Let’s make an adventure. Remember that I’m doing that? Sorry it’s taken me a while to get back to this series. But I am back to it. And ready for the next step. And I’m going to skip the usual longness and ramblingness and most of the introductionness and just jump right in. Well, after I do a recap. Because it’s been a while.
It’s Been a While; Here’s a Recap
I’m building an adventure to show you how to use my Double-Impossible Adventure Building Checklist. Specifically, I’ve been working on the last part of a three-part adventure. You don’t need to remember why I’m working on the last part first, but you do need to remember this part is the dungeon-crawling part. The heroes have tracked kobold raiders back to their lair in a formerly abandoned mine and want to recover the thing the kobolds stole from the dragonborn holding them in thrall. The heroes have to enter the mine, find the dragonborn, and get the thing. But if the heroes raise a ruckus and the dragonborn realizes they’re coming, he’ll attempt to flee into the wilderness. Once he does that, the adventure is over and the heroes have failed.
Previously, I came up with a roster of baddies to stock the dungeon with. A good mix of kobolds and their pet drakes. Now, I need somewhere to put them. I need a dungeon.
A Dungeon… With a Twist
I call this kind of adventure – or this scene anyway – a dungeon with a twist. It’s a dungeon – a self-contained environment the heroes can explore at their own pace – but the heroes can’t win the adventure if they treat it like a dungeon.
In a normal dungeon, the players wander from node to node – from one room or encounter space to another – as they wish. And they keep going until there’s nothing left alive that isn’t a PC and nothing valuable that isn’t in a PC’s pocket. That’s dungeon crawling.
But in this dungeon, the heroes have to move carefully and quietly. They can’t let anyone raise the alarm. The first time a baddie raises the alarm, the encounters get more difficult and the monsters change their behaviors. The second time it’s raised, the dragonborn starts heading for the exit. And the heroes have to realize that and move to intercept him. Otherwise, they lose him forever.
And the fact that this is a dungeon with a twist means that, while there’s different ways to skin a dungeon, there’s only one way that’ll work for this dungeon.
Three Ways to Skin a Dungeon
Before I can explain the three ways to design a dungeon, you have to understand that every dungeon has two maps. Two scales. But they’re usually combined into one. The macro-scale – the layout scale – shows how all the different spaces in the dungeon join up. It shows where everything in the dungeon is in relation to everything else. The micro-scale – the encounter scale – shows all of the minute details that affect the encounters. It shows precise room interior layouts, terrain features, monster locations, furniture, and other s$&% like that. And it’s got a very precise scale. Usually, a grid.
When you see a dungeon map, it usually shows you both scales. That is, the entire thing is to scale and all the furniture and terrain and encounter-scale crap is visible in every corner of the map. But just because they’re usually shown together doesn’t mean you have to design both together. Hell, you don’t even have to always show both together. You can actually draw a general layout map with a vague scale that shows how all the spaces connect and then draw separate encounter maps for all the rooms. Or just for the rooms with encounters in them. The rooms that don’t have encounters in them can be described adequately with text in the notes. The layout map doesn’t even have to be fancy. It can just be a flowchart. And that can be a really great time-saver. Especially if you’re running a game online. You don’t need to import an entire dungeon map to Fantasy Grounds. Or even draw it. You just need the encounter maps.
Like I said in the heading – and in the segue preceding it – there’s three ways to design a dungeon. First, you can design all the encounter areas in encounter scale and then just fit them together. And that’s actually a really common way to design dungeons. You can usually identify a dungeon that’s basically just a haphazard arrangement of encounter spaces just by looking at it. Check out the map of the Fane of the Night Serpent from Tomb of Annhilation by Wizards of the Coast. It has all the hallmarks. Rooms scattered willy nilly. Long hallways. The works.
For more examples, check out literally every published dungeon in every product released for D&D 4E. Because those maps were basically built from interconnected Dungeon Tile layouts.
Haphazard encounter arrangements are just about the gamiest way to design a dungeon there is. There’s no mistaking such a dungeon for anything other than a space that exists to have encounters in. But on the other end of the spectrum, you have dungeons that look like what they’re supposed to look like. Check out the Haunted House from The Sinister Secret of Saltmarsh that was reused in Ghosts of Saltmarsh.
You can tell that map is supposed to be a house. And you can guess what most of the rooms are just by looking at the layout and the furniture. Those maps make logical sense. They’re structures in the world and they look the part. The problem with dungeons that look like what they’re supposed to look like is that the encounter spaces usually have issues. They might be bland. Lacking in interesting features. Or they might be too cluttered or too small for the battles they’re supposed to contain. Look at room 9 on that map. The kitchen. There’s four giant centipedes in there waiting to attack a party of three to five PCs. Just where the hell is everyone going to stand? How are they going to move?
