Zombie Toast Check it out if you want to see some of my "professional" RPG work.

Friday, December 30, 2011

Improving the World's Largest Dungeon: Random Encounters

One thing I haven't addressed so far in my improvement of the WLD are the random encounter tables. This is probably because I don't tend to use random encounter tables myself. However in an environment like the WLD I can definitely see the value...especially once your group has slaughtered their way through most of the scripted encounters in a region.

So lets see how they do...

First let's take a look at the "Random Encounter Conditions" table. This is used to get a general feel for the situation each encounter takes place in and can determine the initial attitude of the Encounter. Overall, I'd rate this table as "acceptable". You'll probably find yourself ignoring it a good deal of the time but it could still be a good way to mix up the encounters and ensure that they're more interesting than "some goblins come down the hall"

That said I notice that many of the entries on this table are fairly redundant. For instance 5 of the 20 boil down to "the encounter is hostile" and there are several other entries that are either identical or so close as to not make a difference. Many of them are also very circumstantial and would have to be re-rolled or ignored in many situations. So here's a simplified "encounter condition table". As a general rule if the creature encountered is really evil and/or violent (such as demonic outsiders, undead or evil cultists) then subtract 1 from the roll (minimum of 1). If the encounter would normally be benevolent (such as celestials) then add +1 to the roll, maximum of 12.

Roll a d12

1- Hostile. The encounter will attack the party or otherwise attempt to subdue or drive them off, depending on alignment.
2- Hostile. The encounter is completely savage. They will attack immediately and fight to the death regardless of alignment. They are completely non-responsive to attempts at communication.
3- Hostile. The encounter is afraid or cautious and will attempt to escape. They will fight if cornered.
4- Unfriendly. The encounter is wounded or otherwise disadvantaged in some way (1d6 x10% off of their hp, possibly missing prepared spells or limited use powers).
5- Unfriendly. The encounter has something else to do. Perhaps they're chasing someone or something, returning to their home or otherwise engaged.
6- Neutral. The encounter is unresponsive. They may be wandering aimlessly or simply catatonic or magically bound or paralyzed. Usually extreme force, pain or violence will cause them to attack but they may be confused.
7- Neutral. The encounter is not sure what to make of the party and will wait for them to make the first move.
8-Neutral.  Dead. The PCs find evidence or corpses showing that the encounter was killed, destroyed or triggered (in the case of traps).
9- Neutral. The encounter having an encounter of its own. Roll on the random encounter table twice and determine how the two encounters are interacting with each other based on the result.
10- Friendly. The encounter needs the party's help or at least believes they do.
11- Friendly. The encounter would prefer to interact peacefully, although they may still attempt to trick or manipulate the party depending on alignment. Creatures of animal or less intelligence will simply ignore the party.
12- Helpful. The encounter will try and help the PC in some way. Creatures of animal intelligence or less will ignore the party but may inadvertently do something that helps them.

So, lets get to the random encounter tables themselves!

Region A

Okay, well Region A has only a single encounter chart so that makes things easy. The first thing I notice is that it asks you to roll on the table once per hour which seems like it's going to get tedious extremely fast. On top of that I notice that 6 of the 20 possible outcomes is "nothing" (more if you count random sounds and . It seems like it would be easier to roll less often but with a greater likelihood of something happening. I'd say that rolling once every 4 hours seems fine. Honestly though, I'd suggest that you go with something more abstract. Perhaps roll on the table every 2 hours of real-world (rather than game) time. Obviously if a random encounter would be nonsensical (or inconvenient) at that exact moment then it can be delayed.

