Over on twitter, @ericvulgaris remarked that the next step in his D&D campaign was for the party to start delivering mail in Phlan, and I responded somewhat excitedly because this is a wonderful idea, and one that does not show up in nearly enough games. Consider: delivering mail requires:

• A lot of open space, but a predictable core of destinations to return to over time.
• A group that is capable of dealing with that hazards of travel, which are environmental but also may be more direct, like bandits or hazards.
• An employer and a paycheck.

To me, that absolutely sounds like a formula for adventure. You have a unifying element (the job) that is strict enough to justify unity but which still leaves a ton of leeway within it. You have a steady source of challenges and threats, as well as an easy tool to introduce new ones (simply adding a new destination to their route) plus the variability that comes from travel.

What’s more, once you open the door to the mail, you can start thinking about the things that are mail-adjacent. The history of the mail is fascinating, but for those of us with a D&D bent, knights of the post could have a host of opportunities. You can look to historical examples like traveling doctors or horseback librarians for inspiration, and you can extend them to the fantastical. Consider how much it might matter to a town when the only spellcasting cleric they have access to is the one who comes through once a month or so with the mail? And, of course, these rough and ready souls can also be expected to handle the occasional monster.

The adventuring benefits of this model are obvious, so why don’t we see more of it?

Well, first, to give it’s due, there are games that do this. This is the default structure for Mouseguard and Dogs in the Vineyard, and it was one of the default modes of play in Legend of the Five Rings (doing a procession through your lord’s holdings). But for all that, it’s an outlier – a dungeon map is a normal thing, but a delivery route map would still be an anomaly in most written adventures.

Given that, there are three real challenges to implementing this model.

The first is progression: as characters level up, how do you deal with that? Having the stops on their delivery route JUST HAPPEN to have greater challenges each time they come back feels pretty fake, so are they just going to outgrow this?

The second is that it calls for a wider range of challenges, and some of those challenges aren’t challenges at all. That is, the things the characters may be called on will not be limited to fighting. Some of the challenges will be based on skills, so the system needs to support that, but other challenges are more simple. Consider the Cleric sanctifying a well – there’s no roll of the dice or challenge in that, but it’s very important to the fiction. It can be MADE challenging or interesting, but that takes a lot more work than creating a monster encounter¹.

The third is that we tend to design settings to be disposable. Tensions are set up during creation, resolved during play, and then we generally move on². I think we like the idea that we might come back later and see how things have changed, but there is almost no support for how to run such a thing, so the result often falls flat.

Now, I’ll admit, these are non-trivial challenges. If I were starting a D&D campaign tomorrow, I’d need to have solutions to all three, and I don’t think those solutions exist yet. But I also think they’re all solvable problems.

I might do a part 2 with some examples of solutions, but in the short term, let me offer some tips for how I would mitigate these things:

• I would schedule out delivery in game in advance. Not in huge detail, but just enough to say that the next route will be A=>B=>C=A in about 1 month. It won’t matter a huge amount at first, but it will make life easier as we progress, because:
• I would treat a single cycle through the route as an important design unit. When a route it completed, that’s when we would do any downtime-equivalent, and where I would make changes to the route.
• That is also when I would hand out rewards, including XP. Leveling up happens BETWEEN route cycles, and the biggest XP driver is mission success, not monsters killed.
• I would start with a small route and add in stops, but try to keep it to 5-6 stops top. That number is from my gut, so it might change, but it feels like there are only so many places they can keep in mind.
• I would make other mail folk into named NPCs, partly to reinforce the larger setting, partly to create a pecking order because:
• I would make changes to the routes and make those meaningful. As characters level, adding a more dangerous stop to their route makes sense, but so does dropping a boring one. And, critically, that stop now goes to someone else. Maybe they’re the rookie crew (who might need some help from the old hands sometime) or maybe they’re that asshole’s crew who totally snaked the village with the hot springs because it’s such a great place to stop.
• I would enter with very flexible ideas about the group’s duties and try to tune those based on player choices and priorities. That said, if there are obvious gaps, I will happily have NPCs in other crews fill them and use that as a complicating factor.
• The Patron NPC will be distant, but the logistics NPC will be always at hand and constantly annoyed.
• I would re-read Going Postal before I start.

I think it would be pretty doable at the table, so now I find myself dwelling on how I’d make it a product. Thoughts, comments and suggestions welcome.

1. Not that I’m suggesting that good monster encounter design doesn’t require skill, but the simple truth is – with D&D and it’s family especially – they’re just a LOT easier to create. This is especially true for a written product because fights are largely one-size-fits-all, whereas non-fights often rely on things like personalities of and relationships with NPCs, which are more emergent in play. ↩︎
2. An exception to this can be found in city games, where there size of the city allows us to mask this pattern, and there are lessons on re-use to be taken from there, but that is also a fairly neglected model. ↩︎