Megadungeon Sandbox and 4E

4E Sandbox, Megadungeon Style
I want to try a sandbox approach in an upcoming campaign. I like the idea of putting a megadungeon out there, provide a half-dozen hooks or opportunities for adventure, and then let the game unfold. I have a compelling idea for a 4E/FR-based megadungeon that I'm developing; it will take place in Narfell on the frozen shores of Icelace Lake, the ruins of an ancient monastery used by an order of wizards, right on the lake shore. What I want to get from the community are ideas on the theories lined out below.

What is a Megadungeon?
Megadungeons go back to the root of the hobby (1E AD&D and the OD&D little books). The idea is that an entire campaign can be built around a dungeon, and plots emerge through play that keep the party moving forward so it doesn't devolve into hacking and looting. But the key is, the party controls the pace and determines the risk-vs-reward ratio. Not many have been printed, Temple of Elemental Evil , Caverns of Thracia, Tegel Manor. I loved Castle Whiterock in 3.5, too. I had great hopes all of Castle Zagyg would see print. There are lots of places to go learn about them on some of the grognard hangouts; run a search over on Dragonsfoot, Knights and Knaves Alehouse, etc and you'll find some great posts on Megadungeon Theory by fellas like TFoster, Evereaux, map analysis by Melan, etc (those posters stand out to me, but there have been many great contributors).

My goal here is not to rehash megadungeon design but discuss how to modify the theories to work with 4E adventure building models.

Sandbox Theory and Dungeons
I guess I consider Sandbox theory the opposite of railroading. The DM puts 'just enough' of the setting out there, with some adventure hooks, and the players make of it what they will. The DM builds on the players choices and the campaign emerges in a more collaberative manner than if the DM started with a Cool Epic Plot™®. JG's Wilderlands is my archetypical sandbox - its just a ton of mapped hexes with a guide and the players are free to wander; if they go north instead of south, there's no story they're ruining, the story will emerge.

Mike Mearls' blog post on Keep on the Borderlands got me thinking why B2 is still so charming; a unique story emerges each time you play it, through a mixture of rumors, hooks, and the player's choices. There's no linear plot to it. I DM plenty of Adventure Paths, and once in a while you just want to punt the baggage, the McGuffins, the Mary Sues, the railroads.

Sandbox and Risk-Reward
Players control the level of risk-reward through their choices; if the DM presents them a staircase that goes from level 1 to 2,3 and 4, and they decide to jump to level 4, a TPK is likely. Similar to Keep on the Borderlands B2, where an unwise group could blunder into the Minotaur maze, the Shunned Cave, or visit the Bugbear before they're ready.

The Five Room Dungeon
A modern theory I like is the five-room dungeon - it's a compartmentalized approach to design (like the 'Delve') - here's a typical 5-Room Dungeon style:

• Room One: Entrance And Guardian
• Room Two: Puzzle Or Roleplaying Challenge
• Room Three: Trick or Setback
• Room Four: Climax, Big Battle or Conflict
• Room Five: Reward, Revelation, Plot Twist

Okay, so what's a 5-Room Dungeon doing in my Megadungeon?! Keep on the Borderlands is a good example of how a dozen independent lairs, loosely connected in a sandbox, can be assembled to form something larger. I'm keeping the 5-Room dungeon option open as a design model, assembling the megadungeon from a series of lairs or delves. It means you don't sit down and design 20 rooms... you sit down and design a handful of loosely connected delves or lairs.

You may note that assembling a larger dungeon from a series of delves or lairs appears to be the model for the 4E published adventures - it's present in H1 where you see the dungeon level 1 broken into areas like 'Goblin Encampment', 'The Tombs', and 'The Caves'; level 2 has 'Hobgoblin Borough' and 'Dungeon Chambers'. H2 and H3 are similar.

Wandering Monsters and Rival NPC's
I think Wandering Monsters add an interesting dynamic to an otherwise static setting. My favorite kinds are when more powerful monsters (maybe from the next level down) are up looking for a snack. (Obviously these aren't all meant to be fights - players need to know when they're overmatched!). The other kind I miss are the wandering NPC parties. Setting up adventurers or other intelligent delvers as rivals can lead to those emergent storylines that develop from the sandbox. Who doesn't remember Gutboy Barrelhouse and Balto throwing down with Arkayn and Abner from the 1E DMG combat example?

Experience Points and Treasure Parcels
This is a tough one for megadungeon design, and I welcome ideas. Here is the issue: you don't want to over-create your levels, and you also need to assume the players won't uncover everything. How big do you make any given dungeon level so it feels 'large' but you don't create an extra thousand rooms, say? How do you employ the 4E guidelines for parcels and items to ensure the group gets the right wealth per level?

On the one hand, 4E might make the treasure parcel piece super easy; maybe you just key it "in pencil" just like you were building B1 (In Search of the Unknown) - first 10 treasure-bearing encounters lead to the first 10 parcels.

Moving the Game Forward
My theory is that through a combination of emergent stories, seeded plot hooks and mysteries, and similar concepts, there will be incentive for the players to delve deeper into the megadungeon as they progress up in level. Keeping the players going vertical and not too horizontoal is an area of concern (highlighted below in problems).

Bringing it All Together
I've developed an overarching plot and framework for the megadungeon (I like the top-down approach, 'Rational Dungeon Design / Let there be a method to your madness' is a great Dragon Article on megadungeon design from way back in the day - Dragon #10 in fact). Once I have a theme for each level, I plan on outlining something like 3-4 delves or lairs to plug in, with room for expandability.

For instance, here is a practical example from the megadungeon I'm currently working on.

I have two potential delves in the upper works (one in the lakeside south part of the ruined monastery, one in the north). The lakeside delive includes a vermin encounter and a human bandit encampment; the encounters on the other side of the ruins include goblins from the Blood Mountain Tribe and their bugbear allies. There are at least 3 ways into the first level of the dungeon from the upper works. (There are also a series of caves in the nearby ravine, and one of the caves leads to level 1 as well...)

Delves planned out for the 1st level include the dragon worshippers (a kobold tribe very similar to the Kobold Hall delve in the DMG with a solo white dragon), a vermin delve (dire rats, fire beetles, stirges, etc), mutant goblins inspired by the roll vs role article on goblins, and finally a delve built around elves, halflings and their guard drake companions (exploring the ruins, these agents from The Good Lands™ have set up a temporary redoubt).

Wandering encounters for level 1 include orc raiders up from level 4, hobgoblins, vermin, and an NPC party. The orcs, the toughest wandering monster, aim to capture/subdue opponents (so there will be no TPK). They sell captured humanoids to the level 6 Shadar-Kai, who remove their captives to the Shadowfell to fight in a Shadar-Kai arena.

There are a handful of plot hooks to get the characters into the dungeon; the biggest one (that launches the game) goes like this: The ruins have existed for centuries as a haven for monsters and bandits, and they've drawn low-level adventurers from time to time to challenge their skills there. But the deeper levels have always held a mystery - the Iron Portals that seal the upper levels from those below. Last year, the famous explorer (haha, I'll call him Arne Saknussem for now...) deciphered the secret of the portals and discovered a large dungeon complex beneath the known levels, leading to the very Underdark itself and the secret laboratories and halls of knowledge for the ancient order. Saknussem returned to the surface with magical trinkets, ancient coins, and items of power; word spread to the lands of Damara, Impiltur, the Dalelands, and even darkened lands like Netheril and Vaasa, and now adventurers of many stripes have traveled to the remote outposts of Narfell to plan their own expeditions into the ruins.

For a more focused plot hook, consider the information from the Orc wandering encounter; there are Shadar-Kai slavers in the ruins that capture humanoids and sell them to fight in the arena. Local barbarian tribes from the tundra use the ruins as proving grounds for young warriors coming of age; warriors sneak into the upper works, descend ropes down a shaft into the great hall on the first dungeon level, and are expected to spend a night in the ruins, returning the next day with some token proving they spent the night. More and more of these groups of young warriors are failing to return; they're targeted specifically by the slavers and taken to the Shadowfell where the barbarians make prized gladiators. Characters investigating this hook would follow a series of trails and ultimately get a chance to enter the Shadowfell, fight in the arena, lead a slave revolt, or otherwise win freedom for the prisoners. (Of course there has to be an arena in the megadungeon after Dragon had those macho Gladiator articles!)

Problems

Okay - so I think there are some problems with megadungeons... some 4E specific, some not.

1. Overdesign:
You want the place to feel "big" - so it makes sense that multiple NPC parties could be exploring at once, and there are multiple ways into and out of the place. However, I also realize that a group will only need 8 or so encounters before they level up (which is probably 2-3 delve areas, 1-2 quests, and a skill challenge). I'm thinking of going with the tried and true approach - barebones notes, barely sketched out for the first pass, and then winging it. If know where the party is going, I'll be able to create a lot more detail.

Anyone else tackle this type of problem - your characters will only likely encounter 50% of your areas, so how do you balance enough detail vs too much?