But despite their hallmark problems, either of those styles is good enough for any standard dungeon. And most GMs and adventure designers – including me – switch back and forth between those styles regularly. If the dungeon I’m building is a recognizable structure, I design it as a dungeon that looks like what it’s supposed to look like. If it’s a gauntlet of encounters or it’s unconstrained by something like a building footprint, I usually build a haphazard encounter arrangement. But dungeons with a twist usually require some extra work.
The gameboard dungeon is – as the name suggests – also kind of gamey. But it’s a different kind of gamey from the haphazard encounter arrangement. It’s gamey on the layout scale. That is, the design is focused on making a dungeon that works as something other than a dungeon crawl. A dungeon whose structure provides part of the challenge. Like a Metroidvaniastyle dungeon or Zelda-esque dungeon where part of the challenge is solving the dungeon’s layout. Or like a dungeon in which the encounters are just a warmup act and the real game is a cat-and-mouse chase.
To be clear though, each kind of design has strengths and weaknesses. Gameboard dungeons are the hardest to design and the most likely to break during gameplay. Haphazard arrangements are contrived. Dungeons that look like things are really constrained. As an adventure designer, you need to learn when to use each kind of design.
And also remember that the best dungeon designs START with one kind of design but incorporate elements of all three. Every dungeon has to end up looking like SOMETHING. Even the Fane of the Night Serpent – with its central worship area, impressive entrance hall, and side areas radiating from the center – implies a temple. It’s not well-engineered or efficiently design considering how expensive and difficult it is to build such an underground structure, but it’s not completely nonsensical.
Even though I’m going to start with a pretty gamey gameboard design, I’m eventually going to make it look like what it’s supposed to look like.
Laying Out a Dungeon in Four Easy Steps… And a Lot of Pre-Thought
Click on the tip jar to leave a tip
Because I’m working on the layout scale, I’m not worrying about filling in the details for each encounter. I’ll do that as I design the encounters themselves. Today, I’m just deciding how everything joins up and where the encounters actually are in relation to each other. Though I will address scaling the map to encounter scale at the end.
To design a layout that works as a gameboard, you have to think about the game being played on it. In it. Whatever. And I’ve already spelled that out. As the players make their way through the mine, the kobolds will try to raise the alarm. If the alarm reaches the dragonborn – twice – the dragonborn will grab his s$&% and beat feet for the exit. If the players handle everything perfectly, they can cut their way quietly through the encounters, picking off kobolds, and catch the dragonborn by surprise. If they f$&% up bad enough, they’ll raise the alarm in the first few minutes and the dragonborn will run. But, most likely, the heroes will trip the alarm the first time about halfway to their goal and the second time pretty close to their goal. And then the mouse grabs the cheese and does a runner.
A good cat-and-mouse chase requires multiple paths from place to place, but not too many. If there’s too many, the cat can only run around at random. If there’s too few, there’s no way for the mouse to avoid the cat.
A good game – any game – strikes a balance between inertia – when you start winning, you keep winning – and catch-up – winners can become losers at any moment. However, a good role-playing game always skews that balance in favor of the players. If the players get ahead, they should win. And if the players start to lose, they should be able to pull out a win.
If the players raise the alarm when they’re close to the boss, they should feel like they have to give chase, but their victory should be a pretty sure thing. Any effort should be enough. Meanwhile, if they screw up and raise the alarm when they walk in the door, they should still be able to win. They might just have to work harder for it. And if they raise the alarm somewhere in between, they should feel like victory is slipping away, but that it’s still in their grasp.
The dungeon’s layout has to provide for all those possibilities.
First, there can only be one way into and out of the dungeon. If the dragonborn has a secret exit, the players can’t head him off. With only one way out – which the players can always locate because it’s the way they came in – the players can just hustle for the door when they realize the dragonborn’s running and try to intercept him there. Second, when the dragonborn cuts and runs, something’s got to slow him down a little. That way, the players have time to realize what’s going on and adjust their approach. The easiest way to do that is to start the dragonborn far from the exit and force him along a circuitous route.
Remember, by the way, that the dragonborn will be adding surviving kobolds to his entourage as he runs. That means, in the worst-case scenario where the players trip the alarms immediately, they can recover just by planting themselves firmly in the exit and letting the dragonborn come to them, but he’s coming with a lot of buddies that will make the fight much more difficult. Meanwhile, the party that managed to clear a lot of kobolds out before they spooked the dragonborn will have an easier fight at the end for their trouble.
And while we’re talking about inertia, remember that you always want the beginning and the end of the game to move faster than the middle. Good games build momentum at the beginning and then speed toward the conclusion once it’s clear who’s likely going to win. I can make that happen by having the paths converge and putting the shortest escape routes nearest to the exit. While the dragonborn has to cover a lot of ground at the start of his flight, the closer to the exit he gets, the faster his escape comes.
