Alternatives to map-and-key

[soviet]

I think there are rules at play in how I assess the fictional state. If the players say they want to sneak across the grounds, I set the DC at 10, and they roll high enough, then I can't decide that they have not made their way across the grounds.

The guard patrol changes shift, make another stealth roll.
The ground here is more gravelly, make another stealth roll.
You're near the inner palace now, make another stealth roll.
You're climbing now? Make another stealth roll.

Likewise if in their infiltration they take a clever set of actions and roll well, the rules of the game (& the world) dictate that they have achieved their result. I narrate that result, but I do not decide it.

There are no specific rules in D&D for 'infiltrating a compound'. There are separate rules for stealth, climbing, perception, etc that rely on the DM to put them together in what seems to them to be an appropriate way. That's a decision.

In the context of the overall challenge, I am never thinking "their goal is to bypass the outer defenses. Have they done that yet?" I think "what is the DC for stealth? Did they succeed? What is the climb DC? What are the consequences? Did they succeed?"

These are exactly the same thing.

I am adjudicating the world as an observer, not deciding as a participant.

The world doesn't exist to be observed. It's all made up. You are making a decision as to which parts of that made-up world you give weight to, and which additional fiction you do or do not introduce.

 

[Gimby]

The guard patrol changes shift, make another stealth roll.
The ground here is more gravelly, make another stealth roll.
You're near the inner palace now, make another stealth roll.
You're climbing now? Make another stealth roll.

Right - I had an example from another thread about this

Let's say we're playing 5e and the rule is that characters can move stealthily, with the DC being determined by the passive perception of any relevant NPCs.

Consider a PC attempting to sneak up on an NPC that is standing facing away from them at the end of a long alleyway that will take 3 move actions to pass through. The alleyway is described as partially obscured with debris with two plausible obstacles in it. Given the passive perception of the NPC and stealth bonus of the PC, the PC has an 80% chance of making any specific roll.

What is the probability of the PC reaching the NPC undetected?

There are a number of ways we could approach this, but I hope everyone would agree that three reasonable ones would be

1) This can be resolved as a single roll - the fictional state is highly constrained, does not change dramatically over the length of the alleyway and the cover described is sufficient to allow a stealthy approach. Probability of success - 80%

2) The obstacles are the most important part of this - they will need to be navigated by the PC and represent possible risks. The GM calls for two rolls, one to pass each plausible obstacle. Due to iterated probability, our chance of success is now 64%

3) The distance is the most important part - it's three move actions, so will require three check to get through - one for each action. Three rolls, and our probability drops to 51.2%

There are of course lots of other ways to approach this, but let's just consider these for now.

All three approaches are justifiable, utilise the stealth rules and respect the established fiction, but we see that the final probability depends in great part on which the GM choses. And the difference is pretty stark - it's the difference (roughly) in 5e between a Dex 10 untrained character and a Dex 18 4th level stealth trained character.

The question basically becomes - how far through a fictional challenge does a single successful check get you?

 

[pemerton]

my play often gets described as map and key

By whom? I mean, I know very little about your play. I don't recall ever reading an actual play thread started by you.

This thread isn't about classifying anyone's play. It identifies a method, one that I believe is fairly well known, and asks about alternative ways of establishing latent scenes/situations, and of progressing scenes/situations in play.

 

[Crimson Longinus]

By whom? I mean, I know very little about your play. I don't recall ever reading an actual play thread started by you.

This thread isn't about classifying anyone's play. It identifies a method, one that I believe is fairly well known, and asks about alternative ways of establishing latent scenes/situations, and of progressing scenes/situations in play.

Given that you said you meant literal map and key, whole tread is rather weird, given that most roleplaying do not involve such. Instead it just involves situations that are described, rather than presented on a map.

 

[pemerton]

It is from the DMG1. The challenges have set of primary skills, and there is no indication that the GM should prevent the players from using those, thus the implication is such that they should frame fiction so that they can always be used.

The DMG says the following (pp 72, 74-6:

More so than perhaps any other kind of encounter, a skill challenge is defined by its context in an adventure. Adventurers can fight a group of five foulspawn in just about any 8th- to 10th-level adventure, but a skill challenge that requires the PCs to unmask the doppelganger in the baron’s court is directly related to the particular adventure and campaign it’s set in. . . .

