The dungeon is the primordial ooze that tabletop RPG’s wriggled out of. There is something elemental and powerful about a mysteriously connected location to explore with treasures and dangers hidden within. It’s just plain fun game design. Many of the layout and design elements that make a dungeon great can also be applied to the sandbox region around it in the form of a point crawl where paths connect locations just as hallways connect rooms in a dungeon.

The way we make a fun sandbox region can be the same way we make a fun dungeon.

What Makes a Good Wilderness Dungeon?

I’ll be using the Alexandrians’ article on Jaqauying the Dungeon as a template for the type of elements found in good dungeon layout. If you haven’t come across it yet, it’s well worth a read. The core thesis of the article is that open exploration focused sandbox design has a number of recurring features that allow for varied, alternative, and interesting routes while exploring a dungeon.
It’s not the only framework for quality dungeon layouts but it’s an excellent one.

The name of the blog post and design philosophy described is informed by his analysis of the designs of Jennell Jaquays- a legendary dungeon designer. She created naturalistic, non linear, and diverse dungeon spaces in the excellent adventure module The Caverns of Thracia (1979) and later as a level designer for the Quake and Halo Wars series of video games.

We can transfer these elements of design practice over to our overland sandbox point crawl to improve the layout of the region and take the engaging elements of dungeon navigation and apply them to a sandbox wilderness region.

For an example case of these Jaquaysing principles in action on I’m using the Cinderstrom sandbox region I’ve been designing using these design ideas. If you’re one of my Cinderstrom players this is going to spoil a whole lot so close the tab and leave this one for the internet strangers to read.

It’s dungeons all the way down (or up)!

Jaquaysing Elements of the Wilderness Dungeon with Example Region

Multiple Entrances

In a traditional dungeon this is a great feature to have because it allows players to approach the dungeon from wildly different angles and completely bypass something they want to avoid or hop into incredibly dangerous situations right away and skip some of the less perilous areas.

In a wilderness dungeon, multiple entrances allow you to have multiple answers to the question of and where your players enter into the sandbox region. Because the region is set in a broader world you can provide several potential means of entry that lead to different play experiences and paths to start your parties experience with the wilderness region to open a campaign. This can also give your players choices also gives them multiple ways of tackling entry into the region from the outside world if they don’t like the dangers or rewards present at one approach in an already running campaign.

Example: The trade road that threads along the region from 704 to 501 offers two points of entry if the party is visiting the Cinderstrom Basin region from a different area of a campaign setting to the North or West.
There’s no other formal paths from outside that I’ve created yet, but the possibilities for wandering in lost and dying of thirst from the Starglass Desert, being awoken with no memories of their past in one of the dungeons as tomb raiders accidently disturb their stasis, or starting in the drinking hall of the settlement in 0402 are all very different introductions to adventuring in the region.

Loops and Branching Paths:

When I run traditional dungeons from other creators this is one of my favorite parts of a well Jaquaysed design. Instead of a linear experience of rooms placed in a chain to be encountered in a certain order, introducing loops and branches allows players to explore in the order and fashion of their choosing as well as use their knowledge of looping paths to bypass areas they don’t want to travel through or ambush some foe from an unexpected direction.

I try to follow this this in the wilderness dungeon by aiming to by limiting the number of dead end paths to just 6 locations of the 49. Most locations have at least three obvious branching path connections to provide for freedom in navigation choices and stop the point crawl from turning into a railroad. I haven’t accurately counted the number of potential loops built into the path navigation but it’s more than 12. Maintaining this complexity of a navigable environment in a wilderness dungeon should keep exploration flexible and sandboxy.

Secret and Unusual Paths

In traditional dungeons these are important because they reward exploration with increased options for navigation and access to undiscovered areas and they work the same in a wilderness dungeon.

The purple line represents the ancient podway network, a relic infrastructure of crystal capsules shooting beneath the earth at great speeds. Upon discovering one of its stops and figuring out how to use the podway it can function as a secret passage to access and connect between the deep and perilous levels of several dungeons.

There are also a few areas that serve as natural chokepoints to funnel exploration through difficult to access paths. In particular the Gray Lake at 0303 and the river canyon block off connectivity to much of the North part of the map with only the Old Bridge over the island between 0402 and 0501 providing connectivity. These constraints on navigation place value and emphasis on exploration to discover these links and open large new areas of the map.

