Image by Thomas Wolter from Pixabay

Many of my campaigns either do not feature “dungeons” or employ transformative conceptualizations to justify their existence within the campaign world, because – to be frank – they don’t make a lot of sense, otherwise.

But there can be other structural concepts that don’t obey all the technical principles of the generic ‘dungeon,’ in other campaigns, that could be considered such if you squint just a little.

These often have singular origins that set them apart from the generic ‘dungeon’.

But sometimes, you just have to hand-wave away the logical flaws in the idea and embrace the concept.

That does not mean, however, that your dungeon designs should not be rational. On the contrary, it places even greater weight on the internal logic of each representative example.
rpg blog carnival logo

This month’s blog carnival, hosted by Plastic Polyhedra, is all about dungeon design, and for the reasons listed above, I was initially hesitant about participating, feeling that I wouldn’t be true to the spirit of the topic – but then I changed my mind, because what I had to offer was an alternative perspective that others might find useful.

And that’s what Campaign Mastery is here for.

Of necessity, this will be a fairly high-level review of the process that I use to create dungeons, when they are appropriate to the game world and the ambitions of the players (obviously, if the PCs aren’t interested in Dungeons, it’s a waste of time creating them).

1. Core Concept

The place to start is always by defining a core concept for this particular dungeon. This central idea will be used to guide subsequent decisions, it provides a focus for all aspects of the design. This is especially important if you have many dungeons in your campaign because there is an ever-present risk of them blurring into each other; you need all the help to can get in making them distinctive.

I’m going to include an example as the process unfolds. In this case, my dungeon Core Concept is ‘defiled temple’. I want this to be a haven for enhanced Undead – especially Skeletons.

2. Constructor – Who and How?

Who built/created the dungeon and how? This is another of those formative questions, or should be. Every subsequent decision should at least consider these answers, if not be actively guided by them.

The heretic Azelfar, now a servant of the Dark God Telvas, returned to the temple of Belaal that banished him a decade prior. With the powers lent him by Telvas, he summoned an Undead Horde to besiege the Temple, eventually killing or driving out the faithful. Claiming the Temple as his own, Azelfar used a wish spell to invert the pyramid-shaped temple; Undead workers completed the transformation before burying the entrances beneath a mound of earth.

— When I started writing this article, I had the basic premise, but not the names or deities concerned.

3. Intermediate History?

Sometimes, this isn’t necessary, but adding in some additional layers of history can greatly enhance a dungeon. Just be careful not to stray too far from the core concept.

Knights of Errich challenged Azelfar, and brought the heretic low, but at great cost to their noble order. Carrying their wounded from the field, they returned to their Citadel, carrying with them some unknown malady that gradually wiped them out to a man. At the same time it slowly emerged that the evil curse on the Befouled Temple had not been fully lifted; creatures of dark nature continued to be sighted in the vicinity. The trade route from Carvelus to Isignerry, which once had swarmed with traffic between these growing communities, slowly choked to nothing, because of the darkness that lay across it. Over time, the exact location of the Befouled Temple was lost, as were the specifics of the trade route over the mountain pass. And so the evil was free to brood, and grow.

— This contains a lot useful information. In particular it raises the possibility that Azelfar survived, deceiving the Knights of Errich. It implies that there are two sources for maps to the dungeon that can be found – the one employed by the Knights (which means braving their ruined Citadel and the disease that may still slumber there, a mini-dungeon of sorts), or a map of the old trade route. The first would probably be more useful, the second is probably more accessible. And it suggests that there may be more denizens than simply Undead in residence – but whatever now calls the Temple home would need to be comfortable in an Undead-heavy environment. Finally, it provides a source of wealth – all those lost trade caravans – which has (presumably) accumulated in the ruined Temple.

4. Recent History

You may not need this if you’ve created an Intermediate history. This is all about motivations – why is it important now that the dungeon be found and cleared? Has it been newly-discovered or does the threat that it poses suddenly seem greater than it was? If greed is to be the PC motivation, why now? Possible motivating discoveries might include the discovery that some long-lost treasure may have ended up there, or the opening of a new (or re-opening of an old) trade route, or some sort of imminent threat that is blamed on the dungeon’s presence. Does someone else decide that the time is right to clear the dungeon, and hire the PCs to do the dirty work? Another old favorite is the notion of workmen disturbing something that was better left undisturbed.

