Spellcraft checks for casting, revisited

[brian c]

In a recent post, I advocated using spellcraft checks for wizard spellcasting. I've seen other similar ideas, and the problem has always been balancing the spellcraft DCs so that low-level wizards are not useless, but high-level wizards are toned down slightly.

The principle behind this spellcraft check, for anyone who hasn't heard of this before, is that although arcane magic should be extremely powerful (and thus it's fitting that wizards are the most powerful class), it is also going to be very difficult to cast the spell correctly, with the proper verbal, somatic and material components exactly right. The best existing mechanism for this in D&D is to make a Spellcraft check.

Besides being a game-balance fix to nerf wizards a little, there's a good fluff background for it. I'm presenting this idea particularly as I will use it in my homebrew setting, where magic is natural for sorcerers and very hard to learn for wizards- I want to use this system only for wizards, and leave sorcerer spellcasting without Spellcraft checks.

As I said before, the problem with setting Spellcraft DCs is that they must remain balanced from levels 1-20. Too low, such as 10+(spell level) or 10+2*(spell level) and high level casters automatically succeed on every check, defeating the purpose. Too high, such as 10+5*(spell level), and high-level wizards are challenged but low-levels have trouble making the checks.

My solution to this is an arithmetic series: DC = 10 + sum(1, n-1) for an Nth level spell. Explicitly, the DCs are as follows

0-level: n/a, automatic cast
1st: 10
2nd: 11
3rd: 13
4th: 16
5th: 20
6th: 25
7th: 31
8th: 38
9th: 46

Another feature of this system is that if a spellcaster fails the Spellcraft check, the spell is NOT considered cast; the character can try to cast the same spell (from the same slot) again later in the day.

To be continued...

[Matthew]

Interesting. So a Wizard 20 with 30 Intelligence, Synergy from Knowledge (Arcana), 23 Ranks invested in Spell Craft and two Feats still has a 25% Chance of Failing to cast a 9th Level Spell?

[brian c]

Spellcraft is not a hard skill to boost. For the following examples, I assume that each character has full ranks (level + 3) in spellcraft, starting intelligence of 18, and 5 ranks in Knowledge (Arcana) for +2 synergy.

Level 1 Wizard- Spellcraft: +4(int)+4(ranks)= +8
Spellcraft DC (level 1 spell) = 10.

Success rate: Needs to roll a 2 (95%). With Skill Focus (Spellcraft), guaranteed success.


Level 2 Wizard- Spellcraft: +4(int)+5(ranks)+2(synergy) = +11
Spellcraft DC = 10

Success rate: 100%, without Skill Focus or anything else


Level 3 Wizard- Spellcraft: +4+6+2= +12
Spellcraft DC, level 2 spell = 11

Success rate: 100%


Level 5 Wizard- Spellcraft: +4+8+2= +14
Spellcraft DC, level 3 = 13

Success rate: 100%


Level 7 Wizard- Spellcraft: +5+10+2= +17
Spellcraft DC, level 4 = 16

Success rate: 100%


Level 9 Wizard- Spellcraft: +5+12+2 = +19
Spellcraft DC, level 5 = 20

Success rate: 100%


level 11 Wizard- Spellcraft: +6+14+2 = +22
Spellcraft DC, level 6 = 25

Success rate: 90% (needs to roll a 3). With Skill Focus (Spellcraft), 100%


Level 13 Wizard- Spellcraft: +7+16+2 = +25
Spellcraft DC, level 7 = 31

Success rate: 75% (needs to roll a 6). With Skill Focus, 90%. With Skill Focus and Magical Aptitude (+2 spellcraft and UMD), 100%


Level 15 Wizard- Spellcraft: +8+18+2+3(skill focus)+2(magical aptitude)= +33
Spellcraft DC, level 8 = 38

Success rate: assuming those two feats, 80%. 100% on level 7 and lower spells.

