D&D 5E Clouds, cubes, and "hitting"

[BookBarbarian]

I do wonder why we keep coming up with such obfuscatory terminology in order to discuss what are really fairly straightforward concepts. Honestly, is anyone in this discussion deriving any benefit from referring to narrative fiction as 'clouds' and game-mechanics as 'cubes', or would we be better off just discussing them in plain English? And god help anyone who joins the discussion six months from now if the terms do gain popularity, and has to trawl through old discussions to puzzle them out.

Bingo. Maybe If I was trying to design an RPG this cubes vs cloud stuff would be useful, but it has no bearing own how I will narrate Hits and HP loss.

 

[Tony Vargas]

This type of RPG is not an exercise in collective story-telling, and there are no mechanics for sharing or negotiating narrative control.

D&D started as a wargame, was eventually recognized as the first RPG, and has been played in many, many styles over the decades. 5e respects that legacy and is meant to be open to being played many different ways. I don't recall Mike Mearls ever excluding collective story-telling.

The player has absolute control over the decisions made by their character, and the GM has absolute control over everything else.

Backgrounds (players can have a hand in designing them) and Bonds/Flaws/etc & Inspiration (DM influencing player decisions for their characters) both seem like exceptions. Probably not the only ones, but presented early and in even the basic pdf, IIRC.

 

[innerdude]

I think there's a distinction to be made, @Saelorn, about the timing and manner of the desired shared narrative control. I don't know about your group, but there's cloud-to-cloud interplay between my group all the time---little details that we come up with in the moment that suddenly just become "true" in the fiction, because the group and GM liked it.

Likewise, I think every RPG group in history has had some of those (not-so-)subtle exchanges between players and the GM like, "So hey, we just spent a lot of time on a boat. We all get a +2 to sailing checks now, right?" *wink wink*, or the more common one for my group, "So, I'm pretty sure we were so awesome last session we get double XP, right?"

And the reason we do this is because in that one-time-in-a-thousand the GM actually agrees, we've now meaningfully changed our character and by extension the fiction. So in your mind, this kind of "reality adjustment" is totally okay; it's the ones that are actually expressed in the game rules / on the character sheet that are problematic?

So what is the dividing line? Is it that cube to cloud should never grant direct narrative control other than through "magic"? Do we accept that altering reality, even for a fictional world, should only arise through something within that fictional world that makes altering reality possible? (I'm genuinely curious about this, by the way; I'm not being rhetorical.)

Something like the following would be very common in a game like Fate, for instance: "You said, Missus GM, that there's been water dripping around and on the ceiling. I'm going to exercise Mechanic X which says my opponent slips on the floor and I get a bonus to attack them."

Whatever these mechanics are, they essentially boil down to "Roll for Reality Adjustment." Is it really the distinction between the player making this call and the GM doing it that's the sticking point?

*Edit: I just re-read the OP, and in looking at the diagram, I think Step 5 is a radical misrepresentation of how much reducing an enemy's hitpoints by 2d6 could potentially affect the fiction. Depending on the circumstances and what was at stake, reducing an enemy's hitpoints by 2d6 could radically reconstruct the entire narrative, even if there was no "Step 6" at all.

 

[The Crimson Binome]

So what is the dividing line? Is it that cube to cloud should never grant direct narrative control other than through "magic"? Do we accept that altering reality, even for a fictional world, should only arise through something within that fictional world that makes altering reality possible? (I'm genuinely curious about this, by the way; I'm not being rhetorical.)

Whatever these mechanics are, they essentially boil down to "Roll for Reality Adjustment." Is it really the distinction between the player making this call and the GM doing it that's the sticking point?

The player controls their own character. The GM controls everything else. That's how an RPG of this type is played. The GM controls all of the NPCs, and is also the one who operates the game mechanics, since they're the only one who knows all of the factors involved.

The player has no ability to alter the reality of the game world, aside from the capabilities of their character. If the player offers advice to the GM about how to run the game or how to mechanically represent a particular in-game reality, then that exists outside of the game, and what happens within the game is still entirely under the control of the GM in their execution of the rules.

If a player did somehow manage to alter the reality of the game world, such as by making suggestions to the GM outside of the game, then it would damage the integrity of the process. The reason why the end narrative matters to the players is because it follows from their in-game choices and the application of the mechanics. Convincing the GM to do things differently, via out-of-game methods, would be like using a cheat code in a video game; it ruins the fun. (At least, it ruins the sort of fun that the game is designed to produce. You might have more fun cheating, if you don't actually enjoy the game as it was designed.)

Something like the following would be very common in a game like Fate, for instance: "You said, Missus GM, that there's been water dripping around and on the ceiling. I'm going to exercise Mechanic X which says my opponent slips on the floor and I get a bonus to attack them."

FATE is not an RPG of the type currently under discussion. FATE has its own arrows, which go left and right and down. It has no bearing on the current topic.

 

[innerdude]

(It also gives rise to a puzzle: what is the point of the players? Why can't their place be taken by some sort of mechanistic decision-making process?)

Ever heard of a little computer game called Football Manager? It's a game where you take the role of a Euroleague football club, but you never once actually control the play "on the field." You're merely the general manager, trying to build the best collection of "on paper" talent, and letting the computer simulate out how your club would fair under various circumstances. Rotissiere baseball, (American) fantasy football, all fall under the purview of this type of game.

The question is, could you construct an RPG around this type of conflict resolution model?

*Edit: [MENTION=6775031]Saelorn[/MENTION], you're correct, the model in the OP is specifically called out as a "classic" RPG model, circa 1990, in other words, D&D 1e/2e. I'll try not to muddy the waters further.

 

[Tony Vargas]

I just re-read the OP, and in looking at the diagram, I think Step 5 is a radical misrepresentation of how much reducing an enemy's hitpoints by 2d6 could potentially affect the fiction. Depending on the circumstances and what was at stake, reducing an enemy's hitpoints by 2d6 could radically reconstruct the entire narrative, even if there was no "Step 6" at all.

Yeah, I was wondering about that one. Like I said, I'm suspicious of the loopbacks, they don't quite make sense. I feel like maybe it's just trying to find fault with 'softer' or more abstract forms of modeling - may have something to do with the whole exercise being a designer's attempt to rationalize the failure of one of his designs..?

 

[pemerton]

I don't buy the 'incidental' bit. Unless he'd also label a DM deciding (based on judgement or arbitrarily) that a given cloud factor (not explicitly covered in the rules) imposes a penalty as an 'incidental' rightward arrow.

These are exaclty the cases being contrasted:

* The rules say I can inflict a penalty on your character, and (incidentally, in the course of doing so) I narrate the fictional cause. A lot of 4e critics see a lot of this in 4e. (I think some of them are wrong, because they're ignoring the keywords, but that's a separate discusssion). In this sort of system, the principal arrow is cube-to-cube ("I made my roll - now you've got to suck up a -2 penalty"). We dont need to invoke any clouds to keep playing. Establishing the fictional explanation for the penalty is incidental, and nothing else follows from it. It hangs there as mere colour.

* The rules say that, if conditions in the fiction are thus-and-so, then a penalty is suffered. In this sort of case, there is no alternative but to consult the fiction and generate a rightward arrow from it.

At the table, the two cases actually unfold differently: in the first, the penalty is established by rolling the dice; in the second, the penalty is established by considering the fiction. These are different things.

It sounds as if the issue is not lack of cube-cloud connections, but rather fear of player latitude in imagining the cloud.

Which is absurd, as the freedom to exercise imagination, much like the freedom of the DM to mod the rules, is necessarily absolute.

No one is afraid of players imagining things.

But - similarly to my response to [MENTION=6775031]Saelorn[/MENTION] upthread - there is a difference between (i) imagining things, and (ii) having those imagings matter to resolution. The former can happen even when playing a non-RPG (I gave some examples upthread). The latter can't.

If the whole game would play out the same whatever the players' imaginings, that does suggest to me that what they're imagining is purely epiphomenal.

This is why stuff like "the ogre hurls its half-eaten cow body at you" doesn't strike me as the acme of RPGing. Because, at that point, we only have colour. Nothing in the clouds has actually mattered to any resolution. And no mechanical resolution has generated anything new in the clouds. It's all just GM patter.

Now, if something had actually happened, in the course of play, that had brought it about that the ogre has a cow that it might throw at the PCs - that is, if the cow-cloud was an outcome of some event of resolution at the table - that would be a different matter altogether. That would be analogous to the weather determination happening then, which results in me suffering a penalty for oppressive heat now. But that was not how the example was presented.