Where it’s not about motivations, this needs to connect the Dungeon to the PCs – how do they know about it? How do they learn where it is located? This frames the prologue and the journey to the dungeon (which may not be without its challenges – anything that doesn’t like living with Undead but doesn’t like being hunted down by man is likely to find security by hanging around the general vicinity).

There are lots of options here.

  • The PCs may be hired to clear the old Citadel, and discover the map.
  • The PCs may be hired to clear the old trade route by a greedy merchant.
  • The PCs may be hired to find out where the evil creatures that are threatening one of the towns are coming from, and what is drawing them to the vicinity – and to do something about it.
  • But my personal favorite is to use the opportunity to make more campaign mythology relevant

    Long ago, before man, Belaal chained Telvas to a rock to limit the spread of Darkness. When the sun sets, Belaal could no longer see his prisoner, who was then able to slip his bonds and rule supreme over the Night, but always as light returned to the world, he was forced back into his bindings. To further constrain Telvas’ influence, Belaal gave to the newborn race of men the gift of Fire with which to light their way. Every now and then, Telvas’ endless struggle against his bonds would bear fruit, and he would escape, and the light of the sun would grew weak or fail entirely, but always he was recaptured by the High Priest of Belaal using the fabled Rod Of Piety. But the bindings grew weak over time, and need to be renewed every half-millennium, lest Telvas escape.

— This raises the stakes considerably. It becomes obvious what the Heretic who created the dungeon was after, and the plot of this Telvas-worshipper may well succeed without intervention. So this not only makes the campaign mythology relevant, it better-defines the dungeon.

5. Internal Structure

Now that we know what we’re talking about, it’s time to settle some of the fundamental parameters of our dungeon – how big is it? How many levels? Are areas isolated, or is the place more like an underground house? An area may be a level, or it may be a cluster of rooms or spaces. Unless it’s truly massive, a single room shouldn’t be an area (though I did this once for the sheer variety).

I tend to stick to the mantra, ‘a purpose to every area, and to every area, a purpose’. These purposes should derive from the central concept and the history created in earlier steps of the creative process.

I’m not particularly interested in how one gets from one area to another at this point in time. That will come later, when the areas get broken down into specific rooms.

We have two design imperatives to guide us – the original design of a functional, reasonably self-supporting, Temple, and what was done to it subsequently, after it became structurally inverted.

  • So, ground level has a chapel for passersby to pray in, cells for the hearing of confessions and the bestowing of healing, and antechambers for the priest conducting services and his assistants. There would be guest quarters for travelers and a store-room for gardening tools. There might also be a crypt for the especially faithful, though most would be buried in an external graveyard that might even be considered a separate area.
  • The first level would have some classrooms for acolytes to learn in, and the functional areas of the temple – kitchens, baths, dining room, etc. This implies plumbing of some sort.
  • The second level would be occupied by the priests who conduct services, who teach the acolytes, and so on. It might also have a small library of predominantly pious writings and a separate dining room.
  • The third level has quarters and a private chapel for the high priest, and – presumably – a hidden chamber containing the treasures of the church. Either the heretic Azelfar couldn’t find it, didn’t know what he should be looking for, or couldn’t destroy or corrupt these treasures. In addition, the mundane work of maintaining the Temple needs an office of some sort, and its conceivable that the high priest would need to hold meetings there. If the High Priest were not personally responsible for the bookkeeping – and that would usually have been the case – there might also be a separate office and quarters for that official. In addition, there would be quarters for a couple of servants (favored and trusted acolytes or junior priests) who assist and care for the High Priest.

By going from bottom to top in listing those levels, the same list provides our foundations post-inversion – from top, down. But we need to think about what Azelfar used each area for. Was Azelfar a Necromancer, or did Telvas do the heavy lifting of reanimating the dead in Azfelar’s service? The first scores points for neatness, the second leaves room for Azelfar to pose some sort of unexpected challenge. The first is also more predictable, which is the main reason why I’m leaning toward the second.

