Making Arcane Magic More Difficult -pls PEACH

[jjordan]

Self-scrubbing for reasons.

[JNAProductions]

If a DM sprung this on me, I'd leave the table. Even if I wasn't playing an arcane caster.

Let me ask you, high-level goal: What are you trying to accomplish, and why?

[heavyfuel]

Why are only Clerics and some Warlocks considered divine? Paladins and Druids are explicitly divine spellcasters, and Rangers are implied to be divine spellcasters as well.

All you're doing is soft-banning casters, since no one in their right mind would ever play one with these rules. You might as well say "no arcane casters allowed"

If you wish to nerf casters, the only possible way to do it is going through each spell and nerfing them individually, there's no blanket solution because blanket solutions hurt non-optimizers far more than optmizers.


If a DM sprung this on me, I'd leave the table. Even if I wasn't playing an arcane caster.

Let me ask you, high-level goal: What are you trying to accomplish, and why?

Seconded in every way.

[olskool]

I'm guessing that your goal is to decrease the power of casters a bit and thereby increase the "grit" of your 5e campaign. I just wonder why you are "complicating" a very good basic mechanic in 5e... the Proficiency/Skill system? We use Skill Tests for casting too. This is our system which gives the typical Caster the same chances of success as a typical fighter might have to hit.

THE DIFFICULTY LEVELS OF CASTING:

The DC varies based on the situation the Caster is working under. The base DC for a typical adventure situation is listed below. To this DC the Caster ADDS the SPELL [being cast*] LEVEL with Cantrips being 0-Level for this calculation. The total DC is then rolled equal to or over for a success. The Caster gets to add any Characteristic Bonus and any Proficiency Bonus to this Skill roll. IF the Spell fails, no Spell Slot is used and the Caster can try again next round. IF the Spell succeeds, the Caster uses up a Spell Slot and the Spells effects manifest. The typical DCs are:

- In quiet or settled environments where the caster can relax = DC 5 + Spell Level
- Under normal adventuring conditions = DC 10 + Spell Level
- Under difficult conditions (unstable footing, under missile attack, or the effects of a Spell that can mess with casting) = DC 15 + Spell Level
- Under extreme conditions (under melee attack, severe conditions or spell effects) = DC 20 + Spell Level

Remember that when you are looking at the DCs, they DO NOT take into account the Caster's Proficiency or Characteristic Bonuses. A typical 1st Level Caster will have an average of +4 to their roll when incorporating those bonuses. This means that the chance of a 1st level Caster succeeding at a 1st Level Spell under normal adventuring conditions is 65% (DC 11 - 4 = 7+ on 1D20 or a 65% chance of success) . This is right in line with a fighter's chance to hit. A 20th Level Caster (assuming they have a +10 Proficiency AND Characteristic Bonus combined) attempting a 9th Level Spell would succeed on a ROLL of 9+ (a 55% chance of success). Considering how World-changing a 9th Level Spell is, that's not a bad chance. That same 20th Level Caster using a 1st-Level Spell would only fail on a roll of 1 on 1D20 (a 95% chance of success). Magical FEATS such as Warcaster can allow a Caster to "ignore" adverse conditions and cast as though they were in a normal situation which can help the Caster who is always getting into melee or being subjected to weapons fire while casting a spell.

We have used this system for ALL our casting and it works just fine.

* Upcasting a Spell increases the Spell's effective level for determining the DC.

[JNAProductions]

I'm guessing that your goal is to decrease the power of casters a bit and thereby increase the "grit" of your 5e campaign. I just wonder why you are "complicating" a very good basic mechanic in 5e... the Proficiency/Skill system? We use Skill Tests for casting too. This is our system which gives the typical Caster the same chances of success as a typical fighter might have to hit.

THE DIFFICULTY LEVELS OF CASTING:

The DC varies based on the situation the Caster is working under. The base DC for a typical adventure situation is listed below. To this DC the Caster ADDS the SPELL [being cast*] LEVEL with Cantrips being 0-Level for this calculation. The total DC is then rolled equal to or over for a success. The Caster gets to add any Characteristic Bonus and any Proficiency Bonus to this Skill roll. IF the Spell fails, no Spell Slot is used and the Caster can try again next round. IF the Spell succeeds, the Caster uses up a Spell Slot and the Spells effects manifest. The typical DCs are:

- In quiet or settled environments where the caster can relax = DC 5 + Spell Level
- Under normal adventuring conditions = DC 10 + Spell Level
- Under difficult conditions (unstable footing, under missile attack, or the effects of a Spell that can mess with casting) = DC 15 + Spell Level
- Under extreme conditions (under melee attack, severe conditions or spell effects) = DC 20 + Spell Level

Remember that when you are looking at the DCs, they DO NOT take into account the Caster's Proficiency or Characteristic Bonuses. A typical 1st Level Caster will have an average of +4 to their roll when incorporating those bonuses. This means that the chance of a 1st level Caster succeeding at a 1st Level Spell under normal adventuring conditions is 65% (DC 11 - 4 = 7+ on 1D20 or a 65% chance of success) . This is right in line with a fighter's chance to hit. A 20th Level Caster (assuming they have a +10 Proficiency AND Characteristic Bonus combined) attempting a 9th Level Spell would succeed on a ROLL of 9+ (a 55% chance of success). Considering how World-changing a 9th Level Spell is, that's not a bad chance. That same 20th Level Caster using a 1st-Level Spell would only fail on a roll of 1 on 1D20 (a 95% chance of success). Magical FEATS such as Warcaster can allow a Caster to "ignore" adverse conditions and cast as though they were in a normal situation which can help the Caster who is always getting into melee or being subjected to weapons fire while casting a spell.

We have used this system for ALL our casting and it works just fine.

* Upcasting a Spell increases the Spell's effective level for determining the DC.

So, a Wizard casts a Cantrip at level 5. They have +7 (+4 Intelligence, +3 Proficiency) meaning they fail 10% of the time-except this cantrip is Firebolt, so they're almost certainly under SOME form of attack. DC 15, then, meaning they fail to cast the spell 35% of the time.

Assuming they succeed, they then need to hit.

Whereas a Fighter just skips straight to rolling to hit.

And it gets worse if they, say, want to cast Fireball. Assuming five enemies with 10 Dex, they'd need an 11+ to cast (DC 18, +7 to the roll), then each enemy gets a saving throw (against DC 15) for an average of 23.8 damage per enemy hit, assuming the spell goes off. With five enemies and a 50% cast chance, that's an average of just under 60 damage from a cast of Fireball-which doesn't really reflect the actual situation. You'd either do just shy of 120 damage across five foes, or fail to do anything at all with one of your two 3rd-level slots.

The biggest issue I have with this is that it's not FUN. This is a game people play to have fun-saying "You have a 50/50 chance of casting your best spell under perfectly ordinary combat situations" isn't FUN.

I'm perfectly willing to have a discussion on individually overpowered spells that could be reigned in. Fireball is a little too good for its level.Clone, Simulacrum, and Wish don't need to be available as ordinary PC spells. There are other examples-but a blanket, unfun nerf of everything from Firebolt to Fireball to Guidance to Bless... No.

[jjordan]

If a DM sprung this on me, I'd leave the table. Even if I wasn't playing an arcane caster.

Let me ask you, high-level goal: What are you trying to accomplish, and why? It's an exercise in working through a way to reduce the power of magic users. I'm not advocating for it and not planning to use it in either of the games I'm running. It was a fun thought experiment and I am trying to balance it and keep it interesting.


Why are only Clerics and some Warlocks considered divine? Paladins and Druids are explicitly divine spellcasters, and Rangers are implied to be divine spellcasters as well. Good point. I'm ambivalent about the recasting of clerics and paladins since AD&D. And druids admittedly are difficult to categorize.

olskool, that's a much simpler system but will fail more frequently and, as JNAProductions points out, still requires the caster to hit the target after a successful cast. That could be addressed, but I was trying to avoid going spell by spell. Admittedly, your system doesn't penalize casters for spell failure to the same extent mine does.

