PDA

View Full Version : Skill-based magic? Something of a nerf...



Grod_The_Giant
2011-10-24, 02:52 PM
I was looking at Kyeduo's Truenamer (http://www.giantitp.com/forums/showpost.php?p=11971747&postcount=1)fix last night, and it got me thinking... what if we applied similar rules to magic? It seems to me that requiring a skill check to cast a spell would:

Make the system a little more consistent- you have to roll to do pretty much anything else, after all. I can't say how many times I've had new players try to roll to cast spells.
Add some flavor- magic isn't easy. As it, the only mechanical difficulty for a spellcaster is surviving the first few levels. With skill checks, even powerful casters can fail (more on this later).
Nerf the power of high-level magic some, by basing the DCs on the spell level.

My idea? To cast a spell, you must first make a Spellcraft check with a DC equal to the spell level squared.

That's it. Simple, right? Sound kind of high? Let's look at it level by level.


Level 1 spells- DC 1. Basically impossible to fail- casters have enough trouble at low levels.
Level 2 spells- DC 4- Again, pretty much impossible to fail, provided you put any ranks at all into Spellcraft.
Level 3 spells- DC 9- Still easy.
Level 4 spells- DC 16- Starting to get up there, but then again, so are the spell levels. Shouldn't be too hard to hit with max ranks.
Level 5 spells- DC 25- Harder, but doable.
Level 6 spells- DC 36- We're getting into epic-level checks, but we're also getting high-level casters. They should be dominating encounters by now, no harm in making them work for it a bit more.
Level 7 spells- DC 49- Mildly ridiculous.
Level 8 spells- DC 64- Ouch.
Level 9 spells- DC 81- ...you know what, you want to throw around Gate and Time Stop, you have to earn it.

The nerf part doesn't really kick in until spellcasters are in full swing. The coolest part about this is that it scales well with non-full casting. A bard, for example, will only even need to hit a DC 36 check- perfectly doable at high levels. A high-level duskblade needs, at most, a 16- he doesn't even need to optimize to hit that.

We might need to give a few extra skill points to Clerics, Sorcerers and the like, who only have 2+Int mod per level and, quite possibly, a low intelligence. Maybe a cravat that lets them use their primary casting stat in place of Intelligence on Spellcraft checks to cast spells.

To balance this out- and I'm afraid I really don't know just how difficult it would be for, say, a Sorcerer to hit those last few DCs- we might simplify save DCs as well, setting the save DC for any spell a caster... casts... to 10 + 1/2 CL + primary casting stat modifier.


Thoughts?

Noctis Vigil
2011-10-24, 02:56 PM
Would this effect SR any? Use this check to overcome SR instead (essentially making SR the magical equivalent of DR)?

Chaos_Laicosin
2011-10-24, 03:44 PM
So a level 41 wizard with 30 intelligence and the magical aptitude feat would still need a headband of intellect +6 (36 000 gp) and would have to spend a week reading a tome of understanding +4 (110 000gp) to get their intelligence up to 40 and their spellcraft up +61. Then they would only have to roll a natural 20 on their spellcraft check to be able to cast wish and get another +1 to their intelligence, or revive someone, or create an item worth up to 25 000 gp!

Treblain
2011-10-24, 03:58 PM
IMO, there should not be any skill-based rules that acknowledge the fact that optimization can get you absurdly high skill checks. It just doesn't work. D&D skills were written without that sort of optimization in mind, and it shows painfully whenever skills are involved. At DCs past 40, the designers just threw up their hands and said, "Do whatever you want."

Making any subsystem that depends on you getting really high skill checks is just going to draw attention to this problem. The numbers just get too far away from the d20 roll.

Grod_The_Giant
2011-10-24, 07:21 PM
So a level 41 wizard with 30 intelligence and the magical aptitude feat would still need a headband of intellect +6 (36 000 gp) and would have to spend a week reading a tome of understanding +4 (110 000gp) to get their intelligence up to 40 and their spellcraft up +61. Then they would only have to roll a natural 20 on their spellcraft check to be able to cast wish and get another +1 to their intelligence, or revive someone, or create an item worth up to 25 000 gp!

Eesh. Ok, that's worse than I thought. I knew the numbers were bad, but not this bad- I've really never played much above 10th level. I was just hoping that "spell level squared" would work, because it seemed so simple...

What about... hmm... spell level times 5? Then we only go up to a DC 45 for 9th level spells. If we have, at level 17, 20 ranks, a +10 modifier (should be doable with the requisite Headband of Int), and Skill Focus, we have a total of +33, requiring a roll of 12 or above to cast the spell, or a little more than 50-50.


