One of the things I talked about in the Game Structures series was the vertical integration of game structures. For example:
The hexcrawling structure delivers you to a hex keyed with a dungeon. Entering the dungeon transitions you to the dungeoncrawling structure, which delivers you to a room keyed with a hostile monster. Fighting the monster transitions you to the combat structure, which cycles until you’ve defeated the monster and returns you to the dungeoncrawling structure.
This can be crudely characterized as: “You explore the hexcrawl so that you can find dungeons. You explore dungeons so that you can find things to kill. You kill things so that you can get their treasure.” This is, obviously, a vast over-simplification. But it effectively drills down to a core structure that exemplifies the basic elements that make these scenario structures tick.
So, in a naturalistic sense, we can ask a wandering hero standing at the gates of a city, “What are you looking for?” But we could look at the same question through a structural lens and ask, “How can we vertically integrate the urbancrawl with other scenario structures?” (And what can that tell us about how we need to structure the urbancrawl itself?)
The structure “above” the urbancrawl is pretty easy to figure out: A hexcrawl brings us to the city, delivers us to the gate, and brings us right back to the question, “What are you looking for?”
Which turns our attention to the structure under the urbancrawl. And for that, let’s start by considering existing structures that we could stick in there.
URBANCRAWL TO DUNGEON
What if we treated an urbancrawl just like a hexcrawl? It delivers you to location-based adventures using a dungeoncrawl structure.
This sort of structure could work well for a Ptolus-style or Waterdeep-style “dungeon under the city” scenario, where the goal of your urbancrawling is to find new and potentially lucrative entrances to the city’s literal underworld.
But while this hypothetical structure could serve as an intriguing patina for a megadungeon campaign, it seems to be primarily interested in providing an alternative exit from the urban environment. What I’m interested in, on the other hand, is finding a structure for actual urban adventuring.
Of course, we wouldn’t necessarily need to do literal dungeons: We could deliver up vampire dens or mob houses or whatever. The mental stumbling block I run into, though, is how to reliably trigger these from simple geographic movement in an urban environment. (But it’s definitely something to keep in mind.)
URBANCRAWL TO COMBAT
What if we treated the urbancrawl just like a dungeon? You crawl into a neighborhood and it triggers a combat encounter just like entering a dungeon room.
The problem I see here is contextualizing this string of violence into something interesting. In a dungeon there is an immediate, physical contiguity which can be used to bind multiple encounters together: The goblins in area 4 are working with the other goblins that can be found in areas 5-10.
By contrast, an urbancrawl is distinct from a dungeoncrawl in that it is presenting selected elements of interest from a much larger pool of information. (As opposed to the dungeoncrawl, which generally presents everything inside the dungeon complex.)
Which isn’t to say, of course, that you couldn’t figure out a way to contextualize urbancrawl combat encounters. For example, if we key a vampire in Lowtown and a vampire in Empire Villa it wouldn’t take much imagination to assume they’re both based out of the same blood den. But since that connection is non-geographical, we would need to figure out a way to “escalate” from the ‘crawl-triggered vampire encounters to the blood den.
(This becomes necessary because movement in a city is, generally speaking, not limited.)
It strikes me that this urbancrawl-to-combat structure would probably work really well for a Dirty Harry-style cop campaign: The ‘crawl becomes your patrol, with the various encounters triggered by the ‘crawl serving as potential hooks into larger investigations.
URBANCRAWL TO MYSTERY
Which brings us to urbancrawls triggering mystery scenarios.
A natural image to pull up here is the cop out on patrol: They walk the streets, spot something suspicious, and the investigation of a crime is triggered. For similar reasons, this might be an 
interesting structure to explore for a superhero campaign: The classic “Master Planner” story from Amazing Spider-Man #31-33 was triggered by Spider-Man simply spotting some burglars trying to steal atomic equipment and then following a string of clues that eventually led him to an underwater base hidden in New York’s harbor.
This sort of “walking the streets and having something mysterious bump into you” is also quite popular in classic sword and sorcery literature, however. Robert E. Howard uses it in a number of Conan stories, for example, including “Shadows in Zamboula” which opens with a quivering voice declaiming that, “Peril hides in the house of Aram Baksh!” as the titular barbarian is walking down a street.
As I mentioned earlier, however, true mystery scenarios don’t play well with a true ‘crawl structure because they’re not holographic in their goals: You can’t solve the second half of the mystery unless you’ve collected the clues from the first half. So let’s lay mysteries aside for a moment and back up a moment.
URBANCRAWL TO LOCATION
“Peril hides in the house of Aram Baskh!”
Actually, I’m drawn back to that quote. Because what’s really happening here is that Conan is being told, “There’s something interesting in the house of Aram Baksh! You should totally check it out.” (The actual character is saying “don’t go in there, it’s dangerous, other people have died”, but from a structural standpoint the hook is saying the opposite.)
So let’s remove the trappings of the “dungeon” concept and instead just deliver up “locations”. That actually sounds familiar: It’s very similar to how hexcrawls work.
So what if we treated an urbancrawl just like a hexcrawl?
If you go exploring through an urbancrawl, what types of locations does it deliver to you? How does it deliver them? Why are you looking for them?
I have no idea. So let’s make a hex map of Ptolus.
What lacks context in a urbancrawl serving up violence? Walk the street of bravos, and you get young fools challenging you until you are feared more than you are hated.
There are many people, and if they notice you, they have an opinion.
A city concentrates wealth, so even if the permanent population has been dealt with, powerful transients may come in and be influenced by weaker locals.
Okay, if you want a lot of combat you probably have to look or it, or make enemies so it looks for you.
A city is easy to visit if the population are allies, and hard if they are enemies.
Imagine a city populated by monsters. It supports monster armies slowly conquering human territory. If the PCs simply wanted loot and XP, it’d be safer to go after monster outposts in mostly human or neutral territory, dungeons. The PCs have a mission of harming or destroying the city, high risk, but high strategic gain. Do they hide out, intrigue to start a civil war, and kill the last man standing? Do they figure out how to bring in an army? Do they start a slave revolt? Do they kill some leadership figures and call it good?
I think you are on to something if you treat an urbancrawl like a hexcrawl.
I also think you will run into problems if your key defaults to combat encounters. Sure, every neighborhood could be terrorized by a monster or have a crime in progress, but that seems to defeat the purpose of being in a city? If that is the sort of campaign you want your players want to play in, why did they leave the dungeon.
Isn’t the fundamental difference between a dungeon and a city that the former is populated by “monsters” and the latter is populated by “people”? Should not the default encounter be interaction?
What if you instead keyed your urbancrawl to hooks? Your players, having left a dungeon to resupply in the city, leave their inn to find supplies or to sell loot (this is a good reason to go heavy on the art objects and light on the cash). They wander around, with or without directions, to find the appropriate shop.
In each hex they pass through on the way, there are one or more distractions: an unusual shop, an interesting public space, a colorful character, a street performance, etc. The PCs can choose to be distracted or to keep going. This way you can use the crawl to reveal things about the culture of the city.
Each distraction has a potential hook to a mini-adventure which may or may not be obvious: the shopkeeper is having a problem with thieves, or something happens if they stop to watch the street performance, for example. They can take the hook, or move on, or maybe go about their business and come back to it later.
You can also have a random chance in each hex of either finding what you are looking for (if you don’t have directions), or having a random encounter, like a parade, or finding a body, or spotting a monster or a crime in progress. For that matter, if you don’t want the bother of detailing the city, you could handle every hex this way.
If your players know that you are including multiple hooks, they can actually choose what kind of adventure they want to have that session: solve the murder, chase down the pickpocket, take on a protection racket, make peace between neighbours, etc.
To follow-up on Beoric’s comment, it can also lead to a useful mechanic where the PC’s enter a hex and have options on how to find information. Some leads may be plainly obvious (a carnival in the neighborhood square, a building on fire, etc.). Others may be found by gathering information: Neighborhoods have broadsheets with interesting information, rumors in taverns differ from those in guilds, and locals may be willing to serve as neighborhood guides. Put it up to the PCs how they want to find information that may not be plainly obvious in each hex.
