What if spell casting was a skill check

[kulosle]

In a lot of other systems, it takes a skill check to be able to use magic. What if you did that in dnd? How high would the DC's have to be to balance it out? Would this be a good or bad addition?

[Segev]

You would see PCs be built to make the DCs trivially, whatever they were, unless the DCs were designed in such a way that it's a treadmill or worse and are geared against maximal optimization.

You can see some of the problem with the Truenamer and his mechanics. Epic Spells have a similar mechanic and different but related problems.

If you make this just as an addition, it either is a pointless tax on the PCs or totally ruins casters entirely.

If you make it replace spell slots, etc., then it either makes casters lose any semblance of limitation or makes them still unable to actually function.


You could build a d20 system to handle this, but you'd have to start without most of 3.5 and PF's material available.

[fishyfishyfishy]

The Truenamer does this. It is generally considered a terrible class without a heavy amount of house rules to fix the DC however.

[kulosle]

Well just to make it clear i'm saying making this an addition to the current system. So still the normal limit of spells per day and all that.

[Segev]

The Truenamer does this. It is generally considered a terrible class without a heavy amount of house rules to fix the DC however.
I actually just started a thread discussing this. If it's JUST the DC doing it, I think it less likely to be an issue than is commonly discussed. However, I could be missing something.

Well just to make it clear i'm saying making this an addition to the current system. So still the normal limit of spells per day and all that.You'll find it's either a non-issue or it makes casters unplayable, as a general rule. Setting the DCs would be so fiddly that it would make optimizing for the skill either required, or nearly pointless. And if it's required, it is either unattainable without massive cheese, or there will be builds that make it a non-issue and those become the (more than) One True Build(s).

[Fax Celestis]

I've done a lot of experimenting with skill-based D&D magic, and the only way I've found that makes any sort of sense thus far is to have it replace the scaling benefits from CL, not per-day limitations or "can I cast this spell at all" checks. And calibrating that scale in a way that makes sense both from a balance perspective and from the perspective of not having to do long-hand division or logarithms at the game table is a nightmare. It's the primary reason I never finished my Cartomancer (http://www.giantitp.com/forums/showthread.php?t=92729) class.

Even then, though, all it really does is introduce some variance into the current system. Sometimes your fireball does 14d6, sometimes it does 20d6.

[Spuddles]

The current magic system is "balanced" around limited castings per day and the assumption that magic should always be better than non-magic, forever.

In such a paradigm, skill checks dont do anything other than reduce uses per day. This then encourages gravitation towards more powerful spells, abusive mechanics, etc. No one likes fireball because there are other more useful options.

If you were to make spells less teleport, finger of death, and polymorph, and more fireball, sonic lance, or meteor storm, then a skill based mechanic would work better.

If all your skill check is going to do is force optimization and enact a chance of failure, you're better off just taking away a wizard's bonus feats, giving him 1+int skill points per level, and a flat 10% spell failure no matter what.

[Flickerdart]

In a lot of other systems, it takes a skill check to be able to use magic. What if you did that in dnd? How high would the DC's have to be to balance it out? Would this be a good or bad addition?
This carries exactly the same problem as every other blanket change to spells that affect how they are activated.

In the current system, there is a huge spectrum of spells. Some are very powerful, some are very weak, and some are right on the dot in terms of power. A reasonable player is thus able to tailor his spell list to the optimization level of his game pretty effortlessly. Instead of Orbs, he takes Scorching Ray. Instead of Glitterdust, Pyrotechnics. Instead of Polymorph, the [Polymorph] spells from PHBII. Pretty simple stuff. He can also keep a few of the heavy hitters around in case of an emergency.

If you implement one of these fixes (Concentration checks to cast spells, arbitrary chance of failure to prepare/cast, etc) then this whole thing falls apart. It stops being practical to prepare intentionally weaker than optimal spells because with any given spell you don't know if you will expend that daily resource and your turn to do absolutely nothing. So the reasonable player, in order to continue being balanced against his party, needs to pick better and better spells - spells that, when they go off, win the encounter, to compensate in magnitude for when they don't go off. So now you have the problem of people using the uberspells when they would normally not use them.