Level 17 Wizard- Spellcraft: +9+20+2+3+2= +36
Spellcraft DC, level 9: 46

Success rate: 55%. 95% on level 7 spells, 100% on lower

Level 20 Wizard, Spellcraft: +10+23+2+3+2 = +40
Spellcraft DC = 46

Success rate: 75% for 9th level, 100% lower level spells.



I think this is a pretty good progression- it also shows that after you've learned a spell it gets easier to cast. There's also the possibility of +Spellcraft items (a special wand or robe), and I estimated the Intelligence bonus rather conservatively I think.


An alternative usage of this system would allow weakened versions of spells if the caster barely misses the DC. If the spellcraft check fails but is within 5 of the DC, the caster may choose to conserve the spell (spell not used, no effect) or use a weak spell with all effects halved. Some spells would require DM discretion on how to be weakened if they do not have any numerical values. The benefit of this alternative is that even a 17th level wizard could cast a 9th level spell with 80% "success" (either full or half).


One last result of this system would be involvement of combast casting rules. Instead of making a concentration check, the Spellcraft check for casting a spell while threatened would be increased by some amount (I dunno, anyone have any ideas?)



Well, that's my proposal. Comments, concerns, questions and critiques are welcome.

Also: post #12 below has additional rules about modifying casting times to increase or decrease the Spellcraft DC

[brian c]

Interesting. So a Wizard 20 with 30 Intelligence, Synergy from Knowledge (Arcana), 23 Ranks invested in Spell Craft and two Feats still has a 25% Chance of Failing to cast a 9th Level Spell?

Yes, and I mentioned exactly that in my second post, along with many more examples.

[Poppatomus]

The issue I have is that it doesn't make the class much less powerful, it just makes them more uneven. Sometimes they will still absolutely win, other times they will be completely useless, and it's largely unpredictable.

What if, rather than making it a flat rate, you tied it to both level and number of spell cast per day? Say you took the old formula, or some other less powerful base check, and instead added to the DC something like:

DC = base + (9*A)+(8*B)+(7*C)....

where each of the letters equals the number of spells of the multipliers level cast that day. you could even make it so that the effect doesn't start until, say, 4th level spells, so 3rd level and below could always be at the base check (allowing a formula that is still non trivial to low level casters that also won't outpace them).

With a system like this you still have some element of the random miss chance effect, but the individual wizards have more control over how fast it stacks up, and thus more direct control over how useful they will be.

EDIT: as I look at the math, that would obviously stack up way to fast, but the idea I think may be sound. Then again, I've got no head for this math stuff, and am way out of my league here i suspect.

[brian c]

The issue I have is that it doesn't make the class much less powerful, it just makes them more uneven. Sometimes they will still absolutely win, other times they will be completely useless, and it's largely unpredictable.

What if, rather than making it a flat rate, you tied it to both level and number of spell cast per day? Say you took the old formula, or some other less powerful base check, and instead added to the DC something like:

DC = base + (9*A)+(8*B)+(7*C)....

where each of the letters equals the number of spells of the multipliers level cast that day. you could even make it so that the effect doesn't start until, say, 4th level spells, so 3rd level and below could always be at the base check (allowing a formula that is still non trivial to low level casters that also won't outpace them).

With a system like this you still have some element of the random miss chance effect, but the individual wizards have more control over how fast it stacks up, and thus more direct control over how useful they will be.

EDIT: as I look at the math, that would obviously stack up way to fast, but the idea I think may be sound. Then again, I've got no head for this math stuff, and am way out of my league here i suspect.


Let me answer your question with an explanatory digression:

Recalling past ideas of this nature (ie chance of spells not working), one of the strongest arguments against is that it nerfs casters too much. Melee characters always have a miss chance as part of their attack roll, but they make up for it by having a theoretically unlimited number of attacks per day. Spellcasters have a limited number of spells, so with really bad luck one could waste and entire day's worth of spells with nothing to show for it. I've tried to eliminate that by making mis-cast spells not be used up.