Begin by describing the situation and defining the challenge. Running the challenge itself is not all that different from running a combat encounter (see Chapter 3). You describe the environment, listen to the players’ responses, let them make their skill checks, and narrate the results. . . .

Thinking players are engaged players. In skill challenges, players will come up with uses for skills that you didn’t expect to play a role. Try not to say no. . . . However, it’s particularly important to make sure these checks are grounded in actions that make sense in the adventure and the situation. . . .

Skill challenges require the players to make rolls at specific times. Call for these checks according to the pace of the narrative and the nature of the challenge.​

The plurals in the second quoted paragraph are ambiguous: they admit of a collective read (all the players proved their responses and roll their checks together, and then the GM narrates the result) or a distributed one (each player, in some appropriate sequences, provides their response to the GM's description, and rolls their check, and then the GM narrates the result, and this provides the description that prompts the next declared action). Given that the distributed reading leads to a good play experience, whereas the collective reading leads to a bad play experience, I think it is the obviously preferable interpretation. The distributed reading is also more consistent with the comparison to combat - where players take turns, and the GM narrates outcomes of each action as prompts for and/or inputs into the next one - and with the statements (i) about checks being grounded in actions that make sense in the situation, and (ii) about checks being called for in accordance with pace and the nature of the challenge.

This seemed pretty evident to me when I first read these rules, and my actual play confirmed it.

So you then agree with me that in skill challenges no solution is better than other and what the players do do not matter? So it is just a mechanical framework to prompt fiction generation?

No I don't. And I'm not sure why you would think otherwise.

Solving a puzzle is not the only way something might matter. The choices the players make during a skill challenge establishes fiction. Given that establishing fiction is why I play RPGs, it could hardly matter more!

Here are two actual examples, from 4e play, of skill challenges:

The starting point
The PCs are low paragon - a dwarf fighter/warpriest of Moradin, a paladin of the Raven Queen, a wizard/invoker, a drow chaos sorcerer/demonskin adept, and a ranger-cleric of the Raven Queen. The player of the ranger-cleric was absent from the session.

The scenario combines elements of Thunderspire Labyringth (a 4e module), Heathen (from a 2008 online Dragon magazine), Speaker in Dreams (a 3E module from WotC) and Night's Dark Terror (a B/X module from TSR), plus some other elements of my own.

The PCs have recently entered a town which is under increasing pressure from hobgoblin and allied raiders. The town is ruled by a Patriarch of Bahamut and a Baron. The PCs are still getting the lay of the political land.

The PCs entered the town as heroes, having saved an affiliated village from being destroyed by hobgoblins. They were lauded by the Patriarch, and invited to join the Baron for dinner that evening.

<snip some play that preceded the skill challenge but provided important information to the players>

Two revelations had the biggest immediate impact. One involved the PCs' principal enemy. This is the leader of the hobgoblins, a powerful wizard called Paldemar (but called Golthar in Goblinish). The PCs learned that in the town he is not known to be a villain, but is apparently well-thought of, is an important scholar and astrologer, is an advisor to the Baron, and is engaged to the Baron's niece. The PCs (and the players) became worried that he might be at dinner that evening. This was a worry for two reasons - (i) they didn't really want to fight him, and (ii) they know some secrets about an ancient minotaur kingdom that he does not, but has been trying to discover. One of those secrets involves a magic tapestry that the PCs carry around with them (because they don't have anywhere safe to leave it).

The second revelation was that the Baron was prophesied to die that night. The paladin had already sensed a catoblepas in the swamps outside the town, and had sensed it approaching the town earlier that day. The priestess explained that a year ago the Baron had been visited by a catoblepas, as a type of forewarning. And the cultist explained that the uprising had taken place today in anticipation of the Baron's imminent demise.

After learning these things, the PCs cleaned up in the cultists' bathroom and then hurried off to dinner.

The dinner
The PCs arrived late, and were the last ones there. On the high table they could see the Baron, and his sister and brother-in-law, and also Paldemar, their wizard enemy. They left their more gratuitous weapons - a halberd for the dwarf and a longbow for the ranger - with the dwarf's herald - an NPC dwarf minion called Gutboy Barrelhouse - and took their seats at the high table. Gutboy was also carrying the backpack with the tapestry.

The PCs also noticed a series of portraits hanging behind the high table. One had a young woman, who was the spitting image of a wizard's apprentice they had recently freed from a trapping mirror - except that adventure had happened 100 years in the past (under a time displacement ritual), and this painting was clearly newly painted. Another, older, painting was of a couple, a man resembling the Baron, and a woman resembling the rescued apprentice but at an older age.

