The Alexandrian

As I’ve mentioned a couple times recently, I’m in the process of playtesting some rules for hexcrawling. This is something I’ll probably be talking about in more detail at some point in the near future, but one key element in testing the system was figuring out how do characters move through the wilderness.

In most traditional hexcrawling systems, this question is resolved by navigating the hex map: You move through the wilderness by indicating the next hex you want to go to.

I, on the other hand, specifically wanted to play “hex blind”. I find the abstraction of hexes useful as a GM for mapping and keying a region, but I don’t want the players to “play the abstraction”. But once “I want to go to that hex” was taken off the table, the question became (a) how do you navigate the wilderness and (b) how can we resolve that in an efficient and interesting way? What we slowly hashed out is that people navigate the wilderness by:

  • By going in a particular compass direction (“I go north!”)
  • By heading towards or following a visible landmark (“We’ll take the road” or “I’ll head toward the mountain I can see on the horizon”).
  • By using a map
  • By aiming for a familiar destination
  • By searching for a location they think is nearby
  • By searching a general area looking for anything interesting.

And so forth.

Eventually I figured out that all of this could be mechanically boiled down to two methods:

  1. Navigation by compass direction
  2. Navigation by visible landmark

Accompanied by some guidelines on “how you find something” in the wilderness (based on whether it’s a familiar, unfamiliar, or unknown location). So far these core guidelines seem to be covering our bases.

(Of course, this assumes that wilderness exploration is the desired mode of play. If exploration isn’t desired, then there are easier/better ways of structuring a journey from Point A to Point B.)


I mention all this merely as a prelude for my main point of interest this afternoon: Navigating the underworld.

Let me quickly explain what I mean by “underworld” by contrasting it to the traditional “dungeon”.

In a dungeon, the players’ navigation of the environment can be handled in an essentially literal fashion: They see a door and can go through it. They see a hall and can go down it. They see a corner and they can turn it. (This is one of the reasons why dungeons are great for new DMs: They don’t need to worry about framing, transitions, or pacing.)

But this only works in dungeons because there is a certain density of cool stuff within the complex. Poke around a corner or two and you’ll find something interesting to interact with: A monster. A trap. An inscription on the wall. Strange tracks. A magical effect. Whatever.

So by “dungeon” I’m referring to any complex where the density of cool stuff is high enough that navigating the complex “room by room” is interesting. It should be noted that this has nothing to do with the size of the complex: A megadungeon can be very, very large indeed. But it’s still a dungeon, because it rewards that “room by room” method of navigation.

By “underworld”, on the other hand, I’m referring to a complex where the density of “cool stuff” is small enough that “room by room” navigation won’t be rewarding. Consider Moria from Lord of the Rings, for example: It’s a vast complex that takes the Fellowship days to traverse, but in all that time they only find a half dozen or so interesting things. Running that on a turn-by-turn, room-by-room basis would be incredibly boring.


Okay, so let’s pretend that we’re standing at the entrance to an underworld complex. The first question the needs to be asked is, “Why are we going in there?”

  • You’re trying to find some specific location within the underworld.
  • You’re trying to get through the underworld to the other side.
  • You’re following the trail of someone (or something) which has gone in ahead of you.
  • You’re just mucking around down there and hoping to find something interesting.

First question: Anything I’ve missed on that list?

Second question: How would you go about doing any of those things? I mean this from a purely in-character perspective, for which a few thoughts occur to me:

  • You have a map.
  • You generally know that your destination is in a particular direction. (For example, in Lord of the Rings Gandalf knew they needed to head generally east and up to reach the far door.)
  • You can follow tracks or other signs.

Other thoughts?

This is all the foundation stuff on which mechanics can later be built. Such mechanics seem to be fairly trivial if you’re willing to remove player agency (“gimme some Dungeoneering checks to see if you find the Golden Crypts”), but more complicated if you want to let the players actually explore the complex by making choices about how they’re proceeding.

Another key element of all this is the handling of transitions to the “dungeon”-type complexes within the underworld. (See Thunderspire Labyrinth or the original D series of modules for a couple different approaches to this sort of underworld exploration.) This seems to be fairly easy if you’re just willing to fall back on the metagame and say, “Okay, you’ve reached the interesting bit.” And then transition back to the more familiar room-by-room sort of thing.

But this can be limiting and jarring. What sorts of things would “naturally” attract greater attention and, thus, provide a more natural transition between exploration modes? Finding recent tracks? Noticing a glint of gold? Encounter creatures?

