On Reader Mail, Cities and Encounter Design

Hey Courtney,

Only just started GMing (well, probably 9 months ago now) and really enjoyed your articles on Adventure and Set Design. They've really broadened my perspective on player agency and allowed me to plan–and most importantly, play–more efficiently and effectively.

Planning linear encounters seems simple, especially the way you do it. I also find planning the power and time structures fairly simple, because these are things that humans already plan with flowcharts and timetables.

What I'm finding most clumsy is planning sandboxy space structures. With dungeons and hexcrawls, this is pretty simple; use a map, code the rooms with numbers. But how do you plan towns, or huge cities? Specifically, how do you label/annotate such structures?

Is there something I'm missing? Please help!


Hi Kalle. This is surely a sticky wicket. I have been thinking very hard about this problem.

Unlike dungeons which, as you note, have received a high degree of development, cities have few representations at all, much less highly iterated or developed ones.

There have been some notable examples though. Vornheim is the most recent and serves as a excellent guide for city as creative backdrop, along with a suite of tools that allow you to handle situations common to cities. It's utility is only limited by the fact that it is plainly, explicitly, always Vornhiem. On the other end of the spectrum is nearly 700 page Ptolus campaign setting which, in true 3.5 style leaves little to the imagination.

Early examples include the encounter tables for City-state of the invincible overlord and city encounters in the Judges Guild ready reference pages. Having used these encounter tables, you're equally likely to run into a petty god or a king as opposed to your average peasant. There's also books from judges guild on villages, each containing a page map along with certain basic information about the city. One of the best of the old supplements is Cities by Midkemia press. And TSR made numerous entries into city supplements during second edition, notably Lankhmar (and little else). The adventures involving Lankhmar are of the standard 2nd edition type and can be easily ignored.

Then there's the dry tournament early Gygax style of T1 and B2, where the settlement is described in exhaustive detail, and actual relevant information is hidden within pages listing the value of bedspreads, curtains, and hidden treasure in niches.

Let's not forget the literal hundreds of supplements describing inns, city sections, specific cities, districts and more. Neigh universally, they are a disappointing way of contributing to entropy.

The Key
It's about purpose. What's a dungeon there for? Looting! Danger! Adventure! Cities are less simple.

This isn't surprising at all. Cities are literally a word for where citizens gather and build things. Along with that comes the whole of human nature: drama, politics, power struggles, oppression, opportunity, families, children, light and darkness–The whole of the human condition. Combine that with the character motivations in the game and how in the hell do you notate that?!

There are a couple of different purposes that cities serve. The fact that the purpose can change over time is what makes city notation so difficult. In my preliminary work, I've come up with a few categories of city purpose that can help dictate how you note them.

The first is the concept of a base, a place where adventure does not occur. This is not well suited for adventure campaign play. It sets up walls the players don't expect to be there, it breaks verisimilitude and really removes a lot of options from play. It is well suited for megadungeon play, considering the focus of that style. The base is represented quite excellently by a menu style, allowing the players to quickly access whatever they need to get on with the play of the game.

Another purpose cities serve is discovery. These are cities and villages stumbled upon while traveling or hex-crawling. They usually provide a safe place to rest and stop with some risk or unknowns involved. They are characterized by a single major feature or two, and generally have one or two issues or quests the players can get involved in. I've done some work on both how to create and notate these, as well as collecting some resources that can help.

A city can be an adventure site. These are the Gygax styled Homlettes, forts on borderlands, and Phandelvers. They are designed to be visited multiple times, with both resources and adventure sites contained within. They are best designed with a traditional sandbox design, with travel between the areas handwaved. In a larger, more complicated city it could be handled like a point crawl to avoid having to spend all that time mapping out non-interesting areas. The travel in a larger city is more risky, lending support to that point crawl random encounter style.

Another purpose a city could serve is as background to another adventure style. Such as a power structure with various other areas and people and their locations being what's at stake. A suite of tools such as Vornhiem's is useful in a situation like this.

And like all campaigns, each of these can change over time. So you might start with an area being one type of site and it might grow and change into another, necessitating a change or expansion in the way in which you've keyed it.

