The cost of magic

[Kapow]

First,
Sorry, that I started this thread and than didn't participate.
Life just hit me hard.

Second,
Lots and lots of really good arguments and ideas here, even in the derails. I can't possibly react to everything.


Back to the topic,

I realized early in the discussion, that, what I consider a cost of magic, may be hard to implement in a class- and level-based system.
All systems (with the exception of the Warhammer systems) I had in mind are point-buy and organic leveling.
So perhaps that's a big part of it?!

What I don't really get is the argument, that, if magic has a chance of failure, it makes mages "un-fun" to play, because the core-ability of the Archetype should just work.
The thief can fail at picking locks, the warrior can fail at hitting enemies, the face can fail at influencing people... Why should the mage be an exception?
Now, I don't say, that death, crippling injury and/or madness should be common consequences (or even at all).

That's why I do agree with the last few posts, that the possibility of balancing risk and result is needed, if you want to implement more serious costs.
In my experience, in most games, that is the case. You can find ways around the consequences like death, but at a cost.
That may be lower power, it may be investing more resources, e.g. skill points, or...
And here I think, that organic/point-buy leveling has it easier.
The cost of magic will be manifest, by either a great risk or by having not being able to invest in all the other stuff (skills, feats, special abilities...) you could get for your character building/developing ressources.

[InvisibleBison]

What I don't really get is the argument, that, if magic has a chance of failure, it makes mages "un-fun" to play, because the core-ability of the Archetype should just work.
The thief can fail at picking locks, the warrior can fail at hitting enemies, the face can fail at influencing people... Why should the mage be an exception?
Now, I don't say, that death, crippling injury and/or madness should be common consequences (or even at all).

I think it's fine for magic to have a chance of failure. I just don't like the failure to be failing to actually cast the spell. The rogue can fail to pick locks, but she never fails to attempt to pick locks; the warrior can fail to hit enemies, but he never fails to swing his sword, the face can fail to influence people, but he never fails to speak with them. If the mage isn't to be an exception, his magic should always work, but sometimes fail to have the desired effect.

[Xervous]

I think it's fine for magic to have a chance of failure. I just don't like the failure to be failing to actually cast the spell. The rogue can fail to pick locks, but she never fails to attempt to pick locks; the warrior can fail to hit enemies, but he never fails to swing his sword, the face can fail to influence people, but he never fails to speak with them. If the mage isn't to be an exception, his magic should always work, but sometimes fail to have the desired effect.

And by fail you mean come up short rather than take MightAsWellRerolld6 damage because lolrandom?

[InvisibleBison]

And by fail you mean come up short rather than take MightAsWellRerolld6 damage because lolrandom?

Yeah, magic failing should be "The enemy dodged your firebolt" or "He resisted the mind control", not "You accidentally set yourself on fire" or "A giant demon appears and eats your face". Just like how a fighter failing is "You missed the enemy" and not "You accidentally stab yourself".

[Max_Killjoy]

Yeah, magic failing should be "The enemy dodged your firebolt" or "He resisted the mind control", not "You accidentally set yourself on fire" or "A giant demon appears and eats your face". Just like how a fighter failing is "You missed the enemy" and not "You accidentally stab yourself".

Agreed.

To me, D&D doesn't feel like it has a good unified system for resisting / defending against magic. Saving throws are kinda a thing, but it feels like every spell has its own little ruleset including maybe rules for resisting or partially resisting.

[Segev]

Agreed.

To me, D&D doesn't feel like it has a good unified system for resisting / defending against magic. Saving throws are kinda a thing, but it feels like every spell has its own little ruleset including maybe rules for resisting or partially resisting.

Eh. It uses one of three rule sets, all of which are pretty well-defined:

1. Attack roll
2. Save for none
3. Save for partial

At its most complex, it might have an attack roll with an additional save for another effect riding on it if the attack roll hits.

[Composer99]

First,
Sorry, that I started this thread and than didn't participate.
Life just hit me hard.

Second,
Lots and lots of really good arguments and ideas here, even in the derails. I can't possibly react to everything.


Back to the topic,

I realized early in the discussion, that, what I consider a cost of magic, may be hard to implement in a class- and level-based system.
All systems (with the exception of the Warhammer systems) I had in mind are point-buy and organic leveling.
So perhaps that's a big part of it?!

What I don't really get is the argument, that, if magic has a chance of failure, it makes mages "un-fun" to play, because the core-ability of the Archetype should just work.
The thief can fail at picking locks, the warrior can fail at hitting enemies, the face can fail at influencing people... Why should the mage be an exception?
Now, I don't say, that death, crippling injury and/or madness should be common consequences (or even at all).

That's why I do agree with the last few posts, that the possibility of balancing risk and result is needed, if you want to implement more serious costs.
In my experience, in most games, that is the case. You can find ways around the consequences like death, but at a cost.
That may be lower power, it may be investing more resources, e.g. skill points, or...
And here I think, that organic/point-buy leveling has it easier.
The cost of magic will be manifest, by either a great risk or by having not being able to invest in all the other stuff (skills, feats, special abilities...) you could get for your character building/developing ressources.

So I think it might be helpful to try to define what sort of costs there might be to learning or using magic, in as system- and setting-agnostic a manner as we can.

Learning magic has, or can have:
- An opportunity cost, in the form of in-game time or system-level build resources your character spends learning how to use magic instead of learning how to do something else.
- An in-game resource cost - spending money to go to a magical academy, say,
- An in-game social cost - people known to be magic-users might be viewed with suspicion or hostility.

Using magic has, or can have:
- An opportunity cost. You used your game-mechanic time or activity unit to cast this spell, instead of casting that one or doing something non-magical, no backsies!
- An in-game or system-level resource cost, such as money spent on components, spell slots, and the like.
- Edit to add: A cost for overuse, insofar as this is distinct from a resource cost. (Overuse by running out of spell slots or going "OOM" (out of mana/magic points) is running out of resources, for instance.) This could be Strain/Drain/Taint, some sort of burn-out, all the way up to some kind of nasty end.
- A chance of failure or of an adverse effect, whether that's you missing your attack roll, your targets shrugging off the effect of your spells, you turning into a potted plant for 10 minutes, all the way up to your mind shattering, you exploding as a rift to another reality opens up, or you inadvertently summoning a TPK monster.
- An in-game social cost - people known to be magic-users might be viewed with suspicion or hostility, or casting particular spells or from particular categories of spell might be grounds for imprisonment or execution or tend to bring out torch-and-pitchfork-wielding mobs.

Depending on the game system and/or the setting/game world, these costs may or may not come across as "real" costs (which I started to try to define but struggled sufficiently with that I'm resorting to "you'll know them when you see them"). Also, I am sure there are other possible costs that I have left out.

---

When we start looking at different RPG systems and different settings with magic that have been represented in RPGs, we can see that there really is no sensible way to say what ought or ought not to be with respect to the costs of magic writ large, not only because of the enormous diversity of systems and settings, but also because different players and tables might have very different preferences.

Imagine a game using the magic of, say, the Harry Potter stories, contrasted with Call of Cthulhu. In one, magic is really quite ordinary in application, with a well-suited wand being all you need for almost all magic, both great and small. In the other, using magic means tinkering or playing with Things That Mortals Were Not Meant To Know, with almost inevitable mind-shattering results. In a Harry Potter game, you're meant to be a magic-user, and in Call of Cthulhu, you really shouldn't be messing with Things That Mortals Weren't Meant To Know at all, and the game rules are there to tell you what horrors lie in store for you if you do. In either game, it kind of defeats the purpose of playing in that setting to change the way magic works.

In a game like D&D 5e, most (all?) settings assume costs somewhere in between Harry Potter and Call of Cthulhu. Spell slots (or, in 4e, usage limits on powers) ensure that you can't spam insta-kill spells in D&D 5e the way you could spam the Killing Curse, and the base rules don't expect casting a spell entails risking permanent madness - mostly, there's a chance the spell will fail if you miss your attack or a target makes its save, and it's only when you're playing with wild magic that there's an off chance you'll blow up or whatnot (some specific spells, such as haste, contact other plane, or wish, notwithstanding).

Unlike Harry Potter or Call of Cthulhu, a game such as D&D encompasses a wide variety of settings. It would be nice if the game included more official support for different "cost structures", as it were, for learning or using magic, in the form of optional/variant rules in a DMG or supplement. Some settings in the game formally include alternate "cost structures", either intermittently (the occasional "dead magic" or "wild magic" zone in post-Time-of-Troubles Forgotten Realms, for instance) or permanently (Dark Sun). Or you could go the other way, much as 4e's at-wills and PF/5e's cantrips are less costly compared to older editions (Prestidigitation was once a 1st-level spell, for instance, and even in 3.X when it was a cantrip, you still had a slot cap). The default can't really be said to be either good or bad, as such - only satisfactory or unsatisfactory for players depending on their preferences.

Now, on top of all that, there's also the matter of how magical and non-magical characters and their capabilities, taken together, fit in with the game setting, or how they compare to one another with respect to game mechanics. In a Harry Potter game, playing a non-magical character is an intentional step down. In Call of Cthulhu, it's the other way 'round (there are enough ways for your sanity to be shattered without layering on trying to cast a spell on top of it, thank you very much!). In D&D 5e, I would say that the design intent is broadly (a) for a mix of magical and non-magical characters to form an adventuring party, with each member contributing roughly equally to the success of the group as a whole over a sufficiently large period of time, all else being equal, and (b) in any given formally-published module, for a mix of magical and non-magical characters whose exact composition cannot be guessed in advance to be able to successfully complete the module. How well 5e fulfills this intent is up for debate, and one's view on the matter might colour whether or how one thinks the D&D magic "cost structure" ought to be modified.

[Pex]

First,
Sorry, that I started this thread and than didn't participate.
Life just hit me hard.

Second,
Lots and lots of really good arguments and ideas here, even in the derails. I can't possibly react to everything.


Back to the topic,

I realized early in the discussion, that, what I consider a cost of magic, may be hard to implement in a class- and level-based system.
All systems (with the exception of the Warhammer systems) I had in mind are point-buy and organic leveling.
So perhaps that's a big part of it?!

What I don't really get is the argument, that, if magic has a chance of failure, it makes mages "un-fun" to play, because the core-ability of the Archetype should just work.
The thief can fail at picking locks, the warrior can fail at hitting enemies, the face can fail at influencing people... Why should the mage be an exception?
Now, I don't say, that death, crippling injury and/or madness should be common consequences (or even at all).

