Monday, 21 July 2025

Populating the megadungeon part three: Balanced vs. randomized

 

"I'll take the giant for 500 XP, Alex."

Balanced encounters vs. Randomized threat levels

The current philosophy of Dungeons and Dragon game design is that encounters should always be balanced. You take the party level and from that calculate how dangerous the monster should be. A traditional megadungeon is actually based largely around this idea: first level down is for first level characters, level two is for second level, third for level three, and so on. By the time they reach level ten, they’re tenth level. 


This, however, is only a general guide, and players will often delve deeper before they are prepared to. If you look at some old dungeon modules, you’ll sometimes find significantly more dangerous monsters on upper levels. I prefer this, as it teaches players that there are some threats to be avoided, and not to think they’re at the centre of the game universe (yet). They have to earn that. But you have to let players know the world is terrifying and that not everything is scaled to suit their skill level. They have to exercise their own judgement in deciding when to engage, when to flee, and when to hide. 


Famous last 1st level words: "Come on, you guys! We can take down a 20th level monster if we use the power of cooperation!"

If you do throw in a dragon squatting in the abandoned Dwarven Hall of Grimiron on the first or second level, be sure you give the players plenty of clues as to what lies ahead. Maybe the kobolds worship it, and have built altars, and scrawled their dragon god’s story on the tunnel walls in front. 


You want your players to have the option to quickly retreat if they realize they’re in deep trouble. 


If players insist on being overconfident, or even outright boneheaded, well… I believe killing player characters should be avoided whenever possible, as players invest a lot more energy and backstory than they ever used to. Dying was a matter of course in AD&D. Now? Not so much. 


But the threat of death still has to be there; it’s a vital part of a fantasy adventure. No risk, no reward! Offer players saves and outs and second chances whenever possible. Remember everyone makes mistakes; after all, seeing characters grow and advance is part of the fun! 


If a player does die, do whatever you can to make it meaningful. 


A player dying during a random encounter is pretty anti-climactic. Dying during the boss fight, saving the lives of the rest of the party? That’s epic. 


Always go for the epic and memorable. Stories can come out of good D&D games, ones that will be told and retold in the years to come. Aim for it.


Add breathing space

Dungeons are more believable when they have ‘empty’ spaces. These help build up tension. However, they can’t be (generally) truly empty. They need items of interest, because players get really bored, really really fast. A couple of empty rooms in a row will do it. So be sure to offer clues as to what is ahead (massive scat outside an owl bear or dragon lair, claw marks, offal, broken weapons, stone statues of adventurers petrified by the basilisk down the hall,  the smell of cooking meat, the foul stench of ghasts, etcetera). Maybe the giant ants build geometric sculptures out of dirt to mark territory between colonies. 


I always feel some anxiety that players will get bored. It makes me throw in more monsters, and more varied monsters, than is believable, even within a fantastical dungeon space. I try to use all the tricks mentioned above to keep it exciting. 


Ultimately, I am only trying to make the mega-dungeon passingly plausible. Enough so that players aren’t constantly stopping to say, this makes no sense! That’s death. So always listen closely to your players. They will tell you as the game goes on what works, what doesn’t. You’ll learn the limits of their suspension of disbelief bubbles. 


Count on that!


Evil avalanches... hmm... why, I could use that!

Player feedback

Getting players to clearly state what they like and what they don’t can be a double edged sword. You don’t want to open up the door to nit picking and negative energy. Most players won’t want to give offence, and may pull their feedback punches, or be hesitant to offer criticism at all. 


If the criticism goes to far, it can kill the vibe. As such, it’s better to ask what’s working and what they’d like to see more of, than what they’d like to see less of. 


If you emphasize the positive, the rest will naturally sort itself out. 


The Rule of Three

The Alexandrian (Of So You Want To Be A Gamemaster fame; it's good, I picked up a copy) puts forward the Rule of Three: one monster, one trap, and one item of interest per room. Or something like that. I can't find the original post, but he does also apply it to clues, which is a very good idea, as well. I’ve tried to implement it in my dungeon. 


