Sphere, Aspect, and Source: Rethinking the Schools of Magic

[PhoenixPhyre]

I'm not a fan of the 8 legacy Schools of Magic. They're inconsistent, mushy, and mostly a legacy hangover.

As I understand it, they are only used for:
* Wizard sub-class feature matching (when you cast spells of ____ school, _____)
* EK/AT spell selection
* detect magic returns the school of the spell detected

Effectively, they're just tags for spells. Not even all magical effects, just spells. Which makes detect magic either really weak (since it only picks up spells in any detail) or annoying to adjudicate (what's the school of magic for a magic item that doesn't cast a spell? Or that casts multiple spells of different schools?). And they're mutually-incompatible in this edition--each spell has one and only one school. It's also no help for people homebrewing their own spells--what school should they be in?
----------
So here's my thoughts towards a non-breaking replacement tagging system. The intent is to support the same use cases as before, but give more and clearer information while not requiring a full rework of the metaphysics. The implementation details (ie which spells/effects get which tags) is still a to-do, this is the general philosophy of the proposal.

Each spell and magical effect has tags from three categories:

Sphere: the closest equivalent to the current schools. In fact, some of the labels get reused. But they're defined more precisely. Ideally, these would be non-overlapping--each spell would have a single sphere.
* Evocation -- produces energy that goes away once the spell is over.
* Summoning -- pulls matter from point A (or plane A) to point B (possibly in plane B). Permanent movement.
* Conjuration -- like summoning, but temporary
* Illusion -- creates a "fake" thing, interactions are entirely in the mind of the observers.
* Transmutation -- changes one substrate into another thing, permanently or temporarily. Requires an existing substrate. Results not necessarily magical.
* Restoration -- restores/repairs life to/in an entity
* Divination -- revealing information that is not present at the location/time of the casting.
* Enchantment -- Altering the mental state of the target.

Aspect(s): These are additional information about the spell's effects. Is it elemental? Aligned (ie connected to the Outer Planes)? Fey? Mind-affecting? Charm? Etc. Spells could have multiple of these.
* Elemental (X)
* Aligned (X) (including positive/negative energy)
* Blessing
* Cursing
* Warding
* ??

Source: This would be set by the effect's creator, not the spell/effect itself. So a light cleric casting fireball and a wizard casting fireball would have different source tags. In-universe, this would affect the exact details of the verbal and somatic components. Other things could key off of this if they wanted, but right now it would be purely informational (ie this ward came from calling on divine power, not a direct twisting of the weave/whatever).
* Arcane: Cast by a wizard/sorcerer/warlock/bard/artificer/EK/AT.
* Divine: Cleric/Paladin
* Primal[1]: Druid/Ranger
* Innate (psionic?): Monk/monster innate casting.

Detect Magic would now reveal one or more of the tags. I'd say it always reveals the sphere, but lets you make an Intelligence (Arcana) check to get the other pieces.
Wizard/AT/EK/etc abilities would key off of tags. Sphere, mostly. But you could have bonus spells for other classes that are selected by tags. Etc.

[Segev]

I really am starting to sound like an evangelist for this, even though I think it at best a good competitor with the normal casting system. However, I think a lot of what you seem to be looking for is addressed by this third party supplement:

http://spheres5e.wikidot.com/start

[Sorinth]

Your sphere's will end up just as inconsistent and mushy. Where does Animate Dead fall into things, what about Bless or Magic Weapons they are no longer Enchantment because they aren't mental states, not sure they fit into Evocation either, or Absorb Elements here you aren't producing energy you are reducing it.

So what's the end game of these changes as it seems purely informational.

[PhoenixPhyre]

I really am starting to sound like an evangelist for this, even though I think it at best a good competitor with the normal casting system. However, I think a lot of what you seem to be looking for is addressed by this third party supplement:

http://spheres5e.wikidot.com/start

I'm not looking to actually change the mechanics of the spell system itself with this--that's a much larger concern. Merely to refactor the foundational classifications onto something more solid than "developer arbitrary choice with no coherent themes other than tradition".

Your sphere's will end up just as inconsistent and mushy. Where does Animate Dead fall into things, what about Bless or Magic Weapons they are no longer Enchantment because they aren't mental states, not sure they fit into Evocation either, or Absorb Elements here you aren't producing energy you are reducing it.

So what's the end game of these changes as it seems purely informational.

The primary end state is mostly informational. Mostly to refactor the classifications so that other changes could build on them. At the point of spell/mechanic interactions, I'd hope that very little changes. But at the level of "describing what you see when you cast detect magic" or "explaining why X subclass gets bonuses to Y spell" or "I want to add a new spell, what 'school' should it be" or "I want to add a new feature that interacts with these spells but not those spells", I want to have a more robust framework.

And the spheres/tags proposed in the OP are neither complete nor final--they're iteration 0.0.1-alpha. The idea would be to iteratively consider each spell, seeing if it fits into the framework or not. If not, adjust the framework (either change the definitions or add a new tag) to help it fit. Then look for duplications and gaps.

Note: if there are spells that don't fit well once the framework works for everything else, that indicates that maybe those spells are doing too much and should be broken into pieces or removed entirely. Even if that means laying traditions aside. That's the value of a stable classification framework--it gives you a comparison point.

[Sorinth]

I'm not looking to actually change the mechanics of the spell system itself with this--that's a much larger concern. Merely to refactor the foundational classifications onto something more solid than "developer arbitrary choice with no coherent themes other than tradition".



The primary end state is mostly informational. Mostly to refactor the classifications so that other changes could build on them. At the point of spell/mechanic interactions, I'd hope that very little changes. But at the level of "describing what you see when you cast detect magic" or "explaining why X subclass gets bonuses to Y spell" or "I want to add a new spell, what 'school' should it be" or "I want to add a new feature that interacts with these spells but not those spells", I want to have a more robust framework.

And the spheres/tags proposed in the OP are neither complete nor final--they're iteration 0.0.1-alpha. The idea would be to iteratively consider each spell, seeing if it fits into the framework or not. If not, adjust the framework (either change the definitions or add a new tag) to help it fit. Then look for duplications and gaps.

Note: if there are spells that don't fit well once the framework works for everything else, that indicates that maybe those spells are doing too much and should be broken into pieces or removed entirely. Even if that means laying traditions aside. That's the value of a stable classification framework--it gives you a comparison point.

I question the value of a framework when you need to remove or alter spells to fit that framework given that the framework itself doesn't provide anything mechanically.

I think what you are getting as it is a question of bottom up or top down approach to making spells. Currently it's top down where the developer makes a spell that they feel is cool/balanced and then slaps on the most appropriate "tags" they can find. In this way there's basically no limit to what magic can do beyond concerns for balance. Whereas you are proposing a bottom up where you create the foundation/framework for magic and then from that framework create the possible spells.

It's not a question of one approach being right or wrong way to go about it, you for sure will get more logically consistent groupings of spells but at a cost of creativity. And I personally don't see the value in having consistent spell groupings outweighing that creativity.

[PhoenixPhyre]

I question the value of a framework when you need to remove or alter spells to fit that framework given that the framework itself doesn't provide anything mechanically.

I think what you are getting as it is a question of bottom up or top down approach to making spells. Currently it's top down where the developer makes a spell that they feel is cool/balanced and then slaps on the most appropriate "tags" they can find. In this way there's basically no limit to what magic can do beyond concerns for balance. Whereas you are proposing a bottom up where you create the foundation/framework for magic and then from that framework create the possible spells.

It's not a question of one approach being right or wrong way to go about it, you for sure will get more logically consistent groupings of spells but at a cost of creativity. And I personally don't see the value in having consistent spell groupings outweighing that creativity.

I'm trying to reverse-engineer a framework that fits. Sure, if 90% (or even 50%) of the spells don't fit into a coherent framework, there's a problem and you'd need to scrap the whole idea. But I, for one, find the idea that "spells can do anything, there's no rhyme or reason, it's just limited by the developer's grasp on balance" to be horrific. "Creativity" in this case really means that you get a few spells that everyone uses and the rest molder. And that kind of slap-dash, anything-goes, post-hoc spell development makes coherent worldbuilding basically impossible. And makes homebrewing anything interesting much more of a pain than it's worth.

