This article starts with an e-mail. Which isn’t that unusual. See, I used to do this thing where I would ask people to e-mail me questions and I would do a short, weekly column where I’d answer those questions. And make fun of the person asking the question. And give them a dumb nickname. Those were good times.

Every so often, someone still e-mails me a question with “Ask Angry” in the subject, even though I haven’t actually done an Ask Angry column in f$&%ing forever. And every so often, I actually read the question that comes in. And mostly, it’s the same crap I always get asked about how to deal with a problem player – hint: you talk to the person like a f$&%ing adult about the problem, try to resolve it, and if you can’t, you ask the player to find a new group – or how I manage to type with boxing gloves on. But every so often – and the reason I haven’t answered such a question in ages is that we’re like three “every-so-oftens” in so it’s basically “never” – every so often, a question really manages to grab me. And most of the questions that really grab me are “how do I pull off this neat thing at the game table using a combination of predesigned game mechanics and my own GMing brain?” I f$&%ing love those questions.

Interestingly, I think the last time I did this s$%& basically started with “Alien: Isolation in D&D; how do I even?”

Long story short: I got this e-mail in which D. Ben Knoble asked me “how do I pull off this neat thing at the game table using a combination of predesigned game mechanics and my own GMing brain?” Not exactly in those words, but close enough.

Well, Knobleman Ben, you win! I’mma answer your question. Even though I don’t have the background whatever the f$&% Strange Things is and therefore have no clue what you’re actually trying to accomplish. But I figure that’s fair because you don’t know what Daedalian means.

Daedalian Just Means Complex, Not Reality Rewriting

D. Ben Knoble asks…

I’ve recently bought the Stranger Things-themed D&D 5e Starter Kit. I’m a vet of the edition but thought it could make an interesting thematic introduction to D&D for some friends. The module promises to be interesting, but I am concerned about one section…

 

And it just keeps f$&%ing going. Because, even though Knobby Ben somehow managed to remember how to e-mail me and properly ask a question so it gets around my numerous spam-and-death-threat filters, he ignored the most important rule of Asking Angry: GET TO THE F$&%ING POINT.

So, allow me to sum up Knobben’s e-mail.

I bought that Stranger Things box set thing. It doesn’t matter why I bought it or why I want to run it. That’s totally immaterial to my question and it would just bore the f$&% out of your readers, so I’ll skip all that explanation and get to the f$&%ing point. In the module, there’s this maze that I am going to describe in terms of an incorrect understanding of the myth of Daedalus and some insipid fantasy novel series for pre-teens that tried to ride the Harry Potter wave called Perry Johnson or some s$&%. Basically, the party has to explore this maze. But the maze ahead of the players changes if the players ever backtrack or turn around or otherwise take their eyes off of it. If they keep pressing forward, everything is cool. But if they go backward, then the parts they’ve explored change. Just like Daedalus. Except that never happened in the myth. The module wants me to do it with random dice rolls and it also makes me place random encounters around the maze, some of which are required to advance the plot. That seems like a f$&%ing chore. And also I have some vague concern about meaningful decisions and about how the whole thing feels. And my gut is also telling me…

Okay, listen Ben Knobs and Broomsticks, I’m going to cut you off there. Well, I’m going to cut off my sumamry of your e-mail. You’re not really asking a question. You’re just saying that WotC put a bunch of mechanics into a game product that seem like they’d be inefficient to use at the table and only sort of capture the spirit of what they’re trying to do. And that isn’t a f$&%ing news flash. That said, I can help.

Sort of. See, I don’t watch Stranger Things, I haven’t opened that box set that Tiny bought, Tiny is currently sleeping, and I have no f$&%ing no clue what Percival Jameson is. I do know a lot about classical mythology and the myth of Daedalus though. And I gotta say, that’s no help here. Because, look, he just built a very, very difficult maze to hide a man-eating bull monster that was the result of a king’s wife jumping into a fursuit and getting it on with the sexiest bull Poseidon could spit out of the sea, but that’s just because the king was supposed to kill the bull and send it back to Poseidon and he decided to keep it for himself because it was just that beautiful. And that’s f$&%ed up enough. It’s video games and pop culture that have turned Daedalus into a mechanical genius who made steampunk mechanical ever-shifting mazes.

