Sunday, August 28, 2011

The Secret Door Tell

I don’t like blocked text.
I prefer really sparse descriptions for room text.
Just let me improvise.

These seem to be common refrains out here on the interwebs.  I don't mind improvising adventures, characters, plots, but I struggle with improvising dungeons well.  For instance, take the ubiquitous secret door:

The room is mostly empty, except for a door directly across from the way you entered.  Unlike the last empty room you entered, this one has a torch sconce on the right hand wall.

The map indicates there's a secret door in the room - I see that, and need to decide whether there's a tricky way to get it to open.  How often do you quickly improvise a unique way to open the secret door, and then introduce that element into your room description at the table?  Or do you just describe it as an empty room, and if the players happen to search the right-hand wall and find the secret door, their roll to find it includes finding the way to open it?  Or do they have to mess with the sconce?

If you add a clever way to open each secret door (like the moose head on the wall, or the sconce) you just telegraphed to the players to fiddle with the moose head.  If you just rely on a secret door roll... yawn.  Might as well play 3x or 4E and use spot checks.  Ostensibly, the paradigm of old school play shifts these types of problems to player skill, but that requires detailed DM descriptions to allow the players to push knobs and levers in the environment.

I'm finding that detailed room descriptions with lots of things to interact are the exact opposite of ultra sparse descriptors- and 33% of the dungeon are empty rooms.  There's an appeal to sparse dungeon keys and stripped down maps that requires a lot of DM improvisation.  My favorite old school experiences have been when the rooms are detailed enough that even empty rooms have something interesting to do in them, and this avoids telegraphing hidden treasures and secret doors.  I'm using secret doors as an easy example, but is really just the start of a larger conversation about improvisation versus pre-scripted content, and the level of detail (or lack thereof) in the two styles.

At this point, I'm just looking for recommendations from fellow DMs - how do you handle empty rooms, dungeon dressing, secret doors and sparse details in your own games?  I'll consider rephrasing this as a poll, later, but am interested in hearing how folks handle improvising the minutia.  I've seen some random tables on dungeon dressing and secret doors but would love to get shout outs to your favorites.  And empty rooms and empty hexes might as well be interchangeable.

If this were a poll, these are the types of answers I might foresee:

a. I develop some notes for empty rooms, traps and secret doors ahead of time (random tables or not).
b. I generate details during the game with tables.
c. I Improvise details for empty rooms and secret doors on the spot .
d. Empty rooms are empty rooms and secret doors are found with a die roll.


  1. I wrote a book.

  2. Furthermore:

    I think (just like my insistence on player agency for traps) that the same goes for secret doors. The simple answer is use the table from knockspell. :-)

  3. My answer would probably be a cross between "a" and "d." I don't have the aversion to "spot" rolls that many GMs in the OSR seem to have, but I do like rooms that have "stuff" for the players fiddle with and lend some color.

  4. -C - I'll check that out, thanks much for the link. But I'd still have a question - improvise, generate in advance with the table, or generate at runtime with tables.

    Oh - and what is the knockspell reference - a table in one of the mags?

  5. I go by method A, using the awesome book by Mr. -C there.

    I also keep a few pre-rolled descriptions generated from this table handy for when there's a door or trap I haven't detailed beforehand.

  6. I think there need to be truly empty rooms to balance out the rooms with things. Players can't tell whether a detail is being given to them because it A) is important to the dungeon's history, B) is important in finding the MacGuffin, C) is important to avoid the fireball trap they're about to set off, E) is there as dungeon dressing.

    So to place detail in a room because you don't want it to be empty can cause problems for players trying to parse the shared world.

    An empty room clears the mental palate. Players will be that much more observant when they find something. But I always telegraph my traps so my players will be paying attention because of the corpse on the floor.

    Secret doors I try to set up so the players will seek one out because of the map layout ("there must be another room here"), or roll to see if they notice something like a draft. I think I got that idea from Ian at Magician's Manse. I admit neither of these is a perfect solution to the problem.

    One kind of detail I think you can get away with more is thematic, so if the dungeon is wet: various things ruining in the wetness, mildew, puddles and such. Step out of the theme and player's attention will perk up: "Why is this pile of dry cloth here?"

    I think my answer to your question more specifically is that I figure out important stuff before game (trap triggers, clues) and improvise environmental detail. What this means is that my players can probably tell what's important to pay attention to by how I relate it to them. I think I'm okay with this. The other option is to have them fiddling around forever with things that are essentially there for the color, right?