[JNAProductions]

It's an exercise in working through a way to reduce the power of magic users. I'm not advocating for it and not planning to use it in either of the games I'm running. It was a fun thought experiment and I am trying to balance it and keep it interesting.

That is, in my opinion, a bad goal. Or, at least, your method is a poor one-don't reduce the power of casters OVERALL-that'll push people towards the more powerful options exclusively, since if it only works half the time, you need to make that half COUNT.

Nerf the outlier spells.

[Old Harry MTX]

Why don't you change it trying to build an upscale system? Like introducing a system like the one suggested by olskool, that gives you very few chance of failing a spell if casted at its level (maybe DC 8 or 10, against a roll in which you consider your spell's modifier and, for certain schools, your proficiency bonus), but allows you to cast it at a higher level with the same spellslot, adding 5 to the DC for each level of upscaling...

I didn't do any math, so it could be completely unbalanced, but tweaking the base DC and the malus I think you can go somewhere...

[Edea]

...I think you might be better off just banning all 'quick' spellcasting classes entirely, and then opening up ritual casting to everybody regardless of the class they're in.

[Kane0]

The basic goal here is to explore the idea of weakening arcane magic users by introducing a degree of uncertainty: the magic will mostly work, but sometimes it won't. I'm looking at this as an optional overlay that can be added to a campaign. I should add that I'm not advocating for these changes, I'm just exploring the options. If we're weakening arcane magic users then divine magic users need to be weakened as well. I suggest this can best be accomplished by limiting the spells a cleric is able to access. Instead of having access to the full list of all clerical spells, clerics (and some warlocks) should only have access to a subset of those spells which are specific to, and appropriate for, the deity/patron they serve. However, I'm not covering that with this post.

Fair enough. Reducing spell access would be preferable to your goal in my eyes but let's have a look at this anyways. Never know, could lead somewhere.



Performing arcane magic is a skill.

-All arcane spells, but not cantrips, have a DC equal to 7 plus the spell level.
-DMs may scale the difficulty of this skill (a 6 is much easier, an 8 much harder).
-Casters must make a skill check versus their primary casting ability in order to successfully cast a spell.
-Wizards have access to all arcane spells. However, they start with proficiency in two schools of magic. They may acquire proficiency in other schools, or advance to expertise in schools they are already proficient in, by acquiring these as feats. Wizards may acquire and use spells outside the schools they know but they are harder to learn and use.
-All other arcane casters are limited to the pre-established spell lists and may not acquire spells outside those lists except as provided for by class features or feats they may acquire.
-Ranged spell attack roles are eliminated from those spells that require them, FOR ARCANE CASTERS.


Modifiers to Casting

-Casting while engaged in melee? Roll at disadvantage.
-Using a spell focus in addition to the required components? +1 to success.
-Taking extra time to cast? (Typically 10 times the spell takes to normally cast) Roll with advantage.
-Casting as a ritual? Automatic success.
-Another caster who knows the spell is helping you? This is a placeholder. I want to do this but not at this moment in time.


I think you mean ability check, not skill check. And they can be hard to balance, because you have things like Expertise, Guidance, Jack of All Trades, etc that impact them in various ways.
The change in wizard spell access doesn't appear to fit with the rest of this concept, going back to your first paragraph.
Why are Spell Attack rolls being removed? Saving Throws aren't being, and if your goal is to reduce the assuredness of spells you would probably want to keep two points of failure for each spell (not counting exceptions like Magic Missile)



Variable success.

Casters total spell/skill roll minus the DC of the spell.

(Table snipped for length)

This I like, though perhaps a bit too much rolling involved. This looks like the element of risk you're looking for, and 5e doesn't have degrees of success/failure so it's a space that can be easily expanded on. With a bit of refinement and balancing this could be both useful and fun.