IMO, there should not be any skill-based rules that acknowledge the fact that optimization can get you absurdly high skill checks. It just doesn't work. D&D skills were written without that sort of optimization in mind, and it shows painfully whenever skills are involved. At DCs past 40, the designers just threw up their hands and said, "Do whatever you want."

Making any subsystem that depends on you getting really high skill checks is just going to draw attention to this problem. The numbers just get too far away from the d20 roll.

On the other hand, the general idea is to put some kind of stopper on high-level magic abuse that doesn't feel like a brutally arbitrary nerf.

RedWarrior0
2011-10-24, 08:29 PM
Well, that ignores the fact that there are a bazillion and a half other ways to boost a skill check.

Treblain
2011-10-24, 08:44 PM
Just because there are many ways to boost skills is no reason to decide DCs have to be ridiculously high to account for that. It just fuels an arms race between player and DM to have higher skill bonuses. If you're familiar with Truenamers and Zaq's effort to play one (http://www.giantitp.com/forums/showthread.php?t=114269), you find that yeah, if you try, you can reach those high DCs. But it's not enjoyable. Casters will go to any length to avoid the complications of the skill check, so they spend all their wealth, feats, skill points, etc. in optimizing their Spellcraft check.

Which means you might as well have just arbitrarily taken away those feat slots, skill points, and wealth in the first place.

NeoSeraphi
2011-10-24, 09:25 PM
Want to roll to cast spells? Want to give spellcasters a failure chance? How about just a straight up 5% failure chance for all spells per spell level? Bear with me here.

Cantrips/Orisons- Auto Succeed

1st Level- A 5% failure chance is the equivalent of rolling a natural 1 on an attack roll. Suddenly, magic's most basic spells are at the same risk as melee! Good!

2nd Level- Generally speaking, only the most optimized melee combatants will hit an opponent of their CR on a natural 2. This is still fair.

3rd Level- 15% chance of failure may seem high, but we're getting into deadly territory here. Stinking cloud, haste, animate dead, these things are powerful.

4th Level- A 1 in 5 chance of failure. Now, in order to avoid failure, spellcasters will need to use their lower powered spells to finish encounters off, or risk screwing up with their attempts.

5th Level- A 25% chance of failure. Now this is what I'm talking about. If you end up at 20th level with 6 5th level spells per day, odds are good that you'll fail to cast at least one of them.

6th Level- A 30% miss chance is not quite 1 in 3, but its close. Still, with power like harm and flesh to stone coming online, we need to really restrict this kind of things.

7th Level-35%. It's official, we are now in the 1/3 failure territory. But with gems like destruction here, you only really need to use one 7th level spell to take care of things, so you can deal with one messed up casting (It gives your party more chances to shine before you just end everyone)

8th Level- A 40% chance of failure. This is more what the OP was intending for these high level spells, and I don't blame them.

9th Level- 45%. Just under the same chance of flipping a coin. And with this chance of failure affecting spell-like abilities as well, chain-gating is no longer viable, nor is chaining wish for higher stats or making free stuff.

Grod_The_Giant
2011-10-24, 10:16 PM
Want to roll to cast spells? Want to give spellcasters a failure chance? How about just a straight up 5% failure chance for all spells per spell level? Bear with me here.

That...works pretty well, really, and is pretty much impossible to get around. Which in its own way is almost a problem... I don't know, I probably just want to have my cake and eat it too, but... the appeal of the skill check, I guess, is that it doesn't feel like it's out of the player's hands, you know? You have to be skilled to cast the spell. But everyone who's posted has been right- either the DCs are high enough that major optimization is needed to reach them, or they're low enough that a tiny optimization effort lets you blow right through them.

Darn.

Noctis Vigil
2011-10-24, 10:27 PM
Personally, I'd go with the skill check (no offense, NeoSeraphi). Maybe a DC of 10+(spell level*5)? This would give a 1st level spell at 1st level a DC of 15, so assuming you put 4 ranks in Spellcraft right off the bat, you need an 11 or higher to hit it. Seems to me like you would pretty much always cast your highest level spells with roughly the same level of accuracy for most your carrier as a caster with a little bit of optimization. Of course, it becomes fairly easy once you hit epic level, but not too many people actually play that high, and frankly, if they play that high, they've earned it.

