Time for some fun in the post-Forge moonscape.
Jason D’Angelo wrote a very insightful analysis of part of Apocalypse World, and (in a comment) linked to Vincent Baker at http://lumpley.com/index.php/anyway/thread/466 as being something that sets up a key idea for understanding that analysis. In that post, Baker makes three claims:
- “Moment-to-moment assent trumps pre-agreed authority, in every case.”
- “Any well-designed roleplaying game will assign (at least some) authority upfront.”
- “Some very good designers consider the assignment of authority to be the point of rpg design. I do not.”
As I find common for the Forge and its hinterlands, that sounded interesting, but I wasn’t sure what it meant. And it was clear from the first few comments that no-one else was sure either. My heart sank as a mess of confusion followed. Baker took a long time to express himself clearly; his many interlocutors took a long time to ask clear, explicit questions that squeezed said clear expression out of him. Sadly, when I go back to the Forge or its diaspora, this is my usual experience — it’s like there’s treasure there, but it’s sunk deep in a swamp.
There was a happy ending this time, though — having read the whole discussion, I think Baker’s point is fairly simple and quite useful.
What Baker is saying, put simply
Context — it’s 2009, and many in the Forge crowd are obsessed with one of their new tools, assignment of narrative authority. It’s not new — Inspectres, The Pool, Universalis are built around it — but Baker thinks many designers still see design only through that. He thinks many designers are trying to create only rules that assign narrative authority.
And he’s says don’t. Don’t force everything into that crude mould. Instead, make rules that “grant permissions, and set expectations, and make demands, and offer opportunities, and impose constraints, and draw out contributions”. In other words, design games that provoke players to say and do things that the other players will readily assent to, not because the proposing player thas formal authority, but because they want those things to happen. Because they’re exciting or dramatic or funny or horrible. Because they feel right. Because they’re what they came to the game for.
He’s saying that sure, assign authority as a fallback, but minimise the number of times that anyone needs to think about it. If you get players to assent a lot, authority will rarely be needed.
Put another way — because “moment-to-moment assent trumps [formal] authority”, you’ll have an easier ride (as designer, gm, or player) if you rely on inspiring assent rather than enforcing authority. You’ll find it easier to get what you want out of the game. As Baker puts it “Designing a roleplaying game means more than designing rules that we can all agree to play by, and that are playable. It means designing rules that capture us — rules that become a vital part of our experience of play.”
Put a third way, related to the second — “Assigning authority is just one way of many to go about soliciting assent.” and, later, “Assigning authority is a brute way to solicit assent, unsuitable to many of the precise circumstances in which a roleplaying game design needs to solicit assent.”
Now, could Baker have just come out and said that up front? Maybe. But maybe he actually needed all that time and noise to understand the point he was scrabbling at, and render it in concrete human language. I can never tell with him.
One reason it’s hard to follow the discussion here is that there’s a dearth of examples. And those they do have relate to games like The Pool, Universalis, and Poison’d, which were obscure then and are bloody obscure now. But, fortunately, between the original thread and a few things that link to it, illustrative examples can be found.
(There are countless better ones, I’m sure, as we’re discussing a property that man successful games have, but these are the ones I have easily to hand right now)
Expectations and permissions in Apocalypse World 2e
Jason D’Angelo discusses this in part 118 his ongoing Daily Apocalypse series.
Resolving conflicts between players without assigning authority
Baker again — “If two players’ interests come into conflict, you can resolve the conflict by promoting one player’s interests over the other’s – by assigning authority to one player at the other’s expense – or else you can resolve the conflict by reconciling the players’ interests so that they go forward in alignment – in which case there’s no need to assign authority.”
Baker wrote up an example of this from Poison’d with enough context to understand it without knowing that game. Key context that’s not in the first post — the Endure Duress action gives a PC currency that they can spend to fight better later. 
In Apocalypse World 2e, when you use the Seduce or Manipulate move on an NPC and roll well, they do what you want. When you use it on another PC, their player chooses whether they do — but you get a carrot or a stick (both if you roll very well). The carrot is an experience point; the stick is to remove on of the “highlighted stats” for the session, which is itself a way to get experience points. See Jason D’Angelo’s commentary on it for more.
Shaping GM behaviour without formal constraints
Have a look Jason D’Angelo’s analysis of the how AW instructs the GM to design NPCs. That post is about how the game primes the MC to expect something and then also guides them towards it so they don’t really have to do anything at all. It never says they can’t do it, but it both advises them to do it and encourages them to do it.
Some side issues
Baker could be read as implying that all “authority” is assigned by the game. I don’t think he means to. In any case, it’s not true — authority can adhere to individual players as much as to their formal roles. See the quote from Eero Tuovinen at the bottom of Valuable insights into OSR play.
1. Possibly interesting detail — Baker says at one point “It’s also significant that in both [Poison’d and Dogs in the Vineyard], the resource you get [for accepting harm] – fallout in Dogs, Xs in Poison’d – isn’t straightforward. It’s not plain bonus dice to be used any time in the future, for instance; it’s itself an uncertain investment.”