So, more directly answering your question:

Right now, melee characters can sometimes be awesome and sometimes useless, depending on their die rolls. Even a suboptimal Monk/Samurai can do a lot of damage if they roll critical hits and roll max damage on their weapons (okay, maybe not Monk/Samurai, that would never be good). Other times, a perfectly optimized King of Smack or an Uber-Charger can roll a string of Natural 1s and do nothing. By comparison, the Wizard by RAW never fails. He casts a spell, and it's done. Maybe you get a reflex save or a will save, but he never screws up his spell the way that a fighter can screw up swinging their sword. What this proposal of mine does is it transfers some of the randomness of D&D combat to spellcasters, but in a way that doesn't destroy their versatility.

[Darth Mario]

Interesting concept. Quite similar to the Truenamer class introduced in Tome of Magic, I recommend that you look into it. The basic concept behind that is every "spell" has a (fairly easy) DC to cast based on level, but the DC goes up by 2 every time you cast it (and resets at the end of the day)

[Jasdoif]

Spellcraft is Int-based. So Wis and Cha casters, if they want to use higher level spells, need to ramp up that Int score to make those Spellcraft checks that are challenging for a wizard (and probably to pick up sufficient skill points), in addition to the Wisdom or Charisma they need to access those higher level spells.

Effectively you're forcing MAD upon the cleric, druid, and sorcerer while inconveniencing the wizard. I don't think rigging the primary caster deck in the wizard's favor (more then it already is) is a good plan.

[Poppatomus]

Fair Points Brian_C, and I don't disagree with your conclusion about nerfing casters, its the method you've chosen that bothers me.

The comparison to melee is a good one, so I will steal it. Your current system, translated to melee, would not give you AC and BAB. The way your system works, the difficulty to hit another creature would be set by the level of the warrior, not the target. It might even be modified by the weapon (even more so then it already is, -1 to attack for each damage die for instance.) Fights lose their unique flavor, even if the actual number of misses is the same.

The problem In my mind is that you do transfer some of the randomness, but you do it without transferring some of the skill/enjoyability of the game. Every caster now must max one stat (note maxing BAB does not detract from other skills) and faces the same miss chance casting a 9th level spell to try and distract the baalor from the village as casting that same 9th level spell to teach the commoner a lesson.

I just think that it would be better if the miss chance were keyed either to the opponent, or some aspect of the strategy of the caster. Then at least the player feels like they have more control over their own fate.

[Matthew]

Jasdoif makes a good point. As far as this goes, it seems okay to me. It seems like a Caster Level Check would have been easier to handle and I think, overall, I prefer there to always be some chance of failure (perhaps not 5%, but some chance).

[Jasdoif]

I just realized, this only has a marked effect during encounters. Since there's no penalty for failure, you can take 20 on the Spellcraft check.

Depending on the exact intention here, this might be a good thing or a bad thing.

[brian c]

Spellcraft is Int-based. So Wis and Cha casters, if they want to use higher level spells, need to ramp up that Int score to make those Spellcraft checks that are challenging for a wizard (and probably to pick up sufficient skill points), in addition to the Wisdom or Charisma they need to access those higher level spells.

Effectively you're forcing MAD upon the cleric, druid, and sorcerer while inconveniencing the wizard. I don't think rigging the primary caster deck in the wizard's favor (more then it already is) is a good plan.

I'm not forcing MAD upon anyone. I thought I had made it clear that as this is written, it is designed for Wizards only. Sorcerers I have completely rewritten (there's a link to it in my signature, in case you never saw it), clerics are replaced by cloistered clerics, and druids use the shapeshift variant. Clerics and Druids are still good divine casters, but lose their tanking abilities a little bit.

Sorry if that wasn't clear. Once again, this is specifically designed for my homebrewed setting. I presented it here to gather input, and because it is fairly easily adaptable (As Matthew points out you can use a Caster Level check instead) to other people's games.


Fair Points Brian_C, and I don't disagree with your conclusion about nerfing casters, its the method you've chosen that bothers me.

The comparison to melee is a good one, so I will steal it. Your current system, translated to melee, would not give you AC and BAB. The way your system works, the difficulty to hit another creature would be set by the level of the warrior, not the target. It might even be modified by the weapon (even more so then it already is, -1 to attack for each damage die for instance.) Fights lose their unique flavor, even if the actual number of misses is the same.

