Megadungeons, epic character deathtraps that have made or broken a thousand RPG campaigns, are staging a comeback. While it's been upwards of 25-30 years since we first crawled into Castle Greyhawk and Undermountain, the Oughts saw the release of Return to the Temple of Elemental Evil, World's Largest Dungeon, Castle Whiterock, Maure Castle and now Monte Cook's Dungeon-A-Day.
But is this a good thing?
Megadungeons played a huge role in the early days of my World of Greyhawk campaign. The Blackrazor Guild cut their teeth (and paid off their debts) adventuring through the Obsidian Maze (a hacked version of Undermountain with certain changes, most notably the addition the Cult of Death Undying) and then made another small fortune plundering the depths of the Tower of War in the ruins of Castle Greyhawk. We went on to run a successful two-year "Redshirts" campaign featuring low-level guildmembers saving the world from the resurgent Temple of Elemental Evil.
Despite these successes though, megadungeons also helped kill the campaign. We started Maure Castle as a follow up to the Temple of Elemental Evil, running it as a series of one-shots whenever we didn't have critical mass for the regular game. It evolved into a weekly adventure, but it's lack of an overarching plot, combined with it's extreme lethality, led to frustration among the players. The length of 3E-style combats, in which a single high level encounter could take all night, didn't help.
"We're heroes of the realm, with fortunes in our own right," one of the heroes said. "Why in the Hells are we killing ourselves in this dungeon again?"
We abandoned Maure Castle in favor of the Expedition to Castle Ravenloft. Admittedly, this is not a true megadungeon, since it is a mix of wilderness, location and castle encounters, but ultimately it shared some of the same failings of Maure Castle. It had the necessary megaplot (kill Straud, save the kingdom) but the unrelenting brutality of its planned and random encounters wore everyone down. The ability- and level-draining powers of the undead, coupled with the lack of well-stocked (and defended) nearby town, ultimately sapped our will to play the game.
When Good Dungeons Go Bad
What went wrong? There were a couple of things. With Maure Castle, we were playing higher-level D&D 3.5 characters (in the 14-16 level range) and the campaign suffered from all the problems of 3.5 high level play. From "save vs. do nothing for four hours" effects to 90% of the party being unable to damage a particular monster, to the "teleport in, blow things up, leave" work day, we encountered the worst that D&D 3.5 had to offer on a weekly basis. None of us realized this at the time -- it took weeks for the frustration to build to the point where we'd all had enough.
The biggest problem with Maure Castle though was the lack of an overarching story. It wasn't enough for our group to simple keep delving deeper and deeper into the dungeon -- they needed a reason to be doing it. With the Obsidian Maze/Undermountain, they'd had those reasons, first in the form of profits from mapping the dungeon, then from their desire to smash the Cult of Death Undying. The player characters were willing to accept the risks because there were tangible rewards (either gold or revenge).
With Ravenloft, we again ran into the limitations of the D&D 3.5 engine. Undead in Third Edition were a challenge; fighting nothing but undead was a nightmare, and not the good kind. Many of the mechancis that prove problematic in 3E -- including multi-round spell effects that effectively knock players out of combat for half the night -- are integral to 3E undead. It wore us down, and ultimately we gave it up as well.
I can't say that Maure Castle and Ravenloft killed our D&D campaign, but it certainly contributed to it. Many of the issues we ran into with these games were things that 4E was designed to address (speed of combat, save vs. do nothing, characters who could do nothing against certain monsters, the one-encounter work day, etc.). The campaigns ended up exacerbating some players' frustrations with 3E, and they were the ones who found 4E most appealing in the playtest.
These sorts of problems are hardly unique to megadungeons, but I think that megadungeons tend to make them worse.
Despite the problems with our last two megadungeon encounters, there's still plenty about them that I find appealing. They're awesome campaign destinations, and done right, they can spawn the sorts of stories that people will still be re-telling 5, 10 or 15 years down the line. Now I should stipulate that when I say "done right", I mean "done right for my campaign." If you're campaign enjoys hacking and slashing their way through the Monster Manual, that's great ... but it's not so great for my group.
Give them a reason to be there: It's not enough for there to simply be this huge dungeon, the heroes need a reason to be there. Maybe goblins have a lair on the first level, and are using it as a base from which to raid near by towns. Maybe a death cult is lurking on the second level, and occasionally venturing forth to town to steal corpses for their dark experiments. Or maybe it is simple greed: some rich noble is paying them way, way too much gold to map the dungeon to its fullest. A good hook will get them to the dungeon. A great one will keep them coming back.
Use a mix of monsters, traps and puzzles: Pacing is important in a megadungeon. We learned this in Ravenloft; one fight against unstoppable is challenging, two fights is exhausting, three fights is frustrating. Players need safe bases to retreat to, as well as non-combat situations that give the skillful characters a chance to shine.
Seed the Myth: The best megadungeons have a thousand little hooks that entice players deeper into the dungeon. I think Maure Castle actually had these, but module doesn't do a great job of playing up areas like the "Hall of Antiquities". When players know that the Well of All-Heals is located deep within the dungeon, and just might resurrect a fallen party member, or that the City of Eyes is located on level six of the dungeon, it gets them thinking. You don't even need to have detailed these treasures; just a blurb and the occasional dropped reference is enough to pique players' curiosity ... and to keep them digging deeper.
The Necropolis of the ur-Flan
While writing this, I couldn't help but think about what, if any, role a megadungeon might have in my next fantasy RPG campaign. My group will be running a Pathfinder RPG playtest sometime this fall. Known as the Ur-Flan campaign, this will take place long millennia before our existing World of Greyhawk campaign, during a time when the Ur-Flan druid-necromancers held sway over the central Flanaess, and the greatest of their number -- Vecna -- was just beginning to consolidate his power.
I'm envisioning a megadungeon that's in the process of being constructed by the ur-Flan -- a tremendous necropolis being built for unknown purposes. I see it as being a sort of Mount Doom or Death Star of the setting; the momentous horror that looms over everything. When tribesmen are enslaved, it is to work in the mines of the necropolis. When virgins are sacrificed, it is to appease the horrors found in its depths (or perhaps to summon said horrors). I don't know if the players will ever step foot in it, but it will affect them and those around them regardless.
Now where's my graph paper?