(Why include the above paragraph? The Alexandrian's remark about "skirmish game linked by freeform improv" can be just as apt for a GM-driven colour-rich game as for a skirmish-y game with little or no colour. Or, to put it less contentiously, injecting a lot of GM colour won't make the game "more" of an RPG, if all that colour just froths on the surface but doesn't actually matter to the unfolding play of the game.)

Cube to cube -- No. Example: "Because you're the second person to act in initiative order this round, you have to take a -2 to all actions." "Wait, what? Why?" "Dunno, man, that's just what the rules say. There's absolutely no logical reason in the shared fiction it should be this way, but it just is."

Yet RPGs have heaps of this sort of stuff. Hit points are deducted from running tallies, but there is no particular change in the fiction that correlates to this; there is all the stop-motion and turn-by-turn action of cyclic initiative; I would say that spell memorisation in classic D&D comes pretty close to being an instance of this too - yes, it generates an incidental leftward arrow ("You impressed the spell upon your brain"; "Now the spell is wiped from your brain"), but that fiction is bascially epiphomenal (you can't make an AD&D MU forget a spell by clocking him/her hard on the head, for instance).

Cube-to-cube is not per se inimical to RPGing, in my view, provided that it is experienced as incidental to something that actually generates meaningful fiction. Now, "experienced as incidental" is going to vary from player to player, from session to session, from table to table. It would make perfect sense for the exact same series of exchanges, interactions etc to be experienced by you as "That was a pretty skirmish-y sesision today" and by me, or even by you in a different mood, as "That was awesome - I really felt like I was there!"

This is one reason why I think lecturing people over the internet about whether or not they are really RPGing - something which is oddly popular on these boards! - almost always misfires. If you weren't there and part of it, it's pretty hard to know what was going on, and how all that cube-y stuff related to cloud-y stuff that anyone cares about.

the problem I'm seeing here is that there's a huge undefined region within the cube-to-cloud space as to how the cube "outputs" parse into the cloud "inputs". There's a gigantic, massive leap to go from, "I swing a sword. My dice roll is a 22, which is a hit; my damage roll does 16 HP damage to the Ogre. Is it still standing?" to, "I roll my dice, and because I succeeded at my Control Universal Laws of Reality check, the Ogre is now a banana."

<snip>

So how is the cube-to-cloud interaction model supposed to represent both of these potential resolutions?

Perhaps I've missed something, but isn't this about the game rules, plus how particular actions are framed?

Eg, consider the following episode of play (sblocked for convenience):

[sblock]

It then came to the drow sorcerer's turn. In an email a few days ago the player had told me that he had a plan to seal off the Abyssal rift created by the tearing of the Demonwebs and the killing of Lolth, that relied upon the second law of thermodynamics. Now was the time for him to explain it. It took quite a while at the table (20 minutes? Maybe more? There was a lot of interjection and discussion). Here is the summary version:

* The second law of thermodynamics tells us that time and entropy are correlated: increases in entropy from moment to moment are indicative of the arrow of time;

* Hence, when entropy reaches its maximum state - and so cannot increase - time has stopped;

* Hence, if an effect that would normally last until the end of the encounter could be turned into an effect of ultimate chaos (entropy), time would stop in respect of the effect and it would not come to an end.​

So far, so good, but how is this helping to seal off the Abyss?

* Earlier in the encounter the sorcerer had created a Cloak of Winter Storm which, using an elemental swapping item, was actually a zone of thunder (larger than normal because created while a Huge primordial) that caused shift 1 sq which, through various feat combos, was actually teleportation;

* If this could be extended in size, and converted into a zone of ultimate entropy instead of just a zone of thunder, then it would not come to an end (for the reasons given above);

* Furthermore, anyone who approached it would slow down (as time came to a stop with the increase in entropy) and, if they hit it, be teleported back 1 square;

* As to how a zone of elemental thunder might be converted into a zone of ultimate entropy, that's what a chaos sorcerer is for - especially as, at that time, the Slaad lord of Entropy, Ygorl, was trapped inside the Crystal of Ebon Flame and so control over entropy was arguably unclaimed by any other entity and hence available to be claimed by the sorcerer PC.​

But couldn't someone who wanted to pass through this entropic barrier just teleport from one side to the other?

