I can offer a somewhat parallel example from a different system - Burning Wheel.
Burning Wheel doesn't use the same mechanics as AW/DW. Of well-known RPGs, a PC sheet in BW looks a bit like a RQ or RM sheet - there are stats and derived attributes, skills from a
long list - Sword-fighting is a different skill from Mace-fighting; Blacksmithing is a different skill from Silversmithing; Intimidation is a different skill from Interrogation is a different skill from Command is a different skill from Ugly Truth; Great Masters-wise is a different skill from Towers-wise is a different skill from Gynarch of Hardby-wise; etc.
Resolution is say-"yes"-or-roll-the-dice: which is to say, if nothing is at stake in the action declaration the GM just says "yes", the fiction changes as appropriate, and we keep going. If something is at stake - and the main criterion for this is the player-authored PC Belief - the a check must be made. In this case, intent-and-task coupled with let-it-ride apply: if a check succeeds, the task succeeds and the player's intent is achieved; if a check fails, the GM narrates a consequence which
must go against the intent, and which may (but need not) mean that the task failed; either way, there are no retries.
If my declared action for Aramina is
to recall the location of Evard's tower (task) because
I (as my PC) want to know a place where I might find spellbooks and other magical lore, then the first question is
is anything at stake? In this case the answer is an easy "yes", as Aramina has a Belief that
I'm not going to finish my career with no spellbooks and an empty purse! And this is far from coincidence: the BW rulebook gives the following instructions (among others) to players (p 552 of Gold edition):
Use the mechanics! Players are expected to call for a Duel of Wits or a Circles test . . . Don't wait for the GM to invoke a rule - invoke the damn thing yourself and get the story moving!
Participate. . . . If the story doesn't interest you, it's your job to create interesting situations and involve yourself.
So on this occasion I was doing just as instructed: getting things moving, and creating interesting situations, by invoking the mechanics (here, the Wises mechanics).
Next, then, is the question of which skill to check. I helpfully suggest to the GM that Great Masters-wise seems apposite, given that Evard was undoubtedly one of the great masters I (as Aramina) would have learned of during my training as a Neophyte Sorcerer (which is where I acquired my skill in Great Masters-wise). The GM set the obstacle, having regard to how obscure such knowledge would be for someone learned in the doings of the Great Masters. I made the check, and succeeded. And so Aramina recalled what she had hoped to. In due course (by no means straight away) Aramina and Thurgon made their way to the tower.
We can step back and identify what assumptions both BW and DW rest on: they assume that
there is always going to be play taking place, that PCs will be in challenging situations and declaring actions for their PCs; and they assume that it
matters whether, in these situations - both how they begin and how they resolve - the PCs are advancing towards their goals, or failing/being defeated. For the player, the pleasure of play has two parts. One is always occurring: the player is getting to declare actions for their PC and find out what happens. The other is intermittent: the player is getting what they as their PC want.
They actively
do not assume that an important part of play is learning the content of the GM's notes, or solving puzzles/mysteries which the GM has planned in advance and posed to the players (via their PCs). So consider the example someone posted upthread, of the PCs wanting to cross a river. For some PCs, making a Boatwright or Carpentry or Shipwright or other salient crafting check might be the way to tackle that problem. For other PCs, making an Orienteering check, to find a crossing, might be best. For yet others, making a Bridge-wise or Rivers-wise or Trails-wise check might be best. In Thurgon's case, my approach is to make a Circles check to find a friendly former member of his order ready to carry Aramina and Thurgon along and across the river on his raft.
In actual play, some of these possibilities might be more or less optimal, depending on fictional positioning and obstacles (though "optimal" here is tricky, because a skill can only advance in BW if it is checked against a variety of different difficulties; so only facing easy checks means your skills never advance). But all are playing the game. All will be resolved using the same basic process. The idea that some approaches involve "easy mode" or circumvent/subvert challenges is completely misplaced. (I think one assumption being made here is that there can be "no/low risk" checks. I gather those are a thing in 5e D&D play. They're not a thing in DW or AW.)
As I've said, DW doesn't play identically to BW. But I think it's near enough, for present purposes, that the points I've made generalise to it too.
<snip>
So who gets to tell you what you remember? Someone has to author it. My own experience is that coming from the player increases authenticity, ownership and immersion in/inhabitation of character. Whereas being told by the GM what I've learned over the course of my life is like having someone else play my character. Or, as GM, is like being asked to play the player's character for them; which I prefer not to do.
<snip>
The trigger for Spout Lore is the player's consultation of their accumulated knowledge about something, so that's what is sufficient (ie that the player has their PC do that). But given that the GM has to (on a success) provide information that is interesting and perhaps relevantly useful, and those are relative concepts (ie interesting
to someone; useful
to someone
for some purpose), if the player doesn't make this clear in their action declaration I as GM will have to ask. I've done that when GMing 4e D&D, and a player has wanted to declare a knowledge-type check and hasn't made it clear what they are hoping to recall, or to achieve via their recollection. The harder this is - the more I have to pull teeth to get (what BW would call) "intent" - the more frustrating to me as GM.
As far as fleshing out details of what is recalled, that is highly contextual. But it certainly makes my work as GM easier if players play their PCs, including their backstories and memories, given that I've already got plenty on my plate framing scenes and managing NPCs!
<snip>
in Burning Wheel the result of a successful check is that intent and task both succeed, so if the task is to remember where Evard's tower is in order to know the location of a possible source of spellbooks and other magical lore, then if the check succeeds that's that.
In DW the constraint that follows from a success is "interesting" and, if the success is 10+, also relevantly useful. Upthread I suggested a possible narration that might depart from Forges yet honour that constraint. But I also posted, and followed up with a bit of back-and-forth with
@Manbearcat and I think
@Ovinomancer, that the most straightforward way to have something interesting and useful is to have the posited recollection be true!
Another constraint is also at work here in DW: the GM has a duty to be a fan of the PCs. I think it would violate this duty if the GM consistently narrated successes in such a way as to make the PC look like they are confused or haphzard in their memories.
And I think part of the reason for that constraint, in this context, is the following: how does it improve play to routinely substitute
the GM's conception of what might be interesting and useful for the player's?
