A Question Of Agency?

[Manbearcat]

I think @prabe has hinted downthread at a corollary question here: while success should clearly be honoured in the moment, for how long does that success remain valid?

Lots and lots of posts. I've got a lot to say about several topics but not the time to say them, so I'm just going to address this.

HOW LONG DOES SUCCESS REMAIN VALID IN PBTA GAMES (AND THE LIKE)

There are three conditions under which a GM can put an earned gamestate (eg - gained asset/ally, etc) under pressure (pressure here meaning threaten a potential change in gamestate, not outright revoke by fiat):

1) While interacting with the establish fiction, the PCs have actively done something to perturb the nature of their prior earned gamestate (meaning, the GM cannot use their uniquely situated access to deploy offscreen backstory that players aren't privy to...and again, going back to the lead post, offscreen info on Fronts in DW and Threats in AW are low resolution things so GMs can basically "have interesting things to say"; conflicts to frame, threats to telegraph and then put to action if the PCs don't act upon them).

Example:

* The PCs have established an alliance with Gang A who has an alliance with Gang B (who happens to specialize in running cargo from x to y) and the PCs are aware. The PCs attack Gang B's caravan causing monetary and personnel damage to Gang B. Gang A is going to need some answers. Its time for the GM to make a "soft move", meaning frame a looming social conflict with Gang A to right this ship (GM: You arrive back at your warehouse after delivering the stolen cargo. There are motorcycle tire tracks leading off to the road like someone lit out not long ago. The overhead door has a letter pinned to it by a piece of scrap violently smashed through the aluminum. "We need to talk. Meet at the cesspool at dusk. D <leader of Gang A> PS: If you bring guns I'll know.")

2) The PCs have violated the mechanical terms of a hireling/henchman/companion's "social contract". That means they've put the hireling in a spot/bad position.

Example:

You've asked a Minstrel or a Porter to pick up a weapon and help fight off a Wyvern that has attacked camp. Minstrels and Porters ain't warriors so you made the Loyalty move and either got a 7-9 (they'll do it, but they come back with major demands) or they failed with a 6 or lower (which means they aren't doing crap but those major demands are coming). The higher the Loyalty (which comes with consistently paid the hireling's specific Cost - whether that is money or vice or glory or whatever), the less likely this move goes bad. Refuse to address their demands and the hireling is done with you (carry your own overladen sacks of equipment/treasure and compose/sing your own ballads).

3) Your Warhorse is a fine specimen and can carry 12 Load (Encumbrance stat) beyond you. After a successful delve, it has multiple sacks of coin that puts it right at its Load limit of 12 (1200 Coin). On the Undertake a Perilous Journey move on the way back to town, the Scout move failed which led to an ambush with Worg-Riding-Orcs and an ensuing chase. You made your Warhorse Skill move and it yielded a 7-9; Your horse is galloping and huffing and frothing with all its might...its stride tires after a good stretch, fatigue taking it...you can get away but you're going to have to unburden the horse of 1 Load worth of Coin (100 out of the 1200) or you'll exhaust it. Or you can stand and fight...or go off the Old Road and into Bleak Mire in the dead of the night...you know the legends...the Orcs sure as hell won't follow you in there.

Success with a Cost/Complication or some kind of alternative choice that outright removes the orc threat or engages with it.

 

[Manbearcat]

Well, there are a lot more resources that players can bring to bear on these rolls to improve their chances or the results.

But even then, you’re really ignoring the actual math as it’s explained in the book. With 1 die, you have a 50% chance to succeed. A character who has zero dice in an action would normally roll 2 dice and take the lower roll. That character can push the roll so that they get 1 die and then have a 50% chance to succeed at something they aren’t even good at.

Yes, some of the time when characters succeed at something, something else will happen that they likely didn’t want. That doesn’t negate their success.

It’s also no different than your typical combat turn in D&D. When my fighter hits the orc it doesn’t mean he also gets to avoid being attacked in return, or hit by the orc’s archer pal, or charmed by the orc shaman.