But, listen, that’s okay anyway. Because I don’t like to let actual knowledge spoil fun game design. Just like I also don’t let other people’s failures of design influence mine. So, I don’t really care how WotC tried and failed to pull this off or what they really trying to do. I just care about the game problem.

And the game problem, as I understand it, is this:

Design a shifting, labyrinthine dungeon the players have to explore. As long as they push forward through the dungeon, it behaves like a normal maze. The layout of the maze between the entrance the player’s location remains fixed. But if the players ever backtrack and return to a previously explored location, the maze ahead of the player’s location changes. The players should encounter obstacles and monsters in the dungeon. And the players also have to accomplish one or more goals in the dungeon to complete their objective.

Hopefully, that’s a good analysis. Because, if it’s not, there’s no way for you to correct me.

How is it Supposed to Play

Whenever I’m trying to figure out how to handle something in the game, I like to start by figuring out how the thing should play out. How do I see the game playing out?

So, the party enters this labyrinth, right? And initially, it seems like a totally normal dungeon labyrinth. There’s a maze. There’s some monsters in some of the rooms. And somewhere in there, there’s something I need to do or find. As long as the party keeps pressing forward, it behaves like a dungeon. In fact, if the party only ever presses forward, they’d never know there was anything special about the dungeon.

Now, that’s a perfectly typical dungeon setup. In fact, that’s pretty much the default dungeon setup. And it leads to a fairly normal gameplay experience. The party just wanders the dungeon, taking left turns and right turns and whatever until they find the thing. Finding the thing is just a matter of covering ground. Canvassing the dungeon. And that’s a perfectly fine way to play. I bring that up because you, Bekn, made noises about meaningful choices beyond “choosing left and right” as if that’s an important thing. Well, sorry, it isn’t. Canvassing a dungeon isn’t driven by meaningful choices beyond the old “always turn right” technique or whatever. But that’s fine. Because it’s the challenges inside the dungeon that create interesting choices. Things like how to deal with obstacles, how to conserve resources, and when to retreat and rest.

What makes this dungeon different, though, is that canvassing it just isn’t possible. Because once you’ve hit a dead-end and backtracked to a previous intersection, the passage you just returned from is unexplored again. It’s completely different. Consequently, the party is essentially trying to explore an infinity dungeon. Sort of. And this is the first – and most important – thing to understand about this setup. If the dungeon can just generate an infinite amount of content, it’s basically unexplorable.

Consider how the dungeon “works” in the world. There’s two basic ways a rearranging dungeon can function. First is that somehow the dungeon is constantly spawning hallways and rooms. Whenever you open a door, the dungeon “assembles” a passage and a room and connections from the adventure game aether. The dungeon grows like a living thing as it’s explored. And it can always grow another hallway or room or whatever. In that dungeon, if you keep walking forever and the dungeon is never “cleared,” you will ALWAYS find another room.

Alternatively, imagine that the room is just a series of floating “rooms” in Dungeonspace. As you explore the dungeon, the dungeon shifts a room into place and joins it up with a hallway. There’s a finite number of rooms. If you keep walking and walking and the dungeon is never cleared, eventually, you’ll see the whole dungeon.

Now, imagine that somewhere in that dungeon is a goal. That’s what you’re trying to reach. In the first setup, that goal could literally be anywhere. It will randomly spawn into some randomly spawned room when the stars align and the fates demand the percentile dice come up double zeroes. If that’s the setup, there’s really not a lot you can do to hedge your bets. Every time the dungeon spawns a room, you’re basically playing Macguffin lottery. And if you are really smart and savvy, the best way to handle that is to step into a room, see if it has the Macguffin, and then step out, turn your back, count to ten, and let the dungeon spawn another room. You can’t canvass that dungeon. You can only wait for the dice to hand you your victory.