I think urbancrawl being modelled more like a hexcrawl, or at least as a hybrid between hexcrawl and dungeoncrawl has a lot of merit. Some particularly dangerous parts of the city where you don’t have freedom of movement could shift into a defacto dungeoncrawl just by virtue of passing through them to gain access to other areas.
I had to smile at the “cop on patrol” example, since I’ve been reading this series while pondering the thought of an urban D&D campaign inspired by – of all things – finally getting around to watching THE WIRE and realizing that so much of the magic (divination et cetera) that we worry about fitting into early constabulary work is perfect for using as an analogy for modern bugging and telecom surveillance. Rather than running a more ‘conventional’ campaign, it would be interesting to set up a corrupt & run-down city and a “precinct” of town guard for the players and set up an urbancrawl for them – even letting them select ‘cases’ coming in like a hexcrawl group would choose keyed locations to explore for treasure and then see how that affects the groups and locales around them…
Ha, Conan would have been the last character I would have thought of as an example of urbancrawling. While I believe the stories are always only one adventure per city, Conan clearly seems to be walking the streets, constantly looking and listening for a new place worth snooping around.
I agree with Beoric that combat should probably not be the default in an urban crawl. Substitute “encounter” (not the trademarked kind) for “combat” and the options increase greatly.
I’m spit balling here because it dovetails with something I’ve been working on and you triggered some ideas.
I’m working on a social combat system because the game I’m currently working on (“Barbarians of Lemuria”) is missing Reaction Rolls and Morale Rules. So I’m writing them inspired by a) The ‘Witty Repartee’ rules in “Honor+Intrigue”, b) “On the Non Player Character” by Courtney Campbell, c) This page: http://rolesrules.blogspot.com/2010/11/encounter-reactionmorale-table.html?m=1. and d) a zombie RPG called “The Dead” (which has some neat ideas about PC’s building relationships with other survivors to have the kind of drama that ‘The Walking Dead’ does).
Now all put together “Social Combat” will resemble the ‘normal’ Combat Structure.
So the “Vertical Integration Structure” got me thinking. What do you go to cities for? Healing, Safety, Information, Services and Equipment.
So what if the vertical of an urbancrawl is: Hexcrawling -> Urbancrawling -> Social Combat.
Whatever you’ve come here for you can’t just go to the address and get it. You have to ‘deal’ with people.
You also have ‘Wandering Monsters’ (city encounters).
You also have to build Relationships with people through Social Combat to get whatever specific thing you want. (And I can’t figure out how you have an urbancrawl without the players looking for something specific – which is notably different than the other crawls).
Just some ideas….
(In case it’s not obvious – the result of Social Combat will be specific people either liking you more, hating you more, becoming friends, becoming enemies, having sex with you, wanting to fight you, giving you information, etc.)
I should also mention that your rules for Diplomacy and Bribery also inspired parts of it as well!
The big challenge here seems to be the default goal at the larger scale not directly feeding into anything interesting at the smaller scale.
In a dungeon, you’re there to get loot and/or kill monsters, so each room, you’re there to search for loot and/or monsters; it’s dead obvious. In a city, whatever reason you have for exploring isn’t going to have a dead obvious expression in each district or neighborhood or block or building.
This brings me back to Paranoia again. Your default goal, at all times, is to find commie mutant traitors and shoot them. And every sector, neighborhood, and room has commie mutant traitors. Every other red or higher is a mutant and a secret society members. The infrareds may not be, but even then, they’re so ignorant that they’re breaking the laws on a regular basis. That doesn’t mean you need to, or even want to, start a firefight in every room, but it means it’s always at least an option. Which is basically like a dungeon.
As before, I’m not sure how well it would work to play it out that way, and I’m not sure how well it would translate to playing cops in a D&D city, but I think it would be easy to try, and the experience you’d get from trying it might be worthwhile.
Funnily enough, the city of Suderham in A3 “Aerie of the Slave Lords” is written as an urbancrawl: encounters/content triggered by geographic (building to building) movement. The goal is to find the dungeon entrance and trigger the famous no-winb dungeoncrawl. It would be very interesting to hear what it was actually like to play through that round of the original tournament.