D&D 5E On Dungeon Design

[SilverfireSage]

Recently, I've been having a problem. I'm currently a grad student working on getting into medical school, and I love D&D. I have been playing for a little over a decade, starting with my dad's old AD&D books and moving onto 3.5 and 4e, but I've never really felt exceptionally passionate about it until 5e. This edition has gotten rid of everything that I didn't like in previous ones, while simultaneously adding in things I didn't even know I wanted! I must digress however, and get to the real problem at hand.

I suck at designing dungeons.

Or rather, I do not have the time I used to have to design dungeons. But even more than that, even when I use dungeons that other people have created, I've found them to be.... less than exciting. Take Wave Echo Cave in the Starter Set for example. Having made my own content for years and wanting to try out 5E with some professional help, I thought that using a dungeon designed by the people who wrote D&D itself would give me a good idea of what made dungeon design great.

Turns out, I was wrong.

It was a pain in the a$$ to run, with my players having a terrible time tracking where they were on the map, trying to make their own and failing, and the whole adventure turned into "ok, you come up to a fork in the path. Right or left?" "Right." "Ok, you walk through the halls for several minutes and come to a fork in the path. Right or...." It's infuriating, but I thought, hey, maybe this one is just a bad egg?

So I looked online and ran a couple dungeons on there. Same issue. In fact, the thing I realized the most is that the more realistic one gets with dungeons, the more banal the adventure gets. Most castles are just room after room of storage/servant's quarters/bedrooms with little to differentiate between them except their occupant. Mapping them out is a chore for the players and the DM, and finding interesting things to do in between those rooms is the most interesting part. Getting the DM's Guide and looking at the dungeon creation tables was disappointing as well. Instead of offering some cool ideas on encounters that you can build, it's really just a way to construct a labyrinth in a short amount of time with no rhyme or reason! Who builds a stronghold with three simultaneous 20ft passages to the left?? After all of this, I've decided on two things for when I run dungeons:

1. Not map them out. Like, at all. Well ok, map out each interesting room individually, but not the connecting parts. So instead of having to spend hours meticulously designing an large castle, have a dozen rooms that are actually interesting, map them out accordingly, and then handwave the rest. I've already tried this and it seems to be working out fairly well, but it's still not satisfactory. I want to imagine dungeon making like in the Legend of Zelda games, where each dungeon is an intricately designed puzzle of interconnecting parts. Of course, it doesn't take too long to realize that I lack the imaginative capacity of professional game designers who have been designing such levels since before I was born!

2. Map them out, but make them really really small. One of my best dungeons was a tower that consisted of four floors, and a grand total square footage of about 3000sqft. Pretty small on the dungeon side, but in that square footage I was able to design some really cool encounters with interesting monsters, and some of the features of the rooms were actually cool enough to warrant a retread as the players tried to figure out each part of the puzzle. We all enjoyed it, but obviously a 3000 sqft castle isn't really going to cut it in the real world.

So, my fellow ENWorlders, has anyone else had this problem with dungeons in a game called "Dungeons and Dragons"? Is there anything I could do better? Which style do you think you use the most often? Or do you find that Tomb of Horrors style dungeon crawls are really the way to go?

 

[Boarstorm]

I tend to use a variant of option 1. I'll describe twisting corridors and such, but in much the same way I describe overland travel. I don't "zoom in" on the action until they get somewhere interesting.

It has its drawbacks, of course. The players know that if I zoom in on an unremarkable room, they'd better get to searching. But I can live with that.

 

[Astrosicebear]

We use battlemats at my table. Between us, we have 3. If its a large dungeon, I draw it out before hand and cover it, revealing it as the players explore. This works well for large, but not tightly packed dungeons. For those, I tend to rough draw the map, or use a small scale map, and use a side portion of the battlemat for specific rooms.

When we played through Worlds Largest Dungeon we drew the maps on battlemats and then covered them with a sheet of clear lexan, or plexiglass, and colored that black with dry erase. Then erased as we went.


As you said, the more natural or descriptive the dungeon gets the harder it is to map. What ive found is to only draw big set piece encounters. As long as the map denotes difficult terrain and important features, you dont need to be an artist. Get and use the old TSR map legend symbols. Use them for things like statues, secret doors, etc.

 

[SilverfireSage]

We use battlemats at my table. Between us, we have 3. If its a large dungeon, I draw it out before hand and cover it, revealing it as the players explore. This works well for large, but not tightly packed dungeons. For those, I tend to rough draw the map, or use a small scale map, and use a side portion of the battlemat for specific rooms.

When we played through Worlds Largest Dungeon we drew the maps on battlemats and then covered them with a sheet of clear lexan, or plexiglass, and colored that black with dry erase. Then erased as we went.

How did the actual session go?

