Saving throws as analogy for moves

Has this observation already been done?

Another cross post from another forum but I thought it was interesting to hear you gearheads' take on it:

[In Apocalypse World and Dungeon World] every single action is like how saving throws work in D&D.
In D&D, you're talking along, describing what you're doing, how you're walking down the dungeon corridor etc, conversation style, everything's dandy... then whoops, you step in some sort of poison hole, there might be some trouble involved with that, make a saving throw and that'll tell us all the consequences.

In Dungeon World (and AW), every single thing is like this. You're talking describing how you're fighting the monster, chopping along, having fun fighting this horrible giant ogre but whoops, you put yourself in danger in the fight, make a "Hack&Slash" check to see whether you are hurt.

It's All Saving Throws, All The Time. My friend Trix described it as a "consequences system" rather than a "task resolution system" (and rather than a "conflict resolution system" that The Shadow of Yesterday uses), but I think of it as ASTATT.

It works because the action economy is built around it. Monsters deal damage when they deal damage. If you say "I go stand in front of the ogre and look at it intently in the eyes", well, you might get smashed straight up. Similar to how there is no saving throw in D&D if you just go jump off a cliff deliberately. But if you go "I go stand in front of the ogre and then I bring up my dagger and try to cut its fingers off before it can smash me [whoops, that got a little graphic, I apologize]", well, you've got to make a roll and the consequences of failing that roll might be that you get smashed because you weren't fast enough or good enough with the dagger.

Also, this is why the name "move" is a little strange; when I first heard of the moves my thought immediately went to moves in video games, things that you can trigger by pushing buttons in the interface; things that after a while become the interface.

But... in the ASTATT model, the moves are kinda the other way around. Your "interface" is what you are doing in the fiction, and the mechanics are engaged from that. Maybe I'm overstating it...?

Kind of. The moves "trigger" by the fictional situation, not the other way around. You don't say, "Oh, I'm gonna roll+Hot to get him to do what I want." Instead, the MC looks at what you're trying to do narratively, and if that situation sounds like manipulation, he or she will say, "OK, cool, roll+Hot."

The important thing is that the moves are both descriptive (they tell us what has happened) and prescriptive (they tell us what's about to happen). The descriptive part is what you're talking about with your "saving throw" analogy, but the prescriptive part is (IMHO) more important. It's what keeps the moves snowballing, and describes how the fictional positions of the various participants change. The MC moves in particular tend to work this way.

Also, different moves have different structures. Do something under fire definitely has aspects of "save versus bad stuff happening to you," but seize by force assumes that bad stuff is already happening to you. As soon as the move is triggered, you know you're taking harm. The question then becomes one of "what are your priorities in this conflict" which has a very different feel mechanically than a "saving throw." Sure, one of your options is to reduce the damage you are taking in return, but depending on the situation, that option might not be important to you. Similarly, seduce or manipulate is all about "what is it going to cost me to get what I want?"

I don't think it's completely analoguous. Stepping into a poison trap and then rolling to see if you get hurt does not happen in Apocalypse World as the moves correspond to real actions in game's fiction. What could happen though:

GM: "You see a narrow corridor in front of you, void of anything remarkable. It is jarringly different from the other corridors you've walked through before reaching this point. (Announce future badness)"
PC: "I have a bad feeling about this. I try to navigate the corridor very carefully and treading lightly.
GM: "Ok! That sounds like doing something under fire. Roll+cool"

Then results would be something like:
10+: Navigated without problems
7-9: Getting through with slight complication e.g. scraped by a dart trap or not getting hurt but triggering some sort of an alarm.
Miss: Getting hurt, getting stuck, getting separated from other PCs by falling through a trapdoor on the floor.

My two cents.

Kind of. The moves "trigger" by the fictional situation, not the other way around.

Right, I'm thinking "to do it, do it", and if the MC doesn't know what happens they'll ask you to make a roll.
Some things have a rule to always be "uncertain" in this way so that the MC just can't decide something.

For example, there is some sorta corridor with a red/white checkerboard floor where the red squares are made out of very thin glass with poison pools under them. The MC tells the player about how the floor looks but not what it's made of. What do you do?

And the player goes "I'm scared, I'm just gonna run through it as fast as I can" and the MC says "after a couple of steps you hit a red square hard and it shatters, and you smell poison. What do you do?" and the player goes "I try to pull my foot up from there before getting hurt" and the MC says "OK, roll act under fire" because she doesn't know how successful the foot-up-pulling is going to be.
7-9 get hurt by glass, 6- get poisoned
But if the player described themselves as like throwing in some hard rocks in the corridor, there would be another outcome and another move engaged, or maybe no move would need to be engaged.

But that's how I use saving throws, too. They flow from the fiction, from the player's actions.


In my original example I had "then whoops, you step in some sort of poison hole". If they don't step in the poison hole there's no need for a saving throw; it flowed out of the player's action.

Conversely I don't just have them roll for everything they do. The roll isn't the doing... the roll is to figure out the consequences of their actions.
I've been lucky playing under four amazing MCs (including Jonatan and Simon J B on these forums) and they were all great with this. I just played "to do it, do it". And then we engaged the mechanics to find out the consequences. I think that's my philosophy when playing AW; to just do things, to trust in the system and just do things, not constantly looking through the movelist to see what "buttons to press".

I think that's my philosophy when playing AW; to just do things, to trust in the system and just do things, not constantly looking through the movelist to see what "buttons to press".

Then I think you're doing it right. :)

I think that's my philosophy when playing AW; to just do things, to trust in the system and just do things, not constantly looking through the movelist to see what "buttons to press".

I don't think that this approach is entirely coherent with the way Apocalypse World is played, and I don't think seeing moves as saving throws is coherent either.

Why? Mainly, because Apocalypse World is not a generic system. And all PbtA games that have had success (Read: that I know of, obviously :) ) are not either. All these games play very differently. They create different conversations. Yet what differs between these games are, almost exclusively, the moves. As such the moves must have a role in shaping conversation.

Note: It could be that it is the difference setting that steers these conversations in different ways. Sure. But PbtA games usually don't have anything relating to the setting in them, and certainly not enough to justify the difference in play. Apocalypse World basically just tells you: "Scarcity and the Psychic Maelstrom". Dungeon World tells you "Fantasy" and Monster Hearts tells you "Monsters and Teens". To further support the importance of moves over setting: if one considers two different games of Apocalypse World, where the setting is the same, yet all of the playbooks are different, I think they will also observe two very different types of conversations.

So the moves have a role in steering conversation. Big deal. How do they work? Your suggestion seems, to me, to be: "Moves are triggered by the state of the conversation and modify its state when they occur". Does this model account for the behavioral differences between players? Partly. You could say that the reason the conversation is more violent when there is a Gunlugger in the group is that violence succeeds more often when it comes into play (A better "violence saving throw"), thus becoming more noticeable. Some moves (Like Battle-Hardened) also replace non-violent acts with violent ones. Further steering the conversation.

But clearly that's not all that's happening here. My Driver will gravitate towards its car before any moves happened. And, more importantly, my Brainer would never have tied a threat to a chair and started getting freaky if they did not have the moves "Deep Brain Scan" or "In-brain puppet strings". The process here is not: "Oh I see you bound Jake in a dark room, and started dancing that weird dance, did you know you could use this to your advantage?". It is "I wanted to gain advantage and as such I went to some length to get Jake bound in a dark room". This is very un-saving-throwy to me.

I think that moves create a metagame of fictional positioning. But that positioning is deeply active and the players probably are and should be looking at the buttons they want to press. I don't think Apocalypse World can be properly summed up by what it reacts to. It very much encourages behaviors too.