Peter: I think the whole concept of mechanical advancement actually pushes a game away from a fiction-based (as you use the word, or perhaps simulation-based is acceptable) space toward a trope-based or arc-based space.  Here's why: (warning, LONG)

Let's start by talking about what supernatural or paranormal elements do to a game setting.  Obviously the moment you introduce them, "realism" as it's most strictly defined goes out the window. But there's such a thing as "verisimilitude" and "coherence."  You can make a world that *feels* real--you can combine suspension of disbelief with abstract reasoning to imagine what would happen in a world that wasn't like ours: you have to even without supernatural elements in *any* sci-fi setting.  You have to be able to "simulate" the effects of economies and technologies that never were to run AW at all.  That's verisimilitude.  And you can make sure the world works "The same way every time"--that's coherence.  And those standard's don't go out the window because you introduce a psychic.  You can and should strive for a coherent world even with magic and monsters in it. 

I'd argue, however, that something slippery happens when you apply the concept of coherence to magic: a change in what it means to work "the same way ever time."  See, with a futuristic motorbike, you can straight-up DEFINE it's gas mileage, it's acceleration, the weight it can pull, if you really want to.  And indeed once an MC has told you it takes 8 hours to drive from A to B, it had damn well better take 8 hours the next time unless you're redlining it or the weather is bad. 

When it comes to something like "open your brain" though, there aren't any real-world benchmarks to compare it to.  You can't write out the mass and power of the Maelstrom, you can't say it's like a cross between a Honda and a Mazda, and you don't expect the results one character got from opening his brain to be reliable or replicable.  So how do you prove to your players that you're "playing fair?"  First, you give your world's Maelstrom a consistent theme, like ghosts or emotions or secrets or mental disorders.  Your players will describe their game's Maelstrom to their friends by analogy, not to real objects, but to other fictional texts.  They'll say, "our MC plays the Maelstrom like the underworld from Earthsea" or "like the Force" or "like the First Evil on Buffy" or whatever.  Second, you make sure that even though what happens to Marie when she opens her brain may never have happened before or again, it feels appropriate--that after the fact it seems like it should have been predictable. 

But that sense of propriety is, I would argue, an arc-based sensibility.  We come up with "reasonable" supernatural scenes by modeling them on our favorite trances and haunting from movies, which used them to advance themes or resolve character arcs. 

So what does this have to do with the advancement rules?  Well, here on Earth I'm pretty comfortable saying that people essentially never suddenly gain a level in badass.  Someone does not, from one day to the next, acquire Ice Cold or Good in the Clinch or even a +1 to hot.  One especially does not, bleeding and forsaken by all allies in the wasteland when the raiders are coming, set his jaw, acquire NTBFW and murder the lot.  (Notice I invoked a trope there).  Sudden, dramatic change in capabilities are a fictional trope used to move the spotlight around an ensemble cast or resolve a previously irresolvable plotline. 

The playbooks themselves are drama-based objects, created to bestow players with the powers of fictional protagonists for no in-universe reason.  A Battlebabe's abilities, or a Driver's, are every bit as supernatural as a Brainer's.  And the advancement rule is also tailored for the sake of drama: by giving when dice are rolled (and dice are only rolled when the shit hits the fan), the advancement rules actually assure that PCs usually *will* pick up new moves while they're broken, bleeding, or abandoned.  For that reason, I think "when dramatically appropriate" would be a pretty good approximation of advancement as it stands. 

If that's your philosophy, why have advancement rules at all?  Why not just have the rule be, "players add a new move when they feel it's dramatically appropriate"?

All rules ultimately come form the MC, since even though Vincent wrote them, you're endorsing them and choosing not to hack them.  By having advancement rules at all, you're implicitly stating that there is a desirable pace for advancement to go. 

To look at it another way: advancement, by its very nature, changes things.  The entire point is to alter the relationship between the characters and their environment.  That means that sufficient advancement destroy the existing situation and replaces it with a new one.  That's great if you (collectively) felt that the old circumstances were mined out, and things needed to be shaken up to keep life interesting.  On the other hand, during the First Session you and your players presumably set up a situation that you liked and found interesting to explore.  If character advancement destroys the situation while people were still having fun with it, that seems like a bad thing. 

Or, for yet another angle: most games have an agenda beyond simply simulating the lives of the characters.  There is an end toward which events trend.  The MC may not know what it is ahead of time, but the point is that the story of a given game is finite.  Vincent himself acknowledges this--a comment I've seen him make several times on this site is that some advancement options propel games toward their end.  The Ungiven Future advanced moves, especially the 12+ Manipulate and Open Brain results, change the game in fundamental ways that tend eventually to end it. 

If like Vincent, you believe that the game won't last long once Players can make Allies and see through the maelstrom, then your game probably won't see more than about 10 advances.  And that means that the XP bubbles are a countdown clock.  If you want your game to last, say, 7 sessions, you can't give out 2 advances per session.  Are we such slaves to the "play to find out what happens" rule that we can't take action to prolong the life of a game we enjoy?   

Not at all against choosing to roll the original stat as a fix for an individual game group, but I want to point out that the printed text does not say "you may use cool instead of hard" (or whatever), but "you use cool instead of hard."  If rolling the original when the fiction demands is a necessary fix, then it should go in the text next time revisions happen.  