I bet that seems like a lot of s$&% to consider just to draw a map of an abandoned mine for some lizard monsters to temporarily occupy. Well, it is. But the truth is, it all just comes from thinking about how the adventure or scene should play – how it should feel – and asking yourself how the dungeon layout makes that happen. To be fair, I have an advantage right now because I already designed the map. I’m just explaining why I think I made the decisions I did. I thought about how the scene should play out, came up with a basic idea for a layout, and then started sketching. The more of this s$&% you do – and the more you immerse yourself in the thought processes of other designers – the more likely your brain is to spit out good designs even if you can’t articulate precisely why they’re good designs.
Anyway, that’s the thought process. Now, I’m going to start drawing.
Step 1: Sketch a Layout that Suits Your Needs
For the overall shape, I chose one of those circular maze-type patterns with the entrance in the middle. That works because the outer circles are bigger than the inner circles. So, as you move away from the center, the paths get longer. They get more spread out. Because I didn’t want too many paths, I made the thing just three rings deep and I only put two paths between each ring. I put the dragonborn on the outside, but I gave him an extra little arc so he has two directions to choose from when he steps out his door and a way to bypass the party unless they are actually standing on his welcome mat. And I placed the bridging paths between the outer ring and the next as far from each other as possible to slow the dragonborn down.
Step 2: Plop Down Encounters and Empty Rooms
With an idea of the general layout, I decided where the heroes would run into major and minor planned encounters. Fortunately, I knew how many of those I’d need. Two major and four minor. That’s enough to fill two sessions of play – especially given the players won’t necessarily run into all of them – and also enough to give the players ample opportunities to f$&% up and trip the alarm.
I placed the encounters so that the heroes would run into a minimum of three encounters on their way to the dragonborn – two minor and one major – if they lucked into the most efficient route. And I assure you that when the map is done, it’ll be a lot harder to spot the most efficient route. Especially from the inside. If the party lucks into an efficient route, they have to screw up two out of three encounters to make the dragonborn run. But each extra encounter they run into – or each complication that comes up as they dawdle – increases the odds of a screwup.
I also put a few ’empty’ rooms on the map. They give me the chance to add things that aren’t encounters to the map later. Interesting events, bits of environmental storytelling, treasure, whatever. Some might end up truly empty or might seem interesting but turn out to be empty, slowing the party down to build some tension. Generally, I leave at least as many empty rooms on the map as I have minor encounters, but I often add more. In this case, though, there’s not much environmental story to tell and the party isn’t really hunting for treasure, so I don’t need that many.
Step 3: Re-Sketch the Layout to Make it Look Like Something
Your initial layout is always going to look a little artificial. Contrived even. No one really builds anything in the shape of a concentric maze unless they live on Myst Island. And the circular shape isn’t the important part anyway. What’s important is the way the paths connect the encounters and how they get longer as you get farther from the exit. I could take this whole layout and resketch it anyway I want as long as I keep the connections between the rooms and the distances between them the same. I could make the whole thing into a ruined castle shape if I wanted to.
But this is a mine. [[ And… it turns out I don’t know as much about the history of mining as I thought I did. Originally, I made some statements about pre-modern mines that turned out to be wildly inaccurate. And got called out. Rather snarkily, but not unfairly. And while you really shouldn’t be expecting to learn anything about the history of any major industry other than the gaming industry from me, it is my fault for speaking categorically. Which is why I’ve edited this part. – Angry ]]
This is a mine. And this particular mine – I’ve decided – is an unimpressive, accidental mine where the locals discovered some exposed ore in some natural caves produced by erosion. They dug the exposed ore out and then the miners proceeded to dig extra tunnels, following the veins until they gave out. There wasn’t much ore here, which is why the mine ended up abandoned. So, essentially, I just need some jiggly jaggly natural caves with some smoother, straighter mine tunnels shooting off here and there.
And my sketch mostly already works for that. It’s basically just a series of tunnels connecting larger caves. The tunnels need to be windier and I need some mine tunnels spiking off, but otherwise, I’m golden already.
Step 4: And Then Draw a Beautiful Map
Once you have a layout that works for the game you’re running but also looks like what it’s supposed to look like, you just draw a beautiful map over it. Widen the lines into halls or tunnels or whatever, draw the rooms with room shapes, and so on. If you sketched everything lightly in pencil, you can just draw the real map over the sketch and then erase. If you have a light table like I do, you can put your sketch under a blank sheet of paper and draw a dungeon over it. Whatever works. Don’t be afraid to adjust as you draw the final layout to make sure the dungeon looks like what it’s supposed to look like.
I made the tunnels windy and bendy. Partly because it makes them look like natural caves, but also because it prevents people from just shooting fleeing kobolds in the back. If the heroes want to stop a kobold from raising the alarm, they have to work for it. Notice also that if I hadn’t shown you that concentric circle maze sketch, you’d probably never know it was there. I told you I’d make it hard for the players to spot the most efficient route. That’s fair game though considering the best strategy to intercept the dragonborn is just to wait in the entrance for him to show up.
Scaling Down
Like I said, I’m not building encounters now. I just wanted to figure out the overall layout. But you might have noticed that I have given no thought at all to a precise scale on this map. Well, the more astute among you probably noticed that all of the tunnels are about the same width and the rooms are never more than about eight or ten tunnel-widths across. I’ve drawn a lot of maps. I can sketch to a relative scale pretty well.
Either way, though, you might wonder how I’m going to build encounter scale maps from this. Well, there’s a couple of things I COULD do.
First, I COULD just say, “close enough is good enough.” I could draw encounter maps for the chambers I need and just freehand them. Get them as close to the proper size and shape as the layout map suggests. And I do that a lot, truth be told. No one but me will ever see my layout maps anyway.
Second, I COULD do all this sketch work on graph paper and keep a close eye on the scale of things. Except I can’t do that now. Because I didn’t. But I could have done it from the start. And, like I said, I did freehand to a rough, relative scale anyway. So I sort of did sketch this on graph paper. Just graph paper only I could see.
Third, I COULD scan the map I drew and slap a grid over it. That’s easy. Watch. I’ll do it right now.
Fourth, I COULD do all this planning and sketch work on a computer so it’d be easy to resize it and add a scale. But that’s something I can’t do now. I had to start that way.
Fifth and finally, I COULD scan the thing into a computer and then trace it in my favorite piece of mapping software and thus end up with an electronic map I can easily match to a proper scale and then add details to. That’s also pretty easy. Here. Watch. I’ll do it with ProFantasy’s Campaign Cartographer 3+ and the Dungeon Designer add-on. It’ll just be some bland blobs for now, but it’s just a layout map. Who cares.
And that fifth option is the canonical ending. When I start building encounters, I’ll be doing the mapping work in CC3+. On that exact map. But I did plan and sketch the whole thing on paper first. That’s how I roll.
I can’t tell you how you should roll. Or what tools to use. Or what works best. I can just tell you that you’ve got to know your tools and what you can do with them. If you don’t have any tools to help you sketch a layout and turn it into a map and transition from layout to encounter scale, you’ve got some homework to do. Find some tools, learn how to use them, and then get good with them. Because that’s what it takes to map a good dungeon.
Nice talk about types of maps. I’ve been bothering too much with Gamey Maps, and too little with cohesive spaces lately.
That said, I rather like seeing how the skeleton (of Angry’s map) gets dressed up, hidden form the world, yet its presence, the influence, the INTENT is still visible to an active observer. Then again, I just like seeing things get made, and all the steps along the way.
I didn’t even know the types existed, my knowledge only extends to “Jaquaying” dungeons (simplified line-route maps to help find and remove linearity in designs) and challenge/encounter types. And that Step 3 to Step 4 transition blew my mind a little- and Angry always complains he can’t draw quite so good.
How do the intersections play out at the table? When I design dungeons similar to this one, the actual movement throuh the dungeon plays out underwhelmingly:
DM: “You come across a fork in the path, where do you go?”
Players: “Uh, dunno. Right?”
With many intersectons, this gets boring. I have tried adding additional informarion to the options (sounds coming from paths, entrance sizes, signs, etc.) but it rarely works out as well as I hoped.
When PCs first enter a dark dungeon where they don’t know what to expect, they might as well flip a coin to go right or left. A group of PCs in a mine would not know, absent other information, if going to the right or left is better so they can only go with trial and error.
But as the PCs gather more information about the dungeon then you can give them more detail about their choices. You can say the intersection goes to the right or left and remind them (or not) that the goblin they interrogated earlier said the boss’s chamber is to the right.
This can happen as they explore the dungeon as well. One thing I do is give rooms in the dungeon readily identifiable and consistent identities. So my current dungeon is a tomb complex and it has an ossuary, embalming lab, tombs of attendants, tombs of kings, and a burial area for slaves. Once those rooms are identified, I use those names consistently, and I will provide information on how the rooms sit in relation to each other. So I’ll tell the players, “you’re in the great hall with the tombs of the kings to the east, tombs of the attendants to the west, the entrance hall to the north, and the embalming lab to the left.” Giving each room a strong identity that you reinforce helps players (and you as GM) the relationship between the rooms. Consider telling someone how to get to you pizza place. You may say to head east 500 meters, turn north for 150 meters, and then heading east again for 200 meters or you might tell someone to head down the street to the east, take a left at the blue house, then take a right at the playground, and then you’ll be at the pizza place.