Sub Regions

This wilderness dungeon is divided into subregions by obvious biome- areas that are distinct environments and offer different challenges and rewards. This mirrors the different content, dangers, and random encounters found in levels or sublevels of a traditional dungeon.

The Sagebrush Steppe is the largest region and serves as a hub to connect each of the other regions. It’s also the least dangerous area and holds the keep of Nabtaar in 0402 that offers the quickest access to the megadungeon of Cinderstrom. The Mastodon Plains and Howling Crags also have smaller settlements, though they are significantly more dangerous areas. Finally the Mist Haunted Hills, Rib Peaks, and Starglass Desert offer the most danger and lie along the periphery and no “areas of civilization”. The regions were inspired by the West Marches campaign idea of increasing difficulty as one gets further from settlements though it’s not adhered to religiously.

This mirrors the press your luck gameplay of player characters knowing that traditional dungeons get more dangerous (but more lucrative) as you get to deeper and deeper levels but used here in the horizontal plane instead of vertical depth. PC’s know when they’re entering a new region as the terrain changes with the random encounters and they can decide how much risk they want to take as they’re exploring a region.

Landmarks

In a traditional dungeon environment these could be fountains, statues, gaping sinkholes, imposing silver doors, or other features that let PC’s recognize the area when they return. It’s not much different for a wilderness dungeon. Each fork in the point crawl has a location of note. This makes each area where a navigational choice is made, more distinct and easier to remember again. These include curiosities like an ever spouting geyser, a standing stone carved with hawk glyphs, a derelict landship, or bubbling mud pots with giant footprints around them that don’t require too many words to describe and only a few kinds of hidden interactivities. There are also 16 full mini dungeons spread out throughout the region as locations that provide more complex landmarks for players to discover and delve into for traditional dungeoneering in the wilderness region outside the tentpole volcano dungeon of Cinderstrom.

Detailed Paths

“It’s also important to realize that there really can be too much of a good thing: There is a point at which endless loops and countless connections within the dungeon result in meaningless choice instead of meaningful choice. In jaquaying your dungeon it’s important to beware this featureless sprawl of ever-looping corridors.”

Justin Alexander “Jaquay(s)ing the Dungeon”

I think that a limited number of interesting and detailed paths to choose from leads to more interesting gameplay then a vast number of abstracted paths.

I initially wanted to make all my sandbox modules as classic hex crawls due to the liberty of players being able to explore in any direction. My issue with point crawls was a perceived linearity and lack of freedom. What about the spaces in between?

But as I’ve discussed in my last post- a hexcrawl is still a point crawl but one in which there are 6 possible and abstract paths from each point to the adjacent. Why does it matter if I go North or Northwest in the same hex terrain type other than one being a more direct line to my objective? Even if the six paths are actually described (A heap of extra GM prep) then you’re still interconnected to the point of overdose, a surfeit of freedom that turns the navigation of a complex natural environment of a region into a flattened abstraction flavored primarily by noting the dominant terrain type. In the sandbox design so far I’ve trimmed the obvious path options from an abstract and ever-present 6 to a distinct maximum of 4 with 3 apparent path options being more common.

Again I need to emphasize I really do still enjoy hex crawls.
I just also really like detailed paths and how they bring the fun of dungeon style path based navigation into the wilderness.

Examples:

Okay But What if the PC’s Really Want to Make Their Own Path?

Unlike a traditional dungeon, a party can go through the “walls”. Wandering away from paths and easy topographic routes like mountain ridges, valleys, or rivers gets a lot more difficult to make progress.

While following a path I’ve been trying out each route spanning roughly 3 miles or one league.
I prefer this scale of hex or path for exploring detailed but smaller sandbox regions without needing to resort to breaking broad 6 mile areas into sub hexes. Travel times range from 1 hour Watch on open plains to 3 Watches on a plateau of fractured glass. A Wandering Encounter check is made each Watch using the sub regions encounter table.

For “off path” travel folks have worked on a number of cool more complex rules but I’d just say that it takes at least 2x the time following a path to get to the closest location would unless they can fly or something (Try to keep limitless flying abilities out of your sandbox regions if you want to keep time and distance choices important.)

I keep keep a hex overlay on the map for ease of concrete distance measuring if needed by the GM. You could use procedural generation for subregion based locations along the way while rolling encounters. I don’t use any rules for getting lost when I run sandbox games because they’ve never seemed fun for me and the players seem to get plenty turned around on their own, especially if they’re required to map the region paths as they might exploring a dungeon instead of provided an absolutely accurate path map.

