Megadungeon Monday: And Then You Make a Map

Print Friendly, PDF & Email

Banner Patreon 800 x 100

Category Megadungeon 800 x 400

Happy Megadungeon Monday! Now on an actual Monday!

After discussing last week’s debacle, we’re going to pretend it didn’t happen. Ultimately, remember, we rolled our work back. So, let’s all get on the same page. Disregarding the nonexistent Week of Disaster, two weeks ago, we came up with a map of the critical path of our adventure. Each “day” of adventure – and remember that the word “day” now refers to a collection of encounters and dungeon spaces which provide the right balance that a party could reasonably explore them without needed a long rest – each day of adventure is represented as a colored bubble. The color refers to the prerequisite item or plot event needed to access that area of the dungeon. The lines coming off those days represent the path that we expect the PCs to follow from one day to the next. The little colored push-pin like thingies on some of the days represent optional encounter spaces accessible with certain prerequisites. Remember that?

Image (5)

While we’re at it, let’s review the colors. Light gray areas (days 1 to 3) are areas accessible from the start of the adventure. Red areas (days 4 and 5) require the party to discover a mundane key to unlock doors too strong or complex to break or pick. Orange areas (days 6 and 7) require a more powerful magical key that can also remove arcane locks. Yellow areas (days 8 to 10) are blocked by thick thorny roots that exude powerful poisonous gas. The roots wither and die once the party kills the toxic plant monster strangling the great tree. Green areas (days 11 and 12) are accessible through underwater passages that require the party to use magic to allow them to survive and fight underwater. The aquamarine colored areas are accessible only after the party changes the position of the floodgate so as to drain the lower levels of the dungeon and restore the flow of water to the canals in the upper level. Blue areas (days 16 and 17) are accessible only one the party gains the magical ability to create magical gates between two nearby points they can see. Purple areas (days 18 and 19) are sealed by doors that also project magical walls of energy and can only be dispelled with a certain magical artifact. Brown areas (day 20) can only accessed when the entire party can fly. Black areas (days 21 through 25) are cursed with powerful necrotic energy fueled by a vengeful spirit of a betrayed elven hero and can only be accessed when the spirit has been laid to rest. The final area (day 26) is sealed by the power of four elven priests whose spirits have been bound by the traitor spirit. If those four spirits can be appeased, they will provide the means to open the final gateway and allow access to the trapped demon.

In addition, we made some basic conclusions about how big an encounter area should be to provide a fun combat experience. We concluded that small encounters should have at least 64 squares of space (being roughly 8 squares by 8 squares, 40 feet by 40 feet, or 1,600 square feet). Medium encounter areas should have at least 100 squares of space (being roughly 10 squares by 10 squares, 50 feet by 50 feet, or 2,500 square feet). Large encounter areas have at least 144 squares of space (being roughly 12 squares by 12 squares, 60 feet by 60 feet, or 3,600 square feet).

With all of that in mind, our goal now is to start building the basic map of the dungeon. Not in terms of abstract days but in actual rooms with sizes and shapes. But, as we’ve seen in The Week That Shall Not Be Named, we need to take some intermediate steps.

But it’s important we think about things in the right way first.

What Is an Encounter Space?

Now, even though I spent a lot of time clarifying my reasoning about encounter sizes in the Week that Never Was, I think it’s important to discuss one more idea: what IS an encounter space.

When you get down to it, this project – which it is becoming increasingly clear was a stupid, dumba$& idea thanks to my ridiculously high standards for quality design – has required us to think in some really abstract ways. After all, we basically redefined the entire concept of day from a unit of game time to a chunk of adventure space. And we did that by understanding what a day represents within the space of the game. It’s really a measure of resources. And because encounters are measured in XP and that XP is really a measure of resource expenditure, we could create an equivalence between days and encounters and space in the dungeon.

So, we have this idea that there is a chunk of our dungeon called, for example, Day 3. And within that space, the party will have about four encounters of a certain difficulty and perhaps also one optional encounter. Which means Day 3 requires 5 encounter spaces on the map.

Now, in the old days, we used to call those things “rooms.” But we’re trying to be more interesting than that. As noted, anything can be an encounter space. A room. A hallway. A courtyard. An intersection. Two rooms with a big open door between them. A prison room with several jail cells. A kitchen with a walk-in pantry. A common room with three dormitories off of it. A cave. Two caves connected by a waterfall and several natural stairways or ramps. A warren of tightly interconnected tunnels. After all, any one of those spaces could make for an interesting and dynamic single encounter. A warren is a great place for guerilla kobolds or terrible crawly burrowing bugs. A flying creature with a ranged attack would make great use of the space at the top and bottom of a waterfall. And so on.

