Post/Author/DateTime	Post
Eldurin 03-22-05, 03:39 AM	Vancian Magic, i.e. spell slots per level has been very well established, and despite whatever anyone may think of it, it is here to stay for still quite a while. Nevertheless practical ideas as alternatives to it have to be developed. So what do you think magic should be and how do you think it should work? 1. Spell Points The first intuitive alternative is of course spell points, representing a mage's arcane reserves. And even better it is set out neatly in Unearthed Arcana. It has limitations, but it is far more flexible than spellslots, or the mythology behind spell slots. 2. Skillpoints: Knowledge and Hitpoints Someone recently posted in the classes forum a variant of spell caster that worked against skill points in knowledge (i.e. knowledge arcana or knowledge nature or knowledge religion), with a spell casting limited boundary of a kind of sub-dual magic damage that a caster takes against casting spells. Very interesting. 3. Skillpoints: Spellcraft A big part of me would like to do away with spellcasting limits completely. When I think of magic I see the arcane, the strange, the unexplained. I would therefore like something that shifts the emphasis away from caster limits to casting limits. How about a system that bases magic on spellcraft (a fairly redundant skill until epic levels that nonetheless epitomises the spell caster.) Lets say that there are no limits on how often a caster can cast, but on how long, and against a spellcraft difficulty. Infinite casting, but limited against ability and time. Lower level spells would have lower level spellcraft requirements (i.e. Magic Missile DC 3) and higher level would have higher DC requirements) (i.e. Imprisonment DC 20) Above and beyond spellcraft requirements one could make magic more difficult by making casting times longer. (i.e. magic missile 3 rounds) but by taking a spellcraft increase (i.e. DC +4) a caster could reduce the casting time by one round. So I am pretty sure I am not the only one who has kinda been playing with the possibilities. What are your thoughts?
Masterful_Norg 03-22-05, 07:31 AM	I see where you're going with the spellcraft based casting and I like how you want to give that skill some kind of real purpose, but I personally try to stay away from too many rolls of the dice when it comes to player power (and just to save game time) so I think it would be best if casters were limited to taking 10 on spellcraft checks just for simplicity and reliability's sake. I also think that whatever rules one uses for determining spell DCs (if you create your spells) should give enough artistic license so that players can create very unique spells (though I would only allow a spellcaster to create their spells at the begining of each day once again in order to save time).
Gilean 03-22-05, 01:09 PM	Skill checks are a bad idea. Spellcraft ranks a much better. Skills can be munchkinized through too many different ways. Ever heard of that "craft epic spell that boosts spellcraft"-thingy, for example. Or the diplomacy builds.
ArchmageOmega 03-24-05, 04:26 PM	I have also been thinking of variations to magic systems. (Not specifically in relation to D&D, but it still applies.) Basically I arrived at the conclusion that a magic system consists of several independent pieces. First, there is a cost to magic. In the Vancian system, this is spell slots. This can also be mana (possibly of varying kinds, and it might come from the individual, the environment or elsewhere), health, fatigue, time or any number of other things. The Warlock class from The Complete Arcane does violate this. The Warlock is free to use his powers without cost but with hard limits to pwer level. Cost also deals with the power levels possible. A magic system also needs a system for learning new spells. This might be by experimentation, teachers, books, etc. Also, there needs to be a system for constructing individual spells or effects. This can be just a list of spells, spells built from runes, channeled through items, spell seeds (like epic magic), etc. Finally, there is the optional set of rules on how a spell, once made, can be enhanced. In D&D this is the metamagic feats, but there are other possiblities. These pieces aren't strongly interrelated, but a underlying theme would help solidify things. Anyway, that was a lot more than I meant to write. :)
JohnSnow 03-25-05, 03:41 PM	I've given this a LOT of thought. I understand that Vancean magic is one of the "sacred cows" of D&D, but that doesn't change the fact that I've always hated it. One of my favorite ideas ever was "Channeling" - introduced in the 2nd Edition supplement Player's Option: Spells & Magic that Rich Baker wrote. This got a little bit of a second life with the Unearthed Arcana section on spell points with the Vitalizing variant. I always liked the notion of tying spellcasting to fatigue, since it seems like a good metric to keep casters...well...casting. The real balance issues (to me) arise if a caster can either: a) Cast spells AND fulfill the physical role that should be done by a non-casting character. Not that they should be helpless, but you shouldn't be able to cast and fight almost as well as a fighter. b) Cast so many spells that you can overwhelm ANY situation the party faces with your magic. This makes the rest of the party "third wheels" or "meat shields" for the