A critical point is that Azelfar and his forces were able to invade and defile sacred ground – certainly, the chapels if not the entire Temple would have to be considered such. Shielding the Undead against Turning would be both a significant advantage and necessary. That sounds more ‘clerical magic’ than ‘necromancy’ – otherwise, every necromancer would be doing it! And he was a religious acolyte before his perversion / conversion into the worship of Telvas, so that fits.

  • The just-below-ground level is full of traps and disturbing imagery to scare away the foolish. The cells are now home to various ‘guests’ of a non-skeletal nature. These will mostly flee if given the opportunity. Note that if any of them require food, they will have to go out to hunt for it. Some may have created private exits to the surface, and sealed off the doors to the chapel. However, Undead skeletons with bows (and no compunction about using them on ‘allies’) are stationed in the alcoves that used to hold various statues devoted to Belaal at what was Ground level (and are now inset into the ceiling.
  • The second level houses the real defenders of the Fallen Temple. These include enhanced skeletons including some of non-humans. A small band of were-rats have tunneled out to the old cemetery, feasting on the remains there. Their bite conveys the disease which killed the Knights.
  • The third level carries the second line of defense – more powerful enhanced skeletons. This is also where Azelfar enhances Undead as Telvas raises them.
  • The top level contains the throne room of Azelfar, and the most powerful defenders of the ruined Temple. The purpose of this level is to free Telvas and be the base of his rule – so Azfelar’s personal throne is NOT the largest or the highest.

6. Look and Feel

In general terms, ‘Look’ is the cosmetics of the dungeon. What sounds permeate the place? What smells? What are the common distinctive features – the walls, the ceiling? Where does light come from, and what is its quality?

These are important building blocks to the final dungeon.

‘Feel’ is equally important, but far more nebulous and hard to pin down; the terminology is just too broad and general. Take “creepy” – are all manifestations of “creepy” the same? if so, how do you distinguish one from another; and if not, what are the differences? Fortunately, there are all sorts of metaphors that we can apply to grope toward such distinctions – “something always lurking in the shadows” is different from “something is watching you – and hating you” is different from “there’s always something lurking behind you until you turn around” is different from “it makes your flesh crawl”. “Insidious Creepy” is not the same thing as “Brutal Creepy”. Appending each narrative passage with the hint of a scene of unwarranted violence being committed by one or more of the PCs creates the impression of something violent trying to worm its way into their thoughts, a variation on the techniques offered in “The Envelope Is Ticking – Insanity in RPGs“.

The walls are pastel shades of yellow that have faded to bone-white; shades of pale orange and gold that have decayed into pale dirty gray-green, all pockmarked with molds of green and brown and bright, angry red. Many have been defaced with crude, rude, or unholy symbols, splashed on with red paint, black paint, or blood. Creepers descend along every column, erupting from the tiled ceiling and weakening those tiles; many have fallen to the floor and smashed, with enough force to inflict serious harm on anyone they strike. The floors are plastered stone, green and slick with mold and the skeletal remains of small animals. Doors are inset into the walls near the ceilings; in some cases, rough ramps lead up to them, in others, the doorway arches have been roughly broken through. Any furniture that wasn’t mounted to the ceiling – which is most of it – and any other loose material has fallen to the floor, creating layers of detritus. And everywhere is the tinkle of falling water, and the thick smell of rot and decay. Brass fittings still hold long-extinguished candles, now pointed to the ground.

— This builds on the notion that the Belaal was a sun god, represented by the mostly-white pastel tones. Some of the information contained in this high-level summary can be reserved for the description of specific rooms, such as the stuff about doorways (which, I have to admit, was the first thing that I thought of). Most of it is concerned with conveying the impact of the inversion, and the impression that it was virtually instantaneous – down suddenly became up – and the implied violence of that transformation.

The ‘water’ sounds require further explanation. Turn any plumbing fixture upside down – sink, water-closet, whatever – and water will fall from it until the supply has been exhausted. On top of that, what hot air there is will rise, carrying evaporated water vapor with it, which will condense on the walls and pipes (which are presumed to be cooler than the surrounding atmosphere). The result would be perpetual drips and running small streams of water. The dampness would encourage mold and rot, so this notion feeds into other areas of the description.