Share on TumblrTweet about this on TwitterShare on StumbleUponShare on FacebookShare on RedditShare on Google+Digg this

13 Responses to “Thought of the Day: Navigating the Underworld”

  1. Leland J. Tankersley says:

    Just at first blush, I think I’d model this as a kind of flowchart/graph rather than a more traditional map. The nodes in the chart/graph of the underworld would correspond to “rooms” in a dungeon, and the edges (connecting arrows) to hallways, stairs, chutes, and so on. So you’d have the “Maze of twisty passages, all alike” node, which might require some kind of skill checks to negotiate successfully (and maybe some of the edges are only discovered through superior checks or luck). You might even treat them sort of like single hexes in a hexcrawl, with _something_ of potential interest in each, though not necessarily encountered the first time through.

    Transitions/edges would be natural places to make wandering/random encounter checks.

    So for example after negotiating the maze of twisty passages, the adventurers discover the ancient dwarf highway which leads north and down. (Perhaps they missed the hidden shaft which leads down to the pits of ennui.) They follow the highway, which leads to the dwarf guardpost and then continues on towards the mines. The highway itself is uninteresting (long, flat, wide, dark hallway) but you might roll for an encounter or two along it to keep things potentially interesting. Maybe they meet a monster, or come across a side passage, or whatever. There might be many side passages that are just sort of assumed to be present, but as long as the highway itself is plain the party knows where they want to go. (If they are interested in exploring all those little by-ways, though, then you start putting sub-nodes onto the highway between the maze and the guardpost.)

    The guardpost node is an encounter area which might have several exits, and so on.

    (You know, just writing this is making me think of the D1-D3 series of modules, and all of those hinted-at encounter areas on the underworld map that weren’t detailed…)

    This strikes me as a kind of structure that you could expand on as-needed, inserting new areas into the graph as players get interested in them, or as you invent things on the spur-of-the-moment.

    Not actually answering the question you posed, here, but this is what your post made me think of.

  2. Andrew says:

    These are useful reflections.

    Another thing to bear in mind is the concept of “basic lair” that was used in old modules; I just got “Horror on the Hill” and it’s got one for steam weevils, neanderthals, ogre cave, etc. but they aren’t keyed much. Only basic keys. The point is, you can pick it up and use it when the characters go into a “lair.” So in the example of the dwarf highway, maybe say “this is what a standard waystation looks like” as Warhammer (Hogshead edition) did for standard coaching inns, toll houses, and so on.

    As I was thinking through the options you laid out, the exceptions that came to mind could fit into them. For example, if someone is following a mental siren call, that could be following tracks or other signs. If they have a guide, that’s like a map. If they are engaged in spoiler and raiding activity, that’s like following signs. And so on. I think this is a compact and flexible summation.

  3. Vereishek says:

    tossing in my own advice about the Moria example, at least in the time frame, my personal rule is that I NEVER jump more than 24 hours. If you’re searching a cavern or something, and you haven’t found it inside a day, something comes up. I think it helps to prevent the dreaded ‘weeks later, you found it!’ Things happen in that time, players should get to experience it!
    Other than that, I want to hear about this hexcrawling!

  4. Confanity says:

    I don’t see the need for a discrete “transition.”

    If you’re navigating in macro (over/under-world) terms, you describe the surroundings at intervals, or when a natural change occurs. (“When the sun reaches the middle of the sky, you’re still walking through the same unchanging oak forest. What do you do?” or “Suddenly, the trees begin to thin out, and you find yourself looking out into an open field.” It’s only natural for the party to take stock and decide if they want to change their course of action when something in the environment changes, and it’s also natural for the DM to give them regular checks so that the former events don’t seem overly significant on a meta level. (I also roll regularly on an encounter table that includes “nothing” and ‘flavor text’ entries.)

    Seems to me that the difference between that macro and the micro (dungeon) scales would be nothing more than a matter of how much in-game time and space pass between one narrative input and the next, on the DM’s part. Instead of {“You come to a vast open cavern…” [players decide to cross straight through it]… “after half an hour of threading between stalagmites, the cavern narrows down until you find yourself walking in an ancient underground river’s track…” etc.}, it becomes {“You come to a fork in the tunnel…” [players choose a direction]… “after about twenty meters, you find the walls, floor and ceiling in front of you are coated with a thick growth of spongy moss.”}.