About this time the players started talking about the skill checks they wanted to make, and I asked them what they were hoping to achieve. Their main goal was to get through the evening without upsetting the baron, without getting into a fight with Paldemar (which meant, at a minimum, not outing him as the leader of the hobgoblin raiders), and without revealing any secrets to him. In particular, they didn't want him to learn that they had found the tapestry, and that it was in fact 15' away from him in Gutboy's backpack. But it also quickly became clear that they wanted to learn about the people in the portraits, to try and learn what had happened over the past 100 years to the apprentice they freed, and how she related to the Baron's family.

This whole scene was resolved as a complexity 5 skill challenge. It ran for more than an hour, but probably not more than two. The general pattern involved - Paldemar asking the PCs about their exploits; either the paladin or the sorcerer using Bluff to defuse the question and/or evade revealing various secrets they didn't want Paldemar to know; either the paladin or the wizard then using Diplomacy to try to change the topic of conversation to something else - including the Baron's family history; and Paldemar dragging things back onto the PCs exploits and discoveries over the course of their adventures.

Following advice given by LostSoul on these boards back in the early days of 4e, my general approach to running the skill challenge was to keep pouring on the pressure, so as to give the players a reason to have their PCs do things. And one particular point of pressure was the dwarf fighter/cleric - in two senses. In story terms, he was the natural focus of the Baron's attention, because the PCs had been presenting him as their leader upon entering the town, and subsequently. And the Baron was treating him as, in effect, a noble peer, "Lord Derrik of the Dwarfholm to the East". And in mechanical terms, he has no training in social skills and a CHA of 10, so putting the pressure on him forced the players to work out how they would save the situation, and stop the Baron inadvertantly, or Paldemar deliberately, leading Derrik into saying or denying something that would give away secrets. (Up until the climax of the challenge, the only skill check that Derriks' player made in contribution to the challenge was an Athletics check - at one point the Baron described himself as a man of action rather than ideas, and Derrik agreed - I let his player make an Athletics check - a very easy check for him with a +15 bonus - to make the fact of agreement contribute mechanically to the party's success in dealing with the situation.)

Besides the standard skill checks, other strategies were used to defuse the tension at various points. About half way through, the sorcerer - feigning drunkenness with his +20 Bluff bonus - announced "Derrik, it's time to take a piss" - and then led Derrik off to the privy, and then up onto the balcony with the minstrel, so that Paldemar couldn't keep goading and trying to ensnare him. At another point, when the conversation turned to how one might fight a gelatinous cube (Paldemar having explained that he had failed in exploring one particular minotaur ruin because of some cubes, and the PCs not wanting to reveal that they had explored that same ruin after beating the cubes) the sorcerer gave an impromptu demonstration by using Bedevilling Burst to knock over the servants carrying in the jellies for desert. (I as GM had mentioned that desert was being brought in. It was the player who suggested that it should probably include jellies.) That he cast Bedevilling Burst he kept secret (another Bluff check). But he loudly made the point that jellies can be squashed at least as easily as anything else.

While fresh jellies were prepared, Derrik left the table to give a demonstration of how one might fight oozes using a halberd and fancy footwork. But he then had to return to the table for desert.

Around this time, the challenge had evolved to a point where one final roll was needed, and 2 failures had been accrued. Paldemar, once again, was badgering Derrik to try to learn the secrets of the minotaur ruins that he was sure the PCs knew. And the player of Derrik was becoming more and more frustrated with the whole situation, declaring (not speaking in character, but speaking from the perspective of his PC) "I'm sick of putting up with this. I want Paldemar to come clean."

The Baron said to Derrik, "The whole evening, Lord Derrik, it has seemed to me that you are burdened by something. Will you not speak to me?" Derrik got out of his seat and went over to the Baron, knelt beside him, and whispered to him, telling him that out of decorum he would not name anyone, but there was someone close to the Baron who was not what he seemed, and was in fact a villainous leader of the hobgoblin raiders. The Baron asked how he knew this, and Derrik replied that he had seen him flying out of goblin strongholds on his flying carpet. The Baron asked him if he would swear this in Moradin's name. Derrik replied "I swear". At which point the Baron rose from the table and went upstairs to brood on the balcony, near the minstrel.