That's why I do agree with the last few posts, that the possibility of balancing risk and result is needed, if you want to implement more serious costs.
In my experience, in most games, that is the case. You can find ways around the consequences like death, but at a cost.
That may be lower power, it may be investing more resources, e.g. skill points, or...
And here I think, that organic/point-buy leveling has it easier.
The cost of magic will be manifest, by either a great risk or by having not being able to invest in all the other stuff (skills, feats, special abilities...) you could get for your character building/developing ressources.

There's nothing wrong with magic failing. In D&D that's missing the attack roll or opponent makes the saving throw. In GURPS it's rolling above the target number. In Ars Magica it's rolling below the target number. That's not unfun. What's unfun is making the player miserable for doing what you said he could do. Cast a spell you die is not a fair balancing factor. Doesn't matter how mitigating you can make it by lowering the power. You said the player can do that high power big boom spell, he does it, and now he has to make a new character. That's the adding insult to injury. That's the unfun. If you don't want the player to do that high power big boom spell, then don't have it in the game at all.

[Telok]

Yeah, magic failing should be "The enemy dodged your firebolt" or "He resisted the mind control", not "You accidentally set yourself on fire" or "A giant demon appears and eats your face". Just like how a fighter failing is "You missed the enemy" and not "You accidentally stab yourself".

I dunno. What's the difference between fighters missing a 2d6 attack 20% of the time and mages failing to cast a 2d6 damage spell 20% of the time?

Maybe you object to the mage only being able to cast 5/day and a 20% fail casting check and a 20% miss attack roll and a 20% save-negates from the target. But isn't that the same as fighters only getting to swing 5/day and miss 20% and armor negates it 20% and the opponent parries 20%?

D&D says something like "fireball 3/day, 8d6, about 4 targets, save for half", tho 4e was more like "fireball 1/fight, 6d6, about 4 targets, make attack rolls". Fireball vs. swording balance in D&D is predicated on fewer fireballs than swordings. If a game system is all at-will magic and no save for half how are you balancing fireballs against swords? Going "fireball at-will, 8d6, about 4 targets, no save" has obvious issues (not including stuff like 'Mage: the foo' games where everyone is a super powered caster). Trying to balance that against fighter 2d6 on 1 foe at a time calls for something like only being able to cast it once every 9 rounds or something. In that case the "roll to cast" is just shifting the fiction layer description without changing the mechanical outcome.

If there's no difference in the mechanics of effect between "roll skill to cast damage", "roll attack to inflict damage", and "roll save to avoid damage" then where's the issue? Is it just people wanting one style of fiction rather than the other?

Obviously a side effect system is different from this. I'm just talking the "roll to cast" bit here.

[PhoenixPhyre]

I dunno. What's the difference between fighters missing a 2d6 attack 20% of the time and mages failing to cast a 2d6 damage spell 20% of the time?

Maybe you object to the mage only being able to cast 5/day and a 20% fail casting check and a 20% miss attack roll and a 20% save-negates from the target. But isn't that the same as fighters only getting to swing 5/day and miss 20% and armor negates it 20% and the opponent parries 20%?

D&D says something like "fireball 3/day, 8d6, about 4 targets, save for half", tho 4e was more like "fireball 1/fight, 6d6, about 4 targets, make attack rolls". Fireball vs. swording balance in D&D is predicated on fewer fireballs than swordings. If a game system is all at-will magic and no save for half how are you balancing fireballs against swords? Going "fireball at-will, 8d6, about 4 targets, no save" has obvious issues (not including stuff like 'Mage: the foo' games where everyone is a super powered caster). Trying to balance that against fighter 2d6 on 1 foe at a time calls for something like only being able to cast it once every 9 rounds or something. In that case the "roll to cast" is just shifting the fiction layer description without changing the mechanical outcome.

If there's no difference in the mechanics of effect between "roll skill to cast damage", "roll attack to inflict damage", and "roll save to avoid damage" then where's the issue? Is it just people wanting one style of fiction rather than the other?

Obviously a side effect system is different from this. I'm just talking the "roll to cast" bit here.

I don't "roll to cast" IFF all of the following are true
* The number of rolls-to-resolve (per target) is similar to doing something non-magical. If martials get "roll to attack", but on a success they deal damage, while casters get "roll to cast" + "enemy rolls to avoid", that's a problem. Or vice versa. Current D&D has a single roll-to-hit, so if you went all 4e and said that the attack rolls-to-hit on spells as well, I'd be ok with that.
* The resource usage is similar, modulated by effect size. If a caster has to roll to cast his basic attack spell plus using a resource on success or failure, while the martial only has to roll to hit (but doesn't spend a resource), both for similar effect sizes, that's a problem. And vice versa. So being able to cast at will (modulo rebalancing effect sizes) gated behind a skill check to cast might be ok. Or a spell slot + roll to cast, except you only spend the slot if you actually cast it successfully.
* Failure doesn't mean anything more than failure to get the result you want (and potentially burning a resource and/or time). Actively punishing the caster or especially the party for failing to cast means that the whole mechanic will either not get used or will get optimized to a binary--you only cast the spells you're guaranteed to succeed on. And in any case, table lookups are slow. Please don't make me do contingent lookups[1] every time someone casts a spell!