The table has a few other problems as well. First and foremost the results are too specific. If you stick to the table you'll probably run into many pairs of troglodytes playing cards in all sorts of improbable locations. Also, when creating this table the writers seem to have forgotten about the entire point of the Encounter Condition table. The reason the encounter condition table is a good idea is that it significantly multiplies the number of possible outcomes without creating a single huge table. A 20 entry encounter condition and 20 entry encounter table provide 400 different possibilities without having to figure out how to roll a d400. However this encounter table throws it out the window, pretty much each encounter has very specific conditions to start with. This is going to lead to some very repetitive "random" encounters. So here's a shorter, more generic random table designed to synergize better with the Encounter Condition table. Obviously "event" or trap encounters can ignore the table.

Roll 1d12

  1. Some terrible noise echoes down the hallway or a mighty gust of wind. Echoes 5, 50% chance of Drafty. 
  2. A lone humanoid. roll 1d3: 1-Kobold, 2-Orc, 3-Troglodyte. 
  3. A flock of flying, fiendish animals, 1d6+1 of them. Roll 1d6: 1-Darkmantle, 2-Stirge, 3-Dire Bat, 4-Eagle, 5-Vargouille, 6-bat swarm.
  4. A swarm. Roll 1d3: 1-Spider, 2-bat, 3-rat.
  5. A humaniod patrol. Roll 1d3: 1- 2d6 1st level kobold warriors, 2- 1d8 1st level Orc warriors, 3- 1d6 Troglodytes.
  6. It's a Trap! Roll 1d4. 1- Razor-Wire, 2- Basic Arrow Trap, 3-Deeper Pit Trap, 4-Wall Blade Trap.
  7. Someone important: Choose a named NPC the party has not yet encountered wandering the halls alone. 
  8. Pack of fiendish beasties, 2d4 of them. Roll 1d6: 1-Dire Rat, 2-Wolf, 3-Medium Centipede, 4-Riding Dog, 5-Baboon, 6-Small Spider.
  9. A lone fiendish monster. Roll 1d6: 1-Dire Badger, 2-Black Bear, 3-Monitor Lizard, 4-Boar, 5-Large Viper, 6-Lion.
  10. Minor Tremor. The ground shakes (DC 15 Reflex save or fall), dust swirls (concealment for 1d10 rounds). 
  11. Lone scout from Region B...a 2nd level goblin rogue or a 1st level Bugbear ranger. 
  12. Nothing at all. 
So that looks good. Now on to region B.

Region B

Now Region B is a bigger fish than A, at least when it comes to random encounters. The place is divided into several subsections and each has its own encounter chart. Ultimately though they all share the same problems as region A...they're rolled too often, produce many uninteresting results and its results are too specific. So here are my improved versions. Like Region A I would suggest rolling once every 4 hours or once per two-hours of real-world time (excepting the abandoned halls, see below).


The abandoned halls Encounter table has a 50% chance of nothing which makes sense but it makes for an uninteresting table. Instead we'll just roll less often. So roll once per 8 hours or once per game session (or 4 real-world hours if you're into marathon-length gaming).

Roll 1d8
  1. A foul gas pours from a crack at someone's feet. Roll 1d3. 1-Stinking Cloud, 2-Fog Cloud 3-Ungol Dust vapor
  2. Wanderer from Region A, roll on that encounter table instead. 
  3. A humanoid scout, A 1st level Bugbear Rogue or a 3rd level Hobgoblin Ranger.
  4. A lone wolf. A single Worg (40%), Dire Wolf (40%) or Blink Dog (20%). 
  5. Ancient trap. Roll 1d4: 1-Portcullis Trap (cold iron), 2-Bestow Curse Trap, 3-Wall Scythe Trap, 4-Searing Light Trap (5th lvl). 
  6. Former prisoner still lurking about. Roll 1d4: 1-Imp, 2-Quasit, 3-1d6 lemures, 4-1d4 dretches. 
  7. Slimy horror. Roll 1d4. 1-Patch of brown mold, 2-Puddle of green slime, 3-patch of yellow mold, 4-Grey Ooze
  8. Nothing.

In the bugbear settlement itself I strongly recommend avoiding "random" encounters. You should definitely have bugbears wandering the halls or keeping watch but it'll stretch believability to run into random magic, wandering beasts or similar hazards when smack dab in an encampment. For other parts here's a revamped encounter chart.