Unfortunately, I made my dungeon too big, and putting three things in each… after the first 100 or so rooms, that was just too much for the scope and scale I was aiming for. Arguably, the scale I wanted is nuts. Arguably, you only need maybe 10 rooms tops to go from level one to level two, and maybe 20-40 rooms per level after, depending on the monster challenge level and size of your player group. 


So after running headlong into the brick reality wall, I’ve scaled back. Some of my rooms only have 1 thing. Others might have 4. I vary it for the sake of my own sanity and the limit of what I’m capable of without burning out. 

Other posts in the series: 

DM Journal 1: Jumping from AD&D to 5e

Populating the mega-dungeon part one: Mix, match and batch


Monday, 14 July 2025

Populating the mega-dungeon part two: Magic (and story) is your friend

"I don't think this spell does what you think it does!"

Magic is your friend

If you don’t feel like working out a vague ecosystem, just say the dungeons magical. Have the players come across some dedication written by a long dead wizard who forged this magical space. 


Magic can be the answer to almost everything. 


Magic is the Dungeon Master’s best friend.


You don’t need a fully worked out food chain. That’d be an awful lot of work and kind of nuts. There’s no way so many apex predators could coexist in such a small space, there wouldn’t be enough for them to eat. You’d need a dungeon tens of miles across. 


You only need is to establish the semblance, the impression, of an ecosystem. 


...just filled with ravenous man-eating monsters, demons, and pits of green slime.

Backfill a backstory 

Another way to add variety when characters are plowing through an area dominated by one monster type is to add backstory. Sometimes, the need for this only becomes obvious when you are populating an area. 


If the dungeon is old and our orcs from the example are not the original inhabitants, you can layer in elements rom earlier eras. Perhaps the orcs have moved into an old arena, and there are trap doors and a labyrinth beneath for monster holding pens, along with weapons, unspeakable horrors, and undead gladiators. Or they’ve occupied an old prison, or an old dwarven forge. That can spark all kinds of ideas: dangerous abandoned equipment, elemental powered furnaces, cryptic murals, cells with immortal prisoners, torture devices containing undead victims, wall carvings that hint of deeper treasures that the characters are seeking. 


Adding an earlier purpose creates entirely new adventure hooks. Players can take them or leave them. I try to throw in a lot, because many (most) won’t intrigue. But when you have a hit, mine it like a Hollywood executive exploiting a time honoured franchise!


I recommend mega-dungeons be abandoned, repurposed spaces. If you work out the story for the duengon’s earlier inhabitants, it gives you direction and purpose when seeding in elements. The space is imbued with an underlying logic that smart characters can suss out, allowing them to piece together the past, possibly leading to greater rewards.


Getting all that to work can be challenging, and it’s easy to overcomplicate things. That’s what do: I plan overly ambitious backstories that won’t conclude until… uh… maybe level 10, and I start setting them up on level 2. 


By the time I get to level 10 I guarantee I’ve forgotten what the heck I was doing on level 2. 


Fortunately, the players will be your guide, letting you know what story hooks to nurture, and what to let fall away. They won’t remember the threads left dangling because those threads didn’t interest them in the first place.


This also argues for the living dungeon. 


"What will suddenly introducing all this loot into the local economy do for inflation?"


The living dungeon 

Don’t finish the mega-dungeon before you start. Trust me no this. Megadungeons are big, sprawling feats of imagination, almost herculean tasks, and if you try and do it all before players go through, you risk investing too much into a dead end path.


All you need are the first 3 or so levels plotted out. If you’re really ambitious, you might rough out some lower level maps, but you don’t need to plot everything out yet. Plunk in a few monster possibilities so you have some inspiration at hand in case the players get ahead of you.


Instead, listen to your players as they progress through the first 3 levels. See what interests them, and what they couldn’t care less about. Have a bunch of story seeds that you can plunk in, and a vague idea of the resolution, and see what characters latch onto. The mysterious lost tome of the Necromancer? They couldn’t care less about. But the Treasure of the Gorgon Queen? They’re all in! 


