Help Develop a Totally New Magic System [Archive] - Wizards Community

Help Develop a Totally New Magic System [Archive] - Wizards Community

Post/Author/DateTime Post
IceFractal
01-15-05, 02:25 PM In the standard rules, magic in D&D is slot based and the effects are preset depending on the slot used. Magic-users generally know a large number of limited-use spells with consistent effects, and all modification, such as metamagic, affects the slot used. Also, spells are usually prepared ahead of time and casters have an exact number of slots per day. What I'm trying to develop here, hopefully with your help as its a big project, is a totally different magic system. The base ideas of the system:
* Magic is skill based and spell power depends on skill - One mage might be able to blast down a wall with the same based spell that another uses to light a candle.
* Mages tap both from a recharging magic aura and a pool of spellpoints. Weaker effects can be used almost unlimitedly, stronger effects exhaust the power quickly.
* Mages know a smaller number of more versatile spells.

Here's some design specifications for this system:
* For version 1.0, let's focus on a single type of caster, just called the Mage, to avoid confusion. Later we can work out the differences between arcane/divine and spontaneous/prepared casters.
* While there are a number of variables in casting a spell (skill check, magic power spent, casting time, components used, backlash damage, etc), we'll limit it to just one variable: skill check, to avoid overwhelming complexity. Amount of magic power spent will be dependant on the skill check, and casting factors will be modifiable by metamagic.
* A spellcasting skill check will be called a spell check ;) and spellcasting skill points will be called magic ranks, to avoid confusion with normal skills, as they might use a different system.
* Instead of having a different magic rank for each spell, or a single universal magic rank, Mages will have 8 magic ranks, one for each school of magic, as a compromise between detail and playability. So when casting a fireball, the relevant magic rank would be Evocation.

Below is the system itself. Feel free to suggest ideas, give feedback, additions, complaints, etc. on what's been posted, or whatever else you want to do. But please, post something, this is too big a change to do without feedback. :surrender
IceFractal
01-15-05, 02:26 PM Three factors go into making a spell check: Skill (which includes caster level), Natural Talent (relevant ability mod), and Luck (a die roll). Its the relative proportions of these which should be decided. Obviously a regular skill check isn't going to work, because then your effective caster level could be up to +/- 10 and a score of 20 in your spellcasting ability would effectively grant you five caster levels. To solve this, I'm going to try the following system:
Magic Ranks = 3 x caster level for an average school, more for a specialized one, less for an ignored one. The die roll would then be +/- 3 effective caster levels instead of 10. For ability scores, if the modifier is just added, then the effect of a high score is minimized, only ~1.5 caster levels for a 20. If the modifier is doubled, that changes to ~3 caster levels difference between a 20 and a 10. As I want the ability score to make a difference, the current calculation will be:

Spell Check = 1d20 + Magic Ranks (relevent school) + Ability Mod x 2
Magic Ranks (average) = 3 x caster level

With that in mind, we can consider Magic Rank aquisition by the Mage. Since we want a Mage to be able to maintain 3 x caster level ranks in all seven schools (universal is special), lets say a Mage gets 21 ranks each level, to be distributed throughout the seven schools. Now since Universal spells are rare, that school is a special case. If you want Universal magic to be the simplest form, a Mage's Universal ranks should be considered equal to their highest Magic Rank. If you want it to be the deepest and most challenging form of magic then their Universal ranks should be considered equal to their lowest Magic Rank.

Next we consider minimums and maximums. If we just give Mages 21 ranks a level, someone could just put them all into Evocation, have effectively a level 70 evocation rank at level 10, and easily blast everything out of their path. Perhaps we should say that the maximum rank in a school is 5, and the minimum rank is 1. However, that would mean that someone could maximize half the schools by minimizing the other half, which is significantly better than the current system where you sacrifice two schools to gain some benefit in one. Maybe raising a skill should count more than lowering it? Say that raising a skill 1 point means lowering others 2 points? Then to get a add 5 in one school you would have to only get 1 in two others (or only 2 in four others), which is more in line with the current specialization system.

So, to recap:
Spell Check = 1d20 + Magic Ranks (relevent school) + Ability Mod x 2
Magic Ranks per level = 3 in each school
Raising a school 1 point = Lowering other school(s) by 2 points
Maximum Rank = 5 x caster level
Minimum Rank = caster level
Universal Rank = highest rank -or- lowest rank, opinions?

More to follow, please comment. :bigeyes:
Gilean
01-15-05, 02:47 PM Make a new game that's not worth it. And then send it to me. Or to rpghost.com. Or somewhere. If you are not too interested, here are my suggestions:
Spell Check = 1d20 + Magic Ranks (relevent school) + Ability Mod x 2
Magic Ranks per level = 3 in each school
Raising a school 1 point = Lowering other school(s) by 2 points
Maximum Rank = 5 x caster level
Minimum Rank = caster level
Universal Rank = highest rank -or- lowest rank
Do not multiply ability modifier. I suggest either as normal skills, or modified level checks: d20+caster level+relevant ability modifier. A good assumption is 18 casting stat for first level, 23 for 20th without magic.
Level check: At first level, reasonable result is 15. A DC of 15 to cast 1st level spells would mean little over 50% chance of success. At every level, check increases by 1 (caster level check), +1/8 (+1 casting stat oce per 4 levels, 50% chance of changing modifier), + some for magic. A DC of 45 would be generally hard to achieve at 20th level (assuming +5 inherent, +6 headband/cloak/periapt)
Mind Rogue
01-15-05, 04:19 PM How about making casting simply skill based. One idea that I saw somewhere involved making it based entirely on spellcraft. In order to cast a spell, a mage must make a spellcraft check, DC 15 + spell level. However, each time the mage attempts to cast a spell (successfully or unsuccessfully) the DC rises by one. Thus, there is the chance for a long time that the mage can cast her spells, but after a while, she becomes burned out. The problem is that it is pretty hard to make at low levels, really easy to make at higher levels, and the slope between low and high level spells is pretty steep.

An alternative might be to say that your effective caster level is your check result - 10 (and possibly minus the spell level). If your check doesn't at least make it to the requisite caster level for the spell, if fails. Since the spellcraft ranks goes up at about the same pace as caster level, this keeps things mostly balanced, and making higher level spells reduce the caster level helps to keep the balance down (from intelligence enhancements and the like). If you take this second approach, you might consider splitting spellcraft into several skills, one for each school of magic. Then the mage chooses which spells she is proficient with by choosing which skills to learn. Of course, skill points per level would probably need to be raised to at least 4+Int in order to offset this penalty.

You could also try to implement something kind of like the 3.5 psionics. Most of the spells could be converted to use a pool of mana, and most of those could probably find a way to be improved by spending more mana. For instance, most damage spells would be 1d6 damage per mp, while charm person and charm monster could be rolled into one spell, with more mana allowing you to select additional targets or those with a wider variety of creature types.

Finally, I think this would be an excellant opportunity to break the mold and separate the differances between arcane and divine. It always struck me as odd that a cleric wearing full-plate (including two gauntlets) can cast spells perfectly, even though he still is required to gesture for his spells, despite being divinely inspired.
IceFractal
01-15-05, 04:53 PM First, thanks for the fast responses! :D

The problem I have with making it a normal skill check is that since the spell's effective power will be based on the check result, a normal skill check would give too much randomness. For instance, a 10th level caster might cast as weakly as a 1st level caster or as strongly as a 20th level caster ... on every spell. With multiplication by three, you only have a +/- 3 caster levels range from the d20. Multiplying the ability by two is so that it counts more relative to the d20 roll, but I'm open to suggestions on the exact ratios. Another option would be to have it be a standard skill in terms of the bonus, but instead of rolling 1d20, roll 1d6. I just picked this system because it makes fractional increases in spell difficulty (like range: medium -> long) easier when you can add, say, 2 points instead of 1/3 of a point.

To clarify, the skill check isn't a DC for success/failure, it determines how powerful the end result is. So the difference between an apple-sized fireball that mildly burns someone and a huge blast that breaks through a steel gate is simply the roll that the caster acheived.

