The Torchbearer example requires generating facts because the trigger is a skill use with no necessary opposition. In contrast to Apocalypse World which requires pre-established opposition, I.E. it’s conflict resolution.
Where the line got blurry is that implicit threats can constitute conflict but they do so by stacking a bunch of stuff together. My argument against that is really just raising an eyebrow.
More formally I’d try and look at what the escalation line is. I wish we had standardised terms for the following:
Intent – tactics/means. What are the means in conflict with, such that different tactics/means do actually constitute a separate arena of conflict and therefore produce a different meaning within the resolution.
Or more simply.
If I’m buying something and I use my money skill, it suggests that money is the means by which II gain my intent (buying something), and failure means my tactics/means is somehow insufficient. Which leads to switching arenas (ok I’ll steal it instead).
That’s not the case with the TB example because the resolution established nothing about the means. Or if it did. Yeah I get the stuff anyway but failure means the sergeant comes. Then really we’re dealing with what power the GM has to introduce more stuff and nothing to do with whether the means are tested.
A bit more about TB2e in this context:
Resources = money + reputation as a creditor (ie someone who pays their bills). So
testing Resources to purchase some thing can be compared (loosely) to
Acting Under Fire where the "fire" is
can I find a vendor willing to sell me the thing on the basis of my creditworthiness? (From here on, I'm ignoring the
money part, because the rules for adding cash dice to Resources, and then the rules for cash dice absorbing tax on a failure, basically mean that if you spend hard money then you get what you want even on a failure. So in the rest, I focus on the
creditworthiness part of the Resources stat.)
The way that prep works in TB2e is that the GM should already know
whether or not things of the desired sort are generally available in this town - this is managed through (i) the rules for specifying towns, which includes specifying whether or not they have a market, (ii) the column in the equipment list that indicates the default availability of each sort of item, by reference to town size, and (iii) any decision the GM makes to depart from the defaults - this can be done either in advance as part of prep (eg
this town has no market, but the householders will sell you bread and darn your socks), or can be done as part of the negotiation around framing the test (eg
I know you want to purchase an X, and I know that generally Xs are available in towns like this, but because of <insert reason here> there are no Xs available in this town unless you want to try the black market).
This establishes a general trend or disposition - that, in principle, these sorts of things can be purchased in this market by a person who is regarded as creditworthy. But it is the Resources roll that generates granular detail. If it succeeds, that establishes that there is at least one vendor of Xs, who was happy to sell an X to this PC. If it fails, the GM has multiple options and one option that is available is to declare (as a twist) that there are no vendors of Xs, or that no one will sell to the PC (because none of the potential vendors regard the PC as creditworthy), or etc.
I don't know whether you would regard this as generative resolution or not. As you know, I've talked about it through the lens of the implicit. And having spelled out the process a bit more in the preceding paragraphs, I also see it as the GM moving from a "generic" state, an
in principle possibility of finding a willing vendor, to a more concrete particular state. Because the rules are very clear, the player can plan around this resolution framework - it's not as "mad libs" as the example of the brother taken hostage. For instance, the players in my game plan out their purchases carefully, so that they do the harder ones that are more likely to fail, and thus have a greater likelihood of triggering a result of
the market is shut or
you're run out of town, last. At the table, this is straightforward gamism; and in the fiction, we can imagine the PCs - aware that they are not entirely welcome in the town, and that their reputations as people who honour promises and pay debts are shaky (again, this is part of the premise of the game's fiction) - don't make purchase requests that really strain the loyalty and willingness of the townsfolk, until they are willing to take the risk of being told
no.
So to bring this back to your post: I think that it
is the case that, in TB2e, when you test Resources, it is your tactics/means that are being tested. But - and here I think I'm agreeing with you - if the test is failed, it's
not true that it's simply the case that "failure means my tactics/means is somehow insufficient". Or to be a bit more precise, the GM is allowed to retroactively render the fictional situation more precise, and introduce details that explain
why your tactics/means - your appeal to your creditworthiness - is insufficient.
But the way that works is different from how, say,
stealing would work. If you shift to
stealing or
scavenging then (i) a different mechanical framework is applied (because of how town phase works), and (ii) the obstacles are calculated differently(they won't be wildly different - eg Ob 1 Resources won't be Ob 10 Scavenging, nor vice versa- but they'll probably be different enough in a system where even +1 Ob can have a real sting to it), and (iii) the sorts of conditions and twists that would apply are different. In
Vincent's terms, the relationship between
effectiveness and
position and also between
effectiveness and
resources and between
effectiveness and effectiveness (Resources can be taxed on a failure, but Criminal and Scavenging can't be) is a different one for each of these tactics/means.
So while it might seem I'm special pleading for Torchbearer (and, by implication, Burning Wheel) I don't think I am. I think there is a degree of
positionality in their resolution which is not as pinned down or robust as you would prefer - from your perspective I think you're right to raise your eyebrow! - but I don't think it is as untethered as (say) some approaches to Fate or to Cortex or maybe even to HeroQuest revised.
The blurry line would be retroactive establishment of means, which is fine in some cases but fatal in others.
Verly goes to the merchant to buy some food and so we test money. I fail and so we establish that she doesn’t have enough coin on her. But given her usual positioning she could probably give away a gold ring or something.
If Verly fights the goblin and loses and I (or you) establish a reason why. Then we might be in dangerous territory because eliding the precise nature of the why has a different effect on building meaning/theme than a continual justification for loss.
Can you say more about what you have in mind with the fighting example?