Reconsidering Hex Crawls

It may seem odd that I say “reconsidering hex crawls”, given that I’ve never really discussed them here, but that’s what I’m doing.

As with megadungeons, hex crawls are something that never really appealed to me.  The ones I’ve seen are generally presented as list of hexes and the things in them.  However, because there are so many hexes to populate the contents tend to be very, very brief, and honestly get monotonous after a while because of it.

Remember the rats and 2000 copper pieces thing recently?  The primary fault it had was in presenting little more than a roll on a random monster table, along with random treasure. This is roughly what I have come to expect from hex craw descriptions.

Doing Hex Crawls Poorly

Carcosa is one of the first hex crawls I’ve examined in any real detail.  I wish I hadn’t, there are few hexes with anything actually interesting in them.  There are a few, there are some!  But they tend to be buried among all the other random tribes of random colors that want to kill something.  If you’re lucky it’s something particular, if you’re not it’s anything convenient.  The places I noticed that had a little more specific items tended to be tied to sorcerous rituals.

Mini-Review: Carcosa SelectShow

One thing Carcosa does that looks kind of like it can make the hex crawl better is it presents two entries for each hex, and the notes for the hex may include pointers to other related hexes (though it doesn’t say how, it just says “see hex 1219”) and the atrocities that may be committed there… or rather, the spells that can be cast there.  The difference is kind of hard to see.

Apart from the freakish nature of the content, this is pretty consistent with how I have seen hex crawls in the past.  Hex ID, list of stuff in it, up to the DM to spin something from it that makes it worthwhile.  Not a lot better than simply rolling on a random table or two.

Doing Hex Crawls Better

After dredging through Carcosa again, I thought I’d look at another hex crawl resource I have handy.  I’ve been slowly (two issues per month) collecting the NOD periodical by John (Matt) Stater.  Yes, the same guy who wrote Blood & Treasure, my favorite OSR rule set.  Each issue so far comes with a hex map and a bunch of information describing what can be found there.

There are two major differences here.

First, the map is sparsely populated.  John provides only the interesting bits, not every hex has text for the sake of having text.  This makes it much easier to focus on the stuff that is worth having.

Second, where he does provide something, it tends to actually be interesting to me.  The amount of text varies from a few sentences to a page or so (for major things), and there’s actually something to work with that I couldn’t have done myself with a set of dice and some tables.  Well, unless I had some really good tables.  I may want to flesh the material out some more, but I can at least have something to start with.

The difference here is, I suppose, something similar to having a stack of logs, and having a pile of lumber and a set of plans.  I might want to modify the plans before I build with the lumber, if I want to go to the trouble of making things fit better where I’m building, but I could just work with what I start with and get something usable in a short time.  Given a pile of logs I have to decide whether I want to use those logs or throw them out, and if I use them I suppose I’d better draw up my own plans while they are being milled into lumber.

I’m not absolutely certain the analogy holds, but I think it will.  John’s hex crawl information is actually useful, Carcosa‘s… not so much.  Even before the atrocities and murderous denizens.

Improving Hex Crawls Some More

I’ve looked at a couple existing examples of hex crawls.  One I find poorly done, one I find done well enough to be useful.  However, I think the second can still be improved.

I’m drifting a bit into theory here, since I haven’t actually tried this yet.  However, it’s based on things I have done, so I think I’m on reasonably solid ground.

Back to first principles.  Hex crawls seem to be primarily associated with sandbox campaigns.  Player characters can go and do what they want, or try to.  They may suffer consequences, but they generally aren’t required to do any particular thing or follow a specific path. Player agency is a major consideration.

At the same time, you want something that will engage the players.  In my experience even the coolest content becomes less interesting if there are no connections between them.  Even if every hex had something amazing, after a while it starts to feel like you are doing little more than wandering a museum of stuff.  On the other hand, evidence that things are related and interact, even if not everything is yet known about them, tends to draw player interest.  Knowing that what they do in one place can affect other places can do a lot to encourage them to want to be involved.

The dual goals of player agency and player engagement can be met by presenting a consistent world of related entities that can be learned of and affected by the players.

I believe this is critical to good campaign design… and it ties directly into techniques described in my Campaign and Scenario Design series.

How I Would do It

Here is how I would apply the techniques from the Campaign and Scenario Design series here.