And we have two problems. First of all, our dungeon is going to have a lot of fighting. Because we’ve decided that we’re going old-school action dungeon. And that’s what Dungeons & Dragons is good at. But that means our fights have to be interesting and varied. We NEED our encounter spaces to be varied and interesting and give GMs a good excuse to break out the minis and the battlemat.

Second of all, we’ve decided we’re going to have a robust system of random encounters and restocking to help drive the pace of our game. And that means EVERY space needs to provide a potentially interesting space for encounter.

So we need the broadest possible definition of “encounter space.”

And that creates an issue. Because if an encounter can be anything from single room to complex of rooms, hallways, tunnels, and a f$&%ing waterfall, what IS an encounter space? The moment we define an encounter as “pretty much anything,” the word loses all useful meaning.

Back to Video Games: Aggro Radius, Fog Walls, and Other Stuff

When you get down it, the problem is that we’re trying to build encounters without walls. I’m not saying we won’t have WALLS. Don’t be stupid. I’m saying we’re trying to build encounters without using walls as the edges of encounters. And we can look at the way different video games have managed the problem of encounter design without walls.

First of all, not EVERY video game is encounter based. And encounter-based game is a specific type of structure wherein the player encounters a group of enemies in a space, defeats all of those enemies, and then moves on to a new space with a new group of enemies. And, it might be shocking to hear me say this, but a really good example of this design is Call of Duty. Specifically, the one I played the hell out of and was really impressed by was Call of Duty 4: Modern Warfare. It was a GREAT example of encounter design. Even though you would move through an ostensibly open map, the map was divided up into chunks of encounter spaces. And each space had a unique goal. In the hold of the cargo ship, you might have to kill all of the enemies. In the streets of the Middle Eastern nation of Madeupistan, enemies would respawn endlessly until you pushed to a certain invisible checkpoint which would trigger the next encounter. The same happened in the TV studio. You had to fight your way through a series of offices to a specific checkpoint to end the encounter and start the next.

But that still doesn’t help us figure out how to define an encounter space. What about another encounter-based game? One we can’t stop talking about. Metr… oh, wait. I mean Dark Souls. See, the problem with Metroid is Super Metroid actually ISN’T encounter based and Metroid Prime – while it is – is based entirely on rooms. Dark Souls is where we have weird nebulous spaces that are still broken down in some invisible way by encounters. And Dark Souls actually provides two big clues as to what an encounter space ACTUALLY is.

First of all, Dark Souls (and Demons Souls and Dark Souls 2, etc.) use these walls of fog as ways to isolate chunks of the world. Those are most prominently used to isolate boss fights and also to put walls around player-to-player interactions, but they are a pretty big clue as to what encounters are all about.

The other clue is the idea of aggro. Aggro is an abbreviation that became popular among online multiplayer role-playing gamers, most especially among World of Warcraft players. Basically, you have this giant open space and you have enemies wandering around in it. How do you create encounters? Simple, the enemies are programmed with some specific rules for when they become of a player and how they respond. Those rules – which actually vary from creature to creature somewhat – are called aggressiveness or aggro. The most common one is simply the “aggro radius” which is the distance at which an enemy will recognize a player and take hostile actions.

Dark Souls has some pretty obvious aggro mechanics. In some spaces, you can walk pretty close to an enemy before it becomes aware of you. And the aggro mechanics are used to create encounters. The three fire-throwing skeletons above the bridge in the Undead Burg all become aware of the player at the same time and create a gauntlet of fire on the bridge. But the skeleton knight and his axe-wielding buddies across the bridge, even though they are no farther away, don’t become aware of the player until he crosses the bridge.

Of course, if you don’t deal with the fire-throwing skeletons, you don’t have an avenue of retreat from the knight and his axe buddies and two encounters can combine. But the aggro mechanics create this separation between two encounters despite their physical proximity.

Now, D&D doesn’t use aggro mechanics. Instead, it uses awareness mechanics to determine things like surprise and sneak attacks, but it has always sort of left it in the hands of the GM to determine when an encounter actually starts.

Either way, the point is, an encounter is defined by its isolation from other encounters. Either by physical walls or by anything else that keeps monsters from getting involved in other monsters’ encounters.