With one check still needed to resolve the situation, I had Paldemar turn to Derrik once again, saying "You must have said something very serious, to so upset the Baron." Derrik's player was talking to the other players, and trying to decide what to do. He clearly wanted to fight. I asked him whether he really wanted to provoke Paldemar into attacking him. He said that he did. So he had Derrik reply to Paldemar, 'Yes, I did, Golthar". And made an Intimidate check. Which failed by one. So the skill challenge was over, but a failure - I described Paldemar/Golthar standing up, pickup up his staff from where it leaned against the wall behind him, and walking towards the door.

Now we use a houserule (perhaps, in light of DMG2, not so much a houserule as a precisification of a suggestion in that book) that a PC can spend an action point to make a secondary check to give another PC a +2 bonus, or a reroll, to a failed check. The player of the wizard PC spent an action point, and called out "Golthar, have you fixed the tear yet in your robe?" - this was a reference to the fact that the PCs had, on a much earlier occasion, found a bit of the hem of Paldemar's robe that had torn off in the ruins when he had had to flee the gelatinous cubes. I can't remember now whether I asked for an Intimidate check, or decided that this was an automatic +2 bonus for Derrik - but in any event, it turned the failure into a success. We ended the session by noting down everyone's location on the map of the Baron's great hall, and making initiative rolls. Next session will begin with the fight against Paldemar (which may or may not evolve into a fight with a catoblepas also - the players are a bit anxious that it may do so).

This is the most sophisticated skill challenge I've run to date, in terms of the subtlety of the framing, the degree of back and forth (two major PCs with whom the PCs were interacting, with different stakes in the interaction with each of them), my concentration on evolving the scene to reflect the skill checks and the other action while still keeping up the pressure on the players (and on their PCs), and the goals of the players, which started out a little uncertain and somewhat mixed, but ended up being almost the opposite of what they were going into the challenge.

There is no puzzle-solving here. There is no "finish line" that the players are trying to get to. There is an open-ended social situation, which in the earlier parts of the skill challenge is rendered more precise (from the players' point of view) so that, as the challenge comes to its end, they can seize control of it: as it turns out, by provoking Paldemar into conflict by goading him into revealing his true colours and thus, at the same time, retaining the goodwill of the Baron. They also achieved secondary goals, like keeping their secrets from Paldemar and learning more about the painting and the Baron's niece.

The PCs erected a magic circle around the Mausoleum of the Raven Queen, in order to prevent anyone from entering it and potentially learning her true name (backstory here); then rested; then scried on the tarrasque, which they knew to have recently begun marauding in the mortal world, identifying its location and noting that it was being observed by maruts. They decided that, to return to the mortal world to confront the tarrasque they would first teleport to their abandoned Thundercloud Tower, and then take that with them through another conjured portal and fly it to where the tarrasque is.

<snip>

The player of the eternal defender had already noted that, when I read out the description of maruts and their contracts earlier in the session, the only being actually mentioned by name was the Raven Queen. So he predicted (more-or-less in line with what I had in mind), that the maruts observing the tarrasque would be there at the behest of the Raven Queen (who is served by three of the five PCs), to stop it being interfered with.

When the PCs then took their Tower to confront the tarrasque, that was indeed what they found. Upon arriving at the tarrasque's location they found the tarrasque being warded by a group of maruts who explained that, in accordance with a contract made with the Raven Queen millenia ago, they were there to ensure the realisation of the end times, and to stop anyone interfering with the tarrasque as an engine of this destruction and a herald of the beginning of the end times and the arrival of the Dusk War.

(Why the Raven Queen wants the Dusk War has not fully come to light, other than that it seems part of her plan to realise her own ultimate godhood. One idea I had follows in sblocks.)

[sblock]With Ometh dead, it seems possible that those souls who have passed over the Bridge that May be Traversed But Once might be able to return - repopulating a world remade following the Dusk War and the restoration of the Lattice of Heaven.[/sblock]

I wasn't sure exactly what the players would do here. They could try and fight the maruts, obviously, but I thought the Raven Queen devotees might be hesitant to do so. I had envisaged that the PCs might try to persuade them that the contract was invalid in some way - and this idea was mentioned at the table, together with the related idea of the various exarchs of the Raven Queen in the party trying to lay down the law. In particular I had thought that the paladin of the Raven Queen, who is a Marshall of Letherna (in effect, one of the Raven Queen's most powerful servants), might try to exercise his authority to annual or vary the contract in some fashion.