The math indeed is a bell curve with the basic distribution putting about 45-50 % of the outcomes in the Success w/ Cost/Complication range then 25-27.5 % on the tails (or about 2/3 success rate). Obviously as resources are brought to bear, you use your better score, you get help, odds will increase (and vice versa).

This is by design as you noted. Success w/ Cost/Complication is the snowballing engine, the primary content creator, the beating heart of the PBtA games.

 

[aramis erak]

In hockey the odds are even worse: a non-goalie player is doing really well if he consistently scores on 15% - or 3/20 - of his shots on goal (and this ignores any shots that go wide, or are blocked before reaching the goalie). By what you're saying, that player shoud just give up trying to score at all.

The thing is, if one were to play out a hockey game in combat rounds, it's quite possible that there could be 3 shots on goal in one round, but I doubt the average hockey player qualifies for 2 attacks per round, let alone 4, with the puck off his stick. The roll to score against the Goalie's score is better than shots rate.

Likewise, the rounds in D&D were 1 per 6 sec (1 per 60 sec in AD&D1, too lazy to look up AD&D2)... In both, that's multiple attacks for even incompetent fighters.

Assuming zero defense other than blocking, most people can hit a man or his shield about 90% of the time. But landing that blow hard enough to injure? much lower. so the 3-5 presumed swings/strikes in a D&D round are combined into a "did I hurt him?" check, mislabeled as a to-hit. That blows are landing is presumed. (Explicitly presumed, in AD&D.)

 

[Lanefan]

Not quite sure what you're getting at here...

The thing is, if one were to play out a hockey game in combat rounds, it's quite possible that there could be 3 shots on goal in one round, but I doubt the average hockey player qualifies for 2 attacks per round, let alone 4, with the puck off his stick. The roll to score against the Goalie's score is better than shots rate.

???

The rough average of shot attempts per game for a TEAM in the NHL is about 55-60*; with roughly 30 of those counting as shots on goal as they either go in or the goalie stops them, while the rest either hit the post, go wide, or are blocked by someone other than the goaltender. The average number of goals per game per team is about 3 (actually a bit less but close enough for here), meaning the overall average score-per-shot-on-goal rate is close to 10% and the score-per-shot-attempt rate is about 5%

* - as a non-overtime game is 60 minutes of play, that's about one shot attempt per team per minute...which nicely dovetails with the 1-minute 1e round length.

Likewise, the rounds in D&D were 1 per 6 sec (1 per 60 sec in AD&D1, too lazy to look up AD&D2)... In both, that's multiple attacks for even incompetent fighters.

Assuming zero defense other than blocking, most people can hit a man or his shield about 90% of the time. But landing that blow hard enough to injure? much lower. so the 3-5 presumed swings/strikes in a D&D round are combined into a "did I hurt him?" check, mislabeled as a to-hit. That blows are landing is presumed. (Explicitly presumed, in AD&D.)

Sure. But the game concatenates this all down to one (or more if a fighter has multiple attacks) roll per round, under the "this is your best attempt" theory.

1-minute rounds are too long to be realistic. 6-second rounds are too short to allow very much movement or to allow enough time for anything external to develop while the combat continues (e.g. someone running to get help and having that help arrive before the battle's over). As a compromise I went to 30-second rounds in about 1986 and still use that today.

 

[pemerton]

I find it quote odd that posters who are asserting the freedom of the GM to impose consequences by deciding how the world reacts (eg @prabe, @Lanefan) are also expressing hostility to a system that mechanically dictates when consequences are to be imposed and (roughly, at least) what those parameters are.

To me it suggests that the true hostility is not to consequences, but to the removal of GM freedom to impose them willy-nilly and independent of the action resolution process.

I don’t know if it’s hostility, but I do think it’s related to the shift in narrative authority. It’s a change in the classic or standard approach to play. I get that one might have concerns about how that would impact the play experience.

<snip>

there’s always going to be preference involved. Some games or styles will just appeal more to some folks than others.

I absolutely agree that preferences are involved. And concerns about the play experience. But what's the fulcrum about which those concerns are pivoting? As best I can tell - given the expressed desire to be free to narrate consequences, and the expressed desire not to have a mechanical system for rationing/determining that - it is precisely the degree of latitude the GM has to introduce whatever fiction s/he wants.

