This is a repost of something I wrote a year and a half ago. I don’t know how much I’d change now; I’m just posting it here for future reference.
In sandbox RPG design, we’ve got basically two methodologies:
- the hexcrawl
- the pointcrawl
Neither seemed to satisfy the particular aesthetic I was seeking. So here’s another variant, which I see as a kind of mashup of the two.
Here are some principles, with elaboration.
Pathcrawls are hex-blind.
That is, as a player, you don’t think in hexes or squares or grids. You don’t have a hex map. Instead, the GM presents navigational information to you verisimiltudinally, in terms of:
- pathed routes
- and unpathed routes
For example, the GM doesn’t say, “Okay, it’s the next day. Which hex do you enter?” Instead, the GM says, “Okay, it’s the next day; which way do you go? Continue on down the forest road [or whatever], take that game trail you spotted earlier, venture out into the wilderness, what?
Navigation is presented as if it were in-character. Your choices as a player are your choices as a character. Get it? (This also presents a challenge for the party’s mapper, but I think it’s a fun one.)
Paths are polymorphic.
There are different kinds of paths:
- landmark chains
- game trails
- naturally occurring clearances
- supernatural guidance
Paths are prepped and static.
You don’t wing paths. You need to know where they are in your hex or grid and where they lead. This is true for the main paths of your map, the big roads and rivers.
Paths are interesting-thing-connectors.
You intentionally place at least one interesting thing along at least one of the paths in each hex.There is always something cool on a path, and the players will realize this. It replicates the feeling of exploration-temptation in sandbox FRPGs like Skyrim. Sure, you’ve got this quest, and you need to keeping heading south by southwest along the road in order to get your next orders for whoever, but, man, there’s this deer run leading up to a cave in the foothills, and I wonder what’s up there? We could always go check that out and get to the city afterward. . . .
To determine what to place along your paths, you’re going to need some random tables. These will vary depending on your setting. For a baseline, I’d take the dungeon stocking rules from D&D Basic and tweak them a little. Like this:
A ruin is a typical dungeon. I’d make a subtable to determine, in general, what kind of dungeon. Then drop in a module or a pre-made map or whatever.
A lair is something more natural than a dungeon proper, which indicates some past level of civilization. A lair is mostly natural. It could be the home of a dragon, the cave-complex housing the local goblin tribe, whatever. Again, subtables: what kind of lair? How big?
A settlement is some outpost of civilization, whether a lone hermit or a metropolis. I think having big cities pre-placed is a better idea than having them randomly generated, but, if you’re doing it randomly, you could use this result to plug in your big cities. Settlements aren’t necessarily friendly, but they’re not lairs either. Settlements have to be populated. If they’re not, they’re ruins. Subtables: what kind of settlement? How big? What’s the problem here? Possible quests?
A wonder is whatever weirdness you want to toss in. Tie in your supernatural bits or lore right here. Put something unexplained here. Put something maybe even you don’t know the answer to. In a TES-inspired game, here’s where you put shrines that grant you blessings and stuff. Subtables: what kind of wonder? What advantages? What history?
A landmark is some natural or manmade feature that reveals (a) the beauty or horror of the region or (b) the history of the region. There may be treasure here (1/6 chance). If so, hide it somewhere, or leave it out in the open, as you please. You can use the landmark to reveal history evocatively or in detail. That is, there can be symbols that even you don’t understand, or you can use this to give your players some actual detail in the form of, say, hieroglyphics on a ruined megalith.
The (-) and (+) signs indicate an optional feature of the landmark. If you want, a landmark(-) can present a danger, like a trapped or hazardous area: exposed, literally trapped, unstable, haunted, cursed. A landmark(+) can represent an area that offers the PCs are a resource (clear vantage, fortified position, food and water, resupply, secret shortcut). Or you can simply treat either result neutrally, as it seems best to you: it’s just a statue with words on it; okay.
Subtables: what kind of landmark? Natural? Manmade? Terrain feature? Item? Sculpture? Ruin (that doesn’t house a dungeon and thus doesn’t qualify as a ruin proper per #1 above)?
