This entry is part 3 in the series Creating ecology-based random encounters

In this concluding part of my series on encounter tables, I look at Urban Settings, Dungeon settings, and talk about ways of integrating Wandering Monster encounters into plotlines and infusing them with meaning. And I might throw in the occasional new idea in other relevant areas, to boot. A quick catch-up – In part one of this series, I talked about the philosophical grounding of random encounters – the theoretical why’s and wherefore’s that underpin the encounters that result, and the ways and reasons why they matter. In part two, I discussed a way of creating better, smarter, encounter tables by first constructing an ecological model from which to work.

Before we can really dig into these topics, we need to return to conceptual basics for a minute.

What is an encounter table?

Try this definition for size: An encounter table employs randomness to select aspects of an environment or situation for closer examination or discovery by characters.

Getting better results from random encounters

The consequence of this definition makes the process of constructing an encounter table both more useful and more conceptually open to interpretation. The most common form of encounter table – a list of random encounters for a given region or environment – emerges as simply the most elementary version of such a list. The general process of delineating an encounter table emerges as logic steps applied to achieving the creation of such a table; the necessary steps are,

  • Determine the environmental domain within which this table will apply;
  • Determine a criterion apon which to categorize table entries;
  • Compile a list of table entries using that criterion;
  • Using some discrete numeric assessment based apon the criterion, assign a relative frequency or likelyhood to each entry;
  • Use these values to generate the actual table, creating subtables as necessary.

The inevitable result of this process are better results from your random encounters, because there is some rational principle applied to the selection process and some relationship that defines relative probability of encounter. These processes can take into account multiple factors.

The previous article, Part 2 of this series, employed a simplified ecological model as the rational principle, but took into account the size, visibility, concealment, temperament, and the likelyhood of each party to the encounter wishing to engage with the other, as key elements of the defining relationship that determined the probability of encounter. Step by step, it travelled up and down the food chain, determining how much food there was available and therefore what the necessary range was for the creature to exist in a self-sustained ecological balance. It made the unspoken but essential assumption that such a balance would naturally exist, or would evolve to exist – or the encounter domain would long ago have suffered an ecological collapse.

This is, of course, not the only principle from which an encounter set can be derived, and that’s going to be the central theme of today’s article – how to employ different criteria and principles to adapt the ecological technique demonstrated for use in other ways and in other environments.

But first, a word on making the resulting system of encounters dynamic, rather than static – a topic that I originally intended to include last time but ended up putting into the ‘too hard’ basket when time began to run short.

Dynamic Encounter Tables

There are three ways that a static list like an encounter table can be made dynamic:

  • Modifying the population,
  • Modifying the probabilities, and/or
  • Modifying the relationship.

The most comprehensive approach would employ all three to incorporate the consequences of a significant encounter into the table subsequently, based on the outcome of the encounter. Let’s take a closer look at each of these (in a slightly different order):

Dynamic Probability Reassignment

The PCs have an encounter. It has either a decisive or an indecisive outcome – if the creature encountered is killed, that’s a decisive outcome. If the entire party becomes lunch, that’s also a decisive outcome. Every other result that I can think of is indecisive, in that it leaves scope for a subsequent encounter with the same creature.

How badly wounded was the creature or group that was encountered? How significant is that encounter as an individual or as a group of individuals? How intelligent is the creature? What will it learn as a result of the encounter? How will its behavior change as a result of the encounter? How will its temperament factor into these answers? Each of these questions should alter and adjust the probability of encountering the creature again.

Intelligence is the key. The more mindless and instinctive the creature, the less its behavior will change. The smarter the creature, the more likely it is that there will be a behavioral change. This applies even when we’re talking about differences at the animal level of intelligence.

If the encountered creature or pack was badly injured, AND they are a significant entity within the domain limits of the encounter table, they may be unable to hunt as they would normally. They would, instead, be forced to consume smaller creatures than they normally would, and in greater number (because each individual would provide less food). If the creature was also strongly territorial, or simply bloody-minded, the chance of a subsequent re-encounter would therefore increase at the expense of a reduction in the chance of encountering the smaller creature on which it is now forced to feed. If it was forced to feed on smaller prey for a while but had learned to fear – or at least to employ caution – in dealings with character-shaped beings, then the chances of both might go down.

Of course, higher intelligence changes things somewhat, because it introduces abstract reasoning and abstract motivations into the mix. Reencounters may occur motivated by revenge, or perhaps the PCs came into possession of something that the survivors want back, or perhaps they have allies or servants. Or, perhaps, fear and caution might dominate. The survivors might set traps and ambushes designed to injure or kill the characters while placing themselves at minimal risk of further harm.

Quite often, it won’t be necessary to completely recalculate the table. If there is an equitable reassignment of chances, the other individual percentages won’t change in size, only in start and end-point. Even that can be avoided by inserting a new entry for the “extra” chance of one encounter that used to be part of the chance of the other: A% Encounter 1 becomes A-x% Encounter 1 and x% of Encounter 2.

Where both are reduced, you can simply insert “roll again” results in the same way.

Only where there is a disproportionate change, or “roll again” results start coming up too often, does the entire table need to be recalculated. If for example, the chance of a re-encounter is higher than the original chance of encounter, so that the table’s total now exceeds 100%, it needs to be recalculated – unless a prior reduction in probabilities has freed up some “Roll Again” slack that you can retask to the increase.

Dynamic Relationship Reassessment

After every encounter, the relationships between the encounter participants and other entrants on the table should be reassessed, and the population growth / diminishment values reappraised. Some encounter outcomes have a more profound effect on the ecology. A wounded predator can be unable to defend its territory, and a dead one certainly can’t. Whoever the next leading predator is within the domain will step up, and at the same time, a younger Apex Predator from a neighboring region may invade the territory, splitting the original encounter chance between them. It’s my habit to end each encounter label with a (+) or (-) sign to indicate whether the population is growing or declining, because growing populations give rise to the capacity for territorial expansion. Also claiming a share of the spoils will be whatever the predator used to live on, which will experience a population boom. Some of encounter chance increase that reflects that boom will come from the encounter chance of the original predator, and some will probably come from a decrease in the chance of encountering a scavenger, who now have one fewer mouth providing for them. On the other hand, depending on the species involved, the scavengers may grow desperate, which may express itself as an increase in encounter likelyhood.

What is true of an Apex Predator will also be true for a pack whose numbers are depleted in an encounter.

None of these processes happen overnight. The chance for an encounter with a replacement predator starts at the minimum possible – 1% – and rises by 1% per N days until a new ecological balance is achieved.

This all becomes possible through the use of Master and Child Domains, something I’ll get back to in a moment.

Dynamic Population Reassessment

Nature is never static, and an ecology is never fixed. There’s always a new population trying to invade from a neighboring region, if the terrain and climate are suitable. Whenever I’m mapping a region, I’ll consider the neighboring regions, with an eye to those population increase/decrease markers. If there’s a region where the dominant predators are a pack of wild dogs and a neighboring territory has a surfeit of T-Rexes, there will soon be a new sheriff in town. The T-Rex population will expand into the neighboring region and the dog population will divide. Some will be pushed out of where they were, in a direction roughly opposite to that from which the T-Rexes were coming from, biased in the direction of whichever terrain/climate/food combination is most suitable – and what the predators are who already inhabit those regions. Domino will follow domino. The remainder will settle for being less than the dominant predator and simply become the dominant pack hunters – with lower numbers and a greater range, because the T-Rex newcomers will eat a lot of what the dogs used to consume.

Even if there is little or no scope for population changes of this sort, some creatures like Dragons and Wyverns, can skip completely over an unsuitable territory to invade some area completely separated from their original terrain. Part of the assessment of any ecological domain is a description of how the population of that domain is going to change with time, assuming current trends continue.

Another essential element of the description of an ecological domain is the impact of migratory patterns. Many animals have them – everything from birds to deer. This Wikipedia Article serves as a hub to a number of more detailed articles on the subject, which are worth reading. These in turn have effects on the creatures that feed on them, who employ various survival strategies in response – some animals hibernate, others follow their prey, and still others switch their appetites to some other source of nutrition. I’ve never heard of an omnivore which is a carnivore part of the year and a herbivore the rest of the time (a biovore?) but in a fantasy environment such a thing should at least be possible. It’s even possible for different predator populations to alternate dominance with the seasons.

The easiest way to approach incorporating such seasonal changes into your encounter table is to assume that the overall groups don’t change their representation, only the individuals within it. Each table entry thus becomes two-to-four separate entries, one for summer, one for winter, and zero-to-two for the intermediate periods of spring and autumn. Facilitating this approach was a key consideration in the general ecological approach described in the previous article.

Don’t forget to allow for climatic effects like seasonal floods and streams, and for the impact of seasonal changes to vegetation. If the food goes away, one way or another, so do the things that eat it – followed by the things that eat them, and so on all the way up the food chain. It can also be worth spending a moment contemplating how an animal’s temperament might change with the seasons. When factoring in the question of who eats what in this way, it’s usually easiest to start at the bottom and work up, in the opposite direction to that which was employed to initially populate the table.

The Turbulent Wake Metaphor

Overall, even with all these population dynamics, the end result is a relatively stable ecology which changes over time – if you omit the impact of intelligent beings, who (at the very least) include the actions of the PCs. By virtue of their interaction with the environments they pass through, these act as a source of turbulence within an ecological map. They take out an apex predator here, scatter a pack there, compete with almost everything for food, take seeds and nuts and fruit for their own consumption (reducing the numbers of viable seeds to form the foundation of the next generation).

The more tightly confined and regular the pattern of travel, e.g. along a road or routinely-used-by-travelers track, the greater the opportunity is for accumulated disruption of the natural ecology. Consider the picture to the left: it is a false-color image of the turbulence resulting from a jet of fluid (at the left), but it could just as readily be a map of the ecological disturbance of a traveler going cross-country (right to left, with the traveler himself in the middle of the red zone), or of the ecological disturbance along a road (with a village in the red zone).

It can even reach the point where you consider the roads (and a narrow span to either side) to be a separate ecology to that of the surrounding terrain.

Here’s an example: Fruit from orchards to the west of a town are routinely consumed by travelers moving from the town along a road to the east. As they do so, they scatter the seeds from the fruit, a few of which germinate. Over time, the road to the east becomes lined with a small population of wild fruit trees, displacing some of the naturally-occurring trees, which are occasionally cut down for firewood. Different small creatures eat the fruit, who follow in the wake of the fruit tree population. The result is that the ecology to the west of town begins extending a pseudo-tentacle along the road to the east. At the boundary of the fruit-tree populated zone along the road, the two ecologies come into conflict – but one is being sustained as a byproduct of human activity, while the other is ignored, or occasionally disrupted, at best. This gives the new population a competitive advantage that is more than enough to sustain it, and may be enough to permit it to expand. Now extrapolate over a century or two…

Master & Child Domains

I made this point towards the end of the previous article, but in this context, it bears more detailed scrutiny and explanation.

Consider the original table that you derived to be the “Master Table” for the climatic/ecological region. This is the base model for the region. Then look at the entry within the table that has the lowest population – usually that of the apex predator. The range of that predator – which factors in the need for sufficient food supplies – is the size of a child domain. Divide the map (figuratively) of the region up into that many specific child domains, and assign each a “virtual copy” of the master table. The results are remarkably similar to a map of counties within a region or state in appearance, but what you are actually mapping are the population of child tables.

When the PCs have an encounter in a wilderness zone, such as the one highlighted in the map (which is actually a map of Iowa highlighting one of the central counties, Hamilton), the encounter table for that specific child domain becomes real instead of virtual, and the consequence dominos flow outward from it. In this case, because it is a central one, those changes will be relatively small and temporary, with one of the surrounding regions supplying a replacement that is virtually identical with what was there before. It’s sufficient in such cases simply to note that there is a temporary change to that child domain. However, domino effects will still travel to an edge child, as population pressure will push a neighboring example of the affected population into the newly vacant territory, and then another will move into that new vacancy, and so on. In effect, removing a predator from a central child domain ultimately removes it from a (randomly selected) edge-located child domain. The larger the ecological region, the more likely it is that one child domain will have a newly-adult predator of the required type who can find his way into the newly-vacant territory, leaving no trace of the original disruption.

When the number of children is relatively small, as in Map 2, things become more interesting. Either directly or indirectly, the encounter has produced an job vacancy, and all sorts of critters from both the yellow (white) and neighboring (yellow) zones will show up for the interview. There are three possible outcomes:

  1. A representative of equivalent type from zone yellow will emerge to claim the vacancy, and all will go back to normal.
  2. A representative of equivalent type from zone red will move in to claim the vacant position, creating a hybrid zone. Depending on the differences, this might transform zone red into a new yellow zone.
  3. A species that is more dangerous in numbers will step up from either zone yellow or zone red to create a new (and very small) ecological pocket in which they are dominant.

The example maps were carefully prepared to illustrate one of the ramifications. Notice that the child domains of zone yellow are roughly double the size of those in zone white, including the red child domain where the loss has actually occurred. That means that Zone Red is probably not large enough to sustain a stable population of the zone yellow dominant predator – so if that’s the vacancy created by the encounter, it may be “expand or perish” for any newcomer from Zone Yellow. They will certainly have to become more aggressive than they were – and that means that any future travelers will probably have trouble with them. Which may well be the PCs going back the way they came…

Fixed Points in an ocean of Chaos

The boundaries of any ecological population are, as you can tell from the preceding, continually in flux, shifting this way and that like a quivering soap bubble. The major reason why it’s worth tracking all this stuff, even in general and abstract descriptive terms, is the impact that these changes have on the circumstances surrounding those few fixed points in this ocean of chaos: habitations and settlements. If you know that a township has recently stopped having trouble with wolves and started having trouble with trolls, it gives you a massive leg up on the verisimilitude of the settlement and its human population. Unemployed fur trappers. Guard dogs who run in terror from trolls but kept the herds of sheep safe from wolves, and the breeders and trainers who are seeing their own livelihoods diminish and vanish. Townspeople with improbable schemes for driving or luring the trolls away. Tracking the changes and asking yourself, with each rural encounter, how the subject of that encounter would have been impacted, both directly and indirectly, by the changes brings not only the ecology but the settlement to life. They have something to talk about, something to do, and something to react to. Of course, they might not fully appreciate, when hiring the PCs to chase away the trolls, that the reason they are menacing the town is the population of Gila monsters expanding to the south – a menace that the trolls actually shelter the townspeople against. The list of impacts just keeps growing; different needs will frequently influence architecture, for example. Doors built to a certain standard (assuming that the wolves would avoid inhabitations) might be no barrier to trolls. Lamps lighting the entrance to the inn might be an invitation to lunch for the trolls, or they might hate the light and repeatedly throw rocks at it.

The ecology of the wild has a direct impact on the ecology, society, and population of the “civilized” regions that border it.

Expanding the premise

By selecting a different set of encounter classification criteria, the ecological premise can be reframed to generate encounter tables for some very different circumstances. By way of example, I’m going to examine the Urban Encounter and the Dungeon encounter, but (even though they would account for 99% of the non-wilderness encounter sets), this is only the tip of the iceberg. Depending on what you are trying to model (in terms of encounters), there are many others that might be valid or useful. In any circumstance in which the concept of a “Food Chain” is analogous, or even simply a metaphor, the ecological system can be used as the basis of an encounter table. For example, contemplate the ecology of undead, and the flows of positive and negative energy through the prime material plane, or the ecology of different forms of arcane energy. Don’t see how? I’ll make the same suggestions after you’ve read the next few sections and expect to get a different answer when I do…

Urban Encounters I: Food

Applying the common ecological basis to an urban community is ultimately a way of tracking the flow of food through the urban culture – in other words, who eats what, what crops will be most common, what the dominant land use will be in those parts of the neighboring regions that are cultivated, and so on.

Taking the pre-existing Master chart for the region and replacing the bottom layer with the cultivated foods then enables you to track the ecological impact and reaction. If you take away a creature’s food supply and replace it with someone else, they will switch to the new foodstuff if they can (becoming a pest to be eradicated from the region), or they will move, or die out. The shockwaves travel up the food chain from the bottom to the top. Some creatures find themselves in an undreamt-of cornucopia, while others will starve until they vanish from the region – impacting all the neighboring terrains – or find another edible food supply. Like, say, people.

Urban Encounters II: Economics & Obligations

The same principles can be employed to model an economy, from those who generate low-levels of income and live modestly to those who have great wealth, and commensurate obligations. All that needs to be done is to return to the basics of the ecological model and determine, “What’s the equivalent of small plants?”, “What’s the equivalent of a Pack Animal”, and so on. Who are the primary producers, who are the consumers of the wealth generated by the primary producers, who feeds off the wealth that these feeders accumulate, etc. This also includes the criminals, from the petty (scavenger equivalents) (which would also include beggars) to the organized predators of a thief’s guild.

While of low utility on its own, in combination with an appropriate ‘traditional’ encounter chart that puts an occupation or character class on an individual, an economics-derived chart makes it far easier to generate potential encounters in terms of relative frequency; and the subsequent refinement methods for excluding and generalizing to sift the significant encounters from the inconsequential give you a tool for establishing exactly where in the economic hierarchy a particular encounter falls. Wealthy land-owners and destitute serfs, impoverished monks and well-appointed heads of religious orders, all are presented in their correct relative frequencies of encounter.

Urban Encounters III: Politics & Influence

In a similar fashion, you can trace the avenues of political significance and power. Combining the results of such a population assay with those of the previous urban encounter forms permits the correct attribution of those with power and those with wealth, those with both, and those with neither, by occupation or character class. A Fighter with both wealth and political power vs. a cleric with power but no wealth vs. a mage with wealth but no power – all fall naturally into place in their correct proportions within the given population, adjusted for significance of encounter to the PCs. All that remains is to rationalize these attributes into a determination of the status of the individual so described.

In terms of a general urban population, I generally relegate the “politics and influence” results to a secondary state relative to those of wealth. That is to say, either the politics and influence result explains the source of the wealth or its absence, and hence the encountered character’s status within the social hierarchy, or it tells me that the source is something other than evident position within society. Either way, a random urban encounter is transformed into an encounter with an individual.

In contrast, when generating an encounter table for a close population, such as those who may be encountered within a palace or castle, proximity to power is going to be more important than wealth. The entire setting is geared towards the relationship between the ruler and the populace, including his civil servants, advisors, social circles, generals – and irritants. In this situation, wealth either explains or accompanies the political influence, or the influence derives specifically from a source other than personal prosperity. Once again, the result is the delineation of individuals of significance – in appropriate ratio of significance of encounter.

The price of such a narrow focus is that virtually all relevance to the standard ecological encounter table is now lost, though there may be individual reactions to events generated through such a chart. One wolf attack too many, and the Count might order (and possibly even organize) a cull. So there is an interaction but the two exist in isolation. In essence, if you have power, you eat the best and you eat what you want – so long as you can afford it. The Economic flow is the interface between Politics and the emergent ecology of the region.

Urban Encounters IV: Religion & Philosophy

From the very localized, let’s zoom back out again, this time to something approaching the national, or even the international scale. At this scale, the flow of religious authority can be considered analogous to the population of an ecology. Different theologies yield different ecologies, overlapping in some areas, and always competing for the hearts, minds, and souls of the populace. Mapping out where different religions and different deities are most authoritative reveals patterns and settings for doctrinal confrontations. This can have a substantial contributive effect to defining the relations between neighboring settlements, and taboo products can influence trade and commerce. Religious differences and tolerances can color relations between individuals within a community and between rulers and ruled. It can provide a valuable foundation of context for other aspects of encounters.

Urban Encounters V: Knowledge & Ideas

In a similar fashion, it’s possible to map out the spread of new ideas, knowledge, and technologies. A better plough here, a better stirrup there; a community that is technologically primitive in one place against one that is progressive in another. Conservatism and tradition vs. Liberalism and the embracing of innovation. Once again, this can provide valuable context, and when combined with the religion & philosophy charts, historical foundations to an urban settlement. Consider, for example, a township that follows Bartier The Enlightened which is isolated in this respect from its neighbors (the GM rolled a low-probability ‘encounter’ on the religion table when defining the town). When generating an encounter table for the spread of knowledge and ideas from one county or province to another, he happens to roll a result that suggests that this particular region is an arch-conservative one and far more regressive than its neighbors. That, to me at least, suggests that the township still adheres to the “old faith” – and that, decades or centuries before, the worship of Bartier The Enlightened was general throughout the region, but that other faiths have been eroding that support. This would influence the relationships between the different faiths, and those relationships would in turn influence relations between the settlement and its neighbors. These facts assist greatly in delineating both the population in question and its neighbors; once again, they have something to talk about, something to do, something to have opinions on, and something to react to. They are inherently more unique and interesting than ‘just another village’, and that rubs off on its citizens when kept in mind while generating those individuals.

Urban Encounters VI: Alignment & Society

The final ecological equivalent to discuss is one that doesn’t appear (at first) to be appropriately analogous to a dog-eat-dog ecological model – alignment and social structure. It’s only when you’ve conceived of the economic and political analogues, and extrapolated out to the religious and intellectual analogies, that you can start to see an ecological model for these factors. Specifically, acting on the assumption that like will tend to attract like, you can ‘seed’ an empty political map with areas where economics and weak social values (low religious strength) make a relatively welcoming setting for ‘lawless’ elements (chaotic evil), and others where law and order and high religious strengths indicate a more “lawful good’ approach. Like a rising flood, these will tend to expand out, following the lines of easiest settlement – in other words, topographic features – until the gap between them is roughly equal to the population levels of the seeded regions. Observing the resulting maps enables trends which will tell you which initially “chaotic evil” regions would remain lawless and which will be “stabilized” by lawful evil, and similarly for chaotic good. As a general rule of th7umb, lawful regions will tend to be at the centre of a region, or regions that are strongly protected by natural barriers; the more unpredictable life is in a given location (ie, at the fringes of the civilized world), the more socially acceptable the ability to react quickly to changing circumstances would become.

Once you have the predominant alignment of a region mapped out, it becomes simple to extrapolate the impact on the local social structure. They will trend toward either bureaucracy and authority or a more “wild west” attitude of “don’t ask, don’t tell, who you weredoesn’t matter as much as who you are.” The bandits may run roughshod over the local government (such as it is). These will be places where someone’s past can always be left behind for a fresh start. Once again, this can provide invaluable context for individualizing a settlement and its populace, but it’s only really possible once you have the earlier encounter structures locked into place.

Dungeon Encounters I: Light

The dungeon environment is one of the most ecologically problematic settings conceivable – at least until you start talking about exotic planes of reality. And yet, as supposedly a part of the natural world, ecological principles should hold sway there as much as anywhere else.

So, here’s a bombshell: Plants are at the foundation of every food chain, and plants need light. It need not be sunlight, and it can even be an alternative energy source (such as the use of volcanism by Tube Worms like the one shown to the left, but in terms of an organism that can sustain higher forms of life and form the foundation of an ecosphere, light-dependant plants are by far the most efficient, as shown by the fact that they dominate completely everywhere where there is light to drive photosynthesis processes.

In order to form the foundations of an ecology, you need light – unless you want to get creative. There’s nothing to stop you, it’s just an exercise in creativity. Inspired initially by Tolkien’s White Tree, I’ve populated one version of the Underdark with a whole range of albino plant forms – from a white tomato-like fruit to something akin to white paw paws, to a grass-like white moss – that grow in little ecological pockets beneath the earth. There’s even a giant tree-like organism that looks something like an asparagus crossed with an oak, and grows down and outwards until it is several miles across. It extrudes a single small green tendril to the surface which accumulates sunlight for a century or so, building up the energy needed to bloom and create a new generation of seeds; the rest of the plant withers and dies in direct sunlight. In addition, there are always fungal forms.

The basis of any Dwarven or Dungeon ecology has to be either created out of whole cloth, or you have to find some way of getting light into the equation – even if that involves a host of small creatures like worms that consume surface plant matter and convey the nutrients to intermediate organisms beneath the surface. These would be smaller than purple worms, but that’s not necessarily saying much – they would need to be the weight of a cow, and need to travel in herds. Something akin to small mammals, lizards, snakes, or “giant” subterranean spiders would then need to consume these – something more mobile and able to distribute themselves throughout the dungeon environment. The next tier up of the ecology then feeds on these creatures, while the dominant creatures that populate the dungeon then consume both that tier and the small critters that come their way – and the occasional foolish adventurer.

To some extent, the same rules apply to the underground world as do the surface. In other ways, the earth can be considered an especially viscous fluid that certain creatures can burrow or tunnel through, and which can do some of the work of supporting the creatures in question, which removes one of the biggest impairments to creature size in the same fashion as life underwater. So some organisms can be abnormally large. The problem is that rock poses a whole different problem – these creatures either have an intensely acidic coating (to which they are immune) that can break down the rock, or they can’t work in the underground environments that are suitable for more interesting lifeforms.

One of my favorite solutions from early in my days as a GM (when I first started considering this problem) was the Phase Moth. Partially ethereal, these emerged only at night and then, only when the moon was in the sky; with a wingspan of anything up to that of a large dragonfly, they ate vegetable matter on forest floors (fallen leaves) and the occasional bush, then returned to any convenient hollow under the surface when the sun rose. When in the open air, either above or below ground, they emitted a pearlescent light similar to that of the full moon, which was enough to feed additional plant varieties that otherwise would not have survived, so not only were they an insect-level ecological layer, but – in the reverse of the usual insect-plant relationship – they fed the plants which other small creatures then consumed.

I mention these solutions because dungeon populations tend to be ecologically-isolated, and we all know what that means, especially in Australia, where we have wonders like the Kangaroo and the Duck-billed Platypus. And where a vast percentage of the spiders and snakes are exceptionally venomous :) Quite frankly, every solution you can think of will not be enough, because no one solution would be anything even approaching universal.

Dungeon Encounters II: Water

The second key ingredient for life – one that all orders of an ecology require – is water. Mapping out a supply-and-demand chain for how water reaches the higher forms is an absolute essential to constructing a plausible dungeon ecology. And that means that it’s time to get creative again. Everything that was said above with respect to light also applies, in spades, to water distribution. Like the light distributors, this is so important that it will form an additional ecological layer.

Dungeon Encounters III: Fresh Air

At best, air would not circulate very effectively deep underground, unless you find a way to make it happen. The smoke from a single torch can be enough to suffocate whoever is carrying it, several times over. Check out this section of a relevant article at Wikipedia. Heck, even the byproducts of normal respiration can be lethal if not refreshed. Here’s some food for thought: Fresh air contains roughly 21% oxygen and less than 0.04% Carbon Dioxide. Exhaled breath contains only 13.3-16% Oxygen (so 5-7.4% has been consumed by the process of respiration) and 4-to-5.3% Carbon Dioxide. If you were to re-breathe this exhalation, you would have to breath up to 1.55 times as fast to get the same quantity of oxygen into the lungs; and what you breathed out would contain only about 6.2-11% Oxygen, and would have a CO2 content of 8-10.6%. Now, 7-10% CO2 is enough to cause suffocation in anything from a few minutes to an hour – and that’s without allowing for the byproducts of combustion of torches or anything else – according to this section on Co2 toxicity at Wikipedia.

Some form of air filtration is an essential – the best one being photosynthesis – but, even then, you will need some means of circulating the fresh air. Together, these needs constitute an entire necessary layer to any dungeon ecology. You can use green plants on the surface and simply recirculate air from the upper levels of the dungeon, but for reasons I’m about to make clear, this is not an acceptable solution.

There is an additional requirement that makes these requirements especially important to contemplate: if the PCs take out the air circulation mechanism or organisms, they will kill almost everything in the dungeon. They can then replace the air circulation system with something else and loot and pillage to their hearts’ content.

I don’t know about you, but I don’t consider this an especially desirable outcome. In fact, I don’t want it to be possible – at least until the dungeon is cleared, and if I can manage it, not even then. That means making the recirculation mechanisms/organisms beyond the reach of the PCs in some fashion, and hiding them at the very, very bottom of the dungeon. Oxygen security is critical to the ongoing existence of your dungeon.

Dungeon Encounters IV: Mobility

Only once these foundation layers are in place can you establish a realistic and otherwise fairly straightforward “this eats that” ecology for a dungeon setting. But there is a third environmental characteristic that you will need to think about – Mobility.

On the surface, you can move in two dimensions with relatively little obstruction. There is the capability for some species to move vertically, as well. There is a volume to be occupied, within which the mobility of life in general is pretty much unconstrained.

This is not the case when we’re talking about an underground environment like the typical dungeon. You can’t have locked doors without providing some means for food to get in – or without the residents opening those doors and coming out to forage. And when they do come out, many of them have extremely constrained mobility – they can move along existing passages and that’s about it.

Every time you emplace a “fixed” encounter within a dungeon, remember to check on the means it uses to access fresh air, water, and food – and modify the environment as necessary.

Exotic Premises Redux

With those concepts filed away in the back of your mind, do you now see how it would be possible to model an ecology of undeath by tracing the flows of positive and negative energies? How doing so would spark ideas about how and what undead actually are, and the relationship between the living and unliving worlds, and the implications for different forms of the afterlife, which in turn impacts conceptually on Gods, Devils, Demons, and a host of other “Outsider” encounters?

Or how you could construct an ecological chart for arcane energies that explains what they are and where they come from, how they are generated, what happens to those not harnessed by sentients, and what the consequences in your campaign of prolonged spell use might be?

How about a pseudo-ecological approach to the spreading of hair styles? Or women’s fashions? Or the spread of diseases? The only limits are those of your imagination, and your capacity to determine what is going to be useful.

Plot Integration

We’re almost at the home stretch of another biggish article. In this section, I’m going to (eventually) reject completely the definition of an encounter table that I offered earlier in favor of some alternatives that – at the appropriate times – can be even more useful. But I need to wrap up the primary subject of discussion before I can veer off on those tangents. The entire topic falls under the heading of Plot Integration – so let’s get started.

Non-Narrative Encounters

Using either a traditional random encounter matrix or the more accurate ecology-derived version that I have featured throughout this series, it is still desirable to be able to integrate the encounter with the main plot of the campaign as something more than simple random violence. This is so much easier than most people think that it isn’t worth a full article on the subject, which may come as a surprise to the many people who have struggled to do so through the years.

That’s because if you don’t have the magic key to doing so, it can be exceedingly difficult. There is a simple trick to it, though, and I’m about to share that secret with you.

Whenever a random encounter is indicated, and before the encounter begins, before you begin describing it in any way, think about how this encounter is going to relate to the nearest significant location. Make the determination that it WILL relate to that location, and the encounter(s) contained within it, then create a situation and not an encounter. The creature might be trying to prevent the characters from proceeding, or it might be fleeing the location, or nervous while in the vicinity of the location, or subservient to someone/something at the location, or hungry because something/someone at the location has been consuming or scaring off it’s food supply, or have been wounded by whatever is at the location, or driven insane by it, or be searching for it, or may even have attacked and consumed the creature there. Add to the mythos and legend from the adventure backstory as necessary. Did you know that Owlbears like to eat books? Many a time I’ve had one cough up a fragment of a page when encountering the PCs en route to a dungeon; it drives the players nuts to think that things have been eating their potential rewards!

There are so many interaction modes between location and encounter that it’s hard to list them all. The list in the preceding paragraph is by no means exhaustive. But if you figure this out in advance, you can then shape the encounter around this central fact, and suddenly it’s not a mere hack-and-slash for its own sake – it’s plot-relevant.

It’s that simple.

You can even come up with a long list in advance and either roll randomly or pick an appropriate one off the list when the time comes, depending on the nature of the encounter.

And of course, it’s even possible that it’s unrelated (especially true in the case of plant encounters), or is a red herring (also in the case of plant encounters). Or perhaps the leaves of that particular rare herb are absolutely necessary to healing some nasty poison in the location. Or the plant might show signs of having been chewed on or scratched up by some more significant threat.

If you can’t relate it directly to the plot, try to connect it with the backstory of the location. If you can’t do that, attempt to relate it to some other potential random encounter from your list that CAN be related directly to the location in one of the manners described. Eventually you will succeed in transforming the significant into the relevant.

Narrative Encounters I: Relationships & Circumstances

Okay, so let’s look a little further into left field and away from the general concept of an encounter table.

Instead of defining a table entry as a creature to be encountered, why not define an entry as a relationship to be encountered, or a circumstance, or an event? Then choose the creature or creatures to be encountered on the basis of that relationship, circumstance, or event.

The big advantage to this approach is that the creature being encountered, if any, are always in the act of doing something.

The colossal downside is that there are a lot more entries to think about. Any given creature can be doing any one of half-a-dozen things or more, especially if there is a second variety of creature involved. Even something as simple as a predator eating a herbivore: the predator could be tracking, or stalking, or chasing, or killing, or eating, or being driven off by, the herbivore(s), depending on what defenses they have. Or perhaps another predator (who doesn’t much like herbivore) has been hanging around the herd, silently, in ambush, waiting for the carnivore to show himself – especially likely if the third member of the tableau is a hunter, or attempting to remove something that’s become viewed as a menace to the local farmers.

Throw in all the other possible animal behaviors, in all their phases and stages, and you get a glimpse of just how comprehensive any useful table would have to be.

Employing this approach, one Master table would be enough; there would be no need for separate Child tables. You could keep track of the distinguishing features of individual significant encounters for reuse another time around, but there would be so many possible “encounters” that you would need to arrange them in subtables just to give them a 1% chance of eventuating.

A far more useful approach might be to create a general table, and use it as a “random activity roll” in conjunction with the ecology-based encounter system. Not only does the process of creating the ecology get you thinking about the different behavioral entries to go on the list, it also retains the “balanced ecology” benefits.

Here’s another wsy to think of such a “random activity table”: it provides context to wilderness encounters in the same way that the economic and political encounters gave context to urban encounters.

Narrative Encounters II: The B-Plot

It’s possible to extend the concept of the “random activity table” even further, and in this format it also becomes useful as an urban encounter engine.

Come up with a B-plot and dismember it into as many individual scenes as you can. Any random encounter roll indicates that another scene from the B-plot has occurred. If you get to the end of the B-plot with the characters staying in the same general location, start another one.

There is an art to devising appropriate B-plots for this purpose. If the game were a TV show, you could simply switch to whatever location was appropriate to driving the B-plot forward, but you don’t have that luxury; the location will be at or near wherever the PCs are when the scene comes up. They must be geographically independent.

The best solution is to not to make the PCs the stars of the B-plots at all. Instead, have them happen around the PCs, to one or more NPCs. Or to an animal. It takes practice, but it can be a viable – if somewhat soap-opera-ish solution to the questions of random encounters.

Encounters with a purpose: The wandering dungeon and other applications of plot

Another potential to consider is to take the concept of a “random activity table” and revise the entire primary plot so that it ALL takes place through random encounter tables using the normal activities of the inhabitants of the region – whether that be wilderness domain, settlement, dungeon, kingdom, nation, or continent. That means that you don’t know where and when a location encounter will take place – but the creatures encountered at s location will always be doing something relevant to their existences, circumstances, and objectives, and not simply frozen in place waiting for the PCs to show up.

Downside: it’s still more work. Upside: it really does make the plotline a dynamic series of events that make the whole thing seem far more real. Verisimilitude is your reward.

The Worst Possible Solution

Having read this entire series, I think it is pretty obvious why I consider the traditional random encounter table to be the worst possible solution to the problem of applying randomness to an environment with which the characters are interfacing. I am quite sure that most, if not all, of the perceived problems with random encounter tables result from the limitations of the traditional concept. The lack of value and meaning, the problems of the xp-and-loot giveaway, they all go away if the random encounter means something. The trick is to find a way to infuse them with such meaning without making it more work than it’s worth. It’s been the goal of this trilogy of articles to offer a set of scalable solutions to that conundrum.



Discover more from Campaign Mastery

Subscribe to get the latest posts sent to your email.