Spellcasting as a skill mechanic
It's really hard to balance. The DC range on skills is really pretty narrow. If you make the check easy enough that "average" casters succeed their check a reasonable percentage of the time, then your typical min-maxed PC's will never fail.
You also need some kind of penalty in place to prevent casters from simply casting their highest level spell every round.
One of the better compromises I've seen is to make the DC's fairly hard, but then let the PC's gain a bonus on the skill check equal to the amount of non-lethal damage they take at casting time. (though then you have to decide what you want to do about healing spells...)
You also need some kind of penalty in place to prevent casters from simply casting their highest level spell every round.
One of the better compromises I've seen is to make the DC's fairly hard, but then let the PC's gain a bonus on the skill check equal to the amount of non-lethal damage they take at casting time. (though then you have to decide what you want to do about healing spells...)
Arkhandus
First Post
Well, I've got a few things to chime in with.
First, I think a skill-based spellcasting system for d20/D&D was done in a product called....True Sorcery, I think. It was a conversion of many core D&D spells into a skill-based casting system, meant to be compatible with D&D and True20, if I recall correctly. Now, I could be misremembering this a bit; there was another kind of project/product done shortly before or after that by someone else, that might be the skill-based one.
Secondly, the basic things that need to be considered for such a system are....
[sblock]1) There needs to be some means of limiting how many spells they characters can cast each day or whatever, otherwise they'll be using magic for everything, even picking their noses with Mage Hand whenever they're bored. -_- Average joe NPCs would love to try and get 1 level of adept, wizard, or whatever if it meant they could use Prestidigitation or Mage Hand or Mending all the time.
You can impose some limit based on skill ranks (like, no more than 1 spell per day per rank of Spellcraft or Concentration or whatever, for example, maybe saying that every three levels of the spell count as an additional spell, even), or make each successful (or even each failed) spell add a bonus to the DC of all further spellcasting skill checks that day, or may, or have each spell cause some harm to the caster each time (like 1 damage per spell level, that can't be prevented, reduced, or redirected, since it's magical drain/backlash; maybe make it ability damage, like 1 point of Charisma damage per 3 spell levels, rounded up, for spontaneous arcane spells, or Intelligence damage for prepared arcane spells, or Wisdom damage for divine spells; if doing so, the required ability score for spellcasting should probably just be 10, or 3, rather than "10 + spell level").
2) Skill ranks and skill check results vary a lot, so you need to take that into account. A 1st-level sorcerer might have 1 rank of Concentration/Knowledge-Arcana/Spellcraft, or he might have 4 ranks in them; he might have a +4 modifier from Con/Int, or a +0 mod; he might have Skill Focus and Magical Aptitude or whatever for another 2, 3, or 5 points of bonus. It gets even more varied at middle and upper levels.
Yet, the results they can get will also be quite varied. They might roll a 1, or a 10, or a 20, or whatever. Joe the incompetent sorcerer might have a mere +1 on the skill check, but he could still get a result of 21 with good luck.
Bob the master sorcerer (....apprentice) might have a +13 on the skill check, and might've been clever enough to get his buddies Moe, Larry, and Curly to use "aid another" attempts for him because they spent a few skill points on a cross-class rank of Spellcraft/whatever in order to use it trained. Bob will get a 14 minimum on the check, or 20 with those buddies "aiding another", or 22 if he gets his slightly-higher-level friend Billy the Bard to sing a verse with the Inspire Competence bardic music effect. Bob could get a 41 total in that case if he were lucky enough to roll a 20 on the die.
By comparison, Jerry the Archmage might have 23 ranks in the skill and a +4 modifier on top of that, because he hasn't bothered to do anything magic-wise or feat-wise to boost his skills or the relevant ability scores, giving him a range of 28 to 47 on his spellcasting skill check. Meanwhile, Adam the Archmage might have 23 ranks and a +30 or +40 modifier from various stuff because he's so slick and munchkinny, for a range of like 54 or 64 to 60 or 70 on the spellcasting check.
In other words, no matter what you set the check DCs to, know that some PCs will invariably find a way to ace them more or less every time, while others will have a difficult time succeeding most of the time unless you make the DCs very reasonably set for the average joe. 15 + spell level is reasonably tough for low-level characters but easy or automatic success for mid or high level casters.
A check DC of 5 or 10, + five times the spell's level, wouldn't be too bad for low-level spells, but would get out of hand after a bit; like 9th-level wizards trying to make the 30 or 35 check DC for their 5th-level spells (with the right powergaming they just might, maybe, be able to get +25 on the check in order to make it half the time), or a 17th-level wizard trying to cast their 9th-level spells with DC 50 or 55 checks (they can be expected to maybe have +35 on the check, but expecting +40 would be seriously pushing it, and +45 or +50 would require very very meticulous powergaming skillz or an obscure/custom magic item that provides +10 or +20 on the check).
As a sort of general rule, you can reasonably expect a PC to have a total skill bonus in their most important skill equal to 5 + character level + (1 per 4 levels), on average (assuming that it really is important, not just a sometimes-kinda-useful skill). Add 2-10 points for varying degrees of optimization, if any (i.e. factoring in exceptional ability scores, Skill Focus, other skill-boosting feats, masterwork tools in some cases, rare magic items at upper levels, etc.).
3) You also have to determine how often you want the PCs to succeed at spellcasting, whether or not you want higher-level spells to be even more difficult or painful to cast compared to low-level ones (in proportion at the appropriate level, taking into account the greater skill bonuses or hit points/saves/etc. at higher levels), and such.
Also, be sure to include some means of preventing either auto-success or infinite spells; maybe the skill check still causes some harm/drain even when successful, just maybe not as much as when it fails, or maybe any natural roll (that is, before modifiers like skill ranks and ability scores are factored in) on the skill check that comes out equal to or lower than the spell's level automatically fails (or automatically causes some backlash/drain). You don't want somebody to just cast healing spells to completely erase their backlash damage from spellcasting, for instance. If they're going to do so, it should at least have a different price associated with it (.....Cthulhu invades the wizard's psyche and devours it after he's caused himself X amount of magical backlash in one day, regardless of how much of X he healed with magic earlier....).
![:]](http://www.enworld.org/forum/images/smilies/devious.png "Devious :]")
[/sblock]
Thirdly, here's what I got in some notes for a sorta-low/uncommon-magic campaign. This note was only written down for wizards, because I hadn't gotten to writing up the slightly different bits of info for other classes, but it can be pretty easily applied to the other casters. This was for a campaign with no spellcasting classes, but instead feats could be taken to gain spellcasting (it required something like 5 or 6 feats, and an average of 1 or 2 skill points per level lost to those feats' drawbacks, to gain full spellcasting ability over time akin to a caster class).
[sblock]Wizards: A Knowledge (arcana) check is needed to learn a new wizard spell, and in order to learn a wizard spell, the character needs an Intelligence score of at least 10 + the spell's level, as well as at least 2 ranks of Knowledge (arcana) per spell level (1 rank for 0-level spells). To learn the spell is a check against DC 15 + the spell's level. Maximum number of wizard spells known is Intelligence modifier + the number of Knowledge (arcana) ranks possessed, but a spellbook, magerod, or spellstave allows twice that many to be known, if all are stored in the item.
Casting a wizard spell requires an Intelligence score of at least 10 + the spell's level, and also requires a Spellcraft check against a DC of 10 + double the spell's level. Failing the Spellcraft check by less than 5 points results in the spell fizzling after spending only a move-equivalent action in the attempted casting, but it still provokes attacks of opportunity as normal for casting a spell. If casting defensively, the Concentration check is made first, and no Spellcraft check is necessary if the Concentration check fails.
Failing the Spellcraft check by 5 or more points results in spending the entire normal casting time before the spell fizzles, as well as the caster suffering 1 point of subdual damage (nonlethal in 3.5 rules) plus an additional 1 point per spell level. If the caster is immune to subdual damage (nonlethal in 3.5 rules), they instead take normal damage.
Each wizard spell successfully cast, or each one that failed by 5+ points on the Spellcraft check, increases the Spellcraft DC for wizard spells for the rest of the day. This increase is +1 DC per 3 spell levels, rounded up. Cannot take 10 or take 20 on the Knowledge and Spellcraft checks for these purposes. The character can only cast wizard spells they have prepared that day. Only 1 spell per rank of Knowledge [arcana] can be prepared each day, plus a number of spells equal to the character's Intelligence modifier.
Preparing a spell takes 10 minutes. A prepared spell can be replaced by preparing a different spell, but this takes an additional 10 minutes. If the character uses a spellbook, magerod, or spellstave to know more spells than they could through memory alone, then they must have that item present and study it while preparing their spells, or else they cannot prepare any wizard spells. A prepared spell remains prepared until 24 hours have passed since its preparation, or until replaced by another prepared spell.
Casting a spell does not remove it from memory. Only spells from the wizard spell list can be learned and cast as wizard spells. The wizard spell list is not considered to be your spell list unless you have the Wizard feat or the Specialist Wizard feat. If you have neither feat, then only the wizard spells you actually know are considered part of your spell list. The save DC for any saving throws against a wizard spell is equal to 10 + the spell's level + the caster's Intelligence modifier.[/sblock]
Here's how I handle something else like this in one of my homebrew D&D subsystems; a variation on normal arcane spellcasting. The whole thing is very different, but right here I'm just copying the sections on Eldritch Stamina and Arcane Threshold.
[sblock]Eldritch Stamina: Each level gained in the spellbinder class grants 2 points of Eldritch Stamina (ES), which is a measure of how much aether can be wielded in any given 24-hour period without risk of dire consequences. Each arcane spell that is cast (or which is interrupted during casting) may drain points of Eldritch Stamina, but ES is replenished after 8 hours or more of uninterrupted sleep, though only once per day. Whenever the spellbinder attempts to cast an arcane spell, they must attempt a Spellcraft check, with a DC equal to 15 + the spell's aether cost.
This skill check does not affect the success or failure of the spell, it merely determines if they will suffer for ignoring the general limit of their abilities. Any hit points lost by failing this Spellcraft check do not require a Concentration check to continue casting the spell. Failing the check expends an amount of Eldritch Stamina for the day equal to the spell's aether cost, and failing the check by 10 points or more results in 1d4 temporary Constitution damage to the spellbinder.
If the character's Eldritch Stamina was already at 0, or if this failed check reduces ES below 0, the excess amount lost beyond 0 is converted instead into equal parts temporary Wisdom damage and temporary Constitution damage. Round up for determining how much is Wisdom and round down for determining how much is Constitution. This cannot reduce effective Wisdom below 0, so any excess temporary Wisdom damage beyond that is instead converted to additional temporary Constitution damage. As normal, if a creature's Constitution score is reduced to 0 by any means, they die instantly or are destroyed instantly. Any creature that lacks a Constitution score altogether takes temporary Charisma damage in place of temporary Constitution damage, when caused by failing these Spellcraft checks. In addition, such creatures suffer 1 hit point of damage per hit die for each point of Charisma damage suffered this way.
Arcane Threshold: At each level in this class, the character's Arcane Threshold (AT) increases by 1 point. A character has no Arcane Threshold until they gain at least one spellbinder level or any other class level that grants arcane spells. The AT is a gauge which determines just how much aether the character can typically harness for any given spell, and exceeding the AT to cast a more powerful spell is quite dangerous because it goes beyond the character's known limits of magical strain.