[1] ie doing different things based on the roll, such as "failed to cast == roll on table ABC to see what consequences happen", while "successfully cast == roll damage XYZ/impose effect ABC". That kind of flow-chart based logic burns table time and annoys me like nothing else.

[Xervous]

[1] ie doing different things based on the roll, such as "failed to cast == roll on table ABC to see what consequences happen", while "successfully cast == roll damage XYZ/impose effect ABC". That kind of flow-chart based logic burns table time and annoys me like nothing else.

So something like PF2’s degrees of success on spells is a no go for you?

[InvisibleBison]

I dunno. What's the difference between fighters missing a 2d6 attack 20% of the time and mages failing to cast a 2d6 damage spell 20% of the time?

There's no difference between these two mechanics. What you're describing is fine. What I've been objecting to is when failing to cast a spell has additional bad consequences beyond the spell not working.

[PhoenixPhyre]

So something like PF2’s degrees of success on spells is a no go for you?

Depends on the simplicity. If it's written into the spell , then it's (mostly) ok, if limited to a few branches without further rolling (determined entirely by result - TN). But if I have to go find a table and roll on it, especially if there are modifiers to that roll or other mechanics that interact with that secondary roll, no thanks.

So "missed? Q. over by X? a. Over by Y > X? B. Over by Z or more, C", where A, B, C, and Q are things that don't need separate resolution is marginally ok. Annoying and slow, but I can tolerate it.

But "missed? Go roll on table 42-a, using A and B as modifiers, and if they did PDQ, add 3 and if Jupiter is in Sagittarius, take the square root of the result, coercing to integer" is very much not ok. At least if that instruction comes up more than about 1x per campaign and isn't the result of an intentional, plot important choice.

[Quertus]

I think it's fine for magic to have a chance of failure. I just don't like the failure to be failing to actually cast the spell. The rogue can fail to pick locks, but she never fails to attempt to pick locks; the warrior can fail to hit enemies, but he never fails to swing his sword, the face can fail to influence people, but he never fails to speak with them. If the mage isn't to be an exception, his magic should always work, but sometimes fail to have the desired effect.

Yeah, magic failing should be "The enemy dodged your firebolt" or "He resisted the mind control", not "You accidentally set yourself on fire" or "A giant demon appears and eats your face". Just like how a fighter failing is "You missed the enemy" and not "You accidentally stab yourself".

If 4e taught us anything, it's that "rolling dice" is synonymous with "playing the game".

So, the Fighter should have to…

Roll to unsheathe their sword.

Roll to grip the sword.

Roll to remember the correct forms in the heat of battle.

Roll to perform the maneuver.

Roll to choose targets (a botch here could result in hitting themselves or an ally).

Then, finally, get to roll to hit.

Personally, I don't have anything against TPK demons eating the party's face after the wizard miscasts, or after the Fighter miss-swings. I think it's great to have the Fighter have to ask if it's really worth it to swing that sword, or whether maybe they shouldn't give Diplomacy one more chance.

But if you're playing a Fighter because you enjoy swinging a sword, *and* don't like to take such risks, well, maybe that's not the game for you.

[Telok]

Depends on the simplicity. If it's written into the spell , then it's (mostly) ok, if limited to a few branches without further rolling (determined entirely by result - TN). But if I have to go find a table and roll on it, especially if there are modifiers to that roll or other mechanics that interact with that secondary roll, no thanks.

So "missed? Q. over by X? a. Over by Y > X? B. Over by Z or more, C", where A, B, C, and Q are things that don't need separate resolution is marginally ok. Annoying and slow, but I can tolerate it.

But "missed? Go roll on table 42-a, using A and B as modifiers, and if they did PDQ, add 3 and if Jupiter is in Sagittarius, take the square root of the result, coercing to integer" is very much not ok. At least if that instruction comes up more than about 1x per campaign and isn't the result of an intentional, plot important choice.

Ah, I think I see. Extra penalty on failure and bad tables, right?

How about failure to cast having no penalty, a 100% safe lower power casting option, regular casting having 1/1000 or less chance of the penalty plus only happening on extra good rolls and ways to mitigate even that, and a high power "overcharge" casting option that did invoke the penalty? Couple it to something where "saves", when they came up, were rolled against the "to cast" total (incentive to sometimes want to keep the high rolls but not a need to), and the bad effect tables are a button built into the game book pdf with results that never make you do more rolls or look-ups. More acceptable?

Really it's OK to not like a particular balancing set up. I'm not a fan of what D&D 4e did. But I try to assess the magic systems based on the game they're part of. D&D is a very "no drawbacks" system, CoC is all "magic bad, but you might use it once an adventure", WH40K is "dangerous but powerful", and most supers systems are "build your own casting system". So D&D 4e succeeded in it's mechanical goals, I just don't like what I see as it's making magic into another way to sword things.

[PhoenixPhyre]

Ah, I think I see. Extra penalty on failure and bad tables, right?

How about failure to cast having no penalty, a 100% safe lower power casting option, regular casting having 1/1000 or less chance of the penalty plus only happening on extra good rolls and ways to mitigate even that, and a high power "overcharge" casting option that did invoke the penalty? Couple it to something where "saves", when they came up, were rolled against the "to cast" total (incentive to sometimes want to keep the high rolls but not a need to), and the bad effect tables are a button built into the game book pdf with results that never make you do more rolls or look-ups. More acceptable?