But instead the argument developed along different lines. What the players did was to persuade the maruts that the time for fulfillment of their contract had not yet arisen, because this visitation of the tarrasque was not yet a sign of the Dusk War. (Mechanically, these were social skill checks, history and religions checks, etc, in a skill challenge to persuade the maruts.)

The player of the Eternal Defender PC made only one action in this skill challenge - explaining that it was not the end times, because he was there to defeat the tarrasque (and got another successful intimidate check, after spending an action point to reroll his initial fail) - before launching himself from the flying tower onto the tarrasque and proceeding to whittle away around 600 of its hit points over two rounds. (There were also two successful out-of-turn attacks from the ranger and the paladin, who were spending their on-turn actions in negotiating with the maruts.)

The invoker/wizard was able to point to this PC's successful solo-ing of the tarrasque as evidence that the tarrasque, at least on this occasion, could not be the harbinger of the end times whom the maruts were contracted to protect, because it clearly lacked the capacity to ravage the world. The maruts agreed with this point - clearly they had misunderstood the timing of celestial events - and the PCs therefore had carte blanche to finish of the tarrasque. (Mechanically, this was the final success in the skill challenge: the player rolled Insight to see what final argument would sway the maruts, knowing that only one success was needed. He succeeded. I invited him to then state the relevant argument.)

That was the end of that session. Due to a conference commitment that I had for June 18, our next session was four weeks later (three weeks ago tomorrow). It was a short session, and it encompassed two events. First, the PCs defeated the tarrasque. With no need anymore to worry about the maruts, and hence focusing all their attacks upon the monster, it lasted maybe three more rounds. The sorcerer/bard unleashed a Climactic Chord, which always tends to bring things to a resolution (just as it says on the tin); and there were multiple crits from the cleric/ranger, a crit from the invoker/wizard, and I think a crit also from the fighter. Another thing I remember was that, for his Climactic Chord attack, the invoker/wizard charged the tarrasque and attacked it in melee with his Rod of Seven Parts (this may even have been the crit).

In this instance of play, the choices made by the players - about how to deal with the maruts, and how to relate that to their fighting of the tarrasque - obviously mattered. The players goal, as their PCs, is to stave off the Dusk War: and here, they do that in part by helping to show that the time has not yet come. Other choices would obviously have yielded different possible fictions, and different outcomes to the skill challenge.

 

[Crimson Longinus]

An analogy might be combat. We can use the basic to-hit rules to throw a solitary punch at someone, or fire an arrow at a target. In principle you could run a whole combat like this, using to-hit and damage rules to resolve individual attacks with everything else like who goes first and who moves where subject to pure GM adjudication. Some RPGs already work like this. But we don't, we have a whole combat sequence, with initiative rolls, action economy, movement rates, maybe a grid and miniatures, etc. Why? Because that structure creates a transparency of stakes and positioning that makes things more fun and (with the hit point rules) creates a definitive end point for when the battle is over.

Combat rules are structured rigidly to allow a tactical mini game. I would not want all of the game to work like this. Though of course the combat minigame has far more tactical depth than skill challenges, if combat rules were just "the side which first lands six hits wins" I doubt a lot of people would care for it. And of course you can "bypass" some or all of the combat rules by various ways if fictional positioning and the action taken warrant such, you can kick enemies into chasms, intimidate them to surrender, collapse structures on them etc.

 

[FrogReaver]

This thread isn't about classifying anyone's play. It identifies a method, one that I believe is fairly well known, and asks about alternative ways of establishing latent scenes/situations, and of progressing scenes/situations in play.

In the very first line of your OP you go much further than identifying a fairly well known method, you say most play defaults on an assumption of map and key play.

So seriously…?

A lot of thinking and discussion about the play of RPGs seems to default to an assumption of map-and-key play.

 

[pemerton]

The notion that there is some finish line and progress bar eludes me. The constraints on the DMs narration are the fiction, including unrevealed backstory if any and the unspoken guiding principle that the possible narrations should be trimmed down to only the most plausible ones (assuming they are within the same order of magnitude of likeliness. Sometimes this will leave a single narration, and others it will leave a roll to decide.

Then we repeat this process again and again. It can produce progress, setback, instant success, instant failure, or even ‘mostly’ status quo.

