Spell composition

[Indigo Knight]

I disagree with your proposition that spells are not balanced against each other.

Are you smoking something?

Bigby's Hand Vs. Mordenkainen's Sword
Chromatic Orb Vs. Burning Hands
Witch Bolt Vs. any cantrip
Blade Ward??
Conjure Volley Vs. Fireball
Friends and Glibness - 8 levels??
Unluck Vs. Daze
Rary's Telepathic Bond Vs. Telepathy
Weird Vs. Phantasmal Killer

That's only from what I can remember. If I spend more time on it then I could come up with a list of useless spells and bestest spells. Most optimization guides dedicate a chapter on it, and for a good reason. Since there are numbers in the spells, then they can be compared, and when compared you witness this weird disparity. True that there are situations where the superiority of some spells increase. But again, those are situational.


Most spells (that I've seen) would aid a specific task rather then any roll. 'Any roll' would seem to be a very powerful unit. Abilities and saves usually gets +2 as the smallest bonus. Attacks and skills get +1 as bonus. And damage dealing - an extra die. So create a spell such as Owl's Wisdom would give +2 to wisdom. If you want more then you need to stack the Unit.


The way I drafted the system, the first 3 picks are able to convey only one answer each. Picking both Melee and Sight contradicts itself. The effect stage, however, needs a way to portray; sometime a mixture of effects, a varying degree of power and combination of both. Unless I find a way to compartmentalize them to smaller stages.
In the meanwhile I wrote down (to myself) that I can combine up to number-of-caster-levels of units, whether duplicate or not, and must have casting-ability-minus-10 of spell level or higher to cast it.
Besides the number of units to combine, your assessment is correct.



I'd vote for minimizing Harm unit instead of buffing Detect. I don't see that detect needs a buff and it would be messing with something that works (I think).


It's an interesting idea. Having the sense grant... well... sense to the caster instead of sensing an area. It does mean, as you noted, changing to Focus (or Short).


In a grander look of things: I think that all of the math is skewed. Detect Evil is the only spell that came out normal. The other example combinations get way too pricey for a non-epic Wizard. I can't tell if it needs a subtraction or division or something else.

[aimlessPolymath]

Oh geez.

Is this for fifth edition?

I've been assuming 3.5e/Pathfinder this whole time, which throws all my numbers wildly off.

I'm going to need a little while to reevaluate my numbers (and a new source for monster statistics), but I think I can still contribute.

[Indigo Knight]

Well...
To be completely honest - I love 3.5 more then 5th and I only talk about 5th because I'm kinda force to play it with my players.
I'll be more then happy to discuss these in the 3.5 surrounding. Just know that I have zero experience with paizo's material.


Also, this little experiment started somewhere 6 years ago. Close to when 4th playtest material was handed out.

[aimlessPolymath]

Alright, we're back on familiar territory then- the mention of 5e spells threw me for a loop.

Right then.

Bigby's Hand Vs. Mordenkainen's Sword
Chromatic Orb Vs. Burning Hands
Witch Bolt Vs. any cantrip
Blade Ward??
Conjure Volley Vs. Fireball
Friends and Glibness - 8 levels??
Unluck Vs. Daze
Rary's Telepathic Bond Vs. Telepathy
Weird Vs. Phantasmal Killer

That's only from what I can remember. If I spend more time on it then I could come up with a list of useless spells and bestest spells. Most optimization guides dedicate a chapter on it, and for a good reason. Since there are numbers in the spells, then they can be compared, and when compared you witness this weird disparity. True that there are situations where the superiority of some spells increase. But again, those are situational.

I am extremely confused by how you seem to be comparing spells of different levels and pointing out how they are different? Obviously, spells have internal differences, given how each of them are different. But there are absolutely general trends of similarity and points of comparison.
I mean, you compared Friends and Glibness. Are you surprised that a spell which ends after 1 minute automatically, affecting only one creature, at which point the target realizes what happens, is lower level than Glibness, which lasts for an hour, affects all checks you make, and does so more powerfully? They have similar effects, but the manner in which they're applied is completely different.

In 3.5, the trend is that spells deal 1d6/level, with higher level spells affecting more creatures. Mass versions of spells are 4 levels higher, when they exist. Summon Monster and Summon Nature's Ally summon roughly the same CR of creatures. Charm X spells are 4-5 levels lower than Dominate X spells.

Most spells (that I've seen) would aid a specific task rather then any roll. 'Any roll' would seem to be a very powerful unit. Abilities and saves usually gets +2 as the smallest bonus. Attacks and skills get +1 as bonus. And damage dealing - an extra die. So create a spell such as Owl's Wisdom would give +2 to wisdom. If you want more then you need to stack the Unit.

Yes, my hypothetical Assist seed was meant to be an example, and obviously the exact implementation would be different.


I think the system goes awry when more than one multiplier is applied, as they scale, well, multiplicatively. For example, Invisibility, the not-mass version, is here a 6-unit level spell. In fact, literally any spell which affects another creature in any way is at least a 2-unit spell, and so unavailable at 1st level. There's a problem!

