Megadungeon design - first principles

The Caverns of The Rift are a megadungeon. There's no real question about that as far as I can see. It would be extremely difficult to work out exactly how big it is, though Level 1 contains 12 different sublevels or zones and 443 rooms, Level 2 contains 16 sublevels and 695 rooms, Level 3 is 14 zones and 499 rooms, and Level 4 is 4 sublevels (one of them huge) with 359 rooms. All those levels have corridors that go of to areas waiting to be further expanded, if necessary. Weirdly, the PCs completely bypassed Level 4; having just about started to explore Level 3, they decided it was too dangerous and went Level 5 instead. No, I don't understand it either.

The first 4 Levels are pretty much completely fleshed out with just over 2,000 rooms (maybe 50 of these are just notes rather than full descriptions)... 5, 6 and 7 are substantially finished (maybe 75%) and Levels 8 and 9 are partially sketched out. There is a staircase - it's almost legendary, and called 'the Endless Stair' - that connects many levels directly, but the PCs haven't found it.

Level 10 is a vast subterranean world, a kind of Underdark-ish hexcrawl (more a pointcrawl really) called the 'Abyssal Realms' (there's even a blog label for that, though to be fair I haven't posted about it for a while). That is done differently; I haven't detailed individual rooms for most areas in the Realms, more like encounter areas, but it contains something like 60 unique locations. These range in size from individual small encounters to subterranean cities. Unlike the caves of the Rift, which are quite tightly-constrained in an E-W direction (the caves the PCs have gone to stretch for about a mile along the south side of the Rift), and somewhat in a N-S direction (the south wall of the Rift is the northern edge of any tunnels, and they stretch... an undetermined distance to the south), the Abyssal Realms stretch for dozens of miles under the mountains in all directions. If the PCs make it there, the campaign will have to change, with the PCs' 'hometown' changing from the above-ground Rift City to somewhere underground that they can use as a base of operations (and place to pick up new party members). There are a few contenders for a new underground base for further exploration. But as I say, the PCs didn't reach it, and may never, as the Rift City campaign has ground to a halt. 

So, what's the point of it all? Originally when I conceived of the caverns in this huge valley called 'The Rift' I had in mind something like the Caves of Chaos, but 500 times bigger. A series of caverns over many levels that were linked together, that had factions and stories and lives, in the side of a vast valley. That seemed like a cool place to go adventuring, and provided one of the necessary elements that I'd identified for the game - an adventure site that would be suitable for repeated visits from a local base. That implies a megadungeon.

Image from Dream by Wombo, using prompt 'valley with caves'

This I think is actually my fourth megadungeon. The first was something I started nearly 30 years ago and it didn't get very far - a Dwarf complex somewhat inspired by descriptions in Magician by Raymond Feist, and Song of the Dwarves by Thorarin Gunnarsen, that I was reading in the early 1990s, I think. I had a few plans but did very little work on it. That was folded into my second megadungeon, Silvergate, the abandoned Dwarven city that got somewhat further but has still stalled (or maybe been abandoned... by me rather than the Dwarves this time, though it also has a blog label). I think I'll probably go back to that and try to push it forward at some point though.

The third megadungeon I think never even made onto the blog. I don't know why to be honest, it would have been good to throw some ideas around here. As it is, it exists only on Facebook. Again it's a stalled project, an attempt to co-operatively build a megadungeon. There's been some serious work put in by one contributor - not me. It's called 'The Labyrinth of Nodnol' (link here for those on Facebook) and is based on the London Underground/Overground map.

Then came The Rift. A few of us were talking about starting a D&D campaign and I offered to DM it. I didn't really have any ideas other than that the structure needed to account for people not being able to make every session, so the idea that the dungeon needed to be near the PCs' home base came about. This would allow them to make periodic visits to the dungeon, rather than having a campaign based on travelling from one dungeon area to the next (or even, travelling overland to a series on minidungeons). Perhaps that will be the format of the next campaign; perhaps it won't. The other option of course is just a huge underground exploration, but I wanted to get round the problem players not being able to make consecutive sessions and their PCs being stuck in limbo while life went on a round them. So, I decided on a structure that says that the PCs get the hell out of the dungeon between sessions.