For example, imagine you have a common room with three dorm rooms. And each dorm room contains two soldiers. If the soldiers hear adventurers in the common room, they burst from their dorm rooms and attack. That’s ONE encounter. And it is created by the fact that all of the creatures in that particular chunk of space get involved once s$&% gets real.

On the other hand, imagine a common room connected to three dorm rooms but each of the door rooms is down a 100-foot hallway for some stupid reason. Add a few closed doors and it’s unlikely the soldiers will all “become aggressive” at once. Instead, the PCs will encounter the soldiers one at a time in their dorm rooms.

So, we’re going to define an encounter space as a chunk of the dungeon wherein, if the PCs attract ANY attention, they attract the attention of ALL of the inhabitants of that chunk of the dungeon. That is to say, an encounter space is a space where all of the inhabitants act as one unit against the PCs.

On Random Encounters and Restocking the Dungeon

It might seem like splitting hairs, but the idea of breaking the dungeon down into encounter spaces – and I fully admit it’s not a new idea even in D&D and 3rd Edition, 4th Edition, and Pathfinder all had various ways of breaking an area into encounter spaces that played into the tactical encounter format that became so popular – the idea of breaking the dungeon down into encounter spaces will actually serve us really well when we design the random encounter and restocking system. If we isolate chunks of the dungeon specifically as “encounter spaces” rather than considering them in terms of “rooms,” we can create a manageable mechanic for random encounters and we can also carefully control how many rolls are likely in a given day and thus manage how many encounters the players are likely to see. Obviously, we’ll talk about this in more detail later.

Metroid Shows Us the Way

“Okay, okay,” I hear you saying, “we get it, encounter spaces and days are spatial measures of the dungeon, but this isn’t making a map!” You’re right. The question is, how this helps us create a map? We can’t go room by room. That would be crazypants and it would probably lead to an entire week of lost work! But we CAN go by encounter spaces.

I mentioned that Metroid Prime is encounter based, but it uses the approach that a ROOM equals an ENCOUNTER. And that sort of defines the space in Metroid Prime. You have big, interesting rooms connected by long squiggly mostly empty hallways. And, to be honest, that was originally how I was viewing THIS project. I figured any given section of MY dungeon would probably look a little like this map of, say, the Chozo Ruins from Metroid Prime:

chozo ruins map

But I realized that wasn’t the best way. But I noticed something as I was replaying Metroid Prime 2: Echoes. Sure, the map was a collection of rooms-as-encounter and squiggly transition spaces, but when you zoomed out to the “world map” as it’s called, you got this 3D rendering of the different areas of the game as a bunch of hexagonal spaces. Basically, the map was broken down into abstract chunks and those chunks contained the details of “rooms-as-encounters” and “squiggly transitions.”

Prime Map 1

Prime Map 2

Now, what I realized is that I had basically already argued for a very small variance in the size of “encounter spaces.” And I had basically argued that all of my encounter spaces could fit inside a square that was somewhere between 40 feet by 40 feet and 60 feet by 60 feet. In fact, the size didn’t really matter at all. If I just accepted that every encounter space would be a squarish chunk of the dungeon I would later fill with a room or several rooms or a courtyard or a waterfall cave or a warren of tunnels or whatever, I could just start allocating squares to the days.

For example, I could build up a “day” as just a bunch of arbitrary squares of “encounter spaces.”

Example of a Day

With some graph paper and a pencil, I could just start building up days out of squares. Sure, I’d still have to futz and fiddle a bit to make sure there was enough space to get days within days. But if all I’m doing is chunking up squares and I have an eraser, it’s a lot easier to build up a dungeon.

On my teeny tiny graph paper, I ultimately decided that an “encounter space” would always be three squares by three squares. Later on, I’ll decide how many squares those squares are equivalent to on an actual map. Initially, I thought that the scale was 5:1. That is to say, a one square by one square space would five squares by five squares of tactical space. But now I’m thinking 4:1.

If this scale stuff is confusing right now, don’t sweat it. It won’t matter until later. For now, just understand I knew I needed to build days out of squares and I have this teenie-weenie graph paper, so I decided each “encounter space” was three squares by three squares. We’ll worry about sizes and scales later.

Why not just use big graph paper and make encounter space one square? Well, you’ll see.

But, hold on a sec. Here’s an interesting question: how many encounter spaces does each day need? You might think we already have the answer. You’d be wrong.

