Dungeons Part Three: Running a Dungeon

July 18, 2014

So, now we’ve covered some basic theory about the design, and laid out the fact different groups want very different kinds of challenge from a dungeon delving game… let’s go into more specific applications to play.  (if this looks like a lot… well, it is. Like I said, dungeon design is level design is game design.  Doing it well requires a lot of forethought and there is a reason the videogame industry PAYS people, full time to do this kind of stuff…)

On the Grid, To the Table

Depth vs. Novelty

Aside from constructing your dungeons to meet your players’ particulars about Exploration/Logistics/Tactics, you also need to think realistically about whether your group wants a longer, megadungeon or a short dungeon they can clear quickly, and, how long that should go.  People really like having benchmarks to go by and a sense of accomplishment.

One of the problems is that a lot of old school dungeon building advice assumes weekly games, long sessions, over years of play… and ends up teaching people to “put the cool stuff” deeper in the dungeon.  (It’s also why I think the modules aimed at higher level play were generally pretty popular – the cool stuff kicks in at the beginning much more often).

Prep only what you need so you don’t end up wasting a ton of time on something that’s not going to show up in play.

Engagement Density

How fast will the players encounter challenges, puzzles, or rewards in the dungeon?  This isn’t so much how far apart are they spaced physically, as much as it’s a question of action time at the game table – “Oh, they should get through this room and into the next with 10 minutes of play.  That fight should take 30 minutes.”

This is a highly variable thing, but you can see that density is much higher if the dungeon is laid out simply, if the challenges/puzzles/rewards are tightly clustered, if there’s not a lot of obstacles from point A to point B, if there’s not a lot of traps or secret doors or other obstacles slowing movement or forcing the players to be extra cautious.

One of the old school pieces of advice is “Most rooms should be empty” which, may be more realistic, but it’s not particularly fun to have most of the rooms empty – it means a lot of wandering around to find the fun, and also, if the game features traps, ambushes, or cool rewards hidden away, a lot of time spent searching and cautiously entering every room.  Empty rooms are generally not fun.  It also means you have to either spend more time coming up with creative descriptions of non-interesting areas, OR give it a barebones boring description as well…And, you’re competing with every other kind of game your players could be playing right now – so you want pretty high engagement density.  (Check out Spacing The Threats below for a bit on the strategic issues with this, too.)

Obvious Exits and Highlighted Secrets

When a videogame is doing it’s job, you have a good idea of where to go, or what to do next.   When a videogame is doing poor design, you’re lost, you’re stuck, your not even sure what you should be trying to do.  Videogames have quickly recognized that players need adequate information to make meaningful choices, and poorly communicating that makes it feel like you’re being cheated rather than challenged.

This is “obvious exits” – big colorful paths pointing the way, a giant gate, a glowing door, or more subtle, lights right at the entrance, etc.  In roleplaying, be sure to point out the things that are obvious and important to the characters – make “obvious exits”, whether that’s literal exits or things worth interacting with.

The other trick videogames do is for many games, there’s often tiny clues or giveaways that something is a secret area worth poking at.  A statue out of place among similar statues, a place where all the flowers are colored differently for no apparent reason, etc.

Why prep a bunch of secret things if no one ever finds it?  Unlike a videogame, you’re not going to have thousands of players running through your dungeon, nor will people be able to play alone and poke at the dungeon, so you actually want to give better odds of finding stuff for your group just to make it worthwhile.  These highlights do not have to tell the group HOW to get into/at the secret, just that it’s worth investigating.

Dungeon Pitfalls: Red Herrings

Lots of dungeons and dungeon advice seems to have Red Herrings – things that look important, but turn out to have no secret at all.   This is a terrible thing to do.  You end up wasting your players’ time, and it teaches them NOT to look or investigate on things that seem weird, but to pass them up for things they KNOW are worth their time.    It’s like the Empty Room of features.

From the Grid to the Players


While it’s easy to see the dungeon as mazes + monsters, it’s also fun to consider how the dungeon can be a piece of world building for your game.  As the players wander through an area, give them descriptions based on their characters’ understanding of what they’re seeing.  Fill your dungeons with history – the Dwarf recognizes the carvings of a lost Dwarven Kingdom, a human sees old shields with insignias from an army that was defeated in her mother’s lifetime, etc.  Give more details to players whose characters have history or knowledge skills as would be appropriate.

If you can, make these relevant to actual conflicts/characters they meet now, not just random filler.  Maybe they find out the friendly wizard in town, his family used to live in this keep before it fell into ruins.  Maybe the Cult of the Red Savior is actually a terrible misunderstanding of an Elvish mural from a famous play…

Groups that love lore and history will eat this up.  This pushes on some players’ love of Exploration, but it bores people who don’t really go for setting and world building stuff.