The problem In my mind is that you do transfer some of the randomness, but you do it without transferring some of the skill/enjoyability of the game. Every caster now must max one stat (note maxing BAB does not detract from other skills) and faces the same miss chance casting a 9th level spell to try and distract the baalor from the village as casting that same 9th level spell to teach the commoner a lesson.

I just think that it would be better if the miss chance were keyed either to the opponent, or some aspect of the strategy of the caster. Then at least the player feels like they have more control over their own fate.

That is a good point about distracint a balor vs. teaching a commoner- let's try something new. This is off the top of my head, but hopefully you get the idea and know that it can probably be tinkered with.


Casting Time:

For any spell with a casting time of less than "Full-round", the caster may increase the casting time one step (free -> standard -> full-round) in order to lower the Spellcraft DC by 5, or two steps to lower the DC by 10. In this way, a Quickened spell may be cast as a standard action at -5 to Spellcraft DC.

For spells with a casting time of "full-round", the caster may increase casting time to full-round plus a standard action in his next round, and by doing so lower the Spellcraft DC by 5.

Alternatively, a spell with casting time of "full-round" may be cast as a standard action by increasing the Spellcraft check by 5.


This adds an element of strategy that I think you were missing. If the spell is too hard to cast, you could choose to take your time and you're more likely to cast it.


Edit: And yes Jasdoif, casting outside of battle you can take as long as you want to cast the spell perfectly (take 20)

[Poppatomus]

I'm not forcing MAD upon anyone. I thought I had made it clear that as this is written, it is designed for Wizards only. Sorcerers I have completely rewritten (there's a link to it in my signature, in case you never saw it), clerics are replaced by cloistered clerics, and druids use the shapeshift variant. Clerics and Druids are still good divine casters, but lose their tanking abilities a little bit.

Sorry if that wasn't clear. Once again, this is specifically designed for my homebrewed setting. I presented it here to gather input, and because it is fairly easily adaptable (As Matthew points out you can use a Caster Level check instead) to other people's games.



That is a good point about distracint a balor vs. teaching a commoner- let's try something new. This is off the top of my head, but hopefully you get the idea and know that it can probably be tinkered with.


Casting Time:

For any spell with a casting time of less than "Full-round", the caster may increase the casting time one step (free -> standard -> full-round) in order to lower the Spellcraft DC by 5, or two steps to lower the DC by 10. In this way, a Quickened spell may be cast as a standard action at -5 to Spellcraft DC.

For spells with a casting time of "full-round", the caster may increase casting time to full-round plus a standard action in his next round, and by doing so lower the Spellcraft DC by 5.

Alternatively, a spell with casting time of "full-round" may be cast as a standard action by increasing the Spellcraft check by 5.


This adds an element of strategy that I think you were missing. If the spell is too hard to cast, you could choose to take your time and you're more likely to cast it.


Edit: And yes Jasdoif, casting outside of battle you can take as long as you want to cast the spell perfectly (take 20)

Exactly the kind of thing I was talking about. As I said in my first post, I have no kind of head for numbers, but I like the way you've tried to incorporate the idea. Fits very well with your fluff from the beginning, and lacks the needless complexity of my suggestion.

[Jasdoif]

Ahh, I missed the part about it being only for wizards the first time I read it. Sorry.

For adapting to other classes, I would recommend something like...Concentration, instead of a caster level check. You can get modifiers on your Concentration check pretty easily, and the cap is level+3; with a caster level check a level-20 wizard would need +6 in CL boosts to even take 20 to cast a 9th-level spell.


Hmm. Speaking of this, I had an idea for an alternate method of coming up with the DCs.

Caster level + (3 * spell level).

So a CL 20, 9th-level spell has a DC of 47. Pretty similar to your table. And a CL 1, 1st-level spell has a DC of 4. A 1st-level wizard with 4 ranks in Spellcraft automatically makes that. Challenges high level casters without penalizing low level casters, sounds good.

