Life in the Wide World – Random Encounters

Last time, I explained why I like random encounter tables and feel like they add to both the “realism/simulation” factor but also add to the story by being bellwethers of the players’ interest.  Now, I’ll talk about how I create and use them.

Scope

The Pathfinder Bestiary has some generic random encounter tables in the back, but obviously in most cases you’ll want something tailored to the place you are.  The Paizo APs do a pretty good job of providing basic random encounter tables for the various cities and regions in which they take place.  You’ll want to put work into these tables proportionally to how long the PCs will be there.

More specific is always better, it’s just a matter of how much work you want to put into it.  The good part is, if you do this right the work you put into it is heavily reusable from campaign to campaign.

1. City Random Encounter Table Template – consider putting some meta-thought into this so you can use it to build other charts easily.  “01-05, Wimpy Local Monster 1”, for example.
2. Generic City Random Encounter Table – to use anywhere when you don’t have anything more specific on hand.  Some are provided in the Pathfinder Bestiary.
3. Chelaxian-type City Random Encounter Table  – you can use this in a broad swath of locations – Cheliax, Magnimar, Korvosa, etc.   If you have a couple of these ready to cover the most likely sectors, you are ready for a lot.  “Port City Random Encounter Chart,” with maybe some entries that say “City Guard (lawful city)/Pirate Press Gang (pirate city)”  would cover a lot too.
4. City of Korvosa Random Encounter Table – use in Korvosa.  Often there’ll be one of these provided in an adventure or AP you can modify for use.
5. City of Korvosa Sewers Random Encounter Table – use in the sewers.  You might have a Generic City Sewer Random Encounter Chart that this inherits from too.   Sometimes an AP will publish tables down to this level of detail.
6. City of Korvosa Midland District Random Encounter Table, During the Troubles – you can vary the charts by time as well.  If the city is under martial law, you’d expect the encounters to be pretty different.
7. Gold Goblin Gaming Hall Random Encounter Table – don’t be shy about making charts that are very very specific to a small location, as long as the PCs are spending a lot of time there.  In the Second Darkness AP, it’s likely the PCs will end up running a gambling establishment and will be working in it a lot.  That might merit a very specific “who wanders in and/or causes trouble on a given night” chart just for that one building.

Scaling

There are two philosophies of encounters in D&D.  The first is to always scale encounters to the PCs.  The other is to allow for the entire range of realistically possible encounters without regard for the PCs’ abilities.  The latter, though nice from a realism/sandbox play point of view, can be a little dangerous.

I thought the Scarred Lands products from Sword & Sorcery had an elegant solution to that problem that well served both realism and scaling needs.  They assigned CRs to actual locations that reflected the average CR of encounters there.  Then PCs should, with a modicum of care, be able to find out how dangerous a region is.

• “Foundling’s Green?  A naked virgin with a sack of gold could wander the cornfields safely there.  Someone thought they saw a goblin 5 years ago and the Count sent troops; turns out it was just a real ugly sheep.”  CR1/2.
• “The Schwartzwald?  Well, you wouldn’t want to wander around in it alone, there’s some dangerous things in there, but groups of loggers work those woods.  One group like that went missing six months back, though.”  CR3.
• “The Forest of Screaming Skulls?  No one has ever entered it and returned.  The Hellknight Order of the Dragon sent in a hundred knights to pacify it; none came back out, but their mothers all died of heart attacks that very day.”  CR 16.

You can vary CRs by smaller scale, of course – the Bowery has a higher CR than the Merchant District.  Of course in most normal human cities, even “rough” areas aren’t going to have super high CRs, or else people’d get wiped out.  The crime-riddled streets of Riddleport only have an average CR of 1 on their encounter table.  This advice mostly applies to hostile/monster encounters; there may be EL5 bands of city guards about, of course, or by crossing the street and going into the Collegium you can probably toss a rock and hit a CR8 wizard, but they are not (usually) eating peasants.

I’d like to give a shout out to “Wilderness & Wasteland – Scarred Lands Encounters,” a Sword & Sorcery supplement all about building encounter tables.  It’s worth picking up.

Anyway, once areas have average CRs, PCs know where they can go safely and what’s risky, and  your campaign flavor is backed by the rules.  It allows scaling by self-selection, in the same way that World of Warcraft does with its zones.

Type

Similarly, there are two philosophies on random encounter tables.  One, the “hostile encounter” theory, holds that they are only for monsters or people likely to provoke random encounters.  These consist solely of entries like “Bandits, bulette, owlbear, King’s Guards, orc warband.”

The other holds that they should also include “friendly” encounters, events, and other stuff. “Crying orphan, dead soldier, abandoned shrine, loose horse…”  These can be used as hooks or just to provide flavor – this example list communicates “war-torn countryside”, for example.

I personally prefer the latter, though the former require less work, as they’re a subset of the larger, more inclusive ones.