  5. Guy Incognito says:

    One mode of travel that may just be folded into one of the others is that you are trying to avoid/evade/get away from something. Be it a location or pursuit. In that case you don’t really care where you are going, just getting away.

    HTe other thought that comes to mind is that the “underworld” is pictured in my mind as having less options for movement. In the overworld you can move NSEW(barring magic) but in the underworld, you can go the direction the cave goes, so it may feel less open to the players.

    I know I am presenting problems but not solutions, sorry.

  6. quirthanon says:

    You might like to check out this thread. Someone has discussed a similar idea about overland rounds.

  7. Leland J. Tankersley says:

    I like the ideas in the Paizo thread; think I may *yoink* that for a game this weekend in fact. I have a 1 hex = 1 mile map of the local area, and the 3.5 overland movement rules map fairly nicely to this:

    Take number of squares moved by a character in a double move (e.g., 8 for a 20′ speed). That is the base movement in miles for a single action (half a day’s march).

    When moving, spend base movement hex-by-hex. When following a road or trail, spend 2 per hex in desert, spend 1 per hex in forest or plains, and spend 1.5 per hex in other terrain. Cross-country, spend 4/hex in jungle, 1.5/hex in plains or tundra, and 2/hex in other terrain. Rugged mountains should maybe cost more; those are just the base 3.5 travel numbers.

    I have a bunch of 1 hex = 1 mile maps from the Great Campaigns of the American Civil War series of wargames that I may see if I can lift some swathes of terrain from.

  8. Brad says:

    These are some great ideas that I might borrow from in the future. I read an article a while back describing a system very similar to this called West Marches. The structure of it makes a lot of sense and results in a very rich world. You might take a look at it and see if you find any new ideas:

  9. Justin Alexander says:

    Re: Paizo thread. Son of a….

    I thought I had finally put the finishing touches on my exploration rules, but the concept of overland rounds is just too useful not to nick it. (Which will mean ripping out large chunks of my system and re-designing it.)

    On the plus side, I suspect that the basic structure of overland rounds can simplify the core of wilderness exploration down to a point where I can incorporate a lot more detail into L&L than I thought I would be able to without violating my design goals.

  10. tussock says:

    Navigating the wilds is very different to exploring them too.

    In navigation, someone’s worked out what mountain to watch for, and which river to follow to get there (and where to cross that river on various occasions, after accounting for the winter floods). Trails are marked by piles of stone, regular tree carvings, or other semi-permanent features. To get through the pass you follow a particular river to a particular ridge line. You “head north” toward an E-W flowing river, because you can’t miss it, and it leads to the next city.

    When exploring, you make those notes, and leave those marks once you’re sure they’re a good guide. Meanwhile, you see a range of mountains, hit deep swamps and impassable cliffs, and may spend days searching out a safe river crossing. Every ridge or tributary that leads nowhere is another wasted day, and you have to keep climbing to see where you’ve gotten to. Eventually you stumble on a fair pass, maybe not the best, and you “head north”, into the desert that never ends, and you die, all one day west of a river in a long hidden valley. Exploring kills, get a guide.

    The underworld should be vastly worse, tunnels that get too small, go underwater, have great drops, are ever so three-dimensional, and you can’t see anything past a few metres ahead. The maps on the net of big cave systems take people decades to explore out, leaving very careful notes of each small exploration for others to follow, just getting into a tiny new section each time down. Unless the Derro, Drow, Svirfneblin, or Illithid have marked a trail, or you’ve a trusty guide, you’re just going to go ’round in circles for weeks and be lucky to make any distance at all.

    Following a trail underground is the same as above, follow the marked trail on the lookout for named features, but you don’t need exploration rules unless you want to bypass a sentry post or something, and who knows where that side passage goes anyway, and even if it is a bypass you could be lost for weeks.

    So you use magical pathfinding, because D&D has that and it’s impossible without it. Then it’s just extreme terrain and you do a couple miles a day, so work the random encounter chart (including special terrain) and count off the expendables until you hit the destination.

  11. Rob says:

    Seconding the note above about the flowcharts. I ran a campaign that partly took place in a Moria-sized dwarven ruin once, and this is how I did it. The whole place was absolutely lousy with random mushroom farms and looted clanhold, but most of them were pretty similar, and thus boring. So the only places I ever bothered to map where the interesting ones, and I just connected them with graph lines in a big flowchart. Never show a map like that to your players, though. Some of them just won’t get the concept of how to navigate by it. Even the ones that do usually find that it’s too clinical, and sucks the mystery right out of the place.

