The Lost Art of Dungeon-Crawling

There is a certain type of adventure that in recent years seems to have fallen out of popularity: dungeons.

Picture courtesy of Pixabay.​

"I Look Up!"​

The release of several old D&D modules got me looking at some of these old original adventures, and they are quite eye-opening on the subject of dungeon design. These are the first games of D&D we played and while some are a little dated, it's easy to see why they kept us playing. While almost everything in those adventures was dangerous, there was magic and mystery in the rooms you found. There were rooms with strange orbs suspended from the ceiling; mysterious indoor gardens full of medicinal plants, poison and monsters; ghostly feasts that share a tragic history; and mysterious keys guarded with fiendish traps.

I think I know why dungeons fell out of vogue. Way back in the early 80s we discovered city adventuring. Modules then became quests or investigations across a cityscape full of NPCs and role-play opportunities with all manner of details and cultures. This new way of gaming outside a maze opened a whole new sandbox. This change in adventure design opened new vistas for adventurers, but crowded out the traditional dungeon crawl as a result.

A Return to Form​

Luckily, in recent years we have seen a more interesting return to dungeons. More designers are coming back to them and trying to break the myth. Mork Borg has its share and a other ‘old school’ games have sought to blow the dust off the idea of raiding underground facilities. Its fun to dive into these lairs once again, and a simple diversion from what has become the usual kind of game. While I’m certainly more on the side of narrative play and character interaction, sometimes it is nice to know that you just need to pick a door rather than work out the villain’s plot and craft an elegant plan (that one of the players may just ignore anyway).

If you are thinking of crafting a dungeon of your own, here's a few pointers.

Give the Place A Reason​

Whether it is an old ruin or an underground laboratory, make sure the dungeon has a reason to exist and some sort of history. A hole in the ground isn’t very interesting so give it a back-story, even just a small one. It might be a tomb, an old ruin creatures have taken over or a lab where magic went wrong. It need not be especially clever, just as long as you can place it in your setting.

A Dungeon Need Not Be an Actual Dungeon​

What you are creating is a place full of rooms linked with doors and corridors, so it need not be underground. A house or a castle is basically the same, as is a sky city, large airship, underwater citadel or even a walled in town (put a roof on real world Venice and you have an epic dungeon).

Don’t Construct It with Only One Path​

When you are making a lot of cool stuff it is very tempting to make sure none of it gets missed. But you should avoid the temptation for having only one path through the dungeon that takes in every room. If the player characters miss out rooms 34-48, you can use them in the next adventure. Nothing is wasted. But if you insist they follow one path you are ruining the fun of exploring a dungeon and taking away the agency of choice. If you offer several different paths, when they enter the room of certain death you can point out with a clear conscience that they didn’t have to open the black door with the skull on the front.

Corridors Are Rooms Too​

Don’t reserve encounters just for rooms. They can happen anywhere in the dungeon, in corridors, on stairwells; anywhere the player characters don’t expect one.

Add Some Mystery Not Just Monsters​

While you will need a few monsters to fight to gain some treasure, put in traps and just weird stuff too. Not everything need be deadly, just something weird to make the player characters think can be fun too, if only to cross a room (the Crystal Maze will be a big help here). With magic in the world you can put some very odd places in a dungeon. Just imagine something that would look strange and enticing when they open the door and then figure out what it does. It might be a room full of glass spheres, a garden with odd looking plants, a table set for a feast with only statues as guests. The weirder the room the more the player characters will be intrigued.

Make Sure There Are A Variety of Encounters​

This relates to the above; don’t rely on one sort of encounter. Make sure you have a mixture of traps, monsters, weird rooms and role play encounters. Try to avoid having the same type of room twice in a row if you can.

Don’t Skimp on the Role-Play​

Even dragons might chat; just because it is a dungeon doesn't mean there are opportunities to role play. Trapped creatures, intelligent monsters under a curse or a contract and even the odd guard might be talked to as easily as fought. You can let the player character make this decision, by who they choose to attack on sight. But remind them that they can talk their way out of situations as well.

