Improvisational DMing

Our g+ game was, I think, an excellent example of my favorite kind of play.  I ran this adventure, which was written communally over at the goblinoid games forum here.  I ran this once before at last GenCon, and it was very interesting to play it again with a different group.  To start, this adventure has several unique features that I think greatly add to requiring a good amount of improv at the table:

  1. The basic model is a inverted pyramid, with several different small sub levels (A, B, C, E, F), with the best stuff at their nexus (D).
  2. The contents of each room vary wildly, as you would expect from something written communally by so many authors.  Each room though also contains a lot of interesting features to be examined.
  3. There is a shocking lack of consistency.

Basically, the combination of lots of interesting stuff that makes little sense when combined is just the motivation I need to start improvising a the table.  As the players encounter unrelated stuff and start asking themselves probing questions as to why certain things are the way they are, I find myself altering the remaining contents to ensure it answers those questions, though not necessarily in the way the players may or may not be conjecturing.

Let me go through it room by room, and you can follow along and see what I’m talking about:


The players start by trekking up the hill where I describe the entrance to the tower and the creepy old tree across from it.  Both groups I have run this for have taken the time to examine the tree, found the second entrance, and opted to use it.  That’s kind of interesting.  I suspect it’s perhaps ingrained in D&D players that sneaking in the back door is always preferable to the frontal assault.  I suppose that makes sense for a game that essentially about tomb robbers, assassins, and spies.

F 14

The players examine the room with the weird mushrooms and ponder the gnome statues.  They then discover the kobold hiding in the closet, and rather than kill him, they question and then hire him to be their guide.  Brilliant!  The last group immediately attacked the kobolds, which pushes the adventure in an entirely different direction.


The newly hired kobold guide leads the players right to his home, provides an introduction to the kobold chief, and thus paves the way for making this into a roleplay encounter rather than a combat.  The kobolds reveal their desire to steal the gnome statues.  The players quickly make up some nonsense about doing such requiring consumption of a magic herb found above ground, and cleverly describe poison ivy.  The kobolds fall for it hook, line, and sinker.  They send 4 kobolds out to find the magic herbs, and recompense the party by giving them two kobold guides and what little info they have about near neighbors.  They describe the scary bats below, and the scary spiders to the south.

The party questions them closely about what they will do with the gnomes, and now I have to start inventing.  Thinking of rooms 4 and 6, I start describing a band of orcs they trade with, and how those orcs are very angy at them as they traded some gnome statues with the orcs for a couple cows (likely stolen from farmers back at town) which the kobolds promptly ate.  Once the statue vanished the orcs demanded recompense, which the kobolds clearly don’t have, thus the kobolds are trying to figure out how to get and keep some gnomes to pay their debt.  I’m already thinking at this point of converting level C to a more traditional orc lair on the fly, though haven’t decided if room #6 will contain live orcs or not.  It’s possible the dead orcs in there only died recently, thus alleviating the kobolds but they don’t realize it.  Or maybe the orcs are still around.  I’ll decide when the party gets there.

(Minor point of interest: room #15 is one of my own contributions and I was pretty proud of it.  I’m super excited to see it treated this way.)


The party decides the bats sound less scary than the spiders.  They go down the stairs and position themselves well in the doorway to the stairs to fight the bats two at a time rather than all at once.  This is an excellent tactic and may be the only reason the party succeeds, though the bats do kill and partially eat both kobold guides.


The party discovers the creepy altar and stairs, and decides to deal with this later.  They move on.


The party sees the carcass and disturbs the flies, but quickly retreats back towards E12.  I roll that they only anger 3 of the flies, which they fight and barely survive in the corridor between E12 and E13.


The party decides to take the stairs down, finds the webs, and sees one of them moving.  They decide to apply fire and crossbow bolts, actually bringing one of the two ogres there all they way down to half hit points before they start really fighting him.  That said, they held the line for a few more hits, but once the ogre landed a solid blow on their front line fighter they decided it was time to get out of there.  Actually probably a wise choice, but sadly it meant missing out on all that sweet sweet ogre treasure.


At this point the party decided it was time to head back to town to rest and recuperate.  While doing so, they discussed hiring on a hireling to replace their lost kobold guides, but instead ended up sending the magic-user over to the bar to draw out a potential fighter and charm him.  Brilliant idea, but very risky in my mind.  Once they were alone and the spell was cast, I think our friend the magic-user would have been in serious trouble had the victim failed his saving throw.  Good luck was with them though, and they returned to the dungeon with additional companion.


The party now decides to try and enter from the other side and see if they can’t find those orcs they heard all about.  Looking at the map I see the giant killer bees described as living on the first floor, which is subdivided into two rooms, and it’s vague which room the bees are in.  I decide they are in the farther room as those orcs must get in and out somehow.  The party here’s the buzzing and decided to go immediately for the stairs and avoid the bees.


