Building Encounter Tables

The "random encounter" table is a long standing staple of D&D, injecting some uncertainty into play while at the same time serving as a window into the game's world (what kind of creatures live in the area? how common are they? etc...) But creating a good encounter table can be difficult. Too often, DMs and designers throw together a collection of creatures with little thought to how they might interact with the setting and one another, and fail to include setting elements that are otherwise pervasive. An encounter table is a "simulationist" construct, and as such requires more from the DM or designer than simply wondering what would be "fun" to fight between the "real" adventure scenes.

There are four basic issues at hand with a random encounter chart: what is the scope of the encounter chart, how likely is an encounter to occur, how likely should any encounter type be to occur and, most importantly, what kinds of encounters should be included. As these questions are answered, the encounter chart will take shape. Once it is complete, the encounter chart will not only serve as a tool for use during play, but also as a part of your general world building process, both informed by and informing your setting and the PCs' place in it.

Scope
The scope of an encounter chart includes the climate and terrain type, the geographic area, the time of day, possibly the time of year and other, less concrete aspects that are dependent on your setting. One can build an encounter chart for a particular forest, for example, or that same forest during the day versus the night, in the summer versus the winter or even for only during unusual events like solstices or magical conjunctions. But encounter charts can be used to represent less obvious game settings as well: an encounter chart just for royal/noble balls and courts; an encounter chart for guild wizards' towers, regardless of location; an encounter chart for the one night of the year where the doors to the land of the dead are flung open.

In choosing the scope for the encounter chart, start with a discrete element such as a contiguous, homogeneous terrain type (a forest or a particular city neighborhood) or a thematically coherent element (such as night of the full moon or festival days). Even if you choose a contained environment, consider how particulars like the time of day (or night), the season or weather will affect the encounter table. In the case of thematic encounter tables, these may also apply (the docks quarter of a city is a very different place in the day than in the night) but cultural impacts must be considered as well (the bazaar during festival times, for example).

Try not to be too broad or too narrow in your scope. Too broad a scope breaks plausibility. For example, having one encounter chart for a trade route that runs for hundreds of miles through multiple kingdoms and terrain types is too broad. Consider instead creating an encounter chart for the road in various terrains or nations, or adding a range of options to your usual encounter charts for those things when the PCs are traveling the trade route. Making encounter charts that are too narrow dramatically increase the workload of the DM, even if they promote variation and detail in the setting. Finding a balance is often a process of trial and error, and the definition of “too broad” and “too narrow” varies from DM to DM and even setting to setting.

Probabilities
Determining the likelihood of any encounter occurring and the likelihood of a particular encounter occurring seems simple at first, but in reality it can be complex and a wrong decion can result in unintended consequences: characters can end up with nothing happening for long stretches, or they can be constantly beset by encounters. Moreover, the scope of the encounter environment will have an impact on the probabilities, both of whether an encounter occurs and what types of encounters are more or less likely. More so even than scope, finding a balance of probabilities is more art than science. these few guidelines are presented as just that: guidelines. In a best case scenario, they give the DM a reasonable place to start.

When setting the likelihood of whether an encounter even occurs, and how often a check is made, consider the population within the scope – not the general population, but the population of unusual things. Encounters are by their nature out of the ordinary events. A forest encounter chart, for example, need not include squirrels or deer or wild grapes. These things are everywhere in a forest and characters are assumed to be “encountering” them all the time. The orc raiders, elf wardens and owlbears, however, are unusual denizens of the forest. If the forest is wild and untamed, on the edge of civilization or protected from loggers and settlers by a druid protector, the population of these unusual elements is higher. Conversely, the lord's hunting grounds with a regular staff of caretakers will have a much lower population of unusual elements and therefore a lesser chance of en encounter.

Generally speaking, areas or situations with a low population of unusual elements produce an encounter only 1 in 12 times per 4 hours, while those with an extremely high population will produce an encounter 1 in 6 times out of every hour. Most areas and situations fall solidly in the middle with a 1 in 6 or 8 chance every 4 hours. Of course, the probability may change with time of day – many monsters are nocturnal (as evidenced by infra- or dark vision) – without necessarily changing the composition of the encounters.

The next thing to consider is the probability of any given individual encounter type occurring. For monsters, “frequency” entries are of tremendous help here (though any particular area of circumstance may change those values – undead in a graveyard, for example), as are discussion of creature or character type ecology and/or sociology. Of course, most important is the DM's view of the setting and the scope of the encounter chart. In any case, a relatively wide distribution with variable probabilities is best, representing the variable nature of probable encounters within the scope.

Consider using 3d6, with one entry per numerical value. Rare encounters will occur at the extreme ends of the range (3 and 18 represent about a .5% chance of occurring), while more common encounters will bunch in the middle (10 is about 19% likely). Because we are all familiar, through character generation, with the general probabilities of the 3d6 curve, it isn't necessary to think too hard about the specific probabilities. Encounters the DM wants to have happen – based on preference or plausibility – irregularly are kept to the “edges” of the range and those he wants to have occur more regularly will automatically happen more often. The same distribution of likelihood can be generated with other bell curves of course, or non linear percentile charts. In some cases, the DM may desire more or fewer entires, so different curves or ranges may be required.

While I consider myself a “simulationist” and tend to design encounter charts around the setting, as opposed to the PCs, probabilities can just as easily be determined by “gamist” elements or “narrativist” ones. For example, in the former case the DM might create an encounter chart with 2d10 as the base range, but with entries going up to 30 or 40, where a modifier to the die roll based on character level is applied. More powerful or otherwise more intense encounters are placed on the high end of the probability spectrum, ensuring that low level PCs don't encounter unbeatable foes. In the latter case, encounter types with a strong connection to the campaign “story” will be more likely to occur, even if, in the setting, they are rare.

One last note as it pertains to probabilities: the use of sub-tables can be useful in expanding the range of a relatively small encounter chart. For example, an encounter chart might have an entry for “bandits” that leads to a sub chart indicating the bandits race, nationality and/or motivation. Be aware, though, that sub-charts do strange things with probabilities – if there is a 10% chance bandits will be encountered, and 1 in 6 of those bandit groups are orcs, then there is only about a 1.6% chance of encounter orc bandits. Go too deep with sub tables and specific encounters start becoming “one in a million” shots.

::continued in next blog entry::