Wilderness Jaquaysing Case Study #1: Slumbering Ursine Dunes

Slumbering Ursine Dunes is one of my favorite classic point crawls and the author invented the term “point crawl” back in 2012 to describe this type of map and exploration format.

Scale: 150-300 yards between locations.
Number of Locations: 25
Multiple Entrances: Yes, two great staircases ascend the high dune ridge.
Midpoint Entry: Possible, access or egress from the river is possible
Loops and Branching Paths: Yes
Secret and Unusual Paths: A boat in a dungeon can be pooled into the Cosmic Void, The docked Barge of the Eld allows for dangerous travel to strange locales, and a pit leads to the Cold Hell. Most of these secret paths lead to locations outside of the Dunes.
Sub Regions: Mostly no, though the content within locations 23,24, 25 is linked by being adjacent to the ancient reservoir.
Landmarks: Yes
Detailed Paths: Yes. Routes are given descriptions that are succinct while giving players enough information to inform their navigation choices.
“13. Ironwood Grove. The trail broadens out and cuts through a 50-foot-long ironwood grove to a central four-way intersection . Fine delicate human bones are tied with silk rib-bons on a tree just off the intersection. A well-maintained trail with a split-log fence on its sides runs south from the intersection. Two narrow, unremarkable trails head north and southwest while the path to the east is faint and weed-choked.”

Wilderness Jaquaysing Case Study #2: Desert Moon of Karth

Desert Moon of Karth was my first published work and was inspired by modules like Hot Springs Island, A Pound of Flesh, Ultraviolet Grasslands, and Slumbering Ursine Dunes. It was my attempt at doing a sandbox space western crawl in Mothership and I think it holds up fairly well. But through the lens of Jaquaysing a wilderness, it has some weak areas.

Scale: 246 mile circumference tiny moon region, travel time measured in number of 6 hour Watches.
Number of Locations: 8
Multiple Entrances: No, unless the ancient orbital defenses are deactivated or destroyed somehow, then every site is a potential entrance landing a spaceship.
Midpoint Entry: Yes, the lone entry point at the space elevator to Larstown is located in a central location surrounded by other points of interest.
Loops and Branching Paths: Yes
Secret and Unusual Paths: No, present only inside dungeon/detailed locations.
Sub Regions: No, it’s all in a big desert- though the map provides some visual inspiration for coral strewn valleys and narrow passes.
Landmarks: Yes, each point has a notable landmark ranging from a fallen monument head to a calcified organic hive tower that will inform party of where they are on repeat visits.
Detailed Paths: No. I think this is something I’d tweak if I ever do an updated version of Karth. It’s up to the GM/Warden to describe the paths and give the players meaningful information towards what they’re interested in which would have been nice for me to include.

On Effort

In one of the Discord servers I shared another of my blog posts in, someone pointed out that my sandbox production process was a big heap of work. They’re absolutely right, I wouldn’t recommend this as the default method of starting a sandbox region. For the average sandbox game, procedural generation and improvisation can take a larger role in exploration while creating just a starting town, an initial dungeon, and the immediate surrounding area help to minimize DM burnout. For me this high prep prior, minimal prep during the campaign makes sense because I’m working on RPG material for published region as a pickup sandbox that’s useable at the table for other GM’s.

As Playful Void’s response and several others pointed out- this strategy of creating bespoke paths for a point crawl is prep heavy and doesn’t allow for using pure random generation at the table. This method is ideal if you have the energy and time to do a good deal of front ended prep instead of improvising most area contents.. Though I will say, if you who have the time for and enjoys creating a bespoke dungeon of 50 rooms, I think it’s a comparable amount of effort to make a sandbox wilderness dungeon with 49 locations of interest for a lot of reward on detailing the wilderness region.

However I don’t think this method of prep is only useful to people publishing for others. Unlike spending time prepping plots and overarching narrative arcs that rely on railroading players to follow a preset story fleshing out the connections in a sandbox provides long term returns to your time investment.

Once you have a detailed wilderness dungeon region that you’ve at least roughed in you now have a trademark locale- a place that you can pull out and run with minimal effort for players at the drop of a hat. The method should help create a fleshed out and interconnected sandbox you can pull out and run quickly for years to come.

What are your thoughts on the merits of drawbacks of creating a wilderness dungeon this way?