Skills Challenges force the fiction to be updated without regards to the above, by constraining the possible path lengths, the normal flow is broken, such that we can look at a skill challenge and see that if the ‘default method noted above’ had been used for resolution instead then the goal in the skill challenge would have had a chance of being achieved with fewer checks, each more responsive to the fiction (plausibility and likelihood of both initial lead up and outcome properly considered). Thus the crux of the criticism.

Here are some examples; they frame the situation as one of overcoming an obstacle or resolving an obstacle:

People keep saying that skill challenges are about fiction, but to me it seems obvious that this is not true. If the actions taken regarding fictional positioning are such that they would overcome the obstacle with few rolls this does not matter

It's about whether the player's action declarations make a significant difference in resolving the obstacle or not. In the skill challenge case, they will always need N checks, so clever solutions or approaches the GM hadn't thought of don't mean anything. As long as an action is minimally acceptable, one success, one tick.

The 4e DMG (p 72) uses different terminology ("goal" and "obstacles"):

What’s the goal of the challenge? Where does the challenge take place? Who is involved in this challenge? Is it a stand-alone skill challenge or a skill challenge as part of a combat encounter?

Define the goal of the challenge and what obstacles the characters face to accomplish that goal. The goal has everything to do with the overall story of the adventure. . . .

It’s not a skill challenge every time you call for a skill check. When an obstacle takes only one roll to resolve, it’s not a challenge. One Diplomacy check to haggle with the merchant, one Athletics check to climb out of the pit trap, one Religion check to figure out whose sacred tome contains the parable - none of these constitutes a skill challenge.​

I don't want to put too much weight on mere terminology. But I do think it is significant that the DMG sees skill challenges in terms of overcoming obstacles - introduced by the GM, via their narration of framing and of consequences - to realise a goal. If the skill challenge succeeds, no more obstacles are to be presented by the GM; if they are still there in the fiction, the PCs overcome them.

 

[FrogReaver]

The above is very close to the standard Narrativist model for situational play but I want to emphasises a few things.

I’ve gotten that vibe before about my play as well, but it’s usually quickly dismissed

One big difference is that you almost always want to start with, at the very least, an implied connection between the agendas of the player characters and the NPC's. In other words, having an adventuring party roll into town tends to produce weak sauce. I mean it can work as long as everyone is genuinely and actually committed to nothing happening. No saying one thing 'we have the agency to not bite hooks' and yet doing another 'well there would be no game it we didn't bite, It would be rude to the GM to waste their work, and so on.' For anyone serious about this style I'd actually kill the whole idea of adventures, parties and plot hooks. Think character centric drama instead.

Why though? If the premise is this band of adventurers travel together and we are finding out about their deeds then surely adventures and parties are part of the situation and characters that should be honored. I mean dungeon world tends narrativist and it’s intended to evoke typical d&d style situations.

Also on ‘plot hooks’, I’m not sure if introducing situations that may speak to the characters/parties is any different than a narrativist GM introducing situations speaking to core character personality traits, save that those personality traits are generally not strictly codified in d&d. Maybe a difference of ‘may’ vs ‘does’?

What makes this different to play that may 'seem' similar, is that we really mean it when we say the situation is the basic unit. The GM isn't introducing new stuff for purposes of pacing or to try and resolve threads.

Really? Because I think I’ve been told at least indirectly the exact opposite. The whole point is to push the conflict to a head. ‘Rising conflict…’ etc.

To do so is to put their hand on the scales and essentially give themselves story control. As @pemerton mentioned earlier, the atomic components are the various characters. Therefore be very wary when introducing new characters.

Is that any different than putting your hand on the scale by forcing situations speaking to specific character traits and then escalating them again and again till the situation comes to a climax?

 

[Crimson Longinus]

I don't want to put too much weight on mere terminology. But I do think it is significant that the DMG sees skill challenges in terms of overcoming obstacles - introduced by the GM, via their narration of framing and of consequences - to realise a goal. If the skill challenge succeeds, no more obstacles are to be presented by the GM; if they are still there in the fiction, the PCs overcome them.

But that is not what the rules actually do. Rules do not care about obstacles, they care about the number of successful checks. Thus if the players manage to overcome the obstacles presented in the fiction with few checks, new obstacles must be presented so that the check quota can be fulfilled. Similarly reaching the required number of failures must result the goal of challenge failing, regardless of whether that would causally follow from the actions taken. In either case the fiction is moulded to follow the mechanics, not the other way around, lessening the impact of engaging with the fiction.