 

[wedgeski]

A dungeon "crawl", to me, does tend to suggest a level of practicality, repetition, and "realism"...a place that the players can creep their way around, mapping as they go. I've had (and still have) players that love to map a dungeon. With those guys at the table, you'd better make sure you're providing the right level of detail, and honestly it's fun to pull out the graph paper once in a while.

My favourite dungeons are Underdark-inspired cave-dives into fantastic, absurd, dangerous cavern complexes populated by all manner of horrors. With that theme in mind, I design simple network diagrams showing the major caves and architecture of the complex, describing what connects where and with only minimal detail. I might label arrows "Up to" or "chimney climb" or "wall traverse" or "swim to" and so-on.

I'll put serious design time into six or seven areas (depending on the dungeon), then to fill in the gaps, I have pages and pages of semi-random hazards and and larger encounters in a ring binder, and a bunch of random tables with which I populate the dungeon as the party goes. I can design one of these hazards in any old spare five minutes (and for any locale, not just Underdark), and I love being surprised by what comes up; my table seems to enjoy it as well!

 

[delericho]

Or rather, I do not have the time I used to have to design dungeons. But even more than that, even when I use dungeons that other people have created, I've found them to be.... less than exciting.

That's not a huge surprise. Most pre-generated adventures just suck. Sad, but true.

It was a pain in the a$$ to run, with my players having a terrible time tracking where they were on the map, trying to make their own and failing, and the whole adventure turned into "ok, you come up to a fork in the path. Right or left?" "Right." "Ok, you walk through the halls for several minutes and come to a fork in the path. Right or...."

The problem here is that while the PCs are being given a choice, they don't have any context with which to make that choice. And, as a consequence, it's not an interesting choice.

(There's also a problem in that a very great many pre-gen adventures boil down to "go into the dungeon and kill everything you find." Because in that case the most efficient approach is simply to systematically move through the rooms until everything is dead. If instead you give the PCs a specific goal, preferably with some sort of time limit, then their behaviour changes - that sort of systematic approach becomes less than optimal.)

In fact, the thing I realized the most is that the more realistic one gets with dungeons, the more banal the adventure gets.

So don't build 'realistic' dungeons.

Most castles are just room after room of storage/servant's quarters/bedrooms with little to differentiate between them except their occupant.

With places like castles, or cities, or Star Destroyers, it's probably best not to map them in detail but instead to divide the place into key zones and have the connections between zones be fairly abstract. So instead of having 20 rooms that are all "servants quarters" you instead have one large zone that contains all those rooms, and gets dealt with in one go.

After all of this, I've decided on two things for when I run dungeons:

1. Not map them out.

2. Map them out, but make them really really small.

It really depends what the purpose of the 'dungeon' is. If it's a labyrinth to be explored then it is certainly appropriate to map it out. And if the players have difficulty with their map, that's a feature of their characters being in a confusing mess of a dungeon!

But for fairly 'known' places, you're probably better taking your "don't map" approach. As I said, divide the castle into several zones, and assume the PCs can probably move from one zone to another fairly easily.

Is there anything I could do better?

In a 'classic' dungeon, I recommend the following:

- Most rooms should have 3 or 4 exits. This means that at each point the PCs have a choice: two routes 'forward' or they can go back the way they came.

- When presenting the PCs a choice, you should always try to provide some context when making that choice. "Left or right?" isn't an interesting choice - might as well toss a coin (or adopt an SOP). But if left is a quicker but more dangerous route, with right being safer but longer then that gives them a chance to think about why they choose as they do.

- Try to give every room/area a distinctive landmark. It's much easier to map the place, and keep things straight, if you can refer to "the room with the lion's head" or "the room with all the corpses" or whatever.

- Give the PCs a goal other than "kill everything". Even better, give them three mutually-exclusive goals. That way they will have to decide for themselves which approach to take - do they wipe out the lizardmen in order to claim the magic treasures they have, or do they do as their lord has requested an ally with the lizardmen in order to guard against some other threat?

- Don't worry about realism too much. Mostly because it tends to be less interesting than the alternative, but also because even if you do put in all the work it's likely your dungeon won't actually be all that realistic anyway

 

[Astrosicebear]

How did the actual session go?

With World's Largest Dungeon? Great book, sessions went well, took us about 6 months. Map system worked well for us. Wasnt invasive at all. When we got to a room with an encounter, DM would move the over-map, and draw the room.

 

[txshusker]

It was a pain in the a$$ to run, with my players having a terrible time tracking where they were on the map, trying to make their own and failing, and the whole adventure turned into "ok, you come up to a fork in the path. Right or left?" "Right." "Ok, you walk through the halls for several minutes and come to a fork in the path. Right or...." It's infuriating, but I thought, hey, maybe this one is just a bad egg?

