps make sense.
3) Living. The dungeon has a history. It's inhabitants have a history that preexists the players. If the players aren't clearly the first intelligent beings to find the dungeon, then the dungeon has clearly been modified by the presence of other inhabitants and explorers. One shot, non-resettable traps are often sprung long before the players get there. The dungeon has aged, and signs of the various uses it has been put through over time are discoverable. If the dungeon inhabitants need food and water, some evidence of how they come by it is available. If the dungeon is closed, some minimally plausible ecosystem exists.
4) Interactive. The dungeon is designed so that every room has some skill which is employable in the room to some advantage. There should be something for the player to mentally play with in every room.
5) The Map is Narrative. All RPG designs are flowcharts of some sort. Even the most complicated adventure reduces to a flowchart. An event driven story is a flowchart. A map is a flowchart. How the players move through the map is the particular story that they tell. The good dungeon designers make the map in such a way that the resulting narrative makes sense. The great dungeon designers do that with a map that isn't linear, and which still has a climatic story if the players come in the back door.

The best dungeon designers - Gygax, Hickman, Cook, etc. - succeed on those grounds to various extents. If you look at some of the best modules ever written, in many cases they come down to having a great dungeon, and often that comes down to a combination of a great map and a good villain.
 

Celebrim: I concur with your list except for #2. believability is not a necessary feature. Dungeons are the "weird fantasy" element of sword and sorcery in D&D. Even if the place had a purpose centuries ago when it was built, that purpose has long since been lost and replaced by a nightmare realm of horrors. Those intervening years have filled the dungeon with all manner of "illogical" monsters, traps and treasures. Dungeons aren't designed (in the game world, I mean) so much as they arise from the inherent weirdness of the world in which the PCs live.
 

Celebrim: I concur with your list except for #2. believability is not a necessary feature. Dungeons are the "weird fantasy" element of sword and sorcery in D&D. Even if the place had a purpose centuries ago when it was built, that purpose has long since been lost and replaced by a nightmare realm of horrors.

I'm having some difficulty understanding exactly how this argues against believability. If the dungeon exists in a 'wierd fantasy' universe, its perfectly logical and reasonable that it has become a 'nightmare realm of horrors'. I'm not sure exactly though how this implies 'that purpose has long since been lost'.

Those intervening years have filled the dungeon with all manner of "illogical" monsters, traps and treasures. Dungeons aren't designed (in the game world, I mean) so much as they arise from the inherent weirdness of the world in which the PCs live.

You'll have to provide some examples. I'm not understanding you. If the world is inherently 'wierd', then its going to give rise to wierdnes of the same sort. This does not imply to me 'illogical' monsters traps and treasures.
 

Celebrim said:
I'm having some difficulty understanding exactly how this argues against believability. If the dungeon exists in a 'wierd fantasy' universe, its perfectly logical and reasonable that it has become a 'nightmare realm of horrors'. I'm not sure exactly though how this implies 'that purpose has long since been lost'.

You'll have to provide some examples. I'm not understanding you. If the world is inherently 'wierd', then its going to give rise to wierdnes of the same sort. This does not imply to me 'illogical' monsters traps and treasures.

You specifically mentioned understandable purpose and logical traps, neither of which are, IMO, necessary. Whatever alien intelligence first carved these halls left no clues as to it's original intent, and the mad artificer that made them his home 10 generations past had neither rhyme nor reason for his malevolent machines. Etc...
 

You specifically mentioned understandable purpose and logical traps, neither of which are, IMO, necessary. Whatever alien intelligence first carved these halls left no clues as to it's original intent, and the mad artificer that made them his home 10 generations past had neither rhyme nor reason for his malevolent machines. Etc...

I think you have to be very careful with that. Even a truly 'alien intelligence' must have some logic to it be considered intelligent, since intelligence only means 'having an appropriate response to the situation'. It might not prioritize or think like us, but it must have some goal which is meaningful from the perspective of the intelligence. As the DM, you are I think under some obligation to imagine how the alien intelligence thinks.

And a mad artificer that makes malevolent machines is itself potentially a logical explanation for dungeons of the 'lethal fun house' sort. It's some attempt to explain the dungeon and, if it is a real explanation, creates certain contraints and narratives about the dungeon that would fulfill my requirements of good dungeon design.

What I think you should avoid doing is making these things a trivial explanation that fronts for you true reasoning, namely, to create things that will challenge or kill the PC's. Players are typically quite intelligent people, and if you make a dungeon where the real reason behind everything is some metagame consideration, don't expect them to buy into some explanation of 'alien intelligence' and 'mad artificers'. They will know that the real dungeon designer is you, and they will see the dungeon in that light as a product of your desires and having no reality outside the metagame.

That is not to say that metagame considerations shouldn't inform your design. You still have to think to yourself, "How will the players use their powers when they encounter this situation? Will the player character's abilities render this situation trivial or impossible, and am I ok with that, or do I need to rethink the design to decrease or increase the challenge? What do my players enjoy in an encounter, and what is there skill level?" But those questions ideally aren't the only ones that inform not only what appears in your dungeon but how it and in what context it appears.
 
Last edited:

What I think you should be avoiding do is making these things a trivial explanation that fronts for you true reasoning, namely, to create things that will challenge or kill the PC's. Players are typically quite intelligent people, and if you make a dungeon where the real reason behind everything is some metagame consideration, don't expect them to buy into some explanation of 'alien intelligence' and 'mad artificers'. They will know that the real dungeon designer is you, and they will see the dungeon in that light as a product of your desires and having no reality outside the metagame.

I must spread some XP around before giving it to Celebrim again.
 

TV producers came up with Doctor Who's TARDIS and Captain Kirk's Enterprise as means of getting a continuing cast of characters into a wide variety of different situations. The original-model D&D dungeon serves an analogous purpose, offering a potentially endless supply of "mystification, enjoyment, excitement, and amusement in the challenge of the myriad passages of the dungeons."

A note on nomenclature: "the dungeons" in this context denotes the same as "the dungeon", its connotations emphasizing the variety, extent and multi-leveled nature of the works.

There is more to it than one novelty bang after another, though. There is another layer of game that may not be immediately apparent, as in a video game that at first glance is just another "shoot 'em up" but has a deeper puzzle-solving element. The added layer is that exploring the map is itself a strategic undertaking.

There is yet another layer in the "overworld". This is for players at once a less alien environment and a less manageable one. There are pockets of hospitality, separated by monster-infested wilderness. Affairs, as in our world, are not neatly sorted into "levels", and clearly walled-in paths are exceptions to the rule of being able to wander in any direction at any time.

The whole nine yards is not to everyone's taste, but I think problems most often arise from a DM's neglect of one aspect or another. It is easier for players to find their way to their own kind of fun if a multitude of ways is readily available.
 

Remove ads

Top