As promised, some post mortem thoughts on Steam Mountain - a dungeon I ran for almost a year, across sixteen sessions and three floors. A moderate effort, but none-the-less the largest dungeon I've made, and by the end one that utterly exhausted me. And because I am me and I am here, that means I gotta chost about it.
What is Steam Mountain?
Steam Mountain's my attempt at making something larger than a normal dungeon but something smaller than a megadungeon. I worked out a plan for 61 rooms spread across five floors filled with a handful of factions and several different adventuring parties. At the bottom of the mountain was a trapped flame spirit - Forneuuma, a fiery frog that the mountain had been dropped on ages past to prevent it from burning the lands around it, and whose eruptions caused steam blasts that gave the mountain it's name. At the top was a white dragon - Whitemedal, whose frigid breath could encase you in unbreakable ice that took special, expensive unguents to melt, and who held a memory-filled brooch that the adventurers need to save the frozen fairy country they are trapped in. It's where a half-dozen or so story threads that had built up over two years of almost weekly play were going to come home to roost as well.
Building Steam Mountain
I started Steam Mountain as something else all together. My expectation as I started work on it was to design a sort of crafting and jousting challenge - repair an airship using resources around Steam Mountain, trying to hide it from Whitemedal till it was repaired, then take to the skies and fight the dragon for the brooch and to save missing friends. As we wrapped-up the adventure preceding it, I pitched the idea to my players and was told "Uh... we don't want to fight a dragon. Can we steal from him instead?"
This happens around November, 2021, as I get ready to move across country and take a new job. I make the trip from sunny Florida and snowy Maine just as winter's about to hit, and start prep on Steam Mountain. I onboard onto a new job, adapt to an entirely new environment, and our group plays out the little bit of adventures I have done until I have to call a break around March, 2022, to seriously buckle down and just write.
At this point, I've made a few decisions about how I want to approach Steam Mountain:
- It should be a dungeon. One way to think of a dungeon and dungeon games is they're heists. You build a location, you set a goal (usually treasure), and the players need to find a way to get the treasure out of the location. Perfect for stealing from a dragon!
- I want it to be large. I'd built out moderately sized dungeons previously, but I wanted to see if I could push into the beginnings of megadungeon territory.
- I didn't want folks to camp out inside the dungeon. I wanted it to have a rhythm like an old dungeon crawl or a video game, entering the dungeon, pushing as far as possible, opening shortcuts, and retreating to rest at base.
- I wanted to use inventory slot encumbrance mechanics. Instead of giving items a weight and tracking how much someone could carry by strength, I wanted to try something closer to what Knave and Mausritter do - you have a set number of slots and once those are filled you can't hold any more items. We tried something like this in the adventure preceding Steam Mountain, so it wasn't too hard to get buy-in from the group there.
- I wanted to use the Event Die for random encounter checks, specifically the version found in Errant, by Ava Islam. Each exploration turn you roll a d6 and that tells you whether resources like torches are spent, the party needs to take a break, something in the dungeon's environment changes, or a random encounter occurs. When combined with limited inventory, the hope was to create a resource cycle my other adventures lacked, pushing the party to weigh whether to continue a dive or retreat and restock.
- I wanted to limit the party size. The group at this point was four players and about a dozen different NPCs of varying degrees of mechanical complexity and narrative necessity. It is a lot to run, a lot to keep up to date, and a lot to make all those NPCs feel interesting. I put a cap of taking one NPC into the dungeon at a time, and decided the rest would stay at base.
- I wanted to try writing the dungeon using bullet points instead of my more traditional method of writing descriptive room text followed by a bunch of bullet points for details. You can see what my traditional approach looks like over here.
An example of the bullet point style I ultimately settled on for writing Steam Mountain.
All this was meant to create a loop. The party would travel into the dungeon, make progress, run low on resources, retreat to base, and then decide where to go next and who to take with them. Nothing fancy, but potentially very fun.
Putting the Work In
Once we ran out of game to play, I put the campaign on hold and spun-up a side-game I hoped would be pretty simple to run in the meantime. I took some old B/X D&D adventures, set them up in Roll20, and pitched the group on playing Old-School Essentials. Three of the four players agreed, and one dropped out. Not long after, her spouse also dropped out. The side-game chugged along, though, and we cleared through B11 King's Festival and started into B7 Rahasia. All the while, I dealt with a miserably intense summer at my job, managed an increasingly unstable FFXIV free company, and chugged away at Steam Mountain.
By the tail-end of October I had two floors built-out, a third partly outlined, and the activities of a dozen NPCs sorted out. My approach is to run things once I have them half-way done, knowing if I don't get started I risk abandoning a project all together. Every player returns, I lay out the rules for the dungeon, and we kick into gear.
Running Steam Mountain
Things started off fine. Folks agreed to the rules constraints, picked an NPC to journey into the mountain with, and headed in. Anticipation is high!
Annnnnd then... the torches start going out. I rolled the Event Die and the first several rolls were "light sources deplete." This immediately frustrated one of the players. I tweaked things after that session, changing Event Die rolls to once every three turns instead of once every turn. This seemed to work better and moods improve.