Constantly travelling to the same place also implies lots of entrances. The PCs discovered (I think) 16 entrances into the caves, which variously take them to Levels 1 (7 entrances), 2 (4 entrances); they also discovered one entrance each for the 3rd, 5th, 6th, 7th and 8th levels. As I say, this is from memory, there may have been more they discovered.

I used a variety of methods to generate the content. A lot of it has come from some random dungeon generators, particularly Donjon's (link) AD&D Dungeon Generator. Other Dungeon Generators are available, and some have been used, including the old WotC Generator (like much of the free content on Wizards, this seems to have been taken down sometime around May 2022) and Save versus Total Party Kill's 'Dungeon with x Rooms' (link here) - it only generates short room descriptions so I've also used maps from Dyson (link) and Paratime (link), among others, as well as creating a lot of my own. All of these sources have been smushed together to create the (approximately) 2,400 rooms so far detailed/mapped out in the Caverns of the Rift.

But, now I'm going to start a new megadungeon, though as the Rift Caverns had 2,400 rooms and were not finished, the projected new megadungeon with 365 rooms looks a bit tiny. This new megadungeon is my response to the "Dungeon23" project. This is a kind of design challenge that basically asks for a dungeon-room a day, every day through 2023. The original post is here, though I found it on a Facebook post from someone else. There is a hashtag, #dungeon23, but I don't do twitter and that probably hasn't worked anyway. In the original post, the 'rules' are pretty simple - "Megadungeon for 2023. 12 levels. 365 rooms. One room a day. Keep it in a journal."

I don't have a nice journal, but I do have a blog. I'm not promising to post every day but I will try to post much more regularly than I have been doing. This timetable does give momething to aim at.

I first read the original post today, but heard about this a couple of days ago. The original says (words to the effect of) 'generators are fine, the point is to get it done and produce something'. But, I've built a megadungeon (substantially) with generators. I want to exercise my grey matter. I *could* just go to 'Save versus Total Party Kill' and ask the dungeon-room randomator to produce a 365-room dungeon. Then all I'd have to do was map it. That would save loads of bother. Or I could get Donjon's generator to produce 12 dungeons of about the right size and tweak anything from there. Both of these would be valid ways given the original guidance to produce the required output, but I want to see if I can do this from the power of my own brain. I may well at some point fall back on the generators, but I want to resist that as long as I can, so I can get an idea of what I can do by relying as much as I can get away with on my wits.

I have an idea about the overall design and how I'm going to 'pattern' the rooms/days, and I have an idea about how I'm going to populate it wit monsters, traps and treasure (it will involve using Moldvay pretty heavily) but I have almost no ideas about what I will actually put in as content, except that room/day 1 will be an entrance. That's all I know right now, except for one thing that I will tease you with because I don't want to say just now. I have for a long time wanted to construct a certain type of dungeon and this will be that type. That's all.

Hopefully come the New Year I'll start posting more about this. In the meantime, happy dungeonering, folks!

End of an Era

 Well, it looks like the Rift City campaign is no more, and 'The Wandering Monster Table', the experiment to bring open-table D&D to Leicester is over, after a little over 5 years.

Brigham's Mule Sanctuary, his retirement plan, interpreted by 'Dream by Wombo' (look it up, I've had fun playing with it) AI art generator

I've mentioned in previous postings on the blog that it has morphed into something quite different from the original conception, becoming something more like a regular gaming group, and also (primarily because of covid, but also due to to work commitments and people moving away) migrating from meatspace (and not just face-to-face but in a publicly-accessible space) to online. However, the Rift City campaign was conceived as a drop-in game and in theory at least remained so until the end. And that was the point really, to give people a regular game that didn't involve committing to be at every session.

This was why I adopted the structure I did for the campaign - adventuring in the Rift took place during the day, and at the end of the day the PCs high-tailed it back to town to lick their wounds an count their loot. Then, 'the next day', a potentially new constellation of adventurers would brave the caves and return in glory or defeat.

It's been fun (mostly), and some hard work to keep it going. But over the last couple of sessions, only two players have turned up - Brigham and Ugli, who were the players of Polly the Magic User and Gibbet the Thief in the very first session. We decided if no-one else wanted to come to the next session, we'd knock it on the head. I put out a call on Facebook (most of the players, something like 25 of them, who'd been to previous sessions are on there) but no-one committed to coming. So, I've pulled the plug.