* On his turn, the sorcerer therefore spent his move action to stand from prone (I can't now remember why he had started the session prone), and used his minor action to activate his Cloud of Darkness - through which only he can see;

* He then readied his standard action to help the invoker/wizard perform the mighty feat of Arcana that would merge the darkness and the zone into a visually and physically impenetrable entropic field, through which nothing could pass unless able to teleport without needing line of sight.​

Unfortunately, the invoker/wizard wasn't ready to help with this plan, and had doubts about its chaotic aspect. On his turn, he instead rescued the paladin and fighter PCs who had become trapped in the Abyssal rift (by casting Tide of the First Storm to wash them back up onto the top of the PCs' Thundercloud Tower).

<snip>

The paladin then used his turn to bodily pick up the drow and carry him into the control circle of the Tower (at the drow's request).

<snip>

The drow's turn then came around. He used his move action to fly the Tower up and out of the two zones (darkness and thunder). He then used a minor action to cast Stretch Spell - as written, a range-boosting effect but it seemed fitting, in spirit, to try to extend and compress zones to create a barrier of ultimate, impenetrable entropy. And then he got ready to make his Arcana check as a standard action.

Now INT is pretty much a dump stat for everyone in the party but the invoker/wizard. In the case of the sorcerer it is 12 - so with training and level, he has an Arcana bonus of +20. So when I stated that the DC was 41, it looked a bit challenging. (It was always going to be a Hard check - if any confirmation was needed, the Rules Compendium suggests that manipulating the energies of a magical phenomenon is a Hard Arcana improvisation.)

So he started looking around for bonuses. As a chaos mage, he asked whether he could burn healing surges for a bonus on the roll - giving of his very essence. I thought that sounded reasonable, and so allowed 4 surges for +8. Unfortunately he had only 2 surges left, so the other half of the bonus had to come from taking damage equal to his bloodied value - which was OK, as he was currently unbloodied.

He scraped another +2 from somewhere (I can't remember now), brining the roll needed down to 11. The dice was rolled - and came up 18! So he succeeded in converting his zones of darkness and thunder into a compressed, extended, physically and visually impenetrable entropic barrier, in which time doesn't pass (and hence the effects don't end), sealing off the Abyss at its 66th layer.

The unfortunate side effect, as was clarified between me (as GM) and the player before the action was declared, was that - as the effects never end - so he can never recharge his Cloak of Darkness encounter power or his Cloak of the Winter Storm daily.

A modest price to pay for cementing the defeat of Lolth and sealing off the bottom of the Abyss from the rest of creation.

[/sblock]We know that the successful die roll (a cube-event) results in the Abyss being sealed (a cloud-event) because that is the action the player declared for his PC.

The cloud is an input into this resolution (eg it establishes the Hard DC; it establishes that the character can manipulate chaos/entropy and create a zone of impenetrable darkness; it establishes that, by giving of his chaotic essence, the character can increase the degree of entropy). The cubes including - setting the DC from the DCs-by-level table; noting that each surge is a +2 (this guideline is found in the 4e DMG 2); expending the Stretch Spell ability; and of course the rolling of the die. The cloud that results from the success also feeds back into the cubes - the character loses a certain ability, because it never ends and so can never be recharged. (In combat, throwing a non-returning weapon involves a similar fiction-driven depletion of an ability.)

I don't see that this sort of case is any different from the attacking case, except that 0 for legacy reasons - typical RPGs tend to have more canonical rules for combat and more handwavey rules/guidelines for other sorts of stuff.

 

[pemerton]

I just re-read the OP, and in looking at the diagram, I think Step 5 is a radical misrepresentation of how much reducing an enemy's hitpoints by 2d6 could potentially affect the fiction. Depending on the circumstances and what was at stake, reducing an enemy's hitpoints by 2d6 could radically reconstruct the entire narrative

It's clear that Baker is not claiming that nothing ever follows from a hit point reduction - as is shown by step 6 (a hit for 8+ is a knockback).

If hit points drop below zero, then the hit might be a kill.

The only point is that, in general, a change in hit points does not in-and-of-iteslf require any change in the fiction.