---

I also don't completely accept Vincent's assertion that it's not the MC's job to worry about how fast characters earn XP.  If I understand him correctly, the assertion is that since failure is as interesting as success, having a "more powerful" character does not make the game more fun, and in any case different characters have such wildly different competencies that discrepancies in the number of avdances aren't really noticeable.  (Is a Hocus going to be jealous of the Gunlugger because he has more advances?  No, they're each doing their own thing.)

I'd agree that in a game where player's don't regularly go all out on each other, they all have their own schticks, and there's a little maturity, trying to keep the party "balanced" in the way you would when playing D&D or something isn't really necessary.  But there's a bigger problem with advancement: it undermines the premise of the game.  

The Character Creation session does a lot of setting the tone for the game to come, and which playbooks are chosen can be very important in that regard.  If nobody takes one of the Leadership playbooks--say you have a Driver, a Gunlugger, and a Brainer--then often the PCs will end up a little less attached.  The game can be more mobile and "pulling up stakes and leaving" is a viable respons eot some problems.  On the other hand, with a Hocus, Chopper, or *especially* a Hardholder, the story is naturally going to end up revolving around the social groups the Character have created.  Most of the game reports on this site revolve around the troubles of a PC Hardholder's holding.  

Well, almost every character has the option to pick up a holding or some followers as an advancement, and the faster the advances come, the more likely you are to see that.  Unfortunately, adding a new holding has a major effect on the entire game, not just that player.  The overall mood can change, lots of new NPCs may need to be invented, and existing threats may need to change shape to remain relevant.  The same effect can happen with certain more prosaic moves, too.  Picking up "not to be fucked with" can make formerly scary raiders much more manageable.  

The bottom line is, the faster advancement happens, the harder it is to predict the outcome of events, the faster existing threats will get churned through, the more of the MC's prep time will be wasted, and the more work will be required to keep up with the advancing story.  Keeping the reigns on that is absolutely the MC's perogative.  

This post is about a potential problem I see reading through the rulebook.  I haven't encountered it in play, but I'd like to hear from anyone who has (or better anyone who successfully avoided it), and ideally from Vincent for his take.  The difficult is the interaction between
Stat-Substitution Moves and Stat Highlighting

My understanding is that highlighting stats is intended as a way to influence behavior.  That is, if another player wants to see you take more risks and put yourself out there, he highlights Cool.  If the MC wants you to pay attention to the Maelstrom, he highlights Weird.  If someone wants to steer the session towards a focus on cultivating relationships, one highlights Sharp and Hot. 

But, let's look at an example character, Wrench, a driver who aims to be a one-man "problem-solver" and protection racket.  He starts off with Weather Eye.  For his first advance he picks up Ice Cold so he put the pressure on his victims.  For the second, he snags Easy to Trust so as to arrange better pay for his labors. 

Let's look at the way the basic moves break down for him:

Cool: Act Under Fire, Go Aggro, Manipulate/Seduce
Hard: Seize by Force
Hot: Nothing
Sharp: Read a Sitch, Read a Person, Open Brain
Weird: Nothing
(Hx): Help/Hinder

Now let's say a new session starts and it's time to highlight stats.  Two Problems arise:

First, you can't really manipulate his behavior by dangling XP any more.  Let's say you want him to go commune with the spirits of the wastes, but he's been hanging around in town politicking.  Well, highlighting Weird won't work, because he doesn't roll Weird anymore.  But highlighting Sharp won't do it either, because he can still rack up the points by sticking with his negotiations.  It's similarly impossible to encourage him to be more or less manipulative. 

The second problem is that there's a risk of the highlighting process becoming more adversarial.  If you highlight his Sharp and Cool, he'll get XP for nearly every roll he makes, which defeats the point of the mechanic and puts him ahead of the group.  But there's no longer anything "interesting" about highlighting Hot or Weird; doing so serves no purpose except to screw him over.  He only ever uses three stats now, which means there are only 3 stat combinations that even make sense, making it hard to follow the "try something different every time" suggestion. 

This isn't the only build affected; it's possible to use Weird for almost everything, for instance.  Anyway, I'm not sure what to do about it, but tentatively suggest awarding XP based on the default stat for all basic moves, not the one actually used.   

As noted Above, Brainers certainly don't have any kind of monopoly on the Maelstrom.  In fact, the class you probably want if the Maelstrom is your deal is Hocus, followed closely by Angel or Skinner. 

But there are good options for anyone to get involved.  Here's all you need to do to get people thinking about it

1: Start highlighting people's Weird
2: introduce some threat moves that forcibly open brains, so you can dump some interesting info in their heads.  Something they'll want to follow up on, which will get them interested in more explanation. 

Then, subtly steer them toward Improvement options that lead in a Weird direction.  Basically every character type has at least one move that relates to the Maelstrom.  Particularly impressive are the Savvyhead's Augury, the Skinner's Lost, and the Angel's Healing Touch (which can end up opening her and her partner's brain). 

Finally, remember that you can take 2 moves from other playbooks.  This means you can get a Brainer going as something tacked on to anyone with decent Weird, but it also means you can build up a no-shit psychic from any chassis. Say you're a Skinner.  Get Lost and Hypnotic, then pick up Followers/Augury, Oftener Right, and Healing Touch.  3 Advances from chargen and you have a creepy prophet who can call you to his temple, heal you, and dispense oracular insight.