Megadungeon Monday: Prelude to a Map

Banner Patreon 800 x 100

Category Megadungeon 800 x 400

Happy Megadungeon Monday! Or whatever f$&%ing day it is. I KNEW I would regret that Megadungeon Monday thing. Look, I know the Megadungeon has been gone a long time. But it’s back now. Let’s not dwell on that. Okay? Just be happy you have a f$&%ing Megadungeon article again.

Now, let me tell you some of the things that I love. One thing I love is spreadsheets and math and analyzing patterns. You’ve seen that. You’ve seen a LOT of that crap. But I also love drawing maps and designing spaces. And, so far, you haven’t seen much of that at all. We got close, but then, I was all like “but before we map, let’s deal with this whole big checklist of things.”

The thing is, I got into this project because I wanted to draw a big f$&%ing map and have it, I don’t know, pinned to my walls or something. And I’m starting to jones for map drawing. Map drawing is particularly fun because it’s something I can do AWAY from writing all the articles. Then, I can just take the maps and post them and describe them. Less words, more drawing fun whenever I want.

Now, you might recall that we took a long tour over two entire articles of the basic layout and critical path of our dungeon. So, there’s no reason we couldn’t just start drawing the damned thing. We have the basic outline, right? I mean, yes, there are questions to be answered yet about how certain areas will work and how gates will function. Before the floodgates are opened, for example, what is it exactly that blocks the passage from day 1 to day 8? How does one fly or teleport between zones? And so on. But, the thing is, we can work a lot of that s$&% out WHILE we design. Mapping the dungeon will get us thinking about the individual locations in the dungeon and how the different spaces are really connected. And anything we really can’t figure out, we can leave blank. After all, we know where it has to go and we know how much space it has to fill, so we can leave a void in the map and say “fill in later with an area that requires water flowing through it to be passable.”

In other words, we can just start mapping if we really want to. And I really, REALLY want to.

Well, that’s a lie.

We CAN’T yet. But we almost can. Today’s article is all about getting ready to just map some s$&%. And like every other thing that we’ve done so far with the Megadungeon, that means we have to figure out HOW to map the thing. Or, more accurately, we need to establish some rules.

More than a Map

Here’s the deal: the map of our Megadungeon – and, in fact, the map of EVERY adventure – has to be more than a map. We tend to think of maps as just diagrams of spaces to explore. But, by this point, you’ve probably realized that the map of an adventure is inextricably tied to its plot progression. Our map, for example, is visual representation of the flow through the whole story. See, this is where I reveal a very kooky truth about designing and running RPGs. And it’s one of those things that I say that SOUNDS backward and crazy, but actually reveals something really subtle and important when you think it through. Here it is. Get ready. Ready?

No matter how you design them, all adventures are linear play experiences.

Now, I know some of you are already calling bulls$&% on me. Others of you are smart enough to realize I’m saying something important, but you’re scratching your head trying to figure out what that thing is. And a few of you got it. And a few of you are already nodding because you know it already.

The thing is, when any given group of players sits down to play an adventure, they are going to play out the game as a sequence of choices, encounters, obstacles, and events. They go to this room, they kill some kobolds, they turn right, they go to that room, they deal with a trap, they go to that other room, they find some treasure, and so on. The truth of the matter is that all adventures play in a linear fashion. The only thing that makes some adventures more open than others is how many paths a given party rejects.

From the inside, our Megadungeon is going to feel big and expansive, stretching out in a lot of different directions. And that’s exactly what we want. But we’ve designed it to guide the players along a particular path through the adventure. We can do that because we know that any given play experience is linear anyway. We just need to make it LOOK like the party is choosing their path.

Obviously, we’ve taken a very hard-line approach to some of the guidance. We’re just outright locking certain portions of the dungeon behind barriers and requiring prerequisites to access them. But we’ve also mentioned some softer guidance. For example, we’ve talked about the idea of a critical path and about using that path to control the sort of challenges the party faces. We’d like to empower the party to handle difficult challenges by teaching them to handle the different pieces of those challenges first. In addition, we’ve talked about places where we need the party to choose to visit one location over another or remember a location and return to it with the piece of puzzle.

So, from that respect, the map is a pacing tool. And that pacing goes deeper than just gating off parts of the dungeon. Within individual plot arcs and sections of the dungeon, we need to have some sense of how the players will explore it. And, more to the point, we need to control that exploration to some extent. And, where we aren’t in control, we need to make sure the players won’t deviate too far from a good narrative pace.

But the map isn’t JUST a pacing tool for the adventure. The map is actually representative of a physical space. And that physical space has a purpose and a history. The map helps tell the story of the game and also helps reveal the backstory. And the map serves as the setting for the story.

Beyond THAT, the map also provides the mechanics for individual encounters. Every encounter – combat, trap, or obstacle – plays out in a space on the map. In fact, zoomed in to the finest level of detail, the map of the dungeon is the map on which all the individual encounters play out. So, they have to provide interesting and fun gameplay opportunities.

Ultimately, the point is that we could just map a good, well-paced adventure. Or we could just map a setting that exists in the game world. Or we could map a series of fun-to-play encounters. But we need to map for ALL of those things. Our map has to serve a lot of purposes. And we’re not going to pull that off without setting a few ground rules for ourselves. So, let’s dig into the very specific needs of this specific map and figure out what we need to do and how we need to do it.

A Space for Encounters

Let’s start with the last idea first, because it’s actually the easiest to deal with. We just have to be aware of how we’ve made things very difficult for ourselves. Because, yes, we did. We’ve imposed a very difficult requirement on our encounter spaces without even realizing it.

We’ve decided that any given “empty space” in our dungeon will have a chance of holding a random encounter if the party ever returns to it. See, normally, when you design a dungeon, you can differentiate between encounter spaces and non-encounter spaces. The encounter spaces are the places where you plan for combats. And because you know what’s in there, you can design the space to accommodate the type of fight you want to have in it. And we’ve talked before about how important it is to consciously design the battlefield for the fight that’s going to happen in it. The non-encounter spaces can be, well, simpler. They don’t need to be designed as battlefields because battles probably won’t happen therein.

