D&D General What Should Magic Be Able To Do, From a Gameplay Design Standpoint?

[Corinnguard]

An alternative to a spellcasting skill:

Language as a skill for casting spells with a verbal spell component. As for the somatic spell component, there is the Performance skill. Lastly, the Nature or Survival skill for finding and preparing the right material component for the spell you want to cast.

 

[THEMNGMNT]

I'm not sure if this answers the OP or not...

I really like spellcasting in Shadow of the Demon Lord/Weird Wizard. It's grouped around themes. So there's necromancy, fire magic, air magic, healing, time magic, rune magic, etc. If you're a spellcaster, you get to pick a limited number of themes. What that means is you can be really good at some stuff, but not all the stuff. You probably can't cast time stop, fireball, cone of cold, synaptic static, and fly. You've got to pick a couple.

To me, this approach sets a boundary around on spellcaster power and also better models genre fiction.

I wish this is how D&D spellcasting worked.

 

[aco175]

I could limiting the selection of spells in the PHB to only several per level. Take the most common or basic for each class level and have those as the ones to choose from each level and have the others as rare spells you find or research. This limits the power of all spells, but may also be a penalty to casters depending on how each table plays.

 

[EzekielRaiden]

In brief?

Dramatically less than what it can currently do, at least in the hands of any single caster.

Having magic that can do lots of powerful things is fine. Having magic that lets any single character do lots of powerful things is less fine.

Dreamscarred Press's Spheres of Power rules (for PF1e, I haven't used the 5e ones) are a beautiful example of how to preserve most of the interesting utility and diversity of magic, while very thoroughly rebalancing things so casters DO NOT rule the roost.

TL;DR: The titular Spheres are sort of like Priest "spheres of influence" from 2nd edition. You must know the base sphere before you gain access to more powerful things with it. Further, you must invest Talents into those spheres in order to actually learn those more powerful things, and using the more advanced abilities costs Spell Points. Both talents and SP are limited resources, so most spellcasters need to specialize in order to be truly effective. However, zillions of specializations are possible and any single character might finish their 1-20 run with several distinct specialties, perhaps 5-7 (assuming a full-caster class with lots of talents, like Wizard or its pure-sphere equivalent, the Incanter).

As a great example, the Light sphere is a surprisingly diverse and strong option, with a great variety of offense, buff, and debuff effects, and plenty of non-combat utility. Alteration (the equivalent of transformation spells in 3.x/PF1) is a powerful but expensive school in terms of both talents and SP, you get what you pay for. However, if you choose to play as the Shifter class, you naturally get Alteration sphere stuff (and ways to make it better/easier/cheaper), and you can take an archetype that gives you Light sphere powers as well, meaning you can specialize in just those two spheres and actually be a quite effective character, or you can start with a solid foundation of those spheres, and grow into new ones as you gain levels and thus more talents.

Magic has plenty of flaws in the 3e/PF1e/5e structure. But its greatest flaw is being able to do so many powerful things all with a single character, and often all during a single day, or at least with 24 hours' notice.

 

[TiQuinn]

My biggest issue with D&D magic is that certain spells undermine rules in the game that I think makes for less interesting games, or narrows the kinds of games you play.

Examples: Guidance in 5e as this always available spell that is just an immediate bonus - why wouldn't you cast it as much as possible? There's no constraint on it's use.

Silvery Barbs - A frankly bland spell that serves only to deny one creature a successful roll and give advantage to another. It's a non-interesting mechanic that leans into advantage - a mechanic that I'm feeling gets overused.

Zone of Truth - A single spell that needs to be considered any time you want to run a mystery game due to the high possibility of circumventing any mystery.

Tiny Hut has been discussed to death. To me, it trivializes any kind of danger or risk involved with resting with little downside. A 3rd level spell that could easily be a 6th level spell.

 

[Indagare Nogitsune]

Magic should allow players to do something remarkable while also ensuring other players feel needed as well. A basic question about any spell ought to be "will this be fun for anyone but me?"

So, for instance, the party lacks a Rogue. A Wizard with the right spell selection could do that role. However, if the party has a Rogue, the Wizard's Player should work with the Rogue's Player to ensure the Rogue isn't overshadowed.

This isn't that hard, though. A Wizard {Rogue} could do things with the official Rogue. Or could take up almost any other role needed except healer.