Whenever the spellbinder attempts to cast an arcane spell with more aether than their current AT, they must attempt a Will saving throw (DC = 10 + total aether cost of the spell). Failing this save results in the character being stunned for 1d4 rounds after the spellcasting attempt, and the character's AT is temporarily reduced for 24 hours, with the reduction equal to one-half the amount by which the save was failed, rounded up (but AT cannot be reduced below 0). Also, failing the save results in the spellbinder suffering subdual (nonlethal in 3.5 rules) damage equal to the total aether cost of the spell, which then requires them to make a Concentration check against the damage or else lose the spell.
Success on the saving throw still results in half this damage, rounded up, so even succeeding at the save doesn't necessarily prevent them from needing a Concentration check to finish casting the spell. Succeeding at the save does prevent their AT from being temporarily reduced though, and does prevent the stun. The subdual (nonlethal in 3.5 rules) damage cannot be otherwise reduced or prevented, as it is direct magical feedback from trying to draw in and control too much aether. If the caster is immune to subdual (nonlethal in 3.5 rules) damage, then this feedback is normal damage instead.[/sblock]
First, I think a skill-based spellcasting system for d20/D&D was done in a product called....True Sorcery, I think. It was a conversion of many core D&D spells into a skill-based casting system, meant to be compatible with D&D and True20, if I recall correctly. Now, I could be misremembering this a bit; there was another kind of project/product done shortly before or after that by someone else, that might be the skill-based one.
Secondly, the basic things that need to be considered for such a system are....
[sblock]1) There needs to be some means of limiting how many spells they characters can cast each day or whatever, otherwise they'll be using magic for everything, even picking their noses with Mage Hand whenever they're bored. -_- Average joe NPCs would love to try and get 1 level of adept, wizard, or whatever if it meant they could use Prestidigitation or Mage Hand or Mending all the time.
You can impose some limit based on skill ranks (like, no more than 1 spell per day per rank of Spellcraft or Concentration or whatever, for example, maybe saying that every three levels of the spell count as an additional spell, even), or make each successful (or even each failed) spell add a bonus to the DC of all further spellcasting skill checks that day, or may, or have each spell cause some harm to the caster each time (like 1 damage per spell level, that can't be prevented, reduced, or redirected, since it's magical drain/backlash; maybe make it ability damage, like 1 point of Charisma damage per 3 spell levels, rounded up, for spontaneous arcane spells, or Intelligence damage for prepared arcane spells, or Wisdom damage for divine spells; if doing so, the required ability score for spellcasting should probably just be 10, or 3, rather than "10 + spell level").
2) Skill ranks and skill check results vary a lot, so you need to take that into account. A 1st-level sorcerer might have 1 rank of Concentration/Knowledge-Arcana/Spellcraft, or he might have 4 ranks in them; he might have a +4 modifier from Con/Int, or a +0 mod; he might have Skill Focus and Magical Aptitude or whatever for another 2, 3, or 5 points of bonus. It gets even more varied at middle and upper levels.
Yet, the results they can get will also be quite varied. They might roll a 1, or a 10, or a 20, or whatever. Joe the incompetent sorcerer might have a mere +1 on the skill check, but he could still get a result of 21 with good luck.
Bob the master sorcerer (....apprentice) might have a +13 on the skill check, and might've been clever enough to get his buddies Moe, Larry, and Curly to use "aid another" attempts for him because they spent a few skill points on a cross-class rank of Spellcraft/whatever in order to use it trained. Bob will get a 14 minimum on the check, or 20 with those buddies "aiding another", or 22 if he gets his slightly-higher-level friend Billy the Bard to sing a verse with the Inspire Competence bardic music effect. Bob could get a 41 total in that case if he were lucky enough to roll a 20 on the die.
By comparison, Jerry the Archmage might have 23 ranks in the skill and a +4 modifier on top of that, because he hasn't bothered to do anything magic-wise or feat-wise to boost his skills or the relevant ability scores, giving him a range of 28 to 47 on his spellcasting skill check. Meanwhile, Adam the Archmage might have 23 ranks and a +30 or +40 modifier from various stuff because he's so slick and munchkinny, for a range of like 54 or 64 to 60 or 70 on the spellcasting check.
In other words, no matter what you set the check DCs to, know that some PCs will invariably find a way to ace them more or less every time, while others will have a difficult time succeeding most of the time unless you make the DCs very reasonably set for the average joe. 15 + spell level is reasonably tough for low-level characters but easy or automatic success for mid or high level casters.
A check DC of 5 or 10, + five times the spell's level, wouldn't be too bad for low-level spells, but would get out of hand after a bit; like 9th-level wizards trying to make the 30 or 35 check DC for their 5th-level spells (with the right powergaming they just might, maybe, be able to get +25 on the check in order to make it half the time), or a 17th-level wizard trying to cast their 9th-level spells with DC 50 or 55 checks (they can be expected to maybe have +35 on the check, but expecting +40 would be seriously pushing it, and +45 or +50 would require very very meticulous powergaming skillz or an obscure/custom magic item that provides +10 or +20 on the check).
As a sort of general rule, you can reasonably expect a PC to have a total skill bonus in their most important skill equal to 5 + character level + (1 per 4 levels), on average (assuming that it really is important, not just a sometimes-kinda-useful skill). Add 2-10 points for varying degrees of optimization, if any (i.e. factoring in exceptional ability scores, Skill Focus, other skill-boosting feats, masterwork tools in some cases, rare magic items at upper levels, etc.).
3) You also have to determine how often you want the PCs to succeed at spellcasting, whether or not you want higher-level spells to be even more difficult or painful to cast compared to low-level ones (in proportion at the appropriate level, taking into account the greater skill bonuses or hit points/saves/etc. at higher levels), and such.
Also, be sure to include some means of preventing either auto-success or infinite spells; maybe the skill check still causes some harm/drain even when successful, just maybe not as much as when it fails, or maybe any natural roll (that is, before modifiers like skill ranks and ability scores are factored in) on the skill check that comes out equal to or lower than the spell's level automatically fails (or automatically causes some backlash/drain). You don't want somebody to just cast healing spells to completely erase their backlash damage from spellcasting, for instance. If they're going to do so, it should at least have a different price associated with it (.....Cthulhu invades the wizard's psyche and devours it after he's caused himself X amount of magical backlash in one day, regardless of how much of X he healed with magic earlier....).
[/sblock]Thirdly, here's what I got in some notes for a sorta-low/uncommon-magic campaign. This note was only written down for wizards, because I hadn't gotten to writing up the slightly different bits of info for other classes, but it can be pretty easily applied to the other casters. This was for a campaign with no spellcasting classes, but instead feats could be taken to gain spellcasting (it required something like 5 or 6 feats, and an average of 1 or 2 skill points per level lost to those feats' drawbacks, to gain full spellcasting ability over time akin to a caster class).
[sblock]Wizards: A Knowledge (arcana) check is needed to learn a new wizard spell, and in order to learn a wizard spell, the character needs an Intelligence score of at least 10 + the spell's level, as well as at least 2 ranks of Knowledge (arcana) per spell level (1 rank for 0-level spells). To learn the spell is a check against DC 15 + the spell's level. Maximum number of wizard spells known is Intelligence modifier + the number of Knowledge (arcana) ranks possessed, but a spellbook, magerod, or spellstave allows twice that many to be known, if all are stored in the item.
Casting a wizard spell requires an Intelligence score of at least 10 + the spell's level, and also requires a Spellcraft check against a DC of 10 + double the spell's level. Failing the Spellcraft check by less than 5 points results in the spell fizzling after spending only a move-equivalent action in the attempted casting, but it still provokes attacks of opportunity as normal for casting a spell. If casting defensively, the Concentration check is made first, and no Spellcraft check is necessary if the Concentration check fails.
Failing the Spellcraft check by 5 or more points results in spending the entire normal casting time before the spell fizzles, as well as the caster suffering 1 point of subdual damage (nonlethal in 3.5 rules) plus an additional 1 point per spell level. If the caster is immune to subdual damage (nonlethal in 3.5 rules), they instead take normal damage.
Each wizard spell successfully cast, or each one that failed by 5+ points on the Spellcraft check, increases the Spellcraft DC for wizard spells for the rest of the day. This increase is +1 DC per 3 spell levels, rounded up. Cannot take 10 or take 20 on the Knowledge and Spellcraft checks for these purposes. The character can only cast wizard spells they have prepared that day. Only 1 spell per rank of Knowledge [arcana] can be prepared each day, plus a number of spells equal to the character's Intelligence modifier.
Preparing a spell takes 10 minutes. A prepared spell can be replaced by preparing a different spell, but this takes an additional 10 minutes. If the character uses a spellbook, magerod, or spellstave to know more spells than they could through memory alone, then they must have that item present and study it while preparing their spells, or else they cannot prepare any wizard spells. A prepared spell remains prepared until 24 hours have passed since its preparation, or until replaced by another prepared spell.
Casting a spell does not remove it from memory. Only spells from the wizard spell list can be learned and cast as wizard spells. The wizard spell list is not considered to be your spell list unless you have the Wizard feat or the Specialist Wizard feat. If you have neither feat, then only the wizard spells you actually know are considered part of your spell list. The save DC for any saving throws against a wizard spell is equal to 10 + the spell's level + the caster's Intelligence modifier.[/sblock]
Here's how I handle something else like this in one of my homebrew D&D subsystems; a variation on normal arcane spellcasting. The whole thing is very different, but right here I'm just copying the sections on Eldritch Stamina and Arcane Threshold.
[sblock]Eldritch Stamina: Each level gained in the spellbinder class grants 2 points of Eldritch Stamina (ES), which is a measure of how much aether can be wielded in any given 24-hour period without risk of dire consequences. Each arcane spell that is cast (or which is interrupted during casting) may drain points of Eldritch Stamina, but ES is replenished after 8 hours or more of uninterrupted sleep, though only once per day. Whenever the spellbinder attempts to cast an arcane spell, they must attempt a Spellcraft check, with a DC equal to 15 + the spell's aether cost.
This skill check does not affect the success or failure of the spell, it merely determines if they will suffer for ignoring the general limit of their abilities. Any hit points lost by failing this Spellcraft check do not require a Concentration check to continue casting the spell. Failing the check expends an amount of Eldritch Stamina for the day equal to the spell's aether cost, and failing the check by 10 points or more results in 1d4 temporary Constitution damage to the spellbinder.
If the character's Eldritch Stamina was already at 0, or if this failed check reduces ES below 0, the excess amount lost beyond 0 is converted instead into equal parts temporary Wisdom damage and temporary Constitution damage. Round up for determining how much is Wisdom and round down for determining how much is Constitution. This cannot reduce effective Wisdom below 0, so any excess temporary Wisdom damage beyond that is instead converted to additional temporary Constitution damage. As normal, if a creature's Constitution score is reduced to 0 by any means, they die instantly or are destroyed instantly. Any creature that lacks a Constitution score altogether takes temporary Charisma damage in place of temporary Constitution damage, when caused by failing these Spellcraft checks. In addition, such creatures suffer 1 hit point of damage per hit die for each point of Charisma damage suffered this way.
Arcane Threshold: At each level in this class, the character's Arcane Threshold (AT) increases by 1 point. A character has no Arcane Threshold until they gain at least one spellbinder level or any other class level that grants arcane spells. The AT is a gauge which determines just how much aether the character can typically harness for any given spell, and exceeding the AT to cast a more powerful spell is quite dangerous because it goes beyond the character's known limits of magical strain.
