I love the Discworld novels. They’re fantastic, and I enjoy the heck out of them, but like many fans I am partial to a particular subset of them. In my case, I’m a huge fan of the City Watch books. Nothing else comes close (except perhaps the Moist von Lipwig stuff, which I’m willing to acknowledge as a fair second). Now, there are many reasons I love these books, but there’s a particular element about them which I think is very relevant to setting design.

See, I should also note that I’m a big fan of cities. I have purchased city books for games I will never play because I’m always fascinated to see how people present cities and urban adventures. I am, by and large, disappointed. The bulk of city products tend to be a handful of really interesting pages about the shape of the city, then piles and piles of pages that turn the city into some sort of above-ground dungeon.

Now, thankfully, this is less common with more modern products (and as noted, the recently released Neverwinter Campaign setting is not half bad) but what’s gotten me thinking is the difference in how Ankh-Morpork (the city of the City Watch novels) is presented. Certainly, there are a handful of places (The Unseen University, the Watch House, the Opera House, the river Ankh, Guildhouses and so on) but the city is primarily defined by trends and people.

Now, people are kind of an obvious part of things, albeit one that is often undeserved in setting design. It is possible to design a setting that is almost entirely characters, but it’s trickier to do it in a way that preserves player agency. That is, there is a strong tendency for a character-based setting to become about the NPCs rather than the PCs. This is where the Forgotten Realm soften go wrong, and it’s where Amber often (but not always) went right by firmly tying the characters to the setting NPCs.

Trends are a little more interesting, because they apply to people and to places equally. They’re the answer to broad, semi-specific question – where would a high class scribe work? How do people on the street respond to a mugging? Things like that. I can answer those questiosn about Ankh-Morpork because they’re the parts that get detailed in the books much more than the specific drilldown of street addresses.

The thing is, most good city books have this information, but they tend to present it in a very different manner. They provide raw data from which a savvy reader might be able to extract trends, but only rarely do they make that leap of abstraction themselves.

Now, I’m a firm believer that one of the most useful things a published setting can do is let the GM answer when a player wonders “What’s here?” It’s a good, often relevant question, and I’ve seen a few products that have sustained the level of detail to actually be able to answer it explicitly (The Birthright campaign setting for one, the old Thieves of Tharbad city book for MERP for another) without being ENTIRELY overwhelming, but it’s a lot of work.

The alternative would seem to be to arm the GM with the understanding to know the answer, or to create an answer that is consistent with the greater whole. But how do you convey that without a series of bestselling novels? The answer, I think, demands experimentation.

(Oh, and we will be getting back to prep in 4e, but let’s just say that one is a many-headed beast)