Also, anything a PC can do with magic an NPC should be able to do as well. I think that this will get overlooked if we just focus on the PCs. They can, even should, face enemy spellcasters. Those spellcasters ought to be able to cast anything the PCs can cast.

 

[Kurotowa]

I'm firmly of the school of "Limits breed creativity". Having constraints on your tools or resources forces you to get inventive to accomplish your goals. And being inventive is more fun and satisfying than just hitting your simple obvious "I Win" button. Therefore, I agree with others that while magic should be able to do almost anything, a specific character should have a much more limited array of options.

If I could make one change to D&D's magic system, it would be a clearer delineation between spells that are your highly optimized field combat tools, and ...let's call it ritual magic where you do a big slow magical working for less immediate and clearly defined results. I would love it if there were clearer guidelines for ritual magic type deals where you're trying to ward a palace or bless a bloodline, rather than being a life or death battle with a gargoyle.

 

[Tonguez]

Magic should require the same Skill test as any other class attempting to do something given enough time, personnel and technology - fireball is found in a vial of alchemist fire, planeshift generates enough energy to open a planar wormhole, even reversing gravity is an advance use of magnetogravitic tech

 

[EzekielRaiden]

Also, anything a PC can do with magic an NPC should be able to do as well. I think that this will get overlooked if we just focus on the PCs. They can, even should, face enemy spellcasters. Those spellcasters ought to be able to cast anything the PCs can cast.

This principle is not as universal as you are claiming.

Perfect symmetry between PC and NPC rules has a very key issue: NPCs do not have the same gameplay purpose, contribution to the experience, nor (in essentially all cases) length or quantity of presence in the gameplay as PCs do. Demanding that anything PCs can do, NPCs can do, always no matter what, is a recipe for many ills, among them DM burnout, DM-player arms races (e.g. "scry vs scry" behavior and the resulting never-ending cascade of magic defenses and counters), and severe issues with encounter design (because NPC spellcasters strong enough to not keel over from a particularly vehement PC fart will be bringing far too much magic mojo, able to drop an entire day's worth of spells in a single encounter.)

It's a lovely idea. In a purified no-story, all-"come what may", sandbox/hexcrawl game experience, it's even productive, because the whole point there is that there is no arc, no point nor direction, there is only the places you chose to go and the things you happened to see (or not see) there. But such an aggressively "no story, just events" playstyle is pretty unforgiving, and most players are looking for at least some degree of satisfying narrative or at least satisfying conclusions, which requires that the experience be shaped to suit that end, to at least some extent. That doesn't mean this playstyle should be left by the wayside, to be clear, but it should be seen for what it is: an uncompromising requirement that, if followed in full, tends to exclude a number of things players really enjoy about playing tabletop RPGs.

 

[payn]

This principle is not as universal as you are claiming.

Perfect symmetry between PC and NPC rules has a very key issue: NPCs do not have the same gameplay purpose, contribution to the experience, nor (in essentially all cases) length or quantity of presence in the gameplay as PCs do. Demanding that anything PCs can do, NPCs can do, always no matter what, is a recipe for many ills, among them DM burnout, DM-player arms races (e.g. "scry vs scry" behavior and the resulting never-ending cascade of magic defenses and counters), and severe issues with encounter design (because NPC spellcasters strong enough to not keel over from a particularly vehement PC fart will be bringing far too much magic mojo, able to drop an entire day's worth of spells in a single encounter.)

It's a lovely idea. In a purified no-story, all-"come what may", sandbox/hexcrawl game experience, it's even productive, because the whole point there is that there is no arc, no point nor direction, there is only the places you chose to go and the things you happened to see (or not see) there. But such an aggressively "no story, just events" playstyle is pretty unforgiving, and most players are looking for at least some degree of satisfying narrative or at least satisfying conclusions, which requires that the experience be shaped to suit that end, to at least some extent. That doesn't mean this playstyle should be left by the wayside, to be clear, but it should be seen for what it is: an uncompromising requirement that, if followed in full, tends to exclude a number of things players really enjoy about playing tabletop RPGs.

Not my experience at all in over a decade of running APs in PF1, but im not getting into it. I love me some NPcs made just like PCs and never experienced the issues (except increased workload, which I developed some great work arounds). So, none of this is to be taken as matter of fact, but as potential pain points.