Again in deference to the supposed orc clan down here I reduce the amount of honey in the room below.  As cool as that room description is, I figure the orcs must get in and out somehow.  The party also checks the stairs for stick prints, which I figure must be there as well.  Clever players — I quickly devise a way to eliminate those tracks later in the dungeon, a line of boots in the hall just outside room #2.  I mean, who wants honey all over the floor of their lair?


The party makes an attempt for the bracelet and fails miserably.  The thief is stuck in the rubble.  They come up with a clever way of fishing him out using ropes and a 10′ pole, so I grant them a speedier escape than the 2d4 hours mentioned, but at the price of some additional damage to the thief.  Also, I was laughing too hard along with the rest of the party at the image of the thief being dragged out by his ankles from a pile of rubble to not push that through.

Stairs to C

First the party finds the boots, and debates whether or not they should take off their own boots.  They decide to go ahead and do so.  I love all the crazy theories they come up with regarding the boots, which were simply a dodge I invented to excuse lack of tracks further in the dungeon.  I start to ponder how I might further leverage trouble for the now barefoot party, but we never get anywhere I can use it.

Another invention I can personally claim credit for is the narrow stairs and fragile knotted rop, inspired by an actual staircase I encountered in England.  Not a real trap, just a dangerously old construction.  The party uses extreme care in examining it at first, sending a guy down slowly tapping with a pole and secured by rope, and given the slow careful pace he makes it down with no problem.  Then he comes back up, and they all proceed down at normal pace, avoiding the fragile rope, and thus naturally one of the fighters goes tumbling down the stairs and crashes into the door at the bottom.


Between the crashing dwarf and the collapse of the ceiling above, I decide any orcs below are well aware of the party’s approach and start to devise what kind of reception they might have planned.  I figure there are orcs laying in wait behind every door in the passage beyond.  As soon as the party encounters one group, the rest will all leap out to attack, hopefully surrounding the party.  They don’t quite get this off though, as the party opens the very first door, leaving their rear ranks still in the stairwell blocked by their second line of fighters.  Still, we have a pretty enjoyable fight with the front two lines facing two directions — into the room and down the hallway.  A well placed sleep spell ends the fight pretty quickly.

And that is where we had to end for the night.  What is in rooms C5-7?  More orcs probably.  Is the same stuff in there as is written in the adventure?  Maybe.  To be honest, I haven’t even read those rooms in advance.  I figure as they enter those rooms I’ll read what’s in there and maybe tweak it a bit on the fly to have some more orcs.   What about level D?  I have no idea.  I’d be just as excited to find out as the players.

This is what I’m getting at with all my talk of improvisational play.  I get just as many kicks playing as the players do.  I’m just as surprised to find what they discover as they are.  Yeah, I sort of know what’s there in advance, but through play the players are actually encouraging me to change it, and I greatly appreciate all their efforts.

Now, the trick will be figuring out if I can write a similar adventure from scratch entirely by myself, without the need for random conflicting ideas from other people.  I’ll have to give that a try sometime in the future.  I wonder if there are any little tricks I might try to add more chaos to my design, but still give more interest to the room than usually comes from random monster, treasure, and dressings.  Any ideas would be appreciated.

3 thoughts on “Improvisational DMing

  1. I’ve tried this method with success. Take a bunch of index cards and on each write a compelling encounter, room, or NPC (I choose color coded ones so yellow are encounters, green are NPCs, etc). Shuffle them. Then when PCs enter a new room or hear a rumor about something in the dungeon, pull a card. It’s more developed than a bare bones random table but less structured than a carefully constructed map. And the size provides a natural limit to any attempt to over specify.

  2. I had a great time, and I was going to say that I actually thought it was a quite well-designed dungeon — feeling there were enough advance clues in places to at least take a stab at making guesses on how to approach or avoiding bats, bees, spiders, and orcs.

    As usual (without having read the module) I’m surprised at how much was improv’d on the fly (and executed so well!). I’m thankful for the dungeon structure that does actually fix monster locations in advance, so that our questioning has the opportunity to establish choices on where we hunt next.

    The other thing is that I actually feel our old group has gotten to be better D&D players than we were 10 years ago (in the PHB strategic sense).

  3. I’m thankful for the dungeon structure that does actually fix monster locations in advance, so that our questioning has the opportunity to establish choices on where we hunt next.

    Yeah, I think that’s the key difference between what I want and what Adam is describing. I suppose I might be able to use Adam’s method to pre-create a fully fleshed out dungeon that still contains some elements that lack consistency. Or perhaps even a 90% complete dungeon, with a couple rooms left somewhat empty to fill in on the fly. It’s forcing myself to overcome little inconsistencies that I think really opens the door for me to do the improv stuff I love.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.