  7. I see where you're coming from, Telecanter. The "disinterested" referee should care more about camouflaging the secret doors and traps; you're approach implies that if the players are listening and thinking about what the referee is saying, they should have a reasonable chance of figuring something is going on. And the only way to separate the noise is if there are some actual empty rooms with minimal dressing.

    It's a fair point; I had a hearth "big enough for a man to step inside" in a recent game, and that line was enough of a hint that players later came back, extinguished the fire, and searched for a secret door. Distinguishing a man-sized hearth from a regular fireplace was enough of a tell.

    The nugget about actual empty rooms and mental parsing is a good one to lift though; there's still hope I can learn some things out here.

  8. Of course, if you look at the Holmes rules (like I've been doing), finding secret doors functions as an "always on" automatic ability (for elves, at least) that the DM must check for during movement even if the player isn't actively searching:

    "If elves pass by a secret door or passage, roll a six-sided die and a 1 or 2 means they sense something there. If the party is searching for a secret door then an elf will locate it on a roll of 1 to 4, other characters on a roll of 1 or 2. Of course, the Dungeon Master will lessen these possibilities in lower levels of the dungeon."

  9. To each their own. I'm happy to play any style of descriptors.

    However, for my own DM writing I keep it as concise as possible. The general title such as "bedroom" or "library" gives the DM all the basics for embellishment if needed. I do include a quick note for smell and sound in addition to a brief visual description because it's an important part of player immersion.

    Random dungeon crawls are fun, but are, by definition, easy to make on the fly. Therefore, when I make adventures, I never do them. Having recently bought and read LotFP, I found myself agreeing with the sparse use of "monsters" and cohesive adventures.

    When I get the germ of an idea for an adventure I mentally play with it a few days. Next, I draw the map of the locations and flesh out some characters. Third, I write a description for each area based on my understanding of that germ. So, all the areas have to hang together with this one idea. Rooms remain empty if the story in my mind dictates that they should be empty. As a result, many of my areas have no special item or encounter, per se, because "special encounters" or items are created through the synergy of the overall story, DM's abilities, and player immersion not solely by (the writer’s) invention.

    I say these things above academically, you can see how this translates into concrete modules on my blog at where I have about ten free finished modules listed under the "products" tab.

  10. I'm using A) above in my Megadungeon for most of the traps and secret doors. There are a few, though, that are just there if they happen to search for, and will open if found. The map should provide clues that a secret door could be around there.

    If there's something important behind a secret door, there WILL be some clue to its existence, if the players are paying attention. Of course, that something could be a treasure map that they need to find, first.

    As for rooms, I do tend to keep some rooms as just plain, completely empty rooms, but most will have at least a bit of description, some more than others. And sometimes that "this room is completely empty" line can function as a tell.

    One more thing I'm doing is that some secret doors can be opened automatically if found, others can be found, but not necessarily opened unless the proper action is done (and if the action is done before the door is found, surprise, it opens like in an old movie).

  11. Beedo, try posting descriptions of dungeon rooms to twitter, off the top of your head. I try to come up with "random" descriptions of creatures and characters and rooms and such while I'm doing chores, just to keep in practice for when I'll need to make something up on the spot.

    OK, it doesn't have to be twitter, but the 140 character limit will help in practicing concise descriptions.

  12. Usually when I'm running a dungeon, for each (often arbitrary, but hopefully relevant) section I decide on the common features of the walls and doors. This gives me something to generally describe to the players and shorthand it as we go, "this room is like the last, though the limestone bricks lining the walls show there was a fire here" and so on. I try to incorporate the traps and secret doors into the common features just like the trap/door designer would have. I give further clues if the players ask, "why yes, there seems to be a lot of wear in the floor near that blank wall."

  13. Red - I use your approach as well - bland descriptions, but yielding further info when asked - but Telecanter's point about player's parsing information is pretty intriguing. I'm kicking around a follow-up post that weighs that approach.

  14. I think this is a really solid question, and it's been gnawing at me since I read it a while ago. In short, my answer would be (d). Partly I think the "twist the moosehead" thing feels more like Victorian/Cthulhu parlor detective work, and overall the effort of detailing every room so as to mask the couple of secret doors is not worth the payoff.

    I probably wouldn't have replied except this week I was also reading the Conan story "Jewels of Gwahlur", where he runs into two secret doors, which are just flush with the wall, discovered by the mismatched crack, and opened by prying them with his sword. Having just read it while thinking on this, I think that suffices for me.

    At Project Gutenberg Australia: