As the party is about to venture forth to Mickleheim, I've started giving some thought to wilderness/sandbox adventure, which I've intended from the start to be a facet of the Fellhold campaign. I cheated a little bit up front by asking my players to do me the courtesy of agreeing that their characters had come together with the intent to explore Fellhold, at least initially, and we'd go from there. This let me devote my initial focus to having mapped and stocked dungeon locations (as it turns out, far more than I've needed so far - before play, I overestimated the speed with which the party would explore). Now they're heading to the big city to sell off some of their more specialized loot, and hopefully get some better prices on fancy things. I was prepared to let them make this trip safely, and only get into the difficulties of wilderness travel later, but they assumed that the journey would involve some play, so who am I to stop them?
So, I'm giving some thought as to how best to handle this. As mentioned in previous posts, I thoroughly like An Echo, Resounding's philosophy of building in later domain level threats and interactions into early play, both because it makes the setting seem more "alive", but also so that when the players start concerning themselves with higher goings on, it won't have come out of nowhere. However, as also mentioned above, I find Dungeon World's fronts (as modified from Apocalypse World) to be a fantastic tool, and so I want to modify them to work here. As such, I'm going to do a bit of thinking out loud about their value and how that will fit into an Old School approach.
First off, for those not familiar with fronts and threats from either AW or DW, they are a fantastic bit of gaming technology thought up by Vincent Baker that quite elegantly reconcile GM prep and vision with wide open player action. The way they manage this is to have a general characterization of the threat (say, a Cult of Demon worshippers) which may be part of a larger "Front" (like eastern front or western front) such as the Incursion of Chaos, or whatever. The GM decides what these things want, and what they will do if left to their own devices, and then figures out some specifics and when/under what circumstances they will happen. So far, nothing that different from what GMs have been doing since ever. Here's the paired techniques that make these something special: first, the GM is specifically charged with making sure these things the baddies are doing will in some way threaten or intrude upon the player characters, and secondly, that these things only happen if the players don't do anything about it. That's where the key difference with pre-planned story lines comes - the GM does not under any circumstances pre-plan a guaranteed outcome about anything that involves player action.
So, examples are more useful than vague descriptions. Under a post Ravenloft/Dragonlance idea of "story" in an adventure, I would probably decide that there's a Demon worshipping cult, and that they are gonna kidnap the mayor's daughter, and then the players are expected to go rescue her, and the climactic sacrifice will be about to happen whenever the players show up. On the other hand, with the fronts and threats system, I'd create a threat about the Cultists, describe what they're like, and what they're prone to do, and come up with a few special things they might do differently than other folks (call upon their Demon lord, make strange pronouncements, whatever) as cues for play. Then I'd decide that if left to their own devices, they'd kidnap and eventually sacrifice the Mayor's daughter. Here's the key thing, though: the timing of those things will be based on the internal logic of the Cultists, not on what seems appropriate to a preplanned idea of the adventure.
So, if the players are astute and figure out the cultists plan and wait to ambush them before they even kidnap the mayor's daughter, that's what happens. If the players don't particularly care about the mayor's daughter for a month or two, well, she got sacrificed while they weren't paying attention. At their most basic, fronts and threats are simply ways of organizing what happens outside of the players' immediate influence, and how to react when they shift their attention to them. For this Old School game, a further difference will be that I'm not particularly interested in pushing on issues of character ("man, this really threatens his core beliefs, let's see how they hold up!") and more interested in pushing on resources, strategic and tactical decisions, and exploration ("man, this really threatens that keep they're building as their new headquarters, let's see how they deal with it!").
What I'm running into in applying these ideas now is that they tend to treat maps as things that should be flexible and vague, whereas there is a lot of Old School value in the map being a key part of the game, with locations just as if not more important than whoever happens to be there right now. Tying threats to locations limits some of their flexibility in responding to player actions, but I feel like without developing theme as a goal, location may provide the answer to "when to push this thing instead of that other thing." And it is because of this desire to link place strongly to play that I put forward last time the idea of "nega-steadings" for dangerous places like lairs and ruins. I didn't have much time last week to sit down and actually put those ideas to use, but I'm hopeful that I'll the time this week, and so hopefully I'll have something more like a finished idea in a few days.