But, in our dungeon every space could potentially be a battlefield. And if we design a space that can’t accommodate a battle or is just very boring to have a battle in, it’s going to suck when some group has to have a fight in that space. So we can’t under design our non-encounter spaces. We have to design every space to provide a dynamic, exciting, or interesting battlefield.

And that’s not all. The other problem is that we don’t know what types of battles will even take place in many of our encounter spaces. Sure, in the rooms with planned encounters, we know about the FIRST encounter to happen there. But subsequent encounters could be anything. Literally anything. After all, with our roster system, different types of monsters will be popping up throughout the dungeon throughout the entire course of the adventure.

So, we have to design every single space in our dungeon to provide an exciting location for a fun battle without having any sense of the battle that will take place in there. Let’s call that rule 1.

Rule #1: Every space has to provide a good place for an exciting and dynamic combat without knowing what battle will take place inside.

As a general rule, that works. And it might seem hard, but we can do it. As we design, we’ll look at the different ways to pull that off. But, there’s a corollary to that rule. Two, actually.

See, here’s the thing: a combat that is designed in tandem with the space that it takes place in is always better. When you create the battle and battlefield to work together, you’ll always have a more fun experience. Right?

Now, we have some combats that are going to be MAJOR battles. Set-piece battles. Boss battles. Big, interesting fights. The fight with the dragon, with the plant monster, with the demon queen, and so on. So, let’s establish a corollary to rule 1.

Rule #1a: Any space that has a boss fight or set piece encounter will be designed alongside the encounter that goes inside of it.

Now, sure, there’s a chance the idiot players will wander back into the dragon’s lair and end up fighting some giant rats in a giant, cavernous space filled with interconnected pools for an aquatic dragon to use to basically teleport around the room, and that rat fight will probably be pretty dull. But that’s okay. We’re willing to make that sacrifice to ensure the dragon fight is really, REALLY cool.

The second thing has to do with pacing. See, part of the reason to use a critical path style of design is so that we can pace the encounters along the path. For example, before the party has to fight a complicated battle with three different types of kobolds and their lizard pets, we want to make sure they’ve had a chance to fight the individual creatures in simpler environments so they can develop tactics. That’s the whole Megaman approach we talked about previously.

Now, the problem is we’re going to be jumping forth between designing the map, designing the creatures, and designing the encounters. And, for reasons that we will discuss in a minute, we don’t want to leave too many blank spaces on the map. Thus, we want to be able to map any room without knowing the monsters that will go in it. We can leave the boss fight rooms blank, fine. But we can’t leave every encounter blank until later.

So, we’re going to add another corollary.

Rule #1b: Any space with a fixed encounter will be redesigned or tweaked to work well with the encounter that gets placed in there eventually.

That is to say, on the encounter level, our map will be a rough draft. When we start sticking the monsters in, we can erase, move, or add features to work with the encounters we’ve specifically designed for that room if we need to.

A Setting with a Story

The second purpose of the dungeon map is to provide a setting and tell a story. The funny thing is that we’ve done a lot of work to pull this off already. We just need to keep it in mind.

See, for a large map like this, there’s basically two ways we can break down the work. We can design the map for each adventure day and build basically in the order of the adventure. Or we can design the map for each geographical zone. Both have advantages and disadvantages. Mapping day by day focuses us on the critical path and the pace of the story. But mapping zone by zone ensures the zones have a more consistent feel. And those geographical zones really do tell a lot of the story. And that’s what I meant by saying that we’ve already done a lot of the hard work telling the story. Just the fact that we’ve divided our dungeon into zones and those zones are inextricably tied to the history of the place ensures the map will tell a good story. We just have to make sure we don’t break it.

Now, we COULD simply create a rule that says we have to map zone by zone and be done with it. But the tradeoff isn’t actually that simple. Before we create any other rules willy nilly, let’s talk about the third goal for our map and see if that helps us arrive at a solution.

Control the Flow of the Adventure

The third thing we want our map to do is control the flow of the adventure. But what does that mean? Well, it means a few things. Some of them we’ve already done, but a few we’ve never even thought about before now.

We’ve talked enough about pace and flow already for you to realize that all the work we did designing the critical path and optional spaces and all of that other crap – that’s the pace and flow of the adventure. In that respect, as long as our map conforms to the outline we’ve already drawn, most of the pacing is already handled. Except, we’ve got a few problems. And they are called players.

Remember above how we said that, although the adventure will play out as a linear sequence of scenes, we don’t want to constrain the players to such a degree that they realize there is only one way through the adventure? Well, a lot of that comes down to the fact that, as the party explores the dungeon, they gradually have more and more freedom about where and how to explore next?

Well, there are times in our dungeon when we need the players to recognize where they should go next (or choose between a few options) without just forcing them on that path. We need them to remember locked doors and go back when they have the key. We need them to realize that new paths are open to them and to decide which of those new paths to follow. And that’s tricky because players, bless their f$&%ing souls, have free will.

But that’s not all. We’ve also designed a random encounter system that specifically exists to drive the players to be efficient in their explorations. Explore, sure, there are rewards for exploring. But explore with purpose. Don’t flail. Don’t be aimless. If they wander too aimlessly or too carelessly, they will expend resources and get no rewards back.

The thing is, you can’t ask people to be strategic and make smart choices unless you also give them a way to actually figure out which choices are smart and to implement strategies. That is to say, if you want to punish bad behavior and encourage good behavior, you actually have to empower people to choose between good and bad behavior.

Look at it this way. Imagine that you have two doors in your dungeon. And you want the players to go right. So, you decide that if they go left, they will suffer a crapton of damage. If they go right, they get a treasure. Now, the players come to the doors. They are identical doors set side-by-side at the end of the hall. Without any clue as to which door will hurt them and which door leads to treasure, they basically just have to choose at random. In that case, they haven’t made a good or bad choice, they’ve just acted randomly and gotten lucky or unlucky.

If we want to reward efficient exploration and punish careless, aimless wandering, we have to give the players the power to know the difference. Basically, that comes down to the idea that we have to make the dungeon easy to explore.

Now, some correspondents in my comment section have already observed this problem like a dozen f$&%ing times. They’ve pointed out that I’m putting a lot of faith in my players to understand and remember the layout the dungeon and to make smart decisions based on that understanding. And I’ve said repeatedly “don’t worry, we’re going to empower the players to do just that.” And then people have called me a dumba$&. Well, we’re here now. It’s time to talk about this. And when I’m done, some of you will owe me an apology.