If all else fails, you can literally signpost things. If the orcs share a cave complex with a terrible drake then the orcs might very well have a sign up at the entrance that says “orc tribe dis way” with a big red arrow. If your players ignore that then the consequences are on them.
Never forget the natural inclination of the player is to try and explore every part of the dungeon. Even if they’ve achieved their goals in the dungeon, the adventures can be assumed to continue to explore the dungeon. So you kinda need to either take that behavior in stride or give them a very, very good reason to remain on track.
(Apologies for having initially posted this in error below DnDistant’s comment.)
Watching you design maps and encounters is pretty fascinating. You have many strengths, but I believe that to be one of your greatest. Makes me miss Megadungeon Mondays :D. Still, awesome article and thanks as always.
When PCs first enter a dark dungeon where they don’t know what to expect, they might as well flip a coin to go right or left. A group of PCs in a mine would not know, absent other information, if going to the right or left is better so they can only go with trial and error.
But as the PCs gather more information about the dungeon then you can give them more detail about their choices. You can say the intersection goes to the right or left and remind them (or not) that the goblin they interrogated earlier said the boss’s chamber is to the right.
This can happen as they explore the dungeon as well. One thing I do is give rooms in the dungeon readily identifiable and consistent identities. So my current dungeon is a tomb complex and it has an ossuary, embalming lab, tombs of attendants, tombs of kings, and a burial area for slaves. Once those rooms are identified, I use those names consistently, and I will provide information on how the rooms sit in relation to each other. So I’ll tell the players, “you’re in the great hall with the tombs of the kings to the east, tombs of the attendants to the west, the entrance hall to the north, and the embalming lab to the left.” Giving each room a strong identity that you reinforce helps players (and you as GM) the relationship between the rooms. Consider telling someone how to get to you pizza place. You may say to head east 500 meters, turn north for 150 meters, and then heading east again for 200 meters or you might tell someone to head down the street to the east, take a left at the blue house, then take a right at the playground, and then you’ll be at the pizza place.
If all else fails, you can literally signpost things. If the orcs share a cave complex with a terrible drake then the orcs might very well have a sign up at the entrance that says “orc tribe dis way” with a big red arrow. If your players ignore that then the consequences are on them.
Never forget the natural inclination of the player is to try and explore every part of the dungeon. Even if they’ve achieved their goals in the dungeon, the adventures can be assumed to continue to explore the dungeon. So you kinda need to either take that behavior in stride or give them a very, very good reason to remain on track.
If I were a pretend elf with all the time in the world, I’d build my mine like that.
Seriously though, I sit here nodding along happily but the real challenge is doing this myself. Well differently of course, I’m designing a different adventure..
Thanks for the post. I have a brief question and a longer one.
Briefly, you say there are three systems to use when designing a dungeon, the disconnected encounter dungeon-as-dungeon way, the realistic place-as-dungeon way, and… what’s the third? Is the gameboard dungeon intended to the be third system? Apologies if that’s super obvious to everyone who does not have a migraine.
Secondly, how does one indicate to the players essential adventure-specific elements? In other words, do we tell the players that the dragonborn could flee if enough alarms are raised, and if so, what means are there to communicate that information? I may be missing something (again, migraine), but I’m not aware of where in the adventure plot the PCs are informed that the dragonborn could flee and alter the climatic encounter. Something like kobold alarms might be sufficiently intuitive that the players could reasonably figure out the consequences of their actions, but it what about where the actions of the PCs and the results are more attenuated?
One option is to explain the results of the PCs’ actions when the consequences arise. “Haha,” the dragonborn cackles, “my kobold servants altered me to your presence so I was able to prepare this ambush.” I really like that generally as it shows the connection between what the PCs do and what the antagonist does.
However, it does not proactively inform the players that the PCs’ actions could lead to this result. GMs shouldn’t hide the sausage, at least not all time, on things like this; we have a better game when players can anticipate consequences and act in accordance. But my problem is figuring out how to communicate that to players to they take such anticipatory actions.
On the second question, you could give clues. A Kobold that runs away could alert others, this probably obvious to many players but you could hint at it when you describe that he escaped.
In this example I think the kobolds also prepare for intruders after the first alarm so you could describe them as such when the PCs encounter them. Add a line “the other kobold must have alerted them” if you want to make it almost explicit.
If they don’t pick up on what’s happening despite everything, there’s always the possibility of not catching the bad guy.