I want the world to respond to the PCs' successes. I want Duke Fornyard to resent them for succeeding in swaying the king; I want the merchant thief Iltan to mark them as potential marks after a profitable dungeon raid; I want Winter's Fang to notice that Auriqua is no longer under the Tundra Queen's protection and take interest. To take examples that at least would fit into the campaigns I'm running.

<snip>

Other than having an NPC do something as a result of what sounds like a Perception-type check (and I'll trust you on that being a legitimate way for that to shake out in that game) this isn't wildly different from how I GM.

As @AbdulAlhazred has posted (forcefully) in this thread, the world is a fiction. So "the world responds to the PC's successes", as you are using it, equals the GM establishes the fiction that s/he wants. In RPG systems that foster player agency, the GM's ability to do that is constrained - by dice roll outcomes (eg DW, AW, BW., 4e combat and skill challenges), or by a combination of those plus rationed resources (Marvel Heroic RP/Cortex+ Heroic, with its Doom Pool), or by principled restraint (the "set in stone" of DW's Discern Realities, Let it Ride in BW, in general honouring successful action resolution if the players haven't re-staked it).

To pick up a concrete example, what does the merchant thief Iltan marks them as potential marks actually mean? Assuming that you are not referring here to an actual encounter (and by you reference to "other than having an NPC do something as a result of what sounds like a Perception-type check" I take it that you're not), then it is not part of the shared fiction. It's just a notion the GM has. A Burning Wheel GM can have the same notion, but should be having regard to the principles of the game: (1) does it speak to a PC's player-authored Belief? (if not, don't do it); (2) if it's something adverse, introduce it as a consequence of failure.

Assuming that (1) is satisfied, then when can Iltan's plans be brought onto the stage? An obvious opportunity would be a failed Wises, or Circles, or perhaps Resources check - the PC is looking around for the <info, informant, goods, etc> and is having trouble finding them, when lo-and-behold, a cloaked NPC steps out of the shadows: Iltan can given you what you're looking for . . .

This is player agency building and helping to drive the fiction - Iltan is not just a conceit the GM has and plonks into the fiction to determine its direction; but is an element in an unfolding story driven by the PCs' protagonism.

 

[pemerton]

the GM can't have any ongoing mysteries or overarching secrets that the players may or may not one day find out about.

It's quite easy to have mysteries in RPGing that have a high-degree of player agency.

I've already posted an example from my BW play: who is Rufus's master? and perhaps why does he want wine?

Another example from the same campaign: while exploring Evard's tower, Thurgon found old correspondence that implied that Xanthippe, his mother, is Evard's daughter. That was a shocking revelation!

But you are correct that the mysteries and reveals will not be unilaterally driven by the GM; with the play in respect of them just consisting of the players trying to work out what the GM has thought up. That is a consequence of the players having agency! And hence of the focus of play being something other than learning what is in the GM's notes.

she can't have her world/setting do anything in a cause-effect way absent the PCs' direct involvement; i.e. can't have something happen that isn't a consequence of anything involving the PCs but is instead either pre-scheduled or by sheer random chance (e.g. a volcano erupts, a pirate ship sails into harbour and opens fire on the town, etc.). That there's too much constraint on when and how consequences may be imposed and what form those consequences may take. That consequences cannot be imposed on an outright success even when logic would say they likely would occur.

This is exactly what I posted upthread: you object to the limits on the GM that flow from player agency.

In a player-driven game events are not narrated just because the GM thinks they would be interesting to him/her.

There can be pirate ships and volcanoes - but these will be used in accordance with the sorts of principles I, @Manbearcat and others have described upthread.

Who gets to determine the weather conditions when the PCs wake up each morning?

Luke Crane discusses this in the BW rulebook: if a character chooses the Weather Prediction ability it's because s/he wants to be able to do something involving a certain sort of weather, and its often easiest just to let him/her have that weather.

If the weather is not at stake, then the GM can narrate it just as s/he might narrate the colour of the flowers on a windowsill. A skilled player might then play on that fiction, but given that we know it's not what is at stake that would be means, not ends. Just as if the GM had narrated red flowers, and then a PC needs to dress as a clown, the player might know how his/her PC can colour his/her cheeks red.

 

[pemerton]

If a player rolls and succeeds, I always frame the result as a complete success. The reason I do so is that I feel that anything other than a complete success is me as GM being mean and robbing the player of their victory. Why I feel I need to do this is because I only call for rolls when a complete failure would also be an interesting result. If neither, or both, complete success or complete failure is not an interesting result, then I simply tell the player they automatically succeed or fail. Thus far I haven't had too many fits of rage from players as I am very prone to auto-pass, and almost never auto-fail. I'm probably a super pushover GM I guess, I just like to see the PCs be awesome and make the players happy. Anyway, I'm prepared to get flamed to death now...let it begin!!!

Burning Wheel is a complete success system. That's a fundamental principle stated and emphasised very clearly in the rulebook.

But it's not a complete failure system. Action declaration requires both task and intent (eg as we head through the outskirts of Auxol I'm keeping an eye out <task> for Rufus <intent>). The general advice for GM narration of failure is to focus on intent - so the GM is entitled to narrate failure as success at task, but never intent. Of course sometimes task fails to - in BW that's up to the GM.

In AW and DW, on a failure the GM is entitled to make as "hard and direct a move as s/he likes" - ie s/he can narrate whatever s/he wants that follows from the established fiction. This may be failed task, failed intent, or both. "Success with complication" happens in those systems on a 7 to 9 result - most moves elaborate in more detail what this looks like so eg an attack action might allow dealing damage on 7+, but unless it is 10+ the PC also suffers damage in return; an escape or avoidance action might succeed on 7+, but unless it is 10+ the escaping PC "brings something with him/her" eg one opponent follows, or s/he takes a wound on the way out, or something more elaborate or specific appropriate to the fictional context.

To relate this to what @prabe and @Lanefan have been posting - a 7 to 9 will generally be the context in which the GM might establish those adverse consequences (you've brought Pup to heel, but the followers may not be going along with it; you've stolen the jewel, but you catch a hint of a scrying viewer shimmering in the air as you make good your escape); whereas a 10+ is more likely to be genuinely free-and-clear.

That's how these PbtA systems use mechanics to generate rising action, climax, pay-off etc (on a probability curve rather than strictly deterministically); whereas Burning Wheel leans more heavily into GM judgement in this respect, and so does a system like Classic Traveller which (in most of its subsystems) doesn't have the mechanical sophistication of the PbtA games, leaving the referee to decide what consequences should look like and to manage pacing as part of that.

 

[Crimson Longinus]

But I didn't give an example of a GM giving motivations and plans for Rufus.

I gave an example of the GM narrating an encounter with Rufus, in which Rufus says certain things.

This is (in my view) a very big difference. As I see it, it is the difference between (1) the GM coming up with a story on his/her own, and gradually revealing it to the players - the paradigm RPG for this approach that I know of is CoC, but a lot of D&D also seems to be played this way - and (2) the GM presenting a fictional situation which is pregnant with possibility that the participants care about (because the GM has built on what those participants have signalled that they care about) and relying on the play of the game, including the action resolution mechanics, to determine what happens next.

Semantics. You're describing the same thing with different words. "GM presenting a fictional situation which is pregnant with possibility that the participants care about (because the GM has built on what those participants have signalled that they care about)" is literally what setting up a plot hook is.

The GM can introduce Rufus collecting wine for 'the Master' with none of the following questions being answered: who is the Master? is he related to Thurgon, or Aramina, or Evard (whose tower Thurgon and Aramina not long ago burned down), or Thurgon's fallen order (the Knights of the Iron Tower, specified in backstory as a component of PC build)? Why does the master want wine - for a dinner party? for a sacrament? because he's an alcoholic? because he wants to bribe some orcs from attacking Auxol?

Those things - and of course indefinitely many others - are all put in play by the encounter with Rufus that I described. I don't know the answer to any of them. Neither does the GM, and he doesn't need to in order to frame and adjudicate the encounter (which he did).

Where in the south has Thurgon's brother gone looking for glory? The Hold of the Sea Princes? The Amedio Jungle? The south of the Pomarj? Did his wife go with him? Again, none of this is, or needs to be, known in order for the play that occurred to occur.

But sooner or later someone has to determine these things.

What's one way that we might learn whether the younger brother's wife went with him? If I declare a Circles check for Thurgon to meet her. If I succeed, we know she's still in Auxol. If I fail, one narration of that failure could be She's not here, she's gone south with her husband. That would be player agency in action.

That is clearly narrative level power. Your character in the setting cannot affect where people in the setting are.

This is all illustrative of the technique described by Paul Czege that I quoted upthread:

I frame the character into the middle of conflicts I think will push and pull in ways that are interesting to me and to the player. I keep NPC personalities somewhat unfixed in my mind, allowing me to retroactively justify their behaviors in support of this​

And both in itself as a technique, and in its relationship to player agency, it's quite different from the GM giving motivations and plans for the orcs.

Yes, this is 'the quantum reality' illusionism relies upon. We discussed this earlier in this thread in the case of the lying/truthful NPC.

But he fact remains that at some point someone has to decided certain things, be it it the GM, the player or some random chart. At some point it (presumably) will be revealed who the 'master' is, and what they want, and someone has to decide that! If the player decides that, they're assuming narrator stance, if the GM does, that's the GM setting up plots.

No. Action declaration: As we ride through the outskirts of Auxol, I keep my eye out for Rufus. It's five years since I've seen him - I wonder how he is doing? Resolution: make a Circles check.

The basic structure of declaration and resolution is no different from a Streetwise check in Classic Traveller c 1977, or Gather Information in 3E D&D.

Difference here seems to be that you want the player have the power to dictate the reality rather than to merely discern it. Like you basically want the player to be able to declare "I try to find a person X" and on a successful roll that person manifests.

What can I say - I don't agree that there is that "shitton" of other stuff that the GM has to make up. If the players haven't declared an action for their PCs, or aren't looking for the GM to frame a situation that will spur them to do so, then what is the GM doing worrying about things?

Perhaps concrete examples would help me work out what you have in mind?

Let's go back to our earlier example. Characters are deciding whether to go to Grim Chasm or the Gnarly Forest. Who decided that these places even exist and where they are?

The characters try to recall what they know of these places. Who determines what there is to be known and how easy it is to know?

They decide to go to the Gnarly Forest. Who determines what they meet there?

It is somehow determined that they find some dead spiders in the forest. They decide to examine the corpses and their vicinity. Assuming that they're successful, who determines what they will find out?

It is somehow determined that the orcs have skilled the spiders. It was also earlier somehow determined that orcs are not native to this area. Now the characters wonder what the motivations of the orcs are and prepare to a confrontation with them. Who determines what the orcs want and why they are there there?

 

[Crimson Longinus]

Another example from the same campaign: while exploring Evard's tower, Thurgon found old correspondence that implied that Xanthippe, his mother, is Evard's daughter. That was a shocking revelation!

And who decided that this was the case?

 

[pemerton]

Semantics. You're describing the same thing with different words. "GM presenting a fictional situation which is pregnant with possibility that the participants care about (because the GM has built on what those participants have signalled that they care about)" is literally what setting up a plot hook is.

No it's not!

I'm going into my PDF files to find a D&D module with a plot-hook <looks in folder> - here's one, H3 Pyramid of Shadows.

Page 4 has, on the left hand side, an Adventure Synopsis:

Karavakos desperately wants to escape from the Pyramid of Shadows. . . . Karavakos has been sending visions of the pyramid, its location in the natural world, and promises of power associated with the pyramid. The adventure begins when the player characters encounter the pyramid and are drawn into its timeless depths. From that point on, Karavakos​

encourages the adventurers to destroy the splinters of his life force so that the power each possesses returns to him. With his power fully restored, Karavakos plans to perform arcane rituals that will set him free. . . . Presenting herself as an ally, Vyrellis guides the adventurers toward the Sanctuary of Light and urges them to destroy Karavakos—all of the splintered versions of him as well as the true wizard—and​

win their freedom in the process.​

​

Over the course of the adventure, the player characters explore the bizarre halls and chambers of the Pyramid of Shadows, fight its hostile inhabitants (including the splinters of Karavakos’s life force), collect the three keys needed to open the Sanctuary of Light, and finally face the true Karavakos in a pitched battle—with death or eternal imprisonment the price of failure. Along the way, Vyrellis also guides them to collect the splinters of her life force from the gemstones that hold them in hopes of restoring herself once she is freed from the pyramid.​

That's a plot, with NPC with pre-authored motivations and plans.

And the right hand side of p 4 has some Adventure Hooks:

If the adventurers experienced the events of H2: Thunderspire Labyrinth, then they discovered a map among Paldemar’s possessions.​

​

The map shows a glowing pyramid within a lush forest . . .​

​

OR​

​

Wherever the adventurers happen to be prior to the start of this adventure, a local wizard, scholar, or sage has been troubled by strange and compelling visions every night when he or she sleeps and dreams. In these dream visions, the tiefling wizard Karavakos appears to the dreaming mage and whispers about the power and secrets waiting within the Pyramid of Shadows. . . . The local wizard or sage is intrigued by the visions and anxious to claim the promised power. He or she is also suspicious of these dreams, and as frightened by the implications as he or she is desirous to fulfill the impulse to follow the dream.​

​

Hearing of the exploits of the adventurers, or perhaps knowing them as friends or colleagues or acquaintances, the wizard/scholar/sage asks them to look into this matter.​

There are some more like this on p 5.

The episode I described from my Burning Wheel session has nothing in common with this. There is no hook into the GM's pre-authored adventure. There is a series of action declarations by me, the player, for my PC and his sidekick, and the GM responds to those as the rules and principles of the game call for. Going in, neither of us knows anything about Rufus beyond what was in my PC backstory:

Thurgon’s father is deceased, but his mother Xanthippe (now 61 years old) still lives on the estate. So does his older brother Rufus (40 years old)., the 9th Count of Adir (although for the past 66 years that title has counted for little, having been usurped by others). . . . Although Auxol is now owned by servants of evil, the family continues to manage it. Xanthippe ensures that the estate serves as a bolthole for refugees. Rufus is sympathetic to their plight, but sees them ultimately as someone else’s problem. His interests are more mundane (it is fairly common knowledge that he has a 3 year old illegitimate son with a middle class townswoman).​

Coming out, we now know that Rufus is serving 'the master', who needs wine and that Rufus is ashamed and that Thurgon and Aramina could not snap him out of his shame, nor cow him into giving them some coin.

There is nothing like the structure of plot and plot-hook. This goes all the way back to @chaochou's post upthread:

in the case of player agency, the context is this: who is creating the purpose of the character?

That‘s the matter in question. Let’s say the GM creates the purpose of the character(s). If the players object to their predetermined fate, you have force. If they are unaware, you have illusionism. If they are aware, but don’t object (such as when a player accepts a ‘hook’ for a scripted plot line) then you have participationism.

Player agency is player freedom to create the purpose for their character and for the game content to begin, and grow, from that ongoing act of creation. It’s not one and done, the purpose can and should change as the game state changes through resolution. When the game follows the player’s protagonism in this way, then there is agency, and it’s completely obvious.

******************************************************

But sooner or later someone has to determine these things.

But all those things - like who the master is, why he wants wine, where Thurgon's younger brother has gone, whether or not his wife has gone with him - can be determined in just the same way as the encounter with Rufus was resolved: that is to say, as part of the back-and-forth of action declaration, action resolution, narration of consequences, and principled addition of further framing elements.

Another example from the same campaign: while exploring Evard's tower, Thurgon found old correspondence that implied that Xanthippe, his mother, is Evard's daughter. That was a shocking revelation!

And who decided that this was the case?

The GM, as part of the process of action resolution.