Really it's OK to not like a particular balancing set up. I'm not a fan of what D&D 4e did. But I try to assess the magic systems based on the game they're part of. D&D is a very "no drawbacks" system, CoC is all "magic bad, but you might use it once an adventure", WH40K is "dangerous but powerful", and most supers systems are "build your own casting system". So D&D 4e succeeded in it's mechanical goals, I just don't like what I see as it's making magic into another way to sword things.

Personally, still WAY too much mechanics for me. And anything that requires tool assistance is a non-starter for me. Because I'd much prefer if the books (including the pdfs) stayed away from the table entirely. And electronics especially. If players have to have electronic devices with particular versions of the thing up just to run their character, no thanks. And inevitably it's one more thing I'd have to juggle as DM. Especially since those tables strongly constrain the acceptable worldbuilding.

Hot-path (ie the ones you do multiple times per session) resolution mechanics IMO should be dead simple, even if that means you sacrifice some fidelity or precision. One roll against a static TN (I don't care as much the details of that roll, whether dice pool, roll and keep, or flat dice + modifier) or at most an opposed roll. And preferably any modifiers are written down and only change when the character gains power (ie between levels or on buying new stuff in a point-buy). Everything that can be precalculated should be. Anything requiring multiple rolls or chained logic/decisions during resolution itself is right out.

And honestly, I'd prefer just flat out reducing the power of spells/magic straight up, at least in a D&D context, rather than bodging on some kind of penalty and adding to the cognitive load that a spell-caster has to juggle. Because that kind of load makes their turns (already the longest turns generally, due to the plethora of options) exponentially longer and makes playing them even less new-player-friendly. And D&D spells are already too strong/versatile in many cases IMO.

In a game where magic was the exception (ie no caster "classes", magic is not assumed), there might be some wiggle room. Because then spells wouldn't be a hot-path issue. They'd be effectively a plot point. Do you do this ritual (which is its own part of the adventure with multiple steps and choices that everyone gets involved in)? What steps do you take to keep it safe? Do you just wing it? What kinds of collateral damage are you willing to accept? Etc.

[Anymage]

In a game where magic was the exception (ie no caster "classes", magic is not assumed), there might be some wiggle room. Because then spells wouldn't be a hot-path issue. They'd be effectively a plot point. Do you do this ritual (which is its own part of the adventure with multiple steps and choices that everyone gets involved in)? What steps do you take to keep it safe? Do you just wing it? What kinds of collateral damage are you willing to accept? Etc.

I really want to second this.

Casting spells is your main shtick and you're expected to do it regularly during every play session? The worst sort of backlash system I can get behind is Shadowrun, where you might get a bit winded and can risk an injury to push through more power. Spellcasting might be more or less reliable, but any harm from failure should be pretty token.

Big plot spells? We had the talk about Ritual Caster quite a few pages back, and here I'm okay with high risks and high costs. You want to summon and bind a powerful demon to serve you? I'm totally okay with that demon coming out to eat your face if you bung the binding roll. That's not regretting having taken an action when your turn in combat came around, that's taking a plot risk and finding that it isn't guaranteed to go your way.

[Duff]

For D&D, there are costs to most magic. But in most D&D the costs are "off screen".
For better and for worse.
Wizards spend years training to learn their art
Warlock mortgage their soul with the posability that their patron will demand things of them
Clerics must do their duties to their god (spread the word, lead the rituals in temple etc)

Sorcerers don't have any of those costs.

So ...
If you wanted to play D&D but have costs, the least work option would be to simply bring those costs onto the stage. Wizards need to get their down time before they can learn new spells, Warlocks patrons tell them to do things - maybe errands, maybe things they don't want to, maybe plot related.
Clerics loose time and or money on their religious duties.

Maybe have sorcerers pay the price through a complicated family life - You're family has history and you have to deal with that.

If you want to balance that, it might be as simple as "full casters have an advantage anyway, no change made" or might be some extra spells/uses/powers.

For other games theres the whole gamut available - cost can be consistant or variable and can be always paid in the same "currency" or it might be different for every single casting.

[Telok]

Personally, still WAY too much mechanics for me.

Yeah, sounds like D&D magic is pushing the limits of your preferences for complexity and rolls. That's fine, those are your preferences. I've taken the magic system I described and had it run faster and slicker than any D&D magic but 4e. But that's also because I've worked to streamline it a bit and rewrote the bad effects tables to be more in tune with the game flow and less random chaos instant death. It's not something I would ever try to stick into a D&D game because D&D doesn't do "cost of magic" any more.

Perhaps the important thing is it's a magic costs system fit to the game system and setting (plus my version is now set up so the player says "I overcast by 40" and I punch 4-0-enter on a tablet then read off the result, less work than some D&D saves+resists). I mean, personally I'm no longer sold on D&D magic simply because the "cost" of phenominal cosmic power is having 2/3rd the hit points of a fighter and not using a two handed sword (and even that's negotiable with multiclass/magic items).

However that's just a preference. The D&D magic system works (in general, outside of specific spells and occasional gaps in the casting rules) for D&D. The CoC magic system works for CoC. The WH40K magic system works in those games (with a few similar issues to D&D of course). I wouldn't try swapping the magic systems without expecting massive shifts in the tone and style of the games. Which, naturally, some people won't like because they play those games for that style of play.