That’s the wrong information though. Showing the kobolds raising an alarm will only indicate that the kobolds will go for help, and then when the next fight is harder it confims “escaping kobolds = harder fights”. Having encounters start to scale in difficulty based on alarm status implies this behavior will continue, not that if you anger them enough they get scared and run away.
Any time you’re thinking about clues, the most important part is “what conclusion do I want my players to draw from this?”
If you want the players to draw the conclusion “the dragonborn will flee if the alarm is raised” you need a clue that implies that.
Have a bad guy escape once and your players will always consider the possibility of that happening.
Easier said than done in 5e. I had the notion to have a giant escape the PCs and collapse the room behind him. Turns out, a single bad roll against a monk’s stunning fist can really put paid to that idea.
Certainly, in my group the ranger rolled a crit with hunters mark and colossus slayer and he murdered the poor girl. Made the final boss fight easier though. Maybe I should have said “have the bad guy attempt to escape”.
Maybe, if the players appear to realise the target might get away hints are not necessary. If the players have no idea this might happen giving some indication is fair. They may have picked it up from the run up to the dungeon after all..
So you’d rather keep your players in a constant state of anxiety because you’re too lazy to telegraph anything and they need to anticipate every eventuality at all times?
Yeah… Sounds “fun”…
Wow… false dilemma and strawman wrapped up in one. If you can’t discuss this like reasonable adults, you won’t be allowed to discuss it all.
There are reasons you may not want to make certain things too obvious. Part of the design goal was to build on a possibility for failure that didn’t involve a tpk. Escape of the bad guy is just that. If you make it too obvious that he will escape the players will focus solely on that and it becomes a non-issue.
As Jesse pointed out here and Angry has said elsewhere, once the PCs are in the same room as the bad guy it’s basically over. If this happens because the players figured out how to catch him, good. If this happens because you told them he will run and oh yeah this is the only exit available you’re just giving it away.
How much hinting is enough and how much is too obvious depends a lot on the players. If you had a baddie escape in a previous adventure, they’ll consider it a possibility without you having to do anything (if they’re paying attention). If this is the first time they encounter this possibility the adventure becomes harder and you may need to drop a few hints.
As an example, that Gallstaff guy in the Rembrandt hideout of the 5e starter set is supposed to try and escape if his familiar warns him. Not saying this is how you should approach this, but an escape attempt is not so unprecedented that players have no reason to expect it.
“Have a bad guy escape once and your players will always consider”
…I disagree. Player tactics seem (in my experience) to almost never coincide with greater strategy and any tactics or goals of the villain need to be at least hinted at if not telegraph before they reach the dungeon- then again, maybe I’ve thrown too many curve balls at them.
In the mental space of this adventure though, the initial raid shows the villain making efforts to steal the book or macguffin or whatever (it’s been a bit since that article), that’s the first hint. A second hint might be something like evidence during the investigation or the dungeon crawl that this dragonborn isn’t from around these parts and has had other, potentially bigger, plans elsewhere in the land. Rumors from the alderman and reeve, bits of decaying banners captured from two kingdoms over, some mysterious nonfunctional artifact that dates to an ancient precivilisation in some far-off locale, kobolds in a state of preparing for travel, etc. Anything that says “the dragonborn clearly has important business elsewhere and doesn’t intend to stay for much longer in this hole in the dirt”
Fair enough, those are subtle hints and would do the job without giving too much away.
In response to this entire thread: I think this wholly depends on your party. If you have strategic party who regularly anticipates and prepares for these kinds of things, just let them see the Kobolds raising the alarm and you’re done. On the other extreme, if you have a bunch of meatheads who will never even try to anticipate this and would not have fun if you gave them anything other than pins to knock down with a bowling ball, then have Kobolds shouting to each other “IS THE DRAGONBORN WE ALL SERVER READY TO FLEE IF THIS GOES BADLY!?” from the moment they walk in. For all groups in between, moderation. The players can sneak and listen in to conversations, interrogate prisoners, and observer the routed kobolds’ actions to figure things out. As always when GMing, adjust the adventure and its difficulty to suit what your group prefers. A while back, Angry talked about GMing hats vs Designer hats. I think a lot of this conversation amounts to trying to solve a GMing problem with a Designer hat on.
While I agree you have the gist of it – and that’s a very apt summation – I’ll also point out that it is part of the designer’s job to anticipate the user’s needs and facilitate them. And you can’t assume any given GM is savvy enough to make the proper adjustments for meatheads. It’s one thing when you’re designing something for your own group of players, but another when you’re designing for any given group of players and a GM that isn’t you. And it’s good practice to always assume the latter, even if you’re doing the former, just because it improves your design abilities across the board.
