On Dungeon Design and The Big Room

I am a regular reader of Gnome Stew.  They generally give good GM advice over there; I've found a number of tips and ideas from their regularly-updated columns.

That being said, something about today's "Playing in the Big Room" article rubed me the wrong way.  The central argument of the piece is that you should start your dungeon design with The Big Room and work your way outward from there.  The Big Room is, literally, a Big Room -- the large room at the center of the dungeon.  Here, larger monsters can be placed, mounds of treasure can be stashed, and The Big Fight can happen.  Here, you can put that crazy monster you've always wanted to place in a game.  You should also dress the room up with cool terrain, NPC's, and atmosphere.

This sounds great, right?  What's my problem with this?

My problem isn't with the order one puts the dungeon rooms on a page.  I don't think that matters.  Nor is it with the desire to work in cool ideas, interesting terrain, or fun encounters.  My issue is with the assumption, explicit early in the article, that the dungeon should be designed for the players.

"There should be something to challenge each party member’s specialty ability or skill set, as well as something to amaze, to amuse, to frighten and  perhaps something it would be best to evade or avoid."

The Big Room then becomes:

"the centerpiece of the dungeon level. It often is where the PCs encounter the big bad evil dude, dudette or monster. It can be — but not necessarily is — the climatic point of the adventure. Which means, hopefully, the Big Reward is somewhere nearby." (typos in original)

No big room here.

I don't think dungeons need a centerpiece.  Nor do they need a Big Bad.  They do need things to fight.  And they do need things that are best to avoid, given the party's condition at any given time.  And I hope they have amazing, amusing, and frightening things.  I am also all for cramming fun and interesting things in your dungeon in whatever way you want.  I just think that should be done independently of any concerns about an adventuring party.

There are two reasons for this.  First, it's not as if whomever "actually designed" the dungeon in your fantasy world did so for the sake of murder-hobos out for loot (unless they did, which is cool).  They did it to keep prisoners in, or to keep the goblins out, or as a place where they could store their mad magical creations before setting them loose on the outside world.  Or maybe it's all just slightly acidic water eroding soft limestone.  This is not a plea for hyper-realism.  We're talking about mega-dungeons, after all.  It's simply to point out that, in making a dungeon, a logical approach to take is to assume the perspective of whomever would actually make the dungeon.  Whomever that is is not making the dungeon as a scaling series of challenges for an approproiate level party.

Secondly, I think assuming party composition in dungeon design limits creativity.  Now, instead of thinking about what's cool, scary, or interesting, one is worrying about having enough traps to challenge the rogue while not hamstringing the fighter, while creating enough big open spaces so the evoker can actually use all those spells he's prepared.  "Do what thou wilt" should be the mantra of dungeon design.  Maybe you don't need or want a big room.  Maybe there is no big bad evil guy.  I actually think that's what the Gnome Stew article is getting at with it's advice about The Big Room and the sharks with laser beams moment.  Sharks with laser beams don't care about party composition.  I just think that's contravened by this earlier assumption about overall dungeon design.  Designing with the party in mind contravenes the idea of tactical infinity, insofar as bulding challenges specific to certain party specilizations often limits solutions to those challenges to the invocation of those specializations.

There's also the very real possibility that when you're designing a dungeon, you don't have a party to design for.  That is, you're not making the dungeon as part of some exisiting set of adventures, you're making the dungeon because you have some free time and it's fun to do.

Of course, you could roll on some tables and randomly generate all of it, in which case this is all moot.

My point here isn't really that The Big Room idea is a bad one, it's just that the article somewhat contradicts itself by 1)encouraging gonzo ideas in construction of the big room while 2)beginning with the assumption that design needs to proceed from party composition.