Meanwhile, the powergamers that like to sling around Polymorphs in games of any optimization are still going to be able to do it with impunity, because they have the knowledge and motivation to get their characters as strong as possible already, so hitting a skill check isn't really going to be all that hard.

A rule that hurts reasonable people and doesn't affect unreasonable people is not a rule that should be implemented.

[lesser_minion]

In a lot of other systems, it takes a skill check to be able to use magic. What if you did that in dnd? How high would the DC's have to be to balance it out? Would this be a good or bad addition?

It's not an inherently bad game mechanic, but as has been stated repeatedly in this thread, 3rd edition D&D wouldn't work with it.

[kulosle]

A rule that hurts reasonable people and doesn't affect unreasonable people is not a rule that should be implemented.

This is a very good point, and even if you could find the "perfect dc" (to me this means that it is hard to always pass but still possible with good optimization and takes up a reasonable amount of your resources to do so. This of course would have to scale by spell level.) does still hurt reasonable people a lot more than unreasonable ones.


It's not an inherently bad game mechanic, but as has been stated repeatedly in this thread, 3rd edition D&D wouldn't work with it.

So why does this work in some systems but not dnd? Should it be a key ability test instead? like an int test for wizards and a cha test for sorcerers because those are a lot more difficult to optimize and most of the ways of doing so are obvious to non optimizers as well. Or maybe instead of having all spells of the same level having the same DC maybe certain spells should have higher/lower ones depending on its power?

[Flickerdart]

So why does this work in some systems but not dnd?
I don't know of any RPG that has such an unintentional disparity in its options, and such a high premium on actions. When spells are roughly the same power, and wasting an action isn't of much consequence, then this kind of thing works fine.

[Spuddles]

As I outlined above, and Flickerdart has explained, it doesnt work in dnd because the magic system uses a highly limited resource with a plethora of options, from beyond useless to what the hell were they smoking.

If you had far more spells per day and they were all around fireball's power, it would be a fine mechanic. It could actually be used fairly well to limit a blaster caster from having too much DPS, as it would basically be a fighter making attack roll vs. AC.

A system like truenaming should work, though truenaming has too many restrictions. A hybrid warlock-truenamer like class could sit comfortably at the bottom of tier 3 with the warlock and crusader.

[Fax Celestis]

So why does this work in some systems but not dnd?

At least part of it has to do with how the skill system in D&D is made. At low levels, the d20 is a huge variance, and your expected results can range between 5 and 35 dependent upon circumstantial bonuses. At higher levels, this range increases to 5 to 100+, based upon how focused you are at the skill. The d20 plays a much smaller part at this range, which also means that the DCs have to be so finely tuned that even one point of bonus unaccounted for means a significant rate increase.

The only way one could calibrate D&D's skill system to account for this would be to directly scale the DC off of the caster's skill (including bonuses)--or to constrain the bonuses that apply to a skill. In both instances, you could frankly just make it a level check at that point.

[kulosle]

So again what about a key ability test?

[Fax Celestis]

You run into the low-modifier variance problem where the luck of the roll plays gigantic part in the variability.

[Maginomicon]

In my main game, I have casting involve a Concentration check in the process, but the Concentration check is by-default trivial unless certain conditions are in play. Specifically, I have this thing called "Spell Burn" and the more Spell Burn you have, the higher the DC gets for the Concentration check. It works out pretty well considering the real penalty of having Spell Burn (that being, ability burn in the relevant casting stat).


Every time you cast a spell, you must first succeed on a Concentration check described below. Failure means the spell fails. This checks happens before all other Concentration checks involving casting a spell. The Concentration check to cast a spell is...
DC 5 + spell level + your current amount of Spell Burn for that class.

