Megadungeon Mastery III: How Large Was My Level

(Continued from Megadungeon Mastery II: Rise and Fall of the Great Underground Empire.)

How large is it? Relatively large, apparently.

As often happens when trying something new, when I decided to build my first megadungeon I said to myself, “I’ll try to be different and original and do things my own way!”

Naturally, this led to problems.

Right now I want to focus on matters of scale. Ironically, this sounds like a brief subject but actually covers a lot of ground. I’ll start with a few things I learned about designing and drawing megadungeon maps.

1) Don’t make your levels too large. In designing the dungeons beneath the Chateau d’Ambreville, I sketched out a huge, elaborate castle and decided to put entrances under various towers. In order to fit all these entrances onto one level, I printed out maps on 11″x17″ sheets of paper, then folded them in half so they could fit into my binder. But there were problems!

• 1a) If you fold your maps in half, they’ll fray and tear along the fold. This is no fun, and requires lots of fiddling with adhesive tape to keep them together. It’s better to make smaller sub-levels that each fit onto one sheet of paper, and connect them with long corridors.

• 1b) Large maps can get out of control. Once you’re trying to fill in a huge map, you may realize that now your themed level now has 200 rooms and after filling in 50 of them, you’re stumped as to what to put in the other 150.

• 1c) If your levels are too large, it’s hard to keep track of what’s where. This can be important when trying to figure out how nearby dungeon inhabitants will react to the PCs and their trail of theft and murder! (I’ll include more detail on these issues in a future blog post.)

Note that you can always provide more level-appropriate encounters by making an additional level — a “sub-level” — at the same depth, connected to the rest of the dungeon by a single stair or passage. Multiple themed sub-levels can be strung together to generate the effect of a huge dungeon level while avoiding many of the problems inherent in huge dungeon levels.

2) Use graph paper with large squares. Maps are about more than walls and doors! They’re an invaluable resource for marking down other details: furniture, whether doors are locked, tracks on the floor, light sources, odors, etc. And at five or six squares per inch, there’s just no room for those details. Even the traditional four squares per inch may feel a little cramped. Sure, you can note these things on your map legend, but if too much of this information is written in the legend rather than drawn on the map, it’s far too easy for key details to slip your mind in the heat of play.

* * *

I’ve mentioned that the number of rooms on a level is significant. This isn’t simply a matter of aesthetics. Rooms contain monsters and treasure; in a sense, they’re buckets of XP for the player characters to fill themselves from. Furthermore, each deeper level is supposed to have tougher monsters and bigger treasures (on average) than the previous level, so that higher level PCs are encouraged to keep delving deeper to get the XP needed to continue leveling up at a decent rate. So each level should contain enough treasure for all of your PCs to level up if they clear it out — plus some extra to account for PC death, level drain and hidden caches that they fail to discover. Too little treasure and they can’t level up; too much treasure and they’ll hang around on that level instead of going deeper. (The specifics of how much XP comes from monsters and how much comes from treasure varies by edition.)

The problem here is that the size of the level roughly determines the number of encounters, which take time at the table and increase the chances of losing XP to PC death. So if you make a really big level, you either need to add more treasure or reduce the number and/or intensity of the encounters.

Reductio ad absurdum: You decide that you need to put 20,000xp worth of treasure on the first dungeon level to ensure your PCs can level up. You’ll distribute this roughly according to the OD&D guidelines: 1 encounter per 3 rooms, half of all encounters have treasure, one-sixth of rooms without encounters have treasure. (This comes to treasure in roughly one-quarter of all rooms.)

• In one instance, you design a 4-room dungeon level containing a single treasure: a 20,000gp gem. Unfortunately, there’s no worthwhile combat challenge available for this purpose; either you’re giving away treasure like Monty Haul, or the monsters are strong enough to kill the PCs (so why did you put this encounter on the first level?), or it’s so well-hidden that no one will ever find it. Failure!

• In another instance, you design a 4000-room dungeon level with 1000 treasures, each averaging 20gp. Aside from the difficulties of designing over 1000 first-level encounters, it will take your players forever to wade through enough of these encounters to get an appreciable amount of experience, and they’re bound to lose PCs faster than they can bring in treasure to level up. Failure!

So the amount of treasure on a dungeon level needs to be enough to level up all of the PCs, as modified by the number of encounters required to obtain that treasure.

You should be able to calculate this by determining two key variables: how many sessions you want to run before the PCs level up, and how many rooms your group typically gets through in one session. The latter number will vary a lot, of course; sometimes the party manages to cover large expanses of the dungeon by wandering from empty room to empty room, while at other times the party runs right into a big set-piece battle where clearing out a single room takes up the entire session.

So if your party averages four rooms per session (for example) and you want them to level up roughly every ten sessions, if one in four rooms has treasure in it, then you want ten of those treasures to be enough to level up. Moreover, if the party tends to miss about half of the treasures they run across (because the treasures are well-hidden, or they use up half the treasure on raise dead spells, or whatever), you could ramp up the treasures further, so that only five treasures will suffice to level up.

And remember: what’s right for your group isn’t necessarily what’s right for other groups. Some players enjoy mapping complicated dungeon levels, seeking out carefully hidden treasures or unraveling intricate tricks and traps. Others don’t. Unless you have access to a sufficiently broad player base that you can find players who like playing exactly what you want to run, you need to adapt your dungeon design to the needs and desires of your players. After all, the fun is the thing!


9 Responses to “Megadungeon Mastery III: How Large Was My Level”

  1. 1 OtspIII
    July 2, 2012 at 4:15 pm

    When calculating the amount of treasure needed for a group to level up (with setbacks), would you say that a dungeon level (including any sub-levels of equal difficulty) should try to contain exactly that much treasure or should that just be a minimum? The rate at which treasure hauls go up in value on lower levels should be enough to make sure the PCs don’t linger on safer levels too long, especially when those ‘safer’ levels still have Save or Die spider bites and wights and carrion crawlers ready to threaten the party with big losses. As long as the density of treasure stays at the ideal level, I don’t see a problem with a big level having way more treasure than a group could need to level up.

    Also, having a good deal of unexplored space on each level the PCs have progressed past is handy for those nights where only a few people can make it and the party doesn’t feel up for exploring quite to the depth they’re currently ‘on’.

  2. July 2, 2012 at 4:31 pm

    When calculating the amount of treasure needed for a group to level up (with setbacks), would you say that a dungeon level (including any sub-levels of equal difficulty) should try to contain exactly that much treasure or should that just be a minimum?

    Ideally, the “setbacks” you’re taking into account should be a worst possible case — TPK, missing several key treasure caches, etc. This buffer means that in practice, the PCs should actually be able to acquire significantly more experience points on each level than they need to level up, which allows them to skip ahead to the next level before the current level is fully cleaned out. (And if your worst-case treasure allocation proves insufficient — the party suffers multiple TPKs and new PCs lack the power to continue — that’s what level restocking is for.)

    The rate at which treasure hauls go up in value on lower levels should be enough to make sure the PCs don’t linger on safer levels too long

    Absolutely. Note that this is tricky when dealing with the first and second dungeon levels, since it costs the same amount of XP to go from second to third level as it did from first to second, so strict adherence to these guidelines would mean that the second level has the same amount of treasure as the first level but tougher monsters. You want to bump up the treasure on the second level enough to encourage PCs not to linger on the first, but if you bump it up too much, that might encourage PCs to linger on the second level rather than moving on to the third.

    Also, having a good deal of unexplored space on each level the PCs have progressed past is handy for those nights where only a few people can make it and the party doesn’t feel up for exploring quite to the depth they’re currently ‘on’.

    Note that this is also partially accounted for by the practice of level restocking. Even if you completely clear out the lower levels, eventually something else will move in, and an understaffed party can go pick on those new inhabitants.

  3. July 2, 2012 at 6:39 pm

    Yes, there are quite a few connected bits in dungeon stocking—far more than I realized before digging into the topic. If you’re making a dungeon for your own group, you know the number of players. Trying to publish a megadungeon for groups of unknown and various sizes would get pretty sticky. Restocking may be the best answer to too little treasure; awarding fraction experience to a party on too shallow a level may or may not a good answer to too much treasure.

    P.S.—Your “continued from…” link is broken. It should be this.

  4. July 2, 2012 at 6:43 pm

    @Paul: Good point about the publishing problem! I think that can be handled with some basic guidelines in the dungeon notes, however.

    Also, thanks for catching the broken link! It’s fixed now.

  5. 5 Charlatan
    July 4, 2012 at 2:19 am

    I just want to say thanks for the insightful and practical series this is shaping up as. I eagerly await the next installment.

  6. 6 John Wood
    July 20, 2012 at 1:23 pm

    Eric, my imagination is going wild with ideas of filling this map for a Gamma World 2.0 campaign that I am running. Any chance you could send me a file for this map? I am giddy about the thoughts of 10 different factions embroiled in a civil war inhabiting this underground world you have created and our heroes needing to navigate with brains and brawn through it.It would be so awesome!!!

  7. 7 felipe
    January 24, 2014 at 8:03 pm

    The game show host’s name was Monty Hall.
    Monty haul is the game situation that was then associated with giving away lots of prizes (treasure).

  8. 8 Jim
    November 10, 2014 at 3:28 am

    Would you email me a copy of your map?

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Past Adventures of the Mule

July 2012
« Jun   Aug »

RPG Bloggers Network

RPG Bloggers Network

Enter your email address to subscribe to this blog & get email notification of updates.

Join 1,053 other followers


%d bloggers like this: