Realistic Consequences vs Gameplay

[Ovinomancer]

The rulebook I'm looking at says, "When you Go Aggro on someone, roll+hard. On a 10+, they have to choose: force your hand and suck it up, or cave and do what you want. On a 7-9, they can instead choose 1:"

That tells me it's not the player choosing, it's the GM, so the player really has no control over the outcome. The example I laid out is book-legal, though it'd be horrible GMing--and I never said it was otherwise.

Yeah, the point is that the player forces the GM to pick one of the two. This doesn't exist, ever, in 5e. Either the GM has the NPC agree with your demand, or the NPC suffers whatever you've threatened. The Burgomaster calling for the guard is neither of these -- it's not a book-legal move by the GM.

The key is force your hand and suck it up -- this means they choose the "or else" and that "or else" happens to them. The nature of AW means that Go Aggro requires an 'or else.' That's missing in the OP, but given that the player in question immediately tried to take the Burgomaster hostage, I went with that as the "or else." That's the nature of Go Aggro, on a success, you either give in or you suck up the "or else."

Oh, yeah, they're different games, aiming at generating/enabling different stories, and I have no doubt that Baker played in some deeply dissatisfying games in other systems before he wrote his own.

I don't think I've been unclear that published adventures, especially adventure paths, are problematic for character agency--for the characters mattering much, even--and I also don't think I've been unclear that 5E is designed to enable play through published adventures. So, it's probable that many players' experience of 5E is going to be ... less than ideal. I think, though, that there's support in the game to play differently; I don't think 5E is limited to that sort of play.



The rulebook I have says "Force your hand and suck it up." BurgerMaster calls for guards and other party member attacks (tries to take him hostage). That genuinely doesn't sound all that incongruent to me. Yes, it's bad GMing in AW if any result of Go Aggro will have that result, and I've never said otherwise; I've just said it's possible to GM that way (and that the play examples in the book don't discourage it, actually seem to suggest it).

What did the Burgomaster suck up? What bad happened because he refused? Your example has no bad for the Burgomaster. It has good -- the odds shift in his favor. Choosing this outcome isn't just bad GMing -- it's not following the rules.

And, no, the example in the book follows this exactly -- the GM chooses to have the NPC not accede and so he gets brain fried. Since the specific move used doesn't require going loud as part of the Go Aggro (which usually does), the PC was able to leave the brain fried NPC without starting a fight with the henchmen. Seems like exactly what needed to happen -- now the NPC has a serious level of Harm, which will make any future engagement easier for the PCs until the NPC can reasonably get help (if the PCs, for example, don't press for a while, I can see that Harm rolling off).

Yeah, I understand the mechanics, and I understand the ... rationalization of the mechanics--how the mechanics are meant to reflect/shape the emergent story.

OTOH: A 5E DM could randomly roll to determine if a door was trapped--the old school-ish random dungeons seem a likely application for this. He'd be finding out if it was trapped about the same time as the PCs. That's not my prefered playstyle, but it's not meta the same way as having it hinge on the outcome of a Perception check (or the equivalent).

Randomly determining if a door is trapped is nothing like what I described play in PbtA as, regarding the fiction of a trapped door. It's not random.

Yeah. The GM's job is to place obstacles in the characters' way, and to present plausible opposition. That's ... pretty close to universal (there might be edge cases but I don't think they're the focus of discussion). Without the obstacles and/or opposition, there'd be nothing to center a story around--no decisions or actions that mattered. I think my sense is that having the world exist in a more or less objective sense (to use the most-current example, that door is trapped) makes it clearer that the GM is neither the obstacle nor the opposition; that seems to be true for me as a player, as well as as GM.

It's plausible I'm bouncing as much off Baker's writing as the game mechanics, but I did come to a similar conclusion about Fate (that the game needed the GM to be more antagonistic than I wanted to be), and that game is written ... more conventionally--and yes, I remember (I think) that you don't think Fate goes far enough.

I wouldn't say I've ever put my players through the wringer, but I would say the Masked Ones killed Imaktis, and more the Tundra Queen seems to have drawn their ire. The players have been coming back every other week for more than two years, so it's tempting to say they're digging it. There's probably some fundamental-ish difference in how we look at the stories that emerge from play, and the elements thereof.

FATE doesn't go far enough in telling you how to play it -- it's wishy-washy. As I run a 5e game you'd be hard pressed to find isn't by the book, I don't have a problem with either style of game. FATE just doesn't give enough insight into how it works and so appears to support multiple playstyles -- and it does, to a degree, but if you bring a D&D mindset, it's not going to work well. That's my gripe with FATE -- it just soft pedals that it's actually a different game, so people bounce off of it.

Let see, the last few campaigns in 5e I've run -- a Big Plot game, which was a cosmic mystery, full of deep backstory to uncover and plotting tightly; a hex-crawl exploration game of a prison plane, not plotted but mapped pretty well; and a Sigil-based Planescape game where I don't yet know who the villain of the campaign will be, or what will be the focus, despite having run it for a bit over a year. I don't have a "way" I see how stories emerge, I see lots of "ways." And I'll use every one of them, if they're fun. But, when I run/play Blades, and when I look at PbtA games, I see how they're used to tell stories and I use that way when I run those games. I do different things when I run 5e.

Regardless, how you run games like FATE and Blades is not antagonistic, or petty jerkry. It's actually far more disciplined and constrained than most of the other games I've tried, especially 5e. Not to say that individuals aren't disciplined when running 5e, but the system has very little discipline. It says, "the GM decides," and pretty much leaves it there, maybe with some vague handwaves at technique. You can clearly see this in the official adventures, which are all over the map in approach and certainly don't leverage the ruleset very well most of the time.

 

[Ovinomancer]

I can find examples that contradict this--at least at the author's conscious level. Just off the top of my head: King has said for decades that he never meant for the kid to die at the end of Cujo, and Alice Walker explicitly thanked the characters in The Color Purple for showing up.

No, you cannot find example of a character having agency. What you can find examples of is an author surprising themselves as they imagine the character and make choices for it.

I agree that confusing what the players do with what the characters do makes for problems. Those problems seem inevitable in games that allow (or encourage) PvP, which is why I pretty much don't allow it at my table. I don't think it's as confusing to think of the characters as having agency as you at least seem to, though, in any sort of fiction.

Treating characters as separate from their authors definitely leads to confusion, such as the argument that D&D supports agency for the characters. Again, I invite you, next time you play, to not declare actions for your PC, to not make choices for the PC, and see exactly how much the character does on it's own. You being surprised by what you decided in no way means the imaginary construct of the character made a choice.

By the way, this isn't a trap argument -- there's not something waiting to go, "aha, now you agree there's no character agency, you have to also accept...." I like looking at where moments of agency exist in rulesets, not because it's a way to make some games better or worse than others, but because the game is where the agency is. If you can point at those moments, you can better understand how to bring any given game system to bear on them in a way that improves the experience. Or find out what doesn't work, and avoid that. Either way, clearly analyzing where and how agency exists can only make playing a given game better. More agency isn't better -- it's different. Players in Blades enjoy more agency in lots of ways, but that comes at the cost of having to do more of the work to make the game play. If you can see that, and know that, then you can get more out of Blades. If you don't, you can still get lucky. Neither outcome is any defense against just not enjoying the game, though, and that's a valid response.

 

[prabe]

No one in this thread has posted actual play examples that would bear on this.

You want play examples? See attached. Yeah. it's 16 pages. When I say my wife takes thorough notes, I am not kidding. Sorry. While there are moments--and this is something of a culmination of an arc--I almost never set them up or otherwise GM for them. The players declared zero facts about the world or the NPCs in this session. I'm curious if you'll think it looks like RPG-as-Puzzle, and what kind/s of agency--and how much--the players had.

I'll answer any questions needed for clarity.

 

[Ovinomancer]

You want play examples? See attached. Yeah. it's 16 pages. When I say my wife takes thorough notes, I am not kidding. Sorry. While there are moments--and this is something of a culmination of an arc--I almost never set them up or otherwise GM for them. The players declared zero facts about the world or the NPCs in this session. I'm curious if you'll think it looks like RPG-as-Puzzle, and what kind/s of agency--and how much--the players had.

I'll answer any questions needed for clarity.

The fiction of play is well captured there, but the mechanics are not. There are a few places where I can see a check probably happened, but nothing about how the game actually ran is there. In other words, there's nowhere to identify moments of agency because that information is missing.

It is clear that your players are well engaged and enjoy your game. That's great!

 

[prabe]

The fiction of play is well captured there, but the mechanics are not. There are a few places where I can see a check probably happened, but nothing about how the game actually ran is there. In other words, there's nowhere to identify moments of agency because that information is missing.

It is clear that your players are well engaged and enjoy your game. That's great!

I am sincerely grateful, almost constantly, for the players I have at the tables I DM for. What may not have come through was the extent to which they keep me engaged.

As to checks where dice were rolled--what I presume you mean by "mechanics" in this context--the ones that are clear to me are the tests to get library cards, and the stealth vs. perception stuff in the House of Masks. Those are probably the ones you saw. It is plausible-shading-to-probable that those were the only dice that hit the table outside of the combat. As I'm pretty sure I've said upthread, I do a lot of autosuccess and a lot of roleplay (and a very little bit of autofail--the example I used was the party trying to get a read on a godlike being). That implies there might have been checks interacting with the Orcphans, with the priest in the temple of The Joyful, at least twice with the Cracked Shields, maybe a check to recognize some spells when described by a non-spellcaster (that might have been rolled, actually, if the players didn't get it), with Black Irnod, maybe with the herbalist. And of course the combat where 5E does get a lot more granular. You can do a lot in 5E without calling for any checks, if you know the characters and you know the world.

 

[pemerton]

Ah, if only it were that easy to get the characters to do what you want them to do, say centuries of novelists. Sure, if you're writing some sort of allegory, you can yank your characters around a fair amount, because they're symbols and not people; anything else, though ... they'll fight you if you try.

This is all metaphor.

I'm trying to have a literal discussion about how RPGing can work.

In the OP example, it is the GM who decides the result of the PC's insult, and also the result of calling for the guards. And those decisions appear to be constrained only by prior, unilateral decisions the GM has made about those various NPCs. The player has no opportunity here to establish fiction, or even to constrain the fiction that the GM establishes.

That may be good; it may be bad. The fact that the OP posted about the session not playing out as well as hoped suggests that, on that occasion at least, it was not perfect.

In the AW play example, it is the player who is able to force the GM to choose whether Isle does what Marie wants, or instead has her brain fried. It is the player who is able to force the GM to decide that Plover is the most dangerous of the NPCs will Mill is a non-violent child of 12. That is a higher degree of agency in respect of the shared fiction. And it is very transparent as to how and why the GM is making those decisions - ie they're forced by the rules of the game.

If, in my BW play example, the Scavenging check had succeeded then it would have been the player who authored the presence of the mace in the tower. Because it failed, it was me the GM who got to author what happened next. And I wrote in the black arrows.

We can see here different systems, which take different approaches to framing checks - BW uses "say 'yes' or roll the dice", AW uses "If you do it, you do it"; 5e D&D uses "if the GM wants to allow a check, s/he can" - and different approaches to establishing consequences - BW uses intent-and-task; AW uses various approaches for various moves, but the examples we've been looking at are force the GM to make a decision; 5e D&D uses GM decides as s/he sees fit in light of previous unilateral unrevealed authorship,

These different approaches produce different degrees of player agency in respect of the shared fiction. And not in some metaphorical sense - in the literal sense of who gets to decide what is the case, and what happens next, in respect of the shared fiction.

I'll try to keep it as system-neutral as I can, and I will explicitly not be talking about published adventures--especially long ones.

The GM establishes fiction by framing the scenario for play. This can stretch all the way up to worldbuilding, but it needn't do so. It usually will include at least one instigating event--which can be a parallel to the "Declare Badness" MC moves in AW. It usually will include at least some facts of the setting, such as an apocalypse, or Elder Gods, or things of that nature, but there are games that are effectively in the here-and-now.

<snip>

The GM is usually responsible for at least most of the NPCs that will appear, and is expected to prepare them (there are differences what "prepare them" means, system-to-system).

The players establish fiction by creating (or generating) the main characters, and by determining those characters' actions. Some (or most) of those actions will only need to be declared; where more is needed--where the outcome is in doubt--there will be some system of resolving that doubt, such as rolling dice or drawing cards. Like the GM, the player is limited by the rules of the game. The player will at least be able to narrate their character's actions if the resolution is in their favor; in some systems they are able to declare outcomes or other facts in the fiction (such as the Crown of Revel being in the box).

This is not system neutral.

There are many systems where the GM is not expected to prepare NPCs. Apocalypse World is one - see eg the actual play example where the GM makes up details about Mill as part of the process of action resolution. Classic Traveller is another - see eg the rules for resolution of Streetwise checks (Book 1, p :15 "The referee should set the throw required to obtain any item specified by the players (for example, the name of an official willing to issue licenses without hassle = 5+, the location of high quality guns at a low price = 9+).")

Your characterisation of the discovery Crown of Revel is not system neutral either. I don't know o any system in which a player can declare such an outcome. (I'm sure there are some; but I don't know them.) In Burning Wheel as in D&D a player can declare "I look in the box for the Crown of Revel". That is not a declaration of "a fact in the fiction". It is a declaration of an action, no different from "I stab the orc". What differentiates the two systems is how the action is resolved. In D&D, as typically played, the GM is expected to decide what follows from that action declaration, canonically at least by reference to his/her notes. In Burning Wheel, assuming that there is table consensus that the Crown might be in the box given the established fiction, the GM is expected either to "say 'yes'" or to set a difficulty for an appropriate check (which could be anything from Scavenging to Box-wise to Crown of Revel-wise depending on the details of the context).

I don't fully understand why you seem so hesitant to address this as a matter of action resolution: I look in the box for the Crown of Revel. If you won't consider that action resolution, and various ways that a system might go about resolving it, you will not be able to understand how Burning Wheel works, or even how Streetwise works in Classic Traveller.

You want play examples? See attached.

<snip>

I'm curious if you'll think it looks like RPG-as-Puzzle, and what kind/s of agency--and how much--the players had.

I read the first page closely and skimmed the next five. It doesn't record anything about the procedures of play, so I can't tell for sure. What follows is conjecture based on your accounts upthread of how you approach RPGs.

My understanding from the list of Dramatis Personae is that the GM was playing the child Turlk and that a player was playing the character Joybell. I therefore conjecture that the player decided what questions Joybell asked Turlk, and that the GM made all the decisions about what Turlk said in response.

Two phrases stood out i particular on that first page: we can’t glean from that where they’re from and we have no way of knowing where their village was. My guess, reinforced by your reply to @Ovinomancer, would be that this ignorance of the relevant elements of the fiction resulted from the GM making unilateral decisions about what Turlk knew and was able to convey.

If my guesses are correct then yes, this looks like RPGing-as-puzzle-solving, and I would say that the GM had almost all the agency in respect of the content of the shared fiction.

This impression is reinforced by a quick look at p 2, where another character who appears to be a NPC controlled by the GM - Jorly - provides information about the Cracked Shield tribe. This then appears to shape the next sequence of play - "We headed off to the Cracked Shields".

Reading on: while it's not clear, I gather that the GM made all the decisions about the compound and the elder called Rask. And decided to provide the players with information about The Masks. The sense of play involving solving puzzles is reinforced by this bit at the bottom of p 4: "We recognized those as Vicious Mockery and Toll the Dead -- which means psychic and necrotic damage. That confirms what Barnett told us about necrotic damage being good against them"

Then, very similar to @Lanefan's hypothetical upthread, we have a description of a street which I assume was all decided by the GM. Thus it would be the GM who established that the street has no place "at all helpful for Fiona and Orryk hanging out for a couple of hours and observing the place."

On page 6 we are told about "one of the most important conversations of Joybell’s life". This all appears to be driven by the GM - eg the idea of "vendetta" which I gather is the crux of it seems to come from a NPC being played unilaterally by the GM.

I didn't read the remaining 10 pages. The consistency of what appeared to be going on in the first 6 pages suggests that they are representative enough.

 

[Numidius]

No I don't think it should be done, unless that's how you guys really love to play. For me and my group it would not only destroy a large portion of the game's mystery, but would drag us kicking and screaming out of immersion, turning D&D into a gamist game. I don't want to play a game of numbers when I roleplay. If I wanted that, I'd play a board game.




I think it sounds like a horrible system(for me) that makes Blades more about being a game than roleplaying the character and immersing yourself in the story. The less often mechanics pull me out of the story the better.



So what. So AC has a mechanic. Without a specific score being given for that dragon, the players are more likely to make a bad decision with the dragon than with the Baron. One AC 16 is beatable. The other AC 22 is not. The players aren't going to know which is which from the description, "The dragon is armored." All of their knowledge of attack bonuses, spells and damage don't matter all that much, since they have no number to compare them to. Hell, that statement could even mean that this dragon wears some sort of barding that makes it even harder to hit.



You're basically arguing that the player not knowing whether they need a 14 or higher to hit or need a natural 20 to hit, gives them an idea of their odds and how things work. That's ridiculous. The odds vary so wildly between those two points that any group that relies on them thinking that they "have an idea of the odds." deserves the TPK that they will eventually walk into.



Fat lot of good that will do the PCs' corpses if they walk into a dragon fight needing natural 20's to hit.


It's even more clearly defined. Instead of the wildly vague and destructive 14 to natural 20 to hit, they have crystal clear knowledge that the Baron is insane and will be very highly likely to have them tortured or killed if they insult him.



This is wrong. With only the knowledge in the OP, I know with crystal clarity to not even attempt intimidation. Trying to intimidate someone who is insane and would react with lethal force to an attempt at intimidation would be stupid, and I'm not stupid. I also know with crystal clarity that deception is pretty risky, but not as risky as intimidation. Someone that insane and touchy about things will probably react poorly to being lied to, but probably not as badly as if I tried to intimidate him. Persuasion would absolutely be the best way to go, IF I even want to risk a conversation with a madman, which I probably don't.



Sure I do. Intimidation = AC 22. I'm very likely to end up dead and take my party down with me. Deception would AC 18ish. Possibly winnable, but still risky. Persuasion would be AC 16. We can win this one, but it's not guaranteed. It's the least risky.

I have a clear enough picture of the Baron to make those assessments and assign AC equivalents to the social skills.


Which is why I'm not going to do something so stupid as to insult a crazy, insult sensitive ruler and risk a reaction with no roll. If the players use their brains even a little bit, it's really easy to avoid auto failures in social situations. That leaves only auto successes and having to roll the dice.

I find your reasoning insightful from your perspective as DM.
I nonetheless think that one strenght of a "Gm decides resolution" is the possibility to indulge in scenes like the OP and have a tense dialogue with a "mad" ruler and the Party on his methods of ruling, with a crescendo of sensations of menace, threat, conflict, without the urge to rely on dice rolls or auto-fails immediately.
Basically, old good roleplay exploring character of an Npc... and PCs, of course, who'd be going to use their Brains both in and out of char, in order to determine difficulty of eventual rolls, while also elaborating an actual first person speech in front of an audience, including the menacing guards of the Baron.

Brains, will, presence, maybe a bit of show off to intimidate the guards first... something like that.

 

[Lanefan]

Been meaning to get back to this for a while...sorry about the delay.

Says whom?

Are you able to understand that there are different ways of playing RPGs?

I'm able to understand that there's different ways of playing different types of games. That said, there's a very real chance I define RPG differently than you do.

Who has agency over the fiction in a play? Not the actors. In the case of a prop. either the director, the producer or the playwright, depending on the detais of the production.

Even if the actors are writing their own lines (or improvising)?

So can you not see that, if a RPG is approached the way that you describe, the players are not exercising agency in respect of the content of the fiction?

Whether or not one has a certain preference, isn't the analysis crystal clear?

But aren't you simply saying here that you prefer a game in which the players do not have agency in respect of certain aspects of the shared fiction? Such as the contents of boxes that their PCs open.

Fundamental disagreement here: determining the heretofore unknown contents of a box is not an agency players get to have in an RPG unless a player's PC put the contents in there in the first place.

A game that gives players that agency has moved away from what I see as an RPG (in which one Plays a Role, that being of your PC) and into shared worldbuilding, which is something very different: a player is no longer simply playing the role of a character in a setting but is also given the responsibility of determining elements of and within that setting, which any player worth his-her salt will very quickly take blatant advantage of.

There's a reason a GM's role includes referee.

Are you really saying that you are unable to comprehend that there are other approaches?

Comprehend? Yes. Accept as being valid? Not so much.

This is a bit like having a conversation about which side of a car the steering wheel is on, and which sort of turn yields an obligtion to give way, and having someone respond to an Australian that, yes, I understand, the steering wheel is on the left and one yields when turning lefft. As if they are literally unable to comprehend that there are parts of the world that have different having different traffic conventions than those that prevail in North America.

Again, difference between comprehension (I've driven numerous times in the UK) and acceptance. I mean, when in Rome do as the Romans do and all that, but from a Canadian standpoint I reserve the right to say it's nuts even as I do it; and don't blame the British at all if they say the same in reverse while over here.

Are literally unable to comperehend that there are approaches to RPGing in which the action resolutoin I look in the box fro the Crown of Revel is determined by a check, with success meaning that the PC finds the Crown in the box when s/he looks, and failure meaning that the GM narrates something different from that which is in some fashion adverse to the PC?

I'm dubious about accepting that as a valid way to roleplay, in that there's no internal setting consistency, no continuity, and therefore nothing to base any long-term in-character thoughts and-or memories on.

On reaching a new valley: I look in the valley for the village of Terynia. Action resolution succeeds and suddenly there's a village there; but for some reason we were never told about it before the trip even though in theory it's been there all along; and had we known or even been able to speculate about its existence sooner we might very well have done things differently.

