Kowolski Nested Encounter (KNeE) tables

In last Sunday's post, I commented on the "nested" design of Zzarchov's random encounter tables used for wilderness sections in his various modules. I've dubbed these "KNnE tables" and propose to review/analyse them briefly herein, before I create versions for the two wilderness sections of the AtG project and/or it;'s successor RttWWP...

Introduced in Zzarchov Kowolski's 2013's Gnomes of Levnec (GoL), the "nested" encounter table has become a staple of his adventures, evolving over the course of the last 5 years into it's most recent incarnation presented in 2018's Shadows of the Forgotten Kings (SOtFK). 

The design harnesses the randomness of three different dice (d8 / d6 / d4) to generate combinations of three different components of a *wilderness* encounter:

• d8 "the Where" ­ the Notable Landmark ie. location of the encounter
• d6 "the What" ­   the Encounter itself eg a monster or even a noise/silence or visual effect 
• d4 "the Weird" ­  a Weird modifier to the combination or additional feature

This approach results in a staggering 8 x 6 x 4 = 192 combinations, comparing favourably with simpler "single dice" tables or the massive centile dice based tables from many older supplements. Even without the "the Weird" modifier, this creates a respectable 48 potential landmark / creature combinations, providing a greater richness to a wilderness encounter table than I've seen previously. A major aspect for me is the prominence of the random encounter location or setting for the scene in the table ­ this is an aspect often neglected in wilderness encounters but implied in a "dungeon" encounter, so it's a welcome inclusion!

So let's start with the basic 2013 format, as generated by the three dice rolls: 

KNeE Table v1 (circa 2013, Gnomes of Levnec)

As you can see from the above diagram, although the system has the ability to generate 192 combinations, at it's basic level it only requires 18 different ideas ­- 8 locations, 6 encounters and 4 weird effects. Yes, that's only 6 encounters (not necessarily creatures either), but "the Weird" roll allows the ability to modify the creatures, introduce a variation and/or combine creatures so that the small number of creatures doesn't seem to be a huge barrier. Creating separate KNeE tables for different "biomes" (such as in 2014's Under the Waterless Sea, UtWS) goes some way to addressing this as well, although it does mean more design work unless there is a significant overlap. This latter aspect is used with effect in UtWS, but at the cost of increased complexity and the need to cross­reference between the tables / flip back and forth across pages.

Note: it's worth remembering that these tables are designed to complement a "wilderness" section of an adventure, so given the low rate of random encounters per day of travel, I think for the relatively short overland journeys of most modules this should provide sufficient options. For a longer jungle expedition such as a Tomb of Annihilation­-sized trek through Chult not so much.

So let's have a look at the basic elements from the GoL KNeE Table:

Locations: 8 choices may seem like a lot relatively speaking, but providing variations or repetition can create a degree of complexity and emphasise the prevailing terrain. For example in GoL, two of the locations (rolls 1 and 4 on the d8) are repeated (rolls 2 and 5 on the d8) with a minor variation, so there's really only 6 unique locations ­ this makes the table easier to build but also introduces an element of increased frequency into an otherwise flat probability single dice roll.

Encounters: only 6 choices forces some consideration, but I don't think this is a bad thing ­ random encounters/creatures should preferably tie into the greater module somehow, either to foreshadow set encounters or if possible draw on and enhance the atmosphere of the area the adventurers are passing through. To me, this is just good basic design perhaps, but I like how this restriction, like the limited syllables of a haiku almost, offsets any temptation to just create the massive and often uninspiring tables seen in many an official line product. In addition, Kowolski uses non­creature elements: (eery) silence / nothing, the sound of a distant creature, a trap or even a visual environmental effect (eg school of brightly coloured fish UtWS) to provide a non­combat interaction. This breaks away from the bland "bandits or beast on the road" style of most published wilderness encounters that seem more about attrition damage or resource depletion ­ in itself not a bad element, but one that is often "hidden" and not properly accounted for.

Weird: this is by far my favourite random element. In some ways this mechanic is a thematic successor or variant to Lion Rampant's original Whimsy Cards (and more recent Storypath Cards) ­ a random non­dice based mechanic from the first two editions of Ars Magica that influence the plot of an adventure, allowing creativity and adding ideas to the story. Mostly this just an additional feature or item (eg. skull on a spike, abandoned equipment or cache, bubbling spring, shrine), an added environmental element (eg. silence, visual effect), or an additional monster. Sometimes, however, the result flips the whole encounter (eg. everything is dead, vermin infest the area) and/or creates a sense that there is agency in the game other than the players (evidence of a past encounter).

Trips and Dubs: this design element provides a variation or addition to the random combination of elements, allowing for "special" encounters 6 times out of 192 (or 1 in 32 / ~3%) when the dice come up triples (all 1's, 2's, 3's, or 4's) or *if the d6 and d8 result are the same* (two 5's, two 6's). In some modules, these encounters are completely different from the result that would be obtained by otherwise combining the random elements. In the earlier modules (GoL and UtWS), the DM / referee is left to combine the elements by cross­ referencing, but in SotFK the combos are now referred to as "Triples" and "Doubles" and the design has been improved to provide a full description of the resulting combination instead of leaving the prep undone.

Note: an expanded "doubles" option can be created if the doubles can occur on any dice, not just the d6 and d8 taken together. This adds two extra options to the 5's and 6's result plus 3 options for each of the results when the numbers are 1­4 ­ 14 extra options of a special encounter, tripling the chance for each specific combination of elements. For extra depth, the three different combinations (d8+d6, d8+d4, d6+d4) can have different results, but this may make the KNeE table unwieldy in practice and would require a lot more time to design.

Max / 18: this is the rarest result at 1 in 192 (~ 0.5%) and in the various modules offers a chance to stop being lost (GoL), encounter a more complex mini­boss style encounter depending on the biome (UaWS), or spot the spires of the lost city (SotFK).

Later KNeE Tables

In UtWS, each of the three underwater "biomes" has a different KNnE table, although there is cross­referencing between the various tables and some individual elements are reused, causing some confusion when reading off the combined result.

KNeE Table v2a (circa 2018, Shadows of the Forgotten Kings)

SotFK adds a completely new element, however:

*Sequentials: this is similar to the "Trips and Dubs" above, appearing in SotFK (2018) and essentially adding 4 "special" encounters when a sequence is rolled on consecutive dice eg 1­2­3, 2­3­4, 3­4­5, or 4­5­6, raising the number of "special" encounters to 10 (~1 in 20 / ~5%).

Note: if the exact order or dice the numbers appear on is irrelevant, the first three sequences can result from multiple die rolls and even the fourth sequence has two possible permutations (d8­d6­d4: 6­5­4 and 5­6­4). This latter interpretation adds greatly adds to the chance of a special encounter and each of the "sequentials" occurs at a higher frequency than the simple "triples" or "doubles" ­ there are 18 permutations, increasing the chance of a "special" encounter to ~1 in 10. If the "expanded doubles" option is used, there are 24 in 192 options (1 in 8 / ~12%) that produce a "special" encounter, which seems roughly equal to or better than older standard encounter tables based on d8/d10/d12 or a centile dice roll. However, this may be much more time intensive to design and only worthwhile for an extended overland trip. 

(Complex) KNeE Table v2b (circa 2018)

But What About Actions? 

One potential weakness of Kowolski's method is that despite all its apparent randomness, there is no random "verb" or behavioural element to the encounters ­ although a creature encountered may be in an interesting locale and affected by "the Weird" element somewhat, the behaviour of the creature (or phenomenon if relevant) needs to be pre­specified or otherwise generated by the DM / referee's imagination. For the "combination" encounters ("triples", "doubles", and "sequentials"), Kowolski seems to have incorporated a starting behaviour other than "attacks immediately", but for the other random combinations this element is generally missing or reverts to the default "attacks". 

This is perhaps less of a problem than first might appear if recognised. For many groups, this may be a relatively minor consideration in actual play given the "the Where" and "the Weird" elements greatly improve on the otherwise bland default "XdY creatures" with summary statistics of most wilderness encounter tables, but it does appear a potential opportunity for improvement. There are a few ways to address this concern, but each adds a degree of complexity. 

The first option is to use other characteristics of the numbers rolled (eg evens vs odds, totals, or the result of a particular die) to determine one of a couple of behavioural options outlined in the description. This does make the "the What" encounter description longer but adheres to the underlying Kowolski central philosophy of generating the whole encounter from a single roll of the dice. 

The second option involves an additional randomised element such as those outlined in Sham's d100 based Monster Business post (and collated PDF).  I'd note that Sham's option requires two additional dice and a perhaps overly complex table, possibly better used for determining behaviour for set encounters or "special" combinations or used to prep in advance. So I favour 1D8's "What are those wandering monsters up to?" post, based on a 2d6 roll. I accept that as originally designed it still requires at least an additional d6 (preferably different coloured or otherwise notable) and also adds significantly to the complexity, unless each encounter is prepped in advance: 

Sample Wandering Monster Behaviour Table

(roll 2d6, excerpted from the now-defunct "1D8" blog) 

• 2    Returning to their lair to heal up after a fight. (reduce hit points) 
• 3    Fighting with another creature. (roll up another monster) 
• 4    Returning to lair with a prisoner. (roll up a prisoner) 
• 5    Returning to lair with treasure. (roll treasure as if they were in their lair) 
• 6    Just passing by on the way to somewhere else. (not looking for a fight, roll for reaction) 
• 7    Defending territory. (war party looking for invaders) 
• 8    Hunting for food. (quiet, wary, and hard to surprise) 
• 9    Chasing after another creature. (roll up another, weaker monster) 
• 10  Running away from another creature. (roll up another, nastier monster) 
• 11  Building a new lair. (digging a hole, setting up camp, making a lot of noise) 
• 12  Groggy from too much partying and looking for a place to sleep it off. (easy to surprise) 

d8 plus d4 dice roll distribution

Note: the 2d6 roll preferentially and perhaps intentionally weights options 6,7, and 8 (arguably the most common behaviours) given the bell­curve effect of two similar dice. A potentially better option is to just total the d8 and d4 results from the initial three dice thrown and use that on the same table ­ this avoids needing an extra "different" d6 and flattens the frequencies somewhat. The average is still 7 (and now 6 and 8 in equal frequency), resulting in "defending" as the most common behaviour on the table above, but options 5 and 9 increase to similar frequency and the interesting options of 4 (returning with a prisoner) and 10 (running away from a nastier creature) are more likely. 

Using the Behaviours Table above does require some additional interpretation if an additional creature is required or if the behaviour is seemingly incompatible (eg an animal with treasure), but I think the "total d8 and d4" option has potential, even if only by assisting with determining the "special" encounters generated by the various default Kowolski combinations. 

Summary: KNeE Table Design Principles 

Distilling the above, I've come to 5 Kowolski Method principles for quality KNeE Tables: 

+ You should never need to roll additional dice. Any additional variation (eg. encountering an enemy on the near or far side of a stream for a location) or totals (eg. the number of creatures in the encounter, the radius of a feature such as a glade or width of a river) is derived from either whether a particular dice is even or odd, the sum of two or more dice or some other explicit means such as "if the d6 is 1-4 X, if 5-6 Y" or "if the d8 is higher than the d4 X, otherwise Y". This allows the encounter to be procedurally generated with a single combined roll rather than needing a series of dependent rolls and is a key design feature. 

+ Encounters must be relevant to the setting. Not actually an isolated principle for a KNeE table, this is just good design practice and should be a given, at least according to Bryce Lynch - the encounters should tie into not only the biome but also link into the other elements. One definition of a good random encounter is one that fits so well into the adventure that it is indistinguishable from a set encounter to the players. 

- Creature behaviours need to be predetermined. As with the second principle, this should be evident but adds somewhat to the design work, unless offset by the "d8 plus d4 total" method suggested above that remains consistent with the first principle. Some consideration of this is essential to avoid the otherwise default "attacks on sight" implicit in many older lesser quality encounter tables. 

+ Building the encounter for play should be easy. In terms of the simpler KNeE table in GoL this is true - you just describe the scene add the creature or encounter and then quickly modify it by the simple "Weird" eg a roll of 6-4-2 results in a spaced apart area with rocks strewn about that is eerily quiet apart from a bubbling spring. However, for the more complex tables in later adventures such as SotFK, the modification by the "Weird" can be a bit confusing to adjudicate on the fly and take significant time, as noted in this review. This can be partially offset by following the next principle, but adding a random behaviour element will reduce the ease of using the table in actual play. 

+ Combinations to allow for predetermined set "special" encounters. For me this allows the mixing of both worlds - randomness is available but with predetermined interesting encounters that can be chosen to save time, to fit in with a particular plot or character story, or just to suit the DM's whim at the time. Effectively these are "set" random encounters, and if designed properly utilising interesting creature actions should be indistinguishable from keyed encounters in their quality.