Thursday, October 11, 2012

Lessons from Running Mega Dungeons

A few years ago, I ran a summer-long campaign using the Stonehell dungeon.  The lessons learned were invaluable, and heavily influenced some of my decisions in structuring my current campaign, The Black City.  Some of the things I got right; there are plenty of things that still need refinement.  Running your campaign using a megadungeon, campaign dungeon, or tentpole dungeon (pick your term) as the centerpiece offers a lot of benefits, but it's not a risk free proposition.  There are a couple of basic things you have to absolutely get right, or it's going to crash and burn on you.

Feed the Player Choice Engine
Choice (or agency,as the cool kids call it) is the magic engine that makes the mega dungeon work.  Information is the fuel.  The megadungeon needs an interesting background to engage the players, and they need to learn enough about the dungeon to plan their own capers.  Making blind decisions is no better than a coin flip, and when the players don't have the right information, they're choices are meaningless.  OSR luminary Matt Finch said it much better than I could on a comment here at the Lich House:  Running a better sandbox game.  Read and gain 1 point of Wisdom.

Information tools include things like campaign background, rumors in town, clues picked up within the dungeon, rumors from other adventuring parties or dungeon inhabitants, treasure maps, and so forth.  These tools are tried and true.  The players need to be able to plan, and their plan needs inputs.

I would elevate this as the # 1 reason, bar none, mega dungeons fail - failure to feed the engine.

Interesting environments develop engagement.  You've all heard the advice - make sure your megadungeon has some distinct levels and areas, to limit repetition and monotony.  Design set piece areas, that are intricate, tactical, or otherwise very cool.  Add factions to the dungeon so the group can engage in non-combat roleplaying, or a bit of politics.  Make sure the rooms have interesting encounters that provide a challenge, or the chance to learn something (even if its an empty room).  This is all common advice, good advice.  My list is not exhaustive, and one can certainly argue more or less special ingredients need to be in that sauce.

But engagement requires information, first and foremost, and that means feeding the engine.  Putting in a wicked cool 'Fountain of Serpents' as a major set-piece location is less valuable when the player characters don't hear about it, learn why it's cool, and get enough information to plan a mission to try and find it.  Multiple competing factions are irrelevant if the group doesn't discover how to take advantage of the rivalries.

The Dwimmermount Controversy
Tenkar over at Tenkar's Tavern recently ran his crew through a level of Dwimmermount (Closing the door on Dwimmermount) and it spawned discussion there, and on TheRPGSite.  The players didn't like it - and ink has been spilt.  Dwimmermount discussions always seem to have popcorn value, since it's a high profile project that has veered off the road a bit.  James Mal has one of the largest pools of OSR readership, perhaps encouraging a bit of schadenfreude - I just know, mention Dwimmermount, and folks come out swinging hatchets.

However, if players aren't excited to step foot into the dungeon or plan a targeted excursion, they clearly don't have the right information.   I can't say if the drafts are to blame - I'm a project backer for Dwimmermount, but haven't bothered with the downloads yet.  Tenkar's group did explicitly set out to run the levels they used "as is", without heavy improvisation or enhancement by the DM.  It's spawned lively discussion about empty rooms, evenly distributed copper coins hidden in rat poo, and nothing less than the very death of the megadungeon format!

No published megadungeon can capture the magic of the campaign that birthed it.  Running a game session is performance art, and the more sparse are the published notes, the more improvisation is required to bring it to life.  The manuscript needs to fire the imagination and give the DM something to work with, too.  Keep your popcorn handy, where will this saga go next?  I should I probably get off my keister and actually download some of the evidence, experience the drama firsthand.  I'm still a popcorn muncher.

I had expected to spend time talking about what's going wrong and right with the Black City, but I'll have to save that for tomorrow.  My soapbox is about to collapse under the weight of self-importance, and I've got enough bruises this week already.


  1. When I started our campaign with the Holmes' dungeon (expanded to mega) as our tent-pole dungeon I fed the engine with plenty of rumors and background. It got them going and in fact it's been hard to make them leave. I've had to push them a bit to get them into other adventures. I have broken up the dungeon crawling a bit with some wilderness but the players seem fine with how things have progressed.

  2. Top notch post. I think you said it better than Matt.

  3. Part of the current problem with Dwimmermount (1st level is all I can speak of) is that much of the dungeon dressing (statues, mosaics, ghostly images, etc) is dependent on setting information that isn't available yet. Gods and historical personalities lack substance without meaning and purpose.

    From what I gather of the weekly updates, there will be background and immediate setting material for the vicinity around Dwimmermount, but that doesn't appear to be in finished form at this time.

    I'm not taking another look until it's nearly complete ;)

  4. One problem with the internet is that people tend to be impatient when they have the opportunity to get information quickly. The project is not yet complete, so it can't really say anything general. I look forward to seeing the final book. I doubt that it will herald the end of the megadungeon, either - that all seems to be driven by people who hated the concept to begin with (because they had bad experiences with homebrew ones or because they hate anything that is different than what they themselves make or whatever). There are several counterexamples, both OSR and actually old-school, to argue against that idea (Stonehell, Rappan Athuk, Castle Greyhawk in the version that wasn't just people heaping derision on Gygax, and so on).

  5. Erik - this has been Tenkar week! It's fantastic that you've generated a week's worth of commentary across many old school blogs and discussion boards this week. For that, take a well deserved bow. I love the chance to get out the popcorn (and participate, in a small way).

    Reserving judgment until we see the whole thing is prudent. I seriously question the premise of running something like a megadungeon "as is", where we know there's so much detail omitted that the DM is required to improvise and own the material. My Stonehell game fell flat at times, and I've done my best to internalize the reasons and adjust. My expectation is the finished product needs to provide enough inspiration to get a prospective DM on the right path, and that's how we should judge it.

    For that matter, the big guy himself should start paying some attention. It can be tough at times, but healthy criticism is to be expected when the backers are funding the creative process and seeing the project unfold. Capitalizing a project creates a different relationship than a pure pre-order or retail purchase. In the end, everyone wants a great book. (Well, I certainly do).

  6. I never even saw the rpgnet thread ;)

    It's not that James didn't give the "empty rooms" visual contents,but without the background info they will (hopefully) be linked to later, they had little value. In some ways, and truly empty room may have worked better in this case.

    Lets see how tomorrow night's game session goes. away from Dwimermount and into The Eagle's Den...

  7. Thanks for the cool Megadungeon post. I agree that it's important to design the dungeon in a way that keeps the players interested. Another concern is that major areas of the dungeon may go unused, so the design/stocking should account for that.

    I did recently manage to create my own large dungeon using the Dizzy Dragon generator. I was able to save it off to a giant .jpg file and even printed it out as a 4' x 8' banner.

    Here's my Megadungeon.