[Pex]

I dunno. What's the difference between fighters missing a 2d6 attack 20% of the time and mages failing to cast a 2d6 damage spell 20% of the time?

In D&D, the %miss chance of spellcasting is missing the attack roll or opponent making the save. It already exists as equivalent %miss chance as the warriors. What you're asking for is an additional miss chance of the spell not working at all and then if it does work the miss chance of it working comes into play. That's not fair. The warrior doesn't have to roll first to see if he could swing a weapon before swinging his weapon.

I really want to second this.

Casting spells is your main shtick and you're expected to do it regularly during every play session? The worst sort of backlash system I can get behind is Shadowrun, where you might get a bit winded and can risk an injury to push through more power. Spellcasting might be more or less reliable, but any harm from failure should be pretty token.

Big plot spells? We had the talk about Ritual Caster quite a few pages back, and here I'm okay with high risks and high costs. You want to summon and bind a powerful demon to serve you? I'm totally okay with that demon coming out to eat your face if you bung the binding roll. That's not regretting having taken an action when your turn in combat came around, that's taking a plot risk and finding that it isn't guaranteed to go your way.

That's why I never cast Planar Binding in D&D. I don't like such drawbacks, but I can get over it if it's for very specific niche spells that are more often BBEG plot devices. My character never casts such spells and plays normally, but I'll continue to grumble about those spells. When such drawbacks of similar effects based on the spell is a possibility for every spell you attempt to cast that's where it's punishing the player for doing what he's supposed to be doing. That is not a fair balancing factor.

For D&D, there are costs to most magic. But in most D&D the costs are "off screen".
For better and for worse.
Wizards spend years training to learn their art
Warlock mortgage their soul with the posability that their patron will demand things of them
Clerics must do their duties to their god (spread the word, lead the rituals in temple etc)

Sorcerers don't have any of those costs.

So ...
If you wanted to play D&D but have costs, the least work option would be to simply bring those costs onto the stage. Wizards need to get their down time before they can learn new spells, Warlocks patrons tell them to do things - maybe errands, maybe things they don't want to, maybe plot related.
Clerics loose time and or money on their religious duties.

Maybe have sorcerers pay the price through a complicated family life - You're family has history and you have to deal with that.

If you want to balance that, it might be as simple as "full casters have an advantage anyway, no change made" or might be some extra spells/uses/powers.

For other games theres the whole gamut available - cost can be consistant or variable and can be always paid in the same "currency" or it might be different for every single casting.

So the cost of playing a spellcaster is not being able to play the game because you have to study or deal with endless frustrations. That's unfun.

However, that is what Ars Magica allows for. Everyone actually has two characters - their Magus and their Companion who is not a spellcaster. The idea is you play your Companion for a while because your Magus is studying. Out of all the party only one player plays a Magus, and you take turns who plays their Magus. If the game system is built around this it may work, but not for people who want to play a spellcaster all the time and they're not having BadWrongFun for wanting that. Ars Magica doesn't have to play this way. When I played it everyone was a Magus, and a player played his Companion when he felt like it.

[Satinavian]

In D&D, the %miss chance of spellcasting is missing the attack roll or opponent making the save. It already exists as equivalent %miss chance as the warriors. What you're asking for is an additional miss chance of the spell not working at all and then if it does work the miss chance of it working comes into play. That's not fair. The warrior doesn't have to roll first to see if he could swing a weapon before swinging his weapon.

That only works for damage spells and debuffs. Utility, buffs, healing, summoning etc. don't have such opposition.

[Pex]

That only works for damage spells and debuffs. Utility, buffs, healing, summoning etc. don't have such opposition.

Yes, and nonspellcasters have their own autosuccess Nice Things such as Advantage when you want it, can't roll below 10, gain two extra attacks, and take another Action in addition to healing yourself.

[Max_Killjoy]

I really want to second this.

Casting spells is your main shtick and you're expected to do it regularly during every play session? The worst sort of backlash system I can get behind is Shadowrun, where you might get a bit winded and can risk an injury to push through more power. Spellcasting might be more or less reliable, but any harm from failure should be pretty token.

Big plot spells? We had the talk about Ritual Caster quite a few pages back, and here I'm okay with high risks and high costs. You want to summon and bind a powerful demon to serve you? I'm totally okay with that demon coming out to eat your face if you bung the binding roll. That's not regretting having taken an action when your turn in combat came around, that's taking a plot risk and finding that it isn't guaranteed to go your way.

Even if casters are a player thing, I think scaling the effort and cost in proportion to the spell's power is worthwhile.

Everyday spells vs big plot spells should, IMO, be a dividing line in many more settings/systems than it is right now.

[Glorthindel]

I fall very much on the side that having serious (and potentially character-ending) drawbacks can be very fun, and certainly enough systems are out there do it, and are highly successful (WFRP / Dark Heresy are successful games and lean heavily on this for their casters).

In my mind, there has to be three specific elements that make such a system work.

- Spellcasters cant just spellcast. If spellcasting carries a potential risk to the character, they can't just be relegated to torchbearer and battlefield scenery when they aren't casting. If they have negligible hit points, cant wear armour, or weild a decent weapon, then of course the only thing they can do is cast, and if casting continually shafts them, then they aren't gonna have fun.