The Mapping Question

The first thing I’ve been asked is whether or not the players will be required to map the dungeon as they explore. And the comments have generally been very down on mapping. People seem to think that mapping is a ridiculous thing to expect. Some have pointed out that it’s stupid to assume that adventurers COULD even map a dungeon. They use words like “unrealistic” and “metagaming” and “surveying equipment,” which just demonstrates what dumba$&es they are. Others have pointed out how much they, personally, hate mapping. And therefore, mapping is a terrible thing for players to ever have to do.

First of all, players are NEVER forced to map. But in this case, it is a very good idea. Yes, we SHOULD encourage the players strongly to map the dungeon. Just like, in a murder mystery adventure or a complex intrigue plot, the GM encourages the players to take notes. After all, a dungeon map is just site-based notes. There is literally nothing wrong with expecting and encouraging mapping.

The problem with mapping is that mapping always ends up being a giant, complicated pain in the a%$. When people think of mapping, they think of the GM painstakingly describing 15 foot by 35 foot rooms and trying to figure out ways to describe how there are three doors in the eastern wall and two in the west and a passage in the southwest corner. And that’s assuming all the rooms are squares are rectangles. If the rooms are irregular caves and weirdly shaped rooms with alcoves and nooks and bits that are half collapsed, it becomes a nightmare.

The problem is trying to be exact is very complicated. But in a complex dungeon, you sort of HAVE TO be exact because rooms are interconnected. Unless you are precise with your sizes and distances, you might not recognize that the four-way intersection you just entered from the north is the same four-way intersection you passed through from east-to-west before. Dungeons are filled with a lot of featureless space and it’s sometimes hard to recognize when your path loops back in on itself.

At that point, the easiest thing to do is for the GM to just map FOR the players the way lots of video games do. Most exploration-based video games have automaps. And there are several good reasons for that. Yes, maps make exploration more efficient and they empower players to escape to safe areas when they need a break and make good decisions. But they also make the site more real. A dungeon without a map is just a disjointed collection of rooms and encounters. But when they get added to a map, they give a sense that there is a real space that brings all those rooms and encounters together. Moreover, the map plays into the idea of the game as being ABOUT exploration. When you add a space to your map, when you fill in the blank page with rooms, you’re actually seeing the results of your exploration. It’s a progress indicator. As surely as an XP bar filling up (or XP points counting up on a character sheet) provide an important sense of progress, so does filling in a map.

Put simply, our adventure NEEDS to be mapped to function. That’s pretty much the only way. We need it to control the flow. We need it to empower efficient exploration. And we need it to make exploration feel like a real, tangible goal.

But how…

Designing a Dungeon to Map

Once upon a time, mapping was pretty much a default assumption. As a GM, you KNEW your players were going to map your dungeon. Hell, in the Mentzer Basic D&D Box Set, the Player’s Book actually TAUGHT you how to map in the same chapter where it taught you how to play the game. I s$&% you not. And people got GOOD at mapping. Of course, back then, all dungeons were drawn on graph paper. Even caves. They were just weirdly rectilinear. I mean, look at the old maps of the Caves of Chaos from module B2: Keep on the Borderlands. Those are natural f$&%ing caves. That’s some pretty strange erosion patterns for a karst landscape, no?


And so, players got GOOD at mapping. And as they got good, GMs and adventure designers USED that. First of all, they started using the layout to hint at things like where secret doors were hidden. Second of all, they started designing traps and tricks specifically designed to confound mappers. To add challenge to the mapping experience. For example, teleportation traps that would silently and invisibly send you from one intersection to another identical intersection. Or trapped intersections that would confuse you about what direction you were going in. And as time went on, some adventure writers purposely designed “Mapper’s Laments,” dungeons designed to be really difficult or impossible to map.

Now, I’m not saying “times were sure better when players could map their way through a teleportation trap while under the effects of a confusion spell.” That was just how things were, then. People were working within the medium to create challenges. And mapping just isn’t de rigueur nowadays. You can’t count on getting even one player who is an expert surveyor-draftsperson with a love of graph paper.

But there’s an important lesson in all of that. Dungeons can be designed for maps. And dungeons can be designed to be impossible to map.

Imagine, for example, a dungeon in which every room was square and the same size and every corridor between the rooms was exactly the same length. Players could very easily make a very accurate map of such a dungeon. It just wouldn’t be very exciting. And, the thing is, a lot of strategies you might use to simplify the mapping process really take hold over the design of your dungeon.

But maybe we can find a halfway point? Maybe part of the problem is the dungeon, but part of the problem is the map.

Designing a Map that Works

What is REALLY needed of a map? And when I say map, I mean player map. What form should a player’s map actually take. Should it be an exact copy of the GM map with exact distances and sizes? Or is that level of detail really necessary?

Once upon a time, David Noonan, my favorite game designer ever, was on an episode of the D&D Podcast. Sadly, you can’t get the episode anymore. I so wish I could. I would pay money for the old episodes of the Official D&D Podcast. Basically, the stuff from 2006 to 2009. In the episode about Dungeons, Noonan talked about going back and poring over his old player maps of Undermountain, which was a massive Megadungeon adventure. He’d drawn the map in character. So he’d made notes on it about what was in each room and he’d nicknamed the rooms and stuff. And he observed that the map had become a kind of visual recap of the adventure. “Here’s the room where we fought that ogre,” or “here’s that door with the strange star symbol we couldn’t open” or “here’s the room filled with illusionary webs that summons spiders.”

Now, THAT is what we want out of a map. We want the map to serve as a visual record of exploration, to remind the players where things are, and help them keep notes about what to go back to later. And NONE of that requires f$&%ing graph paper. Honestly, all it requires is a goddamned flowchart.

See, a player map only has to serve to show how rooms are connected and provide a space for writing down what happened where. It just has to be a set of good gaming notes. And that means, when we decide how to design our actual map of the dungeon, we only have to make it easy to map as a flow-chart, as a series of lines and boxes and circles.

How to Pull It Off

And that leaves us with two questions. How do we design a dungeon that can be mapped easily? And how can we get the players to map it right?