This can be seen by considering the ubiquitous act of D&D play:

"Roll to hit."
"16 after mods - is that enough?"
"Yep - how much damage"
"12 points - does that kill it?"
"Nope, it's still standing. Who's next?"​

Nothing about any of those exchanges demands moving from cloud to cube. And that doesn't change even if everyone piles on the colour:

"Roll to hit."
"I lunge and take a mighty swing - 16 after mods - is that enough?"
"Yep - how much damage"
"It really is a might swing - max damage, 12 points - does that kill it?"
"With a might lunge, your blade sinks deep into its flesh. But it's not enought to kill it - it's still standing. Who's next?"​

The extra colour doesn't actually change anything fundamental about the relationship between cubes and clouds. (And to allude back to my previous post: this is why the ogre throwing the cow bits isn't - in my view - some breakthrough in RPGing. After the GM's initial narration for colour, the actual events of play can unfold exactly as above.)

Contrast, say, the casting of a classic Trasmute Rock to Mud spell. That can't be done without engaging the fiction - first, you have to identify which rock you are transmuting, then it's behaviour has to be inferred from the fiction (eg on a ceiling/roof, it will fall; on a slope, it will flow; etc), and then the fate of those in it has to be inferred from the fiction too (eg, per the spell description, heavy armour > sinking and suffocating).

Yeah, I was wondering about that one. Like I said, I'm suspicious of the loopbacks, they don't quite make sense.

See above. If you can move from mechanical state to mechanical state without need to go via the fiction, then you're moving from cube-to-cube.

This is why the higher ground is different - that requires engaging with the fiction. (Unless the "higher ground" has become completely "mechanise" as a spot of advantageous ground on the battle map - this is a sort of limit case to which I think some 4e might be prone, as I think I noted upthread.)

I feel like maybe it's just trying to find fault with 'softer' or more abstract forms of modeling

Which exercise? And what fault?

Vincent Baker is using scenario 1 to explain and motivate the anlaytical framework - it's not until scenario 3 (which I didn't copy and paste, as it concerns a system quite different from D&D, that we get to the game - In a Wicked Age - that he wants to analyse.

And no one is finding fault in anything. Lots of RPGs will have cube-to-cube events. That's not a flaw in a game, unless it is only cubes-to-cubes.

may have something to do with the whole exercise being a designer's attempt to rationalize the failure of one of his designs..?

No one's trying to "rationalise" anything. He's a designer. Who made a game that, in retrospect, doesn't work. He's trying to work out what went wrong, so he compares it to a trad game (which doesn't have the issue) and to DitV (his most successful game, which also doesn't have the issue).

When a poster asks "Also, you said that those who have the most fun are those who insert these rightward arrows from elsewhere. If that is the most fun to play, why not include these arrows inside the game's mandatory structure?", Baker replies, "Pretty much because I designed the game before I figured it out."

He's not "rationalising". He's analysing and diagnosing, so he can avoid the same mistake in future designs. I think that's part of being a successful designer.

I do wonder why we keep coming up with such obfuscatory terminology in order to discuss what are really fairly straightforward concepts. Honestly, is anyone in this discussion deriving any benefit from referring to narrative fiction as 'clouds' and game-mechanics as 'cubes', or would we be better off just discussing them in plain English?

The "clouds" and "cubes" come from the fact that it is a picture/diagram: clouds are a (not-unintuitive) symbol for imagined stuff, and cubes for real-world stuff like dice.

I think the diagrams are helpful. The fact that, as an upshot, "clouds" gets used instead of "shared fiction" is tolerable.

 

[Tony Vargas]

These are exaclty the cases being contrasted

I see no case of a step being labeled cloud-to-cloud because whether to impose a modifier was left up to the DM. Latitude shouldn't be equated with non-existence, IMHO.

No one is afraid of players imagining things.

Afraid, resistant, whatever... labeling a mechanic that explicitly gives players latitude in how they imagine the fiction 'cube to cube' strikes me as off.

If the whole game would play out the same whatever the players' imaginings, that does suggest to me that what they're imagining is purely epiphomenal.

Ah, but it wouldn't play out the same, because those imaginings are different. It might use the same cubes, but rolling against a DC of 15 with a bonus of +6 is using the same cubes whether it's a diplomacy check, attack roll, or save.

The Alexandrian's remark about "skirmish game linked by freeform improv" can be just as apt for a GM-driven colour-rich game as for a skirmish-y game with little or no colour...

So just as wrong, either way. Sure.

Yet RPGs have heaps of this sort of stuff. Hit points are deducted from running tallies, but there is no particular change in the fiction that correlates to this; there is all the stop-motion and turn-by-turn action of cyclic initiative; I would say that spell memorisation in classic D&D comes pretty close to being an instance of this too - yes, it generates an incidental leftward arrow ("You impressed the spell upon your brain"; "Now the spell is wiped from your brain"), but that fiction is bascially epiphomenal (you can't make an AD&D MU forget a spell by clocking him/her hard on the head, for instance).

