Dungeons & Dragons creators Dave Arneson and Gary Gygax built their campaigns around huge dungeons that grew and changed. Megadungeons span enough rooms and levels to become the focus of an entire campaign without ever being fully explored. These megadungeons enabled Dave and Gary to run campaigns for dozens of players. On any day, they could host games for whoever happened to show up for a session. (See When Megadungeons Ruled Dungeons & Dragons.)
Even though the megadungeons under Greyhawk and Blackmoor became the foundation of Dungeons & Dragons, such dungeons rarely see play anymore. Why not?
1. Players never saw any examples. Originally, Gary thought that players would never pay for published dungeons. After all, players could easily make up their own. Despite this belief, TSR distributed the first published dungeon, Palace of the Vampire Queen. Strong sales proved Gary wrong, and so he set to publish his own dungeons. (See 9 facts about D&D’s first standalone adventure, Palace of the Vampire Queen.)
But Gary’s megadungeon under Greyhawk Castle seemed impossible to capture in writing. As adventurers explored and plundered, the dungeon changed constantly. New monsters wandered in to take empty rooms. Whenever the players’ attention turned, the layouts of old levels subtly changed. Entire new levels appeared. Most of the content lay in one-line descriptions, or worse, locked in the heads of Gary Gygax and co-dungeon master Rob Kuntz. Decades later, Gary wrote, “If we handed over the binders containing the maps and the notes, I don’t think even the ablest of DMs would feel empowered to direct adventures using the materials.”
So rather than attempting to capture Greyhawk Castle, Gary opted to publish adventures that he had created for D&D tournaments at conventions. For instance, the official D&D tournament at Origins ’78 ran the G1-3 adventures. The choice to publish such adventures changed the development of the game. D&D players everywhere saw Gary’s published adventures as a model. Instead of patterning their games after a megadungeon like the one Gary played at home, players imitated adventures created for a few hours of competition.
In 1990, TSR finally published WGR1 Greyhawk Ruins, its first megadungeon in print. “There are more than two dozen levels of horror and treasures. Run into brutal foes and gain uncountable wealth-nearly 1,000 separate room descriptions in all!” Gary had left TSR five years earlier, so fans hoping to explore his actual creation felt disappointed. James M. Ward and other veterans of the Grayhawk campaign still at TSR gave insights, but the dungeon even lacked the Great Stone Face Enigma of Grayhawk that Gary himself drew for the first D&D supplement.
The Ruins of Undermountain followed 7 months later. Undermountain appeared in a box with maps and with booklets that sketched out encounter areas. This outline mirrored the terse descriptions and evolving notes that Gary Gygax used for Greyhawk Castle, but the sketch failed to satisfy DMs accustomed to publications ready for play.
Perhaps locking a megadungeon in a book kills it. Printed pages cannot capture the dynamic essence of those original levels.
2. The ecology and rational of megadungeons seemed ridiculous. From they start, players struggled with the logic of megadungeons. Where did all those monsters get their food or leave their waste? Where did the creatures and treasure come from? Every dungeon master invented an insane wizard as an architect for their game’s underground sprawl until the notion became trite.
In the little, brown books, Gary suggested dungeons with layouts that always changed and grew to “maintain freshness,” but that made the megadungeon even more implausible.
Then Gary published adventures that featured a logic sometimes called Gygaxian naturalism. Monsters had lives of their own that involved feasting, scheming, sleeping, and everything but waiting for heroes to come kill them. Rather than wandering monsters living in defiance of reason, we saw giants and drow in their steadings and vaults. For many players, the giant- and drow-series adventures set an example that killed the megadungeon.
Soon, any DM peddling a megadungeon had some explaining to do. For instance, The Ruins of Undermountain kept to the insane wizard trope, then added magic that continuously gated in fresh monsters from across the Realms, and deep entrances that allowed creatures from the Underdark to well up.
3. Play styles expanded. Sometime in the middle of the 70s, for the first time ever, a party of adventurers visiting the inn met a hooded stranger with a job that needed doing. D&D expanded beyond a series of dungeon expeditions aimed at claiming treasure. Players began to favor games that mixed action with story. Waterdeep: Dungeon of the Mad Mage rates as my favorite megadungeon in print, but when I ran it, the group longed for a story and for motivations beyond a hunger for treasure.