In the second setup though, the fact that there are a finite number of rooms and one of them is going to have the Macguffin – especially if it’s a specific room – it IS possible to canvass the dungeon. It IS possible to see every room if the dungeon doesn’t reset on you. Now, depending on the number of rooms, it may be really hard. But it is still possible. While a smart, savvy group will recognize that it is still possible – in theory – to just play Macguffin lottery by putting their fighter in and taking their fighter out and then turning themselves around and doing it all again, it is also at least POSSIBLE to find the Macguffin by covering as much ground as possible in a single run. And that’s what you want.

Ultimately, you want the experience to feel like the party has to keep pushing forward if they want to succeed. If they are forced to go back, it’s a disaster. And if they do have to go back, because of a dead-end, say, they want to backtrack to the nearest intersection only. No further. If they have to go all the way back to the beginning, they basically get to erase all of their progress. You don’t want them spending forever playing Macguffin lottery. You want them racing the dungeon.

Now, the way to prevent Macguffin lottery is to require progress. Make sure the Macguffin can’t spawn anywhere. Make sure it can only spawn beyond a certain “depth” or whatever. But hold that in the back of your mind for now. We’ll come back to that. We have to solve the problem of the infinity dungeon vs. the lego dungeon.

And we’re going to solve it partially with the metagame.

Metagaming for Fun and Profit

Now, many GMs consider the word “metagame” to be a very bad word. And those GMs are f$&%ing stupid. Because, first, the metagame is just an inextricable PART of every game. There is no way to remove the metagame or get away from it without totally destroying the game in some way. Just ask any player who has been told they are not allowed to use a certain plan because “they’re metagaming.” That ain’t fun. But I’ve talked about that before. The second reason why it’s stupid to treat “metagame” as a bad word is because it’s a very powerful, very useful tool. If you can bring the metagame into alignment with the game world, you can communicate a lot with game mechanics without the players ever knowing you’re telling them anything.

Case in point: when the players enter the rearranging labyrinth, assuming there are no thematic elements to communicate otherwise, they could either assume the dungeon is an infinitely spawning labyrinth OR they could assume the dungeon is an assemblage of finite rooms. Now, a good GM will communicate which way it actually works to help the players, sure. But a better GM will communicate it to both the characters AND the players.

See, the players can see what you’re doing. You may be working behind a screen, but you’re not invisible and inaudible. If every time they open a door, you start rolling a bunch of dice and flipping through tables, and then you tell them what kind of room or passage is on the other side, they are going to figure you’re generating the dungeon from a random table. And that communicates to the players – subconsciously if not consciously – that you are generating the dungeon as they explore. Which means the dungeon is infinite and spawning as they explore.

But what if you were doing something else instead? What if every time they opened a door you pulled a card off the top of a deck of cards? What would they think then? They would recognize that the dungeon is made out of a fixed number of rooms. And if they know the Macguffin is “deep” in the dungeon, they will assume it’s “deep” in the deck. You see where I’m going?

Now, if you pair the card draws with some sort of thematic flavor text that describes something happening as you draw a card, you bring the metagame feel in line with the feel of the game. For example, when the players pull the lever that opens the door, before the door slides open, they can hear clunking and metal shrieking and grinding of massive machinery. Or they can hear stone rubbing against stone and feel the shaking of the dungeon as things shift into place.

And they can also hear and feel that stuff whenever they turn their back. And they will also see you take a card off the map you’ve been building behind your screen and slide it back to a random place in the deck. Which tells them the dungeon is f$&%ing with them every time they turn around.

So, that’s the first bit of mechanics I’d use: I’d get a bunch of index cards and draw map elements onto each one that connect up like geomorphs. And I’d use that assemble the dungeon as they play. And I’d make damned sure the players could see exactly what I was doing. Well, not exactly. The map itself would be hidden. But they can see me playing solitaire with dungeon cards as they explore. And I would keep the “draw” pile visible, by the way. Because as the players watch the deck dwindle and know they are covering ground, they feel like they are making progress deeper into the dungeon. Which is exactly what the characters would feel inside the labyrinth. They’d have a sense of going deeper, of time passing, of fatigue setting in, of tension rising, and of dread at the thought of turning back now when they’ve come so far.