Roll 1d10

  1. Bugbear patrol. 2d3 Bugbears.
  2. Crude bugbear trap. Roll 1d4: 1-Camouflaged Pit Trap (20 ft), 2-Bricks from Ceiling, 3-Javelin Trap, 4-Poison Dart trap
  3. 2 bugbear scouts (1st level rangers)
  4. Goblin scout. A single 3rd level goblin rogue (75% from the Empire, 25% from the rebels). 
  5. Important NPC. Pick a named NPC, they're here with an appropriate escort. 
  6. Ancient Trap...a Glyph of Warding (Blast, 5th level). 
  7. Blockage. The party encounters a barrier. Either a section of collapsed rock or an abandoned barricade. 
  8. Goblin invaders! A 2nd level hobgoblin fighter leads 2d6 1st level goblin warriors. 
  9. An alarm in the form of a Shrieker Fungus. 
  10. Nothing.


Just like the bugbears I'd make sure to avoid making encounters in the empire random....and really that's actually this entire area with the exception of B89 and B90. And I've already suggested that B90 be changed to a "civilian" area for the goblins and it would be pretty ridiculous to just have a table for B90. So my suggestion is to ditch this encounter table entirely.


So, the Maze is kind of unique when it comes to random encounters...since those are in fact the only encounters. As I suggested in the last post it's best to treat the whole area as an abstract puzzle rather than having the PCs simply march through. To recap: Traveling through the Maze involves 30 minutes and a DC 20 survival roll. On a success the PCs have navigated from one end of the maze to the other. On a failure they've lost their way and must spend 30 more minutes and another Survival roll to get through. Roll for an encounter every 30 minutes. Of course if the PCs are checking carefully for traps then they're going slow (taking an hour between each check instead of 30 minutes).

However, I never actually looked over the random encounter tables themselves...and it features some truly bad editing...for instance options 6-9 involve finding a trap...but there's no traps listed at all. So here's a new, improved set of encounters. There's also the Maze-specific encounter condition table.

Roll 1d10 for the encounter then 1d10 for the encounter condition

  1. Annoying Traps. Roll 1d3: 1-Portcullis Trap (Cold Iron), 2-Stone Block Trap, 3-Wall of Thorns. Either way the trap blocks the passage, increasing the Survival DC by 5 unless bypassed. 
  2. Deadly Trap. Roll 1d4: 1-Fussilade of Darts, 2-Lightning Bolt (5th level), 3-Glyph of Warding (5th level), 4-Spiked Pit Trap (40 ft).
  3. Fearsome Trap. Roll 1d4: 1-Fear, 2-Confusion, 3-Phantasmal Killer, 4-Sepia Snake Sigil.
  4. 1d6 Vargoulle. 
  5. 1d3 Shadows
  6. Maze of Illusion. The character navigating (the one making the survival checks) must make a DC 17 Will save or become tricked by illusions, increasing the DC of the next survival check by 10. 
  7. A lone Wight. 
  8. A lost humanoid of your choice. 
  9. A group of goblin rebels navigating the maze. 2d6 1st level goblin warriors. 
  10. Nothing. 

  1. A zone of complete Silence
  2. A cloud of vile stench. Stagnant Air
  3. Patches of Brown Mold dot the walls. 
  4. Deep Darkness. 
  5. The area is coated in slime. Hazardous Footing 15.
  6. Eldritch evil fills the air. Fear 12. 
  7. Horrific echoes fill the hall. Distracting Noises 3.
  8. Thick layers of moss and slime. Bog. 
  9. Intense cold fills the air. Extreme Cold
  10. An ill wind blows...Drafty. 