4. Megadungons can feel monotonous. Even the biggest megadungeon only shows a tiny corner of the giant canvas that D&D worlds can offer. In a campaign limited to a single dungeon, kicking in endless doors to fight and loot can start fresh and thrilling but often becomes a tiresome slog. Even those of us who like dungeon crawls want to see some daylight and a plot.
5. Computers do megadungeons better. In 1979, computer games like Akalabeth and Temple of Apshai took gamers into megadungeons and started an electronic-gaming genre. Dungeon crawls limit players’ options, so they offer an easy premise for a computer game. (See How the Dungeon Powered the Success of D&D and the First Role-Playing Games.) With a computer DM, players can explore anytime. Digital dungeons offer faster play and better graphics. For players who just want to visit a sprawling underworld to kill monsters and take their stuff, electronic games probably offer a better experience.
Can a megadungeon work today?
A clever design can avoid the problems that pushed megadungeons out of play.
A story-centered game can take PCs into a megadungeon to accomplish more than looting. For instance, when Monte Cook created his superdungeon The Banewarrens, he paired it with overarching plot. Players don’t raid the Banewarrens just to loot. Instead, the story leads to objectives that require missions into the place.
Many megadungeons avoid monotony by introducing levels or zones centered on unique themes such as crypts, flooded sections, or fungus gardens. Even the levels under Castle Greyhawk followed themes that grew more exotic at deeper levels.
A megadungeon design can add intrigue and roleplaying by borrowing a page from The Keep on the Borderlands and adding factions of monsters. Players can join a side or play one against another. Factions under attack will bring reinforcements, creating more interesting battles, and giving players a reason for caution. The stories “Red Nails” by Robert E. Howard and “The Lords of Quarmall” by Fritz Leiber helped inspire the concept of dungeon exploring. Both yarns centered on feuds and intrigue.
A megadungeon (and a live DM) can create player agency and tests of ingenuity that no computer can match.
Although good design can yield a megadungeon that proves fun to play, ordinary dungeons can bring the same advantages. Today’s gamers tend to create megadungeons to foster nostalgia or to enable episodic play.
I think that the most important reason for Megadungeons to be mostly a thing of the past is point 3. When D&D started to change from a roll-play to a role-play experience, Megadungeons became a hindrance, because it’s exceptionally difficult to tell a good story within such a limited space. Also the original dungeon is a very gamist play experience and when players started to care more about the narrative, the old story of “kick in the door, kill the enemy, loot the treasure” wasn’t enough anymore. Now I’m obviously biased because I probably would never have started playing D&D when confronted with a Megadungeon and all the gameplay that that structure involves and never had the wish to give it a try. I mean it’s cool to wander through Moria, have one or two climatic encounters and escape that dungeon while still alive, but in D&D terms, that would be having an endless array of encounters against Orcs, and while this is kinda cool to see in a film sequence like in the Hobbit, at the game table that becomes stale very quick, just because it takes so friggin long. But than that’s just me, and I’m probably an exception to the norm.
I think the first thing that is needed to successfully run a megadungeon is players who want to play a megadungeon. And not just heard about the idea of giant dungeons and thinking it sounds cool, but actually wanting to play an infinite treasure hunting dungeon crawl. If the players are not really looking for that, I don’t think any amazing execution by the GM will make it work.
There certainly are numerous players who are genuinely interested in seeing for themselves what these old campaigns from ancient times played like, but if players don’t already have this desire, it seems very hard to sell them on this idea.
When you get a group of people together to play D&D, what can you really offer them that makes playing a megadungeon more (or at least equally) appealing than something more story focused?
Unless you have people coming together who all want to play a megaungeon already, I don’t think you can make a megadungeon work for a group, even if you really would love to run one.
I agree with you Wormys_Queue, but the funny thing here is that my first D&D campaign (2nd ed) actually was within Moria. And we died before being able to get out. Maybe it created a yearning for this kind of megadungeon and that’s why I like it so much today? Who knows.
That being said, I think one “solution” for what you speak of is also change the function of the megadungeon with relations to the world and vice-versa. In my current open table campaign, I made the “error” (with regards to this subject) of having a megadungeon that’s too “neutral” and a world that’s too enticing (huge hexcrawl with bunch of different stuff to do). Since people are not used to MD, they prefer to explore the hexcrawl instead, even if it’s less lucrative and more dangerous.