[Segev]

I'm trying to reverse-engineer a framework that fits. Sure, if 90% (or even 50%) of the spells don't fit into a coherent framework, there's a problem and you'd need to scrap the whole idea. But I, for one, find the idea that "spells can do anything, there's no rhyme or reason, it's just limited by the developer's grasp on balance" to be horrific. "Creativity" in this case really means that you get a few spells that everyone uses and the rest molder. And that kind of slap-dash, anything-goes, post-hoc spell development makes coherent worldbuilding basically impossible. And makes homebrewing anything interesting much more of a pain than it's worth.

I just disagree with you that the current framework "doesn't fit." It works fine for what it does. I don't see value added for doing the work to relabel all the spells by a new organizational system.

[PhoenixPhyre]

I just disagree with you that the current framework "doesn't fit." It works fine for what it does. I don't see value added for doing the work to relabel all the spells by a new organizational system.

The current system is entirely ad hoc. Why are spells labeled the way they are? Because tradition and because we felt like it. It's also impossible to extend--you can't decide "what should detect magic report if what it's looking at is a magical effect that wasn't caused by a listed spell" with any kind of fidelity. Which makes detect magic not very useful.

Plus, you lose all thematic matching beyond the very most basic. Which makes wizard subclasses, EKs, and AT get all sorts of kinked up and inconsistent.

That said, this isn't a high-priority project. More a thought experiment to see if there's low-hanging fruit. Prompted by the "we should use the Magic colors instead" thread. I'm not actually planning to do any work along these lines anytime soon--I have much more pressing demands for my homebrew time. Including figuring out how to represent mecha-cyborg gnomes in a way that makes sense in-universe[1] and are easy to run mechanically at the power level I want them at. Which I need for...well...tonight.

[1] I know they have to exist in this one particular semi-quasi-demi-plane. But how, and what, and the details are still not certain.

[Segev]

The current system is entirely ad hoc. Why are spells labeled the way they are? Because tradition and because we felt like it. It's also impossible to extend--you can't decide "what should detect magic report if what it's looking at is a magical effect that wasn't caused by a listed spell" with any kind of fidelity. Which makes detect magic not very useful.

Plus, you lose all thematic matching beyond the very most basic. Which makes wizard subclasses, EKs, and AT get all sorts of kinked up and inconsistent.

That said, this isn't a high-priority project. More a thought experiment to see if there's low-hanging fruit. Prompted by the "we should use the Magic colors instead" thread. I'm not actually planning to do any work along these lines anytime soon--I have much more pressing demands for my homebrew time. Including figuring out how to represent mecha-cyborg gnomes in a way that makes sense in-universe[1] and are easy to run mechanically at the power level I want them at. Which I need for...well...tonight.

[1] I know they have to exist in this one particular semi-quasi-demi-plane. But how, and what, and the details are still not certain.

I suppose my problem is that your proposed solution doesn't seem to solve any of the issues you raise, here. It will still be arbitrary. It will still require deciding on what "sphere" a given effect that isn't specifically a defined spell would fall under.

We already have discussion over whether a given spell should be in X school or Y school (3e's decision to make healing a subschool of conjuration, vs. other editions' choice to make it a necromancy, for example), and these discussions are based on actual understanding of why spells fall into various categories. The schools ARE defined, if sometimes fuzzy at the edges, and I don't think any classification system of fictional magic that doesn't use a Sanderson-esq bottom-up hard magic "physics" to build it will resolve that issue.

[PhoenixPhyre]

I suppose my problem is that your proposed solution doesn't seem to solve any of the issues you raise, here. It will still be arbitrary. It will still require deciding on what "sphere" a given effect that isn't specifically a defined spell would fall under.

We already have discussion over whether a given spell should be in X school or Y school (3e's decision to make healing a subschool of conjuration, vs. other editions' choice to make it a necromancy, for example), and these discussions are based on actual understanding of why spells fall into various categories. The schools ARE defined, if sometimes fuzzy at the edges, and I don't think any classification system of fictional magic that doesn't use a Sanderson-esq bottom-up hard magic "physics" to build it will resolve that issue.

The goal is to nail down the spheres (especially) to very concrete, easy-to-apply categories. Or to determine that it can't be done without 50-bazillion categories, which is another option.

A secondary purpose is to just flat out give more information. So you're not limited to just saying "uh...it's evocation (which encompasses enough that it's not really giving anything useful[1])", you can say "it's an evocation aspected to elemental fire, cast by an arcanist". Without having to tie everything to existing spells or read through all the details of every spell every time. And help build thematic lists quickly.

[1] eg--why the heck is wall of stone an evocation? It produces a permanent, physical object of stone. Contrast that with the description of the evocation school

Evocation spells manipulate magical energy to produce a desired effect. Some call up blasts of fire or lightning. Others channel positive energy to heal wounds.

"Manipulate magical energy" is so broad that anything can fit into it. It's a useless definition.

[Segev]

The goal is to nail down the spheres (especially) to very concrete, easy-to-apply categories. Or to determine that it can't be done without 50-bazillion categories, which is another option.

A secondary purpose is to just flat out give more information. So you're not limited to just saying "uh...it's evocation (which encompasses enough that it's not really giving anything useful)", you can say "it's an evocation aspected to elemental fire, cast by an arcanist". Without having to tie everything to existing spells or read through all the details of every spell every time. And help build thematic lists quickly.

Allowing it to grant more information isn't a bad idea. "It seems to be cast by a wizard" vs. "It seems to be cast by an artificer," and "it seems like it's got a lightning type effect" are all good things to be able to tell people. I think you need less to focus, then, on what spells fall where, but on determining exactly how difficult you want it to be to get various pieces of information. Should detect magic just give all of that, or should it require an Intelligence (Arcana) or (Religion) check or...?

[PhoenixPhyre]

Allowing it to grant more information isn't a bad idea. "It seems to be cast by a wizard" vs. "It seems to be cast by an artificer," and "it seems like it's got a lightning type effect" are all good things to be able to tell people. I think you need less to focus, then, on what spells fall where, but on determining exactly how difficult you want it to be to get various pieces of information. Should detect magic just give all of that, or should it require an Intelligence (Arcana) or (Religion) check or...?

For my personal style, that latter part is pretty simple. I actually run detect magic as way more free, giving lots of additional information. But deciding what information applies takes effort. And I don't like effort. This is more about making my rather free-form, labor-intensive system more internally consistent and helping me worldbuild better.

Rough guess, I'd say "you get the sphere for free, then DC 10 Intelligence (Arcana|Religion) to get the basic aspect (so "elemental" or "aligned"), with DC 15 giving you the full aspects and DC 20 giving you everything." Whereas for identifying a spell on the fly, getting the source would be easy as long as there are somatic or verbal components but getting the aspects/identity would be harder. So I'd probably say "DC 5 Intelligence (Arcana|Religion)[1] for the source, DC 10 for a basic aspect, Xanathar's DCs (15+spell level) for the identity." It's basically a form of partial/non-binary success. You don't get the identity, but you can tell something about it.

Or about deciding that there's no consistent framework here and I need to beat things into a much harder shape and commit to a lot more pruning and labor. CF the part about not liking effort.

[1] with advantage for picking the "right" one for the source.

[Segev]

For my personal style, that latter part is pretty simple. I actually run detect magic as way more free, giving lots of additional information. But deciding what information applies takes effort. And I don't like effort. This is more about making my rather free-form, labor-intensive system more internally consistent and helping me worldbuild better.

Without spending the effort ahead of time to identify what each spell individually would have available to tell, I don't think you're going to be able to avoid having to make it up on the fly.

[Dr. Cliché]

I appreciate the idea but I don't think your system will really solve much.

As an example, consider a spell like Fear. Is it an illusion spell or an enchantment spell under your system?

"Illusion -- creates a "fake" thing, interactions are entirely in the mind of the observers."

Check. You're creating an illusion of a creature's worst fear.

"Enchantment -- Altering the mental state of the target."

Also check. You're making them afraid.

Hmm.


Another example would be Planar Ally. Depending on the creature you summon and your ability to bargain, it could end up serving you only temporarily (making it a Conjuration spell) or it could become a permanent party member (making it Summoning). Does the sphere change retrospectively, depending on your bargain?


Then you've got more esoteric stuff like Soul Cage. The 'primary' effect doesn't seem to be covered by any sphere, and the secondary ones could make it fall into multiple categories (it can heal you, so Restoration, but it also has a Divination ability).


I don't want to just trash on your idea, I'm just trying to get across that there are always going to be spells that don't fit neatly into any one sphere, regardless of whether we're using your system or the current one.

The reason is that these systems are descriptive, rather than prescriptive. If you want a system with more defined categories then it needs to be implemented from the ground up. Because you need to create and enforce rules for each category. You need to know what the hard limits of Transmutation magic are - what a transmutation spell can or can't do. Otherwise, you'll just end up with categories bleeding into one another.

[Sorinth]

I'm trying to reverse-engineer a framework that fits. Sure, if 90% (or even 50%) of the spells don't fit into a coherent framework, there's a problem and you'd need to scrap the whole idea. But I, for one, find the idea that "spells can do anything, there's no rhyme or reason, it's just limited by the developer's grasp on balance" to be horrific. "Creativity" in this case really means that you get a few spells that everyone uses and the rest molder. And that kind of slap-dash, anything-goes, post-hoc spell development makes coherent worldbuilding basically impossible. And makes homebrewing anything interesting much more of a pain than it's worth.

World building is perfectly coherent within the given system. Magic is simply magic and there are many different ways to access magic, the resulting spells/effects are the same no matter how you access that magic because they are simply a method to access the same thing.

In fact the schools of magic aren't real and have no bearing on how magic works, they are simply ways people decided to classify/group similar seeming spells. Think of it as biological classifications of species, genus, etc... they aren't real rules that life actually has to follow, they are things we made up as a way to organize the differences/similarities we see. Wizards/Scientist study the thing, they notice similarities with other things and group them together with a name. Some are easy to classify because they are very similar, and other are Platypus like things that don't fit.

[PhoenixPhyre]

World building is perfectly coherent within the given system. Magic is simply magic and there are many different ways to access magic, the resulting spells/effects are the same no matter how you access that magic because they are simply a method to access the same thing.

In fact the schools of magic aren't real and have no bearing on how magic works, they are simply ways people decided to classify/group similar seeming spells. Think of it as biological classifications of species, genus, etc... they aren't real rules that life actually has to follow, they are things we made up as a way to organize the differences/similarities we see. Wizards/Scientist study the thing, they notice similarities with other things and group them together with a name. Some are easy to classify because they are very similar, and other are Platypus like things that don't fit.

I disagree--if magic can do anything, then magic can do everything and the answers to "why X" collapse to "because someone cast a spell". And the answers to "why not Y" collapse to "no one has cast that spell yet." Which mean you can't actually build anything solid and are constantly playing catch-up. I prefer being able to ask the setting questions and having the answers follow from the assumptions and frameworks in place. Instead of just being "lol, something something a wizard did it."

If magic is literally anything, if the only limits are what a particular writer writes, then magic is a total deus ex machina. And that removes the entire weight and verisimilitude of the world. Limits are important. And as it stands, there are no limits. No notion of what magic can and can't do. Which, I think, is 99% of the magic/non-magic imbalance issue. This isn't a total solution, but there has to be a framework, some things that can be said "thus far and no further".

[Sorinth]

I disagree--if magic can do anything, then magic can do everything and the answers to "why X" collapse to "because someone cast a spell". And the answers to "why not Y" collapse to "no one has cast that spell yet." Which mean you can't actually build anything solid and are constantly playing catch-up. I prefer being able to ask the setting questions and having the answers follow from the assumptions and frameworks in place. Instead of just being "lol, something something a wizard did it."