Learning spells.
[INDENT]Arcane spells are hard to learn. Each spell has a task DC equal to the level of the spell times 10 and a learning period equal to the level of the spell times 2. After each learning period (during which the character spends the majority of their time studying the spell) the caster may make an attempt to learn the spell by rolling against the DC. They subtract their roll from the task DC and when it is reduced to zero they have learned the spell. You can increase the difficulty by adding materials costs (simple gold investment per learning period) or adding material requirements (go find some powdered dragon claw). Cantrips are feats/features not spells for the purposes of this solution.

The exception to this rule is when an arcane caster goes up a level. At this time they get 2 new spells without having to go through the learning process. This is the result of practical knowledge acquired through the practice of magic.

Again this looks like a change to spell access change that has been snuck in. I personally have a bias towards the spell access angle instead of the spell certainty angle so it's hard for me to critique both at the same time.

[Grod_The_Giant]

It's an exercise in working through a way to reduce the power of magic users. I'm not advocating for it and not planning to use it in either of the games I'm running. It was a fun thought experiment and I am trying to balance it and keep it interesting.
Okay... but what, specifically, is it about magic users that makes power reduction necessary? Why limit the changes to arcane casters? And, most importantly... is the problem severe enough to be worth curtailing player fun?

Because that point is critically important-- whatever improvement to game balance you get has to be worth the added irritation to spellcasters. And make no mistake, an added spell failure chance (however you reckon it) is an irritation. It's not an interesting tactical tradeoff, it's not much of a roleplaying hook, it doesn't add an interesting wrinkle to how you play, it just means you waste a not-insignificant number of turns.

When it comes to game balance, take it from someone who labored for a long time (http://www.giantitp.com/forums/showthread.php?t=329161) on a much more imbalanced system (https://forums.giantitp.com/showthread.php?266559-Tier-System-for-Classes-(Rescued-from-MinMax))-- it's more rewarding all around to improve weak options than it is to weaken strong ones.

[Morphic tide]

Given the fluff, I've pretty much always gone for flipping from this sort of skill-based magic, because the vast majority of the complexity is supposed to be in the preparation. The big thing is that the issues stem from a few different spell, so the rules ought to be punishing to the hyper-utilitiarian, but only a mild bother to the lobber of Fireball. Having the inconsistency lie in preparing the spells at the start of the day is a big advantage because it isn't a sudden action loss like this suggestion that utterly ruins combat contribution when it goes off, but rather declaring that you can't cast some particular spell until you get another chance to prepare them.

However, the 5e system doesn't really have large enough progressions to do any sort of skill-adjacent casting mechanic, because Bounded Accuracy is literally designed to prevent the dice rollovers that are required to have early-game inconsistencies become assurances, and its only innate spell scaling is a matter of outright spell levels that is extremely harsh to use as penalty space; stacking your own modifiers will break things all over the place. Could be built around the 3rd-level workhorse and half-casters, leveraging Bounded Accuracy to have low levels be generally consistent (70%+), the half-casters are pretty much free, while the stuff that makes full casters snap open whole campaigns (6th-9th) becomes a gamble.

[Kane0]

So I think it might be helpful to start at the simplest level and work our way up, as there are lots of extra moving parts being considered all at once.

Casting a spell requires a successful Ability check of DC 7 + Spell level, except cantrips.

Then we decide on whether this remains a straight d20 roll, Casting stat check (like Concentration) or a skill check (Arcana/Nature/Religion depending on class). This is in order of the number of other mechanics that can influence the result.

Starting with the ability check, at low levels you're looking at about a 75% success rate which drops as you level to around a 50/50 chance. If you're adding proficiency bonus as a skill check it becomes about 85% to 75% at most stages of play not counting sources of extra bonuses, advantage, etc.

But what if we don't use either, and use a fixed value on a straight d20 roll? Say any levelled spell has say an 80% success rate, a 5 or higher on the d20. Far less ability to influence things (really only things like Lucky and Portent this way) means less we have to factor in and casters won't be getting worse at casting spells as they level up nor be required to have proficiency in a certain skill.

And if spells are always more likely to be cast than not, we can manipulate the degrees of failure/success much more freely too. Say a 1 always stuns you, a 2-4 is a fizzle with no other loss, a 5-10 is a success with no other benefit, an 11-15 boosts the range or duration by half and a 16-20 gives you a free upcast by one slot.

So you have a 20% chance of doing nothing with your spell slot, but also a 50% chance of it actually being better. That's the kind of odds that players will like, and you can adjust the numbers to get just the right amount of balancing you want so it's just the right amount of buff/nerf.

[jjordan]

Okay... but what, specifically, is it about magic users that makes power reduction necessary? Why limit the changes to arcane casters? And, most importantly... is the problem severe enough to be worth curtailing player fun?
-The power discrepancy.
-I didn't limit it to arcane casters, I just addressed arcane casting in this post. I mentioned pared down spell lists as a way to limit the power of divine casters but didn't want to add it to this post. I also mentioned, but deleted, a piety system of some sort and I'm interested in checking out what MOoT has to say on this point..
-No. But making this an optional overlay that doesn't require detailed changes (editing individual spells) means that it can be discussed in session zero and implemented if people find it interesting and not implemented if they don't. It can also be used to implement a style of magic only used by NPCs in a campaign.


Because that point is critically important-- whatever improvement to game balance you get has to be worth the added irritation to spellcasters. And make no mistake, an added spell failure chance (however you reckon it) is an irritation. It's not an interesting tactical tradeoff, it's not much of a roleplaying hook, it doesn't add an interesting wrinkle to how you play, it just means you waste a not-insignificant number of turns.Which is super important to one style of play (the mainstream style of play), but not all styles of play.


When it comes to game balance, take it from someone who labored for a long time (http://www.giantitp.com/forums/showthread.php?t=329161) on a much more imbalanced system (https://forums.giantitp.com/showthread.php?266559-Tier-System-for-Classes-(Rescued-from-MinMax))-- it's more rewarding all around to improve weak options than it is to weaken strong ones. At a certain point you end up turning martials into superheroes in order to give them power parity with casters. Magic should be strange, powerful, and awesome. It should be amazing. It should be difficult and if you do it wrong there should be a possibility your brain squirts out your ears and makes Jackson Pollock paintings on the walls. Martials shouldn't be able to do the things that magic can do and strengthening them until they can makes the game less fun.

Kane0 and olskool, your systems are both better than mine by the objective measure of simplicity and arguably better than mine on the details. I'm paying attention and reconsidering some points.

[Morphic tide]

-The power discrepancy.
Except that the discrepancy is in versatility, the combat situation, which is to say the blunt power situation, is rather close to even. In many respects, the casters are actually worse off, because they're hit extremely hard by the expected encounter balancing function, such that a five-encounter day takes considerable efforts to give a good showing for, and six or more will leave them wrung down to cantrips if they try to keep up in any way. The problem lies in stuff like Teleport and Charm, immense utility functions utterly absent from the Martials, and the fact they have for some reason been given a total monopoly on area damage.

Seriously, the deliberately-overpowered Fireball does less damage than an Action Surging Fighter at 5th level, and 1st and 2nd level spells aren't competing with just the blunt Greatsword Fighter for damage. The problem is specifically because casters get to be "strange, powerful, and awesome", because everything strange or awesome rapidly gets filed under "yep, that's magic". Anything beyond "swing sword" has been shoved under "yep, that's magic" for so long that it's been over a bloody decade and people are still bitter about Tome of Battle!

[Kane0]

Except that the discrepancy is in versatility, the combat situation, which is to say the blunt power situation, is rather close to even.

Indeed, which is why i prefer limiting spell access first.

[Edea]

4e already tried putting all of the utility functions into a class-agnostic ritual system, and people got mad. I remember hating it at first, too, but that was because all the rituals cost resources you could never get back; 5e removed that part.

