September 15th, 2024
Let me tell you about how I used to understand and run dungeons, the problems I faced and how I've come to run them now. It’s a journey from naivety to misery to joy.
My entry point into GMing was D&D 4th Edition. If you know anything about it, you may be able to predict what my sessions used to look like.
4th edition has an obsession with balanced, tactical combat encounters. It plays on a grid with all distances described in terms of grid squares, and all other mechanics similarly codified into quantised chunks like "healing surge", "close burst", "move action" and "epic tier".
This isn't necessarily a bad thing if what you are looking for is a fantasy tactics game. My friend group are all gamers with, at the time, a sprinkle of tabletop RPG experience. Our preconceptions for RPGs came largely from the video game space, and so to our eyes, 4E's mechanics seemed unsurprising. This was true most of all for myself, and I was the GM.
So, picture my nubile GM mind wading into the pages of a 4E adventure book and preparing the first session for my friends. What greeted me was rigid structure. Just as every mechanic was quantised, so too were dungeons. A dungeon was a collection of encounters: discrete chunks of gameplay, each with a tactical map of clearly designated zones of "difficult terrain", pre-defined starting positions for combatants and a script for their tactics.
It was like a screen from Bauldur's Gate. The PCs hit a trigger that begins the encounter, and are now very firmly in combat. Roll for initiative. The NPC script begins to execute. Players go into kill mode.
What our group learned from this structure is a routine. We equated TTRPGs with this dynamic. You wander around looking for triggers, enter encapsulated, balanced combats that you solve with options on your character sheet, heal up, rinse and then repeat. Eventually, you hit all the triggers, and the adventure is over. Level up!
What I learned from this is a little more specific. I learned that an adventure, and notably a dungeon, is a static, scripted experience. This is what led me into trouble.
I don't imagine any of you have ever tried to wrangle a lively cat into a bag. I certainly haven't, because I value my well-being. But I can imagine what it might be like because I have run 4th edition adventures for crafty gamers.
Here's the problem: you arrive at the session clutching your bundle of scripted encounters and sit down at the table with your friends. Your friends are awesome, and they are either poring over the player's handbook, drooling over all the cool shit their characters will be able to do when they level up, or (God willing) they're dreaming of all the cool directions the story could wander once the game kicks off.
The problem is the hunger. Your players are hungry for progress. What stands between them and progress? Your encounters. Your slow, tactical, rigid encounters. Remember what I said about 4E being fine if what you want is a fantasy tactics game? Here's the rub: someone who wants tactics is someone who thinks tactically. They will jump into the encounter's play space, for sure, applying their tactical thinking with delight. But they're hungry.
Soon, my players began to cast their tactical eyes to the horizon, beyond the walls of the encounter. They thought, "the most efficient move here is to not deal with this combat at all, is there a way around?". This is like if you had drugged a cat and snuck it into a bag. It wakes up and eventually realises it's confined and it wants out. It starts clawing and prodding and testing for weaknesses. You'd better hope your bag is up to the task of holding it in.
My players started—heaven forbid—to roleplay.
"Let's fall back and try a different route."
"What if we just ask the mayor to help us?"
"Let's recruit the orcs to clear the infested halls!"
And there was little old me, with naught but my simple understanding of the game, my list of encounters and my scripts to follow. I didn't have any encounters for those things. Some of those aren't event combats. Who the hell is the mayor?! I didn't have anything for that.
But I tried. I tap-danced and sweated and fumbled my way to the end of the session and was left exhausted. I’d just performed the genesis of a world in my brain on the fly, based on nothing. I'd never done that before.
Let's make sure that doesn't happen next time. I'll just need to take into account special cases in the script! If the players fall back, there's a pursuit encounter. If the players go to the mayor, there's a negotiation encounter. If the players...
I'm sure you can see where this is leading. Every session, my pile of encounters got heavier. I looked a little more tired, and my voice was a little weary. As the players described their actions, I sat rigid, praying I'd anticipated their course. Many times, I did, but sometimes I did not. Next time, I would...
Here is the final and most insidious consequence of the tactical table. The GM is also partial to tactical thinking. And what was I hungry for? Order. Safety. Adherence to the script. So, what tactics did I employ to ensure that? What material did I present to ensure that goal?
I began to shape the things players experienced to limit the possibility space.
“You're fugitives now.”
“That city is sealed off to outsiders.”
“The enemy is a cult impossible to be reasoned with.”
You get the idea. These are all ways of saying "no" to the players. Don’t get me wrong, “no” has a place in the GM’s toolkit. “No” helps define the theme and boundaries of your campaign and it can push PCs out of their routine if things are getting stale, but ultimately “no” should be the exception. How often does someone need to say “no” to you before you start feeling frustrated and act out? Our sessions began to feel like an exercise in answering that question.
