Collapse Your Tables
Sometimes Less is More
Nothing in this post is new. As with anything to do with randomization and tables in the OSR, it’s been discussed to death already. But this is about my own personal experience learning a lesson that seems to already be common wisdom among more experienced designers. This is about how nested tables are kinda bad and cumbersome and better off replaced by a single table. To quote one example from Playful Void:
If I have to roll multiple times on different random lists or tables to get a single result, it’s usually boring and time consuming. Almost all tables are better off being a giant array with very specific results
I didn't listen to that advice.
For the past six months I've been playtesting an adventure in Perils & Princesses. It takes place in a big forest and involves a lot of traveling around the forest with chances for encounters to occur, so I wanted to provide for a wide variety of encounters to avoid repetition. The two obvious ways to solve for this, to me at least, would be to either make a big list of encounters or make a smaller list with nested options to inject variety. In the past, I've more often opted for making big d20 lists in this case, but this time I decided to try a nested table.
The base table was a d6 during the daytime or a d8 during the nighttime to allow for night-only events involving werewolves. You also needed to add a modifier at night based on the number of active werewolves to nudge the probability toward those results. Each of the base results was a different NPC type: wolves, werewolves, werewolves+wolves, "The Beast," you get the idea. For seven of the eight results, there was then an additional d4 roll to determine specifically what they were doing. For one of the eight results, random forest creatures needing help, there was instead a d10 roll for specific creatures with needs. For two of the results, there were then additional rolls to determine which specific NPC from the faction that was being called up was involved in the encounter.
Describing it all like that makes it sound like a bad idea, right? Sure there's a huge variety of possible encounters, but that's just a lot of dice rolling.
Turns out it was a bad idea. Who knew?
Not only did it take probably half an hour total of game time each session just resolving encounter rolls, but we didn't even end up with that much variety. Somehow they got through like 7 of the 10 forest creature encounters while only seeing maybe one each of the other types.
Luckily, this can be done better.
For reasons I can discuss more later this year, I've now been converting the adventure over to Cairn 2E. Since I already needed to rework the tables for the hazard die procedure there, I took the chance to overhaul it completely. It's now a table with 28 entries on which you roll a d20. The werewolf encounters are 21-28, and you add a modifier for the number of active werewolves at night to again weight the results toward those entries.
I only ran one session with the Cairn format table before we wrapped up the playtest, but it already felt a lot simpler to work through.
Now, none of this is to say nested tables can't have their place. There's a natural appeal to them to a certain type of person who likes how things can be organized in that structure. But, for me at least, the time spent resolving them at the table just isn't worth it in actual play.