Thanks players for turning up and running through the games I put on, I hope everyone had fun... but it's time to find other things to do every first Sunday of the month. Ciao!

Rift City Campaign Session 60 - out of the Rift!

I'd like to start with an apology to anyone still reading this. It seems I've put two posts on the blog in the last year. I've been very bad at keeping it up to date. I originally started the blog to discuss fantasy gaming, mostly Warhammer and related topics, but after I did, I started playing D&D again and that became the main focus. But lately I've been posting neither D&D nor miniature-gaming content. So, I'm sorry to anyone who feels disappointed. Perhaps no-one is, in which case, that's OK, but in case anyone is, you have my apologies, and at least the expressed intention to do better.

I'm not really sure why it's tailed off. I have been continuing to run the Rift City Campaign, and messing about with a bunch of other things; I just haven't been good at talking about them. But, with the Rift City Campaign, and the 'Wandering Monster Table' project, now 5 years old (we had our first session in August 2017), it seemed like a good reason to sit down and say something about it. So I will.

First, the 'Wandering Monster Table' itself. This was originally conceived as a monthly drop-in game in a (relatively) public space. Our first six sessions were in a pub, on the second Sunday of the month; we had to move for a month when there was an event on for one month (there's a comedy festival that takes place every February in Leicester, and the pub was a regular venue); we met at the house of a couple of people from the group for one session; but after the festival, we went back to the pub for about another year.

Then, the pub we met at closed down and we had to find another venue. So, for a couple of months, we went back to the house of the players we'd been to previously; then we found a different pub, for about another six months.

After this however, I got a new job that involved travelling on Sunday evenings. This meant that we had to move the session, to Sunday afternoon, when most pubs are serving food and much less keen to let some weirdos use a couple of tables. But shortly after this, Covid happened and all the pubs were closed anyway. Also, people couldn't meet face-to-face. So, we started using Discord.

So for half of its life, the 'public drop-in game' has existed as an online game instead. It's still a drop-in game, in theory, but I don't think we've added any players online that hadn't come to face-to-face sessions. I'm not saying the project is a failure - it certainly isn't, it's the longest-running campaign I've been involved in bar none; three people came to the first session (apart from me) and two of them are still regular players; and hopefully people who've played in it have enjoyed it (I certainly have) and will continue to enjoy it; but it must be admitted that it doesn't really fulfil the original brief any more. But, that's OK; who in the balmy days of the Summer of 2017 could have predicted Covid? Not me at any rate. Maybe, at some point, we'll get back to face-to-face meetings in a public place; maybe we won't.

Anyway, that's the 'Wandering Monster Table'... what about the campaign itself?

At the previous session (in June) the PCs had decided that they need more magic items - specifically Bags of Holding, as they'd found what turned out over two sessions to be several dragons, and were trying to find ways of carrying back the piles of coinage that tend to be found in dragony hoards.

There are no 'magic shops' per se in Rift City. Not quite true; there are no general magic emporia, but there is an Alchemist who has opened a potion shop, and potions are sometimes available from some of the temples too, but the supply overall is erratic. Perhaps I've done this 'wrong'; maybe it would have been more sensible to have a magic shop in Rift City, but my thinking has been that magic should be somewhat more difficult to access than say torches or iron rations. It *should* make it more mysterious and 'special', but perhaps it's just more annoying.

However, the party knows that there is a more substantial city about four days' journey away. 'Rift City' is a somewhat rough frontier town, built on the momentary profits of exploiting the adventurers coming back from the monster-infested tunnels in the Rift - it's sort of like a gold-rush town. It hardly has much infrastructure. The 'big city' is actually a real city - much bigger than Rift City and, the PCs hope, with better amenities. 

So, I worked out how far away it was and estimated how long it would take to get there. In effect, it's four days' journey through mountainous terrain, unless something goes very wrong. I wondered about this, rolled up some encounters for different times of the days involved, and knocked together a rough-and-ready plan of action. My estimate was that we could do the journey in one session, then the PCs could have as many sessions as they liked in the city, and then come back to the Rift if they wanted to, or use the city as a basis to go off elsewhere.