Let’s tackle the second question first. Because, in this day and age, a lot of players might not even think about mapping. Now, we COULD just tell the GM “okay, tell your dumba$& players they have to map the dungeon, or they will probably get lost forever,” but that’s a sucky approach. Besides, then we have to tell the GM to explain exactly HOW to map the dungeon in a quick, efficient way. OR, we can take another page from the book of all the video games we are stealing from. We can just put the players in a situation where they can figure it out for themselves.

Imagine if, very early in the dungeon, the players find an explorer’s corpse. And that explorer has a map on him. It’s unfinished. He didn’t get very far. But he maybe discovered a hidden thing that the map could point to, like a treasure map. But, more importantly, the map is mostly a blank, open piece of paper with just the start of a map. And what if it’s mapped the way we want the dungeon to mapped. Nodes and lines, notes, nicknames for rooms. Maybe it helps the players early on. But we also tell the GM to tell the players, flat out, that they can write on the map if they want. It’s theirs.

You can pretty much count on a player at the table taking possession of the map and taking a pencil and continuing it. And because the map already demonstrates how to map the dungeon usefully, the player will just follow that guideline. Moreover, maybe, as time goes on, other maps can be scattered around. Kobold maps. Ancient elvish diagrams. Whatever.

So, we’ve taught the players how to map and given them a head start. After that, if they don’t do it, getting lost or confused is their own damned fault. Right?

All that leaves us to do is make sure the dungeon CAN be fit into that map style. Basically, we have to make sure that exactness isn’t needed. And that connections between rooms are easy to understand. In other words, the map can be intricate, but it can’t be complicated. For example, we might decide that each level of the dungeon is flat. As long as you’re on the first floor of the dungeon, there won’t be any overlaps or passages that go over or under other passages.

Rule #2: Every level of the dungeon is flat with no rooms or passages above or below others.

And we might decide that there are never more than four ways out of a room and they always leave by the four cardinal directions.

Rule #3: Every room can have no more than four exits and each exit must leave in a certain cardinal direction.

And we might also decide that the tunnels and passages between rooms, while they can bend or turn, always follow cardinal directions. And, for that matter, we might even decide that intersections are always rooms. Passages between rooms don’t have junctions.

Rule #4: Passages between rooms always follow cardinal directions and never have junctions.

Now, you might worry about how boring that’s going to make the map. But I’m going to show you something pretty neat. First, look at this map of the Chozo Ruins section of Metroid Prime. That’s a pretty neat map, right? That looks fun to explore, doesn’t it? I can assure you, it is.


If you look really closely, you’ll find that it MOSTLY follows MOST of the rules I laid out. If you want proof, look at this strategy guide map of Metroid Prime.


The top brown area of the map is Chozo Ruins, the same section as in the 3D map above. Notice that it’s basically just a lines-and-nodes map? There are a few spaces where it breaks my rules, but it’s close enough that you can see what I mean.

Just those rules alone make it easy to map the dungeon without a lot of detail. But we’re not quite done yet. Because one of the other problems with mapping – as mentioned – is that the players don’t always realize when they’ve doubled back or come around to a place they’ve already been from a new direction. And a lot of the precision that players strive for is a result of trying to recognize exactly that. Much of the rest comes down to trying to figure out where secret doors might be hidden. But we’re not going to play THAT game.

The trick is to make each room in the dungeon recognizable. But that’s not enough. Because we also want the players to be able to keep easy notes about the different locations or to remember the locations. And one simple rule – which comes from combining two different ideas – can help us accomplish both of those. And it can also help us make sure our encounter spaces are always interesting places to have encounters. And it can also help us emphasize the larger-scale structure of the dungeon by highlighting the differences between the different geographic zones.

Once again: let’s talk about Metroid. Other games do it, but Metroid does it very well. Especially Super Metroid and Metroid Prime. As noted, those games are divided up into geographic regions. And, as noted, those games involve intricately interconnected maps. And, as noted, those games are all about exploration as accomplishment. And two accomplishments really emphasize how well you’re exploring. The first is when you pass from one geographic region into another. For example, when you find your way from Brinstar (the overgrown rocky jungle caves) into Norfair (the fiery lava caves and ruins). The second is when you come into an area you’ve already explored from a new direction, thus completing a loop. For example, when you escape from Norfair after finding the Hi-Jump Boots, Ice Beam, and Super Bombs and return to the surface region of Crateria, where your spaceship is parked. The question is: how does the player KNOW they’ve accomplished these things?

Well, the passage from region to region is shown off in the games visuals. Each region has a unique visual design and color palette. Brinstar is green and gray. Norfair is red and brown. And because the color palettes are so strongly emphasized and rarely deviated from, the instant the palette changes, you know something is up. Now, in most places, Super Metroid uses elevators to also mark the transition from region to region, but those are actually about allowing the Super Nintendo console to unload one set of sprites and color palettes and map data and load another in its place. The regions themselves are delineated by the graphical changes. The elevators just allow the computer to keep up with those changes.

Now, Metroid Prime’s regions are even more visually distinct. And, in addition, each room tends to have a recognizable feature that sets it apart from other rooms. Something about the shape of the room or the specific features and landmarks in it make the rooms distinct and recognizable. Super Metroid does it too, but not to as great a degree. And Super Metroid has a lot of “samey” spaces that can easily be confused for other spaces.

The problem is that those solutions are entirely in the visual design. And D&D has NO visual design. The players can’t actually see the spaces they are in. Now, you COULD rely on flavor text. If you’re dumb. Because flavor text doesn’t do jack s$&% by itself. The problem is that the level of detail you’d need in the flavor text to actually pull this off is far more than the average player’s attention span (three to five sentences, MAX!), so that’s a no go.

Well, that’s actually not QUITE true. You could use flavor text, sort of.

We talked before about the idea of building encounters out of a limited set of resources. The Megaman solution. You remember. I mean, you should. After all, we JUST mentioned it again a few paragraphs ago. Well, just like “less is more” when it comes to building encounters, “less is also more” when it comes to building battlefields. That is to say, most battlefields are best if they use only ONE specialized terrain element and the rest of the terrain is more standard stuff like walls or rough ground or whatever. For example, there might be a massive statue of an elf champion in the middle of the room. That’s a neat feature that provides cover, breaks lines of site, and stands out as unique.

