What makes Felltower hang together?

During the BSing on Friday night during our S&W game, we talked about megadungeons that hang together. That got me thinking, what makes Felltower hang together?

It certainly does, if the proof is by longevity. We've been playing our DF game since September 2011, and in the megadungeon I wrote since April 2012.

We've had 42 sessions in Felltower, which Douglas Adams told us is a pretty significant number. Despite that, there is still stuff to do.

Why does it work?

Here is how it looks from the GM's side of the screen.

It is Cohesive. Even the really obscure weirdness and odd bits strewn through the dungeon are there for some reason. The layers on top of each other aren't just thrown there. Well, they are - but I threw them there with a vision.

Back in the old days putting down things for challenge, not with a plan ahead, probably flew a lot more than it does now. My players aren't coming into a game without an experience of fantasy role-playing, and we can all get "random dungeon to loot" by firing up any number of video games. It all has to feel like it will ultimately make some kind of sense, if you can just get to the right angle to look at it.

What this means is that all of it - the cone-hatted cultists, the orcs, the stomping demon lord, the fortress entrance of the dungeon, the mix of level layouts and construction, etc. - ultimately it all has some meaning. Nothing, even if placed by random rolls or encountered by wandering monster rolls, is really random and meaningless.

I often make rumors up just as I go. There isn't a master list of 100s of them. But they fit what I think is ahead and what I think people would think about what's ahead. They feed information and feed hooks to the players.

So even if the PCs can't figure out the why for everything, they can for some things. From there it's enough that they know there is a why? Even if that why is "this is what it looks like after adventures do something that made sense at the time and then left them remnants" it's there.

Corollary: It Has Some Game Mechanical Sense. Joe the Lawyer was complaining about something in a published adventure that basically threw him right off his game - an encounter that's basically a bit of cute nonsense but would take a variant version of a spell castable only by a ridiculously powerful wizard, which is then used for a totally trivial purpose. I did my best to avoid the error of putting in weirdness and cuteness that, according to our rules-set, doesn't make a lot of sense to do. That helps it hang together, too, because the rules support the dungeon's experience as encountered as well as as-played. Encounters might show extreme magic being used for small problems, but in ways that have consistent game-mechanical sense and which directly tie into further revelations about the dungeon. It's not just a red herring that takes breaking the game rules as known to the players to happen. So the rules feed the cohesion.

It's Big Enough. The megadungeon is big enough that there is a lot to do. More than the players can ever do, but anything they want to do is "deep" enough that they can get some enjoyment out of it. But it's also not so big you can never finish any particular task or goal. The orcs aren't endless. The draugr have limited numbers. There are only so many statue rooms and missing statue heads and unopenable doors. There are just so many pools. There are limitless adventure possibilities but the adventures themselves are not endless.

It's a bottomless pit of discrete, potentially solvable puzzles and completable challenges. Like the hobgoblins - remember them? Gone. The demon-apes? Banished. The lizard men? Temple smashed and the lizardmen routed. Things come, entertain, and then are ended through PC action. There are dozens and dozens more things to do. The dungeon won't ever end. But parts of it will pass.

It's Varied. There is a lot of variety in my dungeon. Not a huge variety, but a large variety.

A too-large variety can be bad, because you don't get a sense of cohesiveness. It can be tiresome. A too-small variety is bad - you just feel like you're doing the same thing over and over.

I think I got this right. There is enough variety that no one can really relax and say, "It's all orcs" or "it's all monster-types" or "it's all undead barbarians." It's all of them, but with enough regularity in the encounters to reward their knowledge, experience, and tactics.

There are fights, puzzles, tricks, traps, and things to explore. And all of those vary. Trends appear (cohesion) but encounters aren't cookie cutter replications of each other. Players bored of orcs know they have to work hard if they want the game to only be fighting orcs, not that they need to work hard to find some variety.

Problems are Player-Active. Or to put it another way, There Are No Cutscenes. The players don't walk in and watch NPCs, or illusions, or whatever, do stuff. They do stuff.

This is not to say I don't use in media res or have things that do stuff on their own. There have been paintings that change over time, traps and statues that seemingly activate or replicate themselves, and the occasional ghost going through the motions (although I think they bypassed that.)

But it's sparing. It's all based on the players doing things. You don't go from room to room looking at the paintings or looking at the history or watching the NPCs do something. In fact, most of the "show the history" or "read history at the players" bits in my dungeon actually require the players to go get that information. Magic, research from clues, examining paintings, setting up shop to watch for something, talking to the ghosts - if you want to watch NPCs do stuff, you have to actively choose to do so. You'll be rewarded, perhaps, if you do, but cutscenes don't come in and take over your controller and make you watch just as the fighting and fun got good.

I keep such things short and punchy, too, which is what I learned from Borderlands 2. A 10-second cutscene that screams "This woman is worth talking to!" or "This fight is going to be awesome!" beats a 2-minute one showing you stuff you need to know. Especially if it leads to a reason to seek out knowing more. Everyone likes to read the book they chose, not the one chosen and handed to them. I try to make all of my dungeon encounters demand action from the players and then reward that (the players - the PCs sometimes suffer.)

I Wrote It. This is a big part. It's a dungeon that is intimately known to me. I know what I wrote, where it is, often when I wrote it, and why. I know what it grew from and where it is going to. The whole place is mine from end to end.

This means whatever sense I'm trying to convey isn't a question of author -> GM -> players but author/GM -> players. We've removed one level of the telephone game, and one level of confusion. All that stands between my vision and their seeing is my ability to summon the right vocabulary to convey it to them.*

Knowing it intimately means I can ensure isn't a whiff of "I don't get this part either" or mis-representing something to the players.



From the GM's side of the screen, I think this is why we've done 42 sessions in our temporary, let's-play-for-a-while DF game's megadungeon. My players probably have more insights, and maybe they'll chime in (or maybe not.) But I must be doing something right. I think what I outlined about is why it's hanging together so well.



* Not a small issue. One thing I've learned teaching, as well as learning from teachers, is that you need to find the words that convey the most meaning to the student. The same sentences said to different people will produce a different effect. I can still remember the day, after many, many class hours, when my MMA coach said exactly the right thing in the right way to turn a particular from something I couldn't do except by luck into my A-move. Settling on a proper vocabulary for describing and conveying dungeons isn't easy. I'm not even done doing it. But it's critical, and deserves some thought and time.