Keep track of Spell Burn for each spellcasting class you have, as well as the relevant ability score which determines the Save DC against spells of that type. Once you accumulate new Spell Burn in a spellcasting class such that your total Spell Burn of that class reaches a higher multiple of 4, you also take 1 Ability Burn for the ability score for which the Save DC is determined for that spellcasting class (ex. for Wizards it’s Intelligence and for Clerics it’s Wisdom). Keep track of which spellcasting class caused each point of Ability Burn.

If the amount of Spell Burn you possess in a particular spellcasting class is not a multiple of 4, Spell Burn is eliminated for that class at a rate of 1 per hour of rest until the amount of Spell Burn for that spellcasting class is equal to a multiple of 4. From then on, when a point of Ability Burn that was caused by Spell Burn for that spellcasting class is recovered, that spellcasting class eliminates 4 Spell Burn.

I can go into more detail, but it would get rather involved and complicated, so PM me if you're interested in learning more.

[gorfnab]

Here is a previous thread about using a skill to determine spell casting. (http://www.giantitp.com/forums/showthread.php?t=290711)

[mregecko]

What you're essentially doing is trying to introduce a failure chance into casting.

Since it's based off a flat-scaled DC (skill, ability, whatever), people are inherently going to either A.) Optimize against the check, making it essentially 0%... or B.) Not optimize the check, in which case they probably will have a predictable chance of failure (depending on whatever DC formula you choose).

If you cater to (A) and make the check very difficult, then you will screw the (B) group over. If you cater to group (B), then group (A) will have essentially zero chance.

The two options that would make sense to ME are:

1.) A scale of diminishing returns. This would probably be very difficult to implement, and cumbersome to run. But say there's a base failure chance of (5%)*(Spell Level).

Then you have a table with (Spellcraft Roll) vs (Failure Modifier)


ROLL MODIFIER
1-5 -5%
6-10 -10%
11-15 -15%
16-20 -18%
21-25 -20%
26-30 -22%
31-35 -23%
36-40 -24%
41+ -25%

This means that even 20th level spellcasters, who should be making 40+ on their spellcraft checks, still get a 20% failure chance on their highest level spells... But everything 5th level and under is a guarantee.

(Again, I think this is a cumbersome way of doing it)

2.) A flat (or scaling) percentage chance of failure. Maybe 15%, or 2-3% per spell level.

Please note, I think spell casting is just fine as it is. These are just (my opinion) better ways of implementing what you're talking about.

-- Mr

[Gemini Lupus]

There is a 3rd party splat book that has an option for skill-based spellcasting, the Advanced Player's Guide from Sword & Sorcery Studios. I use it in my group and it seems to work very well. I don't have the book on hand, but I think that the DC for 9th level spells is 39 and the book states that you cannot use skill focus to increase your bonus to spellcasting checks. The Spellcasting skill uses the caster's casting stat as the key ability and there is a soft limit on spells per day. You have your regular limit as prescribed in the PHB, but you can cast more than just that many spells. Once you go beyond your "limit" for the day, casting spells of that level becomes more difficult. The DC for spells of that level increases by 2 for every spell cast beyond the limit, until you fail to cast a spell of that level.

Not for everyone, but my group seems to like it.

Them's my 2 coppers.

[Lafaellar]

An idea would be to force a skill check that is easy in the beginning but increase with every spell you cast. So instead of saying "I have 2 7th level spells remaining" it is more like "uh, I have to roll a 12 or higher to pull of another 7th level spell".
But what it actually does is replacing a strictly designed resource managing system with a randomiser.

[Ninjadeadbeard]

The Truenamer does this. It is generally considered a terrible class without a heavy amount of house rules to fix the DC however.

Fixed.

And I've always thought that skill-based magic could work in dnd if it worked like the Truenamer...except with reasonable DCs and heavily nerfed spells compared to the current selection (though even then not too many need it). But to keep skill ranks in check you'd have to remove magic items that boost casting stats, among other cheese.