Illustration I found googling 'fantasy mountains', to post on the group's Facebook page: © Wizards of the Coast by Alayna Danner

I've discussed at various times on this blog what a 'quest' campaign would look like and how it could be made to work with an open table like The Wandering Monster Table - most thoroughly in this post from March 2018. The short version is, as long as the PCs finish the session somewhere relatively 'safe', where people can chose to stay and new people can join the party, an open table should be able to accommodate a campaign involving PC travel. The trick is ending the action at a point where there can be a reasonable 'changeover'.

As I mentioned, I'd estimated that the journey from Rift City to the city called 'Selen' by the inhabitants of the Rift could be accomplished in one session. I really should remember Helmuth von Moltke's dictum 'no plan survives first contact with the enemy', and not just remember it, but also rigorously apply it to D&D as soon as I get any barmy notions like 'the PCs will probably do this, it should take that long'. No. Doesn't happen.

First, though the players were all agreed to go to Selen, there was no consensus on how they were going to go. I had assumed (me and my assumptions) that they would just set off and go along the road. But, instead, they decided to attach themselves to a caravan. That's OK, I know caravans come and go, some of the early plots and hooks around the Rift concerned the raiding of caravans by Kobolds and the theft of supplies destined for Rift City. Caravans are definitely a thing in the vicinity of Rift City.

So I rolled up a random caravan. I don't have any specific rules for creating caravans - maybe I should write some - but there are Merchants in the D&D Expert book. There caravans can be of different sizes, so I randomised creation and rolled one up. I got the largest size of caravan I could, about 20 merchants and 80 guards; but I don't see wagons making it through the mountains very easily (even with the relatively-OK roads that I assume exist... it doesn't matter why but 'OK roads' seem reasonable in context), so I decided this caravan would mostly be mules - a giant mule-train, basically. I randomly determined when this caravan would be setting off, and as luck would have it, the PCs only had to wait one more day.

Some of the PCs spent this day buying up all the gems and jewellery and platinum pieces they could get their hands on in Rift City, the idea being to take portable wealth to Selen. Other PCs, not so much. They mostly just stuffed money into sacks.

Some of the PCs decided to sign up as guards for the caravan. They negotiated a price with one of the merchants to provide two armoured humans, two armoured Dwarves and an armoured Elf. The party at this point consisted of Brigham and Halvor (Clerics around 8th Level), Heedor and Ugli (Dwarves around 8th Level) and Jade, a new PC (Elf of 7th Level). The price was lower than the player doing the negotiating had hoped for, but reflected market conditions; pay (3gp each for the humans, 5gp for the Dwarves and 6gp for the Elf) represent their cost as 'Heavy Foot' mercenaries. Only not for a month, for four days. PCs have to provide their own food rather than it being provided, and there is some danger on the way for sure; it isn't garrison duty in a sleepy backwater. So the pay is much higher than the minimum to get soldiers. But crucially, there are lots of people who could do the job in Rift City (the place is awash with 1st-Level Fighters, Elves and Dwarves) and the organisers of the caravan weren't going to pay loads for 'special' guards. So the five PCs were offered 22gp to be caravan guards.

However, others of the PCs decided it would be fun to get mules of their own; so, some of the other members of the party bought seven mules, mostly to carry large sacks of gold.

This led to some problems with the caravan. The organisers had agreed to pay for five guards; but on the morning that the caravan was due to depart, two guards and three merchants with seven mules turned up.

A quick re-negotiation with the merchants led to a revised offer; 8gp for the Elf and one of the Human (Cleric) heavy infantry; the remaining PCs paid 15gp for up to two mules to join the caravan (the PC with three mules had to pay 30gp). Also, these PCs hired a mule-handler-cum-extra guard, a Dwarf called Gami.

At last the PCs and the caravan were ready and the whole travelling roadshow set off through the mountains.

According to my rolled encounters, there was nothing in the morning. So, the morning passed uneventfully. I had rolled both an evening encounter, and a situation of some consternation at the end of the day. I'm trialling some rough wilderness creation rules for precisely this sort of situation, and my results were that there would be an encounter with snakes (Giant Rattlers, in this case) and the 'safe' area at the end of the day would be a village of ... Troglodytes. Hmm. Think my procedure needs tweaking a bit.

