new player/GM exploring the system

Yeah, I'd have had both of them at the bottom of the shaft too.  It's just funnier that way.

Yeah, I got a laugh out of that one.

Side note, is there any way to account for task difficulty? For example, scaling a dangerous cliff vs. scaling a dangerous cliff in the middle of a blizzard without proper equipment. Do you just throw a modifier into the mix?

DW discourages adding situational difficulty modifiers to the basic Roll+Stat. If another PC is Aiding or Interfering, there's a +1 or -2, and a number of conditions or setup moves give a +1 forward (i.e. modifier to the next roll) or +1 ongoing (i.e. modifier lasting until a specified condition), but there's no task difficulty modifier in the RAW.

So instead of saying "this is super hard, take a -3", what the GM can do is say "you can't do it at all, unless you do X first" or "you start to do it, then thing Y happens that you have to deal with before you can finish it," or what Munin said, making the consequence for failure much nastier in the harder case. Maybe the GM says "there's no way you can climb this cliff face without any equipment" and someone else has to Spout Lore to think of a way to improvise some gear or Discern Realities to find an easier way up. Maybe the GM says "sure, roll Defy Danger with STR to start making your way up... okay, good, now, you get about halfway and then start to panic, roll Defy Danger with WIS to see if you can keep it together..."

but there's no task difficulty modifier in the RAW

What's RAW?

Maybe the GM says "there's no way you can climb this cliff face without any equipment" and someone else has to Spout Lore to think of a way to improvise some gear or Discern Realities to find an easier way up.

I thought about that, but that could/would actually swing the odds more than a flat modifier, because then you have to succeed at both checks. Let's say you have a situation to which you figure a -1 penalty would be appropriate. No matter the skill level, a second check is going to be more punishing than the penalty. And, consider these three situations for example:

1) An average strength fighter is trying to wrestle a weak goblin to the ground.
2) An average strength fighter is trying to wrestle an equal strength fighter to the ground.
3) An average strength fighter is trying to wrestle a larger, strong troll or whatever to the ground.

Even if you throw in a second check to compensate for the added difficulty for the troll, you're still looking at the same odds for situations #1 and #2. So, lets say you say that for the weak goblin, you treat a 7-9 as a 10+. Well, then you have the problem of the results either being very good, or very bad, and no in between.

Don't get me wrong, I'm not saying we need to go full sim here (god forbid). And, sure, the subjective categories for success/failure gives you some wiggle room, but I'm not sure it covers the full scope of potential environmental variables.

What's RAW?

Gamer shorthand for "rules as written."

It is absolutely true that calling for another check is fundamentally identical to making the task harder by adding a modifier.  The trick is to make each roll consequential.  Don't say, "OK, you can scale the cliff, but you'll have to Defy Danger using DEX three times to do it," because while it might be more realistic it's not at all interesting. If you want to use that kind of an approach (multiple rolls to make the complete task harder), make sure that you vary what it is that you're asking the players to roll for.  In other words, break the task down into smaller pieces.

So for instance, say you're climbing that icy cliff without proper gear in a blizzard.  And say you get a partial success on your first roll.  So in this case, maybe you get a worse outcome - instead of the options presented in my last post, one of the things the GM could say is that you only get partway up the cliff.  You might be tempted to say, "OK, I'll roll again to keep climbing," but see above under uninteresting.  It's important for the GM to structure the narration/fiction such that simply "trying again" isn't obvious or maybe even not possible.

In this case of a partial on roll+DEX, maybe the GM says, "you make it about two-thirds of the way up, catching your breath as you come to rest on a narrow ledge.  But as you move to continue your climb, a big chunk of icy rock comes away in your hand, precipitating a bit of a rockfall.  You come through it OK, no more than a few cuts and scrapes, but that safe route you'd planned from the ground is now gone.  You'll have to find another.  What do  you do?"  At this point, the player has some options.  Maybe he chooses to Discern Realities to find a new way up.  And if he gets a 10+, cool, maybe he finds a safe way up to the top.  The GM has a choice - maybe the success in discerning reality can stand in for the rest of the climb and the player makes it to the top, or maybe it reveals some other way that might take another roll (e.g. the player spots another way, but it will force him to drive a dagger into the rock with his bare hands and use it as an attachment point to swing over to a better route - he can roll+STR to Defy Danger, taking the +1 forward from the success he got discerning reality).