Make Every Door Worth Opening​

If you do the job right, each door the player characters come across will fill them with a mixture of fear and anticipation. What lies beyond this door, a trap, a fearful death, untold riches or wild magic? If a room or encounter doesn’t’ feel that interesting to you, cut it from your dungeon. Maybe consider it a little and use it later on when you’ve made it work better. A dungeon need not be a sprawl, and a shorter one has the advantage of potentially allowing the player characters to escape and try another one some other day.

 

[iserith]

I wouldn’t consider them just advice. That sells them short. Procedures are important to reinforcing and supporting an intended style of play. If you disregard the procedures in basic D&D, the game gets much more dangerous. The dungeoncrawl stops working because the consequences for making a mistake go from failure (possibly including some PCs deaths) to total party kill. After all, if we can’t trust the procedures, then nothing stops the GM from ganking the PCs or preventing their escape.

Obviously, people wanted to do things other than crawl dungeons for treasure. That’s how we ended up with generous death rules and other changes to mitigate the problems that occur when you dispense with the procedures. If 5e had really wanted to be the modular edition, it would have supported pluggable procedures, so you can do a dungeoncrawl or epic fantasy or whatever, and the game’s mechanics would support and reinforce that.

Right. D&D 5e requires a bit of hacking to make it work well for dungeon crawling. The rules are sort of there, but not really, and what rules they do have are spread all over the place. I had to tighten them up a bit in the area of exploration tasks relative to time and wandering monster checks, but did not want them as tight as Basic (or OSE, which is a game I'm a player in now). I find the procedure in OSE, for example, to be kind of boring and repetitive, whereas I don't experience that in my D&D 5e game procedure (which some of my players who also DM use now). That said, I'm somewhat new to OSE, so I don't want to completely trash it. Perhaps with more experience I'll notice something that I don't. For now it's a bit too much in the "other direction" for my taste.

 

[overgeeked]

Right. D&D 5e requires a bit of hacking to make it work well for dungeon crawling. The rules are sort of there, but not really, and what rules they do have are spread all over the place. I had to tighten them up a bit in the area of exploration tasks relative to time and wandering monster checks, but did not want them as tight as Basic (or OSE, which is a game I'm a player in now). I find the procedure in OSE, for example, to be kind of boring and repetitive, whereas I don't experience that in my D&D 5e game procedure (which some of my players who also DM use now). That said, I'm somewhat new to OSE, so I don't want to completely trash it. Perhaps with more experience I'll notice something that I don't. For now it's a bit too much in the "other direction" for my taste.

Can I ask what procedures you do use for dungeoncrawling in 5E?

 

[transmission89]

Right. D&D 5e requires a bit of hacking to make it work well for dungeon crawling. The rules are sort of there, but not really, and what rules they do have are spread all over the place. I had to tighten them up a bit in the area of exploration tasks relative to time and wandering monster checks, but did not want them as tight as Basic (or OSE, which is a game I'm a player in now). I find the procedure in OSE, for example, to be kind of boring and repetitive, whereas I don't experience that in my D&D 5e game procedure (which some of my players who also DM use now). That said, I'm somewhat new to OSE, so I don't want to completely trash it. Perhaps with more experience I'll notice something that I don't. For now it's a bit too much in the "other direction" for my taste.

And that’s perfectly valid. B/X is more “game like” in its structure. That might not appeal to all and like most things B/X, is easy to hack and modify to taste.

 

[delericho]

Perhaps a hardback is the most efficient way to present material from an economic or quantity of info standpoint but...

Yep, I very much meant only from a quantity of info standpoint!

Is it efficient though? In my view, it’s bloated to fill page count. Is anything meaningful or integral added to curse of strahd vs the original i6 Ravenloft?

If it was a 64 page softcover without the extraneous cruft... compare I6 Ravenloft (not exactly terse itself!) with Curse of Strahd.

I certainly agree about the comparison between the original Ravenloft and CoS. Indeed, somewhere on the site there is probably a review I did of CoS that highlighted exactly that bloat as a weakness.

However, we should be wary of reading too much into the comparison. The original Ravenloft is a genuine classic, but it shares its format with such adventures as "The Forest Oracle" and "Quagmire!" which, being charitable, are not.

I would suggest therefore that simply changing the format probably wouldn't change the average quality of the offering. And given a choice between spending $50 on a 250-page hardback of mostly-mediocre material that I can spin out into a year of gaming, or spending $75* on a boxed set of mostly-mediocre material that I can spin out into nine months of gaming, I'd probably take the hardback.