I believe that if the MD is the main setpiece and that every other locales is directly linked with it from near or afar, it completely changes the dynamic and really push players to go there. And my reasoning with this comes from (ironically) a 5th ed. adventure I ran: Curse of Stradh. Since everything in CoS is linked with Castle Ravenloft, the actual delving (although not that big) of the Castle becomes memorable: everything you do outside the castle “leads” one way or another to the castle. That’s, IMO, a genius move and that’s what I think a MD campaign should be. At least, that’s what I’ll do next time!
Wormy I think you are largely attacking a straw man. There never was a switch from “roll Play” to “Role Play”. People were Role Playing their characters from the get go. That hasn’t changed a bit.
Expanding on what Wormys_Queue says there, it seems to me the big question is Why would you want to center a game around megadungeon? (Other than “This is a drop-in game with little to no continuing plot.”) If you look at fantasy and mythology as a whole, dungeons of any sort play a relatively minor role, and megadungeons are even less common — and if you do run into such a thing, it is as likely to be inspired by D&D as the other way around.
“This is a drop-in game with little to no continuing plot.” Exactly why we use it. I don’t have a group that can commit to play every other week. A have a pool of 15-20 players, most of them with full-time jobs and children, that just want to spend some hours playing D&D. It’s really easier to have a megadungeon with some quests at hand, than prepare a new one-shot adventure everytime we play. We don’t have a problem if the session needs to be interrupted, or if some player wants to leave earlier. The characters can just go back to the home base to get ready for the next expedition, spend their treasures and discover some new rumors.
Me 2. Working for me almost 20 years.
I think megadungeons are where CRPGs take center stage. There are so many player created modules for NWN and Baldur’s Gate that run with the concept, but it’s something that’s much harder to get right and keep player interest in a PnP gameplay setup.
I think you are spot on in your analysis.
When I started playing, late in 1981 and onward, I never considered a megadungeon as an option precisely because I never saw one. The one-shot modules were the only model, exactly as you say. I learned to string scenarios together in an evolving epic in dialogue between characters and campaign world, and soon flew solo without modules. I hardly ever ran dungeons when there were so many ways to run adventures.
I remain fascinated by the record of the earliest gamemasters who each had one dungeon that slowly expanded through play (and I ramble about it here: https://lichvanwinkle.blogspot.com/2020/05/one-dungeon-per-dungeon-master.html).
Mark Swanson, a prominent D&D zine author of the ’70s, described his beginnings in D&D as “trapped four years in a gilded hole.” He describes exactly what you say: he got bored of dungeons, using almost exactly the language of the previous commentator, Wormys Queue: “Bash the door and kill the monsters, break the traps and then march on.” Even wilderness surroundings bored him until he finally found fellow players, new play styles, and ideas to create the “fantasy epic” he had always craved. This was still in the ’70s. His testimony is in DW#1. He ends by suggesting he wants to start all over with a re-design, re-imagining fantasy role-playing. I think that this encapsulates a pattern we see repeated over and over and over… Look at all the clones, each the product of loving care.
This is the way to do in homebrew. All dungeons are one.
And Megadungeon are not to “clear” all rooms.
The problems pointed in the article are design problems.
And here i was excited to use the Mad Mage Megadungeon with my players 🙁
Nah, i actually run a bunch of quests that are reasons for they to get in there and get to certain specific levels or rooms on such levels. I dont really want them to run the whole Undermountain. 🙂
I hope they will like that game twist from the first 6 levels of storytelling.
I had another idea that worked pretty well in 4th edition : a reverse megadungeon. If anyone read the Deathgate Cycle by Margaret Weis it run on a similar idea.
The characters starts in a prison that is actually a full level “city” and there are no monsters. The prisonners founded a city, had kids, … a simili civilization with a lot of barter. The level just over it has monsters from level 1 that serves mostly as game and food for the city so many people get to the 1st level. The higher you go the harder it gets and every few levels there is a one way portal that prevent them from returning to the lower levels including the city. Those levels usually have a small hub near the portal where people that fear going foward will stay at and might help the characters. It is known from the start that if a prisoner make it to the outside of the prison they are considerer free.
Of course, my prison had 21 level so when they would get out they would be Epic (in 4th ed term) but sadly the attrition of the prison is so high that none got through it yet. Their character died on level 17.
