I've talked in several places about how much I like random encounters, and about how the Ancestral Burdens campaign that I've been running all these years really is almost entirely derived from random encounters that captured the players' interest and blew up into a "plot." At the moment I'm getting ready to close out Chapter 13: The Worm of Hatred, which has been all about temporarily pausing one of the longest-running of these random encounters turned plot threads, concerning a particularly evil werewolf named Berger Cole. Between the events of the past couple of chapters and my expectation that the PCs are going to head into a new section of Tolrea's Mereflow Valley region, I've been busy over the past few days, updating my existing tables and building new ones. Since I'm going to be replacing what I have with new material, this seems like a good time to pull back the curtain, so to speak, and explain how I've made the campaign run for the past ten years or so. I'd be curious to know whether my practices as discussed here are consistent with what other DMs use. Incidental comments over the years have made me think that a lot of DMs feel as if random encounters are a distraction from the "point" of the game.
Anyway, when I build an encounter table, I usually don't tailor it to match the average level of the party that I'm going to run it against. Instead, I use a status quo approach that is meant to give an opportunity for the PCs to encounter what's actually present. That doesn't mean that I don't have some wiggle room to avoid presenting encounters that I know will definitely result in the whole party getting wiped out, but it does mean that if something is roaming around that's virtually guaranteed to kill the PCs if they get into a fight with it, and they DO fight it, the campaign is going to come to an abrupt end. I don't particularly want that to happen, but I think the game is more exciting for everyone if a misjudgment about when to talk or flee versus when to fight could spell the end of the game. And I don't have guaranteed-fatal monsters roaming around just for the hell of it. Instead, I try to dream up an ecology of sorts, and I use that to decide what belongs on the table for a given stretch of real estate. One consequence of this policy is that the PCs can encounter creatures that want nothing to do with them. A herd of wild horses, for example, is likely to leave the PCs alone unless bothered. And most of the time, I expect the PCs to leave them alone—but since they're there, why not see if you can capture one and tame it? My general rule of thumb for random encounters is that I try to know why the encounter is there, how it will affect or react to the PCs, and what it'll do if the PCs avoid interacting with it.
With all that in mind, I'll explain the methods I use to build a encounter table that creates plots instead of serving as a distraction. At some point before this discussion is over, I'll also provide an actual table that I used for this campaign.
First of all, I look at the region I'm building for. What's the climate look like? The core rules recognize three basic climates: cold, temperate, and hot. And in general, what I do from there is follow a "one-step" rule, kind of like the core rules' approach to cleric alignment. If it's a cold region, then I'll populate the table mostly with creatures that reside in that climate, with a few temperate creatures mixed in. The same if it's a hot region. And if it's temperate, then usually I decide whether it's a cool or a warm version of temperate, and I'll intersperse appropriate cold or hot critters accordingly. This flexibility gets me some extra variety without making the table come out as gobbledygook.
I tend to ignore climate specifications associated with playable humanoid races unless the race has some kind of weakness to heat or cold, because (for example) hobgoblins are supposed to be native to warm hills, but their organizational habits include associations with ogres (temperate hills) and trolls (cold mountains). As you might be able to guess from this, I'm also relatively easygoing about the kind of terrain that's inhabited by playable races, short of some instance where the creature needs to have water in order to breathe, or shelter from the sun in order to avoid bursting into flames. I also consider myself at liberty not to include creatures that I arbitrarily decide are not endemic to a region. There are no dinosaurs in the Mereflow Valley, for example, even though several of them are supposedly native to temperate plains regions.
After determining climate, I figure out what the terrain is like. From that I derive a list of creatures that falls within the limitations of terrain and climate. Very often I follow a one-step rule for this, too, so that you might find hill-dwelling creatures on plains that are near hilly areas, or amphibious creatures in mountainous areas that have a river running through them. That's definitely what happens for the tables I use in Ancestral Burdens, because the Mereflow Valley is a small region with a lot of variety in its terrain. It's roughly the size of Louisiana, Arkansas, and Texas combined, but it contains all of the major kinds of terrain: forest, hills, marsh, mountains, plains and aquatic are all represented, and there's a desert region right next door. If the campaign were to move into an area in some other region that is composed almost entirely of open plains, there'd necessarily be a lot less diversity on the encounter tables for areas near the center.
After I've decided what climates and terrains I'm allowing on a table, I also decide which extraplanar critters, if any, are present. We won't get too far off track into the particulars of the Tolrea setting, but this is where I determine what "spirits," if any, have made their way into the mortal world. If I'm building a table for some other setting, I may skip this step unless I know that there's a spellcaster or other magical influence in the area to account for the presence of otherworldly critters. Similar concerns pertain to how many creatures with "any" for their environment are likely to make it onto the map; many of those are undead, and I like to have a clear idea of where they are coming from. In general, I try to have a "why" to explain any supernatural or exotic creature's presence on this part of the table, because that means I can take the random encounter for that creature and (potentially) blow it up into a plot. The plot concerning Berger Cole is an ultimate result of this policy.
Up to this point, I will not have assigned a percentile chance of encountering any of the creatures on my list, or a defined mechanism for saying how many of them are going to appear. It's just a list. Before I address those things, I consider what any intelligent races might have done to alter the natural environment around their habitations. If the area is made up of plains bordered by hills but it has been settled and tamed into farmland, then usually I decide that there should be fewer roaming monsters because of local patrols. But the trade-off is that the availability of food, goods, and money draws bandits, as well as raids by various monsters that collect treasure. If there's an unfriendly settlement or organization based in a neighboring area, then quite often the outbreak of hostility leads me to include them on the table, as well.
I always finish up my list with a set of NPCs and other special encounters. These can be anything from local farmers, hunters, or trappers who're on their way to market for supplies or to sell their harvests, to a wandering monk, to the aforementioned bandits or raiders, to a caravan or band of religious pilgrims. This usually is also where I include patrols by the local government or other authorities, and I use this space, likewise, for encounters that are meant to advance any plots that the PCs have stirred up but which are not necessarily their immediate priority. As a general rule, these encounters are pre-constructed in terms of the NPCs' stats and gear, but they usually aren't fleshed out for personality and history. I handle those things on the fly. These NPC encounters always have a separate file where I keep a list of rumors, goods, and so on that the PCs might get hold of through their interactions with the encounter.
Encounter distance 6d6 x 40 ft.
Verdant/civilized: hourly encounter chance 10% Day Night Encounter CR/EL 01-10 01-03 1d3 farmers (human com 1) and wagon with 2 mules 1/2 to 1 11-20 04-13 1d2+1 shepherds (human com 1) with 1d3 riding dogs 1 to 3 21-30 14-18 militia patrol (Keriel's Turn) 4 31-40 19-21 mercantile caravan 7 41-45 22-31 2d8+8 goblins on worg mounts 9 to 14+ 46-48 32-37 1d10+10 orcs, plus 2 3rd-level sergeants and 1 leader
of level 1d4+2) 8 to 9+ 49-50 38-42 1d2+2 ogres 5 to 7+ 51-52 43-47 1d4+1 hill giants 9 to 12+ 53-57 48-50 Velvet Purse caravan 10 58-59 51-55 2d8+8 hobgoblins 5 to 7+ 60-61 56-60 2d8+8 kobolds plus 1d3+1 dire weasels 5 to 10+ -- 61-62 vampire 7+ -- 63-66 1d4+1 vampire spawn 6 to 9 62-63 67-68 1d6-1 wights 3 to 9 64-70 69 religious pilgrims 7 71-75 70 wandering monk 8 76-80 71-72 Cleansing Flame agents 7 81-100 73-100 no encounter --
Once I've populated my list, I go through and figure out how many creatures of a given type are likely to show up. Most roam around in groups of several different sizes. Just for example, goblins can show up as a gang, band, warband, or tribe. You've encountered a warband, so far. As you'll see when you look at the sample table, there was no option to encounter any of the other groupings for goblins, and that's a decision I reached because (up until this point in the game) the goblinoid clans have not pressed close enough to the settled heartlands of the Mereflow Valley for you to encounter anything else. A gang includes only 1d6+3 members, which is so few that they'd face almost certain defeat if they were caught roaming where they shouldn't be. Goblins advance by character class, so I could have decided to use a gang of elite goblins . . . but that'd be like sending in Seal Team Six. A band is much larger, but it includes noncombatants. So you wouldn't expect to see them roaming around in hostile territory. A tribe has the same problem, although it may include several HUNDRED goblins rather than a few dozen. That left me with the warband, which consists of 2d8+8 goblins on worg mounts: fast, stealthy, and relatively small in number. Perfect for causing trouble and gathering information. This phase of my planning is also when I decide whether I'm going to advance some or all of the creatures in a group. Once I know how powerful and numerous a given monster is, I calculate a CR/EL range for that entry on the table.
Finally, I assign a percentile chance of having an encounter with each monster or NPC grouping. I always set it up with both day and night columns. Obviously, some creatures (like vampires and spectres) can only be out at night. So they get a "--" entry under the day column. Most of the time, I also place "--" under the night column for any creature that I think should only operate in daylight, like hawks or giant bees. Some monsters also get a note that they are only valid encounters when the party is moving; web-spinning spiders are a good example of this. Some encounters are more common in the day than in the night, or vice-versa, but possible in both. Insofar as I ever "tailor" an encounter table, I make encounters more likely or less likely based on how dangerous they are, or how interesting I think it would be to run that specific encounter (as I write this, I'm thinking of an encounter with a homebrewed "slitherweb" that I am running in the Cataclysm campaign, which got a couple of extra points on the table because I thought it would be fun).
Sometimes (as in the example table) I run out of encounters before I run out of percentile chances. If that happens, then most of the time I just have a line on end of the table that goes "no encounter." This is particularly likely if I'm preparing an encounter for a relatively safe area. I periodically revise these tables to reflect changes in the status of a given area; recent events in the plot of Ancestral Burdens have led me to revise the level of danger upward, although it still isn't a radical change. As the Mereflow Valley spirals closer to war, players can expect that subsequent drafts will contain higher and higher chances to encounter larger and larger groups of hostile humanoids. If war breaks out, its resolution could in turn bring about a wider prevalence of monstrous encounters, both as an outcome of dead combatants rising as undead, and as a result of population losses that will cause the abandonment of once-settled areas. If civilized people aren't hunting down monsters, they'll tend to breed and take over.
In play—and again, you can see this on the example table—what I'll do is roll 1d100 per hour, and check against the hourly chance for an encounter. If there is one, then I roll 1d100 again to see what kind, and work from there to flesh out numbers, strength, and distance. This is all stuff from the DMG 3.5, but there are no exemplary tables in there to walk you through the whole process. Once I have a starting distance, I use that to set Spot and Listen DCs for both the PCs and the creatures from the encounter. In most cases I set the base DC equal to 10+Dex+size modifier+check penalty+Hide or Move Silently for the largest creature for each side of the counter, and then adjust the DC to account for distance, and then I secretly roll Spot and Listen checks to see who (if anyone) sees the other side first. This is easily my least favorite part of a random encounter, because the math gets a little complicated, but I like the way it plays in and out of character, so I put up with it. The biggest downside is that it (arguably) penalizes the PCs if they have any kind of sizable mount or pack animal, since the base DC to spot most horses or mules is around 9. But in most cases, I think that having mounts is worth it because the PCs are able to travel significantly faster, and this reduces their overall chance to encounter things because they aren't on the trail for as long.
As I said, I'm curious to know if this is what other people do, or whether I'm weird.