Imagine if each of our rooms had something that made them unique that could be spelled out in one, single sentence. Like that statue. Or a pair of big waterfalls, side by side. Or three natural stone-arch bridges over a river. Or whatever. Imagine if those features lent themselves to a single, simple descriptive sentence. Or even a phrase. The players could use that feature to nickname the room and to remember the room and each encounter space could be built around that unique feature. Three bridges that join together create neat chokepoints for melee combatants without impeding ranged combat and give an advantage to flyers, right?

The trick is that, as we start to sketch out or map, we give each room a name. Metroid Prime actually does this. If you go into the map and highlight a room, you’ll see that each room has an evocative name. And the reason is that the name helps you remember specific rooms. Some spelled out the purpose of the room (like The Gathering Hall or The Furnace). Others spelled out a major feature (like The Sunchamber or Fiery Shores).

The thing is, we can use that trick to help us make our map more memorable. If we give each room an evocative or descriptive name, it will help us design encounters in that room and it will also help the GM call attention to the major features of the room. And, if we tell the GM how to do it, we can even employ a metagame trick to help the players remember the names of the rooms.

See, I often have multiple different versions of the same creature in combat. It’s a habit I picked up since 4E. So, I might have the goblin that’s fast and outflanks and backstabs. And I might the goblin that’s tough and can take a beating. And I might have the goblin that has a bow. And I give those creatures nicknames: skirmisher, knight, and sharpshooter, for e.g. And then I use those names to refer to the creatures. “The goblin skirmisher darts behind you and stabs you for three points of damage. Next up, the sharpshooter takes aim and fires, but he misses.” See? I just slip the names into the flavor text. And then, the players start using the names too. It helps them remember which creatures do what and gives us an easy way to distinguish one from another without relying on things like “the goblin with the bow” or “that fast guy with the knife who keeps stabbing me.”

And we can do the same thing in the Megadungeon. If we give the GM the name of the room, “The Twin Waterfalls,” for example, we can encourage the GM to use that name himself when discussing the room. “The room is dominated by Twin Waterfalls…” for example, in the flavor text. And if we ALSO label the example maps we give the player with these nicknames, we establish a pattern for the players to follow. They will start naming the rooms after the feature that the GM emphasizes. When the players return to these rooms, or when the GM is narrating a trip through several rooms to return to a previous area, the GM can say “you turn south and pass by the Twin Waterfalls, continuing across the Three Bridges, and finally reach the Statue of the Penitent Warrior.”

Without explicitly setting down rules for it, just through deliberate usage (assuming we explain it to the GM), the GM can establish a language for talking about the dungeon that also serves as a mnemonic device and, at the same time, focuses our design. In fact, we could start mapping simply by giving every encounter space an evocative name that we make up on the spot, and then actually doing the detail mapping later.

Rule #5: Every room must have an evocative name that speaks to either a major landmark or the obvious purpose of the room.

The final problem is how to link everything together and to make things feel like a contiguous whole but also make different geographical regions distinct. And we’re going to accomplish this in two ways. The first way is one we’ve already sort of touched on. We’re going to give each region a unique palette. Of course, we can’t use color palettes like a video game. But, what we can decide on is some themes that the whole region will have in common. For example, the entirety of the Desiccated Sanctuary might be dry, sandy, lifeless, and withered. If every room has sand or dead plants or dry air or dust or is made of sandstone or something, that serves as a contiguous palette. In addition, if we have certain obstacles or challenges unique to a given zone that we can add to encounters in that region, we can further tie the encounters together. For example, the Great Tree, choked with poisonous plants, might feature a particular poisonous thorny vine obstacle in several rooms. The Crystal Caverns might feature outcroppings of razor-sharp crystals that will seriously cut up anyone who blunders into them. And we can further come up with lists of different furnishings or dungeon features common in the region. The Crypts of the Ageless will have sarcophagi, offering bowls, coffers, candle stubs, and, of course, bodies.

Rule #6: Each region will have unique descriptors that emphasize the theme and unique hazards or obstacles that can be used only in that region. Each region will also include a list of furnishings and features that fit the theme of purpose of the region.

Finally, in order to really tie the zones together, we have to give some thought to the larger geography of the place. If we have a room that has waterfalls, a room that has bridges, and a room that has a lake, for example, it makes sense for rivers to connect those rooms. After all, the water has to get from one to the other. So, even though it’s EASIER to map each room as its own unique encounter space, to bring things together, we’ll start each geographical region or zone by planning the larger features that connect different rooms. For example, the Fiery Abyss is going to be fixated on a lava lake with a temple at the middle where the Demon Queen lives. So, we can actually start by drawing (roughly) where the temple is and where the lava lake is and where the lava tubes are and then drawing our rooms OVER those features. In other words, we will plan out the macroscopic geography of entire zones and then map individual rooms over those features.

Rule #7: Design the macroscopic geography of regions first, then map individual rooms over those features.

If we follow all of those rules, we should be able to map and design the whole space.

Design by the Numbers

Now, it probably seems a little silly to try to figure out a set of rules for mapping out the entire dungeon. After all, most people just map their damned dungeons. But then, that thoughtlessness is why most players hate trying to map dungeons. And even if it does make sense to establish the rules, it might be a little unclear as to how they will really work. So, let’s see if we can’t end by summarizing the rules and spelling out the logical process for mapping the dungeon.

First, we need to work on the regional level. We take a particular region of the dungeon and figure out the descriptive themes of that regions, which we then boil down into a few useful adjectives. We need to think about some specific hazards or obstacles we might encounter in that region. And we need to figure out some furnishings and features that might be found throughout the region.

Then, we need to figure out, based on the descriptive themes, where the large-scale physical structures of the region are and sketch them into place.

Once we’ve done that, we need to divide the region into encounter spaces using the critical path map we’ve already designed. Each encounter space must connect to no more than four other encounter spaces and each of the exits from each space must lie on a different cardinal direction than the others.

Next we need to identify each encounter space with an evocative name that suits the geographical features that encounter space is built upon and that utilizes the themes of the region in question. We must pay special attention to rooms that we know will contain major encounters like boss fights.