It was a pretty cool concept to run. The levels even if they were in a “dungeon” were sometimes more like huge caverns that felt like they were outside in a forest… so it wasnt just door bashing / monster killing / treasure looting.
Just writing this here makes me want to go revisit this setting in 5th.
Why would you want to center a game around megadungeon?
Because the characters are actually wannabe generals for tabletop wargames happening on the surface of the game world. The player was creating an army on the surface by use of followers and retainers and creating a general for that army by leveling him up and getting rich to pay them down in the dungeon. The dungeon was a means to an end, not an end in itself: get rich, level up, and wage war above!
BTW, I played Temple of Apshai, and the spin offs. Super cool game and at the time there was nothing like it. Wouldn’t rate now, though. You moved a stick figure with a stick for a sword (literally) through a black screen between white lines that represented walls. Keys on the keyboard represented which actions you could take. Numbers 1 – 9 were how many steps you could move in one turn. Bats were actually flying “V”s. Absolutely groundbreaking!
Good post overall. I particular agree with the point behind #1 but I would say it isn’t true that Greyhawk is somehow un-publishable (if I’m reading the meaning there right). Don’t be surprised when you see it in print in a few years. Gygax never published Greyhawk because it is very big and a lot of work to expand short notes into room entries. He never quite got around to it, although Castle Zagyg upper works was a start in that direction. And of course, Arneson did publish his megadungeon – twice.
I’d argue rather strongly that B4, The Lost City, was TSR’s first megadungeon, and B4 is a good example to demonstrate that the problems you discuss in #2, #4, and #5, while very real, have nothing to do with dungeons, regardless of size or type per say, and everything to do with POOR DESIGN. A well made megadungeon like B4, or even Blackmoor as Arneson ran it in the early ’70’s or Jakalla as Barker ran it, is all about a teeming, living underworld with factions and plotting and opportunity. If players are finding it boring and dull then blame the game the DM is running – not the concept of a giant, vibrant, underworld labyrinth.
Regarding point 3, well, I don’t really understand. Characters were being given missions before D&D existed and throughout the early 70’s. There were always overarching villains and their minions (Hello Egg of Coot, hello Saint Stephen, Hello Zuggtmoy). Gygax, Arneson, Barker – any of the early DM’s you care to name developed plots and goals for their players. Like the mission to destroy the artifact in the city of Father Dragon in Blackmoor or The Great Vampire Hunt, to name a couple off the top of my head. Now, I’ll grant you that there were no doubt plenty of people running pointless monster zoo’s, and that it was largely TSR’s fault for not providing more examples of more creative adventure-crafting in those earliest years. I think, to be honest, that Gygax simply assumed that people would create imaginative adventures and didn’t need to be hand held in that regard. Time demonstrated otherwise.
Thanks for weighing in. Your comments always fill gaps I missed and I value your insights.
Don’t forget that TSR was the not the primary source of dungeon material in the period 1977-1980. Judges Guild was very popular at the time and were selling many thousands of copies.
* Referees who bought The First Fantasy Campaign got some ten levels of Blackmoor Castle.
These dungeons were lightly keyed (to the point of uselessness) but the MAPS of a castle and ten levels below it were included and perfectly usable, but the concept itself was very clear, and in 1978 I had no trouble figuring out how to do a megadungeon from this example.
* Referees who bought Judges Guild’s City State got something like eight or so dungeon levels (plus the really intricate Sunstone Caverns). These were completely unkeyed, but the maps were fantastic for the era, and again.
* TSR sold the Monster and Treasure Assortments which provided hundreds of simple monster/treasure encounters pre-keyed (plus the dungeon geomorphs, but these maps were crap compared to the Judges Guild ones).
This meant that I could easily combine the Judges Guild and TSR products to produce large megadungeons c. 1978-1979 – and I did so. I’m sure plenty of others did as well.
Moreover, Judges Guild also produced several multi-level dungeons that were fully keyed. The loosely themed DARK TOWER (1979 or 1980) with four main levels and at least two extra sub-levels had 65 pages of dungeon encounters; while not quite a megadungonen in contemporary standards, it was probably the closest to one in print at the time. THIEVES OF FORTRESS BADABASKOR with five levels in 30-ish pages was also fairly sizable.