To determine how many to place along your paths, you’re going to have to take into account the size of the hex you’re dealing with. I’d say a default of 1d6 things per hex is fine. If you want it really packed like a TES game, you could get away with 1d6 things per mile. But that’s at the extreme side of the spectrum. For my game, I’m going with 1d6 things per 24-mile hex.
For each thing, you’ll roll on the above tables. Then you’ll place them, as in the image at the top of this post. For this hex, I rolled 1d6(3), and so I place three things, arbitrarily, along the paths. In this case, I place one thing on each of the roads in the hex and one thing on the westernmost branch of the river. The settlement of Branchmont was already there on the map; it doesn’t count against my 1d6 things.
Pathed locations always have a path that leads to them.
That is, not every (or even most) pathed locations need to be right on the path or even visible from it. They can be off in the wilderness, but the key is that there is a path from the path that leads to the location, whatever it is.
You can call this particular kind of path a lead. Here are some leads:
- game trail
- landmark chain
- naturally occurring clearance(s)
- supernatural guidance
- actually on the path
- visible from the path
- audible from the path
- smellable from the path
For example, your ruin you rolled up could be actually on the path, as in, in the way, or just off to the side, or you could say it’s visible from the path: the sculpted head the open mouth of which is the entrance is on a mountainside that the PCs can just make out through the tree cover and the fog.
Pathed locations may lead to further paths.
Each generated location has a 1/6 chance of leading to another path. This second-order path either:
- meets up with an extant path, creating a detour
- meets up with an extant path, creating a shortcut
- leads to another location to be generated
Unpathed areas are unprepped and dynamic.
You don’t know where things are, if there are things of interest at all, in unpathed areas. Since going off-path is generally a bad idea, you don’t want to waste your prep there.
Instead, create a random encounter table of appropriate locations. When the characters explore the unpathed wilderness, add a chance to encounter a random location to your normal chance of having a random encounter.
For instance, if the characters normally have a 1/6 chance of having an encounter per “turn” of wilderness exploration (and I like a “turn” of 4 hours), add a 1/6 chance of a location encounter from your unpathed location encounter table. This effectively creates a 2/6 (1/3) chance of something interesting occurring in the unpathed wilderness: either something from your general encounter tables or a special “what kind of place am I going to find out here?” encounter.
Unpathed areas are dangerous.
There are two primary dangers of going off-path:
- getting lost
- finding things
Check for getting lost however you want, say checking for a 1/6 chance at the beginning of an exploration turn to see if they’ll get lost that turn and in what direction they veer off course. You keep this to yourself, of course.
When I say finding things are a danger, I mean that, in general, stuff should be nastier and weirder in the unpathed wilderness. Rewards should also be greater out here.
Unpathed locations remain unpathed unless the PCs take action.
If the PCs want to find the location again, they must take some action to leave a path back to the location they’ve discovered, which you should then place on the hex, permanently. Otherwise, give them a 1/6 chance of finding it again for every wilderness exploration turn they expend trying to locate it again.
Navigation in unpathed locations is by compass direction.
When the players tell you they want to head off into the trackless wastes, ask them what direction they’re going (if it’s not obvious). Keep them navigating this way, and be sure, if they get lost, not to tell them they’re lost but to track the variance of where they think they’re going and where they’re actually headed.
Use my simple overland movement rate chart.
This thing is perfect for tracking movement in a pathcrawl scheme. Use it for great justice.
Miles per hour, walking. Find the intersection of your path and load (q.v., below).
To handle forced marches and stuff, just think: what would tire them out? If they’re pushing their limits, just give them a choice: rest now, or have penalties when the baddies come to visit.
The scheme is conceptual. Translate into the terms of your game as it suits you. For example, road and clear grassland would be easy terrain; swamps and jungles would be dire. Stuff in between is hard.
Light armor and no armor and minimal gear is easy encumbrance/load. So leather armor and no armor is easy. Chain and banded is hard. Plate is a dire load for long distance travel.
For carrying stuff, a few things well secured is easy. Some awkward stuff and heavy stuff strapped onto you is hard. Carrying something really awkward or heavy (a person, say) is dire.
Horses! For simplicity, keep them at the same rate, unless you’ve got some kind of pony express or want to run the thing until it dies. They can carry more than you can without increasing their load. Maybe give oxen and such a base of 2.0, but the load they can carry is way higher than you; so it kind of evens out, right?