The main thing to keep in mind is that you want the frequency of encounters to be lower if you’re using an all-hostile encounter table; you can ramp up the frequency the more other stuff is on them.

Populating the Table

What, besides monsters, can go on a random encounter table?

• Normal animals or plants – threatening, tasty, valuable, or entertaining
• Natural phenomena or events like sandstorms, stampedes, marching strikers, or an overturned dung collection cart
• Traps like bear traps or pungi stakes
• Diseases like malaria or yellow fever
• Largely friendly people like traveling merchants, con men, pilgrims, or prostitutes
• Largely unfriendly people like city guards, roving berserkers, or bloodthirsty natives
• Local features like totem poles or caves
• Happy treats like ivory, an outcropping that contains valuable metals, or a dropped coin purse
• Specific people the PCs know
• Pretty much whatever else you could think of

Now, I will say that more isn’t always better.  Especially if you evolve your tables over time, you can easily get 100 different things on a given table.  But this runs the risk of cognitive sprawl – it’s hard to work up 100 encounters ahead of time and so they are more off the cuff, and it’s less likely to have anything recur – and from both a story and a game point of view, you want some recurrence.  Recurring threats help show theme more strongly than totally different ones every time and may become a plot element, and also PCs like to get better at fighting things.  The first time your party encounters a yellow musk creeper, it may take them a while to figure out there’s a plant behind all the zombie business.  So the second time, they can know what’s going on, target the plant, and feel proud in showing off what they’ve learned.

It’s easy enough to have a big long list of random encounters and cross ones off.  So if you have your big jungle encounter table, for a given trek through the Mwangi you might cross off a bunch of it and use a subset, just so  you can prep some of them a little, allow for recurrence, or even just not have to carry ten different books with monster stats in them to the game.  And if you’ve added a cool new monster or NPC it’s more likely to come up, and if there’s just monsters you feel more like running this week and ones you don’t…  If you have your table stored in Excel then it’s easy to cut, paste, and remove/alter entries to taste in very short order.

Chance of Encounter

Historically in D&D this has been a fixed rate; “x% per hour” or other time increment.  (Bizarrely, in Pathfinder they seem to not have addressed chance of encounter at all.)  I kinda prefer to make it dependent on the PCs’ skills.  Someone with a good Survival check (or, in the city, Diplomacy/Gather Information) should be able to avoid a lot of encounters with their knowledge of the ways of the wild/city.  In addition, the chance should probably be lower if the party’s staying still and higher if they’re romping through the underbrush.  It’d also be nice to take into account the scenario where the PCs want to provoke random encounters, for whatever demented reason happens to be at hand.  (You could argue PCs acting as a highly motivated city guard and looking to enforce the law could be simulated by provoking random encounters.)

Cautious: If you’re staying in one place or otherwise trying to avoid encounters, you check once every twelve hours.  You can travel overland  cautiously by moving at half speed.

Normal: If you’re moving about normally, you check once every six hours.

Aggressive: If you are looking for trouble deliberately, or otherwise moving about in an intrusive manner (searching the wilderness for a dungeon entrance, foraging, checking every inn in town for someone) you check once every three hours.

The DM may also modify the frequency of checks based on what’s going on – if there’s an active battle in the vicinity, it’s reasonable to say that the area is hot enough that there’s random encounter checks every hour.

Use skill checks to determine random encounters.  In the wilderness, use Survival, and in the city, use Diplomacy.  The DC to avoid an encounter is 10+the CR of the location.  So for example, in a CR3 location, a Survival check of 13+ avoids an encounter during that time period.  Feel free and combine this with other rolls of this skill; for example overland travel usually provokes Survival checks.  If someone’s rolling their Survival check to avoid getting lost, you can just compare that roll against the encounter DC and have it serve double duty.  You can always deliberately “pull” this check if you’re looking for trouble and “take zero,” so to speak.

Optionally, you can do a little more granular work and roll on the encounter table first and then check against the CR of the specific random encounter rolled.  Then you can let Knowledge checks be used to assist – if the encounter rolled is a bear, then PCs with Knowledge: Nature could roll to assist the Survival roll to avoid (or provoke) the encounter.    This rewards domain specific knowledge, which is always a plus.  It also allows for some discretion in “looking for trouble”.

Example

Let’s say the PCs are out looking for magical beasts to capture for some dude in town that’s paying well for them, but would like to otherwise avoid pointless fights.  The PCs decide they’re looking aggressively during the day (one check every 3 hours) but taking it easy at night (one check every 12 hours).  During the day, they decide to “take zero” and provoke an encounter on a d20 roll under 10+ the encounter EL.  If you roll an encounter with orcs, for example, which they don’t want to deal with, let them make Knowledge: Nature checks to “assist” the Survival roll in the direction they want.   Let’s say two of them have Knowledge: Nature and make their DC10 assist rolls, meaning a total d20+4 on their check to avoid the encounter.  If they might encounter an owlbear according to the chart, the same Knowledge: Nature assist rolls would let them subtract two from their roll, so if both make their assist they would roll d20-4 versus the same encounter.