* $75 is, of course, just a guess. Though, in the UK at least, it's probably going to be on the low side - boxed sets, being games, suffer from import duties and other taxes that books are exempt from.

 

[iserith]

Can I ask what procedures you do use for dungeoncrawling in 5E?

Basically I differentiate between "traveling" and "exploring." Traveling is moving around the dungeon and, unless other stated, you are alert to dangers. Marching order matters for noticing traps or hidden monsters at the front of the party. Passive Perception applies when there's uncertainty and a meaningful consequence for failure. Speed and distance travelled is mostly estimated, not as rigorously tracked as OSE. I tack on an extra 10 minutes and possibly a wandering monster check when it seems like they've travelled about that long/far.

When the players decide they want to have the characters stop to explore an area, then we go into "exploration mode." The players all state what tasks they want to accomplish in the given area (about 1000 square feet) - find traps, search for secret doors, pick a lock, loot, keep watch, etc. - and we resolve those over the course of 10 minutes of in-game time. I then roll for a wandering monster (or possibly it ticks toward the time when I will do so, if on hourly timer, and/or it gets closer to some other deadline/time constraint). Anyone who wasn't keeping watch is automatically surprised if the monster is a stealthy one. (Not all monsters are, but I make it a point to include some percentage of the wandering monster table that are.)

That exploration process plays out until the PCs move on. Notably, this process doesn't "feel" like a process in game because naturally I'm describing the environment and the players are describing what they want to do followed by my narration of the results at which point I loop back around to describing any changes in the environment. And so on.

Mostly this procedure is meant to make sure that there are meaningful choices to be had within the core resolution framework and that spotlight sharing is maintained while also creating tension with time and potential conflicts. I also find that if you do this in the context of wandering monsters never having treasure and/or monsters not being worth XP, it increases the incentive to take precautions or be more strategic about how they engage in these tasks.

 

[overgeeked]

Basically I differentiate between "traveling" and "exploring." Traveling is moving around the dungeon and, unless other stated, you are alert to dangers. Marching order matters for noticing traps or hidden monsters at the front of the party. Passive Perception applies when there's uncertainty and a meaningful consequence for failure. Speed and distance travelled is mostly estimated, not as rigorously tracked as OSE. I tack on an extra 10 minutes and possibly a wandering monster check when it seems like they've travelled about that long/far.

When the players decide they want to have the characters stop to explore an area, then we go into "exploration mode." The players all state what tasks they want to accomplish in the given area (about 1000 square feet) - find traps, search for secret doors, pick a lock, loot, keep watch, etc. - and we resolve those over the course of 10 minutes of in-game time. I then roll for a wandering monster (or possibly it ticks toward the time when I will do so, if on hourly timer, and/or it gets closer to some other deadline/time constraint). Anyone who wasn't keeping watch is automatically surprised if the monster is a stealthy one. (Not all monsters are, but I make it a point to include some percentage of the wandering monster table that are.)

That exploration process plays out until the PCs move on. Notably, this process doesn't "feel" like a process in game because naturally I'm describing the environment and the players are describing what they want to do followed by my narration of the results at which point I loop back around to describing any changes in the environment. And so on.

Mostly this procedure is meant to make sure that there are meaningful choices to be had within the core resolution framework and that spotlight sharing is maintained while also creating tension with time and potential conflicts. I also find that if you do this in the context of wandering monsters never having treasure and/or monsters not being worth XP, it increases the incentive to take precautions or be more strategic about how they engage in these tasks.

That sounds cool, but it doesn't sound drastically different than earlier procedures. Unless you're reading out the steps of the old-school procedures when you run them, you could (and as far as my experience, most people did) run the procedures much as you describe. Now we have terms like "fiction first" or "immersion" to give the game the feel of a living, breathing world, but a lot of people were doing that long before the terms appeared.

I bolded the bit above to point out that the same can be (and often was) done with the old-school procedures.

 

[iserith]

That sounds cool, but it doesn't sound drastically different than earlier procedures. Unless you're reading out the steps of the old-school procedures when you run them, you could (and as far as my experience, most people did) run the procedures much as you describe. Now we have terms like "fiction first" or "immersion" to give the game the feel of a living, breathing world, but a lot of people were doing that long before the terms appeared.