That said, there’s an unaddressed assumption in some of these comments that the party MUST clearly and unambiguously KNOW what’s going to happen instead of being given the opportunity to figure it out and being screwed if they don’t. It IS okay to set up a puzzle that the players can fail to solve. That’s part of creating a challenging adventure. And the difficulty of the puzzle is determined by how easy or hard it is to figure it out based on the number and clarity of the clues. And while I know this now invites people to jump in and comment on precisely how easy or hard is the RIGHT easy or hard, there actually isn’t a right level of challenge that fits every adventure and every game. Which is why the text of a final project should empower the GM to tweak – not set, necessarily, but tweak – the difficulty for their group. So, after you write up the adventure with whatever baseline difficulty you set, you provide a clear bit of text that explains the challenge and tells the GM how the “puzzle” works and how they might make it easier or harder in the way they present the dungeon.
Signal it in the briefing. Not only is it the obvious place, but you also get to develop details of the dragonborn’s character and skill-set.
Have a hardbitten ranger with one leg say “Six months ago I trapped that damn dragonborn in a set of caves near VyrdStone. I killed his gang, but he … got away. Not before shooting a needle of Drok poison into my leg, mind you.”
“He’s a slippery one – that scaley b*stard will try to escape the mine at the first sign of trouble. Make sure you don’t let him get away!”
Just from what Angry has been saying in earlier articles, they might capture a kobold during the raid on the town they can interrogate, who can give clues. Also irrc he intends to make it obvious that the kobolds are packing up in order to leave the place, indicating that the players don’t have all the time in the world.
Thank you for paying attention.
I’d thought he mentioned the packing thing, glad it slipped into my thieving subconscious. Kind of wondering how well the morale/route/surrender system would work, even monsters that I have run when the situation feels right (because I didn’t have morale mechanics in place) tend to get shot down unless the party knows they need more information.
Do you reveal the full map to the players?
Good God, no. Encounter maps as you need them, but the map of the whole dungeon? If players want to see that then they can draw it for themselves as you describe it.
Players will weep at this injustice, but the players can just choose not to be lazy and pick up a pencil to make a map.
One thing you can do (especially with irregularly-shaped “rooms” like this which can’t really be described in terms of simple dimensions) is keep the full map behind a screen, and as the PCs explore the dungeon, sketch out the parts that they can see bit-by-bit on a sheet of tracing paper laid over the top of the map.
Or you could just make them draw their own map, like Jesse said. It all depends on how hard you want to make it for the players.
Getting players to ante in on the party cartographer job is a task these days, but it comes down to being a rigid enough GM not to give in to complaints. I am not as strict as I ought to be…
If you *have* a full map (not just a series of linked encounter rooms) then – yes.
The passages between the main rooms are a critical part of a living. immersive dungeon.
That’s where enemy reinforcements & patrols & wandering monsters move around the dungeon, and where the enemy lay traps or place dire warnings.
Those are the places where the players can hear the distant Ogre bellowing, or discover strange tracks, or find the discarded bones of a prospector, or stand and cast Divinatory spells, or wait for the thief to come back from his scouting trip, etc etc.
Leave no ‘dead’ places. Put just a little bit of dressing into each linking passage – a seam of mined-out ore, a broken pickaxe half eaten by rust, a dead centipede hung from a hook – and you get to breath life & narrative into the whole.
For those who are interested in a light “box” for tracing, you can now get a thin panel lit by LEDs that is much easier to store and cheaper than the old-style lightboxes.
I realize this is tangential to the main point of the article, but… I would view that kitchen fight not as a problem, but as a tactical challenge. Perhaps after two PCs have entered the kitchen, the centipedes swarm out and surround the one in front. How do the PCs deal with the cramped quarters? Do they retreat into the hallway so the centipedes can only attack them one at a time? Do they all try to press forward into the kitchen so that they can gang up on the centipedes? Do the ones in the hallway use ranged attacks (which might mean the centipedes gain cover from the PCs in the room)?
I would put half the centipedes inside the kitchen, and then have the rest crawl out from holes in the ceiling or floor *outside* of the kitchen once the fight starts.
“The two Centipedes utter a strange, shrill hiss. Roll Nature Lore! You realise with fear that they have released Kill Pheromones that will summon the rest of the Centipede pack to a feeding frenzy! It looks like more Centipedes will be along any minute!”
That way we get a multi-directional fight that also changes over time (some enemies turn up later). And we get a cool detail about how dumb Centipedes can arrange impromptu ambushes.
Yes, it is an interesting single encounter. It’s sort of okay as a second encounter too, but by encounter three in a single adventure this gets tedious. If Saltmarsh only has one two encounters like the giant centipedes in the kitchen, that’s cool. If they have more, it will get tedious.