If magic is literally anything, if the only limits are what a particular writer writes, then magic is a total deus ex machina. And that removes the entire weight and verisimilitude of the world. Limits are important. And as it stands, there are no limits. No notion of what magic can and can't do. Which, I think, is 99% of the magic/non-magic imbalance issue. This isn't a total solution, but there has to be a framework, some things that can be said "thus far and no further".

And reclassifying spells into sphere's instead of schools address that how?

[PhoenixPhyre]

And reclassifying spells into sphere's instead of schools address that how?

It starts down the road of building a framework for "spells can do X" in the future. As well as pointing out spells that are either missing or that are bad fits for coherency.

Is it the right path? Dunno. Probably not. But throwing up our hands and saying "yay, magic can do anything with no limits at all" just reinforces the current mess. And I mean that very much more for the thematic, narrative, and worldbuilding aspects than the game-mechanical balance aspects, which I care relatively little for.

[tKUUNK]

I like it.

On one hand I'm glad it's not part of the official 5e rules, because it could get a bit cumbersome. Especially for n00bs. I can see the heads exploding already like back in the good old days....thac0....3.5e.....[[sigh]]

But man, here's something I can blend into my own games, and it sits on top of the existing rules with zero difficulty, and adds some flavor to magic. Heck yeah!

As a DM, I tend to be a bit more descriptive with the information players gain from Detect Magic already anyway. Nice framework you lay out here.

[PhoenixPhyre]

I like it.

But man, here's something I can blend into my own games, and it sits on top of the existing rules with zero difficulty, and adds some flavor to magic. Heck yeah!

As a DM, I tend to be a bit more descriptive with the information players gain from Detect Magic already anyway. Nice framework you lay out here.

If this is all that comes of it, that someone else considers it useful even at an informational level/helps them be more engaging to their players, then I'll consider it a success.

[Sorinth]

It starts down the road of building a framework for "spells can do X" in the future. As well as pointing out spells that are either missing or that are bad fits for coherency.

Is it the right path? Dunno. Probably not. But throwing up our hands and saying "yay, magic can do anything with no limits at all" just reinforces the current mess. And I mean that very much more for the thematic, narrative, and worldbuilding aspects than the game-mechanical balance aspects, which I care relatively little for.

Honestly I have a hard time imagining how a framework would actually accomplish any kind of spell building. I mean I can see how you could use it to prevent certain themes, like say time travel. But I can pretty much guarantee that there will be spells that will fit in the allowed themes that would be disruptive to theme/narrative/worldbuilding. Maybe if you have some concrete examples?

There's also something to be said for magic can do anything is needed for many plots. If you have created all sorts of rules about what magic can/can't do then aren't you preventing the trope where the PCs have to race to stop some evil cultist from completing a world ending magic ritual?


That said, if certain spells are world breaking or don't fit the theme of a campaign then by all means ban those spells from that campaign. As an example if you want to ban Leomund's Tiny Hut, you could make up rules about how force magic is inherently temporary in nature and so is mostly reserved for instantaneous effects like Eldritch Blast, and only very high level magic can get it to last more then a few minutes so LTH is not supported by the magic. Or you could just say LTH is banned because it doesn't fit the narrative/world of the campaign.

I guess I disagree that the system is a mess where magic can do anything with no limits. There are limits, yes the limits are knowledge and power based. Much in the same way an AK-47 isn't on the weapon list, it's not that it's impossible and can't work, it's that nobody has the knowledge of how to make one.

[Segev]

It starts down the road of building a framework for "spells can do X" in the future. As well as pointing out spells that are either missing or that are bad fits for coherency.

I don't see how it does this any more than the existing schools do. You could use the existing schools exactly as you intend to; you'd just have to start eliminating some spells. You could also wind up with your proposed categorization system breaking down in exactly the same way the existing one does, with "spells can do anything" spells being shoehorned into one sphere or another.

[PhoenixPhyre]

Honestly I have a hard time imagining how a framework would actually accomplish any kind of spell building. I mean I can see how you could use it to prevent certain themes, like say time travel. But I can pretty much guarantee that there will be spells that will fit in the allowed themes that would be disruptive to theme/narrative/worldbuilding. Maybe if you have some concrete examples?

There's also something to be said for magic can do anything is needed for many plots. If you have created all sorts of rules about what magic can/can't do then aren't you preventing the trope where the PCs have to race to stop some evil cultist from completing a world ending magic ritual?


That said, if certain spells are world breaking or don't fit the theme of a campaign then by all means ban those spells from that campaign. As an example if you want to ban Leomund's Tiny Hut, you could make up rules about how force magic is inherently temporary in nature and so is mostly reserved for instantaneous effects like Eldritch Blast, and only very high level magic can get it to last more then a few minutes so LTH is not supported by the magic. Or you could just say LTH is banned because it doesn't fit the narrative/world of the campaign.

I guess I disagree that the system is a mess where magic can do anything with no limits. There are limits, yes the limits are knowledge and power based. Much in the same way an AK-47 isn't on the weapon list, it's not that it's impossible and can't work, it's that nobody has the knowledge of how to make one.

Part of my mental frame here is that I don't build a campaign world per campaign. I have a single world that has multiple ongoing, more-or-less sandbox campaigns, and the campaigns influence the world, "living world" fashion. So I have to have consistency not only within a campaign but between campaigns. Which means I can't get away with ad hoc, campaign-specific reasoning. I have to think of the world as an independent entity with its own needs, built on a solid foundation so that I can extrapolate on the fly without writing myself into corners.

I don't see how it does this any more than the existing schools do. You could use the existing schools exactly as you intend to; you'd just have to start eliminating some spells. You could also wind up with your proposed categorization system breaking down in exactly the same way the existing one does, with "spells can do anything" spells being shoehorned into one sphere or another.

The goal would be to have the spheres have a single, clear purpose. So anything that would have to be shoehorned is a red flag (either for the framework or the spell). You certainly could turn the current schools into spheres, but (as I noted about evocation) you'd have to radically rewrite the schools. Because currently they're a bunch of words that don't actually impose any constraints. "Manipulate magical energy"--every spell manipulates magical energy by definition. So either way, you're at "come up with a coherent set of descriptors, whatever you call them". Or give up on a framework entirely. Which (IMO) would stink.

[Segev]

The goal would be to have the spheres have a single, clear purpose. So anything that would have to be shoehorned is a red flag (either for the framework or the spell). You certainly could turn the current schools into spheres, but (as I noted about evocation) you'd have to radically rewrite the schools. Because currently they're a bunch of words that don't actually impose any constraints. "Manipulate magical energy"--every spell manipulates magical energy by definition. So either way, you're at "come up with a coherent set of descriptors, whatever you call them". Or give up on a framework entirely. Which (IMO) would stink.

Eh, while you'd have to reclassify some spells, Evocation "manipulates magical energy" in the sense that it creates fireballs, acid bolts, lightning swords, etc. If it's a construct or damage-expression in energy form, it's Evocation. This isn't any "reworking" of it, just a clarification of a short phrase. It's not "manipulating magical energy" in the broad sense you're interpreting it to be any more than using a disguise kit to create a false image of yourself as a nobleman is Illusion just because Illusion creates false images.

[PhoenixPhyre]

Eh, while you'd have to reclassify some spells, Evocation "manipulates magical energy" in the sense that it creates fireballs, acid bolts, lightning swords, etc. If it's a construct or damage-expression in energy form, it's Evocation. This isn't any "reworking" of it, just a clarification of a short phrase. It's not "manipulating magical energy" in the broad sense you're interpreting it to be any more than using a disguise kit to create a false image of yourself as a nobleman is Illusion just because Illusion creates false images.

But that's more than just a clarification--that's fundamentally different nature. And you'd have to do that to all schools (because they're all vague and jello). The point being is that the current schools aren't enough of a framework to be more than (slight) thematic support. And if you're rewriting them all fundamentally, tying yourself to that legacy baggage seems to be a bad idea. Because it sets expectations.