I bolded the bit above to point out that the same can be (and often was) done with the old-school procedures.

Yes, they are not dissimilar, but what stands out to me with my experience with OSE is that the resolution framework is much more consistent in D&D 5e and that makes a difference in my view. With OSE, various tasks are resolved in different ways and it's just kind of a mess comparatively.

 

[S'mon]

The work has still been done. It's just been done by someone who isn't you.

If you employ a cleaner does it make cleaning less laborious?

Most of the material in my current game I make myself. But it's fractal - until the players take an interest there is just a base seed, from which I can grow as much detail as desired.

 

[overgeeked]

Yes, they are not dissimilar, but what stands out to me with my experience with OSE is that the resolution framework is much more consistent in D&D 5e and that makes a difference in my view. With OSE, various tasks are resolved in different ways and it's just kind of a mess comparatively.

Sure. But we're talking about the procedures of dungeoncrawling, not how something like picking a lock or searching for traps is resolved. I get that they feel differently when it's a 1-in-6 vs a roll under % vs a skill check, not saying they feel the same at all, but that isn't the procedure for moving through a dungeon.

Things like having 60 10-second rounds in a 10-minute turn. And the Order of Events in One Game Turn, the eight step process for tracking time and movement in a dungeoncrawl from B/X. Detailing what can be done in a round vs what takes a turn. You commented that this approach was too rigid for you so I asked what you did instead. It sounds like you do basically the same thing only you're a bit looser with time and you focus on description and keeping the game focused on the characters and what's happening around them instead of dryly running through a list of procedures.

Keeping things immersive vs dryly running through a procedure is a DM style thing, not a problem with the procedures themselves.

 

[kenada]

Basically I differentiate between "traveling" and "exploring." Traveling is moving around the dungeon and, unless other stated, you are alert to dangers. Marching order matters for noticing traps or hidden monsters at the front of the party. Passive Perception applies when there's uncertainty and a meaningful consequence for failure. Speed and distance travelled is mostly estimated, not as rigorously tracked as OSE. I tack on an extra 10 minutes and possibly a wandering monster check when it seems like they've travelled about that long/far.

When the players decide they want to have the characters stop to explore an area, then we go into "exploration mode." The players all state what tasks they want to accomplish in the given area (about 1000 square feet) - find traps, search for secret doors, pick a lock, loot, keep watch, etc. - and we resolve those over the course of 10 minutes of in-game time. I then roll for a wandering monster (or possibly it ticks toward the time when I will do so, if on hourly timer, and/or it gets closer to some other deadline/time constraint). Anyone who wasn't keeping watch is automatically surprised if the monster is a stealthy one. (Not all monsters are, but I make it a point to include some percentage of the wandering monster table that are.)

That exploration process plays out until the PCs move on. Notably, this process doesn't "feel" like a process in game because naturally I'm describing the environment and the players are describing what they want to do followed by my narration of the results at which point I loop back around to describing any changes in the environment. And so on.

Mostly this procedure is meant to make sure that there are meaningful choices to be had within the core resolution framework and that spotlight sharing is maintained while also creating tension with time and potential conflicts. I also find that if you do this in the context of wandering monsters never having treasure and/or monsters not being worth XP, it increases the incentive to take precautions or be more strategic about how they engage in these tasks.

What you’re doing in 5e also works in OSE, but the GM has to be the one doing it. I used a variant of OSE’s dungeon exploration procedure when I ran PF2. When we first started playing online, I tried to be very rigid about handling turns. It was terrible. What ended up working for us was using the natural back and forth to delineate turns. If the PCs needed to travel a longer distance than a turn would allow, I just montaged them together (like I did wilderness travel).

We did have some growing pains switching to OSE (mostly VTT-induced), but the approach I used in PF2 worked there every including the need to rest. When it’s time for that, I work work it that into the narration and ask them if they wanted to take a break (usually yes, but sometimes they want to press on towards a goal). I also track all resources (except ammunition) myself using a simple exploration tracker sheet I put together. It’s really tedious for players to have to check that stuff off every turn, and it’s easy enough for me to tell them what they consumed during downtime.