Other Factors

There are other things to keep in mind. There are no cell phones, no maps, no cars, no useful information sources on what's around the next corner. That means for anything beyond a small village or hamlet, travel within the city can be difficult. A city the size of Phandelver doesn't have this problem. You can stand in the center and in a few minutes know generally what each building or place is. But when you get much larger, travel time, dangers, and information can be unknown.

Imagine being dropped into Chicago on foot in a random place with no maps, narrower streets and no cars. There's no public police force and no easy way of contacting the guard. How many buildings are locked? Where can you rest? How threatening are your environs? You come off as an outsider, and unless you are in the appropriate section of the city will likely be treated badly by the locals.

Large cities, really large cities of the Baldur's Gate, Invincible Overlord, or Waterdeep type are not places you can just hop out your door and head to your destination. I've found that the Judge's Guild type encounter tables along with urban skill rolls to determine travel time rather useful in this regard. I wouldn't bother with mapping such large cities, except in the broadest and most general way.

Small and medium sized cities can be handled much like mini-hexcrawls with broad background maintained, but exploration and contents determined randomly as they explore local and distant neighborhoods. The shared discovery and mapping of uncharted territory can be a fun exercise, as long as players have pre-existing goals. 


There is still a lot of work to be done in this area. I suggest checking out a few of the works linked above for rough ideas about how to generate and key cities as adventure sites. (I am not affiliated with any of the products, nor receive any revenue from their purchase).

Figure out what the purpose of your city is and then figure out what information you need to minimally generate in order to make the situation fun for your players. The design of how exactly to do that is still in flux and has a great deal of room to grow.

Hack & Slash 


  1. I too, have a problem with towns or cities. It’s not so much about what purpose they serve, but rather, how to run them as an adventure. The Curse of Strahd is a perfect example of this. The town of Vallaki is keyed perfectly well, with an abundance of information on specific sights within the town, important NPCs the players might encounter, as well as some plot threads that develop depending on which establishments the party enters.

    However, where everything seems to fall down is: how exactly is a DM expected to run all the various encounters in it? Dungeons, as the original poster mentioned are far easier, because even though the layout may have branching paths, the players still move from one room to another in a linear order. They can either go from point B or point C from point A, but not point D. However, in a town or city, there are no such restrictions. Parties can move from one encounter point to another in any order.

    Now, it’s not the chaotic order that’s particularly the problem to DM… it’s how should a DM present each keyed site or event as players walk through it? Should a DM establish “zones,” wherein the party is alerted to particular sites (or events) when they get within range? Or should the DM take a more heavy-handed approach and introduce some NPC that points them in the direction (that most likely leads to an encounter)?

    The problem I face with many adventures that feature a town or city is that they don’t properly describe how a DM should run them. Each mini-adventure found within them is either all-inclusive (their encounter begins and ends at a site or specific event), or they are a jumble mess that require DMs to impose structure (an orderly approach for the players to move from encounter to encounter until all they are all complete).

    Have you done anything in the games you’ve played to tackle this beast?

    1. Yeah, that's the issue. You've hit the nail on the head.

  2. This comment has been removed by the author.

    1. This comment has been removed by the author.

    2. This comment has been removed by the author.

  3. So, I’ve been chewing on this in the back of my mind for a couple of days, and wanted to bounce off of someone how running a town or city, as an adventure type, might work. The reason I throw in “as an adventure type,” because it seems to me that there are two firmly established site-based formulas that everyone is familiar with: keyed areas (like dungeons, castles, ruins, etc. that have a relatively linear progression) and hexmaps (such as wilderness, oceans, etc. that cover large areas and are resolved in a nodal sort of way — that is, players can travel from point-to-point if a path of their choosing).

    So, how would towns / cities be treated differently? The key difference is — as far as I can tell — is that because towns and cities have encounters open up *only after* the players trigger an initial encounter area. For example: players can wander into a small farming hamlet. The only areas they can encounter during their preliminary walk through it are: the inn (location), the chapel (location), and a barn that’s caught on fire (location, but also an event). They can resolve any or all of them in the order that they wish. They walk into the inn, a drunk bandit picks a fight. That location resolves and ends. They go to the Chapel, meet the village priest, who then asks them to pick up a newly-fashioned holy symbol from the Blacksmith. He doesn’t have full payment ready, so he’s hoping the party can persuade the Blacksmith to give them the holy symbol with the coin the priest has provided. This is now a location that has opened up a new encounter that wasn’t originally available to the party. Lastly, the party makes their way to the burning barn. Well, because they visited the other two locations first, the barn has now burned to the ground. The farmers think that goblins were to blame, and ask the players to investigate a nearby warren. However, if the players had gone to the barn first, they would have discovered the evil sheriff responsible as he made his escape, so that location has two branching encounters from it… depending on the order that the party visited the locations.