The Importance of Empty Space

We made this big f$&%ing spreadsheet, right? And that told us that Day 1 has 6 encounters and Day 8 has 5 encounters and whatever. And then we made this flowchart and added “and also Day 3 has 2 optional encounters stuck on it and Day 17 has 3 optional encounters stuck on it.” So, it would stand to reason that we know exactly how many “encounter spaces” fill each day, right?


It is actually incredibly important to leave empty space in the dungeon. Every dungeon has empty rooms. Every dungeon NEEDS empty rooms. And there’s a few reasons.

First of all, empty rooms are a pacing issue. They provide spaces where the players can make decisions like “left or right” or “examine the statue” or “search for secret doors” without having an encounter (be it an obstacle, a monster, or a trap). In terms of game structure, they are breather spaces. Spaces where the players can relax for a minute and keep playing the game without the game trying to kill them. That’s really important because players can get fatigued by endless action.

Then, too, our dungeon is not JUST encounters – combats, traps, and obstacles – our dungeon also contains treasures and discoveries and history to explore and all sorts of other crap that makes it more than just a combat slog. And while, yes, some of that will be hidden behind encounters – kill the monsters and then loot the room for example – not everything that is worth discovering needs an encounter standing in front of it trying to kill people. That s$&% needs space.

In addition, if all of our encounter spaces are used up by encounters on the critical path EXCEPT for the side passages that lead to the optional encounters, there really isn’t an element of exploration anymore. That is to say, the dungeon just becomes too linear. There’s no interesting spaces to find.

Finally – and this is kind of a subtle one – we’re relying on the probability of a certain number of random encounters each day. We’ve minimized the XP effect of that, but we’re still planning on it. If we’re going to key random encounters to encounter spaces, we need the players to be wandering through empty spaces now and again so that they can bump into unplanned encounters. Sure, backtracking and leaving and reentering the dungeon will account for some of that, but the thing is, if we know how much space in the dungeon starts off unoccupied, that will help us set initial probabilities for random encounters in a way that we like. The random encounter system has several jobs to do and, if random encounters never happen, it won’t be doing them.

Now, you could say that most of the empty space could just be side chambers added on to the sides of encounters. It doesn’t need to be “encounter spaces.” But, again, we’ve decided the party can get jumped by monsters ANYWHERE. Every space in our dungeon is an encounter space. And frankly, again, that makes the whole thing easier to manage anyway.

So, roughly speaking, how much empty space does a dungeon need? Well, there’s no real formula and there’s not really a good way to answer that question. But, to some extent, it really doesn’t matter. We’re going to backfill our random encounter system based on how much space we have and how likely we want encounters to be. As for discoveries, side passages, breathing space, and so on? Well, you could argue just about any number you want. The real key is that we’re going to have to make every space interesting. The more empty spaces we include, the more pressure we’re putting on ourselves to do SOMETHING in that space. Because, when we’re talking about empty encounter spaces, we’re not just talking about empty rooms. As we’re going to talk about in the coming weeks, it’s very important for every space to be interesting, even if it is devoid of encounters.

The most important thing to realize though is, from now on, it’s going to be VERY hard to go back and change something. Now that we’re actually allocating physical space in the dungeon and on maps, we’re locking ourselves into decisions. Having TOO MUCH empty space is easier to fix than not having enough. After all, we can simply FILL excess empty space with story details and bits of lore and trivia that aren’t valuable enough to count as major discoveries. Or just make them pretty. Scenery porn, if you will. Or we can just make smaller spaces and longer connecting hallways to erode the empty space. But if we don’t have enough empty space, we’re going to have a hard time adding more.

In the end, my gut tells me that 30% to 50% of the dungeon should be empty encounter spaces. In other words, if an adventuring day consists of anywhere between 4 and 7 planned encounters, that adventuring day should also have anywhere from 3 to 6 empty spaces. Roughly speaking, I would expect any given adventuring day to comprise 8 to 12 encounter spaces.

Obviously, this will vary somewhat because we also need to consider how the physical spaces work. For one example, Day 4 has to be long enough to move the party all the way from the Desiccated Sanctuary to the Great Tree. Looking at our flowchart, we can see how the physical spaces have to come together. But we also have to consider other factors, like the shape of the area and how linear or open the design should be.