Making Progress and Finding the Macguffin

Now, if that’s all you do, you’ll have a pretty fun and creepy shifting dungeon that the players will know is f$&%ing with them. Mission accomplished. But there’s still a little bit of a problem with it. And the problem is that, while the players will know they are making “progress” by watching the cards get pulled, there’s no actual sense of progress built into the adventure. As it stands right now, the adventure is just “wander deep enough to find the Macguffin.” Beyond that, it’s just a random dungeon. All that shifting stuff doesn’t really change anything unless the party has to retreat. Moreover, there’s the issue of what it means that the Macguffin is actually “deep” in the dungeon. How do encounters even work in the dungeon? In fact, I guess we should address that before we go any further.

How Do Encounters Even Work in the Dungeon

Let’s talk about encounters. Without encounters, this is just about wandering into a maze long enough to win the Macguffin lottery. Something has to happen in the dungeon. Partially because otherwise the game is boring. And partially because you really want the party to want to retreat.

See, if the party is never forced to backtrack, the dungeon is just a normal dungeon. It never resets. It never does anything crazy. Beyond that, though, if the party never WANTS to retreat, they never feel conflicted about it. The thing is, what makes this dungeon special is that it really demands that the players get as far as they can all in one go. In a very visible, very obvious way, they can see that if they retreat, the dungeon will completely reset. And for that reason, you want them to WANT to retreat. You want them to get tired out. You want them to be afraid to push forward, knowing they will invalidate a lot of their progress.

To that end, the first room of the dungeon should be a safe room. Or else it should be safe to rest right outside the dungeon’s entrance. The party should be allowed – invited, even – to rest right outside the door to the labyrinth. If they choose to rest because they are overextended, they are paying a hefty price and they know it. They are going to reset the dungeon.

Which means, first of all, that you need to have encounters in the dungeon that will drain the party’s resources. There need to be monsters in the dungeon. And traps. And all the normal dungeon stuff. But how to handle encounters? Are the encounters “fixed” or are they random?

By fixed encounters, I mean that the encounters are keyed to the rooms they occur in. When you draw Room #3: The Altar of Demogorgon, the undead high priest of Demogorgon comes with it. And so does the high priest’s treasure chest. In that setup, when you make up your little cards with the map elements on them, you put an encounter right on the card. And also when the players find that room again after the dungeon resets, it’s got a dead undead high priest and an empty treasure chest.

Now, that really helps drive home the point that the dungeon is just an assemblage of finite rooms being moved into position as the party explores. And it helps the players feel as if they are actually having an impact on the dungeon. But…

The problem with that setup is that, eventually, the party can clear the dungeon. Which means every time the party retreats and resets the dungeon, they know the new dungeon is easier. So, they can just chip away at the dungeon. And the fact that it resets is trivial. Because, eventually, the dungeon will be empty. And it also means there’s never any danger of resting in the dungeon. There’s no reason to retreat. It’s safe to sleep wherever you are.

The alternative is, of course, to have a random table of encounters. After you pull a card and spawn a room, you roll randomly on a table to see what’s in there. That ensures the dungeon is always dangerous because it keeps respawning monsters and if the players want to make progress, they have to keep pushing. But it also undermines the idea that the dungeon is finite and it’s just being rearranged from existing elements.

So, my suggestion is to do both. Have a certain number of rooms that have specific encounters on them. Things like the undead high priest in the altar chamber or the riddle fountain that will heal or poison the characters or whatever. And whenever the players make changes in a room, make a note on the card so the room remembers that the next time it spawns. Once those encounters are cleared, they stay cleared.