Dungeon Politics

Not all of the monsters are working together.  They might even be in the same army/cult/whatever but have internal rivalries.  Where there are intelligent creatures in the dungeon, you can assume many of them either have alliances, rivalries, or out and out hatred for each other, and when the PCs enter the mix, all these groups will respond to what happens.   Smart players can make allies or get monsters to take each other out.    There might be creatures who need to be rescued or protected – in which case you have a heroic situation int he making.  You can set up all the usual issues of politics, needs, and pressure among the monsters and just make sure it gets communicated to the players as you play.

Groups that love social roleplaying and deception schemes love this kind of thing.   Players who don’t care and simply want to kick in doors won’t care.

Leaving a Mark

Dungeon delving is disruptive… things get damaged, destroyed, entire monster groups get driven out or wiped out.    If this dungeon is going to be a focus of play for many sessions, take notes of this.  Note how the players change the setting and how the creatures react.  Be sure to remind the players as they return through areas, all the little (or not so little) marks they’ve left:

“You still see the arrow shaft near the top of the door where you fought the Ghost Knight.”

“The stench of the rotting giant lizard is nearly completely gone, scavengers have picked it over pretty well.”

“It’s easy to find the secret door again, you just look for the charred blast marks on the stone from the fireball fight a few weeks ago…”

While it’s easy to just have superficial things, also make and expect objects for the players to break or use to change the possible ways to MOVE in the dungeon itself.  Maybe the players used magic to form a tunnel between two rooms that didn’t exist before, or they toppled a giant statue over a chasm to serve as a bridge.

These little things are just FUN, and it gives the players something they don’t get from videogames – a chance to realize the dungeon isn’t an unchanging, automated area to run through.

Adapting Threats

This is a dial – on one end, the creatures in the dungeon do not change or adapt in response to the players.  This is pretty much like a videogame at this end.  On the other end, the dungeon dwellers set up traps, move, cave in areas or build barricades, set up complex ambushes, etc.    Now, the dial doesn’t just affect how dangerous the dungeon is, but it also affects the engagement density…

See, one of the benefits to a longer dungeon delve is familiarity with areas you’ve already gone through and relative safety to run in or out of those areas.  If the areas are repopulating with new creatures, or the other dungeon dwellers are already moving in and changing things, it means previous “cleared” areas are not actually cleared at all.  (I’ll go into some options of what this can look like in the next post.)

It’s realistic and dynamic, and that part is fun, but it also means the players might feel like they’re getting nowhere in beating the dungeon.   The time it takes to get through areas you’ve run through before stops being fun and becomes a chore if the threats are too dense or it happens all the time.

Talk to your group!  Some folks really like the idea of a living world and having to adapt to meet it, other folks feel like it makes play feel like an unrewarding struggle.


Logic Puzzles vs. Context Puzzles

Puzzles are hard to do in tabletop play for two reasons – puzzle games are normally a solitary activity.  You can take all the time you want to mull over it, put down the videogame or walk away from the puzzle book and think about it while you go about your day and come back and then solve a puzzle.   When you’ve got a group of people at the table, everyone’s time is on the line and you can’t quite do this the same way.   This is the reason most puzzles in tabletop games are side options – they open an extra area, or give you some extra bonus, but they’re not the core of play.

You end up with two types of puzzles – logic vs. context puzzles.  A logic puzzle focuses on elimination or procedure to create the puzzle – things like the fox, the hen, and the corn across the river kind of things.  You know what your end goal is, you’re just not sure what order to do things in.  Or process of elimination type puzzles, which again, you know there’s one correct answer, you just have to figure out all the ones that aren’t it.

The one problem with logic puzzles is that your group is either skilled with them or not.  If they are, they solve these in short order and the puzzle becomes a minor chore rather than a fun thing.  If they aren’t it requires more work, and then comes the question of how much work is still fun vs. crossing over the line where they just decide to ignore it.

Context puzzles depend on context outside of the game – stuff like riddles, knowledge of greek mythology, etc.  These never work that well because not only does it depend on the players having that context, it also depends on the players thinking in a completely different context than the world the game is taking place in – so they might know the answer but not think of it from that alone.  I see people post about “the players didn’t get it! It was SO OBVIOUS.” and it’s like… no, actually, context puzzles are never really obvious.

Spacing The Threats

One point of Engagement Density is monsters.  How tightly packed are the monsters?  If there’s a fight in one room, why don’t they all run and mob the party right away?  This is the issue of spacing threats, and I believe, one of the reasons old school dungeon advice often recommends a lot of empty rooms – to give the party a better chance at survival.  That said, I still think the empty room technique is bad for game play.

If the monsters are not cooperative with each other, many will probably stick to certain safe areas, or only come out some time after things have quieted down, looking for carrion or things to scavenge.  Intelligent monsters might hear the commotion and decide the best thing to do is hide or fortify defenses until they know what’s going on.