Post/Author/DateTime	Post
IceFractal 01-15-05, 02:25 PM	In the standard rules, magic in D&D is slot based and the effects are preset depending on the slot used. Magic-users generally know a large number of limited-use spells with consistent effects, and all modification, such as metamagic, affects the slot used. Also, spells are usually prepared ahead of time and casters have an exact number of slots per day. What I'm trying to develop here, hopefully with your help as its a big project, is a totally different magic system. The base ideas of the system: * Magic is skill based and spell power depends on skill - One mage might be able to blast down a wall with the same based spell that another uses to light a candle. * Mages tap both from a recharging magic aura and a pool of spellpoints. Weaker effects can be used almost unlimitedly, stronger effects exhaust the power quickly. * Mages know a smaller number of more versatile spells. Here's some design specifications for this system: * For version 1.0, let's focus on a single type of caster, just called the Mage, to avoid confusion. Later we can work out the differences between arcane/divine and spontaneous/prepared casters. * While there are a number of variables in casting a spell (skill check, magic power spent, casting time, components used, backlash damage, etc), we'll limit it to just one variable: skill check, to avoid overwhelming complexity. Amount of magic power spent will be dependant on the skill check, and casting factors will be modifiable by metamagic. * A spellcasting skill check will be called a spell check ;) and spellcasting skill points will be called magic ranks, to avoid confusion with normal skills, as they might use a different system. * Instead of having a different magic rank for each spell, or a single universal magic rank, Mages will have 8 magic ranks, one for each school of magic, as a compromise between detail and playability. So when casting a fireball, the relevant magic rank would be Evocation. Below is the system itself. Feel free to suggest ideas, give feedback, additions, complaints, etc. on what's been posted, or whatever else you want to do. But please, post something, this is too big a change to do without feedback. :surrender
IceFractal 01-15-05, 02:26 PM	Three factors go into making a spell check: Skill (which includes caster level), Natural Talent (relevant ability mod), and Luck (a die roll). Its the relative proportions of these which should be decided. Obviously a regular skill check isn't going to work, because then your effective caster level could be up to +/- 10 and a score of 20 in your spellcasting ability would effectively grant you five caster levels. To solve this, I'm going to try the following system: Magic Ranks = 3 x caster level for an average school, more for a specialized one, less for an ignored one. The die roll would then be +/- 3 effective caster levels instead of 10. For ability scores, if the modifier is just added, then the effect of a high score is minimized, only ~1.5 caster levels for a 20. If the modifier is doubled, that changes to ~3 caster levels difference between a 20 and a 10. As I want the ability score to make a difference, the current calculation will be: Spell Check = 1d20 + Magic Ranks (relevent school) + Ability Mod x 2 Magic Ranks (average) = 3 x caster level With that in mind, we can consider Magic Rank aquisition by the Mage. Since we want a Mage to be able to maintain 3 x caster level ranks in all seven schools (universal is special), lets say a Mage gets 21 ranks each level, to be distributed throughout the seven schools. Now since Universal spells are rare, that school is a special case. If you want Universal magic to be the simplest form, a Mage's Universal ranks should be considered equal to their highest Magic Rank. If you want it to be the deepest and most challenging form of magic then their Universal ranks should be considered equal to their lowest Magic Rank. Next we consider minimums and maximums. If we just give Mages 21 ranks a level, someone could just put them all into Evocation, have effectively a level 70 evocation rank at level 10, and easily blast everything out of their path. Perhaps we should say that the maximum rank in a school is 5, and the minimum rank is 1. However, that would mean that someone could maximize half the schools by minimizing the other half, which is significantly better than the current system where you sacrifice two schools to gain some benefit in one. Maybe raising a skill should count more than lowering it? Say that raising a skill 1 point means lowering others 2 points? Then to get a add 5 in one school you would have to only get 1 in two others (or only 2 in four others), which is more in line with the current specialization system. So, to recap: Spell Check = 1d20 + Magic Ranks (relevent school) + Ability Mod x 2 Magic Ranks per level = 3 in each school Raising a school 1 point = Lowering other school(s) by 2 points Maximum Rank = 5 x caster level Minimum Rank = caster level Universal Rank = highest rank -or- lowest rank, opinions? More to follow, please comment. :bigeyes:
Gilean 01-15-05, 02:47 PM	Make a new game that's not worth it. And then send it to me. Or to rpghost.com. Or somewhere. If you are not too interested, here are my suggestions: Spell Check = 1d20 + Magic Ranks (relevent school) + Ability Mod x 2 Magic Ranks per level = 3 in each school Raising a school 1 point = Lowering other school(s) by 2 points Maximum Rank = 5 x caster level Minimum Rank = caster level Universal Rank = highest rank -or- lowest rank Do not multiply ability modifier. I suggest either as normal skills, or modified level checks: d20+caster level+relevant ability modifier. A good assumption is 18 casting stat for first level, 23 for 20th without magic. Level check: At first level, reasonable result is 15. A DC of 15 to cast 1st level spells would mean little over 50% chance of success. At every level, check increases by 1 (caster level check), +1/8 (+1 casting stat oce per 4 levels, 50% chance of changing modifier), + some for magic. A DC of 45 would be generally hard to achieve at 20th level (assuming +5 inherent, +6 headband/cloak/periapt)
Mind Rogue 01-15-05, 04:19 PM	How about making casting simply skill based. One idea that I saw somewhere involved making it based entirely on spellcraft. In order to cast a spell, a mage must make a spellcraft check, DC 15 + spell level. However, each time the mage attempts to cast a spell (successfully or unsuccessfully) the DC rises by one. Thus, there is the chance for a long time that the mage can cast her spells, but after a while, she becomes burned out. The problem is that it is pretty hard to make at low levels, really easy to make at higher levels, and the slope between low and high level spells is pretty steep. An alternative might be to say that your effective caster level is your check result - 10 (and possibly minus the spell level). If your check doesn't at least make it to the requisite caster level for the spell, if fails. Since the spellcraft ranks goes up at about the same pace as caster level, this keeps things mostly balanced, and making higher level spells reduce the caster level helps to keep the balance down (from intelligence enhancements and the like). If you take this second approach, you might consider splitting spellcraft into several skills, one for each school of magic. Then the mage chooses which spells she is proficient with by choosing which skills to learn. Of course, skill points per level would probably need to be raised to at least 4+Int in order to offset this penalty. You could also try to implement something kind of like the 3.5 psionics. Most of the spells could be converted to use a pool of mana, and most of those could probably find a way to be improved by spending more mana. For instance, most damage spells would be 1d6 damage per mp, while charm person and charm monster could be rolled into one spell, with more mana allowing you to select additional targets or those with a wider variety of creature types. Finally, I think this would be an excellant opportunity to break the mold and separate the differances between arcane and divine. It always struck me as odd that a cleric wearing full-plate (including two gauntlets) can cast spells perfectly, even though he still is required to gesture for his spells, despite being divinely inspired.
IceFractal 01-15-05, 04:53 PM	First, thanks for the fast responses! :D The problem I have with making it a normal skill check is that since the spell's effective power will be based on the check result, a normal skill check would give too much randomness. For instance, a 10th level caster might cast as weakly as a 1st level caster or as strongly as a 20th level caster ... on every spell. With multiplication by three, you only have a +/- 3 caster levels range from the d20. Multiplying the ability by two is so that it counts more relative to the d20 roll, but I'm open to suggestions on the exact ratios. Another option would be to have it be a standard skill in terms of the bonus, but instead of rolling 1d20, roll 1d6. I just picked this system because it makes fractional increases in spell difficulty (like range: medium -> long) easier when you can add, say, 2 points instead of 1/3 of a point. To clarify, the skill check isn't a DC for success/failure, it determines how powerful the end result is. So the difference between an apple-sized fireball that mildly burns someone and a huge blast that breaks through a steel gate is simply the roll that the caster acheived.