But you also want to have some wandering encounters: vermin, undead, slimes, clockwork horrors, whatever. Whenever the players spawn a room that does not have an encounter assigned or spawn a room with an encounter that’s been cleared, something has wandered in. A carrion crawler is eating the corpse of the undead priest, for example. Or giant spiders are passing through on the hunt. These encounters shouldn’t be as big or complex as the fixed encounters and, for full effect, I wouldn’t give full XP or treasure for them. They are the punishment for resetting the dungeon. They ensure the dungeon stays dangerous even as it cleaned out. But not AS dangerous. A couple of giant spiders are easier to handle than the mummy of Thraxx the Malignant, but they still have to be dealt with. If the party does clear every fixed encounter, the dungeon will be EASIER, but it won’t be EASY.

Now, that takes care of respawning. But what about the party resting. Or just dilly-dallying and wasting time. Just use your wandering encounter table for that too. If the party sits on their asses in one room for ten minutes, there’s a chance something wanders in. And ten minutes after that, there’s a chance again. That’s the easy way. If you’re really into doing some extra work, you can actually build a separate table of things the “dungeon” will do if the party sits with their thumbs up their a$&es too long. Maybe, the dungeon moves the room they are in somewhere else. Or maybe there’s some sort of nemesis character, like a clockwork labyrinth guardian, that catches up with the PCs if they don’t get on with it. I do recommend you do SOMETHING to keep the players moving.

So, that’s encounters. Now, back to progress.

Getting Back to Moving Forward

As I was saying before I so rudely interrupted myself, right now, there’s not really any sense of progress built into your shifting labyrinth beyond the “get deep enough for the Macguffin to appear.” And you really don’t even have any way to measure the idea of “deep enough.” Now, that might be okay if you just want this dungeon to fill a session or a half session or whatever. But if you want to really spend some TIME on this dungeon and let it fill a good two sessions and really show itself off as the set-piece to your whole adventure, you want to build in the idea of progression. And you also really want to build in some ways for the players to use or get screwed by the shifting mechanic.

Firstly, there’s the issue of dealing with a card draw mechanic with putting cards back in the deck and maybe reshuffling and how you can make sure the Macguffin will show up randomly, but only at a particular depth. I mean, one easy way to handle it is to deal the deck of map cards out into three equal piles, shuffle each separately, and then stack them up so the pile with the Macguffin is on the bottom. In point of fact, that’s an easy way to make sure specific encounters happen at specific depths. Like to ensure there’s a locked door that appears before the room where the key to the door is hidden. Hold that thought, though. I’ll come back to it.

But that’s still just metagame progress. And you really want a way to convey depth in the game world. Which is why I’d consider – depending on how big I want the dungeon to be – building two or three dungeon levels. Each level would have its own deck of cards. And somewhere on each level is a ladder that leads down to the next level. And each level has its own deck. And, again, the players can see the draw piles. Well, eventually they can. We’ll get back to that too. They can then assume the Macguffin is on the lowest level of the dungeon.

Now, even if they get lucky and find the ladder on one level right away and don’t have to explore any further, they still have to progress through another level and locate the Macguffin, so it’s less likely a lucky card draw will end the dungeon in one room. You could also bury the ladder rooms in the bottom half of each deck – shuffling the two halves of each deck separately and then stacking them up – if you feel the need. I’m not sure that’s necessary though. There is such a thing as going too far.

The point is, by having a couple of levels – be they ladders between different floors or even just prominent doors that lead to an obviously different section of the labyrinth – you build a sense of progress into the dungeon that both the characters and the players can feel. When they cross the threshold into a new level and you start pulling from a new deck, they know they’ve won a minor victory on the road to winning the adventure.

And I think that’s good enough to make a fun little dungeon. Except…

Letting the Dungeon F$&% with the Players and Vice Versa

Even if you do everything I said above – and you should, because it’d be a fun dungeon experience that would really feel like a shifting labyrinth – you’re still not really using the dungeon shuffling mechanic to its fullest potential. The only time the dungeon shuffling mechanic affects the players is when they choose to back out and take a nap. That’s okay, but it’s not great. You could do more.

For example, you should consider building some dead-end rooms into your dungeon map deck that forces the players to retreat to the last intersection. They won’t want to, because with the need for progress and the random encounters will make sure they feel like they are losing ground if they retreat, but they have no choice.