You could also just go strictly with cantrips and rituals on the casting classes and get rid of the quick-cast spell slots, then balance the cantrips a little bit so they're a more suitable contribution to combat scenarios.

Echoing JNAProductions and heavyfuel, I'd also probably pass on the game if any of these spell failure systems were implemented.

[Grod_The_Giant]

Which is super important to one style of play (the mainstream style of play), but not all styles of play.
Uh-uh. The goal of a game is always fun. Different people can have different ideas of what's fun--I enjoy playing demigods in Exalted, and my friend enjoys playing hapless schmucks in Dark Heresy-- but the goal of a game is to be enjoyable to the people playing it. Period, end of discussion. But across all kinds of game, "I get to play less than anyone else" is not fun.


At a certain point you end up turning martials into superheroes in order to give them power parity with casters. Magic should be strange, powerful, and awesome. It should be amazing. It should be difficult and if you do it wrong there should be a possibility your brain squirts out your ears and makes Jackson Pollock paintings on the walls. Martials shouldn't be able to do the things that magic can do and strengthening them until they can makes the game less fun.
And that's fair enough; that's a great vision for a game. But it's not the vision 5e casting was designed around, and trying to shoehorn the square peg of "flexible, reliable casting with no downsides" into the round hole of "difficult, risky, and powerful" is all but doomed to failure. You'll have much better results if you start at the bottom and build your way up.

Take, for example, the alternate low-magic classes (https://forums.giantitp.com/showthread.php?615370-Low-Magic-Casters-Replacements-for-the-Bard-Cleric-Druid-Pally-Warlock-amp-Wizard). Rather than trying to cut down the Bard until it felt low-magic enough, I wrote a Rogue subclass that hits similar thematic points. Instead of adding risk and strangeness to the Wizard, I created a new base class around the idea of weird and/or slow magic. By approaching the problem from a different direction, I was able to introduce "strange, powerful, and awesome" magic in a positive way ("this is your special ability") rather than a negative one ("you lose these abilities").

So replace the Wizard with a ritual-focused class. Replace the Sorcerer with a blood mage who burns away their flesh and mortality to cast spells. Replace the Warlock with a binder who can claim great power at the expense of mental and physical changes. Get wild.

[Old Harry MTX]

If you want an easy and, IMHO, funny way to do that, extend the Sorcerer's Wild Magic unpredictable effects table to other casters...

[jjordan]

Uh-uh. The goal of a game is always fun. Different people can have different ideas of what's fun--I enjoy playing demigods in Exalted, and my friend enjoys playing hapless schmucks in Dark Heresy-- but the goal of a game is to be enjoyable to the people playing it. Period, end of discussion. But across all kinds of game, "I get to play less than anyone else" is not fun.
I'm not going to dispute that the goal of the game is to have fun. I am going to dispute that this takes away the opportunity to play. It changes the way some characters would play. But if players object to this overlay then it shouldn't be used in the game.


And that's fair enough; that's a great vision for a game. But it's not the vision 5e casting was designed around, and trying to shoehorn the square peg of "flexible, reliable casting with no downsides" into the round hole of "difficult, risky, and powerful" is all but doomed to failure. You'll have much better results if you start at the bottom and build your way up. That's both fair and accurate. I recognize this. But rather than engaging in major restructuring of the game (or, to be more honest, moving to an entirely different system) I chose to apply an overlay and accept that it wouldn't be perfect.


Take, for example, the alternate low-magic classes (https://forums.giantitp.com/showthread.php?615370-Low-Magic-Casters-Replacements-for-the-Bard-Cleric-Druid-Pally-Warlock-amp-Wizard). Rather than trying to cut down the Bard until it felt low-magic enough, I wrote a Rogue subclass that hits similar thematic points. Instead of adding risk and strangeness to the Wizard, I created a new base class around the idea of weird and/or slow magic. By approaching the problem from a different direction, I was able to introduce "strange, powerful, and awesome" magic in a positive way ("this is your special ability") rather than a negative one ("you lose these abilities").