If the monsters are cooperating with each other, that’s a much harder set up.  You can have them with lax and poorly considered communication situations (outposts far out in the dungeon outside of hearing distance), drunk or lazy or absent guards, or perhaps other problems preventing them from reaching each other (like other monsters between them).   Even well organized groups may have internal rivalries that get in the way – “Intruders are attacking the east tunnels?  Let them!  If Captain Grimfanger dies, I get command!”

If you are dealing with an organized group of monsters, who do work together well, your threat is multiplied GREATLY.  Assume any encounters will be with many of the nearby groups as well, and figure out how long it will take for them to figure out there is danger, grab their gear, and get down there.  In older D&D, where a round was 1 full minute, that was usually only a few rounds at best.  In newer D&D and many other games, a round of combat is 10 seconds, 6 seconds, or even less.  It’s quite probable that a combat will be over before the reinforcements arrive… buying the party a little bit of time to set up.

Strategic Space: Chokepoints and Flanking

When you design your dungeon, if your players like tactical challenges, you have to consider a variety of space in which to move and fight.  Realistically… dungeons were cramped, small and not particularly great spaces to have fights in.   Realistically, you don’t get wizards or dragons either, so screw realism.  Whereas overall dungeon design benefits from learning from Metroidvania videogame design, most Tactical play design benefits specifically from looking at the level design philosophies in First Person Shooters.

Usually, two types of set ups work well for interesting tactical play, along with gearing fight sets specifically to whatever stunting/tactical bits your particular game system supports:

Open space plus obstacles

A larger room where the whole party can fight.  This is usually going to be 30-40 feet on the smallest end, and up to 100 feet on the large end (with anything bigger than that being so large as to being a “open field” as far as tactics are concerned.

This space will do well with a few obstacles in it that can be used as cover, things to get around, climb on, etc.   This can be support pillars, furniture, book shelves, fallen debris and supports, pits and cave ins, etc.  All of the junk serves to provide places of defense, choke points and things to flank around.  Ranged attackers must deal with cover, melee attackers need to move around these things.  Clever use of powers and magic might alter the terrain some, and monsters may be able to ignore or easily deal with some of these barriers – a swarm of bugs just flows through/over it, a giant monster might casually topple anything while coming after you…

Tunnels and Flank positions

The small cramped areas are fine, as long as there is plenty of paths to flank each other.  It gives the party multiple avenues from which to protect from attacks, and avenues in which to launch them.  It rewards cautious players for covering the sides or rear areas, and it rewards the clever players for finding ways to attack the enemy from the side or behind.  Don’t forget this can include vertical movement – climbing up onto something and jumping down is also a flanking move.

You can adjust a lot of difficulty based on how well the monsters make use of this.  Things that can climb on walls, burrow through the dungeon, are amorphous or outright ethereal are absolutely terrifying threats in tunnel fights.  Think of the xenomorph from Alien and you’ve got a good idea of how bad it can get.  This becomes worse if the creatures not only are good at setting up flanking and ambushes, but regularly use retreating tactics to wear a party down.

Also remember that winning the fight can be separate from beating the opposition.

Fun Traps vs. Crap Traps

The “game” part of any roleplaying game depends on this: what choices do you have to make, and why is it fun to make those choices?

In many dungeon crawls, traps are shit.   The choices you make are not really choices.   It’s either “woops, roll a random die to see how badly you get maimed” or it’s a choice of “Spend a tedious amount of time, searching everything, for traps, all the time, or suffer.”  Those aren’t fun choices.  (Now mind you, if you were playing very old school D&D where everyone gets 10-20 PCs, what happens is a trap takes out 1-2 of your guys, THEN you have a choice about dealing with it with the remaining party.)

Fun traps are mostly puzzle traps.  They’re elaborate deals that take several turns to kick off, and you have to figure out how to get out or disable them while they’re in action.  This requires a lot of forethought into making sure the goals/options are obvious, but the procedures in doing so are more complicated.  For the usual “dangerous dungeon feel”, I prefer Hazards…


A rickety bridge, a partially collapsed wall, a room half flooded…  Hazards aren’t traps in the sense that someone set them up and hid them, they’re obvious things that are a pain in the rear to navigate, or things that are obviously dangerous.   The fun in hazards isn’t information scarcity – you see them, you see they’re trouble, but it’s in the Logistics of getting past them, and how you can use them as choke points or ways of hurting the monsters.

Hazards also require less justification.  You don’t need to explain things like “Why the hell would someone build this falling block trap HERE?  How the hell do they reset it?” etc.  Hazards are just the natural ruin of the place and it makes rather mundane skills like climbing, jumping, balancing, repairing/jury rigging things, using poles, hammers, spikes, ropes, etc. a fun thing to add into your game.

Next: How to build your dungeon starting with Threat Structure as your theme.

If you find my blog entertaining and valuable, consider supporting me on Patreon.

%d bloggers like this: