Random Encounter Tables – my old-school way

I spent ages enhancing the eyes of this giraffe to suggest heightened intelligence, only for most of the effect to be lost when it was reduced to publishable size… Oh well. At least there’s still a hint of it there.
This was originally intended to be part of my recent article, Pieces Of Everyday Randomness, but it quickly grew to dominate everything else in that article. So I’ve extracted, edited, and enhanced it into this stand-alone piece.
Some people are really opposed to the concept of Random Encounter Tables, aka Wilderness encounters, aka Wandering Monster encounters, simply because they don’t advance the plot of whatever adventure they are running, and because they can be difficult to compile.
I think this is unfair; games lose an element of verisimilitude without their inclusion, and that loss weakens those plots and their credibility. I’ll solve the “plot problem” by the end of this article, so that’s not a good enough reason not to do them, either.
Creating a Region-specific encounter table
This was something that I did frequently in my early days of AD&D, and would still do if I had the time. I don’t, but don’t let that stop you. I did my working on a draft table, longhand, and on scratch paper. Here are the steps that I follow (there are a lot of them, but they are simple and quick, at least for the most part):
- The first step is to realize that you actually need three tables, not one. Day, Night, and Dawn/Twilight.
- Take one of these as the default. I usually take Day or Night, and which one depends on which is likely to have the greatest number of “active” entries.
- List all the creatures that you want to have entries on the table as having a significant chance of being encountered in the default list, preferably in alphabetical order, and label them E01 to E-whatever. This table is called the RAW TABLE, and all work that follows is to this table or on scratch paper until further notice. I should point at this stage to a series I did from March-April 2013, Creating ecology-based random encounters, which is dedicated to the subject of choosing those entries on an ecological basis.
- Add any creatures that are to have significant entries in the other time frames but not in the default time frame. You end up with three alphabetic lists – one long, and two short.
- Categorize each of the entries into Frequent, Common, Uncommon, Rare, and Very Rare. This information is usually provided as part of the creature’s write-up, but you have to customize it for time of day, nocturnal vs diurnal, the environment and region that the table is to apply to, and so on. The latter two items go into the heading for the table and on every page of notes so that I don’t get one table confused with another.
- The assumption is that each of these categories is twice as likely to be encountered as the one before it, as a broad rule of thumb. But we need room to manipulate the results, so I start with 4% each for the very rares. So the second column gets 64%, 32%, 16%, 8%, or 4%, respectively, as a starting point. I don’t care what these add up to, especially not that they will almost certainly add up to more than 100% – not at this point.
- Each of these table entries then gets considered for their usual activities. I want to know how likely they are to be out and about, and how likely they are to be in their lair, home base, or equivalent. For the 64s and the 32s, I will also look at the most common activities and usually subdivide these encounters into different groups – “Orc Village”, “Orc Hunting Party”, “Orc Religious Activity”, “Orc Laborers”, “Orc Domestic activity”, “Orc Battle Training”, “Orc Romantic”, and “Orc – Unusual”, for example. The general rule of thumb is to try and get each entry down to 10% or less – so a 64% is likely to require 6-7 subdivisions or variations on the activity, a 32% gets divided into 3, and so on. I might even further subdivide these – “Orc hunting party 2 Orcs”, “3-6 Orcs”, “7-12 Orcs”.
- Look at each of the other entries – the creatures who aren’t likely to be encountered in the default time frame. What are they doing when they aren’t active? If they are in lairs, add an entry for those creatures with the subtype Lair to the night table. Again, think about how likely it is that these will be encountered under the circumstances; they are almost certainly going to be Uncommon, Rare, Or Very Rare, but sometimes there can be surprises.
- I add a “Stealth Modifier” to each of the encounters. This can be positive if the activity or creature is big and noticeable, or negative if it is subtle, quiet, camouflaged, disguised, or naturally stealthy in nature. I will take into account things that the PCs are likely to be on the lookout for, i.e. encounters that are likely to be hostile in nature. I’ll list the modifiers in one column and an updated total in the next.
- If I started with 45 entries on my table, I have a great deal more by this point. The next step is to tweak the encounter probabilities – do I feel that encounter E11 is too likely, relative to those around it, or encounter E32 is too unlikely? If an encounter is too likely, I add 5 to all the encounters above it on the table, if too unlikely I add 5 to it and subtract 1 from all the encounters above it, working through the table from encounter 1 all the way to encounter X. If any encounter gets to 0, I add 1 to all the other encounters on the table instead of subtracting 1 from the encounter in question.
- Same thing, but this time the adjustment is 4 instead of 5, and I work up from encounter X until I get to encounter 1.
- Same as step 10, but this time the adjustment is 3 instead of 5.
- Same as step 11, with an adjustment of 2 instead of 5.
- Same as step 10, with an adjustment of 1 instead of 5. That means that there is a potential adjustment on any given encounter of up to 15% increase or 5% decrease.
- Find the lowest value on the table. If that’s a 1, move on to the next step; if not, subtract enough from every entry on the table to bring it down to 1%.
- Add up all the percentages. The result is likely to be something absurd like 743%. Record this total. That completes work on the RAW TABLE – for now.
- On a fresh page, and leaving 1 blank line for every 100% or part thereof, list all the encounter entries for this time period in order of (adjusted) likelihood, high to low. This is called the WORKING TABLE and will be the focus of attention for some time to come.
- Starting with the lowest-likelihood items, I do a reverse tally until I get to a score of 100%. The entries so tagged go in the “most unlikely” encounter table. I then add a line to the top, “Most unlikely 100%”. As they get transferred into the final version of this table, they get crossed out on the working table.
- Divide all the results by 2, rounding up. This includes the entry just added.
- Look at the lowest remaining entry chance, which is probably something bigger than 1%. Subtract enough from it, and all the other remaining entries on the working table, to reduce it to 1%.
- This is quite likely to throw your relative values a little out of whack. So I repeat step 12…
- …then step 13…
- …and then step 14, for an additional +6% to -3% adjustment.
- Starting with the last remaining entry on the working table, go up the table, ensuring that each entry has the same % chance or more as the one you are currently looking at, i.e. that you don’t go 5,5,7,6,7,8… or anything like that. When you find an entry whose chances are less than the one below it on the table, add enough to it and every entry above it that relative order is maintained.
- Starting with the last remaining entry on the working table, add up the results until you get to 100%. These all go into the “Next most unlikely encounters table”. Again, transfer those entries into that table in the same way as was done in step 18 and cross them out, then add a “2nd most unlikely encounters” table with a value of 100%.
- Divide all the remaining results on the working table by 2, rounding up.
- Look at the lowest remaining entry chance, which is almost certainly something bigger than 1%. Subtract enough from it, and all the other remaining entries on the working table, to reduce it to 1%.
- Again, check your relative values by repeating steps 13…
- …and 14, for an additional +3% to -2% adjustment.
- Repeat step 24 for the remaining entries.
- Repeat steps 25 through 30, compiling a “third most unlikely”, then a fourth, and so on, until you reach the point where the remaining table entries total less than 100% after the division by 2. While there may be a few encounter entries remaining, most will have been spun off into subtables of successively lower encounter likelihood.
- Divide 100 by the total of the remaining entries. Multiply all the remaining entries by the result, rounded off.
- Generate a fresh total of the remaining entries. It’s likely to be a little over or under 100. Making most of your adjustments to the biggest entries (where they will have the lowest relative impact), tweak the values until you get to exactly 100%.
- Using these % values, create a Master Table for the time period.
- Go back to the RAW TABLE. Using the numbers and adjustments already there as a guideline, do the other major time period – so if you’ve done Night, now do Day. In effect, this takes you all the way back to step 5. Bear in mind that creature behavior is likely to be quite different – some of the entries from the preceding table might have 0% chance, and you may need fresh variations added to the table: “Orc hinting party 7-12, 2 guards, rest sleeping”. Generate a new Working Table, extract out a “least likely” encounter table, and so on, until you reach this step once again. Then move on to step 36.
- There are six hours of the day that are “in between” these two distinct time periods, and encounters are a bit of a muddled mix of both. I usually consider the two hours prior to sunset and the hour after, and the hour before dawn and two hours subsequent, to fall into this “Dawn/Twilight” Mix. Others use half this amount. Either way, the process is the same: Start by totaling the day and night values on the RAW TABLE to get a composite value.
- Halve each result, rounding up.
- Note that some encounters at this time will link to subsequent encounter probabilities – for example, a hunting party encountered at dawn is likely to be just setting off, and so indicates a possible village nearby. Adjust probabilities accordingly, and create new variations where necessary, extending the RAW TABLE.
- Some encounters may be more likely at dawn/twilight than at any other time of day. Adjust existing chances and add new ones as necessary.
- Perform Step 9 for any new entries on the table.
- Repeat Steps 10 to 34 to compile your Dawn/Twilight Encounter tables.
When to create a region-specific encounter table
This is quite a lengthy process when spelled out, step by step, but it doesn’t take all that long to actually do. Nevertheless, a lot of the time I simply won’t bother – or I’ll generate a more generic table instead of one that’s been customized to a particular region.
There are two considerations that I take into account when assessing whether or not a custom table is going to be useful. The first is how long the PCs are likely to be in this region – if it’s only a day or two, it’s not worth the effort; if it’s a week, it’s definitely worth the effort. The second is how frequently they are likely to return – a day or two now might not be worthwhile in and of itself, but if they have to spend that day or two every time they step outside their base of operations to go somewhere, that makes the effort worth the investment.
Sometimes, I will take a more geopolitical approach. I might do a single encounter table for a major trade route with a generic entry for “non-road encounters”. The same encounter table thus covers the entire length of the trade route, no matter where the PCs happen to be along its’ length, but it contains a pointer to a separate table for encounters deriving from the surrounding terrain, climate, and region. I’ll often use the same approach for major rivers.
Why to make your own
There are generic tables available. AD&D 2nd Ed had some quite good ones in their Monster Manual, and I still refer to these from time to time when I’m in a hurry.
But no-one else runs campaigns in my game world, with its unique creatures and variations, like “Black Trolls”. No-one else has the combination of monster sourcebooks that I do. There are regions of Fumanor where it is not impossible to encounter a 12′ tall satyr-leprechaun hybrid who wants to commit “suicide by PC”, and assumes that they won’t do so voluntarily, or a spider who’s half rattlesnake and has a necrotizing gaze. It’s not very likely mind you, but it’s possible.
No-one else GMs in my garden, so no-one else’s encounter tables will be exactly right for my game. Therefore, I need to grow my own.
Modernizing the technique
These days, I would use a table in an appropriate electronic document to do most of this work. That would let me copy and paste columns that need totaling into a spreadsheet to do the math quickly, it would let me easily color code variations, and so on. Most importantly, it would enable me to sort by whatever column I wanted to use as an index, and copy and paste to create the subtables, taking a lot of the manual labor out of the job.
Random Encounters as part of the plot
There was a time when I didn’t think I had to worry about this too much. Then a player started running a Druid and asking the wildlife what they knew about the places and circumstances that they were embroiled in. And an elf started talking to the trees, once it was established that they had a level of sentience sufficient for communications, at least at an elementary level. And suddenly, every encounter became part of a bigger picture.
These days, I retain an index listing the answers to one of five questions by random encounter number:
- What can this encounter do to advance the current plotline(s)?
- If he can’t advance it, what will this encounter do to complicate the current plotline(s)?
- If he won’t complicate it, what does this encounter know (NB may be incorrect- or mis-information) about the current plotlines?
- If he has no involvement in the current plotline, are there any plot hooks, clues, or teasers relating to future plotlines that he can throw out?
- If none of the above, what are three pieces of idle (possibly incorrect) gossip that the encounter can relate? Includes what the encounter has been up to lately.
By making these part of the encounter, it stops being plot-irrelevant and becomes a tool for plot- and player- manipulation. What’s more, if the players start heading down a path I don’t want them to follow, or simply don’t see how the clues are supposed to fit together, I can select an appropriate random encounter to use as a conduit to getting information, or misinformation, into the PCs hands.
Finally, by keeping these in a separate list, indexed by “item number”, and keeping separate tables for correlating item number with encounter numbers, I can replace or alter one without changing the other. I can keep “old information” on tap. I can have multiple vectors for the same bit of news or information.
In fact, I can use random encounters as the glue that holds a campaign together, binding adventures into a coherent, cohesive, whole. When they can do all that, why on earth would you ignore them?
Discover more from Campaign Mastery
Subscribe to get the latest posts sent to your email.
January 17th, 2015 at 5:57 pm
[…] Random Encounter Tables – my old-school way Here are a couple of illustrations I did …. Art that Inspires Gaming Thoughts Reconceptualising the Morale Check Me, Myself, and Him: Combat and Characters in one-player games Three Approaches to Challenge Levels Eight Years Of HeroPress! […]
December 4th, 2015 at 1:14 am
[…] on an ecological pattern, as described in the Creating ecology-based random encounters series, and Random Encounter Tables – my old-school way. Within the assigned parameters of what could be there, it was whatever the dice came up with, and […]
May 24th, 2016 at 12:30 pm
[…] Random Encounter Tables – my old-school way […]