Spoiler: System which turned out not to work.

Show

So.

Proposal:
A reworked system of costs, flattening the differences between different classes of targets:

Targeting is now in the following categories and costs:
Object(Void): Cost 2
Object(Surface): Cost 2
Creature: Cost 1.
Creature, Mass: Cost 3 (affects selected creatures within an area)

The distinction between Caster and Creature is equally served by the difference between Melee and other ranges- any spell that you can cast on yourself, you can cast on other people now. Spells like Righteous Might which were "balanced" by being self only were really good anyway (CoDzilla!), so rebalancing them against comparable buffs can happen now.

Furthermore, the difference between willing and unwilling creatures was somewhat pointless. Under most circumstances, a given seed would only be cast on one of the two, and people can choose to fail their save willingly anyway. Increasing the cost of harmful spells is unnecessarily punishing.

Instantaneous/1 round (as appropriate): Cost 1.
Focus and Short spells are combined: You can concentrate on a spell for up to 1 minute/level. When you stop concentrating, it persists for 1 round/level afterwards. Cost 2.
Medium length spell: Lasts 1 minute/level. Cost 3.
Permanent spells are inherently abusable and require some form of material costs, tbd.

Ranges are as you described, but I would make the costs 1, 1.5, 2, and 3. In order to make 1.5 fit, multiply the final costs of all effects by 2.

Under these estimations:
Detect Evil (my version of Detect):
Creature (1) * Melee (1) * Focus(2) * 1 seed = 2 cost, almost appropriate.
Mass Invisibility:
Creature, Mass(3) * Melee (1) * Medium(3) * 1 seed = 9 cost.
Fireball:
Object(Void) (2) * Sight (1.5) * Instant(1) * 1 seed = 3 cost.
Misty Step:
Creature (1) * Melee(1) * Instant(1) * 1 seed = 1 cost. This feels a little too cheap- I might have Warp teleport to a square in range rather than a constant length- so this example spell teleports the caster up to his reach- 5 ft for Medium creatures.

Based on these costs


After looking this over, I feel like the multiplier reductions didn't go far enough, and would further reduce the cost multipliers of Focus and Medium-length spells, to 1.5 and 2.
(or, equivalently: Instant spells are a *2 multiplier, Focus is a *3 multiplier, and Medium is a *4 multiplier)

Then Detect Evil is 1.5 units, Mass Invis is 6 units, and Fireball is 3 units.
Let the limit of effective units available be 1.25 + L/4, rather than L.
Then Detect Evil is available at level 1.
Fireball is available at level 5
Mass Invisibility is available at level 13.

However, this still breaks down under examination.
Invisibility:
Creature (1) * Melee(1)* Medium(2) *1 seed = 2 units- available at level 1.

I cannot create a system in which 1 unit of Conceal is enough to provide invisibility, without making it an effect available at level 1. Otherwise, it's impossible to cast any kind of buffs at level 1.

I really think that for this to work, the actual seeds need to partially scale with level.

[Indigo Knight]

It's not that I'm surprised about them being different. It's that I'm surprised about the difference being so out of line. See, Volley and Fireball do similar damage for having 2 levels difference. For example. Glibness may last an hour, but most situations only call for a couple of rounds. Go to any page about Wizard Optimization; LogicNinja, Dictuum, Treantmonk. And you'll see a breakdown of what spells are a must have and what spells are a trap for new players (their words, not mine). Sometimes, the parameters of the campaign, or this current adventure affect such choices (say, breath water in underwater quest). But more often - the rules themselves makes it so that certain spells are numerically better then others. Just looking at the spell of Polymorph suggests that there was no attention for measuring and comparing to other spells. That's why we have pun-pun and other cheese.




Knobs that we can twist:

• Change the scale of 1-2-3 (like self, natural, hostile) to 0.5-1-2
• Have 'units' that reduce cost, either after multiplication or before. (Extra round casting for -1 cost)
• Swap multiplication (every mult' or only some) for addition.
• Have the end result divided by static number or a dynamic, like caster level. (I'm highly against this. This is not a math lesson).

I second your idea about changing target costing. It has basis. Except the the disappearance of Self. There are spells that were specifically design for the appliance on the Wizard and the Wizard only.
The idea that Short encompass Focus was something I had reservations about.


Let's do the following. Let's pick 4 spells. One with the tiniest buff. Its mass version. A moderate buff (maybe multiple effect?) and it's mass version. I'd suggest not picking any invisibility or illusion ones for now. Those are complicated.


Maybe need to swap the static number for parameters and chart down some functions and graphs?

[aimlessPolymath]

Mmm.
I'm not convinced that we want spells that are balanced by being Self only, but okay.