Plus, it has an easy "weakened spell" mechanic. If the check fails, subtract thirce the spell level from the result. That's the effective caster level. If it's high enough to cast the spell in question, the caster can opt to either cast at the reduced CL, or to hold onto the spell for another attempt.

[I_Got_This_Name]

I'd make it a level + int check. DC 10 + minimum level to cast spell, most likely. On a failure, the spell isn't wasted, but your action is, and you still provoke AoOs. Possibly roll this after concentration and the like. Making it a level check deals with the issue of skills being far too easy to augment.

Alternately, 10 + spell level + number of high-level spells cast earlier that day, where all spells within, say, two levels of your highest available spell level are considered high-level.

Also, a wizard would be using 9th-level spells to teach a commoner a lesson in very few circumstances.

[brian c]

Also, a wizard would be using 9th-level spells to teach a commoner a lesson in very few circumstances.

Hey, I said no pickles on my Big Mac! Meteor Swarm!

[Fizban]

I barely skimmed this except for the OP, but I did notice you said it's only for wizards, which I like. I'm also assuming that you've made your DC's in light of spellcraft boosting items, which works for balance, but I still don't like.

For making anything like this based on a skill check, I advise stating that the skill cannot be affected by magic items, and then working with DC's based on only having ranks, ability scores, and feats.

[Gralamin]

I don't see how a Check of 46 is all that high for a 9th level spell.
Say a wizard starts with 18 int, and gets all the standard boosts (5 increase from level, +5 tome, +6 Headband [total price: 36,000 + 137,500 = 173500]) for a final Intelligence of 34, or a +12 to Intelligence.
Then consider a +20 item (price 40 000 or more depending on specifications)

12+23+20 (item)+2 (synergy) = 57, without anything really cheesy. Heck, change the item for a +9 item.
The total price? ~213500 gp. Easily done by even a 17th wizard, and nothing exceeding the single item half gold limit.

[Matthew]

Apparently, there is a limit on Intelligence Bonuses. +6 from Magic and +5 from Tomes/Level Advancement.

Spell Craft Bonus Items will obviously make a huge difference. I doubt Brian is going to allow them in this context, though. However, it is one argument for making it a Caster Level Check.

[barawn]

A neat flavor addition which I was thinking about was to require Spellcraft checks for preparing spells, as well. Again, not under duress, a wizard could prepare all spells (by taking 20) but under duress, it becomes harder - duress such as less than 8 hours sleep (but really, really hard) or injured (per hp, for instance).

[brian c]

Several people have now talked about having +spellcraft items and how that would affect this mechanic. Allow me to share my thoughts:

No item should give more than +5 to spellcraft, under any circumstances. Perhaps some sort of Epic item or Major Artifact could give up to +10, but +5 is the most I would normally allow. Obviously you can use the RAW to make a scarf of Spellcraft +20, but seeing as this variant mechanic isn't in the RAW it's safe to say that some things have to be changed. Like I alluded to in a previous post, something like a wand or staff might give +2 or so to spellcraft, but too much would be unbalancing.

@Fizban: I did not consider such items in my DC calculations. If I had allowed for +20 items, the DC for a 9th level spell would have ended up being something like 65 or 70.

Addendum: Gralamin said "price 40 000 or more depending on specifications" for a +20 item. One specification could be that +spellcraft items now cost 10 or 50 times as much as before. That would at least be a little discouraging

[Eldariel]

Really, if you wanna make anything based on skills, you must ban skill enhancing items - those things just mess it up. We went through this in another thread about remaking the skill system to make more sense; skill checks are problematic because generally you can get like +6 at most to anything from Magic items; skills allow for +30. Further, the same attribute-boosting items happen to help associated skills too so there need not even be an item for those skills specifically as their boosts are already covered. Same goes to spells; stuff like Glibness just messes DCs up. While at it, such limitations also fix (or at least hinder) a bunch of broken classes like the Artificer and the Incantatrix (to a degree, anyways; the DCs are still somewhat low for high level Wizards).