This is a major reason why I tend to draw dungeons-as-dungeons rather than using realistic layouts-because realistic layouts tend to be cramped in a manner that winds up defining every encounter. The relationship between the size of the room and the type of encounter can get lost in design. I’m sure we’ve all seen modules where the giant is a room that he can’t physically enter. Hopefully the relationship between room and intended encounter is something Angry will address.
But, yeah, unexpected cramped quarters can be totally awesome for an encounter! My PCs recently ran into a bunch of ghouls in a cramped cavern tunnel. Declaring the PCs could only get a single guy to the front while the ghouls could swarm around that guy turned a minor speed bump encounter into a fun one. PCs were doing acrobatics checks to leap off each other, pulling out weapons they don’t usually use, and the like.
Plus, I was able to describe to the table how the party’s light source and poor perception rolls enabled the ghouls to get the drop on them. Which felt like a personal victory as I am trying to provide more narrative explanation tying PC behavior to consequences in game terms that the player readily grok.
The problem in this situation being that, as there are only 9 unobstructed squares in the kitchen, 4 centipedes and 5 players will not be able to move at all! There will literally be no space anywhere in the room to move to. This Isn’t a very good example of a cramped layout.
That’s assuming there *are* five PCs in the party. And assuming that the centipedes can’t occupy “obstructed” spaces by crawling up onto counters and such. And assuming that you wait until the whole party has entered the room before the centipedes attack — but why would you do that? As I described in my original post, it makes for a more interesting and less frustrating tactical situation if they attack when only one or two PCs are in the room.
It doesn’t matter what you’d assume because you’re missing the point so bad, you’re no longer in the same post code as the point. You’re not on the same continent as the point. The point is – and this isn’t a controversial point – that when you design dungeon maps as real spaces, the amount of space left in the rooms tends to be smaller and the encounter spaces less interesting then when you design encounter spaces as spaces for encounter gameplay. I chose ONE SPECIFIC EXAMPLE of a room that is TOO SMALL to contain the stated number of players for the game and the stated number of monsters. One of many examples. I can find others. In that very book. And you could too. But I didn’t think I needed to because I figured the statement “that different mapping techniques produce different kinds of maps with different strengths and weaknesses” was the sort of thing no sane person would feel the need to argue. And it’s a good lesson for anyone who wants to be good at designing maps and dungeons. Which is why people come to my website. For good design advice.
What you’re doing is now ignoring the broader point and the design lesson to defend this one poor, specific encounter from unfair criticism from mean ole angry who said it’s bad and that it should feel bad. And, even after people say, “yes, what you said is true, but Angry’s point is…” and when other people “yes, but when you look at the core assumptions of the game, it doesn’t really work that way…”, you feel like you have to dig in and defend your off-topic non-point.
So, enough. You had your say. And yes, you are technically correct in the same way that a GM can do anything with any encounter they want. But you’re wrong in terms of broader design critique. And you’re officially wasting my time because I have to read all of these comments. This thread is over. Your point is made. Your words are on record. The stenographer has noted it’s way more interesting this way to you and therefore, there’s no point in understanding the different ways to design maps and how to pick the best design method.
They also forgot how large a fighting space is and how the expanded rules on squeezing multiple creatures into a single space entirely permit cramped battlespaces. It’s a moot point belabored for lack of system mastery and, failing that, improvisational bullshittery (aka ignoring the nonreal problem via cinematic description and some handwaving) and the “miscellaneous (un)favorable modifier”. Any GM worth their salt shouldn’t even be concerned. Was the fight fun? If yes then the players probably didn’t even stop to question the potential oddity.
For once I seem to be the one who isn’t chewing the mechanics to death for shiggles.
Can thread over please mean thread over?
…did the system eat your avatar, Angry? I just realised this wasn’t one of the copycat parodyname users.
Sorry. Maybe. I probably did something wrong. I suck at this web hosting thing. And I’m really worn out lately.
I know the feeling, S.A.D. is crap. Take care.
The article isn’t so much saying it’s a problem, as it’s saying “it’s something you need to take into consideration when designing the dungeon”. It is not a “bad” room in and of itself, but it very much depends on what you want to use it for. Angry isn’t saying “don’t use it”, he’s saying “use it when it is the correct thing to do” 🙂
Just remember, the article literally tells us that using all three types of dungeon is something you *should* do. But designing a “gamey dungeon with a twist” in a “dungeon that looks like what it is” would be wrong (generally speaking)
You win the Angry GM Prize for Totally Getting It! A rare award I hand out pretty much once a never when someone recognizes that identifying the pros and cons of different tools and advising people to pick the right tool for the job isn’t the same as saying “this tool is a bad tool; shame on this tool; BOOOOOOOO tool” and then defending it by saying “you know, sometimes, that bad tool can do good things anyway.”