Whenever the spellbinder attempts to cast an arcane spell with more aether than their current AT, they must attempt a Will saving throw (DC = 10 + total aether cost of the spell). Failing this save results in the character being stunned for 1d4 rounds after the spellcasting attempt, and the character's AT is temporarily reduced for 24 hours, with the reduction equal to one-half the amount by which the save was failed, rounded up (but AT cannot be reduced below 0). Also, failing the save results in the spellbinder suffering subdual (nonlethal in 3.5 rules) damage equal to the total aether cost of the spell, which then requires them to make a Concentration check against the damage or else lose the spell.
Success on the saving throw still results in half this damage, rounded up, so even succeeding at the save doesn't necessarily prevent them from needing a Concentration check to finish casting the spell. Succeeding at the save does prevent their AT from being temporarily reduced though, and does prevent the stun. The subdual (nonlethal in 3.5 rules) damage cannot be otherwise reduced or prevented, as it is direct magical feedback from trying to draw in and control too much aether. If the caster is immune to subdual (nonlethal in 3.5 rules) damage, then this feedback is normal damage instead.[/sblock]
Aus_Snow
First Post
I've used Legends of Sorcery. . . along with many house rules. It worked very well. Tweak to taste, as always.
But the nice thing is, you can use just about any d20 fantasy spell, as is, and you don't have to rely on players coming up with totally appropriate and accurate spells/effects on the fly, or whatever. It's nice and simple, and also provides guidelines for low, medium and high magic campaigns, with a range of classes to suit each (with D&D core full spellcasters assumed for the 'high' setting.)
But the nice thing is, you can use just about any d20 fantasy spell, as is, and you don't have to rely on players coming up with totally appropriate and accurate spells/effects on the fly, or whatever. It's nice and simple, and also provides guidelines for low, medium and high magic campaigns, with a range of classes to suit each (with D&D core full spellcasters assumed for the 'high' setting.)
DungeonmasterCal
First Post
Legends of Sorcery is good. I'm using it in a current campaign and I like it for all the reasons Aus does.
Spatzimaus
First Post
In a classic Vancian slot system, replacing core mechanics with skills can be tremendously hard to balance. However, we've done a lot of tweaking for the homebrew my group made up, linked here. The short version:
There are a lot of little pieces of the existing spell system you can replace with skills without imbalancing the system:
> Caster level can be replaced with a skill, allowing you to minimize the loss in effectiveness for multiclassing. (You'd still lose spell slots, of course.)
> Replace the ranged attack rolls for rays, etc. with a skill check (effectively giving casters a full BAB progression when using spells).
> Allow casters to replace their saving throw rolls for certain spells with a skill check (to deflect/disrupt the attacker)
These definitely improve the feel of the existing system, without requiring many extra skill checks. You would need to rebalance the caster class' skill points, though.
But if you're talking about dumping the slot system altogether, we've got an entire class (the Channeler) that uses drain-based spellcasting. That is, instead of spell slots, the caster takes some damage depending on his level and the spell's level (and you then need a Concentration check to succeed in using the spell). There's a pseudo-skill used to make the drain somewhat random, but that's not strictly necessary (and frankly ended up requiring a bit too much rolling).
The main reason I bring this up is because of a common concern you hear: that in skill-based systems, characters become able to use low-level stuff almost at will, and it breaks the system. Well, it's true, sort of. They WILL be able to use the low-end stuff ad nauseum when they get to high level, but that's not always a bad thing, and it only breaks the system if you leave the spell list as-is. An infinite supply of cure light wounds spells sounds abuseable, but at high level most parties I've been in have just made wands of that for the same thing, with only a paltry resource cost. The bigger problem is unlimited stat buffs.
The other side effect is that you, as DM, can't depend on spell depletion to make normally-easy fights challenging by putting them at the end of a day, but I've never liked that method anyway. Too much like railroading.