2. Leaving Room for Expansion:
Haha, I guess there have been a few creative solutions out there - the Greyhawk Construction Company comes to mind, as does 'The Fog' I heard was used in Castle Zagyg. I'm not sure yet... one idea might be to sketch out some large major hallways - "Roads" like in Moria... provide slots for where certain Delve encounters could take place... and place my pre-designed Delves in the slots the characters actually choose to explore. The less detailed Delves get pushed off and I wing-it if necessary. Or I constrain where the players go and use some 'Greyhawk Construction Company' device to limit their options. Even B2 had the rubble leading to the Cave of the Unknown...

3. What if they don't descend?:
So what do you do if the party levels up, they realize they only scratched the surface of level 1, and rather than descend to level 2 they decide "No matter how long it takes, let's clear this sucker first!"? Megadungeons aren't meant to be exhaustively mapped and stocked, it ruins the aura of 'it's too big to be known entirely'. And besides, after a certain point the challenges will be downright boring.

The brute force approach goes like this - once you've advanced, you've 'exhausted' the experience you'll get on this level and you need to go down to level 2. I've seen some of the grognards propose fractioning the experience so the party gets the hint - once you're level is higher than the current dungeon level, you get 50% xp or something like that. Do you invoke the 'Bag of Rats' rule? Ostensibly level 1 fights are still in the range for level 2 characters, they're just considered "easy".

Note: I'm assuming that one way or the other the DM is ensuring the characters have received the necessary info they need to descend, easy access to up/down access points, they've received the correct amount of parcels and items, etc. Therefore the only pragmatic reason the players would stay on level 1 is grinding...

Awesome thoughts. I'm considering adapting Ruins of the Undermountain (read: plagiarizing what I need) for a mostly dungeon oriented homebrew campaign I want to start, but I hadn't thought of alot of the points you raise.

My player in this campaign has a good imagination and is a good sport, but is relatively new to RPG's. I'm afraid if I don't use a fairly structured setting (traditional dungeon) I might lose him! He's a good guy, but I don't want to provoke a "Oh, so you're just making it all up as you go along" response. A good Wizardry style dungeon romp ought to work. As he gets familiar with the mechanics and narrative dynamic, we can layer in more and more role playing and shift away from pure hack-and-slash.

Anyway, the point is that I think your commentary's very relevant to what I'm doing at the moment.

I like where you're going with this.

The "5 Room Dungeon" principle works. I used it in designing my Tekumel megadungeon (the Jakalla underworld) for my EPT (1975) game. Each level is really just a collection of small, self-contained dungeons.

For example, the party started at the first entrance I showed them. They went down into some catacombs (a complete mini-dungeon) and before completely exploring them started to deal with a functioning underworld temple (The Black Abode of Putrescence Triumphant, to Durritlamish). Even though they beat the roughest encounter, they didn't seal the deal (lost their nerve I guess)! Then they moved in another direction and plundered a mini-dungeon called The Tomb of Kalvar the Cruel. It remains to be seen where they're headed next. There was a pit-shaft out of Kalvar's tomb down to the Tomb of Jirek the Relentless (buddies to the very end!) but they were too scared to go down it.

In my own game, each session is designed as a single foray. Theoretically, characters who remain in the underworld at the end of the session have to roll on the Table of Despair ("Your flesh is consumed by underworld denizens!", etc.), which has no positive results. So the game is for whoever shows up, and there's no problem of "Bob's character is here, but Bob is in Seattle" or whatever.

On the question of trying to "clear level 1", I see the point of just reducing or removing XP. You could also make it so darn big that they'll never do it, or just leave enough room for expansion that previously unknown areas open up later. But what you're doing could work.

crash_beedo said:

Experience Points and Treasure Parcels
This is a tough one for megadungeon design, and I welcome ideas. Here is the issue: you don't want to over-create your levels, and you also need to assume the players won't uncover everything. How big do you make any given dungeon level so it feels 'large' but you don't create an extra thousand rooms, say? How do you employ the 4E guidelines for parcels and items to ensure the group gets the right wealth per level?

Click to expand...

You give the appropriate treasure to monsters. It's part of the risk-reward thing.

If the encounter is level 4, give it a level 4 treasure parcel. You could roll randomly (hoping things even out over time) or make sure all parcels have been handed out - i.e. you have 10 level 4 encounters, so you can cover all the level 4 treasures.

Mix it up, too: Hide some treasures, make some encounters treasure-poor, some encounters treasure-rich.

If you do this, then the players know that they will get more rewards for punching above their weight.

Hey, this is a great thread! I haven't signed onto 4E yet, but if I do it will definitely be sandbox style.