    What this looks like, on paper (in addition to the keyed map), is basically a series of mini-adventure paths. Kind of like a tree. The branches of the tree are the order that the players handle the initial encounter locations. Other locations become “active” once they visit them, and can either branch out further depending on how the players resolve the site-based task, OR, in the order that they resolve the branches altogether.

    Is that complex? I’m not sure it’s any more complex than some of the published material I’ve seen out there that merely provides backdrop but no structure. What this solution does is provide an order of events. If this was to be written out, the Inn would be #1, the Chapel #2, the Blacksmith #2a, the Burning Barn #3, the Escaping Sheriff #3a, the Goblin Warrens #3b, etc. This could be expanded upon much greater for a city.

  4. I’m sure my solution isn’t anything new. I’ve just never seen it in a published RPG. World of Warcraft, Diablo, and The Elder Scrolls series have done this formula to death. Locations are basically Quest Givers… but you need to actually go to the starting location for each quest first in order to activate them.

    So here’s the clever part: Let’s say a quest chain has the players go to the barn after it’s burned down, then go to the goblin warrens, then visit their dragon wyrmling leader, etc. until that chain ends. Mission complete. By this point, the players have gained enough XP to level up. Then, and only then, a new series of quests / locations open up that’s available only to players who have completed one or more chains (which have a higher XP challenge tied to them). This way, a town or city could have tiered levels of quest chains so that players can explore more than a fixed XP level. The first set of 3-5 quests could be levels 1-3 while the second tier would be levels 3-5. This would add variety and life to a town / city.

    The important thing is: this is how a DM would *run* a town / city-based adventure. Players basically wander through until they enter a zone triggers (“You see an Inn with a very welcoming painting of an overflowing mug, but you hear angry shouts from within”). They can either go to that activated location, or keep on walking.

    Lastly, I just wanted to mention Angry GM’s “Schrödinger, Chekhov, Samus” article, which brilliantly lays out how to break a location into zones and have encounters areas repopulate (or not) depending on the players’ actions. While it’s ingenious how it helps DM resolve factions based on how players interact with each zone, it falls just short of spelling out how quest chains would work with that system. Could my idea be married with it? Sure. And I’d be curious to see what that would look like on paper.

    Anyway, I know I’ve spilled a lot here, but I’m curious to know what you think… or how it could be improved, or where it falls flat.

  5. This comment has been removed by the author.

  6. Jason Alexander has a full series on running "city crawls" on his blog at www.thealexandrian.net . It's not perfect or comprehensive, but it's got some good ideas and discussion layered in with tons of research from past publications.

  7. Jason Alexander has a full series on running "city crawls" on his blog at www.thealexandrian.net . It's not perfect or comprehensive, but it's got some good ideas and discussion layered in with tons of research from past publications.

  8. As my campaign (currently on hiatus) plays in a metropolis (there are only three of them in this part of the world) I am currently working on how to best handle this. It is impossible and also unnecessary work to describe each building in such a huge settlement (in my world a metropolis has between 25'000 and 1 million inhabitants and covers an area of up to 25 squaremiles). So I work with quarters/wards and have a general template to give it a distuingishable feature, so called vistas. The rest are re-usable tables.
    As I prefere the open landscape gameplay or a mix between the so called sandbox- and railroad-playstyle, I set-up conflicts in advance, from miner to bigger to huge - conflicts in which the players can stumble, can partake or even can solve.

    In this whole build-up I found Red Tide: Campaign Sourcebook and Sandbox Toolkit most helpful as it discusses how to set up sites. Sites are simply places of interest, locations that have something worth an adventurer's time - and this is all what players care about - adventure!

    Hope this gave you some input for your own city design. If you are interested to learn more, feel free to conact me. I love to hear from you.


Related Posts Plugin for WordPress, Blogger...