Augury vs. Spell Lists

So, Plausiblefabulist threw down with this great quote in regards to the maelstrom in AW

From my point of view -- also as an author of fiction -- the way the world's psychic maelstrom works in AW begins to repair the decades of ruination that fantasy RPGs since D&D have wreaked on fantasy.

The psychic malestrom is numinous. It is deeply unpredictable. You encounter it at terrible risk; it can change everything, make anything possible, and you have no idea, in advance, how, or what the cost and consequences will be.

In other words, it acts like magic used to act in literature -- how magic acts in Middle-Earth, in Earthsea, in Prydain -- and as the miraculous acts in, for that matter, the Hebrew Bible --  and not how magic and the divine function in fantasy ever since D&D turned them into vending machines that produce precisely predictable effects given precisely predictable inputs.

Harry Potter casts third level spells from page 126 of the Player's Handbook. Ged opened his brain to the world's psychic maelstrom.

That got me thinking about color-first gaming and the specificity of moves like this. Most of AW is very color-specific - it is designed to evoke specific images of apocalypse-ness. The maelstrom is the same, but mechanically it's very open ended. Even augury is very open.

I'm working on a hack right now that involves witchcraft, and I was originally designing the magic to have specific moves (like a spell list) in a desire to evoke a very flavor. I'm sure it really comes down to preference and design intent, but I was wondering what people's thoughts were on possible advantages and/or drawbacks to having something like "magic" being more or less structured.

I found that discussion very interesting.

I think it depends on how you expect the players to interact with "magic".  Is magic something to explore, unknown and possibly unknowable? Or is it a science, defined and with definite effects given the right input?

Basically are the players going to expand the world's knowledge of magic? Develop new powers outside? Then the augury way may be better.

If on the other hand magic is set and the best characters might do is find a new application or recover lost lore, then a more specific move based approach would be better.

It may also be true that a lot of this depends on just what magic is capable of in your specific fiction.  If magic is extremely broad and can cover a lot of different effects or situations, a move-based approach is going to be an exercise in silliness.  You'll be better off just riffing on opening your brain and describing the effects on an as-needed basis.  I mean, what's the real functional mechanical difference between a lightning bolt and a disintegration ray?  And is that difference important enough to force a player to opt to spend an advancement on move on one over the other?

On the other hand, if magic is very specific and there are only a handful of things that a user of those powers can accomplish, then yeah, maybe a move for each "spell" (or broad effect type) is OK.

If it were me and I were trying to keep flexibility but give different kinds of magic different flavors, I might split the difference.  So there might be a custom move for divination magic, a custom move for curses, and a custom move for summoning, and a custom move for whipping out a can of elemental whoop-ass.  All of those moves would be flexible internally to allow the players (and the MC) plenty of latitude to describe how they are applied and what their effects might look like.

I just had another idea, which is this.

On the one hand, I'd like to see open-ended magic that begins by presenting cues and imagery to the user, like open your brain does, but leaves it up to the on-table play what that ends up meaning, so that the "magic system" is an emergent result of the characters' actions. As I described in the other thread, that's been working brilliantly in our AW game.

On the other hand, I've heard it said that "advancement" is the soul or throughline of an RPG -- certainly it seems to be where a lot of its values live -- and Powered by the Apocalypse games seem to do a brilliant job creating character motivation through experience and advancements. If your game is about witches and so magic is at its center, and you leave that magic generative and open-ended, how do you also tie it to the advancement system? How do you make the "ways to get experience" part of the game connect with the "ways to affect the world with magic" part, if you're not letting them pick spells from a spell list when they level up.

Maybe something like:

When you try to achieve something you have never done before by altering the world-weave, roll + gifted.
On a hit, you achieve some version of it and will be able to replicate the effect in the future on a successful roll. On a 10 up, choose two limitations or costs it always has -- situations it will never work in, ways that it invariably leaves you drained, dangers it necessarily exposes you to -- the MC chooses two costs or dangers that it sometimes incurs, but not always (and not this time). On a 7-9, the MC picks all four. On a miss, the MC will create a new threat related to what you were trying to achieve.

When you try to recreate a mystical effect you have already once achieved, roll + control. On a hit, you do it, subject to its necessary limitations and costs as established. On a 7-9, you pick one of the possible costs or dangers as well.

When you take an advance, instead of taking one of the listed advances, you may remove one of the permanent costs or limitations of your magic.

I like this.  It leaves things open-ended, but allows players to repeat known effects.  Essentially allowing them to build their own "spell-books."  What might make this even cooler is if there were no way for one witch/mage/sorceror to teach a "spell" to someone else - you just have to know it's possible and try to figure it out for yourself.  That makes magic highly individualized, which I think is fantastic.

I like "you can't teach spells"; also interesting would be to have the teaching of spells be a separate move requiring experience, and have it be risky as opposed to an automatic success. A spell, here, isn't a rote recipe or algorithm, it's a way of interacting with a move. Teaching someone to cast a spell is a separate skill from casting a spell, and it requires that you actually cast it while they're with you, and that they participate, and their participation is risky and can end in disaster.

So perhaps you have to take the advance that allows you to teach magic, and then to teach a spell, you cast the spell (requiring its own roll, risk, and prerequisite situation) and you also roll+something to see how well you did on teaching  -- maybe you roll+relationship (hx/bonds/etc), mirroring the traditional intimacy of master/student relationships -- and a miss there counts too...