I think you have to throw out the idea that characters end up with a certain amount of wealth at each level. Like LostSoul said, monsters have wealth according to their challenge, and the characters are entitled to whatever they can get!

PCs in a sandbox game who get to purchase magic need to make different choices than in a standard D&D game. They need to assume that they will end up overmatched at some point, and invest in 'escape magic' or other means of not ending up a snack. I don't know enough about 4E to know how they would go about this.

anyway, I look forward to reading more on this subject!

Ken

I'm not entirely sure yet, as I haven't played the system enough yet, but so far, I think 4e is a bit less dependent on specific wealth levels, at least the way 3e was. I'd think that its a bit easier for the party to miss a treasure parcel or two, especially the ones that just contain gp. But with the transfer enhancement ritual, its pretty easy for everything that the party finds to be useful.

I guess here is why I'm drawing attention to the wealth problems with the Parcel system...

Game balance expects PC's to have a certain amount of magic items (and money to spend on more, or potions and rituals) - and in 4E, this aspect of the game is managed by the Parcel system. From that perspective, the DM can create treasure and magic items for a series of encounters, literally in minutes.

BUT - in an old school megadungeon, you'd go by things like the % in lair, the treasure types and charts (haha, remember treasure type "Q", stuff like that!)

If you built a level 1 dungeon with 50 encounters, you literally could give every monster an appropriate treasure. And one reason players might not spend time on level 1 any longer is the XP and Treasure rewards they're getting no longer keep them moving forward at an acceptable rate.

In the 4E paradigm, I don't believe XP is exponential any longer... it's a more gradual curve... raising the issue that a level 2 party could still be advancing at a decent pace while whomping on level 1 monsters.

I guess I see the Parcel and Encounter level, XP per encounter, and Dungeon level issues are a bit intertwined. In a normal dungeon - guess what - you've beaten all the encounters, time to go down to the next level. Harder to do in a megadungeon where there are more areas of the current level that could be explored.

Not ready to say 4E is a mess (yet) because we love the game play and I love the behind the screen aspects of it, but the gentler XP curve may cause some problems in the megadungeon environment.

Korgoth said:

I like where you're going with this.

The "5 Room Dungeon" principle works. I used it in designing my Tekumel megadungeon (the Jakalla underworld) for my EPT (1975) game. Each level is really just a collection of small, self-contained dungeons.

Click to expand...

Cool - I'm glad to see the modular approach has worked.

Even if the players "cleared" an area, I'm thinking there could be ways to open up new areas - for instance, their rivals from town (The Black Hats) find a secret door to one of those tombs, but only one of the Black Hats makes it back to town alive to tell the tale. Voila, new "wing" of the dungeon is ready to go, replete with atmosphere and intrigue.

A single monster increase in XP value by 25 from levels 1 through 4. So a level 1 monster is 100 xp, a level 2 is 125 xp.

Take this to the next logical conclusion, a party of 5 monsters goes from being worth 500 to 625 XP.

One step farther, 8-10 encounters of this type would require 4,000 to 5,000 XP (total among 5 players), and for level 2 would require 5,000 to 6,250 XP.

Finally, if players were facing average level 1 encounters despite being level 2, we could surmise 5,000/500 - 6,250/500 means approximately 10-12.5 encounters required to go from level 2 to 3.

Just for posterity, consider going from 3 to 4 with level 1 encounters:
750 (XP for level 3 encounter) * 10 = 7,500; 7,500 / 500 = 15 encounters, 5 extra. This seems to be something of a predicament, so I might advise against dipping them directly into the mega dungeon from 1. at around level 4 or 5, it should be less of an issue.



For treasure parcels I recomend the following: Plan out treasure parcels independant of monsters. Plan them out for every level you plan to let the players progress through. For each encounter that would drop treasure, start with the appropriate parcel for it's level, then find other parcels of equal value and create a "back up" list. This is easiest for magic items since over the course of going from level 1 to level 4, there will be 4 level 4 magic items. After you have created enough of a back up list, use this list multiple times whenever it's appropriate for that encounter. If the first choice is taken, move on to the next until you come to a parcel you haven't given and check it off. If there are no parcels left to give out, the monsters drop either nothing or a cursory amount of gold.

Do the same for things like dragon hoards which contain multiple parcels, but create an independant list of equivalence for each parcel you would use. Then, go through the list for each parcel, and complete as many as you are able to. If some are already fully cleaned, the dragon simply drops less treasure. Naturally, flavoring such as these opposing delver's teams getting to it first (since your party was so busy dawdling around stomping level 1 encounters) will help to add to the sense of the dungeon being "bigger than the party".

Hope these suggestions help.