And Then the Dungeon Happened

And now comes the hard part. As I might have mentioned in the Week That Totally Didn’t Happen, the actually laying out of the encounter spaces on the graph paper was just a bunch of trial and error, erasing and sketching, fitting stuff together, and then coloring in what I liked.

At this point, I don’t want to say “and given all of that, I just kind of drew the whole dungeon.” But that’s sort of what happened. And the truth is, sometimes, that’s how you HAVE TO do it. You just have to sit down and sketch and erase and sketch and erase and sketch and erase until something good comes out. Which is precisely what I did.

Initially, I was working on graph paper. I worked in light pencil, laying out roughly 8 to 12 boxes per adventuring day, each one three squares by three squares. Once I was sure everything worked, I went over the whole thing in ink to add color. So, Day 1’s 10 boxy “encounter spaces” became an abstract pink mutant Tetris box. Why pink? Because I didn’t have a gray marker that worked well. So pink became the color for spaces open from the beginning.

And that is how THIS happened:

Boxes to Day

Note that the downward arrow shows a connect to Day 2 on the floor below.

I then took the entire map and copied it into the Profantasy mapping software Campaign Cartographer. This wasn’t a necessary step, per se, but I think it will be a lot easier to show off the different parts of the map going forward. Campaign Cartographer 3 is a great fantasy overworld, dungeon, and city mapping tool built on top of a very power CAD (computer aided design) engine. I don’t use it much, because I prefer to hand draw my maps, but I recommend checking it out.

Anyhoo, the hand-drawn map looked like this when all was said and done.





And here is the Campaign Cartographer version. If you click on it, you will get a much larger image.

Megadungeon Map

What’s important to note is that, if you look at the flowchart, you can see how it determined where each space had to go. But, what the flowchart doesn’t show is what other passages might exist. For example, did you know that you can probably travel from Day 4 or Day 5 to Day 11? Neither did I! But I know now. And because I know Day 11 requires water breathing (it’s green), I know THAT passage must be underwater. Or it must require some higher power, like the floodgates.

With the blocky representation, I can see how the areas of the map interrelate much more clearly. And that means, I can disrupt my critical path a little. I can allow for shortcuts, side passages, and other connections.

A Tale of Two Maps

But I know all of that explanation is a little unsatisfying, isn’t it? After all, you want a goddamned tour of each day. And you want to know WHY Day 1 is shaped the way it is, don’t you. You want to know how the map got built. Well, next week, we’re going to dispense with the theory craft. We’re going to take a tour of the dungeon, one day at a time. We’re going to talk about how I went from this…

Example of a Day

To this…

Boxes to Day

To this…


To this…

Megadungeon Map

Because even though I was sort of just drawing and erasing and doodling and throwing s$&% at the wall to see what stuck, I was also making secret decisions. I was making decisions about the areas of my dungeon. More importantly, I was making decisions about THIS (click for a larger view):

Critical Path

If you’ve been paying attention, you know what THAT is. So, next Monday, we have a nice meaty TALE OF TWO MAPS.

Banner Patreon 800 x 100

10 thoughts on “Megadungeon Monday: And Then You Make a Map

  1. Love how it’s coming together, but one suggestion – in the future, would it be possible to include the legend on the Campaign Cartographer version? It’s a bit of a pain in the a$$ to refer back to the top of the article to check what each color represents.

  2. The last picture makes me very happy. This is very exciting and informative! I’m definitely looking forward to the tour, and of course the nitty-gritty of carving tunnels and dungeon rooms out of those encounter spaces.

  3. Looking at the map and letting ideas come to mind, I see a particularly cool one. There’s an optional blue area attached to Day 7. IE, an area adjacent to the great tree that you need the portal gun to access.

    It would be cool if that optional area was in the canopy of the great tree itself, and you need to use portals to climb up there.

  4. One issue I’ve always had in designing dungeons is the tension between realistic, living environments (be they fortresses or cave systems or tombs) and the need to separate encounter spaces.

    On the surface of it, putting space between rooms solves a lot of the problems (especially where undead or traps are involved), but I’ve aways had this ‘every creature must have access to food and water without traipsing through some other critters lair because that ends poorly for one of them’ hang up that makes proper dungeon structure damn near impossible. The other side of it, ‘everyone gets along or is part of a single group’ tends to cause the opposite problem, where finding ways for an encounter to conceivably end without alerting the whole dungeon becomes a problem.

    I’m most interested in seeing how the kobolds get separated in the early days…

Comments are closed.