Among the OSR blogosphere, there’s a lot of advice about how to make a sandbox. A whole lot of them involve breaking out a hexmap and populating it, creating a world for players to explore. Even advice that doesn’t focus on a hexmap will usually direct would-be sandboxing GMs to create a lot of points of interest in the wilderness for the characters to trek out to.
This way is different.
I’m running a sandbox style game where the players have access to a ship in which they can traverse the Shimmernight Sea. This opens up a lot of possibilities and pathways, but can be tough to pull off. Smarter minds then me have written a lot on the topic, and I’ve been absorbing as much as I can in order to pull this off. I really enjoy Forlorn Encystment’s post on how much detail a sandbox needs.
There was one important thing I wanted when I started designing this campaign. I want the PCs to visit settlements often, and revisit them over and over again. Each settlement should be worth exploring, spending time in, and engaging with. A good chunk of gameplay should come simply from travelling from settlement to settlement. The way I’ve set up the map forms a little ‘loop’ that the PCs can sail, which I’ll talk about in another post.
The bulk of quests will come from the settlements. There will be many points of interest outside of these main ports, but the lead-ins to those adventures will almost always come from a settlement.
This kind of set up requires something different than usual standards. Here’s what I’ve done.
First, I needed a map. Drawing one with pencil and scrap paper is totally fine, but I wanted to mess around with some procedural generation so I used a world generator. I’ll talk about how I did that in another post. But I ended up with a nice map of an archipelago-filled sea that would work very nicely.

There’s a slightly obscure RPG out there called Technoir. It’s an interesting game that leverages a quantum ogre-style of noir mystery creation. You essentially build the mystery as the players investigate it1. That’s not important though: what is important for our purposes is the way the game sets up the environments that the characters operate in. I’ve taken the elements from Technoir, changed one of them, and used them to build my own settlements.
Each environment in Technoir has a d66 grid that you use to drive the plot of the mysteries. I’ve modified one of the elements, but left the others intact. The six elements I’m using are:
- People: These are specific NPCs that typically need stuff doing and don’t have a force of their own to do it. Hence, hiring the PCs. Alternatively, just interesting enough characters on their own that they justify being a full blown element.
- Events: Things that have have happened in the past, are currently happening, or are brewing for the future. I find this is good ‘connective tissue’ for a settlement and the other elements.
- Factions: A group with sufficient resources that the jobs they might give are different from the individual ‘people’ above.
- Places: Important sites in the settlement that are unconnected to the other elements and used for tense encounters and other fun stuff.
- Threats: People, groups, monsters or other various things that are in some way threatening the settlement, either from within or without.
- Wildcards: Everything else. For Shimmernight Sea, the first wildcard in each settlement is an ancient piece of Kouroi technology.
To start the prep, I filled each settlement with one of each element. I tried to make sure each of them was evocative enough to paint a picture of what the port might be like. After that, I wrote a small pitch detailing the world truths we’d be using and unleashed the document on my players. Being naturally a bit lazy and wanting this to be a collaboratively built world but not spend a session running one of the very cool collaborative worldbuilding games out there, I let my friends go hog wild and stuff the document with more elements in each of the ports. As we spent a week asynchronously adding and sharing things, we were able to bounce ideas off each other and connect the elements across ports thematically.
So, with my players adding elements on their own, that brings me to my next point: this isn’t ‘secretive’ stuff. It can all be player facing, because as the GM your job will be to take the elements and twist them for your own sick and twisted purposes. That also means that your players can tie their characters to the world a little bit more firmly (and as we know, a shared baseline goes a long way towards understanding lore).
Now, I have a sandbox ready to be explored. Each port element can be a hook for adventure, and for me, I’m making sure that’s where the bulk of starting adventures come from. This ensures that the creations we’ve made will be at the forefront of the game and get used.
I’m also using a bit of a faction turn to churn out more adventure and make the world feel alive, but that’s a different post.
This game came before the Carved from Brindlewood games, but is a bit clunkier to set up and get running, and is much less ‘defined’ setting wise.↩︎
Did you enjoy this post? Consider signing up to the mindstorm, my semi-regular newsletter!