Once we’ve done that, we can leave the boss fights alone and focus on the other encounter spaces, designing each room within the space we’ve delineated for it, based on the name and the themes, with an eye toward creating an interesting battlefield and utilizing the furnishings and features for the region to add additional complexity to the battlefields.

Finally, after we design the encounters and place the other major elements of the adventure (like discoveries and treasure), we can return and tweak the design of any encounter spaces that need it.

And THAT will be our process for mapping the entire dungeon. If it sounds like a lot of work, that’s because it WILL BE a lot of work. But the systematic approach should make it a lot easier to come up with a design that works. And in the coming weeks, we’ll start mapping our very first region: The Desiccated Sanctuary.

Banner Patreon 800 x 100

45 thoughts on “Megadungeon Monday: Prelude to a Map

  1. I love the idea of handing out a sample map, and especially like the rider of handing out additional maps as progress rewards.
    I’m guessing you’d just make sure to add a ‘locked door’ icon onto the original map to ensure that they knew how to keep track of impassable spaces?
    My players are really bad at the metagame and need all the help they can get.

    • You might go a step further. The legend of the map has a locked door symbol. The notes next to a locked door include a very subtle hint that the party should find the item linked to that door. Something like “door too complex for scout; didn’t the elves have a key?!” You know, subtle enough that a player might understand.

    • And if I were a moderator, I’d delete it.

      Anyway, Thanks Angry, for another great article. While you’re designing a mega-dungeon, I’m running two adventures (group 2 started last weekend) where all your advice rolled up together is a great help.

      Even before reading this article, in retrospect (based upon my years of experience and having read prior articles here), I find myself following these steps and rules in my designs, where before I never thought about it.

      Group 1 is getting ready to invade a witch’s tower, built on a small island off the shore of a large, mountainous island. Aside from what I wanted in the tower, I did some searching online and found a cool picture of a tower on a small island. Using this as my outer framework, I built the tower, the structures at the base of the tower, and the tunnels inside the big rock the tower is sits upon. I estimated the size of the rock, how high it goes, both above and below the surface of the water, and all that. All from one painting.

      Inside are a mixture of encounters, some are variations of what the party has seen before, and other are completely new for them. In this case, especially, there are no “random” encounters. Everything is planned, drawn, and ready to go, when the party gets there. (We’ll they are close. They are on the bigger island presently, and the tower is just a short distance away to the west.)

    • But, you’re not first at all. How strange.

      It’s almost as if I can make the time stamps say whatever I want. It’s also almost as if I can remove your link because it sure as hell looks like a spam attempt to me. Because I have the power. I win.

  2. Some of this is, as you say, fairly self-evident. The idea of node-maps and forcing a mnemonic feature into every room is genius. As you say – it’ll simultaneously feed the atmosphere, enrich the encounters, and help everyone have a common parlance about the region. That one I’m stealing right now :D.

    Thanks again for providing fantastic guidance for crusty vets and wide-eyed newbs alike!

  3. Dude, I’m so happy that the Megadungeon is back on tracks. Also, great piece of advice on how to prepare a dungeon adventure. Keep up with the great work, man!

  4. Interesting you mention Menzer, because this is very much turning into a teaching dungeon, both for GMs and players.

    I wonder if you are going to introduce faction play later on, and use the dungeon to teach players and GMs how to handle it?

  5. Very cool stuff. No one is thinking about dungeon design in this way and to this granular a level.

    I have mixed feelings about having the players mapping, even with the tools you’ve come up with. I can see the benefit though. My hesitance is probably just PTSD from my player days going through Undermountain. That crazy Halaster, always a trickster.

  6. Very good article.

    I have been putting these ideas into place over the last few months as I build my ongoing campaign. As a result, my regions, areas, and battlefields are much more dynamic.

    I have already addressed the problem of the player overall map in a different way.
    – First, I am using roll20. I create large, sweeping areas and put an entire region in that one roll20 screen. As a result, the ‘Player Map’ and the ‘GM Map’ are identical in layout (though I have a few more things visible in the ‘DM Layer.’ Players are only able to see things that they have gone through using ‘fog of war’.
    – Second, I created a second roll20 game to have a miniaturized map that is purely in box-form layout that is also premade, but I can reveal as we go using ‘fog of war’. In effect, players (and the DM) have two separate windows open with two separate roll20 adventures loaded. One roll20 is just for players to see the overall map (and make notes on it), while the other is open to hold character sheets, avatars, and host the individual encounters.

    Roll20 is not the solution for everything, but as it is the thing that I am using to run my games, I find a lot of answers within the program.

  7. I join those who are so pleased you’re back to the Megadungeon! I admire this system and especially prompting players to map a certain way with an in-game event. Clever and simple.

  8. I actually found that giving my players a sample map lead to the expectation that they would find another sample map around the edges of the first one. Because I didn’t realize they weren’t mapping the dungeon, they were completely lost by the time I pointed it out.

      • In the past, I have been so blatant as to have an NPC stop the players on their way out of town.

        The NPC is a cartographer who is like, “Yo, I heard you might be going to x? You know, I will pay good gold for even a basic map of that place. I don’t have one.”

        In my old HS game he became a recurring NPC who the players used as a very minor patron, resource and even quest giver.

  9. Nice to see another article on this 🙂

    Given your rules, I’m curious how you’ll handle one of the first parts of the dungeon you’ve mentioned a few times now and that seems to break most of them; the part where the players fall through the floor into the room below. (But I’m sure you’ve got a good idea there)

  10. I’ve actually been big in to handouts since I started DMing. It would be interesting to see an article on more stuff like that, but i realize there’s not TOO much one can say bout handouts in general. Usually, most of mine are notes (often left behind by the villain, or someone important; clues and such). Every once in awhile I’ll hand out a “book” (which is just a couple lengthy paragraphs of info on a particular subject. I wasn’t a fan of doing something like that, but all my players are really interested in lore and the story-telling aspects of the game. I can also leave clues within the “book”s themselves.

    I’ve been wanting to do the same thing with a map but wasn’t sure how to get people on board with it. With this advice, I think I can manage. This will actually make certain mapping (for me the DM) easier as far as planning goes.