Pivot to what the characters are intrigued by and alter the dungeon to fit. This doesn’t mean abandoning what you like. You’re just emphasizing elements you already had, to pull the players in more effectively. 


Remember that the dungeon is a shared space, living in the imagination of both DM and players. Everyone has to be interested, if you want to keep your magic imaginary realm alive. 


Another issue is that players level up much, much, MUCH faster than in AD&D. You needed a lot more treasure (you got XP points for treasure back then) and monster kills to advance then. To reach level 2, some classes needed 3,000 XP. Now, everyone only needs 300 XP. 


This means your characters may advance in levels musch faster and more erratically than you might be prepared for. You then need to buff up the dungeon denizens as players have outpaced the challenges you originally set. 


Phil and Dixie
Another fabulous What's New strip with Phil & Dixie, from an ancient copy of Dragon Magazine

Other posts in the series: 

DM Journal 1: Jumping from AD&D to 5e

Populating the mega-dungeon part one: Mix, match and batch

Saturday, 5 July 2025

Populating the mega-dungeon part one: Mix, match and batch

Early era dungeon ecology

Populating a dungeon is always a challenge. You need to keep things varied and unpredictable, yet provide a believable environment so players can suspend disbelief. 

Balancing all the elements perfectly isn’t easy. 

As they say, the no plan survives contact with the enemy (or in this case, the players). 

Sometimes it takes a couple play throughs to find all the weak points in the dungeon’s design. Players being imaginative and innovative, it may require more. 

Many, many more. 

Mix it up

Generally speaking, I like to populate a small cluster of interconnected rooms with a sapient species like kobolds, orcs, or goblins, have an empty room arround the edges, and then populate a non-sentient monster. The sapient monsters are the dungeon population hubs; they organize regions and monopolize turf. They typically won’t tolerate anything right next to them, so there will be a buffer between them and the rest of the denizens. You can put traps or other items of interest in such rooms (such as heads on pikes, warning messages, refuse dumps, scat, stationary monsters, etcetera). 


A sapient species will have a safe and secure water source within their territory. No one wants to risk death at the hands of innumerable dungeon horrors every time they want a drink of water. Water is essential for cooking as well. If they don’t, make it an adventure point: have multiple factions fighting for control over a water source. Gold is all well and good, but you can’t drink it. 


Food and water and safety are existential issues. 


The rest of the dungeon will be a chaotic space. Large and powerful predators will have both a lair and a hunting range. Anything near their lair will be a smaller monster, one capable of evading it, whether by hiding in holes in the walls (like giant rats or ants), in the rafters above, or by some other means. Perhaps it is inedible, like skeletons, golem, living statues or elementals. 


The options are as endless as your imagination!


And if you need help making mix-matched monsters, books are just a click and a few bucks away!


Tear down those doors, Mr. Gorbachev! 

A lot of monsters have no hands. There’s just no way they can open heavy dungeon doors. So if you want wandering monsters, the critters need to be able to move freely through the common areas. The bigger the monster, the bigger the hunting range they need to reach their daily calorie count. They can’t be impeded by locked doors. 


As such, you want the majority of the dungeon halls to be interconnected and unobstructed. 


I try to take this into account, but sometimes I wind up putting a door where I shouldn’t. So long as the logic generally holds, players won’t notice a few gaffes.


I do like a good dungeon wall

Tribal monopolies 

Let’s say a tribe of orcs has settled down in a dungeon quarter. Players can expect to fight room after room of, well… orcs. Everything else will have been cleared out by the orcs, who will work to secure their turf. They’ll then install barricades, traps, and sentry posts around the periphery. Hence, you meet only that kind of monster in their lair. 


This, however, breaks the rule of keeping things varied. 


You can remedy that by varying the orcs (big orcs, little orcs, berserker orcs, sick and contagious orcs, drunk orcs, sneaky orcs, etc.), and throw in some shamans. Presence of a shaman suggests worship of an evil god, so you can throw in a demon haunted altar. Or maybe they worship a snake god, and there are snakes slithering around. Or they have a spider hatchery, and train the eight legged horrors as guards. Or they might have attack dogs. 