 Also, while I'm at it I should try and justify this place's existence. It's a neat area from a challenge/strategy standpoint but it doesn't really have much reason to actually ever be built. Supposedly it's meant to help the celestials in case their prisoners escape...but it's tucked away here on the far side of the WLD and while it is the only way out of the B93 area it still doesn't do the job nearly as well as several solid walls and large heavy doors would. So lets reconsider its purpose. With my re-vamp of the WLD's origin story I claim that the demons/undead/aberrations here were not brought here and imprisoned by the celestials. Rather they are drawn here or actually born from the horrific abomination sealed in the dungeon's heart. The center of the maze here is one area where the concentrated evil wells up and has the potential to spawn demons or create undead. The celestials capped it but the "bleed off" is enough to still spawn lesser horrors. If the place was sealed and walled in it would soon be full to bursting so instead they created a sprawling maze to confuse and delay any new spawn (as well as traps to kill off the weakest and alert the guards) with guards stationed at the mouths of the Maze to slay any who make it through. Obviously such things were long, long ago but the well of evil is still waiting there in the heart of the maze. Fortunately the celestial's seal still holds...for now.

Anyway...enough of that. Back to the tables.


Again it just doesn't make sense to have random encounters when wandering through what is essentially someone's home. So really this area shouldn't have random encounters (or more accurately every random encounter should simply be some rebel goblins wandering between rooms or on guard duty).


 This place on the other hand is perfectly fine for random encounters:

Roll 1d10
  1.  Zombie beast. Roll 1d4 1-Owlbear, 2-Dire Wolf, 3-Minotaur,  4-Dire Ape. 
  2. Undead warriors, the reanimated corpses of paladins who died battling here. 1d3 ghouls armed with longswords, shields and chain shirts. 
  3. Frightful illusion. Treat this encounter as a Shadow Conjuration spell, pick a random creature from the Summon Monster III list.
  4. Rolling fog moves through the area. Treat it as a Fog Cloud spell that moves slowly (1 foot/round) down the hall. 
  5. 1d4 Howlers.
  6. Unlucky wanderer. Roll on the encounter table for Region F (remember my new map).
  7. Traps. Roll 1d3: 1-Cameoflaged Pit Trap (50 ft), 2-Acid Arrow Trap, 3-Wall Scythe Trap (silver)
  8. Lone goblinoid.
  9. 1d3 Wights. 
  10. Nothing

And that concludes the random encounters for Regions A and B. As I complete more Regions I'll make sure to address their random encounters as well.


  1. With your reaction table, I would be inclined to swap the first two entries (completely savage and no quarter sounds more hostile to me than 'attempt to subdue or drive them off', which sounds more hostile than 'afraid or cautious and will try to escape').

    Similarly, I would probably swap entries 10 and 11. 'Prefer to interact peacefully, although they may still attempt to trick or manipulate the party depending on alignment' sounds a little less friendly than 'needs the party's help' (more inclined to play it straight). The last entry sounds more or less benevolent.

    In fact, I might also be tempted to roll 1d10+1 (2..11) or 1d8+2 (3..10) -- or adjust the table so it goes from 0..11 or -1..10 so you don't need the modifiers -- so the outlying results only crop up with particularly nasty or nice creatures show up. I'd probably then narrow the 'neutral' range. Right now you've got three hostile, six neutral, and three friendly; four each but trim one or two of the extremes off except in certain circumstances might do what you want, too.

  2. I mostly set it up that way because I figured a "pure evil" creature would be more likely to be hostile but not necessarily more likely to be mindlessly/suicidally hostile. At the same time a normally "super good" creature would not normally be Hostile (so the +1 prevents that result) but they may be driven mad (suicidally hostile) or driven by fear (hostile and fearful).

    You're right that a better encounter table would definitely have a less even spread of possibilities. Overall it's mostly an attempt to generate ideas to mix up the encounters and give the GM some ideas to spark their creativity. Random encounters are certainly not my forte as I rarely use them myself.

  3. ah, I see. That was not evident to me, so my minds arranged them in the order I did. This makes more sense now.