7. Significant Occupant Ecologies/Societies

I’ve preempted this discussion a bit in my example, but it’s time to get more specific. I try not to think of specific inhabitants at this point in time (even if I have already done so); instead, I focus on ecologies and societies, as implied by the heading. I might have a “Minotaur kingdom”, for example specified as one of my areas; that tells me that the apex predators in that region are Minotaurs. What do they feed on? What wildlife have they domesticated? I try to keep this broad, if I can, but “plant – leaf-eaters – meat-eaters – apex” is a fair start.

    Threat Ranking

    Once I have that list, I rank them in sequence of the logical progression of danger to the PCs. This will be the foundation of my placement within the different areas and the way I structure the flow of encounters within the dungeon. This is also when I think about the relationships between the different groups – in Fumanor, you had to go through a goblin society and a Minotaur society (which preyed upon the Goblins) with Troglodyte slaves (who were preparing to revolt and reclaim their former “lands”) before you could get to the Dwarven society. And below them, Shadow Dragons hunted Giant Spiders, which led to the Drow City that the PCs were actually hunting for. All in the former Dwarven mine-shafts of Mount Ayer. Was this a dungeon? Technically not – but thematically, yes.

It’s at this point that the time required begins to impact on what I can demonstrate in this example. Fortunately, skeletons don’t require too much in the way of ecology and have minimal society; they are more automatons. But I do need to think about the nature of the enhancements and the implication that not all of them are human skeletons.

I’m thinking that I want something tougher than human but human-sized; something bigger; and a dragon skeleton, as well as the basic Skeletal Warrior.

Bugbears fit the first, but the second is more problematic; the obvious choices (cyclops, hill giant) are too unintelligent, while most of the other choices are too powerful. But it might be fun to apply their regenerative abilities to a skeleton, so I’ll go with Troll for choice two. But I also want to give them all enhanced resistance to turning, automatic ‘curse’ (opposite of ‘bless’), and weapons of +2, and +3, respectively.

The Dragon Skeleton will need no such weapons, but the resistance to turning would still be good, and let the dragon have multiple attacks per round and maybe even a breath weapon of some kind (even without lungs). Maybe “Cloudkill” as a breath weapon.

Finally, we have the reanimated skeleton of Azelfar, who is closer to a Lich than he is to a skeleton, possessing intelligence, the ability to cast clerical spells, and a +4 mace, plus various other magical goodies. And, like a Lich, I’ll give him a soul jar – to turn him, you have to actually direct the Turning at the jar, which you will have to find first.

The next population to be considered are the were-rats. They exist on worms, birds, and small animals, but those are barely enough, so they scavenge for rotten meat in the old cemetery.

Next, we have the population of other undesirables. Dire Rats, Bats, Carrion Crawlers, perhaps an Ooze – those are all off the top of my head. I don’t want another variety of Undead, because that would distract from the skeletons. In reality, I would look through my various Monster Manuals (etc) for ideas; I’d probably want half-a-dozen or so. The criteria are that the creatures should be ‘comfortable’ around Undead, and living below ground.

Finally, we have the population of critters living in the wild around the dungeon. These would live on the traditional small herbivores like squirrels and rabbits, but would avoid the dungeon itself. Any of them that are intelligent may be able to offer valuable intel to the PCs if captured – a Troll requesting a parley might be interesting, but I’ve already tapped the “troll dispenser” for skeletons. Once again, I’d go through the books. I have in mind that the region would now be very overgrown with lots of old-growth forest – it’s had hundreds of years, so the undergrowth should be quite dense.

8. Amenities: Air, Water, and Sewerage

There are four things that most creatures require to survive – food, air, water, and some way of disposing of their sewerage. Food has already been covered under the ‘ecology’ heading, so that leaves the other three.

Air circulation is always a hazard below ground, and needs some solution – in general, each campaign should have its own bespoke solution to the problem (even if that solution is to ignore it).

