On Towns in RPGs, Part 5: Building a Playable City

In the first article in this series, I embarked on an ill-defined
quest to figure out what, if anything, a town map is actually for in tabletop play.

In the second,
I took a look at the common metaphor comparing towns to dungeons—unfavourably.

In the third,
I proposed an alternate metaphor: that cities are more like forests than
dungeons.

In the fourth,
I looked at how forests are used in D&D to see what we could use when
thinking about cities.

Now, we’re going to get to the nuts and
bolts of designing cities for use in D&D.

Think In Terms of
Districts, not Distance

No player is ever
going to remember, or care about, the actual distance between their current
location and the tavern they’re trying to get to. Similarly, they won’t
remember, or care about, the roads they have to cross to get there.

The absolute
most you can hope for is that they’ll remember and care about some of (but
not all of) the neighbourhoods they have to go through. In Terry Pratchet’s
Ankh-Morpork, the Shades is an extremely memorable and dangerous area. Like
Pratchett’s characters, players are going to avoid it wherever possible and yet
always find that they have to go through it. Planescape: Torment’s Hive and Fallout:
New Vegas’s Freeside have similar qualities. If you grimly tell the
players: “the quickest way to the princess is through—oh, dear—the
Shades,” they’ll have a reaction to it.

Don’t overdo it with districts; keep the
number small enough for them to be memorable. I’d recommend seven as an
absolute maximum, but as few as three is perfectly acceptable. Lantzberg,
from City of Eternal Rain, only used
three (one each for lower, middle, and upper class—end elevation). A district
can be as big as you like; feel free to simply scale them up for larger cities.

Forget Thee Not House
Hufflepuff

It’s no secret that in JK Rowling’s Harry Potter series, only two of the
four houses matter at all. If you’re not Gryffindor or Ravenclaw, you’re lucky
to get any screentime at all. However, if they were simply cut from the series,
then Hogwarts would feel terribly small, as if it were built solely for Harry
to gallivant around in, and not part of a living, breathing world. Your city
can’t just have people to tell your players who to kill and people to be
killed, it needs someone to clean up the mess after, also. From a narrative
standpoint, these people don’t matter, and will rarely be mentioned, but they
can be used to pad your world out. When dividing up your map into districts,
include a few that, as far as you’re concerned, will never see an adventure, and give it maybe one or two notable
characteristics. These are areas that are primarily residential, or involve industries
not relevant to adventure (i.e., anyone other than an alchemist, blacksmith, or
arcane university). Feel free to leave these places utterly devoid of points of
interest.

In the adventure written for Lantzberg, for
instance, there’s little to no reason to ever visit the castle at the peak of the
hill. It’s there for verisimilitude (someone
has to be in charge) and for the GM to hook later adventures to (which I’ll
elaborate on in my next point), but mainly it’s just there to make the city
seem larger. Similarly, most of the buildings in Castleview are manors of rich
and important citizens, each one of which might have any number of use for a
band of adventurers, but only a handful are actually fleshed out. After all, it
would hardly feel like a living, breathing city if every single building was
tied into a single adventure, would it?

Gaming is full of Hufflepuff Houses: the 996
Space Marine chapters that aren’t lucky enough to be Ultramarines, Blood
Angels, Dark Angels, or Space Wolves; D&D fiends that are neither lawful
nor chaotic, Morrowind’s Houses Dres
and Indoril, and any of Homeworld’s Kushan
other than Kiith S'jet. This isn’t laziness; they’re there for a reason: they
make the world feel larger.

Leave Room to Grow

Try to design a city large enough, and
versatile enough, that once the current quest is wrapped up, you can inject
some more content into it without serious retconning. This is part of where
your Hufflepuff-tier-neighbourhoods come in—maybe one of them has been under
the heel of a violent gang the whole time, but the party never found out
because they never went there. Once the players have started to clear out your
adventure ideas and points of interest, there’s still plenty of room to pump
some more in without the city bursting like an over-inflated balloon.

The map I posted earlier probably
represents the upper limit of how detailed you should make your city. A GM
could run a few more adventures out of Lantzberg, but a long-running campaign
would probably benefit from a bit more room to breathe.

A Few Key Details

What are the kinds of things a DM really needs to know about a city? D&D3.5
had little statblocks for cities and settlements that broke down the
demographics of different areas, but that’s probably more granular than is
actually necessary. Remember—every bit of detail that you include has the
potential to distract the GM from finding the fact they actually need. It isn’t
for instance, particularly important to know that 12.5% of a neighbourhood’s
population are halflings while 54% are elves, but it might be useful to know
that a neighbourhood has a notably large elf population and an often-overlooked
halfling minority.

Who are the Watchmen that
the Watchers Watch?

One infamously common thing that comes up
in D&D is the city watch. It’s shadow looms large over every action the
party, and your villains, will take, so it’s worth thinking about them a little
bit. Its best to err on the side of making them too weak rather than too
strong, as a powerful, well-organized law enforcement group can really put a
damper on the opportunities for adventure. The counter-argument is that if the
city watch isn’t strong enough to threaten the party, then the party
effectively has the run of the city; my preferred answer to this problem is to
give the local lord a powerful knight or champion who can be used as a
beat-stick against major threats to law and order (like the PCs) if need be,
but can plausibly be busy enough with other problems to leave some for the
party to handle.

When deciding who the local authorities
are, almost anything you can come up with is more interesting (and historically
plausible) than a centralized, professional police force. Here’s a few
examples:

A
militia organized by local guildsA
local gang that provides protection in exchange for money and doesn’t want
outsiders muscling in on their turfA
semi-legitimate religious militant orderA
mercenary group funded by a coalition of wealthy merchants (who just so happen
to overlook their own crimes and corruption)

Don’t get too bogged down in their stats;
just pick a low-level NPC from the back of the Monster Manual and write down who they work for. Different
neighbourhoods can share the same organization, but try to prevent a single
organization from policing the entire city.

By breaking up law enforcement by district,
you also prevent the entire city dogpiling on the party when they break a law, like
you see in video games. If the party robs a house in the Ironworker’s District,
they can lay low in the Lists, where the Ironworkers’ Patrol has no
jurisdiction, until the heat dies down.

Points of Interest!

All those numbers you see scattered over
D&D cities? Now’s the time to add them. Each one should correspond to a description
in a document somewhere. These descriptions can be as long or as short as you
wish. For example, on the short end, #1 from Lantzberg just has this to say:

However, and I won’t get into too much
detail for fear of spoilers, some of those numbers are elaborate, multi-page
dungeons.

While you should endeavour to keep the
number of districts low, there is no ceiling to how many points of interest you
should put into the city. Don’t burn yourself out. If you can come up with six,
put in six. If you can come up with fifty, put in fifty.

A point of interest can be anything from a scenic
overlook to a toll bridge to an elaborate sewer system packed with kobolds and
giant rats and treasure. They can be as fleshed out or as minimal as you are
comfortable with. There’s a sweet spot that varies from GM to GM, as if you
include too much detail you suffer from information overload as the party
approaches the point of interest (sixteen pages of description, for instance,
for a single shop is less than helpful), while too little information might
lead to you having to do too much on the fly. I like maybe one to three
sentences per point of interest, or per room in a point of interest if it is important
enough to warrant its own map (I typically only map dungeons).

Random Encounters

I’ll write a series on
handling random encounters later, but for now, breaking up encounters by
district is a convenient way to do it. More dangerous districts, for instance,
might have muggers or even monsters that attack (especially at night). If
you’re going to use random encounters in your campaign, creating a table for
each district lets you use your local colour to affect actual game mechanics.
Castleview, for instance, is very safe due to constant patrols by the Lady-Mayor’s
Watch, while the flooded Lists are full of man-eating fungi, ghouls, criminals,
and who knows what. This lets you follow the age-old advice to “show,
don’t tell.” You don’t have to say
“this area is full of crime,” you can show the players this by throwing some criminals at them.

This post has already
gone on way longer than intended. Next time, we’ll use what we’ve learned to
answer the original question and make better town maps.