NeoSeraphi
2011-10-24, 10:36 PM
That...works pretty well, really, and is pretty much impossible to get around. Which in its own way is almost a problem... I don't know, I probably just want to have my cake and eat it too, but... the appeal of the skill check, I guess, is that it doesn't feel like it's out of the player's hands, you know? You have to be skilled to cast the spell. But everyone who's posted has been right- either the DCs are high enough that major optimization is needed to reach them, or they're low enough that a tiny optimization effort lets you blow right through them.

Darn.

I understand your point. I was offering a pretty straightforward, consistent chance of failure that offset the ridiculous quadrilateral scaling of spells. But by putting it in the hands of the player just makes them put more work and WBL in order to get the same effectiveness of the spells that we have now.

Yitzi
2011-10-24, 10:57 PM
Skill-based spells are a bad idea unless you ban (or at least vastly increase the cost of) magic items that boost the relevant skill. Otherwise spell power is far too variable based on how much money you invest into it.

Making spells skill-based moves things further away from the goal of "unoptimized characters can still survive, optimized characters don't dominate everything", often irredeemably so.

Grod_The_Giant
2011-10-24, 11:35 PM
I understand your point. I was offering a pretty straightforward, consistent chance of failure that offset the ridiculous quadrilateral scaling of spells. But by putting it in the hands of the player just makes them put more work and WBL in order to get the same effectiveness of the spells that we have now.

No, no, you're right. I'm not sure I would ever use either variant in a campaign, but of the two, I think yours is probably superior.

jiriku
2011-10-25, 01:04 AM
The difficulty in making any change to the spellcasting mechanic is that so much of D&D's superstructure sits on top of it. A dice-based spellcasting mechanic works extremely well in certain other systems, like AEG's 3rd- and 4th-edition L5R, but in that system, the mechanic was set in stone before the core rulebook even hit its first printing, and everything else was built on that foundation. Not so in 3.5. Adding any core mechanics is like wiggling a block loose from an extremely tall Jenga tower and hoping for the best.

Kobold-Bard
2011-10-25, 01:52 AM
Maybe combine the two: NeoSeraphi's scaling failure rate, with a skill check to reduce it by a certain amount for an individual spell spell?

Qwertystop
2011-10-25, 07:46 AM
I was offering a pretty straightforward, consistent chance of failure that offset the ridiculous quadrilateral scaling of spells.
Quadrilateral scaling? never heard of that before. :smallbiggrin:

Maybe combine the two: NeoSeraphi's scaling failure rate, with a skill check to reduce it by a certain amount for an individual spell spell?
Looks good to me.

Zeta Kai
2011-10-25, 08:49 AM
either the DCs are high enough that major optimization is needed to reach them, or they're low enough that a tiny optimization effort lets you blow right through them.

Which illustrates the problem with skills in the D20 system: Because there are so many ways to boost a skill check, & not everybody knows about them (or doesn't use them all because some don't fit a character concept), it is just about impossible to set DCs that work for all players. A high DC won't provide even a mild challenge for an optimizer, but will be far beyond the capabilities of a character run by a non-optimizer. The bonuses from different sources can be combined to make the check pointless, but if they aren't used (or used "poorly"), then the check is hopeless.

The only way around this would be if the skill check were for a completely new skill, one that did not have bonuses from countless splatbooks released over the years by different authors. This new skill would have to be accessible to all casting classes equally, even partial casters. Essentially, it would/should be a retooled CL check, tied to the character's primary casting stat.

Elfinor
2011-10-25, 09:24 AM
The only way around this would be if the skill check were for a completely new skill, one that did not have bonuses from countless splatbooks released over the years by different authors. This new skill would have to be accessible to all casting classes equally, even partial casters. Essentially, it would/should be a retooled CL check, tied to the character's primary casting stat. I've never had a super-hardcore optimiser in group but there's probably still ways around it (mind blank ATM) - unless it was not phrased as a skill check as such, a CL check is a way around that and gets rid of the whole "You must put points in this skill!" vibe (and adding skill points to classes) but... (http://community.wizards.com/go/thread/view/75882/19872350/Raising_Caster_Level). That still sounds like an excellent improvement though. If I may be so bold as to make some additional suggestions to add to that improvement:
Spellcraft provides a synergy bonus
The ability to increase skill checks through the use of rituals (represented through other skill checks, foci, XP/material component cost, caster backlash and/or increasing casting time etc.) and a relevant Knowledge (Arcana/Religion/Nature) synergy bonus for the ritual aspect only.
EDIT: Backlash (spell level squared nonlethal damage?) on a check that's failed by 5 or more could be another interesting addition. Maybe WIS/CON damage for the highest level spells or 10+ points failed checks.