You'd also need to make sure the system had a method of mitigating "alpha strikes". That is, in 3E a wizard can dump all his 8th and 9th-level spells in a single fight, and the only thing holding him back is the knowledge he'd need those spells later. You need something to replace that; for the Channeler, we made the drain be of a new damage type (Mental), something normal cure spells don't heal. Instead, there's a separate line of spells that repair that damage, and they only do it slowly. That forces casters to pace themselves a bit more, relying mostly on their mid-level spells.
All of this means that unless the fights have NO breaks between them, your party will go into each with effectively full resources. Everyone will be healed and buffed, with casters ready to unload their best spells in the first round. Bottom line, it WILL change how your campaigns play. Whether that's a bad thing is up to you and your players. Personally, my group liked it far more.
There are a lot of little pieces of the existing spell system you can replace with skills without imbalancing the system:
> Caster level can be replaced with a skill, allowing you to minimize the loss in effectiveness for multiclassing. (You'd still lose spell slots, of course.)
> Replace the ranged attack rolls for rays, etc. with a skill check (effectively giving casters a full BAB progression when using spells).
> Allow casters to replace their saving throw rolls for certain spells with a skill check (to deflect/disrupt the attacker)
These definitely improve the feel of the existing system, without requiring many extra skill checks. You would need to rebalance the caster class' skill points, though.
But if you're talking about dumping the slot system altogether, we've got an entire class (the Channeler) that uses drain-based spellcasting. That is, instead of spell slots, the caster takes some damage depending on his level and the spell's level (and you then need a Concentration check to succeed in using the spell). There's a pseudo-skill used to make the drain somewhat random, but that's not strictly necessary (and frankly ended up requiring a bit too much rolling).
The main reason I bring this up is because of a common concern you hear: that in skill-based systems, characters become able to use low-level stuff almost at will, and it breaks the system. Well, it's true, sort of. They WILL be able to use the low-end stuff ad nauseum when they get to high level, but that's not always a bad thing, and it only breaks the system if you leave the spell list as-is. An infinite supply of cure light wounds spells sounds abuseable, but at high level most parties I've been in have just made wands of that for the same thing, with only a paltry resource cost. The bigger problem is unlimited stat buffs.
The other side effect is that you, as DM, can't depend on spell depletion to make normally-easy fights challenging by putting them at the end of a day, but I've never liked that method anyway. Too much like railroading.
You'd also need to make sure the system had a method of mitigating "alpha strikes". That is, in 3E a wizard can dump all his 8th and 9th-level spells in a single fight, and the only thing holding him back is the knowledge he'd need those spells later. You need something to replace that; for the Channeler, we made the drain be of a new damage type (Mental), something normal cure spells don't heal. Instead, there's a separate line of spells that repair that damage, and they only do it slowly. That forces casters to pace themselves a bit more, relying mostly on their mid-level spells.
All of this means that unless the fights have NO breaks between them, your party will go into each with effectively full resources. Everyone will be healed and buffed, with casters ready to unload their best spells in the first round. Bottom line, it WILL change how your campaigns play. Whether that's a bad thing is up to you and your players. Personally, my group liked it far more.
Nifft
Penguin Herder
The only example I've personally analyzed in depth was the Truenamer. While that class shows that it's possible to balance the skill mechanic, it also shows that doing so restricts your build options greatly, since you must focus everything on maximizing that one skill check.
Cheers, -- N
Cheers, -- N
GreatLemur
Explorer
Yeah, I've got some pretty serious conceptual problems with any class or magic system that relies on a single ability score to power its central abilities. It's just bad design to leave as optional something which absolutely ever character of that class will want (and, in the case of skills, will want at max ranks, probably with some feats to cheese it up even further).Nifft said:
So, yeah, I'd advise using level checks instead of skill checks. There's a lot less in the way of min-maxing possibilities there, and--when you think about it--doesn't it just make sense that the ability to cast spells would be pegged solidly to caster level? I mean, what are levels in a primary caster class describing, anyway, if not exactly that?