On top of that, you could also build rooms that, by their nature, for the players to explore and then backtrack. That’s a normal thing that has to be done sometimes in dungeons, but in this dungeon, backtracking comes with a lot of baggage. For example, you could have a locked door that requires a key. The locked door is fixed on a card. And once the locked door spawns, there’s a random chance that each time the party spawns a room, the key is hidden in that room. That ensures the key will never spawn before they find the door. You could also have a room with bars or a gate that divides it. The party can see something on the other side of the gate they want or another exit, but they can’t get to it except by coming into the room from another entrance. Or the switch that opens the gate might be accessible – visibly – from another entrance to the room. Obviously, once a door key is found or a gate has been opened or unlocked, it’ll never impede passage again. The party won’t have to do that s$&% every time. Just the first time. Even if the dungeon resets, gates stay unlocked and the party still has the keys they found.

Now, once you’ve introduced s$&% like that, you also introduce the possibility of spawning an impossible configuration. The party might be blocked from progress by a locked door down one passage and a dead-end down another. And in a normal dungeon setup, that’d be a disaster. But in THIS setup, there’s no problem. The party just has to backtrack to despawn the offending rooms and then explore again to see some new rooms.

That MIGHT sound like the Macguffin lottery technique I mentioned earlier. And, yes, it is something like that. And it MIGHT also sound like it would frustrate players to have to pull back because of an impossible dungeon. But in this case, it is not only okay, it’s actually good. First, it’s okay because the rooms the party has been through remember what they’ve done. Obstacles stay bypassed, traps stay disarmed, and fixed monster encounters stay dead. So, they aren’t erasing all their progress. Especially if they only retreat a couple of rooms. Second, though, it’s good because the players are making a conscious choice to use what they know about the dungeon to finagle with the layout. They can’t control the rooms that come up, but they can purposely despawn rooms that present obstacles they can’t overcome yet. Which means they get to feel like they are cheating. And as long as you use tricks like dividing the dungeon into multiple levels with different decks and you bury ladders and Macguffins in the second half of every deck, they can’t cheat enough to break the game. Just enough to feel smart.

If you’re feeling really creative, you can even include rooms that play with the shifting mechanic more or allow them to bypass it in limited ways. You can include safe rooms on lower levels where they can retreat to go to sleep without having to go back through all the previous levels. You can include some kind of “locking” room that allows them to backtrack for a certain amount of time without resetting the layout. Or rooms that shift and rotate when entered and swap with other rooms on the “map.”

Since the rooms are just cards, it’s easy enough to reshape the dungeon on the fly. Assuming you have enough room behind your screen and you make your cards small enough. Obviously, there might be a little logistical issue there you have to get around.

If you do all – or most – of that, you’ve got yourself a damned good shifting dungeon with a sense of progress that screws with the players and lets the players screw back. As long as you run it right.

Revealing this S$&% to Your Players

It’s still possible for all of this brilliance to totally flop in the execution. So, let me wrap up this overly long answer to a question I’m not qualified to address by talking about tutorializing and showmanship. As I noted above – or at least STRONGLY implied – it’s important for the players to have a basic understanding of what the hell is happening in the adventure. That’s partly so that everything makes sense and partly so they understand how their decisions might screw them over and partly so that they can try to come up with clever plans to sabotage the whole damned thing. For example, they might wonder what happens if they backtrack backwards. That is, what if they walk backwards and keep their eyes on the dungeon as they retreat. Does the dungeon shift when they backtrack or just when they take their eyes off of it? And for that reason, I suggest you design your map so that lines of sight are broken whenever rooms have to spawn or despawn. Say, exits are doors that open or close with big levers. And you might also want to spend some time thinking about what happens when the party splits up. They might try to “hold” the dungeon that way. Which is precisely what you want. When the players come up with a good plan to sabotage a major encounter or set piece in a way that makes logical sense, it tells you you’ve done a good job. And also, it should totally have a reasonable chance to succeed.