- Mitigation Measures. Flat "every cast has the same chance of backfiring" sucks, since there is no incentive to restrain yourself for safety. Being able to reign the power in to mitigate most (if not all) the risk, whilst being able to crank up the firepower for a commesurate increase in risk allows the player to make a balanced decision, and makes the game more fun than "shall I cast: yes/no". The risk has to be in proportion with what the character is trying to do - lighting a bonfire with a six-inch long tongue of flame shouldn't carry the same risk as bombarding a football-pitch sized area in hellfire.

- Got to be worth the effort. If you are going to risk your characters life with a cast, it better be worth the risk, the damage level and effects have got to be game-changingly powerful, because the law of averages says that some day those big casts are gonna catch up with you.

I have a story of a Psyker death in my Dark Heresy campaign, that highlights some of the potential fun to be gained out of the sort of system, but which also highlights some of the drawbacks:

My Dark Heresy cabal was storming the hideout of a coven of Witches. Going in, the party knew they were facing likely 4-5 enemy Psykers, and the party Psyker knew from past experience that the more powers going off in a close proximity, the thinner the veil got, and the more likely things would go sideways, and that in particular, Witches tended to exude a (un)natural aura that thinned the veil in their proximity. So going in to the fight, he knew this was a very high risk situation (and furthermore, something he was aware of, but was unknown to the rest of the party, a few sessions previously a Psychic mishap had 'broken' his own Imperial Psyker conditioning, so he also gave off the same veil-thinning aura as the enemy Witches). One of the other party members botched a door breach, and walked into a force-lightening storm that took him straight out of play, so the Psyker decided 'gloves off', and went in full lightening hellstorm. For several rounds the game was chaos, the veil was getting shredded to pieces, and random psychic effects accompanied nearly every cast for the last few rounds, but the party Psyker accepted that the encounter was just too dangerous, and he was willing to run the (increasing) risk of sacrificing himself to prevent the party suffering any other losses. However, luck was on his side, the party won through, the enemy Witches were wiped out, and the rend in the veil restored. However, that first character taken out had been bleeding to death all fight, and with the background psychic effects receded, the party Psyker walked over to him, and used his weakest power, on possibly its lowest strength to stop the bleeding long enough for proper medical attention to be provided, and with a pop, he was sucked screaming into the warp, never to return.

While the player couldn't stop laughing for several minutes due to the absurdity, the encounter highlighted both how such a system can go so right, and how it can go so wrong. He walked into that fight as death incarnate, pulled out all the stops, and accepted the risk of doing so. It was glorious to watch. He was convinced he was going to die during that fight, and had made peace with that. However, while the system definitely allowed him to increase his risk, it clearly didn't sufficiently allow him to mitigate it, as the spell cast that killed him was at the power level of something that could have been achieved with a skill check (indeed, another character who was unconscious a couple of rooms away had something like a 90% chance to stop bleeding on a Medicine skill check), and the result was way out of proportion with what he was trying to achieve.

[PhoenixPhyre]

Even if casters are a player thing, I think scaling the effort and cost in proportion to the spell's power is worthwhile.

Everyday spells vs big plot spells should, IMO, be a dividing line in many more settings/systems than it is right now.

I agree and try to do this even in D&D.

I'm working on a system where most of the "utility" and many of the "non-combat" spells aren't spells anymore. They're on no-one's list. Instead, they're what I'm calling incantations. Ritual performances that anyone (of the appropriate level) can do, given knowledge, time, and the appropriate components. Certain spell casters get bonuses to conducting certain incantations--a cleric of a life god can do Restoration (all the Lesser/Greater restoration spells rolled into one multi-faceted incantation) faster than most...at the cost of a spell slot. A wizard might be able to cast a different incantation cheaper (but at the same time cost) OR faster (at the cost of a spell slot).

So what makes spellcasters special is that they're the only ones who can do quick magic. Want someone blown up? Spellcaster. But your friendly neighborhood priest can cure that disease, given plenty of time and cash, despite not being able to cast any real spells.

IMO, that's way better for worldbuilding (gating things on knowledge, time, and money for the most part) while still letting the spellcasters feel special and cast their spells.

And even grander magics are actual plot points. I had a quest that involved gating in a meteor (like a real one, not that mediocre meteor swarm, but a significant asteroid) to drop on a temple way over there that was infested by something real nasty. The party ended up protecting the NPCs who did the ritual while hordes of demon creatures attacked and even fought off a (weakened) avatar of a demon prince who tried to stop them. I had the NPCs do the ritual so that the PCs weren't stuck doing nothing while I played NPCs protecting them. In a different circumstance I might have had one of them be doing the ritual while the rest of the party did the fighting.

[Max_Killjoy]

I agree and try to do this even in D&D.

I'm working on a system where most of the "utility" and many of the "non-combat" spells aren't spells anymore. They're on no-one's list. Instead, they're what I'm calling incantations. Ritual performances that anyone (of the appropriate level) can do, given knowledge, time, and the appropriate components. Certain spell casters get bonuses to conducting certain incantations--a cleric of a life god can do Restoration (all the Lesser/Greater restoration spells rolled into one multi-faceted incantation) faster than most...at the cost of a spell slot. A wizard might be able to cast a different incantation cheaper (but at the same time cost) OR faster (at the cost of a spell slot).