spellcaster. This is bad and un-fun. The INTENTION of the spell slot system is obviously to address these concerns. The reality of how well it's addressed point a) is obvious *cough*cleric*cough*. In practice, it handles point b) not a whole lot better. The normal reality is that the party operates until the casters run out of their most effective spells and then RUNS away until later. Not that the casters are forced to provide "support" until a challenge arises that ONLY they can deal with. I'd say the ideal system would force casters to conserve their highest level powers for when they're really needed. Spellcasters shouldn't be nuking the raiding party with a fireball unless it's a monstrous horde and the fighters are getting tired. Setting things up that way gives the fighters a chance to "shine." Since spellcasters aren't usually in melee, they ought to have a resource that they use up the same way that fighters use up hit points. One obvious idea is for their spells to cause some form of fatigue. That way, just like the melee combatants who wade into combat, the spellcasters are expending a resource when they cast their spells. I've seen a few variants. Vitalizing is good, but another one I like is from Fantasy Flight's Midnight, which I've mentioned before. The basics: 1. Midnight does away with ALL the PC classes that cast spells. No wizards, clerics, druids, bards, sorcerers, paladins or rangers. It also dumps the monk for flavor reasons. 2. Spellcasting becomes feat-based, so that any character can learn spells. Spellcasting now requires a feat (Magecraft) to be able to cast spells, and a feat (Spellcasting {school}) for each school of magic. 3. Spellcasting comes in 3 distinct "flavors," based on what ability score "powers" your casting. Spiritual channelers use wisdom, hermetic channelers are INT-based, and I think you can figure out what ability Charismatic channelers use. 4. Spells take spell energy to cast equal to their level, so a 1st level spell costs 1 spell energy. Casters get spell energy equal to their spellcasting ability score modifier. When a character runs out of spell energy, they can cast spells by causing spell damage. Spell damage is temporary CON damage equal to the spell energy cost. It goes away with a good night's rest. Rituals and talismans exist which lower the spell energy cost. Some specialized classes also get extra spell energy. 5. Casting level = character level. The spell energy system automatically makes dedicated spellcasters are better spellcasters. Characters can learn spells of a level equal to one-half their character level (rounded down). So you have to be 2nd level to cast 1st level spells, 4th for 2nd, and so on. 5. Learning new spells, beyond the few you learn from the spellcasting feats (or from certain classes - see below), takes XP. 6. A new class is created (Channeler) that is particularly adept at spellcasting. The Channeler is based loosely on the core Bard, with Medium BAB, 6 skill points per level, bonus magic feats, a granted ability (which depends on the "flavor" of channeler) and improved "facility" with magic. Specifically, Channelers get free access to a few schools of magic (as they level up), spell energy equal to their class level, a couple of free "spells known" every level, and reduced XP costs to learn spells. Characters with more levels in Channeler than any other class can add "1" to their character level before dividing by 2 to determine the level of spell they can cast. Or, put another way, Channeler-types round UP instead of down. So a Channeler (or primarily Channeler) can cast 2nd-level spells at level 3, rather than waiting until level 4. So, essentially, the spellcasters get a FEW spells "for free." That is, without tapping their CON scores. And they get more as they go up in level. Now, I admit this system might be a little "harsh" or "low-magic" but it's pretty customizable. You just give the casters more spell energy. This is a pretty fundamental change to D&D, and I admit that it does away with the arcane/divine distinction as well as the Vancean system, but if you'll read my sig, I think you'll see exactly how much that upsets me. By the way, the Midnight designers considered having those "extra" spells do subdual damage, but abandoned it because of the "restore full potency by casting spells on myself" problem. Of course, if you adopt the sidebar variant from Unearthed Arcana that all healing just converts lethal damage to subdual, that's not a problem. And you can just change restoration (and its variants) so that casters can't cast it on themselves. That creates a system somewhat similar to what was done in the late, lamented Wheel of Time game. Using ranks in spellcraft instead of character level to determine which spells you could know is one variant you could add to this. I'm not sure what formula I'd use though (Ranks/2 - 1?). Another would be to allow the character to make a concentration check (or some kind of saving throw) to cast a spell (once your energy is gone) without it causing spell/subdual damage (whichever system you were using). Just some ideas I've been kicking around. Interestingly, the feat for school access (per Midnight) means that low level characters tend to have a far narrower "range" than their higher-level counterparts. As they go up in level, their "bag of tricks" increases. Again, as it should be, IMO.