Water is even more necessary than food – humans can live a week without food, perhaps longer, but a lack of drinkable water kills us off in three days. Fortunately, there are natural supplies of water than can be incorporated into dungeon design quite easily – but you must remember that each society that needs water to sustain life will need access to that source, or will need some other source.

The more answers you can think up to these two environmental requirements, the more practical dungeons become in a campaign. While players can ignore the question just as easily as the GM can, it always strains the credibility and suspension of disbelief; whereas, providing a specific solution ‘scratches that itch’ and lets players relax into their roles that little bit more.

Getting rid of sewerage is not something most GMs think about, but it’s not only a source of disease, it’s a source of nutrients for many plants and insects, and an important part of any ecology. On top of that, some waste, when exposed to water, becomes extremely corrosive – bat guano is notorious for this, but it isn’t the only example. All these effects impact on the dungeon. For example, in the Shards Of Divinity campaign, there were a number of scarab-like beetle species that liked shadowy places in the undergrowth and efficiently recycled natural waste. Some of these never went underground, some never emerged into the light, for example the Ivory Beetle. Snakes and Birds fed on these beetles (so there were a lot more of them), and higher organisms fed on the snakes and birds.

Skeletons don’t have any such needs or problems. I’ve already covered the water angle, anyway, but in the upper level, rainwater will penetrate any holes in the roof (which used to be the floor) and pool. The were-rat tunnels provide ventilation on the second level. Below that level, however, there are greater problems. Azelfar would have solved this problem while alive, though his solution may have fallen into decay since his apparent death.

To make all this work, I would need some rules on bad air (there might be some in a game supplement or I might have to write some house rules), with minor effects for level 3 and crippling, potentially fatal effects for level 4. And I would need to come up with some form of air circulation. My first thought was a water-wheel driven fan, but that would require knowledge of the layout of the dungeon, post-flip, before arriving. My next answer is a vine with leaves that continually flap to drive off oxygen and replace it with carbon dioxide; this is a solution that could be applied to multiple dungeons within a campaign because it’s naturally-occurring – but it doesn’t quite fit with the idea of letting the PCs find and reactivate the air machines that will permit them to survive on the fourth and final level of the dungeon. That suggests a mechanical solution of some kind – maybe a bellows worked by a couple of skeletons?.

Clearly, more thought is needed on this subject than I have time to devote to it in this example. So I’ll leave it there – something I would not do if I were doing this for real.

9. Space Allocation

Take the ecologies/societies from 7 and locate them in the areas from step 5, if you haven’t already done so, in order from strongest to weakest, on the premise that the strongest would have first choice of digs, followed by the second, and so on. That doesn’t necessarily mean that they are in the deepest part of the dungeon, it means that they are where they will enjoy the most satisfied life. That might be next to an underground river, or close to a tunnel or passage to the surface, or near a deposit of ore, or close to a food supply. The specifics will vary from ecology to ecology.

Once they are emplaced somewhere, the immediate thing to do (before moving on to the next ecology / society) is to think about how they would modify the area. What changes would they make? What would they ruin, and what would they improve? What resources might they need, and how would they go about obtaining them? This is the foundation of defining the occupants role in an economy, and it can give social groups surprising depth.

I’ve already covered most of this in earlier steps, so I can pretty much skate right past it, now.

10. The Dungeon Narrative

Edging closer to a completed design, the next thing to do is think about how the dungeon will play out as a plotline. I find that if I think of each dungeon as telling a story, I can make it a lot more interesting and interconnected for the players. As well as an overall general narrative summary, I will consider six specific aspects of the story.

    A: Approach

    What’ s the story of the journey to the dungeon? What do I want the characters to be running short of (if anything) and how can I arrange that? BA Felton, in a famous Knights Of The Dinner Table, had the players trek through a jungle populated with apes that threw stones with pin-point precision, and I can’t think of a better example. Sometimes, many low-level encounters are better than one big one! The problem with such is that they can be tedious – something to work hard on avoiding. And that’s where thinking of the journey as a story can be helpful – use the tedious encounters as content and more interesting encounters as punctuation.

    B: Interior Style

    This isn’t about the narrative look-and-feel of the dungeon (we’ve already covered that), this is about the style of narrative that you will use within the dungeon. Short, snappy sentences will make the characters jumpier. Long, drawn out descriptions with lots of ells and sibilants and a gentle cadence will rock them to sleep. There are other tricks that can be employed, but you get the idea.