Then we lost one player. Life had gotten busy and she had to drop off. No ill will had... but that's when I start wondering "Shit. Did I make the wrong adventure for this group?" A few sessions later, her spouse also leaves. Again, no ill-will, but also, I can't shake the sense that dungeon crawling isn't what she was interested in either.
Around this time, I also notice that the party just keeps taking the same NPC with them. Turns out it's the one who has a power that shorts out the entire inventory system - her very long hair is a portal into a fixed point in the Astral Plane, and a very convenient place to store items. On top of that, she's fun to run, useful in a fight, and has a very strong story hook - her twin brother's captured by the dragon at the top of the mountain. I bring it up, and the two remaining players acknowledge they keep taking this one character, Thalia... but also point out they don't feel safe taking the other available characters. One is hunted by a bounty hunter wandering the mountain, another is effectively a five-year old, and third is... well, just not that interesting.
Also around this time, I start to run out of steam. Work has caught up with me and I have less-and-less time each evening to work on the third floor. I keep making progress, but it's slow. One of the two remaining players also starts looking into a move of her own, so sessions become increasingly infrequent. Eventually, I finish the third floor, but I start to run dry on ideas.
We carry on, though, and the three of us have fun. But I can't shake the burn-out and eventually, in late summer of 2023, I ask the group if they're fine with me off-ramping us from the rest of Steam Mountain. They agree, and we also figure out how to scope back the game - we figure out what NPCs its fine to sideline and we sketch-out what adventures probably will get us to the finish line of the campaign. We play out the last floor I have finished, hit some big story points, and on Sep. 9, 2023, we wrap-up Steam Mountain!
By the time we wrap, I've spent close to two-years conceptualizing, building, and running Steam Mountain.
Lessons Learned and Take-Aways
I think the big one is I tried too many things all at once, and tried it when I did not have the free time to put in the work. Too many systems, too large a dungeon, too many other things outside of this game vying for my time and energy - all of that created a huge amount of burn-out for me and led to me asking if we could off-ramp the story.
I think if I hadn't tried too much, I could have more reasonably designed a large dungeon. I planned for 61 rooms, and I managed to create 45 of them. That's well over two-thirds of a dungeon! I am going to carry forward the organizational techniques (heavy outlining, a spreadsheet tracking each room's contents) and already am applying lessons about making maps to my D&D game.
An excerpt of the spreadsheet I made to outline Steam Mountain. I tried to track treasure, clues, encounters, and more and it ended-up helping a lot when creating maps and notes.
I learned I don't jive with the bullet point method of adventure design. Inevitably, I just created as much work for myself as with my normal writing approach, but organized it in a way that made it difficult for me to find the notes I need. I run very prep-heavy, but that's because it helps me keep the game at the table running smoothly. Bullet points didn't do that for me. I went back to my normal way of writing for my D&D game and that's gone better.
The party size limit wasn't bad on it's own. It did make navigating the space much more manageable than if the party could take all the NPCs with them. It also gave me space to seed NPCs in the dungeon for the party to find, rescue, and recruit. But as described above, it led to the party taking a handful of specific NPCs each time. By the third floor, they started to vary things up, but I feel like I didn't nail this goal.
The Event Die didn't work out at all. Changing how frequently I rolled helped, but each time a light source depleted after that first wave of torches going out, it just created a bad mood. On top of that, it created more work for me. I normally might include environmental effects and character encounters on one random encounter table. The Event Die splits these into two tables. I built random encounters for three floors, plus two small exterior biomes and (eventually) for if the party took Thalia, the girl with extraplanar hair, with them. That was a lot of work, especially since random encounter writing is some of the hardest stuff for me.
I liked the Inventory Slots encumbrance system, but it works better when you don't introduce it in the middle of a campaign. It encouraged players to use disposable resources instead of horde them, make tough choices on what to loot, find places to store things safely in the dungeon, and retreat every so often to clear-out their bags - all really cool things! But also, every time the group returned to base, the session ground to a halt as the players spent an hour shuffling equipment around a party of five or six characters. We're going to go back to our old approach of "Encumbrance doesn't really matter" after this, but I might try Inventory Slots in other campaigns.
I also stopped tracking NPCs quite as carefully. And told the players this. Our game has downtime and crafting rules, and leading-up to Steam Mountain and a little bit into it, I was having to track that for a dozen characters. I just don't any longer. We also had a discussion and have whittled down the NPC list to those the remaining players are interested in. The other characters are around, but I can let them fade into the background. Honestly, one of the biggest reliefs for me.
So yeah! That's Steam Mountain! I could write so much more - I didn't explain any meaningful backstory to the game or talk about the fun and frustrating story beats - but this is already a lot of writing. My next steps are to wrap-up some other on-going tabletop projects and then I'll puzzle out what comes next for my Beyond the Wall group. We're well over three years into this campaign and are on track to finish it after two more adventures. I may have... lost steam... with Steam Mountain, but I am still excited to play this campaign!
