Saturday, October 9, 2010

Problems With Encounter Tables

Encounter tables exist to provide random features or creatures for the purpose of inspiring play. A random dungeon generator creates the physical conditions of the setting, as well as the inhabitants. A random wilderness generator intends to place population centers and notable features into a mapped hex.

An encounter table should not only include monsters. It should include every kind of feature, event or inhabitant that might be discovered or met by the party.

Which means that first and foremoest, encounter tables need to be divided into those conditions that might affect a party that is remaining in one locale, as opposed to one that is actively on the move. The small village would still experience events such as storms, fog, floods, housefires, earthquakes, death of a leader, increased taxes or new laws; and encounters such as thieves, vermin, inquisitive residents, random associates or outside attacks; but obvious physical features, such as a freshwater lake or a mountain in the shape of a head, would not be included.

However, a travelling party might encounter such a mountain, and many other potential features such as lakes, ponds, chasms, creeks, shrines, depots, crossroads, monster lairs, toll booths and so on. So to begin with, a 'Stationary' table would alternate with the use of a 'Moving' table.

Even at that, Moving tables must be subdivided further according to the sort of terrain, climatological condition, or level of civilization that they supposedly represent. A single table could not be used for both mountains and lowland plain. A single 'lowland plain' table could not be used for lowlands for sub-tropical jungle flatlands and for vast arctic-tundra shelves. A single 'jungle lowland plain' table could not be used for both heavily populated, cultivated lands (such as Benin) and sparsely populated hunting lands (such as pre-Columbian Florida). Moreover, floods occur only in certain locations and at certain times of the year. Volcanos, earthquakes, hurricanes, tornadoes, blizzards, fog, the accumulation of snow, rime or ice, and other climatological incidents are similarly limited.

The same elements would not be encountered, leading to the familiar pattern of 'rolling on the encounter table until getting the result you want.'

Even within an urbanized setting, where the party might be stationary, encounters would depend upon the size of the urban setting (patrons and lawyers do not dwell in villages), the characteristics of the neighborhood (slum vs. posh), the city's economic base (holy city, port, farming centre, military outpost) or the characteristics of its people (bureaucratic, rigid societies as opposed to freewheeling entrepreneurial piracy). Each different kind of urban center requires the recording of different features, different likely personal encounters (leading to adventures built upon differing elements of success) and different hazards which might occur.

The variety of required encounter tables described above is a finite number, and - with tenacity and patience - could be created for use. A secondary problem arises, however, with the variety of items that might occur upon a particular table. If the party should remain for an extended period of time upon an uninhabited tundra shelf-land, there must be enough possibilities entered in the encounter table to ensure constant novelty. Encounter after encounter with wolves pales quickly, as does the fourth mountain indicated that has the shape of a head. Considerable ingenuity must be employed to create literally hundreds of possibilities for every kind of encounter format - a daunting, near-impossible task, particularly if one considers that any encounter table meant for use in multiple campaigns must not be campaign-specific (or edition-specific) in its random results.

Also, encounter tables must be constructed so that certain features or events are so unlikely as to ever happen again that the result itself should be stricken from the table. For example, a meteor strike (such as the Tunguska hit of 1908) is such an unusual event that, although it might conceivably occur twice within one's lifetime, the actual incidence of it occurring a second time in a D&D campaign would be farcicle, ineffective in rousing the party's attention (been there, done that) and ultimately damaging to the memory of the first event.

At the same time, numerous events of a particular, common type - say, snow - can serve to increase the tension as they happen. As such, not all events should be stricken once they have occurred.

Finally, in my experience encounter tables should not be based upon the likelihood of an encounter or an event to occur. While it makes sense constructively, within the game it only serves to reduce the incidence of interesting, highly playable circumstances in favor of dull, common conditions. Once again, the continued result of 'wolves' as opposed to something more challenging. While there is something to be gained by the repeated incidence of mundane encounters, it should be balanced in some degree with things that will not suffocate a party in banality.


  1. Parties with riding beast and pack animals would also have a different range of encounters/events to contend with as opposed to those on foot.

    Hireling/Henchman interaction with the rest of the universe might be worth considering as well.

  2. "Finally, in my experience encounter tables should not be based upon the likelihood of an encounter or an event to occur. While it makes sense constructively, within the game it only serves to reduce the incidence of interesting, highly playable circumstances in favor of dull, common conditions."

    Using the wolf example...

    The party is out in wilderness, crossing a high tundra to get to the forgotten temple and answer an ancient riddle.

    If you roll "wolves" three times in succession and make the party fight the wolves each time, then they are going to get bored. If you roll "wolves" three times in succession and tell the party (1) they see wolf tracks, (2) they have come across a recent kill, and (3) they see a sentinel wolf watching them from the next ridge, I don't think they will get bored. You aren't making them spend time or resources that they would rather reserve for the temple.

    So leave the common events common, just be ready to gloss over them. "It snows another inch over night, but the next day dawns clear and cold."

    I suspect that you could get away with more wolf attacks if there were some justification for them. For example, the party is traveling in the dead of winter, with lots of snow on the ground. The wolves are hungry and desperate enough to take on armed humans.

  3. Fair enough, Derek, that works, and in fact I've done things like that. But it works for one running.

    The wolf example, by the way, comes from the DMG, where for arctic conditions there's a 50% chance of getting a wolf, a winter wolf or a herd animal.

    You can't come up with clever wolf scenarios running after running without your campaign becoming a joke. Fact is, your suggestion could also be employed for one result of 'wolf' ... and the way I like to play, there is almost always some sort of lead in to encounters.


If you wish to leave a comment on this blog, contact with a direct message. Comments, agreed upon by reader and author, are published every Saturday.

Note: Only a member of this blog may post a comment.