[Segev]

But that's more than just a clarification--that's fundamentally different nature. And you'd have to do that to all schools (because they're all vague and jello). The point being is that the current schools aren't enough of a framework to be more than (slight) thematic support. And if you're rewriting them all fundamentally, tying yourself to that legacy baggage seems to be a bad idea. Because it sets expectations.

Again, I disagree; they're pretty clear to me. I just don't see greater clarity in your proposal.

That said, more power to you if it works for you. Don't let me impede your fun. Sorry for raining on your parade. I'll bow out here.

[Garfunion]

I never really liked spell schools tied to spells. I would have rather have Wizard spell schools that add effects to all their spells. For example an evocation wizard adding a damage component to any spell they cast (psychic damage sleep spell) or transmutation wizard shaping the area of effect of their spells.

[TaiLiu]

I agree that the implementation is a little messy and ambiguous, but I think there's potential here. Focusing just on the Spheres:

1. As others have mentioned, some spells seem to fall into multiple spheres. In other words, the spheres are intended to be mutually exclusive, but don't actually seem to be.

2. There seems to be an ontological divide between the Spheres, so they seem inconsistent. Evocation is defined by how it works ("produces energy that..."), while Enchantment is defined by what it does ("altering the mental state..."). You should choose one or the other, I think, or break it up into two categories.

It might also make sense to reduce the number of spheres, just to make it easier for you and your players. Maybe a second very flawed go at Spheres would look like this:

a. Evocation—all spells in this category produce instantaneous effects.

b. Transmutation—all spells in this category produce permanent effects.

c. Enchantment—all spells in this category produce temporary effects.

Of course, with precision comes sterility. My second go seems to be missing a lot of the richness that 5e schools currently seem to have. I dunno if that's inevitable or not.

[PhoenixPhyre]

I agree that the implementation is a little messy and ambiguous, but I think there's potential here. Focusing just on the Spheres:

1. As others have mentioned, some spells seem to fall into multiple spheres. In other words, the spheres are intended to be mutually exclusive, but don't actually seem to be.

2. There seems to be an ontological divide between the Spheres, so they seem inconsistent. Evocation is defined by how it works ("produces energy that..."), while Enchantment is defined by what it does ("altering the mental state..."). You should choose one or the other, I think, or break it up into two categories.

It might also make sense to reduce the number of spheres, just to make it easier for you and your players. Maybe a second very flawed go at Spheres would look like this:

a. Evocation—all spells in this category produce instantaneous effects.

b. Transmutation—all spells in this category produce permanent effects.

c. Enchantment—all spells in this category produce temporary effects.

Of course, with precision comes sterility. My second go seems to be missing a lot of the richness that 5e schools currently seem to have. I dunno if that's inevitable or not.

1. Right. That's the challenge. But remember, the idea is to split the load between the sphere (singular) and the Aspects (multiple). So spells that fit in multiple spheres could either be
a) split into smaller, more single-purpose spells. Which generally isn't a bad thing IMO--spells work best IMO when they do one specific thing (fictionally speaking). Not amorphous "swiss army knife" spells. That also makes them easier to balance against each other, and increases opportunity cost. Win win win as far as I'm concerned. But more work.
b) have one sphere, but many different aspects. And which aspect(s) show up with detect magic would depend on the mode in which it was cast.

2. The wording for these (and even the basic categorization) was not intended to be more than just a vague handwaving starting point. The descriptions were intended as just starting points that need further refinement; it's the idea behind the 3-way split that I was mainly going for. Basic concept, not implementation. But yes.

As for richness--I think the aspects can serve most of that purpose. As they're intended to be much more granular and open-ended, more like "tags" than like hard categories, you can go a long way using that. Some abilities and features could key off of aspects (ignoring source and sphere); others might only care about sphere, while others might only care about source. Or any combination of them. So a wizard might get a bonus to casting Arcane Evocation spells, while a druid might get a bonus to casting Primal Conjuration (Fey-Aligned) spells. Or something.