Part 1 is here.
Yesterday I said, "The adventuring party with a consistent membership and unified goals is the most pernicious construct to good gaming known," and dealt with consistent membership. Today let's deal with getting rid of unified goals.
A unified goal is one which all, or almost all, of the party treats as the most important thing they should or could be doing, and which they all accept responsibility for dealing with. The stereotypical example of this is saving the world, whether that involves fighting off an alien invasion, dragging the MacGuffin of Power to the Vacation Spot of Power or stopping some schmuck from attaining godhood. I'm sure you've all played in games like this. I know I have, tons and tons of them, until it got to the point where every time Curtis would pitch a new campaign to me, the first thing I'd say would be "I'm cool with anything so long as we don't save the world".
Referees tend to rely on saving the world because it's an immediate, simple way to get everyone to buy in regardless of their personal squabbles. Everyone lives in the world, having it end would be a bad thing, so no matter how much of a prick you are, you probably want to stop whoever's trying to destroy it. The idea is that unifying all the PCs in the pursuit of a common goal will prevent conflict between characters.
Except preventing conflict between characters isn't an inherent good, and part of the point I'm trying to make here is that efforts to prevent it distort games and campaigns. Not only does it prevent the kind of games I like from occurring, it also prevents the kinds of games most other people like, especially if you want tons of drama and exciting narrative twists and so on, because the essence of drama is characters in conflict. So by setting it up so that the party is not a site of conflict, you the DM have declared "I am the font of assholes, all conflict shall be through and with me".
And that puts tons of load on you. You have to develop and remember everyone's rivals and enemies, make sure to have them pop up once in a while, figure out what they're doing when they're not around, plus, all of those rivals and enemies have to have their own teams of dudes helping them out, because as soon as one member of the party is challenged or attacked, the rest will jump on board to help out.
Let me suggest that instead of your campaigns having unified goals, especially ones that last the entire campaign, that you should instead establish a strong frame at the beginning, and then mainly rely on temporary, shifting, goals which engage only some of the PCs directly. Once you have this going, it may be interesting from time to time to reunify the PCs in a common, urgent cause once more, but only as a change of pace, and only from time to time, rather than suddenly shifting over to this mode and then running the rest of the campaign like this.
A strong frame is a set-up that establishes how and why the PCs associate without forcing them to do so or containing a goal with it automatically. For example, in Emern, the frame was "You were all part of Don Marengo's expedition to find the lost city of Xapoltecan until a hurricane scattered the entire expedition and left you out here in the jungle with only each other". The very first thing the PCs had to do as they clambered out of the muck and surveyed their ruined camp was decide whether they wanted to find Don Marengo, continue to the lost city of Xapoltecan, or just start heading home. Frankly, if a bunch of them had wanted to do one, and a bunch the other, I wouldn't have stopped them. But I didn't, because the PCs had a discussion about it, which immediately established what kinds of people their characters were. After discussing it, they decided to stick together and find Don Marengo, hopeful that he would help them. And so their long journey to Xapoltecan began.
Q. Why should goals only engage some PCs directly?
A. Engaging only some PCs directly forces those PCs to make the case to their comrades, to enlist their aid, and encourages less committed party members to participate but without eclipsing other opportunities to explore the world. It serves as a test of the relationships between PCs, and that testing generates dramatic conflict without you, the referee having to do anything. It also allows for debate about priorities. If you have two or more things going on, each of which only directly affects a subset of the PCs, they will have to pick and choose. The losers will harbour resentment, the winners will be forced to make concessions (or their arrogance may be such that they do not), all of which stirs the pot.
Q. But how will I keep the party on track?
A. In my experience, parties are excellent at keeping themselves on track, so long as they have reasonable and meaningful feedback on the effects of their actions and clear ideas about what they are capable of doing. The other night, playing the intro to the Thousand Thrones with two totally new roleplayers, the first question during play that one of them asked was "So what's the goal?" I answered this question (as a fellow PC) not by saying "to roleplay your halfling" or whatever, but by saying "Well, we're on board a ship going somewhere, and you're a stowaway, so you might want to use your sneak skills to stay hidden and get food", which seemed to help her without actually directly answering the question (I myself don't know, I'm just a wizard on a boat right now).
If you lay out opportunities, and they are real opportunities that are clearly understood as such by the PCs, then they will follow up on them. One of the most fun things about being a DM is having my PCs say things like "Clearly Rev [my nickname] wants us to follow up on this," or "The Rev must know we'd be interested in this," when in reality I don't have any strong preference or special knowledge at all. I just puff on my cigarette, raise my eyebrows and say "Gentlemen, what's your decision here? Do you want to examine the side corridors or ignore them for now?" and they do their magic.
Q. Why shouldn't my goal be my frame?
A. Frames and goals should not be unified because if they are, the integrity of the frame will collapse when the goal is accomplished. You should especially not do this if you are thinking of the frame being a short, easily accomplished goal that will "bring the PCs together", because that almost never actually works. The PCs accomplish the goal, and then there's a slump where you have to figure out how to weld them together now that there's nothing unifying them.
The frame should present problems or opportunities, not goals. The purpose of a frame is to push the PCs to start defining themselves, their relationships to one another and to the wider world. Frames don't need to be complex, they can be as simple as "You are the four mercenaries who live in this quiet farming town" to "You are the three sons of Errol Dessinger, Warden of the North".
By not linking frames to goals, you allow the frame to persist across goals, which allows you to do what I said earlier, which is confront the PCs with goals that only some of them are directly interested in, but which that subset is inadequate to resolve.