[kulosle]

Well i was always thinking of it being a scaling DC the only problem is figuring out what it would be. It can't just be C+spell level (Constant), it would probably increase exponentially. When i was thinking of it being a skill check i was thinking something like 10+spell level^2. That's close but definitely not right.

Then i started thinking of an ability check. I'm not sure what the dc would be but i think it could be made easier if you did something like limit the temporary bonus you can have to a stat, probably to 6, because only heavy optimizers know how to get passed a 6 any ways. So this leaves you with a stat bonus on or about 4 - 11. But these small numbers means it's almost always up to the dice.

So I've abandoned that idea too. I'm thinking that i want the chance of success for the highest level spell to be 65% ish (i'm still fiddling with that number). Also, i'm fairly certain that the DC should go up exponentially, i have no evidence or reasoning why, but it sounds right.

I have another idea, that i haven't finished yet. But more on that tomorrow, for now i must sleep.

[Unusual Muse]

You might consider playing around with the Incantations variant system from Unearthed Arcana.

[Fax Celestis]

Well i was always thinking of it being a scaling DC the only problem is figuring out what it would be. It can't just be C+spell level (Constant), it would probably increase exponentially. When i was thinking of it being a skill check i was thinking something like 10+spell level^2. That's close but definitely not right.

Then i started thinking of an ability check. I'm not sure what the dc would be but i think it could be made easier if you did something like limit the temporary bonus you can have to a stat, probably to 6, because only heavy optimizers know how to get passed a 6 any ways. So this leaves you with a stat bonus on or about 4 - 11. But these small numbers means it's almost always up to the dice.

So I've abandoned that idea too. I'm thinking that i want the chance of success for the highest level spell to be 65% ish (i'm still fiddling with that number). Also, i'm fairly certain that the DC should go up exponentially, i have no evidence or reasoning why, but it sounds right.

I have another idea, that i haven't finished yet. But more on that tomorrow, for now i must sleep.

To be perfectly frank, you are running into all of the same problems the truenamer has, except you're also making them worse.

[Drachasor]

Rolls to see if you do anything in a round generally make a game worse, in my opinion.

[Ninjadeadbeard]

Rolls to see if you do anything in a round generally make a game worse, in my opinion.

It could work if the roll was to determine if the spell worked "perfectly". Failing a Spellcheck for Hold Person inflicts some sort of Slow penalty instead. Failing teleport causes a Dimension Door effect. Failing Fireball causes Produce Flame. Etc etc.

[Gemini Lupus]

Rolls to see if you do anything in a round generally make a game worse, in my opinion.

Rolling to see if a spell succeeds is not fundamentally different than a fighter rolling to see if they hit an enemy. Why should a spellcaster never have to worry about a spell failing when a fighter still has to worry about whether or not his attack is going to succeed?

[Fax Celestis]

Rolling to see if a spell succeeds is not fundamentally different than a fighter rolling to see if they hit an enemy. Why should a spellcaster never have to worry about a spell failing when a fighter still has to worry about whether or not his attack is going to succeed?

Because it sucks for the fighter too, and his attack is an unlimited resource. Putting a spell failure system onto vancian casting quickly devolves into "how many times can I lose my spell today?"

[olentu]

Rolling to see if a spell succeeds is not fundamentally different than a fighter rolling to see if they hit an enemy. Why should a spellcaster never have to worry about a spell failing when a fighter still has to worry about whether or not his attack is going to succeed?

Saving throws among other things.

[Drachasor]

Rolling to see if a spell succeeds is not fundamentally different than a fighter rolling to see if they hit an enemy. Why should a spellcaster never have to worry about a spell failing when a fighter still has to worry about whether or not his attack is going to succeed?

One, a spellcaster is spending a very limited resource to cast a spell. For another, I'm not a fan of the fighter having fruitless actions either. A "how much do you accomplish check" is better than a "do you get anything done check", imho.