That said, there is also something to be said for surprising the players. There’s a lot to be said, in fact. A lot of emotional engagement in a game comes not from tension, but from sudden changes in tension. And there’s also a lot to be said for keeping secrets. If you’re too clear and too transparent, you rob the players of the chance to figure s$&% out for themselves. I often rant about GMs being too f$&%ing secretive and screwing their players with their lack of clarity and consistency. But it cuts the other way. It’s possible to be so transparent that you keep the players from ever having flashes of inspiration or realization or understanding or dread. Eureka moments make players feel smart. And “oh s$&%” moments make them feel imperiled. Those are good things.

The point is, you want the players to understand what is happening conceptually, but you don’t want to spell it out. And you don’t want them to know anything until the moment they absolutely have to.

But, when it comes to a major, adventure-defining mechanic, you want them to absolutely have to know something as soon as possible. Which means you want to make them engage with the mechanic as early as possible in the adventure in a relatively safe way.

For example, when the players start exploring the shifting labyrinth, they shouldn’t have any clue anything is amiss. Even if they’ve heard rumors of how the dungeon behaves, they shouldn’t get any indications from you that the dungeon is in any way weird. Don’t have any cards visible. Don’t say anything. Just let them start exploring.

Except that you also want to have prepared the initial layout in advance without any randomness. And you want to prepare it so the players start in a room with two ways out. Whichever way they go, they should have one fixed encounter and then they should run into a dead end and get forced to go back to the start. Along the way, they should hear the rumbling and changing or whatever to indicate something strange is happening. If that makes them turn around and discover the rooms they left behind have shifted, good. If not, when they return to the entrance and go the other way, they should encounter a dead end again and be forced to return to the start. And again, they should hear the rumbling and changing. Only let them start making progress in the dungeon once they have actually had to see that the dungeon is different after they backtrack.

And only when you have to spawn a room in a place where they previously explored a room do you take out your deck of cards and dramatically place it in full view outside your GM screen and then dramatically draw the top card and place it down behind your GM screen and start describing.

And if you do that, they will go… “oh? Oh. OHHHH! Oh s$&%!”

And that’s how you know you did it right. And you also haven’t had to stop the game to explain a single mechanic. Nor should you. If they want to know more or they aren’t getting it, let them figure it out through experimentation and experience. Explain nothing. But let them see what you’re doing. When they DO finally get it, they will feel smart. Then scared.

Oh? Oh. OHHHH! Oh s$&%!!!

Now, the reason to include at least one fixed encounter immediately is so that when that card comes up later on during their explanation – and if you really want to do this right, stack the deck so the same fixed encounter comes up pretty quickly – when that card comes up again, they will see how the room “remembered” what they did. The monsters will still be dead. The treasure will still be looted. And they will hopefully understand the ramifications.

But you’re not done yet. If you’re taking all of my advice and doing two or three levels, that doesn’t mean you should lay out all the decks of cards right away. Nope. Wait until the party gets to the ladder down to the next level and only after they climb down and have to spawn a room at the intersection below the ladder do you take out the second deck of cards and lay it out beside the first and dramatically draw a card. And, again, say nothing about what you’re doing.

Tutorializing, showmanship, and telling rather than showing. THAT is how you metagame.

Of course, I do have to recommend that you don’t take any of my advice. Don’t use the cards. Don’t find ways to use the dungeon-shifting mechanic against the players. Don’t let them try to sabotage it. And don’t use theatrics to teach the players by making them experience the mechanic firsthand. Don’t do any of that. Because none of it is true to the spirit of the module you’re running.

Look, Mike Wheeler was a f$&%ing 12-year-old running a game he wrote in his composition notebook during studyhall or some s$&% like that. And it was 1983. This level of game-design sophistication would have probably been beyond that kid. This sort of at-the-table procedural generation combined with metagame theatrics really didn’t exist in TTRPGs at the time. But you know what did exist? Appendix A of the Advanced Dungeons & Dragons Dungeon Masters Guide: Random Dungeon Generation. And he would have been rolling his a$& off on that thing to pull this off.

Okay. Fine. Maybe I’ve seen one episode. You got me.