Tome of Adventure Design

So I'm a longtime gamer but "no-time" OSR player (having spent my youth roleplaying with games like MERP, Champions, West End Star Wars, etc. instead of traditional D&D). Never played D&D till my mid-20s with 3.0 and then 3.5. So, anyway, now that I have introduced myself, let me say that I finally have the opportunity to run Swords & Wizardry with my home group (had the delight to play some Mythus Tower with Matt Finch at Gamehole Con last year). I'm going to let them run 6 traditionally-generated characters through Hall of Bones and then -- I'm not exactly sure.

I own a lot of industry-generated adventure content, but I've always preferred to make up my own stuff. So I've been using Finch's awesome Tome of Adventure Design to "randomly" craft my first home brewed old school dungeon. Finch provides a system for generating rooms and grouping them into arrangements called "Dungeon Areas." I'm certain there can be multiple areas across a dungeon level, and I imagine that a single area could comprise a single level, if the Referee so chooses. But my question is this: does anyone have any guidance on the average size (or depth) of a dungeon and how many rooms, typically, per "level"? The published adventures I own tend to be quite conservative on the small end in dungeon size. An exception of course is Rappan Athuk or The Black Monastery. Any guidance, dungeon builders?

michaelsandar's picture

There's not a real hard and fast rule here. I think you'll find most published modules tend to have smaller dungeons (with a few notable exceptions) mostly due to space and publishing costs.  I think that a lot of it will come down to what you and your players expect to get out of the experience. 

Note that a lot of this is 'in my opinion only'... your players and GM style will vary some of this.

A good small dungeon has between 5 - 20 rooms / areas of interest. Something like this can usually be covered in a session or two, mapping is kept to a minimum, density is usually high.

Growing larger, two or three of these stacked either vertically or horizontally can make for a nice full adventure.  Density can still be high in mulitple areas, but there will also be sections of the dungeon that can be left unused - monsters don't have to be packed into every room.  Mapping can get trickier, there's room for some puzzles and sealed off areas.

Finally, a third catagory is the megadungeon.  These are the big guys, and break a lot of molds from time to time.  Take for instance, density.  Rappan Athuk is pretty dense.  Sure there's a ton of levels, but each level only has a dozen rooms, maybe two dozen.  One of the biggest difficulties in RA (and one of it's best features, imo) is the multiple paths that can be taken.  You can go from the ground floor, to level 1 to level 7, to level 5, to level 4, the 4a and back to the ground floor. 

Compare that to Rob Kuntz's Bottle City.  It's only one level but there are potentially dozens of rooms, hallways, chambers, and mini areas, though only about 20 are keyed.  This makes for a much more exploratory game, where the dungeon and map itself is a big part of the challenge.

Ultimately, a lot of the answer to your question is 'what type of game would my players like'?  My advice - pick the middle option and leave several areas unkeyed or sealed off until later.  If the players like the megadungeon angle, keep building and adding as you go.

Thanks for the advice. This helps a lot. Actually, last night I just finished a three-level dungeon, each level a single area averaging 5 rooms, stacked one on top the other. I took Finch's advice and left about half the rooms "empty" (this was hard, because after rolling odd things to be found in the rooms, I often elaborated them into interesting things). One of my first observations now about how OSR play differs from the "evolved" roleplaying I'm used to is how much OSR gaming is about exploring, finding things out, and the attention to small detail. My experience with other roleplaying has been a general idea of what could happen, much more improvisational, with the emphasis on NPC encounters nearly every time. The idea of puzzles and "unusual things" in dungeons, when elaborated, make the experience so different: I'm learning that I might be able to awe and entertain my players with something other than physical conflict. First session is tonight!

michaelsandar's picture

Just remember to bring backup character sheets.  :)

S&W is pretty deadly, particularly for 1-3rd level characters.  Assure them that they will get more powerful... but it takes time, effort, and no small amount of luck!

Start small but always leave room for more. 

The only limit is your imagination as a DM but starting with a small dungeon is the best way to get things going. A one level dungeon can expand in rooms through hidden passages, sealed doorways, secret chambers; it can creep downward through cobweb choked stairs, pits that open to guardrooms, long hallways, mazes, unworked caverns, streams that become waterfalls  and sink into the earth. The big trick is to not plan more than you can chew and execute and always stay one doorway ahead of your players.

Go mission-based--pick an objective, hit it and split like Jason says! Come back later and hit the next one. Bring lots of gold-plated coppers, food and oil to make barriers to let you escape. Don't underestimate wizard lock and hold portal as monster-prevention methods.

Thanks again, everyone! We'll see how it goes from here.

Everyone had a good time the first session. It helps that I convinced a veteran OSR player into the game; he helped guide the new players. But I was running Hall of Bones, it was near the end of the night, and no one appeared to have any ideas about how to proceed into the dungeon -- or if it proceeds at all. (I'm not going to give it away here, but the players have to work through a Trick before being able to proceed -- I think I read somewhere that that was a bad idea in dungeon design for just this reason: puzzles or concealment can stop short the dungeon delve.) I'm totally fine with it, because I've gone absolutely bonkers designing dungeons! I want to entice my players into checking out The Octahedron Temple of Horror. Or maybe they will prefer to take a job preventing an assassination at the end of The Branching Gardens of the Shattered Princess?

So it looks like I have jumped away from the advice here. Or maybe not. The last thing I intend to do with preparation is to build two more dungeons at the outer edge of a hex crawl area. While whiling away their time in the tavern The Wizard's Maid, they will be assailed with choices. We've already learned that they can go back, delve as deep as they can, retreat, gather resources, try again -- but that the world is not static, and the dungeons certainly may change in their absence.

I also want to point out the best resources I've been using: Matt Finch's Tome of Adventure Design, The Mother of All Treasure Tables, Richard J. LeBlanc Jr.'s D30 Companions, and some name charts from Kent David Kelly. Thanks, everyone!



All right, veteran dungeon designers. More questions...

As I've mentioned, I've been going bonkers with dungeon design. I'm like a kid with a new toy. (I am a kid with a new toy! I'll probably settle down in time.) I've read somewhere (I'm pretty sure it's Finch's advice) that half of the rooms in a dungeon are supposed to be "empty." "Dungeon dressing" is what's used to make them a little more interesting. Okay, well, good. So in time, after I had flow-charted a dungeon level, I used a d8 to determine what might be in each room: 1-4 nothing, 5 trap, 6 monster, 7 trick, 8 unusual.

Okay, so, I was using this method, even though I felt like there were A LOT of traps being generated (Finch again, in my memory, saying don't use too many traps -- it will bog things down), and then I stumbled across the following in Table 3-54: Contents of Special Rooms: "Special rooms are simply the rooms with more detail to them than just monsters, treasure, and/or bits of dungeon dressing. ... In some cases they might incorporate tricks or traps." And the table provides results for Trap, Complex; Trick, Architectural; and Trick, various.

So my question is what do you people think should be balanced ratios of elements in a dungeon? Do any of you have a table or rubric for placing this material? As a final and related question, what is, after all, the average ratio of Monsters to Rooms in your average dungeon? A related question focuses on Finch's mention of monsters, treasure and dungeon dressing being the typical elements of a dungeon room: should these, instead, inform my random die roll? A d6, perhaps, with 1-3 signifying "nothing"? Here's one of the reasons why I ask these questions: I just got done completing a "Level" (as I understand it) totaling 27 rooms (comprising 5 areas of various numbers of rooms); I rolled for Monsters for Dungeon Level 2 in the Complete Rulebook and got...

1 CL 4 creature.

What do you think? Should these rolls actually be made for each "area" of a dungeon level? I rolled four more times and got much more satisfying (though dangerous!) results.

jmj_1975's picture

In regards to Traps, but you can apply the advice to anything really, does your group enjoy locating and attempting to deal with traps, or occasionally missing a trap and having to deal with the consequences?  If your group has a character who is built (however well or poorly) to deal with traps and wants to take the scouting role, so the entire (noisy group) doesn't run into a room of six things -- where they'd be lucky to be able to deal with two of whatever is in that room... then you have a character who expects traps and wants the scout role.

With a dedicated trap/scout character, you want to give that player some spotlight time to shine.  But you don't want the other players to grow bored.  If your session is 25% spent on this one player, with the remaining 75% on everyone evenly (including the 25% player), that's not fair to everyone else.  You could handwave time, and only play out the important bits, once a direction has been chosen.  If you do have the trap guy, spend about as much game time on each of the others in a role that they're designed to shine at, if at all possible.  If the warrior doesn't have an area of expertise outside of combat, where everyone plays equally as far as time investment goes, you can still design a sidequest to run concurrently with the main plot that is of specific interest to that character -- let her shine too.

Conversely, if the group has no trap finding character, and gets annoyed at traps going off that they cannot deal with, you have two basic choices.  

In a status quo scenario, the dungeon is trapped whether the group has a trap finder or not.  In that case, maybe have them meet an ex-adventurer who has been there, or something similar and let them convince him/her to return for a fee or whatever.  In that case, you don't need to play out the scouting at all, just have buddy return and say two hours went by and he said you can go towards the den of the Orcs or bypass that and end up deeper within a chamber filled with impossibly life-like statues... occasionally have buddy tell the group to hug one side of the chamber or whatever.

In the tailored option, your dungeon is constructed according to your group.  If the group has no trap guy and doesn't enjoy traps, then don't include traps or if you do make them especially sparse.  Maybe the group encounters three traps enroute to the boss's room to give the group fair warning that there are traps, but to not place forty traps along their route since they cannot deal with them easily and don't enjoy enjoy them.  The forty being how many you would have placed over a five level crawl, with a group that enjoyed traps.  Then have a couple of traps that the villain uses and is aware of, in his chamber of choice for dealing with adventurers that interrupt his work.  The party has been made aware that traps are a possibility, not simply had three traps sprung upon them without any kind of a trap up to that point... sure they may not expect a trapped pit or acidic gas to spray if you use the column for cover... but they're at least aware that there are some traps within the dungeon, so you're not springing stuff on them out of the blue.  Plus if your group is level six at this point, and you feel a really tough single creature is CR 10, then you can enhance the difficulty in going with two CR 8 options, where one of those is your boss, and the remaining experience value for the other CR 8 is a bunch of CR 4 things including the trap.  Multiple baddies are a lot harder to deal with (due to the action economy), than a single big thing.  If your group is level 6, and an EL 10 battle is too easy then make it 12 or 14; or conversely scale it down, if that would be too hard.  Having six mooks on the field of battle, supporting the boss also gives the AoE wizard/caster type an opportunity to shine and semi-punishes a group that ignores that role.


Thanks for the advice, Jason. To clarify, until (perhaps) I learn better, I'm not tailoring my creations to the composition of the PCs. Why I'm running Swords & Wizardry and not something like, say, Pathfinder, which I ran for about two years until last spring, is because I want the "OSR" experience. Players aren't even getting a lot of choice in who they play: each 3d6 roll is assigned to Attributes in order; from there make the best choices you can for Class and Race, if you want. Maybe don't bother naming them -- high likelihood of death -- and roll up three characters, run as many as you want in any single session. In some ways I'm trying to create a DCC "funnel," I suppose.

And by the same token, what I'm trying to design is a "balanced" dungeon. I wouldn't neglect Traps just because there is no Thief in the party (though there is). The way I understand OSR is that PCs either have the resource or they don't -- that's part of the game, the "tough luck" component (certainly accompanied by an evil grin or a cackling laugh from me, the Referee). And this is where Hirelings are supposed to come into play, right? Think a Thief will be useful next session? Well, you can always hire one! Well, that is as long as you have enough gold and Charisma to pay them and keep them. And that first resource, certainly, is in short supply at level 1. Again, part of the challenge of the game.

Since posting here last I kept digging and came up with dungeon stocking advice in Labyrinth Lord. Daniel Proctor recommends a d100 roll with the following results:

01-30, Empty 15% chance of treasure; 31-60, Monster 50% chance of treasure; 61-75, Trap 30% chance of treasure; 76-00, Unique, Variable

This seems clean to me: equal, 30% chance that the room is empty or contains a monster, with a bit of a possible reward (with some looking about) that the empty room contains some treasure; equal, slightly less 25% probabilities that the room contains a Trap or something else entirely. I think I'll be using this Table from here on.

And oh yeah, one more thing about traps. Unlike my experience with Pathfinder, wherein for the most part only the Thief could search for traps (and because of, you know, Skills, always find them and usually disable them), in SW most anyone taking a Turn to search an area has a chance to find them. Also, in Pathfinder, if a Trap was sprung, it seldom touched anyone (Save to avoid or half, which I suppose can be true for SW too), or if it did, 1d4 damage or whatever was negligible for a character with full hits and many resources. In SW, as I understand it, there's a chance that a trap might never be triggered (2 in 6 chance to trigger).

Another session tonight! We'll see if I can put these principles into practice!

I'm not a believer in featureless rooms. They can be empty if you mean they have no monster or treasure but I always like a room to trigger some type of sensation be it sight, sound or smell. I remember one set of rooms where the player's find it completely empty except for a bloody handprint, then the next has a dried smear of blood leading to a closed door that opens to  a hallway with the smear o blood down its center leading to a closed room. Inside the room is a bedchamber with bloody hand prints all over and a bloodsoaked bed, there is a pool of blood near the bed and footprints lead to a far door, the door leads to a chamber and the bloody footprints lead to a large wooden chest in the middle of the floor, the only thing in the chamber. My players were abut ready to fireball the chest. An empty room shouldn't just be an empty  room, it should have a reason in your design for being empty. It needn''t have anything in it, even blood, but the emptiness shouldn't make your players complacent either, unless that is what you want them to be.