So what makes spellcasters special is that they're the only ones who can do quick magic. Want someone blown up? Spellcaster. But your friendly neighborhood priest can cure that disease, given plenty of time and cash, despite not being able to cast any real spells.

IMO, that's way better for worldbuilding (gating things on knowledge, time, and money for the most part) while still letting the spellcasters feel special and cast their spells.

And even grander magics are actual plot points. I had a quest that involved gating in a meteor (like a real one, not that mediocre meteor swarm, but a significant asteroid) to drop on a temple way over there that was infested by something real nasty. The party ended up protecting the NPCs who did the ritual while hordes of demon creatures attacked and even fought off a (weakened) avatar of a demon prince who tried to stop them. I had the NPCs do the ritual so that the PCs weren't stuck doing nothing while I played NPCs protecting them. In a different circumstance I might have had one of them be doing the ritual while the rest of the party did the fighting.

In general I prefer that, if a system has classes, those classes are about being better at doing certain things, rather than about whether a character can do those things at all -- so I back that approach.

[PhoenixPhyre]

In general I prefer that, if a system has classes, those classes are about being better at doing certain things, rather than about whether a character can do those things at all -- so I back that approach.

A place we (mostly) agree. I don't have a problem with classes who aren't spellcasters not casting spells. That does not mean that they can't do fantastic things (via incantations, rituals, just sheer awesomeness, burning martial spirit, or whatever), just that they're not casting spells. But the vast majority of things should be "anyone can try it, you can do it better/faster/cheaper/etc".

[Pex]

I agree and try to do this even in D&D.

I'm working on a system where most of the "utility" and many of the "non-combat" spells aren't spells anymore. They're on no-one's list. Instead, they're what I'm calling incantations. Ritual performances that anyone (of the appropriate level) can do, given knowledge, time, and the appropriate components. Certain spell casters get bonuses to conducting certain incantations--a cleric of a life god can do Restoration (all the Lesser/Greater restoration spells rolled into one multi-faceted incantation) faster than most...at the cost of a spell slot. A wizard might be able to cast a different incantation cheaper (but at the same time cost) OR faster (at the cost of a spell slot).

So what makes spellcasters special is that they're the only ones who can do quick magic. Want someone blown up? Spellcaster. But your friendly neighborhood priest can cure that disease, given plenty of time and cash, despite not being able to cast any real spells.

IMO, that's way better for worldbuilding (gating things on knowledge, time, and money for the most part) while still letting the spellcasters feel special and cast their spells.

And even grander magics are actual plot points. I had a quest that involved gating in a meteor (like a real one, not that mediocre meteor swarm, but a significant asteroid) to drop on a temple way over there that was infested by something real nasty. The party ended up protecting the NPCs who did the ritual while hordes of demon creatures attacked and even fought off a (weakened) avatar of a demon prince who tried to stop them. I had the NPCs do the ritual so that the PCs weren't stuck doing nothing while I played NPCs protecting them. In a different circumstance I might have had one of them be doing the ritual while the rest of the party did the fighting.

Devil in the details, but I have no issue with this. I like you're allowing for rituals to be done quickly in one round, accepting depending on class and the ritual. That was 4E's mistake not allowing that. It was fine havng the rituals and anyone can do them, but sometimes you needed a ritual Now! It's fine for a remove poison ritual to take a bit of time given a non-fatal debilitating poison, but when it happens in combat you want the victim to be cured immediately.

In any case, the opposite of not wanting a player punished for playing a spellcaster is not having no retrictions on magic at all.

[Satinavian]

Yes, and nonspellcasters have their own autosuccess Nice Things such as Advantage when you want it, can't roll below 10, gain two extra attacks, and take another Action in addition to healing yourself.

The most proper equivalence would be skill use. Which has a chance to fail.

But that was not really my point. I don't need a chance for magic to fail. I just wanted to highlight that does not count as already having such a chance, just because there are defenses against magic.

[Anonymouswizard]

Even if casters are a player thing, I think scaling the effort and cost in proportion to the spell's power is worthwhile.

Everyday spells vs big plot spells should, IMO, be a dividing line in many more settings/systems than it is right now.

I've been growing to like how Advanced Fighting Fantasy handles magic. Outside of combat your actual skill in magic basically only matters of you're on a really tight timeframe, taking a minute to cast a spell will give you +5 to your casting, so even the weakest magician only fails outside of combat one time in thirty six. Then magic is split into five types. Minor Magic is just basic utility stuff. Wizardry and Sorcery are two different flavours of spellcasting more suited to spellcasting, and there are significant differences (Sorcerers just know all their spells and cast from hp, while Wizards omit begin with a handful but get a separate MP stat). Priestly powers are just a handful of big effects you can pull out one a day (mostly). Arcane magic is narrative scale rituals.

Arcane magic is interesting. Spellcasters can already do rituals to make their spells easier to cast, but if those are taking longer than two minutes then something is very wrong. Arcane magic creates effects that change the narrative for everybody in the area, it's what you need if you're summoning a storm, imploding a city, calling Azatoth so that he may step through into our realm, or draining a sea so that you can cross safely. I'm AFB, but being able to take a single rank in the skill requires you to be a very powerful mage (6 MAGIC and 6 Wizardry or Sorcery), I believe the spells are very hard to cast, and at this point the book gives up on providing spells and tells you to work it out.

So yeah, despite the lack of spells it's a cool idea.