Kobolds might raise centipedes; the shaman could initiate new warriors by letting centipedes bite them, bestowing immunity to their poison. When they go hunting, they do so with a horde of the little creepy crawlies climbing over them... and their enemies. 


Stationary monsters with limited reach, like violet fungus, brown mould, and tentacled horrors in deep dark wells work, too: the orcs will know to give these areas a wide berth, but intruders won’t. 


Prisoners that are actually dopplegangers, cunning gnome thieves, and rich potential patrons can add spice. 


Tie them in to the character back story when you're able. 


"Oh crap, it's Aunt Ethel!"


Double and triple threat

To make encounters more interesting, combine two very different monsters into a complimentary threat. Put a brawler type with a spell caster, missile troops with brawlers. Have the goblins travel with giant spiders, kobolds with their pet basilisk, orcs in service to a necromancer, blind grimlocks with a medusa, bandits with dire wolves, ghouls with bat swarms, and so on.


Think about monster relationships: how might they be complimentary? Pair a fast and weak monster with a slow and tanky one. Back up bruisers with a spell caster. 


Goblins have hands and can open doors, spiders can’t, for example. Skeletons can be paired with gelatinous cubes and green slime, as they have no flesh to devour. 


A semi-symbiotic relationship


Dungeon Diplomacy

Sapient factions will naturally compete with each other, and to compete more effectively and dominate their enemies, they will form alliances. Counter alliances will form to resist and disrupt potential hegemons. 


When the characters arrive, the most interesting situation to find is a precarious balance of power, with two or more factions deadlocked. 


The players then become the tie breaker, and they can play the factions against each other for their own benefit. By allying with one group or another, clever players can clear out the dungeon without having to do all the heavy lifting. 


You don’t want to make this too easy for the players, but you do want to reward clever game play. Whenever there’s more than one faction in the dungeon, work out their political relationship, and one or two incidents that have occurred between them to cause the bad blood. 


Maybe the kobolds sacrificed the goblin queen to their dragon-god, or the goblin tribe is composed of former slaves of the bugbears, or ruthless dwarves. Invasive troglodytes might be eating up the dungeon food supply that the goblins were carefully managing. A basilisk might be turning all the feral pigs, the food supply, to stone... perhaps some of the humanoids in the dungeon will want to hire the player characters to kill the beast.


"You have no respect for the dungeon ecosystem!"


Delicious in Dungeon

If you haven’t seen the show, give it a watch, it’s a hoot. The creator of the original manga was very careful in working out an ecosystem and food chain within his dungeon spaces. I’ve always struggled with how so many monsters can be in close proximity without food sources, unless they are eating each other. 


I added feral albino displacer pigs to the Druidun dungeon, along with thick moulds (with various side effect bonuses) eaten by hideous, poisonous centipedes, and hairless dungeon blink bunnies. All three breed fast, and act as food sources. 


Fungus is always great; I put it all over. Easily recognizable types have a variety of effects: some emit knockout spores, and then grow on the slumbering bodies. Others give immunity to poison, or greater magic resistance, or even a few puffs of heartburn fire breath. 


Underground lakes teem with translucent fish. 


Magic spheres allow trees and plants to grow. 


Centipedes eat the fungus, giant rats eat the centipedes, goblins eat the rats, grick eat the goblins, carrion crawlers eat what’s left. Boom! Done. You've got a food chain. Let the players come across a few incidents of it in progress and you’ve established a plausible ecosystem in their minds. 


Delightfully yum

Other posts in the series: 

DM Journal 1: Jumping from AD&D to 5e


Monday, 21 April 2025

Building a megadungeon: Level design

Level 3 of the dungeon, featuring 3 panopticon prisons, old druid refuge to the northwest, and grand dwarven halls to the south.

Once the history is established, you can set about designing the levels. 

First, create the levels according to the original purpose of the dungeon. Is it a mine? A prison for malcontents? A refuge? A storage site for powerful magics? A vault? 