  1. About the time I generate the hex map (maybe before, maybe after, maybe I’d do them together) I would work up a set of major entities, identifying their scope and their relationships.  I don’t need a lot of detail about them yet, just enough that I can understand what they are and how to recognize when they are present.
  2. I would review the list of cool stuff I have available that could be applicable to these things.  This might include modules and similar materials (such as Quinn Conklin’s Toys for the Sandbox), it might include various locations and encounter ideas.  Anything that looks like it could be related to the entities defined, or just look cool and are unrelated.  Not everything has to be related to major campaign elements, after all.
  3. I would associate elements from the second step with those from the first.  Take generic elements and make them specific.  A shrine to the god of healing becomes a Shelter of Albry, while the major temple becomes the Grand Hospice in Enellarn.
  4. I would take generic elements that are not associated with specific places and describe the sorts of places they could be found.  Where a hex entry might say “raiders attacking a village”, you might replace “raiders” with “Oltarean Cossacks” and identify the circumstances under which they might attack a village.  Contrariwise, you might identify several versions — Oltarean Cossacks, a gnollish slaving party, and so on.  I don’t care for encounter descriptions — in hex crawls or dungeons — that present things in media res because I find the situations are rarely so static.  It seems unlikely a particular village will always be under attack by raiders every time the PCs come over the hill.  If I have to change things anyway so they will continue to make sense, I may as well make it easy for myself by building it into the process in the first place.
  5. For each entity, describe ways it can be found. I do not think it enough that I know a particular entity is present in a specific hex, I want to know how to get there.  Even a six mile hex can be quite large to search, but knowing that the Aranea are located in the marsh south of a small lake in the hex means I can trivially decide whether or not the PCs are anywhere close.  Contrariwise, it might be worth known how to avoid the entity.  I’ve seen discussions of when PCs should encounter whatever is in a hex (when they arrive, halfway across, a random distance across, random chance of encountering at all… all have been suggested).
  6. Unfilled hexes can still have random events and the like, but because I know they are random I needn’t feel compelled to develop them in great detail, and nor would I be annoyed by having a half-baked encounter provided for me that I had to flesh out.

I would certainly use a full entity definition for the major entities, and I can imagine an abridged template for use in the hex descriptions.  Many of the elements of the definition might be implied by where the entity is found, others might be by the associations.  For instance, the Oltarean Cossacks would have many of the signature elements of the Oltarean Hegemony already, combine that with signature elements of cavalry and you’re likely done.

Closing Comments

These steps, taken together, should make it possible to develop a sandbox campaign and apply it to a hex map in a way I find both interesting and useful.  More than a physical dungeon where physical paths are often limited, it is helpful to provide alternate paths.  These may be literal paths (following the road from town to town), or it might be an information path (the party is chasing a band of gnollish slavers, and while they are not particularly skilled trackers it is easy to see the smoke of the villages they burn as they leave).  Devising overarching entities can help provide structure to the setting, and thus to the sandbox/hex map as a whole.

The contention between the Oltarean Hegemony and the Empire of the Sun might be a major driver for the campaign if the PCs become intimately interested, or it might become little more than setting color if they don’t.  Either way, by making these definitions it is possible to provide a consistent background.  By giving them goals (the ‘threat’ section of the entity definition) it is possible to predict (decide) what they will do under various circumstances, and it also becomes possible for PCs to change what happens.  This can encourages both engagement and agency.

Now, I haven’t actually tried this yet, but it looks to me like it would go a long way toward making a hex crawl interesting to me.


  1. great post mate, as I’ve been considering hex-crawls for the first time recently. I got sent some great links when I asked about it on Reddit, and I think this could be added to that list. At some point I think I’d like to take a show at doing a city based hex crawl, once I get the basics down a bit better.

    • I’m glad you liked it.

      To be honest, I realized while writing this that my emphasis on ‘hex crawl’ might not have been appropriate. A conversation about hex crawls prompted this post, but honestly I think this can apply to just about any sandbox.

      “Hex crawl” or “node crawl” (“point crawl”) is not an important distinction, and I perhaps should not have focused on it.

      A city crawl can be handled in much the same manner. Identify the powers that be and what’s going on. Identify things you want to have present. Tie them together.

  2. Jack Colby

    A couple points, not to refute anything you said, but to add an alternate view.

    A hex in Carcosa-style is not supposed to be interesting until it comes up in play, at which point it may or may not be, depending what your players do there. That is as it should be, because it is just a tool to use as it becomes necessary. Something that reads like a result from a random table can become a memorable encounter… as DM you have no way of knowing just by reading the entry, it takes action and play to bring that out, and it’s great when it happens, as you never know what’s going on in that hex until it is interacted with, reaction dice rolled, etc. Fun stuff!

    And yeah, you could get the same result with a random table. The nice thing is, you don’t have to, it’s all there for you. If you’ve ever done the random roll thing in advance you know it can be tiring, so this is a good thing to have done and waiting for you to use. Just my take on it!

    • Very nicely done, Daniel. This is a lovely synthesis of the hexcrawl and nodecrawl/pointcrawl.

      When I started reading I figured it was basically what I do with nodecrawling anyway, the edges in the graph can represent paths or roads or the like. You expand on the idea nicely and the integration with hexcrawling is clever and usable. It allows you to associate entities with hexes and identify how they may or may not be found by someone travelling through the hex, while still allowing the ease of travel decisions offered by nodecrawling. That the entire thing becomes fractal, just as nodecrawling does, tickles me a little more.

      Nicely done indeed.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Back to Top