In conversations on other forums about Dungeon World, I often see Defy Danger discussed as a poorly-written move, mostly for its ability to be used with any stat and thus exploited by players (see: the Wizard who dodges an arrow with +INT by determining its trajectory using math or whatever). I’ve seen people hate on this move to the point that it puts them off DW entirely. (I suspect that’s some typical internet hyperbole, but who knows.) In my play, I rarely come across it being problematic from my players; instead, I will fall into the trap of using it when there is a more applicable move instead (often Defend) that I forgot to consider. Is Defy Danger, and moves like it that use multiple stats, a badly designed move? Or are the problems associated with it issues that should be addressed fictionally? Or both?
Is Defy Danger a bad move?
I actually appreciate how letting Defy Danger be rolled with any stat acknowledges the broadness of the move-- it’s a move you see in original Apocalypse World and many other PbtA games being rolled with +Cool or some equivalent, but to me that creates a weird incentive for pretty much any character to take whatever their main stat should be, and then also have a high Cool stat, since it lets you use the “basically do anything that isn’t a more specific move,” move.
I think the obvious thing with the wizard dodging an arrow with +INT is just for everyone at the table to have a certain respect for the fiction and not try to explain their rolls in ways that everyone at the table can pretty intuitively tell feels like a stretch. Now, whether a system being so easily made or broken by its players buying into it or not is “bad” design or not, I dunno, but as you said, it doesn’t seem like something that’s usually all that problematic in actual play.
As for the general topic of moves in PbtA that let you roll multiple stats, I think it can be pretty sleek design, actually. If you think of it compared to like OSR or more “traditional” systems, they basically do the same thing but obscure that somewhat? For example, at least in newer additions of D&D, you can basically attack with any stat you want, it’s just that you have to flip to different parts of the book and call on different sub-rules to figure that out-- attack rolls are made with +STR, unless they’re using finesse or ranged weapons, which is +DEX, or casting a spell, which can be +INT/WIS/CHA, etc. A move in something like Dungeon World that just said like “when you attack something, say how you do it and roll with the relevant stat” wouldn’t be all that different.
I don’t think it’s a bad move per se, but it is probably one of my least favourite ones in any PBTA game, because it does indeed tend to suck up a very wide range of actions and challenges and pour them into a single mould, a mould that I find (a) a bit boring, and (b) lacking in the helpful prompts given in more specific, specialised moves. As a GM I like to have helpful prompts to steer me towards the sort of outcomes a move should generate; and as a player I like to have some idea of what might go wrong if I roll badly. The 7-9 results on DD just don’t help at all with either of these; they more-or-less leave the group on their own to just come up with some not-failure result.
I’m less fussed about the fact you can roll different stats, I don’t see that as a problem. But it is important to me, for the integrity of the game, that you can’t simply bend over backwards to crowbar any stat into any move. Can the wizard use their intelligence to calculate the trajectory of the arrow? Hell no! They have to move out of the way of the arrow, and that’s the method they’re using to evade it. If you’re pausing to calculate trajectory, you’re dead. So the key here is to describe what you’re actually doing, and if that thing doesn’t actually help you with the action you need to take to defy the danger, you don’t get to roll the relevant stat.
So, there is definitely a design approach that takes an unfavorable view on it (let’s see if this summons Sid 
).
But in practice, I’m fine with it? I mean as a player you don’t want to have to roll Defy Danger as the 7-9 result is basically “yeah you do it but on the condition that the GM also gets to make a move (out of order and one that’s bad for you)”.
I mean, the entirety of WoDu is build on it and lots of people enjoy that.
So, players are incentivized to do try to avoid it, they want to roll anything else, every other move is better.
And nothing much breaks if a player really wants to roll an unusual stat for a task. I’d make it a point in the fiction to explore that, actually?
I always see Defy Danger as a reaction the players choose to a MC soft or hard move.
Most other moves is the players deciding to do something, either in response to the MC or of their own idea.
Even if they enter combat intending to dodge, as soon as they help someone else its either protect or aid.
But Defy Danger replaces the function of all saving throws in D&D, so it will come up alot.
The MC can say no, using Int with Defy Danger to work out how long until the archers are in range is different to using Int to dodge the arrow that’s mid flight.
Just for fun.
1d6 Ways A Wizard Dodges An Arrow With +INT
- “I know that type of crossbow always pulls to the left.”
- "I read its harder to hit a moving target."
- "Wait, did I pack my travel copy of Xeniluos Insights or my research copy? Let me check my pack." whoosh “What was that?!”
- "Just like we practiced, don’t wait to hear the snap of the bow before you dodge. Sound is slower than an arrow."
- "Fighter, stand in between me and these archers."
- summons a blast of wind to knock the arrows away
I’d interpret 2 as Dex, 3 as luck so no bonus/minus, and 5 as Charisma.
6 is a casting roll, if its not then you’re doing magic without the in genre cost of the casting move.
But 1 and 4 sure.
I play with a variant of Defy Danger based on Make a Saving Throw from FotF. Despite it allowing various justifications for players who want to use their best stat, often, the situation won’t allow this.
Here it is (note that we play with 2d12 instead of 2d6)
MAKE A SAVING THROW
When you act or react in the face of danger , roll…
+str if you use sheer might…
+int if you think fast, focus, or remember…
+wis if you use intuition or willpower…
+dex if you use speed, agility, or a delicate touch…
+con if you resist or endure…
+cha if you charm, command or impress.
24+: You do it! Take or pass +1 forward.
20+: You do it as well as one could hope.
13–19: You do it, but there’s a consequence… Roll 1d12:
…1–2 Out of the frying pan… Roll a GM move.
…3 Wounded! You take damage (half, 1d4+1, etc.)
…4 Handicap! Burn 1d2 of the ability you saved against.
…5 Turmoil! You take -1 forward.
…6 Cost! You must expend resources (your choice).
…7 Peril! A follower needs your help! (an ally if you have none, or re-roll)
…8 Consequences! You now face a new challenge.
…9 Cruel twist : You must re-roll this Saving Throw at -1.
…10 Setback! The current situation worsens. Say how.
…11 Display! Role-play a downside to your class, alignment, traits, backstory etc.
…12 Embarassment! How utterly shameful or ridiculous. If you’re playing in turns, lose a turn while you compose yourself.
3–12: Mark XP. If the danger you fail to defy has its own consequences, suffer them now. If not, roll a GM move.
2 or lower: As above, but expect the very worst!
This gave me a lot to think about! I’ll share what I think doesn’t work well about Defy Danger.
I think PbtA moves in general do three things well:
- Focus the game on a theme.
- Introduce unexpected outcomes.
- Maintain the momentum of a dramatic situation.
Apocalypse World’s basic moves are all about getting what you want from other people. There aren’t moves to scramble up sheer cliffs, or traverse a pit of spikes. Even if you attack someone, it’s either to get someone to do what you want or to take something from them. The game doesn’t need to say what it’s about. The specific triggers and outcomes continually bring you back to what the characters want from others and how they get it. Act Under Fire is the most vague, but I think it works because it is always about whether you flinch or hesitate when trying to make those other moves. Because DW doesn’t share the same tight focus, Defy Danger becomes a catch-all for a wide range of actions that don’t always drive you toward something interesting.
The moves in Monsterhearts are phenomenal at producing unexpected outcomes. When you try to lash out physically, one of the consequences is “They learn something about your true nature and gain a String on you”. This pushes play in a direction that can be surprising to all players. You can make a character with a straightforward agenda and the moves force you to address the push and pull of social influence that surrounds them and the nature of their inner life. The outcomes are never just success / partial success / failure, but a means to introduce new tension between characters. Keep Your Cool is the most similar move, but it requires a player to tell the MC what their character fears, and a 10+ provides insight on a situation (rather than resolving it), so this can’t absorb the actions of other moves.
If a character fails at Defy Danger, the GM often introduces a tough choice, which leads to a new charged situation. But to me it’s often not clear when to lead with a move (like introducing a tough choice), prompt the players to roll Defy Danger, or call for a more specific move. Can Defy Danger remove a threat completely? If a player misses another move, can they Defy Danger to avoid the consequences? Can they do that with Defy Danger, too? Sometimes this flow works, but I find it less intuitive than in games with a tighter thematic focus. I’d contrast this with Attempt Something Risky in World of Dungeons (even though the moves are similar) because this is the only resolution system in the game. The flow in WoD is always: follow the fiction until they clearly attempt something risky, then roll. Compared to other PbtAs, play is not driven by moves in WoD, but that works because the move takes minimal time to resolve, and always takes you back to the fiction rather than to another move. This makes WoD (imo) similar to a lot of OSR games, in that the fictional details of the environment drive play, rather than the moves. Having more moves would make running WoD more cumbersome without improving it.
To me it seems like Defy Danger is a necessary move to maintain the genre of play that DW aims for (light, character-focused D&Dish adventures) but it doesn’t drive play forward in the way I’ve come to expect from a solid PbtA move.
I am going to piggyback on what @Michael said: I think (like much of DW) Defy Danger is there to follow the “genre” of D&D. (Because of we are being honest, that is the genre DW is emulating - not fantasy fiction - D&D.) But it actually, in my opinion, is going a step further than genre emulation. It is trying to emulate certain feels and rhythms of D&D - specifically roll-for-anything-where-there-is-a-chance-of-failure.
And that’s okay. It’s not my brand of vodka, but it was a design choice they made. It is a fair way outside of some people’s ideas of “what pbta is” (something I’m honestly tired of people claiming exists) but it was intentional I think.
I think though that it is instructional that World of Dungeons has that as its only move. It means you can use it to do anything and everything, which is why I think Defy Danger feels like a dissonant chord ringing out amongst many pbta peeps. People generally feel moves are meant to be specific and Defy Danger is anything but…
I think this and what @Michael said are good points and made me think of a perspective I’ve been trying to keep in general when I look at how games are designed-- You can usually separate a question like this into two different questions: 1. Does this do what it intends to do well? and 2. Do I like what this intends to do?
So as @shane points out, you can ask, is Defy Danger a good move on DW’s terms, which is to say, a move that fits within a game meant to feel like D&D? Rather than “is it a move that feels right in some broader sense of how PbtA as some broader genre should play.”
And then, even if it does, you might not like the idea of a PbtA game that emulates the D&D experience (which is of course totally fair, I’m not a huge fan myself), but that’s different than it being a “bad” move, or bad game, for that matter.
Defy Danger might be a poorly written move only because the trigger for it is so fictionally vague. It’s the classic “when do I call for a roll?” question. It can feel clunky on how to execute because it’s a vague move in order to catch most if not all rules exceptions.
In other words, a partial reason why it’s so vague is because it’s the catch-all move in the game. If dungeon world were an onion, deft danger is the center. Its kind of the ur-move that the whole game orbits around.
Take a look at World of Dungeons if you haven’t yet. That entire game relies solely on defy danger.
So, I’ve found there’s generally people in PbtA that like vague (or suggestive triggers) and people who like extremely prescriptive triggers and specified outcomes.
I myself favour suggestive move triggers and sometimes even outcomes as well. But that is just my preference. Fans of PbtA often conflate their preferences with what is “good” or “bad” but really it’s just how you like to play. Unless you’re applying a design to a critical framework you’ve built, I’m not going to be taking your opinion (regardless of how informed it is) on if something is designed well or not because you like it or not.
I get into a bad habit of doing this as well and am trying to rejigger my lexicon to get rid of that mentality myself.
All of this is just to say, I myself like defy danger just fine. Though I really like future versions of the move in World of Dungeons (and Blades in the Dark), Veil 2020, and The Between in which there’s a built in safety tool more explicit than defy danger. Telling the player the consequence as well as what the danger is and getting on the same page, asking if they want to proceed, and then firing off the move to see what happens.
Ooh, this is something missing from my list of “What PbtA moves do well” above: set clear expectations for success and failure before the roll. I prefer moves that do this for sure.
I’ve started and deleted posts here a couple of times, because I’m definitely one of those players who, as Fraser says, prefers prescriptive (or at the very least descriptive) triggers, and I’ve also not played a lot of DW. (Mostly by choice). But Michael and Shane really nail the issues I have with the move. In particular the idea of “D&D emulation”, and the fact that when you can call it is very fuzzy. The “roll to see if it works” is one of the things I find least interesting about D&D because often (depending on GMs as ever) failing a roll just maintains the status quo and nothing happens. Though with Defy Danger I think often the opposite is true - succeeding the roll often feels like it’s the boring option. And the move just feels like a way to stymie hard moves. I can’t remember if this was brought up in this thread or else where, but I think it would often be more interesting for the GM to just ask a question/impose a cost without the roll.
Regarding the multi-stat thing at the beginning, I think it can be worse the other way round. I’m not a fan of the “act under pressure” move in SCUP being incredibly broad in application, while also being tied to one stat. So I think the multi-stat justification for this in DW is the lesser of two evils 
I’d be inclined to stick to the principle “to do it, you do it”. If an archer 15ft away fires directly at you, I think eyebrows would be raised if you said ‘I try to determine its trajectory and blah blah’. It’s a direct shot, not trajectory! However, if a volley of arrows is fired at you from 300 yards away, looking at the trajectory and planning evasive action would seem completely reasonable.
You could always add “Wizard quotes Xeno’s Paradox and the arrow never reaches her”.
Love it. “I told the arrow that I write the laws of reality in this locality.”
6- Wizard is condemned to run away from an arrow for an eternity of infinitely small time fragments.
The arrow can never reach them though right - coz it’s gotta go half the remaining distance each time…?