    Thanks as f@&%ing always! =D

  11. Rule #1: Every space has to provide a good place for an exciting and dynamic combat without knowing what battle will take place inside.

    Rule #1a: Any space that has a boss fight or set piece encounter will be designed alongside the encounter that goes inside of it.

    Rule #1b: Any space with a fixed encounter will be redesigned or tweaked to work well with the encounter that gets placed in there eventually.

    Rule #2: Every level of the dungeon is flat with no rooms or passages above or below others.

    Rule #3: Every room can have no more than four exits and each exit must leave in a certain cardinal direction.

    Rule #4: Passages between rooms always follow cardinal directions and never have junctions.

    Rule #5: Every room must have an evocative name that speaks to either a major landmark or the obvious purpose of the room.

    Rule #6: Each region will have unique descriptors that emphasize the theme and unique hazards or obstacles that can be used only in that region. Each region will also include a list of furnishings and features that fit the theme of purpose of the region.

    Rule #7: Design the macroscopic geography of regions first, then map individual rooms over those features.

    Seems like a set of golden rules to me. Will be a good guideline for setting up my own rules for dungeon making, though I will have to outsource the actual drafting to another, even using the basic outlines that your players will be given is a fine start.

    Stay classy Angry, and see you at the Tops.

    • Why outsource to another? Time or ability? If it’s time, then understandable. If it’s ability, then google map drawing techniques. There are a million Reddits and other places with tips n thangs of that nature. Map drawing is kind of one of those things that’s only has hard as you make it. Simple stuff can really add life to a map (like cross-hatching your borders to fill negative space). I’ve been into cartography since I was little, so i can understand the frustrations of trying to draw a map and not feeling like you got it quit right…the first 3 times.

      • Mostly both, actually. Besides, when I say outsource, I mean have my sister do it, being the artist of the family. As I have zero talent when it comes to drawing, it may be better that way. And as a bonus, she enjoys D&D, so it would actually be very easy to convince her to do so.

        And I also love cartography, just can not do it myself. I personally find to be quite enjoyable to browse.

  12. Wonderful mega-dungeon article! So much good stuff! I must admit to using maps the PCs find rather backwardly, or perhaps lazily haha. If I feel the PCs need one, I’ll let them find it, but I then treat it like a descriptive resource to be interacted with during use via time and Navigation or Inteligence checks. I feel that when you look at ancient maps, they’re so personal to whoever made them, or even the society that produced them. Like many arbitrarily designate North and South opposite to our understanding, or certain features important to the cartographer, like towns and river systems, are stylized or emphasized (an ancient map of the Tigris and Euphrates Rivers from Iraq during the Abbasid dynasty comes to mind). So keeping this in mind, I feel it can be pretty neat idea to just give the PCs a complete map, but require them to constantly interpret and adjust it according to their own needs. Just an idea, and not suitable to all situations. I find your incredibly detailed and methodical approach really interesting and I look forward to seeing the first renderings of the Desiccated Sanctuary!

  13. I’ve left criticisms a lot on previous posts. So this time I’m happy to have a Megadungeon article. Whenever you get the chance. I love reading about this project. Even if I like to point out things that seem like potential problems.

    • You sound like you might have the same issue I do.
      The more I care about something, the more obvious it’s flaws become.
      Also, the closer something is to perfection, the more annoying its flaws become for inhibiting its potential.

      Basically, the more awesome something is, the more I complain about it. 😛

  14. One thing I hate as a DM is having to describe any non-rectangular room to a mapper who’s trying to keep a high level of detail. The principle of giving each room a name goes a long way to encouraging a simplified node-and-line mapping style.

    I’m not sure where this fits into your design schedule, but one element that can often be a lot of fun is to take a step back from the dungeon and give attention to the local town, considering how it can be connected to the dungeon and not just a generic home base with all the standard amenities.

    One obvious consideration is the kobolds and the dragon, who would have a certain amount of treasure not original to the dungeon. Some of it could have been burgled from the town, for a Return the Family Jewels sidequest, but those hoards can also have a more roundabout function of substituting for whatever original loot disappeared from the dungeon, keeping the PCs from feeling too ripped off if some important item needs to be bought back from someone’s private collection in town.

    This does of course imply that someone else has explored the dungeon before the PCs, which requires considering how far they could have gotten, how many trips they made, and whether they died in the dungeon or decided to quit for some reason.

  15. I don’t know if I’m in the minority, but this whole concept of players mapping is absent from my play experience. I’m fairly new, but most of the other players at my table are all self proclaimed old schoolers and in every campaign across a rotation of 4 different GMs, every time a new room is entered, the GM takes the time to copy it out from his notes onto the big sheet of graph paper in the middle of the table.

    I guess I’m spoiled?

  16. Angry,

    Holy. Crap. I am new to your blog, and I just binge-read your entire Megadungeon blog in a few hours. I am blown away, and so excited to start my own Megadungeon.

    I work at a call center of sorts, and sitting in a cubicle with my type of work requires something to pass the time. Recently I had been getting the itch for some D&Ding, so I dug out my old tomes and some dice. Now, I was raised on the old-school 1st edition D&D. My parents taught me that since I was 8, and I’ve never had the chance to try anything else, mostly because I don’t know anyone who does. Regardless, I decided to roll a completely randomized dungeon. And I mean COMPLETELY randomized, because with 40 hours a week to kill, I have plenty of time to kill. Thus, when my boss walked past and spied me in my cubicle with books spread from heck to breakfast, hunched over a sheet of graph paper with some dice rattling in my hand and my tongue in my teeth, he figured I’d enjoy your blog.

    How right he was.

    However, I am now in a dilemma. I had been working away at my random dungeon for quite a few weeks before my boss told me of your website. Thus, I was already over halfway through with the nitty gritty details by the time I started reading your blog. For the next several hours my books and dice were left abandoned while I devoured the ideas you were proposing on every page. You were addressing concerns that I had been struggling with the entire time I had been rolling said dungeon. You were introducing concepts that I would have never conceived of my own volition that blew my f$&%ing mind six ways to The Day Arbitrarily Called Sunday. And I loved every minute of it.

    Now, however, I am finding it difficult to get back to rolling my dungeon. My storyless, labrynthian, pathetic excuse of a dungeon. I have some really neat things in hear, but I can’t find it in me to scrap the whole thing after all the hours I’ve rolled into it, nor can my OCD stand to leave in unfinished.