Or maybe the player says, "you know what, screw all this crazy mountaineering business.  I'm going to roll+loyalty with my hireling Fafnir to get him to climb the rest of the way to the top and lower me a rope."  And even if the player is fully successful with this roll, the GM is perfectly within his rights to look at the fictional situation and say, "OK.  Even over the howling wind you can hear Fafnir grumbling, but he sets off as you command.  But it's gonna take him a while, and you are completely exposed to the elements up here on this tiny ledge.  Roll+CON to keep from freezing your ass off while you wait."

That's how you make a situation more difficult or dangerous.

But in the abstract, the question you should always be asking yourself when structuring these situations is "what is this roll accomplishing in the story?"  Think about the obstacle posed by the cliff and what it means to the flow of the tale you and your players are trying to tell.  If the cliff is just something you threw in to keep this from being a total cakewalk, or because it just fits the situation that the Wicked Sorceror's tower might be sited on a cliff, maybe a single roll is sufficient.  But if these are the oft mentioned and heavily foreshadowed "Cliffs of Insanity," then maybe besting them is an important plot point and a more in-depth task/conflict resolution is in order.

Does this help?

Does this help?

Somewhat, but feel free to take a crack at the wrestling hypothetical I posted above. :)

I get the whole what's-it-matter-to-the-story thing, but there's a bit of an inconsistency here that I'm having trouble negotiating. On one hand, I've read plenty about how this isn't a sim system. But on the other hand, it actually is. The aspects you're testing aren't your character's propensity for trouble, bad luck, drama magnetism, etc. It's Strength, Dexterity, Charisma, etc. The character breaks down into some pretty sim elements. So, the character with Strength +3 is stronger than the character with Strength -1. When I'm doing a Strength test, it's because the task at hand requires Strength. Nevermind comparing wrestling a dragon to wrestling a puppy. These are both pretty common sense situations. But, if I'm wrestling a GM char that's a little stronger than myself vs. one that is a little weaker, according to the RAW(:D), I have the same odds against both. It's a GM char, so there's no opposed check. My roll doesn't get a modifier. You might have this multifaceted narrative potential in front of you, but the player only has a single port of entry, their own simulated and static measure of general competence.

The multiple check thing is a bad idea, not because of the narrative constraints, but because of the math. For the sake of simple mathematics, let's say we have a character who is average at everything that needs to be tested for the check and has no modifiers from anything else. For a single check, that character has a 58% chance of success. If it's a little complicated and they need a follow up check, then their odds drop to 34%, add a third check and we're down to 20%. If you did this with flat modifiers for each difficulty, your odds would go 58%, 42%, and 28%, respectively. It's less harsh and gives you more play to account for the external environment with which the characters interact.

I think the idea of the narratively-driven follow up checks is clever, especially if you have to tap into different moves to pull off a combo. But the math really bites you in the ass. There's actually a pretty easy fix for this. Add in a "combo bonus". All these skills working in synergy give you +1 to your follow up checks until the larger task is done. So average (+0) for the first check, and +1 for checks 2, 3, and so on. There your odds go 58%, 42%, and 30%. You're actually getting pretty close to the mechanical effect of the flat modifier, but you're using the multi-check narrative approach. The only odd thing about it is that your narrative approach takes three times the dice rolling as just giving someone the -2 to the first roll. And, you're still modifying rolls.  So, really, I haven't accomplished anything here. :P

But in the abstract, the question you should always be asking yourself when structuring these situations is "what is this roll accomplishing in the story?"

I ask (and usually answer) that question all the time. But I'm also asking, "What does this roll say about the world in which my character exists?" Every dice roll isn't just a beat in the story. It's also the in-game universe telling you, "I know what you want, but this is what you are going to get, because that's the way the world works."

Quote

Does this help?

Somewhat, but feel free to take a crack at the wrestling hypothetical I posted above. :)

Sure.  The important bit is to focus not on the wrestling itself, but rather on the context in which it takes place.  Wrestling a goblin to the ground might be relatively easy.  Or not, because the goblin might have the slippery little bastard custom move, which means that attempts to hit him use roll+DEX instead of roll+STR.  And the troll might have crazy-long rubbery arms, meaning that the character needs to defy danger to get close enough to inflict damage.

Furthermore, the important point is what happens on a success.  If you get a 10+ on the goblin, yeah, he's a 38-pound weakling and you're going to be able to keep him pinned for as long as you like.  The guy with equal strength?  Maybe you can keep him down, maybe you can't.  But you have "leverage" on him now should you wish to parlay: "Tap out, man.  I don't want to break anything you're gonna need."  The troll?  Even on a full success, you're not out of the woods.  He's probably strong enough to stand up with you hanging onto him, so now that he's thrashing around the dungeon crashing into walls and columns trying to knock you off his back, roll+CON to defy danger to maintain your hold and keep inflicting choking damage on him.