Purpose drives design. 

Once you have that, you can start laying out the rooms. A prison will need areas for guards, prisoners, latrines, kitchens, waste disposal, barracks, bureaucrats, and so on. 

You'll need to plot out things like ingress/egress points. For a megadungeon, you'll want a number of them. If the original purpose wouldn't have many entry points, add new ones created by subsequent inhabitants, like goblins, kobolds, ankhegs or purple worms. All of them dig tunnels. Xorn might follow ore streaks. Umber hulks might make tunnels following the same, and in so doing connect different areas and create new exits. 

Airflow is also important. Druidun has multiple airshafts going to the surface. Hot air rises, so at the bottom of these shafts are Dwarven Firestones, red hot slabs that cause the air to, well, rise. Shafts may be split into two sides, one heated and one cooled, to generate airflow. 

The dwarves also have many Skystones (which draw in air from the plane of air), and Skygates (gateways placed outside the dungeon, preferably near sea level, which suck air into the dungeon). Magic can fix most things. Floating sunspheres, delicate orbs that emit just enough sunshine to grow plants, can help diversify the underground food supply, which typically is almost entirely dependent on organic matter falling from the surface. There might be lounges where underworlders bathe in the light from such spheres, soaking up vitamin D. 

Dwarven Runecarvers will bring in air and water elementals to manage air and water flows within holdfasts. Earth elementals can be automated to conduct repairs on damaged sections, and these magical beings may persist long after the original dwarves have long since departed.

Water is absolutely essential, and multiple sources are needed on every level. Several should have open access. Watering holes are dangerous areas and key hunting grounds, as such, most creatures won't want to build their lair over one due to the high traffic... unless it's fortified and within established turf. 

phil and dixie
From Dragon Magazine in The Before Time

There are a number of competing priorities when planning a level: 

1) Verisimilitude. You want the space to feel believable, at least on a superficial level. That means food sources like fungus, airflow, water and light sources, etcetera, as described above. Hazards like anoxic areas, methane pockets, moulds, fungus, slimes, and so on help flesh out the environment and add spice. You want to at least suggest some kind of an ecosystem, even if, on closer examination, it wouldn't really work.

2) Functionality. This is akin to 1, but more purpose driven rather than environmental. Dwarven mines and forges should be laid out (vaguely) logically; mines would connect via rail lines to ore refineries. Kitchens would be near the great halls, along with beer and wine cellars. 

Latrines would be separated, so the smell doesn't penetrate the living areas, and beneath (on the level below), there would be the cess pits. Fecal matter is often used as fertilizer, so there might be old, overgrown mushroom farms nearby. 

It doesn't need to be perfect, just enough so that players get the idea that there was purpose driving the construction. Even more importantly, these details can help you draw out your dungeon with purpose. An underlying purpose helps create that bubble of disbelief, transporting your players to another world. 

3) Puzzle & mystery. Functionality is (sometimes) at odds with this. You want your megadungeon to be full of surprises, twists, and breathtaking vistas. Stairs that go down every level, top to bottom, make it easy for characters to move around. This makes sense from a functional point of view, but then there's no quest to find the next set of stairs. 

You can justify deliberately confusing layouts as part of a dungeon's defenses: levels are mixed up to put potential invaders at a disadvantage. This can also inspires traps, ambush points, and other defences. 

Throw in magical forces to create intrigue and mystery. Treasure vaults will be surrounded with misdirection, puzzles, and traps. Dwarves will be especially aware that their vaults can be dug into from virtually any direction, and plan defences accordingly. 

4) Awe. Strictly functional spaces don't require narrow 10 story high temples, labyrinths, or grandiose pillars of spirit infested flames. But you want cool stuff in the dungeon to capture the imagination of the players! A megadungeon has to walk that fine line between believable and mind blowing. That's the goal, or at least my goal. 

Remember that the purpose driven design that determines the layout provides only the foundation. You can overwrite that with all kinds of awesome phenomena that occurred long after.

Dungeons should be dark
Always sage advice from Phil