If you, a grad student, wandered into a dungeon, would you need to map it the entire way to find your way out? Some are more complicated - labyrinthine - but most aren't. With possible Ranger and Dwarf abilities in the group, it's almost impossible for a party to get "lost" in a dungeon without the DMing wanting it so. You could come up with a way that player's mark each corridor they've chosen with chalk and move on. It sounds like there's too much detail on that aspect of the gaming. (If you've random monsters that may erase them, then you can work that possibility into the play). I've played off an on for 30yrs and never actually mapped a dungeon while I played a character.

Most castles are just room after room of storage/servant's quarters/bedrooms with little to differentiate between them except their occupant.

yup. So there must be some purpose for the characters to be there to make all the banality worthwhile. Otherwise, you're just kind of being voyeuristic.

1. Not map them out. Like, at all. Well ok, map out each interesting room individually,.....

I think a DM needs to be as prepared as possible for any choice the players may make. The more info you have at your disposal, the better off you are, IMO. You may not use it all, but when you need it, the preparation makes for a smoother ride. Of course, if you're limited with your time because you actually attend classes, do your work and such, then the extra prep may not be possible. But I would never let class get in the way of good dungeon making.

2. Map them out, but make them really, really small. One of my best dungeons was a tower that consisted of four floors, and a grand total square footage of about 3000sqft....

Large or small, it's really all about the purpose behind the mission, I think. Honestly, medieval castles were very small. Mostly just small keeps with outlying buildings for support. Large castles/palaces are more Elizabethan to Renaissance+.

So, my fellow ENWorlders, has anyone else had this problem with dungeons in a game called "Dungeons and Dragons"? Is there anything I could do better? Which style do you think you use the most often? Or do you find that Tomb of Horrors style dungeon crawls are really the way to go?

I rarely draw my own maps anymore unless they're going to small... I just did a small tower similar to the one you described.. but it took no time at all. I often use things already drawn that I find online or re-purpose maps from other modules that I won't use or aren't using in the current campaign. It is time consuming, coming up with a reasonably logical layout as if you're an architect. I have a 4e Dungeon Delve book that has a bunch of small maps to reuse, ignoring the adventures they're built with.

Pure dungeons are a bit passé, though, I think you're right. An extensive "man-made" dungeon doesn't lend itself to much monster diversity. IMO, the best dungeons are usually the trapped dungeons... from 1e/2e days Tomb of Horrors, as you say, Castle of Inverness, White Plume Mountain are good examples of that. Caverns and outdoor settings can have a wider variety of monsters than pure dungeons, so that's the tendency for me recently. I think 4e went that route, as well. When the players need to delve into "dungeons", they're usually small or pre-drawn.

I plan on running Thunderspire Mountain, a 4e module converted to 5e; looks like a good combination of both. The module itself has a theme, is a dungeon, but it can be expanded into a number of other smaller dungeons while still retaining the mass of a larger piece.

I remember when I bought the 1e boxed set that came with B1. B1 was a drawn dungeon and had a small amount of overview, semi-theme and a few rooms already occupied with monsters. The DM was then supposed to insert monsters the rest of the way. Wow... the party met a group of orcs in one room, a chimera in another 40x40 room, and by the end was facing Demigorgon.... good times, good times.

It wasn't the dungeon layout that was the problem, though, it was my understanding (I was 12) and lack thereof of how one should be set up.

Anyway, to answer your basica question again, I repurpose maps. Then I find another map for my next idea and repurpose it. Then I run a module that fits for the campaign... and repurpose some of it.

 

[aramis erak]

Ways to make it better?

1: make geomorphs.
2: repurpose maps
3: describe more
4: remember that even servants are a threat in 5e - not much of one, but a threat none the less.
5: try some on the surface adventures that don't involve combat

Remember: people defend their homes to insane levels all the time in the real world. Someone breaking into my home will be detained at sword point until the police arrive. I've done so in the past, I'll do so in the future.

Geomorphs can be quite useful. They don't even have to be the full square kind - space hulk tiles make for quick and easy (albeit now quite expensive) dungeons.

When I make geomorphs, I usually start with a dungeon in mind I then draw the rooms at scale as separate pieces. I can then simply drop down what's there. If I plan to reuse them, I laminate before cutting. I've one dungeon I've run 3 times that way, and it makes the issue of player mapping moot, while also allowing for tokens-on-the-map play.


Repurposing is always a good use. Change the entry point and monster content and new dungeon, no drawing.

 

[Rune]

In my experience, if players don't like to map, they aren't into sprawling dungeon crawls, anyway. A few distinct rooms--all with purpose--are sufficient

If you're using a random generator, ignore the corridor charts. Just connect the rooms as you like/looks fun.

Finally, as several others have said, make decision points meaningful. This can be done, even if the characters don't know which rout would lead to certain outcomes. Instead, you can inform their senses. Give clues in the way the air tastes, or sounds of travel, or a rank odor, or slime on the wall.