The snake encounter was fine; as originally I'd done the rolling up when I thought just the PCs were travelling (Helmuth von Moltke is spinning in his grave I presume) there was always the chance that the encounter I planned for them would have been triggered by some other part of the caravan. I just asked the PCs where they were in the caravan ('front, middle or back') and assigned a 2-in-6 chance that that was where the snake attack was. And, it was. Had it not been, I'd have said 'there's a kerfuffle up ahead/somewhere behind you...' but it was nearby. The party toasted the snakes - literally, as far as I remember, I think Jade used Fireball - and that was that. I think someone ate some barbecued snake but can't remember who - probably one of the Dwarves.

My assumption had been for the village at the end of that, that there was a (friendly) village, but the Trogs had overrun it. So, in the evening, as they approached the village, the caravan came across some bodies in the road. As so often happens in situations when I need 'people' quickly, I turned to Donjon's random generators - this case, the NPC generator. I must have also rolled for how many bodies there were, because there were definitely three of them. I read the first three entries, and from the descriptions gave the PCs an idea of the people they'd found - two middle-aged women and one younger woman. I also told them the dead women seemed to be fleeing the village and there was a horrible stench around. Several of the PCs have fought troglodytes before (everyone except Jade I think) so I thought it reasonable that they would recognise the lingering smell.

The PCs decided to Raise Dead on the young woman. It turned out that her name was Astal, she lived in the village with her parents who were farmers, and she had fled when the stinking lizards had attacked and had started running to the north hoping she and her companions could have made it to Rift City, but were caught and brutally attacked... Raise Dead leaves the raised one with 1hp so the PCs decided that Astal should sit on one of the mules and come back to the village.

As the caravan got there, the PCs seemed reluctant to go in (I described it as having an earth bank around it: in my imagination, it's a village inside an old disused fortification, and the road passed between the banks on either side). Some of the PCs turned themselves invisible, but they delayed going in for a while. In the meantime, some of the other caravan guards went inside. By the time the PCs had decided to follow suit, I ruled that the other guards had already dealt with the Trogs in the village so the PCs didn't get any experience for that (or the gems that made up the Trogs' treasure).

The PCs asked Astal if she wanted to go to her house, but she didn't - the potential sight of her murdered parents being too upsetting. So, instead five PCs, an NPC Dwarf caravan guard, a newly-resurrected peasant girl and seven mules all bedded down in a stable. They worked out a watch rota and settled down.

I knew there was going to be an attack of a particular kind during the night. There's a kind of monster from Deities and Demigods called an 'Astral Wolf'. It's in the 'Nehwon Mythos' section, and appears in the Wilderness Wandering Monster Tables for the campaign because, well, everything from the Nehwon Mythos section exists in my campaign. The PCs just don't know it yet because they've only seen the Rift so far. Astral Wolves attack by ambushing people in their dreams and forcing them into the Astral Plane where the wolves then attack them bodily.

To be forced into the Astral Plane, you have to fail a save versus Spells. Of course, all of the PCs made their saves, so they just had some bad dreams about zombie wolves pursuing them through a silvery realm. Gami the NPC Dwarf however, who was on watch, woke the party members to show them Astal, - a Normal Human with rubbish saves, who had therefore been transported to the Astral Realm, where she was attacked by the Astral Wolves, and, having only 1hp from her Raise Dead experience, had quickly succumbed to their attacks. So that was that (for the moment) - the Astral Wolf attack served only to give the PCs bad dreams but insta-killed the person they'd only just brought back.

The PCs went back to sleep as best they could but there were no further attacks that night. In the morning, the party rose and began prep for the next day.

And that's where we ended the session, it having taken massively longer to play that day than I expected. I thought Day 1 of the journey could be 'done' in about 45 minutes - it took nearly three hours. Proof positive that I shouldn't bother trying estimate this stuff, because the PCs will always do stuff I don't expect.

What will the next session bring? who knows? Not me. But I have an idea how the rest of the journey could proceed, if the PCs don't take things in strange and unexpected directions, which they might well; and, if and when they get there, I hope they'll find the City of Selen an interesting place to visit and to adventure in.