    I’ve also been working up another system for wilderness survival hexcrawling that lets you uses a sort of overland round in a sort of “Man vs. Wilderness” combat. You go in with a sort of abstract Preparedness score which factors in your supplies and the skill of your guide and whatnot. Basically, the party wanders through a patch of mostly similar unexplored wilderness and the fun begins.

    You start by writing a list of things that could go horribly wrong in the section of territory the party’s blundered into, and then assign the land a certain number of “Hazard Points”. These are something like Hit Points for hostile territory, instead of hostile monsters. You damage the land’s Hazard Points by successfully dealing with or weathering the things it throws at you. I don’t have these down to a science yet, but I’ve been operating by guessing at how many incidents I want conquering the land to entail, multiplying that number by the number of characters in the party, and rolling that many d6 for the lands HP.

    You’ll need to track the player’s position on a large map of the territory. Hexes are optional; the old tack-and string method works just fine. On your map, you draw out the boundaries of the area, and place any particularly interesting places the way you normally would. While the players are wandering around in the areas between – the broth in the soup – you go in three-hour exploration rounds. You then jump back off into the normal dungeon-time whenever they hit one of the really interesting sites you placed.

    In each exploration round the players get to wander around a bit, and the land usually mounts some sort of attack (chosen via d% from my list of local difficulties) on them. Whenever the players are called upon to make a saving throw or use an ability check to defend against these incidents; or whenever they are trying to ward off or avoid a wandering monster encounter, they get to use their Preparedness score as a bonus to the roll. Bunkering in a shelter or in camp gives you a large bonus to this – the more developed and permanent the camp, the better.

    These attacks can literally be an attack, in the form of monsters. It can also be bad weather, disease, a local environmental hazard, a difficult detour, a natural disaster, or a minor accident that hurts someone, sinks boats, kills or drives off NPC guides, gets the party lost, or degrades their supplies. This last effect slowly whittles down their Preparedness score more and more the longer they stay out of doors. Preparedness can go negative, which I figure is the point at which you stumble into town a ragged mess, missing half your gear, covered in mosquito bites and poison ivy, and hacking your lungs out from pneumonia.

    Each time the players successfully surmount a hazard for the first time, invent a useful preventative measure, or just survive a really nasty ordeal, they gain 1d6 knowledge about the terrain. This works like “damage” against the territory. Once you’ve gained knowledge equal to the land’s Hazard Points, you’ve mastered the area, and its got no more terrors for you. The rounds between the land’s attacks lengthen to a full day apiece, which makes them much safer and easier to travel through. The thing I’m trying to balance now is exactly at what point an old area ceases to be interesting and the players should be made able to easily move on.

  12. Leland J. Tankersley says:

    Just posting to say that, inspired by the “Overland Round” concept from the Paizo thread referenced above, I adapted the idea to dungeon exploration and 10-minute turns. The 1-page document can be downloaded from my page (linked from my name in this comment). There are a few other things I’ve build there, too (house rules and short adventures).

  13. Buddy Richards says:

    A bit late to the party here, but I thought I would put forth how I handle pretty much all exploration in my games.

    I primarily use 5-mile hexes for my overland/underworld mapping or structuring, and populate the area covered within those hexes (or fractions thereof) with 1 Point of Interest per mile, which are included as a generic hit on the random encounter tables I use, and then fleshed out either as they are discovered, or pulled from prep that I’ve already done.

    It’s important when using this method, to understand that you can pace enounter sites (delves), ruins, or whatever else you want in, however you want. This method can also aid in spreading out your prep in that if the party doesn’t enounter something in one hex, if they move on, you can still insert the Point in if your tables indicate they found something.

    In play, it feels very organic, and unless your players are the kind to “clear” areas, can be a nice surprise when they stumble over something in a place they’ve been before. Once you’ve reached your destination or goal, then you’ll obviously no longer have to roll on the tables, and play can proceed however you wish.

    I believe the system works well for the transitioning you mentioned because it prompts the GM in an enounter mindset, and so the setting piece or whichever thing you want there can be described in much the same way as any other encounter…even if it’s a cave opening they just stumbled upon.

    The overland turns in the Paizo article are nice, but not really what I felt you were looking for, structurally, as they seem to assume no encounters (and not necessarily combat) of interest other than visiting a town.

    I feel like, perhaps, not everything I tried to cover came across. I’d be willing to clarifiy anything.

Leave a Reply



Recent Posts

Recent Comments