A DM’s Experiments, Part I: Sandboxing

I’ve recently assumed once again the mantle of Dungeon Master for my Sunday morning D&D group. After two sessions with this campaign, it has become clear that this campaign may be too full of experimental elements. It is very uncertain to me that I will be able to call this a successful campaign, at least not for the first several weeks while I figure some things out.

I have so many thoughts on this that I’m going to break this up into multiple posts, the first of which is about the challenges of running this as a sandbox campaign.

Sandboxing

I’ve made a rule for myself not to introduce any imperative quests, i.e. no conflicts or events that, if the characters don’t intervene, will lead to horrible things inevitably happening, such that the players have no reasonable choice but to follow my lead. I’d much rather follow theirs.

This is new for me. I’ve never run a really sandbox-style campaign. In the past, even if the specific paths were chosen by the characters—or at least perceived to be, with the illusion of choice crafted just so—the goal was still always presented by Yours Truly, and the milestones along the way were likewise placed by me. Through some mix of making the problems too big to ignore and my players being willing to play along, this has worked reasonably well in driving the action forward.

Ideally, in this campaign, the players will decide what they want to do within the setting based on their characters’ stories and motivations and the circumstances I place before them. But while my players may have very good senses for their characters’ motivations, they have much more limited knowledge of the world those characters inhabit. In D&D this is okay, because the default mode doesn’t require much knowledge of the world beside what’s in front of the players: the DM presents to them The Door® and it’s up to them to kick it down.

In this campaign, I’ll have to make sure that not only are there many doors (or at least the opportunities for many doors to be rapidly constructed, hopefully between sessions and not at the table as my players rush down the hallway I didn’t think they would find), but that the players know those doors are (or may be) there. Each door also needs to be interesting enough for all of the PCs to want to open it, or at least interesting enough for some subset that the others will follow along for a while.

The other trouble with having no imperative is that I then need to construct a world that is less compelling. At first I feel like that’s a cop-out, but as I think about it more, it’s almost actually a tautology. To say that a story is “compelling” is literally to say that the audience—in this case, the players—feel that they have to follow it to its conclusion; they are compelled. But if I don’t want to compel them to explore this dungeon or that one, then I need to be careful not to get caught up in my own stories, and not to make my own ideas for how to build this setting out the only interesting bits.

It’s curious that my players have chosen a very socially-oriented goal for the group: to disrupt the governing structure of the kingdom they live in, so that they can root out the corruption within it. That seems like a story that will be rife with political intrigue, deception, and subtlety, which is, I believe, pretty heavily counter to the kill-monsters-and-take-their-stuff tradition of Gygaxian D&D, a tradition that at least some of my players certainly enjoy. (It doesn’t help that the particular kingdom is actually a oligarchical republic, so there’s not just a single king they can kill in the climax.) Yet many of their characters are also heavily social outcasts. It’s like they’ve placed handicaps on themselves, but perhaps that also makes it easier to make their characters hate the status quo: it’s easier to imagine the downtrodden rioting than the bourgeoisie.

Lastly, the players have themselves created some very compelling backstories and motivations for their characters, but in order to capitalize on that, I’ll first have to get over my disinclination to write into other people’s stories. Those stories are also going to make it a little harder—but not too much, in the end, I trust—to implement my second experiment: heightened danger and lethality.


While you wait with bated breath for part two, check out the Obsidian Portal site I’ve created for our campaign, if it pleases you.

5 thoughts on “A DM’s Experiments, Part I: Sandboxing

  1. The one game of note that I have run was somewhat sandbox-y. As the campaign grew, some challenges became somewhat compulsory due to their dire nature, but much of the campaign, particularly at the outset, was largely player-driven (or, at least, that was my intent…any of my players are free to chime in and disagree). I faced two main challenges. One, I’m not very good at improvising an entire session. After a particular goal was achieved, I would ask the players what they aimed to do next. Although this gave me an opportunity to prepare adequately, I suspect it broke some of the verisimilitude. It also led to my second main challenge, trying to ensure that each player had significant input in the group’s choice. If I were to do something like this again, I would spend an entire first session with everyone developing their characters together, developing relationships with one another and other NPCs, establishing group and individual (some public, some private) goals, etc.

    • We indeed had a character building session before starting the campaign. This was where the group’s goal was determined to be overthrowing the government.

      I like to improvise a lot, but I like to do it with some kind of skeleton prepared for where the action could lead. Having stat blocks ready to go is also pretty important.

  2. I’ve run a sandbox Paizo adventure path (Kingmaker) before. At the end of every game session I *always* asked my players what they were going to do next time, and I made a point at the table to get them to come to a consensus before everyone left. This would involve a dialogue along the lines of:

    Me: What are you guys doing next time?

    Player 1: Let’s go explore the mountains.

    Player 2: And kill the dragon there.

    Player 3: I wanted to visit Shieldtown, though.

    Player 4: *packs up his stuff without participating in the conversation*

    [Silence]

    Me: So are you going to go east to explore the mountains and try to kill the dragon, or go north to Shieldtown? I want to be sure we have a clear plan.

    Player 3: We can do Shieldtown later. Let’s get that dragon!

    Player 4: Yeah, that’s fine.

    In this manner I’ve made sure to elicit input from everyone (even the quiet guy), and I’ve effectively made the players tell me what to have ready for next time.

    When next time starts, though, I need to hold them to it and remind them that I had to prepare for something!

    • I usually do this. In my last campaign I ran for this group, which was arguably nonlinear but definitely not a sandbox, I ended sessions with questions both about what they planned to do and what they anticipated to happen next. I’ll be employing it more diligently now.

      • Nonlinear but not sandbox…that is a better description for what I was running, I think. Or somewhere in between, perhaps.

Leave a reply to Riley John Gibbs Cancel reply