Also as a system it's broken as hell the minute the players don't severely self-restrain, which IMO they shouldn't have to do.

Given that, in this context, agency over the shared fiction and authority in respect of the shared fiction or authorship of the shared fiction are all synonyms, I don't understand your contrast.

That the latter two are synonyms (in this context) is evident in the fact that author and authority are cognate words. As far as the first is concerned - if the players can't, via the procedures of game play, bring it about that the shared fiction is or contains (say) X rather than (say) Y, they manifestly are not exercising agency in respect of it.

They're not all quite synonymous.

The last two are, but the first - agency over the shared fiction - is not. As a player I have agency over the shared fiction inasmuch as my PC is part of the shared fiction and I as its player get to decide what that character does and in many cases how it does so. But I don't have authorship or agency over the setting or its elements, and nor should I.

Shared fiction = setting + PC actions/words + NPC actions/words. The GM controls the first and third of these; the players collectively the second.

Story = shared fiction + time [+ story, every time this equation is run after the first time]

The contrast emerges pretty clearly in the OP's situation. The player, in that situation, clearly had the power to trigger the GM to reveal the GM's prior conception of the burgomaster - this is in fact exactly what happened when the insult by the PC led the GM to narrate the burgomaster's response. But pretty clearly the player was not exercising agency over the content of the shared fiction: it seems pretty clear that the player wanted the shared fiction to contain a burgomaster who was cowed or chastised or rebuked, or in some other, roughtly similar way put back into his box by the PC's harsh remark. But the player had no chance to bring this about. Which is to say, the player was not exercising agency over the shard fiction.

As I have said to Lanefan, one may or may nor prefer a game in which players exercise this sort of agency. But I am not talking here about preferences.

Well, yes you are, in that oftentimes a great deal of what you post screams your preferences out for all to hear. I merely try to counter this by doing the same, only perhaps making it a bit more obvious that I'm so doing.

 

[Lanefan]

My point is twofold. First, if the GM is deciding what's in the box then the player is not exercising agency over that component of the shared fiction. And obviously is not. Yet there are posters here - @prabe, @Lanefan - who are asserting the contrary.

Of course not. That part of the fiction is not the players' to decide.

The players' collective agency decides whether one or more PCs pay any attention to the box and whether a PC then tries opening the box. Both of these things affect the continuing shared fiction, in that if the players/PCs ignore the box then the Crown is not found and things in the fiction have to proceed on that basis (and important to note: they have to proceed on that basis whether the GM so desires or not; she has to neutrally deal with the curveball of the Crown not being found even if her story ideas expect that it will be).

 

[Lanefan]

In Burning Wheel as in D&D a player can declare "I look in the box for the Crown of Revel". That is not a declaration of "a fact in the fiction". It is a declaration of an action, no different from "I stab the orc".

In fact in both systems those two action declarations are very different from each other.

"I look in the box" and "I stab the orc" are the same, in that each identifies an attempted interaction with an already-known piece of the fiction: the box and the orc are each already present.

However, "I look in the box for the Crown of Revel" tries to bring in another element, that being the Crown, the presence or absence of which remains yet unknown to the players. In my view, saying the words "for the Crown of Revel" does nothing but add flavour; and while adding flavour is always cool in and of itself it has no other relevance.

Put another way, the only mechanically relevant bit of that declaration is "I look in the box", because the box is all that the PC/player knows to be there. On this declaration, then, the GM resolves the action by saying yes if there's no obstacle to opening the box (and then almost certainly narrates what's inside) or rolls the dice if there's some obstacle to opening the box e.g. it's locked or it's not immediately clear just how the box opens.

What differentiates the two systems is how the action is resolved. In D&D, as typically played, the GM is expected to decide what follows from that action declaration, canonically at least by reference to his/her notes. In Burning Wheel, assuming that there is table consensus that the Crown might be in the box given the established fiction, the GM is expected either to "say 'yes'" or to set a difficulty for an appropriate check (which could be anything from Scavenging to Box-wise to Crown of Revel-wise depending on the details of the context).

As everyone seems to agree that part of the GM's job is to set obstacles and challenges (right? We're all agreed on this?), if the GM has no way of knowing where the Crown is ultimately going to be found how can she lay down any obstacles to finding it? How can she set traps, locks, guards, and other assorted challenges if she doesn't know where to put them? How can she set red herrings and misdirections when for all she knows she might in the process unknowingly be steering the PCs right to it?

This is what I mean when I say giving players control of setting elements strongly fights against - or even outright prevents - setting consistency.