Trying to come up with fair rules for players to research original spells.
What should the Spellcraft DC formula be for researching new spells?
What should be the consequences of failing this check be?
I am not concerned with other aspects of researching spells at the moment (cost and time), but will be in the future.
Opinions welcome!
Somewhat lengthy considerations follow...
In 3.5, the suggested DC for learning a new spell is 10 + the level of the spell. I think this value is patently absurd. A level one character can very easily be optimizied to be able to research level 9 spells 'right away' (note: only AD&D spell research rules stipulate you actually have to be able to cast the spell you are researching; even if we institute this common sense rule, the low DCs that can be met by level 1 characters is still relevant). In Pathfinder, the guidelines suggests a "number of Spellcraft and Knowledge (arcana) checks." This isn't very helpful.
In AD&D, the check for wizards to research a new spell is the same as for them to learn a new spell. In Pathfinder terms, that means "...he must make a Spellcraft check (DC 15 + spell's level). A wizard who has specialized in a school of spells gains a +2 bonus on the Spellcraft check if the new spell is from his specialty school." That would mean it is as hard to invent an entirely new spell as it is to identify a spell being cast, learning a spell from a spellbook or scroll, or preparing a spell from a borrowed spellbook. This, to me, is an absurd result.
I set the DC as 20 + spell level in the first draft of my research rules. This makes it as difficult as deciphering a scroll. In retrospect, this seems too easy. Surely inventing Magic Missile (DC 21 with this formula) should be harder than deciphering a scroll of Magic Missile (DC 21 as well)?
The other thing we can compare it to is crafting a magical item. Crafting a +5 sword (DC 20) is as difficult as deciphering a scroll of Light without the use of Read Magic. Consider that a level 1 Wizard NPC has a +5 (basic) or +6 (heroic) bonus to spellcraft and a 15 PB PC can easily get +7 and then a possible +1 trait and a +2/3 feat (magical aptitude, skill focus) as well as possible racial bonuses, it doesn't seem that unreasonable when it comes to identifying the scroll (30%, 35%, 40%, 50%, 55% chances respectively); it is, after all, meant to represent a rather hard task (read magic notwithstanding). The DC of the +5 Sword seem pretty low as it seems like wizards capable of crafting such items should have rather inflated craft checks. It seems, instead, that these values are purposefully low to make crafting for those who do not meet the prerequisites or who want to craft items more quickly.
Researching a new spell should at least be as hard as deciphering another's arcane writing, if not harder. What things have a DC 25 check (only skill checks considered)?
Getting a natural animal to approach a creature with an unnatural aura
Picking an average lock
Overcoming a difficult obstacle in a chase
Listening in on a message
Climbing a rock wall
What things have a DC 30 check (only skill checks considered)?
Guessing the command word to an item using Arcana or History
Overcoming a very difficult obstacle in a chase
An iconoclast destroying a minor artifact
Find a well hidden secret door
None of these DC 25/30 checks are scaling; that is, they are flat DCs.
So the question is, what should the DC formula be for researching a new spell?
Failure?
What about the penalty for failure? In AD&D, if you failed the character has to spend another week in study before they re-roll. This process continued until the player made their check.
In 3.5, a failure means that the player must start anew (it is unclear whether they wasted all of their money, but this is probably what is intended). If we make an analogy between researching a spell and crafting a magic item, it seems to make sense that failure under this option results in wasting your time and materials.
Pathfinder has research rules:
Successfully researching a new spell requires time and expensive research. An optional system for researching new spells is outlined below.
The research should cost at least 1,000 gp per spell level (or even more for particularly exotic spells) and require both the Spellcraft skill and a Knowledge skill appropriate to the researcher’s class. Wizards and bards use Knowledge (arcana), sorcerers use a Knowledge skill appropriate to their heritage (usually arcana, nature, or planes), druids and rangers use the Knowledge (nature) skill, and clerics and paladins use Knowledge (religion). The actual research process varies by the type of spell, often involving magical experimentation, the purchase and study of moldy scrolls and grimoires, contact with powerful magical beings or outsiders, and extensive meditation or rituals.
For each week of research, the caster makes separate Knowledge and Spellcraft checks against a DC of 20 plus twice the level of the spell being researched, modified according to Table: Spell Research Modifiers. To successfully research the spell, the caster must succeed at both checks. Failure indicates the week was wasted. Spells of 4th-6th level requires 2 weeks of successful research, while spells of 7th-9th level require 4 weeks. The researcher may employ up to two assistants in the research process to assist on the skill checks using the aid another action.
Pathfinder has research rules:
Successfully researching a new spell requires time and expensive research. An optional system for researching new spells is outlined below.
The research should cost at least 1,000 gp per spell level (or even more for particularly exotic spells) and require both the Spellcraft skill and a Knowledge skill appropriate to the researcher’s class. Wizards and bards use Knowledge (arcana), sorcerers use a Knowledge skill appropriate to their heritage (usually arcana, nature, or planes), druids and rangers use the Knowledge (nature) skill, and clerics and paladins use Knowledge (religion). The actual research process varies by the type of spell, often involving magical experimentation, the purchase and study of moldy scrolls and grimoires, contact with powerful magical beings or outsiders, and extensive meditation or rituals.
For each week of research, the caster makes separate Knowledge and Spellcraft checks against a DC of 20 plus twice the level of the spell being researched, modified according to Table: Spell Research Modifiers. To successfully research the spell, the caster must succeed at both checks. Failure indicates the week was wasted. Spells of 4th-6th level requires 2 weeks of successful research, while spells of 7th-9th level require 4 weeks. The researcher may employ up to two assistants in the research process to assist on the skill checks using the aid another action.
Where is that? The core rulebook (at least the printing I have) says...
Independent Research: A wizard can also research a spell
independently, duplicating an existing spell or creating
an
entirely new one. The cost to research a new spell, and the
time required, are left up to GM discretion, but it should
probably take at least 1 week and cost at least 1,000 gp per
level of the spell to be researched. This should also require
a number of Spellcraft and Knowledge (arcana) checks.
Edit: Google seems to indicate those are just house rules suggested by d20PFSRD.com (although I do like them, somewhat).
Don't know what book it's from, I found it on d20pfsrd.
Don't know what book it's from, I found it on d20pfsrd.
Edit: I was wrong. It's in the gamemastery guide (I find I rarely reference that book). Kind of an unfortunate location, but alright. Thanks!
Huh, you're right, the PRD has your version.
So why is d20pfsrd inserting their own house-rules into the middle of stuff copied from the CRB without marking it as such? Ugh.
Huh, you're right, the PRD has your version.
So why is d20pfsrd inserting their own house-rules into the middle of stuff copied from the CRB without marking it as such? Ugh.
It's in the game mastery guide. I thought that they would mention that or something if they give half-hearted rules in the core book, but there you have it!