    Most GMs and authors go for the big narrative hit, trying to create the atmosphere in one move – and there are times when that is the best approach – but it’s often better to set the stage with that initial hit and then build on that beginning repeatedly and regularly. I know one GM whose game room is fitted with a remote-control dimmer; he combines a hushed voice and gradual dimming of the lights to enhance the mood until suddenly the action starts (when it’s back to full brightness).

    C: Variations

    Are there any areas that will require particular variations on this style? It’s a very small nuance with little impact from one particular statement, but repetition and cumulative impact can be significant. For example, if you want to create uncertainty, end each block of narrative with a question, however rhetorical; if you want to create certainty, end with a statement (even when that statement is really a question in disguise – telling a PC that they are contemplating the choice of two paths, for example.)

    D: Setbacks & Triumphs

    While you can’t predict the action or the outcome of any encounter, you can think about the overall flow of the narrative and what the setbacks and triumphs will be. There should be some pre-planned small victories along the way, and some difficulties to overcome. This becomes especially relevant when there are two or more paths from A to B – any intelligent species will secure the least-defensible with traps if they can’t render it impassable or invisible, so as to direct enemies along the most-defensible route, which is where they will concentrate their forces.

    E: Loot

    Most loot is irrelevant in specifics, but there’s usually one or two items that are extremely important to the narrative.

    In the example, that’s the Rod Of Piety needed to chain Telvas once more. Where is it hidden? What does it look like?

    Is there any loot that is specifically to be emplaced to enable a subsequent part of the dungeon – with a submerged dungeon, some method of water-breathing should be made available (however difficult it may be to obtain). You may need multiple options, of varying usefulness.

    F: Climax

    The final piece of the narrative puzzle to think about in advance is the climax – what’s going to be the big finish of the dungeon adventure? And what do you need to do to make sure it happens?

    Telvas awakens and begins to appear on the big throne. The PC cleric gets his hands on the Rod of Piety and discovers that it requires a life to be sacrificed if a mortal is to bind Telvas once again. He prepares to make The Ultimate Sacrifice (TM) when Azelfar reforms (again) and makes one last attempt to wrest the rod away – just as the Rod is activated. Azfelar’s life is whisked away and used to once again bind the dark Deity that unnaturally extended it in the fist place. Thus, Telvas is hoist upon his own petard, and all his scheming has come to naught. Happy ending.

11. In The Afterwards

While this could be part of the preceding step, I think it’s important enough to be out on its own, here. What are the long-term ramifications (if any) for the campaign? Where does the road go from here? What makes this dungeon part of a campaign, beyond the recurrence of player characters? Where does this dungeon fit into the bigger picture?

>My example obviously has a somewhat disquieting aftertaste. The PCs only just succeeded because there was no strong Priesthood here. 500 years from now, if they remove the Rod, who knows where it will be when the critical moment comes? Only if the Priesthood of Belaal is restored and the Rod entrusted to a worthy and reliable high priest will the world be safe – but this time, the kings and queens of the realm, and its military commanders, need to be made aware of why the Temple is important, so that they can protect and sustain it.

12. Enfleshing

The final step is to stop with these design thoughts and get down to work on the specifics, fleshing out each area and each location, each encounter and challenge, and each reward – putting meat on the bones, in other words. The first sub-step is always to put those bones in structural order, of course. This is where you draw your maps (or label someone else’s map), and so on.

You’ve made all the broad decisions; now all you need do is incorporate those decisions into the specifics.

Go Forth and Play!

That’s how I design a dungeon. Now you can do it that way, too.

Oh, what the heck – one more quick (partial) example to drive the point home. A dragon’s lair, methinks – areas include the play area, sleeping area, bathing area, the larder, the scrying pool, front gate security, perhaps a workshop…

989… 990…

I’m still looking for ideas on how to commemorate my 1000th post at Campaign Mastery. just ten to go, the countdown has begun!


Discover more from Campaign Mastery

Subscribe to get the latest posts sent to your email.