Not only were the sessions becoming more strained, but the time between them was growing. As my players fought more and more frantically to get out of the bag, I needed more and more measures to keep them in. Those measures took time to deliberate. I needed to sit there and agonise over every move, like some deranged chess master playing by mail.
Eventually we reached a head. The sessions were so conceited I could no longer look on them with fondness. They were abominations, far removed from the light-hearted fun that marked the beginning of the campaign. I was exhausted, crushed under my now towering stack of contingent encounters swaying menacingly under the dining room light, threatening to collapse on us all. I ended the campaign, not with a bang, but with an exasperated groan.
Where did I go wrong? I was just following the examples in the book! What did Einstein say? You can't solve problems with the same thinking that created them. I had to change my thinking.
Now that I've thoroughly bludgeoned you over the head with my past woes, let's turn to solutions.
I want to return quickly to what I understood to be the structure of a dungeon:
A dungeon is a collection of isolated encounters with scripted combatants.
If this represents the old thinking, what is the new thinking? For this we need to think more in terms of fundamentals. Use common sense. What is a dungeon?
A dungeon is a location in a fictional world full of treasure and monsters.
So, what do the notes for that look like? Let's consider the GM’s window into the dungeon: the key – those pages in the adventure book describing the areas. In 4E, they are encounters, and we have seen how they lead to misery. So the solution is simple: just key the areas with what's in them. It's not an encounter; there's no script. Let things play out more fluidly. Done! Right?
This is good. This is already vastly better than before. Indeed, this is how any other edition of D&D structures its dungeons. This is how all the official 5E adventures I've read key their locations. But I want to tell you a little secret. You and your players are still in the bag. Why? Because it's still static. Your dungeon state is frozen in time.
What if the players leave and come back? Do you present the same room? What if they set the whole place on fire? What if, indeed, they recruit the orcs and send them into the dungeon to clear it out? Where's that in your dungeon key?
Here, we arrive at the catalyst for my big perspective change: the key is not the dungeon. Just as a map is not the territory, the key is not the dungeon. The key is an incomplete description of one state of the dungeon, perhaps its initial state when the PCs first tread into it. Critically, the state, and thus the key, is in flux.
As the players affect the dungeon, so too do they affect the key. Things change in reaction to the sessions. This may sound obvious to you, but to my 4th-edition-trained brain, this was a revelation. They set it on fire? The denizens scatter to non-flammable areas and fight for new territory. Flammable treasures are destroyed. The entire place fills with smoke. Once tactical players realise they can affect things on this scale, a wide smile breaks across their faces and their hands rub together with glee.
Here is the path to liberation: let go of the static key. Give yourself permission to expand it, change it, ignore it, even. Your table will relish the freedom and you will never look back.
By divorcing yourself from the static dungeon key, you've unlocked the joy of player scheming. The cats are out of the bag and are happily pawing at strings, springing into the air and doing whatever else cats love to do (perhaps dying out of curiosity or having nine lives).
But what if I told you it gets even better? If the key represents a potential initial state of the dungeon, we may ask, "Why that state?" If the PC’s actions can change the key, we can imagine some previous characters and their actions leading to the state described in the key in the book.
In fact, we could jump to any state of the dungeon and apply any group of actors to derive a new dungeon state, with its corresponding key.
Now, my tactical GM brain finally has a good problem to sink into. Whenever something happens in a session that requires me to update the key for next time, I turn to this. What if I invent a group of actors, send them into the dungeon, simulate their expedition, and note how they update the key?
Let's take the example of the PCs recruiting the orcs to clear the dungeon for them. I can just take the stats of the orcs and take them room by room through the dungeon and roleplay their decisions. They encounter some goblins. How does that play out? What treasure do the orcs take? What traps do they trigger? Boom! You've got a new dungeon state. I call this trashing the dungeon.
Furthermore, the denizens currently in the dungeon are themselves such a group of actors. What are they doing in between the PC’s visits? How are they affecting the dungeon? Are they fighting each other? Building new traps? Recruiting allies? This way of thinking becomes an extremely versatile tool for dynamically responding to the PC’s actions.
If my previous way of prepping was an endless list of "no’s” I was saying to my players, this kind of prep is a very resounding "YES". But here's the other awesome thing about this approach: it's fun. It's incredibly creatively stimulating and surprising. You discover the new dungeon state. You're not some laboured chess master playing by mail, you're Michael Jackson eating popcorn in the movie theatre.
And when the session at the table kicks off, you're still Michael Jackson! That's right, instead of sitting there terrified your encounters are going to fall apart, you're waiting eagerly to see what state the players are going to push the world in. The line between prep and gameplay has all but evaporated to reveal one beautiful continuous simulation of the secondary world.
Let me end my long-winded diatribe here. If you take away nothing else from this post, take this: 4th edition sucks. No, wait, this: treat your NPCs like PCs. Run sessions for them. That's your "prep". Watch how it turns your game into a simulated world. Watch how it breathes life into everything. Trash your dungeons and enjoy the popcorn.