The discovery of Evard's Tower was (in the fiction) a result of Aramina's memory of tales of its presence in the area, and (at the table) a result of a successful check on Aramina's Great Masters-wise ability. The salience of Thurgon's mother is due to her being a Relationship purchased as part of the process of PC building.

That is clearly narrative level power. Your character in the setting cannot affect where people in the setting are.

Upthread you used the phrases "narrator perspective" and "narrator stance". I made the point that, in fact, I never did anything but declare what my PC is doing - ie looking out for Rufus. Now you are talking about "narrative level power", by which you seem to mean action resolution that can produce outcomes that, in the fiction, are not solely under the causal influence of the player character.

As the example I posted shows, that sort of power does not require any distinctive "perspective" or "stance". And if players never have such power in RPGing, they will have very little or no agency. For instance, whether or not the Orc blocks my sword with a shield is not solely under the control of my PC. So if I can never influence that via action resolution, the GM is always free to narrate the Orc's shield block as a response to my action declaration I attack the Orc with my sword.

The fact remains that at some point someone has to decided certain things, be it it the GM, the player or some random chart. At some point it (presumably) will be revealed who the 'master' is, and what they want, and someone has to decide that! If the player decides that, they're assuming narrator stance, if the GM does, that's the GM setting up plots.

Now you are back to "stance".

There are any number of ways these things can be done. As I've already noted, Classic Traveller (1977) - that game well-known for its radical indie features! - settles the question of whether or not there is someone willing to sell illegal firearms at a good price via a Streetwise check. This doesn't require the player entering "narrator stance". It just requires the player to say "I put feelers out - who here sells illegal guns at a good price?" On a successful check, the referee provides the answer. (Classic Traveller is a bit weak when it comes to advice on the narration of failures, but this one seems easy enough: it could be anything from a visit by the local constabulary, to some toughs come to rough the PC up.)

Like you basically want the player to be able to declare "I try to find a person X" and on a successful roll that person manifests.

That is how Circles work in Burning Wheel. That is how Streetwise works in Classic Traveller. That is how a paladin calling for his/her warhorse works in AD&D. It's not a very radical mechanic.

Let's go back to our earlier example. Characters are deciding whether to go to Grim Chasm or the Gnarly Forest. Who decided that these places even exist and where they are?

The characters try to recall what they know of these places. Who determines what there is to be known and how easy it is to know?

They decide to go to the Gnarly Forest. Who determines what they meet there?

It is somehow determined that they find some dead spiders in the forest. They decide to examine the corpses and their vicinity. Assuming that they're successful, who determines what they will find out?

It is somehow determined that the orcs have skilled the spiders. It was also earlier somehow determined that orcs are not native to this area. Now the characters wonder what the motivations of the orcs are and prepare to a confrontation with them. Who determines what the orcs want and why they are there there?

I don't know. What system are you playing? What mechanics does it involve? What principles apply?

In my Burning Wheel game, a successful Great Masters-wise check established that Evard's tower existed nearby. The GM decided that, when we arrived there, a demon attacked - that's the GM's prerogative in framing, and given that (i) the successful check established that Evard is an evil sorcerer and (ii) Thurgon is a faithful knight of a holy order, it accorded with the principles of go to the players' evinced interests and concerns for their PCs.

In the Prince Valiant game that I GM, the players decided to travel to the Holy Land to go on a crusade. I decided - using my prerogative in framing - that their ships had to land on the Dalmatian coast, so that the last leg was to be undertaken overland. I decided - again using my prerogative in framing - that they encountered the Bone Laird. The players decided to spend a fiat resource (Storyteller's Certificate) to find the locus of the Bone Laird's curse. When one of the players succeeded on a check to interpret the magical signs of that place, I narrated what they learned (not unlike Discern Realities in DW or Read a Situation in AW, which trigger GM narration). You can read the rest by following the link, if you like.

In some approaches, of course, the system and techniques are more like that evinced in H3 Pyramid of Shadows - the GM decides everything in advance and the main thing the players do is declare actions to trigger the appropriate GM narration. This is RPGing with low player agency.