    I don’t know why I felt the need to tell you all of that. Anyway, all of this to say: your blog is awesome. Keep it up. I love the no bulls$&% attitude. Your sarcasm and wit make me laugh. I can’t wait for more on the Megadungeon.

    Looking forward to more attitude.


    • This is a color-coding system I use for random dungeons, it might help with providing structure. I use a 6-color highlighter set, substitute your own preferred implements.

      Green: The main zone. Anything connected by a clear path from the floor’s most accessible entrance.
      Yellow: Secure zone. Any area separated from the green zone by locked doors or traps. Creatures in these areas are more likely to be intelligent, and contiguous yellow areas are more likely to be controlled by the same faction.
      Blue: Secret zone. Any area separated from the green or yellow zones by secret doors or cave-ins. Intelligent creatures might use these areas, or, if undisturbed by outside forces, they are more likely to contain supernatural creatures that don’t depend on sustenance.
      Orange: Bottlenecks. Any room that controls the only route to another large area. Intelligent creatures will want to control these areas but not sleep or stash their loot here.
      Pink: Lair sites. Clusters of rooms with one or two secure entrances. Intelligent monsters strongly prefer these areas as lairs.
      Purple: Reserve for marking locked doors and any arbitrary boundaries

  17. Hey Angry,
    I’ve been trying for a good long time to build an adventure in Pathfinder, and having recently stumbled upon these megadungeon articles I realized that they are absolutely amazing for that sort of thing. Since this is the most recent article, I was hoping I could put a few questions to you here.

    First, and hopefully most important: I’m hoping to do something that’s a lot more aesthetically-open than a dungeon. It’s an overland-hike through a bunch of different terrains that serve basically the same purpose as your seven dungeon areas, throughout which are scattered lots of special locations (and I’ll be very heavily differentiating “location” from “random spot in the wild” when random encounters come up). I have no problem telling people “here’s a large wilderness, but there’s only so many interesting places in it” and having them travel between those places as if they were dungeon rooms, but I want to capture something like the same sense of awesome pacing that you’ve got building in the megadungeon. I’m actually not super worried about it; I think with a judicious use of slightly-higher-CR encounters when I’m not expecting them to go through a “full adventuring day”, I can still make it feel all cool and actiony. Any problems with taking a megadungeon “out of the walls” that immediately leap out at you?

    Second, with a big “that being said”: that whole “full adventuring day” thing. You’re working in 5e, I’m doing Pathfinder. I’ve made my own little XP spreadsheet to help organize stuff, and what immediately sprang out at me is that there’s no “XP per day” chart for PF. The best I’ve been able to find from the internet is that the system assumes a party is equal to about four equal-CR encounters per day. Again, not tremendously tripping about how to design actual encounters, but the fact that every “adventuring day” (in that abstract, encounter-area sense) is going to be the exact same length is making me sweat a little. I feel like the removal of that variance is going to make it more boring, although I honestly can’t say why I think that. Am I pissing myself for no reason?

    Third, and without a qualifying appendant: your megadungeon doesn’t (yet) seem to included any NPCs who the party is not ultimately expected to kill. I’d like to be implementing a system of picking sides in important conflicts in each area, tied into gating with the idea of “you need one of the NPCs to open the way forward, but they each want the other dead”. I’m hoping it’ll give PCs a nice sense of decision-making that keeps them on the critical path but allows for different-flavored encounters down the road (if you kill the magician’s apprentice to appease the gnome mafia, the Magician’s gonna be pissed when you meet him, but the gnomes will intercede for you with the tengu warlord). Does that mess with any of your core principles? My big mechanic for this is, once one of the “progress NPCs” has been defeated, their paired NPC provides only 10% XP if defeated, like a random encounter. So the bastards don’t just kill the whole place and end up like six levels too high. Again, feels like I’m making mountains out of imaginary molehills, but this is my first major endeavor in this vein that I don’t expect to suck, so I’m a bit nervous.

    And fourth, because I just read “Nothing Here but Worthless Gold!”: Pathfinder is very big on using equipment upgrades as a supplemental level-up system. I think I’m just going to create another spreadsheet, like the XP spreadsheet but using Wealth By Level, and use it to distribute treasure exactly on-level with XP. I’ll allow selling at full base price: I’m turning everything into a cash equivalent. Don’t like the flaming sword, turn it in for its full value and buy something else, because then I can control exactly how much “level by GP” the party gets. The biggest concern I have about this is that crafting feats might become absolutely nuts in a closed system like this. Thoughts?

    Thanks so much for explaining things like people are smart and want to understand these games.

  18. I just spent all night reading through these, and all I can say is – holy s$&%. This is really inspiring, thought-provoking stuff, and I can’t wait to see how it all comes together.

    I also hadn’t even realised until you mentioned it that players don’t really map dungeons anymore. I started playing with AD&D and we always had somebody mapping. I’m not sure when that stopped being the norm, but thinking about it now I haven’t seen anybody do that for years. I’m going to have to try and encourage my group to start doing it.

    • He said he’d hopefully start it back up every Monday starting in November. It’s December, sure, but don’t give up on him. There’s a lot of planning for the actual dungeon, translating that to articles, helpful pictures, etc.

      • I think Angry has been clearing up is questions backlog recently, grabbing the low-hanging fruit, so to speak. It’s holiday season and I’ll bet he’s busy with real-life stuff and can’t devote as much time as he’d like to this particular project.

        Though if you asked me, he’s done enough on MegaDungeon that I think it could be wrapped up at the “macro” level in one or two more articles. Of course, if he wanted to go into detail for each encounter, we’d be talking another few dozen. But after the first couple, I don’t know they’d add much value.

        As for when we’ll get the next installment, I wouldn’t be surprised if it doesn’t get posted until some time in February.

        Oh, I noticed PMSWeekly (or whatever your name is) posted questions in comments above. It’s been my experience that unless the article is less than 2 days old, Angry generally doesn’t respond – except to deal with abuse. To that end, I would suggest you find the means to ask Angry via email (it’s hidden here on the site if you know where to look). If your questions are worthy, he may create an Ask Angry article for you.

Leave a Reply