A secret second encounter table
The magic of doubles
Folks are probably familiar with my favourite encounter table format at this point: a d20 table where results of 7+ are ‘nothing’.
1-6: Encounter entries.
7-20: Nothing
This one-roll d20 approach is both a d6 encounter table and the encounter check itself wrapped up into 1 die roll.
These days I tend to construct encounter sequences (a.k.a. encounter chains) for each of these entries, which is how most encounter tables work in Inkvein.

But I’ve been thinking on something: a way to work in very rare encounters without adding dice together (e.g. rolling 2d6 and adding them together) and getting some extra benefits too.
Doubles & probabilities
With this d20 structure there’s a 30% chance of rolling an encounter. It’s a bit boosted over a typical B/X style procedure, but I like that.
So here’s the trick. In addition to rolling the d20, we roll a d4. Using dice of different sizes means we don’t worry about order, which is nice. If the d4 matches the roll of the d20, we go to a second encounter table and choose the entry based on the double rolled. We have 4 weirder encounters that we want to be less likely to happen in this second secret table, if you roll 1 and 1, you’ve got encounter 1.
The probabilities work out nicely. There’s a 5%-ish chance of a double coming up with this structure, meaning these encounters are way less likely.
The individual probability per entry is less than 1% on the second table, this is for rare stuff. So what do we use this for?
Persistent encounters
A simple use is to put encounters that should be rarer in this secondary table: this is where your dragon goes, or your level 20 wizard.
I think what’s cooler is to use this method with encounters that hound the players or roam between regions of a large dungeon or sandbox. We’re talking:
Rival parties or people sent after the PCs.
That dragon that should have a small chance of showing up ANYWHERE on the hexmap because this whole province is her turf.
This ‘special’ table is something you use for all dungeons in your sandbox or all regions on your hexmap. 4 encounters that defy localisation.
In one dungeon (undead themed) I might have:
1: d6 spectres
2: d4 wights and d10 zombies
3: d8 clerics
4: A thief
5: A ghost
6: 3d10 zombies
Then in another dungeon (plant-themed) I could have:
1: d8 dryads
2: d6 vine blights
3: d4 druids
4: A woodcutter
5: A sentient tree
6: 2d6 sentient shrubs
But both of these dungeons can access ‘special’ encounters via those doubles.
1: The Shovelwrights, a crew of tomb robbers.
2: Juelristra the Fireweaver, a pyromancer.
3: Krinvara, a dark paladin and her host of d10 wights.
4: A purple worm, tunnelling through everything.
This way, there is that tiny chance that a purple worm erupts through the wall in the middle of the dungeon, but it’s on a special table you use for an entire big space, so it feels like good prep.
If you’re rolling encounter checks every 3 dungeon turns, statistics tells you that, on average, 300 turns will pass (50 hours!) before a purple worm turns up in a dungeon. That feels appropriately rare.
This isn’t something you want to make for small spaces, a less than 1% chance of an encounter coming up is small for say a 30 room dungeon. But if you’ve got a big space to play with, having these special encounters is a great addition.
For Inkvein?
One possible stretch goal on my mind for Inkvein is rival parties. The encounters in Inkvein already have a few groups like this, but having rival parties that jump between the different regions of the dungeon with you would be really cool. I’m thinking about using this tech if we manage to make that happen!



This feels like the best w.enc rule since 2=DRAGON/12=WIZERD
re: rival parties, I learned a lot from the way Deep Carbon Observatory handles it. In particular having people on your heels who get closer if you dally, and then finding ways to ambush you