So replace the Wizard with a ritual-focused class. Replace the Sorcerer with a blood mage who burns away their flesh and mortality to cast spells. Replace the Warlock with a binder who can claim great power at the expense of mental and physical changes. Get wild.I recognize the approach (and I've admired your work) and I use it. I create supplemental classes (and get derided for it, but that's not an issue I pay much attention to) and I'm always open to working with a player to re-build existing classes to more closely match their vision of play.

[Grod_The_Giant]

That's both fair and accurate. I recognize this. But rather than engaging in major restructuring of the game (or, to be more honest, moving to an entirely different system) I chose to apply an overlay and accept that it wouldn't be perfect.
I think we can do a more interesting veneer, though. How about...


Full-Round Casting: Spells of first level or higher that you cast don't take effect until the beginning of your next turn. If you take damage before this time, you must make a Constitution save, as if to concentrate on the spell you are casting. On a failure, the spell is lost. If the spell targets an area, you must choose the area when you begin casting the spell, not afterwards.
This weakens magic, but it does so in way that keeps control in the players' hands. It's tactical-- you have to choose your spell a bit more carefully, and the whole party has to work together to protect you while you cast it. It's a drawback, but one that can be overcome through skilled play. It feels rewarding when you pull it off, rather than punitive when you don't. Hell, you could probably give casters +2 attack and save DC with leveled spells to balance it out, to add a bit to the reward side of the column.


Sorcerous Initiation: Magic has a cost. You can't just pick up a spellbook and start casting without sacrifice--and to access stronger magic, you need to sacrifice even more. You must make a sacrifice when you gain first level spell slots, and again when you gain third, sixth, and ninth level slots. The following are some examples of possible sacrifices:

Dark Pact: You've made an agreement with inhuman agents. At any point, they can issue you an instruction that you must obey--to the letter, at least, if not the spirit. Until you do, whenever you are not pursuing that goal you have disadvantage on spell attack rolls and creatures have advantage on saves against your spells-- but the penalty is reversed when you are pursuing it. Once an order has been given, your patron cannot issue another order for one month, or until you have gained a level.
Grafted Flesh: You've stitched a piece of a magical creature to your body to borrow its power. You have disadvantage on saving throws against disease.
Madness: Your studies have damaged your mind. Roll once on the Indefinite Madness table.
Persistent Hallucinations: You are plagued by fleeting visions and inhuman whispers. You suffer disadvantage on Perception checks, and on Investigation checks to recognize illusions.
Ritual Scarring : You've branded yourself with occult sigils. You suffer disadvantage on all Persuasion checks, and on checks to disguise yourself.


This really helps emphasize the dangers of magic. The player has to accept penalties, but they're interesting ones-- things that provide plot hooks and roleplaying inspiration, rather than a simple mathematical penalty.


Backlash: Magic can be risky. When you miss with a ranged spell attack, or an opponent succeeds on a saving throw against one of your spells, you suffer a penalty based on the spell's school. If the spell allows you to make multiple attacks or targets multiple creatures, you only suffer backlash if at least half of your attacks miss/saves succeed. Cantrips never cause backlash.

Conjuration: One random piece of nonmagical equipment disappears from your person. Clothes and armor you are wearing, and any item you are currently carrying, are excluded.
Divination: You are blinded and deafened until the beginning of your next turn, or until you are damaged.
Evocation: You suffer one point of magical damage per level of the spell.


Not complete, but you get the idea-- the penalties are small and thematic, and linked to in-game events.

[jjordan]

Updating this message.

I've thought about this. Grod's way fits D&D better. I personally think this is because his formulation of how to implement this fits into the system better and because it doesn't reduce player agency to the degree my overlay does.

I like my overlay better. Obviously I'm biased. But my overlay doesn't fit with the superheroes/video-game characters flavor of D&D which is pretty core to the game. And the fact that I like my overlay better just reinforces the fact that I ought to be playing something else, probably a home-brew GURPS setup, rather than trying to make D&D fit my preferences.