IceFractal 03-26-05, 03:46 PM	I'll post more on this later, but here's a quick idea. I'm working on a skill based system and found it necessary to change the d20 roll to 3d6. The reason is average deviation. Skill ranks increase linearly with level, except at 1st level, so generally, in terms of DCs for spells, +1 skill point = +1 caster level. With a d20, the average deviation is 5, so using a standard d20+skill roll will mean that people are, on average, 5 caster levels higher or lower than normal. This is a lot of variation, really too much unless you want highly chaotic magic. This means an archmage couldn't even cast 8th and 9th level spells consistently and that a novice could sometimes cast fireballs. I wrote a little program to determine average deviation for various dice combinations. Here's the results: Type Average Std-Dev ------------------------------- 1d20 10.5 5 2d10 11 3.3 3d6 10.5 2.42 4d4 10 1.78 5d3 10 1.44 7d2 10.5 1.09 So, as you see, changing the dice type to 3d6 cuts the randomness in half. Using other combinations of smaller dice would cut the randomness down even more, but I don't think they'd work as well, because: 1) 3d6 is faster to add up than 5d3 or 7d2 2) 3d6 is a well known combination, 7d2 is rather odd 3) We do want some amount of randomness, just not as much as a d20 gives
Lina_Inverse 03-26-05, 11:04 PM	a general thing on 3d6 variation,is it changes the value of each point.+1 over the DC-10 is worth alot more than +4.but its intended to be equally valueable.anywho. my variation on vancian magic is simple,casters have spells memorized and spells per day.you can cast any memorized spell freely if you have enough spells castable left. i like spell slots,i dont like spell preping.slots make it so just because youve learned meteor swarm doesnt mean youll never use fireball again.you need to rely on your lesser spells too.
IceFractal 03-27-05, 12:07 PM	Hmm, good point about the bonus worth fluctuation. Perhaps the best solution would be to use a smaller die: Dice Avg Std-Dev --------------------- 1d10 5.5 2.5 1d6 3.5 1.5 1d4 2.5 1 1d10 probably works, or 1d6 if you want less randomness in effective CL. Obviously, the DCs would have to be adjusted, but since we're making them up anyway, it shouldn't be a problem. Now the next item: Stat bonuses to the roll. Obviously, when adding to the roll, each class would use its casting stat. The problem, however, is that doing so at a 1:1 ratio gives casters a significant level advantage. Someone with 12 Int has +1 CL, someone with 20 has +5 CLs. The spell DCs could be adjusted to the minimum score necessary (so for 8th level spells a casting stat of 18 would be assumed, making the DC 4 points higher). The only problem is that this creates a huge gap between multiclass semi-casters, fairly normal casters, and super powered casters with incredibly high Int. If we keep the DCs at the lowest level, super optimized casters will be able to cast 9th level spells quite early, and pretty much never fail a roll. If we bump them up to compensate for super optimized characters, multiclass casters, or for that matter even normal casters, will be totally incompetent, barely being able to cast cantrips. Therefore, I propose that the stat bonus to the roll be damped somewhat, by dividing it by 2 (rounded down). So having a 20 Int would add +2 to your roll. Also, having gotten at least the minimum necessary stat (under the old system) would be assumed, so 9th level spells would assume an Int of 20 (+2), and have the DC set 2 points higher.