Update on June 26, 2017: My thinking on fronts and dangers has evolved. I was right that fronts are just an organizational tool, but I now think grim portents and danger moves should be kept separate. The important thing is to not forget your danger moves. Dungeon World fails to do an adequate job of emphasizing this (to its detriment).
Although they originated with Apocalypse World, fronts were popularized by Dungeon World. I was first exposed to them when I tried to run Dungeon World last year, but I struggled to utilize them effectively. I made one front and then ignored them for the rest of the campaign. However, I like them in theory. Part of the problem is the way in which they are presented by the Dungeon World book.
Fronts are supposed to be an organizational tool for the GM. You spend time between sessions creating new ones and adjusting existing ones so that when you disclaim responsibility (as Apocalypse World puts it), you can look at your prep and react appropriately to what the characters are doing. Did someone roll a miss while undertaking a perilous journey? You could do something boring like have them get lost or take longer, but it’s much more interesting if you can look to your grim portents and make a move that follows. For example, perhaps they still arrive on time, but the town was assaulted by a surprise ogre attack. That grim portents ought to be moves is something of a throwaway comment in the Dungeon World book when it’s really quite important. One of the many mistakes I made trying to learn to use fronts effectively was to treat them as a sequence of events that happen until you reach the impending doom, but that’s problematic. If grim portents happen outside of the characters’ influence, then you’re failing in your agenda to fill their lives with adventure. However, even if they always involve (or can involve) the characters, it’s still too easy to create grim portents that are not as effective as they could be. It’s similar to the problem of writing good next actions in GTD.
Well written next actions in a GTD system should be the next physical thing you need to move forward. An action like “pizza” might make sense when you write it down, but you’ll probably forget what it means when you look to your lists for things to do. One problem this creates is the temptation to ignore the action because you haven’t finished thinking about it. An action like “Order pizza” is better, but it might still be a little too vague unless ordering pizza is habitual. Personally, I like to provide additional details in the form of appropriate direct and indirect objects along with an appropriate action verb: Order pizzas for the party (with the specifics of the order available my reference system). What does this have to do with fronts? Quite a bit, actually.
Fronts are composed of dangers, which are analogous to projects in GTD. Impulses are your principles. Impending dooms are your outcomes. Grim portents are your next actions. Fronts themselves, however, don’t really have an analogue; and I suspect that they’re superfluous. Apocalypse World 2nd edition actually dispenses with them, replacing them with threat maps. A threat map is essentially a default set of contexts in which your dangers (called threats in Apocalypse world) exist. If the characters are traveling north, then you look at the north part of your map for move to makes from the threats there. This is exactly like GTD, where you look to your contexts for things to do. Apocalypse World does use a literal (if abstract) map, but this is really nothing more than a visualization of the relationship between the default contexts it defines.
While there are similarities, there are also differences between the two systems. One of your principles as a GM is to draw maps and leave blanks. Essentially, a project and your next actions should as specific as necessary to create the desired outcome, but the elements of your dangers should be just specific enough to provide the spark needed to keep the game moving forward without being too specific. This seems like a contradiction, but it’s not. Ultimately, the system has to serve your needs, and grim poretents have different needs than purely next actions. What follows are my suggestions for creating better dangers.
A danger’s impulse should be written as an infinitive, simple yet evocative. Like the principles in a project in GTD, the purpose of the impulse is to guide and constrain the grim portents you create for your danger. If the grim portents don’t follow from the impulse, then you should consider whether the danger itself is still relevant to your game. It’s possible that it fizzled out and was replaced by a new danger with its own impulse. This also helps you determine whether a danger is resolved: if the characters are able to undermine the danger’s impulse, then it is resolved.
Dungeon World provides a decent list of impending dooms, but don’t hesitate to create your own. Unlike the outcome of a project in GTD, you don’t want to be too specific. I recommend sticking to the style used in Dungeon World and denoting the impending doom as a noun ending in -tion. It’s worth noting that many fronts created by third parties don’t do this. I think that’s a mistake. By being too specific in your impending doom, you’re prescribing a certain outcome, which undermines your agenda to play to find out what happens.
As I suggested above, you want to be specific when writing grim portents, but you don’t want to be too specific. Use a verb and an object, but avoid further detail. Grim portents are moves; a well-written move should be an command, telling you what you need to do to portray a fantastic world and fill the characters’ lives with adventure when your players look to you to see what happens next. Moves written in the passive voice feel too far removed from the action: consider “put them in a spot” versus “a character is put in a spot”, or more specifically: “Judgement is passed” versus “Pass judgement” (or even “The Light passes judgement”). One side effect of this approach is that the grim portents in your dangers will require regular review to keep them relevant, but I that doesn’t seem like a bad thing. The time between sessions should be spent pre-thinking, so that when you’re at the table, you have the tools you need to keep the game moving, keeping downtime while you think to a minimum. In GTD, this review process known as the weekly review, and it’s the backbone of a well-functioning system. I expect a similar approach to managing dangers will make them really shine.
At this point, I’m inclined to dispense with fronts. At their best, they represent plot threads in your campaign. At their worst, they’re a distraction. How many fronts should you have? Is a front a campaign or an adventure front? Does that even matter? I don’t think it does. The dangers are the driving force in fronts, and they can function just as easily if not more effectively outside of the constraining bounds of fronts. In my upcoming Pathfinder game, I plan to create dangers and tag them with contexts as appropriate. I haven’t decided yet whether to use threat maps specifically, but my intuition is that I’ll likely start out with something a little more specific to the setting we choose. However, since I will be reviewing them regularly after each session, I can always change them if my initial contexts aren’t working for me. One of the things I’ve learned practicing GTD is to always be looking for ways to improve my system. The same goes for the structures I use when GMing.