Obliviax (1e MM2) could suck the spell out of his head, though, if that helps. ;P Seriously, though, memorization was kinda awful because it was so contrary to genre, but, it impacted the 'cloud' it generated, however repulsive the miasma of that cloud may have seemed.

Cube-to-cube is not per se inimical to RPGing, in my view

I should hope not. You've got a fairly knowledgeable RPG designer dreaming up a model to describe RPGs, one would hope he didn't create a model inimical to them.

provided that it is experienced as incidental to something that actually generates meaningful fiction. Now, "experienced as incidental" is going to vary from player to player, from session to session, from table to table.

So why bother with such a proviso?

And, really, I'm still not convinced cube-cube is legit. Take the first example of it, marking off hps. As abstract as D&D combat has always been (and clearly, 2e AD&D is the example game in question), drawing a line between the hit and the damage roll like that is just odd. Concluding that one has a cloud effect and the other doesn't is even stranger.

This is one reason why I think lecturing people over the internet about whether or not they are really RPGing - something which is oddly popular on these boards! - almost always misfires. If you weren't there and part of it, it's pretty hard to know what was going on, and how all that cube-y stuff related to cloud-y stuff that anyone cares about.

You're far to magnanimous.

And no one is finding fault in anything. Lots of RPGs will have cube-to-cube events. That's not a flaw in a game, unless it is only cubes-to-cubes.

Well, that'd be a fault, then, wouldn't it? Fault found.

Though, honestly, I'm still not convinced by either cloud-cloud or cube-cube. Both strike me as abnegation. Cloud-cloud means the game has rejected the situation - it declines to address it, so there's no mechanical resolution. Cube-cube means the cloud is simply being ignored by the players (including DM).

No one's trying to "rationalise" anything. He's a designer. Who made a game that, in retrospect, doesn't work. He's trying to work out what went wrong

OK, sounds like 'rationalizing failure' may have been a tad harsh on my part. Analyzing failure, then.

cubes for real-world stuff like dice.

But also for abstractions like hps, which I'm also not too sure about...

 

[pemerton]

in D&D or any other game which follows the general model under discussion <snippage> the character's skill in swordfighting simply represents the likelihood that the character will succeed when swordfighting. The number on the sheet directly represents the literal skill of the character at wielding a sword.

How much any given player cares about each area of the story is not a factor at any point during the game. During the pre-game (character building), the players declare what sorts of characters they want to play during the game, and they do their best to accurately represent those concepts given the constraints of the system. Wanting to play a character who is an amazing swordfighter is not the same as wanting a lot of narrative control over what happens during swordfights, though. It's not even saying that they want a character who often wins at swordfighting. It is simply saying that they want a character who is good at swordfighting

<snip>

You could even have a player who wants to play a Worf-type character - incredibly strong, yet frequently overcome in shows of direct strength.

I think it's a bit disingenuous to say that someone who builds a character to be good at sword fighting doesn't want narrative control during sword fights. And OF COURSE they want a character who "wins" at swordfighting---because that's their avenue to narrative control. A character who doesn't win at sword fighting in an RPG doesn't generally get to control much of anything in the fiction.

There are always corner cases - the player who builds the 18-STR plate-wearing fighter who hopes nevertheless to lose many fights.

But this whole discussion of "narrative control" is, in my view, a detour. It's using a term - "narrative control" - whose meaning is (to me, at least) unclear, in part because it's being used as a normatively-loaded way of distinguishing categories of game.

My point was more banal: if you build a PC with big numbers in swordfighting, then - everything else being equal - when you declare actions pertaining to sword fighting, the fiction is more likely to end up as you desired it (eg if your action is "I attack so-and-so hoping to kill him/her", your wish is more likely to come true if your numbers are bigger).

I gather it's contentious to describe an action declaration as "a wish pertaining to the content of the shared fiction". But if my PC is wearing gauntlets of ogre power and wielding a two-handed sword, and I declare the action "I attack the goblin", and yet I don't wish for the shared fiction to include the gobling being dead at the hands of my rather formidable PC, then something looks like it's gone wrong (eg my action declaration seems rather irrational).