Bringing It All Together

Here’s what an encounter table might look like.  Let’s say it’s this party’s first trek into the Mwangi Expanse, Golarion’s equivalent to Africa.  I have a giant jungle encounter table including all the entries in the table in the back of the Bestiary.  I decide that I want to stress the more “mundane” threats that the jungle has to offer for this first outing, so I focus on the inhabitants, animals, and diseases of the Mwangi more than supernatural or monstrous threats; on later journeys I’ll crank up those and have the mundane stuff stay there but become a  smaller piece of the pie.

They’re going to be inland only three days or so, therefore probably a dozen or so entries will allow for enough variation with some chance of recurrence.

Kaava Lands, Mwangi Expanse Random Encounter Table (CR 5)

1. Bonecrusher Fever Fort DC 12 (CR1)
2. Sleeping Sickness Fort DC 14 (CR2)
3. Javelin Trap (CR2)
4. Enteric Fever Fort DC 15 (CR3)
5. 2d6 Mwangi Tribesmen Warrior L1 (CR3)
6. 2d6 Jungle Elves Warrior L1 (CR3)
7. 2d10 Aspis Consortium slavers (CR4)
8. 1 army ant swarm (CR 5)
9. 2d4 gorillas (CR6)
10. 1 dire tiger (CR8)
11. Location (roll d6):
    1-3: Mwangi village – 10 warriors, about 30 souls total, starting attitude of indifferent
    4-5: Mwangi village – burned to the ground, no survivors
    6: Mwangi sacred site – any Mwangi with the party will refuse to enter the area, causing two hours of lost time maneuvering around it, and any PC taking any of the fierce little carved masks hanging from trees in the place will get a hostile reaction from any Mwangi seeing them later on.
12. Person (roll d4):
    1-2:  That crazy explorer they met in Bloodcove
    3-4: That hottie that works for the Aspis Consortium they’re all trying to impress

This table carries out my theme for the adventure – it’s more about discovering the area than hacking on wildlife.

Other things to put in your table besides the encounter name and CR – I would have done it above but WordPress is awful at tables, sadly – put the source and page reference, like Paizo does.  A lot of their cooler new monsters are from articles in the APs, and even now I have a dickens of a time hunting down where a monster is.  Writing it down in the table once will save you much flipping later.  Also, consider putting in starting attitudes (friendly, indifferent, hostile), especially useful for people-heavy tables.  There’s a big difference between a friendly “come join us!” merchant caravan and a “don’t come within 200 feet or we fire these crossbows” merchant caravan.

Here’s a more city-focused encounter chart that I might use in my current campaign.  It uses the Riddleport Random Encounters Table from Second Darkness: Shadow in the Sky p.79 as a base, but I’ve added specifics.  The AP says  “con artist,” I’ve come up with some specific hustles and also have some specific con men added in from random places, like Mungo and his Amazing Monkeys from a Freeport supplement.   And the table completely omits certain things that seem obvious, like Riddleport Gendarmes.   Plus, as their intrigue among the city’s inhabitants heats up, I’m adding more specific people and/or spies for specific people.  CR is less meaningful as more of these aren’t (necessarily) combat encounters.

Riddleport Wharf District Random Encounter Table (CR 1)

• 1 monkey (Avg CR 1/6, MM p.276)
• Con artist pretending to be a shanghaied princess (CR 1/2, see notes)
• Con artist – Mungo and his Amazing Monkeys (CR 7, Denizens of Freeport p. 65)
• 1 leper (Avg CR 1/2, SitS p.79)
• 1d4+1 Gendarmes, reasonably honest (Avg CR 2, use Riddleport Thug stats)
• 1d4+1 Gendarmes, looking for trouble and/or bribes (Avg CR 2, use Riddleport Thug stats)
• Harlot, quickwife (CR 1/2)
• Harlot, easyboy (CR 1/2)
• Harlot, Selene (CR 2, Maiden Voyage p.XX)
• Harlot, Lavender Lil in disguise (CR 7)
• Drug dealer, connected to Avery Slyeg (CR 1/2)
• Drug dealer, unaffiliated (CR 1/2)
• Drug dealer, the one Sindawe likes to beat up and take drugs from (CR 1/2)
• Homeless person from St. Casperian’s (CR 1/4)
• 1d8 Drunken pirates Ftr1/Rog1 (Avg CR 3, see notes)
• Street vendor, spy for Clegg Zincher (CR 1/2, see notes)
• Splithog Pauper, in disguise (CR classified, SitS p.66)
• 1 goblin snake (CR1, Pathfinder #1)
• 1d4 Small monstrous centipedes (CR 1, MM p.276)
• …

Conclusion

Random encounters can be fun and useful, and are for way more that just monsters.  Customization is your friend, and it can be done without requiring a lot of work all the time – some upfront work to make some basic tables then lets you do minimal per-session work to customize them.