I get the whole what's-it-matter-to-the-story thing, but there's a bit of an inconsistency here that I'm having trouble negotiating. On one hand, I've read plenty about how this isn't a sim system. But on the other hand, it actually is. The aspects you're testing aren't your character's propensity for trouble, bad luck, drama magnetism, etc. It's Strength, Dexterity, Charisma, etc. The character breaks down into some pretty sim elements.

Yeah, that's actually why I like Apocalypse World better.  Cool, Hard, Hot, Sharp, and Weird are much more anchored in the drama and less in the simulation.

But, if I'm wrestling a GM char that's a little stronger than myself vs. one that is a little weaker, according to the RAW(:D), I have the same odds against both. It's a GM char, so there's no opposed check. My roll doesn't get a modifier. You might have this multifaceted narrative potential in front of you, but the player only has a single port of entry, their own simulated and static measure of general competence.

And this is where custom moves come in.  Wrestling a puppy is almost certainly going to be easier than wrestling a dragon because the puppy isn't fucking terrifying, forcing you to defy danger with roll+WIS if you want to do anything other than soil your armor or run the hell away.

I think the idea of the narratively-driven follow up checks is clever, especially if you have to tap into different moves to pull off a combo. But the math really bites you in the ass. There's actually a pretty easy fix for this. Add in a "combo bonus".

And that is exactly why acting on information gleaned from discern reality lets you take +1 forward into your next roll.

Let's be clear - it's not that the game has no modifiers, it's just that it doesn't have very many.  There's a great combat example that Vincent wrote many moons ago in which the consequence of a character's failure on one move was a -2 on a follow-up move - he was effectively interfering with himself.

All these skills working in synergy give you +1 to your follow up checks until the larger task is done. So average (+0) for the first check, and +1 for checks 2, 3, and so on. There your odds go 58%, 42%, and 30%. You're actually getting pretty close to the mechanical effect of the flat modifier, but you're using the multi-check narrative approach. The only odd thing about it is that your narrative approach takes three times the dice rolling as just giving someone the -2 to the first roll. And, you're still modifying rolls.  So, really, I haven't accomplished anything here. :P

Which is exactly why you should be asking yourself what this roll means.  Is it there to heighten the drama?  Is it there to punish the players for attempting something stupid or crazy?  Or is it there to offer an opportunity with some associated cost?  And most importantly, do the consequences of either success or failure substantially alter the course of the story?  Completing a difficult or costly task gives a sense of accomplishment, which is fuel for the furnace of character development, which in turn is why many of us play these games in the first place.  It's all about giving the players an enjoyable and memorable experience.

Additionally, this totally glosses over the other GM-tweakable knob, which is the follow-up move.  If a character fails a roll, the GM can make as hard and direct a move as he likes.  If you fail while climbing a cliff, I don't have to inflict damage on you.  I can do something better - I can separate you from the rest of your party.  Or I can take away your stuff.  Or I can capture you, which is to say that perhaps even on a failure you might make it to the top, with the GM narrating: "You scrape and claw and there are a few close calls there, but eventually you make it to the top of the cliff. But just as you're about to pull yourself over the ledge, someone sticks the blade of a spear in your face and says, 'Thor's balls, you climb slowly. We didn't think you were ever going to make it to the top. Haul him up and hog-tie him, boys. And he didn't fall, so you owe me 20 silver, Haemish.'"

Generally, the more momentous or important or dangerous the roll, the more dire the consequences are, even if you succeed.  That is where the in-game universe lives.  That is the difference between failing the roll to wrestle the puppy and succeeding to wrestle the dragon. The world works a certain way, by following the fiction and the GM principles, not by adding modifiers to the rolls.  In some sense it is completely arbitrary, and you need to trust that the GM is not abusing his agenda.  But in some sense, it's no different than the arbitrariness of whether the modifier for any given situation is a -1 or a -3 or a +5.  I'd rather fail a roll and suffer the consequences because the story demands it than because the dice demand it.

Don't get me wrong, I'm not saying we need to go full sim here (god forbid). And, sure, the subjective categories for success/failure gives you some wiggle room, but I'm not sure it covers the full scope of potential environmental variables.

It's really not trying to do that.

You wrestle the Orc. We really don't need to figure out how difficult the orc is to wrestle. If you make your roll, it wasn't too difficult. If you blow the roll, the GM makes a move from the list of moves.

If you really want a super difficult to wrestle orc. You can just give him an instinct like "grapple with extreme ferocity and aggressiveness." Then when the wrestling starts you can do something like:

GM: The orc bellows and circles warily. What do you do?
Player: I'm reach in a grab him with a wrestling hold.
GM: As you close with the orc, your wrist is caught in his vice-like grip. This guy is stronger and faster than he looks (reveal an unwelcome truth). What do you do?

It's all fictional positioning. You don't need modifiers. I'd probably excuse myself from a DW game in which the GM was messing around with modifiers the way you suggest.

Yeah, that's actually why I like Apocalypse World better.  Cool, Hard, Hot, Sharp, and Weird are much more anchored in the drama and less in the simulation.

That makes more sense. The system should measure what it's supposed to measure.

Wrestling a goblin to the ground might be relatively easy.

But it might not, and that's what I'm trying to get at. So, say you have a surgery skill/check, and an appendectomy is your middle ground. You can run multiple checks for brain surgery, but what if you're just stitching up a bad knife wound?

Of course, this example is invalid because the system is supposed to be used not for the task at hand (despite that being what it's measuring), but for the narration of the task at hand.

The world works a certain way, by following the fiction and the GM principles

The latter and former are two very different and usually unrelated elements.  In terms of the fiction, it doesn't even matter how the world works, the fiction is going to happen regardless. However, by using in-game abilities to check for metagame elements, you can easily end up with a stream of fiction that can only be held together with non-sequitur.

This is what I'm starting to think.

Quote

Yeah, that's actually why I like Apocalypse World better.  Cool, Hard, Hot, Sharp, and Weird are much more anchored in the drama and less in the simulation.

That makes more sense. The system should measure what it's supposed to measure.

The classic six stats are being used because they're iconic. It's a nod to classic D&D, rather than an attempt to measure with any rigor.

The latter and former are two very different and usually unrelated elements.  In terms of the fiction, it doesn't even matter how the world works, the fiction is going to happen regardless. However, by using in-game abilities to check for metagame elements, you can easily end up with a stream of fiction that can only be held together with non-sequitur.

This is what I'm starting to think.

Have you read the game yet? From your OP it sounded like you had just heard about it. When Munin says you follow the GM Principles, do you know what he means?

Quote

Wrestling a goblin to the ground might be relatively easy.

But it might not, and that's what I'm trying to get at.

Exactly, and that is why in my previous post I said:
"Wrestling a goblin to the ground might be relatively easy.  Or not, because the goblin might have the slippery little bastard custom move."

So, say you have a surgery skill/check, and an appendectomy is your middle ground.

No.  No, no, no.  That's not how it works.

You can run multiple checks for brain surgery, but what if you're just stitching up a bad knife wound?

Of course, this example is invalid because the system is supposed to be used not for the task at hand (despite that being what it's measuring), but for the narration of the task at hand.

Exactly.  Quit trying to simulate the difficulty of things directly.  If something is easy (i.e. stitching up a minor knife wound), don't even bother to roll, because whether the knife wound gets stitched or not probably doesn't matter to the overall story.  Roll for things that are consequential, and contextualize the results of that roll based on the fictional position of the characters in question.

Quote

The world works a certain way, by following the fiction and the GM principles

The latter and former are two very different and usually unrelated elements.  In terms of the fiction, it doesn't even matter how the world works, the fiction is going to happen regardless. However, by using in-game abilities to check for metagame elements, you can easily end up with a stream of fiction that can only be held together with non-sequitur.

The point is that in the PbtA engine, they are not different and unrelated elements.  They are the same element.  If the fiction dictates that a task is very difficult or dangerous, then the consequences for failure (or even for success) are more dire than for a task that is easy.  If the fiction dictates that dragons have iron-hard scales a foot thick, then no amount of success on a simple Hack & Slash is going to produce damage.  That is "how the world works."  The GM then uses the principles to convey this to the players clearly and truthfully (say by revealing and unwlecome truth: "As the dragon rises into the sky, you watch in horror as it shakes off the arrows and quarrels from the militiamen like water from a duck's back.  Even the ballista atop the tower is unable to pierce its metallic scales...").

Dice rolls are both descriptive and prescriptive.  That means they not only tell you something about what has happened, but something about what is going to happen.  They serve to guide the story, not model the physics of the gameworld.

They serve to guide the story, not model the physics of the gameworld.

So, the system, whose elements are modeled on the physics of the game world (Strength, Hit Points, Load), aren't actually used to model the physics of the game world.

You wrestle the Orc. We really don't need to figure out how difficult the orc is to wrestle. If you make your roll, it wasn't too difficult. If you blow the roll, the GM makes a move from the list of moves.

This is key.

Kneller, think of it this way. Say you simulate every factor that might be involved in the wrestling roll: less than an hour since you last ate a meal, so your hands are slippery, that's a -2%. Recent rainfall, ground is muddy, additional -4%. Orc is Knobblegob clan, so his armor has lots of big rivets you can get a grip on, +7%, and so on and so on.

In this system, every roll you make has a different modifier, some positive, some negative, but over the long term, your success odds even out to include the consistent modifier that applies to every attempt -- the relevant stat modifier.

The DW system just puts all that detail stuff after the roll. The 2D6 roll, as you noted, is pretty swingy, and you're gonna miss sometimes even when you're rolling at +2 and you're gonna hit sometimes even when rolling at -2. When you miss the roll, that's when the GM narrates what factors screwed you up. "You rolled a 5? Okay, your greasy hands slip right off the orc's smooth-worn leather armor as you try to grab him, and he sort of does this judo thing and before you know it you're face-down in the mud with the orc's knee in your back."

Quote

They serve to guide the story, not model the physics of the gameworld.

So, the system, whose elements are modeled on the physics of the game world (Strength, Hit Points, Load), aren't actually used to model the physics of the game world.

Almost but not quite exactly right!

In some sense Strength isn't actually how strong you are, it's how good you are at doing heroic things in which being strong might play a factor.  Remember, there are no skills (in the traditional sense) in Dungeon World, so the relevant stat needs to encapsulate a lot of different ideas in one tiny package.  STR is the attribute that governs the Hack & Slash move, for instance, so it also includes some measure of skill-at-arms.  But you could also use it to Defy Danger when trying to cling to the aforementioned dangerous cliff, so maybe in that instance it's a measure of your strength-to-weight ratio.  Or you could use it to bash down a door, so maybe there's an element of mass to it as well.

So no, it's not modeling the physics of the world at all, or at least not directly.  The fact that it's called "Strength" is just shorthand (and a nod to D&D, the system which inspired DW).  It's modeling elements of the story.

If this seems completely nonsensical and counterintuitive it's because compared to traditional rules systems, it's a pretty massive paradigm shift.  I came to Apocalypse World from a pretty hard-core simulationist background.  I was into gritty realism and modeling situations with the rules and all of that other stuff.  I had house rules to make automatic weapons fire more realistic and to handle all sorts of situations that the basic rules (in whatever system we happened to be using) didn't cover.  I resisted AW for several years

But once I gave a thorough read-through and tried to understand what it was all about, everything clicked.  Now I don't worry about house rules or situational modifiers or modeling physics correctly or whatever.  Because it's not important to the mechanics of the system.  In a PbtA game, if you want gritty realism or cinematic action or epic fantasy, that lives in the fiction of your game, not the rules.  And relating this back to one of your first few posts, that's why you see different GMs handling the effects of partial success differently - they are trying to stay true to the fiction that they and their players have created.  The GM is "saying what the fiction demands" (one of the GM principles) and following the story in a way that a) makes sense given the shared verimilitude of the setting and b) is fun (following one of the other GM principles, which is "make the characters' lives not boring").

I encourage you to read through the sections of DW about playing the game and being a GM very carefully.  Don't approach it as just another rules set and gloss over that stuff.  There are some really important differences in mindset that are easy to miss.  And if you want further insight, I encourage you to read through Apocalypse World (though unfortunately there's no Creative Commons for it), because it is even more clearly different from traditional games.

Quote

They serve to guide the story, not model the physics of the gameworld.

So, the system, whose elements are modeled on the physics of the game world (Strength, Hit Points, Load), aren't actually used to model the physics of the game world.

Here's a Paladin move. What physics is it simulating?

I Am the Law
When you give an NPC an order based on your divine authority, roll+Cha. ?On a 7+, they choose one: • Do what you say
• Back away cautiously, then flee
• Attack you
?On a 10+, you also take +1 forward against them. ?On a miss, they do as they please and you take -1 forward against them.

Notice the GM is free to choose to comply, run, or attack even if you roll a 10. If you roll a 6, the GM can choose one of their moves like deal damage or a monster move like attack. What are your odds of being attacked?