Lanefan
Victoria Rules
I've taken the liberty of resequencing a few of the below quotes, to batch some related bits together.
Wouldn't prepping a lot of this ahead of time just be easier?
First, while "begin and end with the fiction" is excellent advice in itself, it seems to assume the players/PCs will always know about whatever fiction you're beginning and ending with. In other words, there's no hidden surprises for the PCs that they won't be warned about in some manner ahead of time as a soft move. That seems a bit unrealistic, particularly when combined with...
Second, the other missing element is time. Does the fiction you're beginning with have to be relatively recent (as in, the players/PCs remember it), or can it be something long forgotten that's only just now coming home to roost? Or, flipped around, does the end-fiction have to follow from the begin-fiction right away or can it wait for a few days-months-years? When you present a threat as a soft move and it somehow gets turned into a hard move, does that hard move need to happen right now or can it be saved up for (maybe much) later?
Both of these play into "think offscreen too". Revenge is a dish best served cold, remember.
This all comes back to my sniper-on-the-rooftop example from (was it this thread? I forget now); where a PC leaves an inn and gets shot by a sniper they had no idea was there, this being a hard move carried forward from some long-forgotten soft move or other event in the campaign's past.
OK, this seems fine; but as this is in theory a low- or no-prep system how can you be and remain consistent in your answers over the long run without doing a huge amount of work both during sessions (note-taking) and between sessions (record-keeping)? I mean, I don't always remember what I told them three weeks ago, never mind ten years; and I'm an awful on-the-fly note-taker.Some of the player moves in DW explicitly limit my behavior. I must truthfully answer any question asked with Discern Realities (which is why it has a defined, narrow list of questions!) If someone rolls Spout Lore (essentially, a "knowledge check"), then I must give an answer that is both interesting and useful when they roll 10+ full success, but I only need to make it interesting ("it's on [the PCs] to make it useful") for a 7-9 partial success. Some of the player moves explicitly empower me, e.g. to continue with Spout Lore, I am empowered to then ask how the character came to know whatever answer I just gave them, and the player must now answer truthfully as well. Obeying these moves, without just deciding not to because I feel like it, is thus openly and explicitly part of play; the players can see that my behavior as GM is bounded too.
Wouldn't prepping a lot of this ahead of time just be easier?
There's a couple of IMO important elements missing there.Some of these are empowering, e.g. Think Offscreen Too is an explicit instruction to include dangers, problems, and events which will only be revealed when the players discover them.
As for hard and soft moves, soft is the default state, and something must happen for a hard move to apply: the players fail a roll, make a decision that ignores a threat, or take a risk knowing that there may be costs, or something similar. I no longer remember your specific intended example of a move that is "too hard," but the Principles and Agendas guide here. Always, with everything you do, you must Play To Find Out What Happens. "Rocks fall, everyone dies" is not playing to find out what happens; it is simply fiat declaring what happens. This is thus forbidden by that Agenda. Likewise, most other forms of doing something fundamentally and permanently ending something (a life, a story, a goal, an item, whatever) are "too hard" to be casually dropped whenever you feel like it: doing so runs afoul of the Agenda mentioned, as well as a few of the Principles (certainly "begin and end with the fiction" and "make a move that follows," among others.)
Likewise, moves that are too soft will fail to actually drive the story forward. Things will just sit in ambiguous "something is about to happen" land. The impending threats need to be actual threats, the opportunities-with-cost equally real in both what they make possible and what cost they will exact if taken. There is certainly a curve of learning how to provide exciting and open-ended challenges. I myself only recently realized that I've been handling monster attacks poorly, making it basically "an attack is incoming that you obviously have to dodge! What do you do?" And of course the players' answer is "I dodge." I'm having to train myself to start thinking before the attack actually rolls out, making soft moves earlier in the process so there is greater tension and more opportunity for difficult decisions and open-ended outcomes. But I realized this...by going back and rereading the book! Turns out it had had the guidance I needed all along, I just forgot it.
Something I keep coming back to here is that "begin and end with the fiction" is a pretty strong guideline, especially when paired with "draw maps, leave blanks." That is, the former says (more or less) "only do things that are well-rooted in the world and understandable through said world." The latter says (more or less), "don't prepare more of the world than you need, and intentionally leave parts of it undefined so they can be discovered later." Together, they actually put some limits on the GM's ability to just enter whatever they want into fhe fiction. There will always be some things the players don't know, that's why you are reminded to "think offscreen too," but that Principle has a shadow in how it is phrased* that must be remembered too: you should usually be thinking on-screen, but sometimes supplement that by thinking offscreen.
First, while "begin and end with the fiction" is excellent advice in itself, it seems to assume the players/PCs will always know about whatever fiction you're beginning and ending with. In other words, there's no hidden surprises for the PCs that they won't be warned about in some manner ahead of time as a soft move. That seems a bit unrealistic, particularly when combined with...
Second, the other missing element is time. Does the fiction you're beginning with have to be relatively recent (as in, the players/PCs remember it), or can it be something long forgotten that's only just now coming home to roost? Or, flipped around, does the end-fiction have to follow from the begin-fiction right away or can it wait for a few days-months-years? When you present a threat as a soft move and it somehow gets turned into a hard move, does that hard move need to happen right now or can it be saved up for (maybe much) later?
Both of these play into "think offscreen too". Revenge is a dish best served cold, remember.

This all comes back to my sniper-on-the-rooftop example from (was it this thread? I forget now); where a PC leaves an inn and gets shot by a sniper they had no idea was there, this being a hard move carried forward from some long-forgotten soft move or other event in the campaign's past.