On Focus/Short - I do think that both modes need to be very cheaply available, though it's fine if they're different. Perhaps each is 1.5 after normalization?
Then Medium-duration Invisibility is a 2nd level effect, but Short duration is a 1st level effect- which actually exists in Pathfinder!



My selected spells are Vanish (Pathfinder- lasts 1 round/level, up to 5), Invisibility, Mass Invisibility, and Greater Invisibility. Maybe
A spell: Level 2
A shorter duration lower level version: Level 1
A mass version: Level 7
A more powerful version: Level 4.


My takeaways: Mass spells are either 4-5 levels higher (which is supported by the Chain Spell metamagic, as well as mass vs normal cure spells), or three times as expensive as single target effects. This particular one fits my example costs.
An extremely short duration costs about half as much (or -1 levels) as a medium duration. This seems fair- note that it is Short, similar to Detect Evil's Focus duration.
Greater Invisibility costs twice as much- 2 levels higher.
Multiplicativity or additivity is not well-defined yet.

Edit: Whoops, didn't see the part about avoiding illusions/invisibility, but it worked out anyway.


A large number of minor comparisons:

Spoiler

Show

Heroism is level 3, but Greater Heroism is level 6; but:
Greater Heroism provides slightly(temp hit points + fear immunity) over double the benefits of regular Heroism, but lasts a tenth the duration.
This is moderate evidence towards linearly scaled bonuses.

Mage Armor is 1st level, and provides a +4 bonus.
Greater Mage Armor is 3rd level, and provides a +6 bonus.
Mass Mage Armor is also 3rd level- this is evidence for "Mass costs 3x as much"?

Enlarge Person is 1st level, but Mass Enlarge Person is 4th level.
The Expansion psionic power is effectively 1st level, but augmenting it to double its effects raises its level to effectively a 4th level power. Perhaps the extra reach makes up the difference? Alternatively, Enlarge Person is just a really good spell.

Magic items provide statistical bonuses with costs that scale with the square of the effect, but character gold scales (approximately) by the 2.9th power of character level. Call it a power of 3.

True Strike provides a +20 bonus to attack for one attack roll; Bungle does the reverse (and applies to skill checks as an alternate mode).

I was wrong- the mass versions of the various Animal's Attributes actually are 6th level, not 5th- they're 5 levels higher than their counterparts.

Mass Fly has a +4 level markup for being a Mass spell. Ditto Charm Monster, Charm Person.

The Cure/Inflict chain have a +4 Mass markup for clerics, but +5 for druids. Collectively, they indicate that additive cost increases are most accurate.

Not a buff spell, but Shocking Grasp deals 1d6/level damage to one creature; Fireball deals 1d6/level to many creatures at range.

Magic Weapon is 1st level, and provides a +1 static bonus. Magic Weapon, Greater provides a scaling bonus, but is level 3. IMO, this provides evidence that a single buff seed (of whatever kind) provides a static bonus, but a second makes the resulting spell scale with level.

Based on a cycle of 3rd party cantrips in Pathfinder, I construct the following equalities:
+2 to the following once, is equivalent (self only):
-AC against a melee attack
-AC against a ranged attack
-One skill check (within a minute)
-One melee attack roll
-One ranged attack roll
Also, +1 to saves for one minute.

Obviously, cantrips =/= regular spell slots, though, since they're at-will in PF.

Greater Resistance is a 4th level spell that grants a +3 bonus on saves for 24 hours. Superior Resistance is 6th level, and provides a +6 bonus. The two are almost linear- Greater Resistance is a level higher than expected.

Takeaways:
Bonuses from spells are almost, but not quite, linear. This surprised me quite a bit! Perhaps the value of a bonus scales similarly to the value of a spell level, even if they're nonlinear?
This is arguably supported by magic item pricing: The price of an nth level spell continuously is 2k * (level) * (2*level -1), which scales quadratically, but a static bonus has a price of n * (bonus)^2 (n is a coefficient dependent on the bonus type), which also scales quadratically.
So there's some link involved:
n * (bonus)^2 = 2k * (L) * (2L-1) = 2k * (2L^2 - L)
bonus = rt(2k / n * (2L^2 - L))
= rt(4k/n) * rt(L^2 - L/2)
~= rt(4k/n) * L

Todo: Compare the various polymorph variants in Pathfinder to measure ability score bonuses.

[Indigo Knight]

When you concentrate on a spell for a turns, stop, and then the spell lingers for b turns you can replicate 4 situations:

1. a = 0; lingering spell.
2. b = 0; concentration spell.
3. a = 0 && b = 0; no spell (d'uh).
4. a > 0 && b > 0; complex spell of lingering and concentrating.

When I went over the spell lists I had a very hard time with couple of spell-types: illusions (mainly shadow spells), summoning, resurrecting, polymorphing and scrying. Treating them requires a bit more then translation. There's a need for a clear definition of upper limit to what each spell can do, what power does levels represent, etc'

It's true, you are right about using Metamagic Feats to see examples for augmenting spells and what each modification worth in levels. Still, having a 2 level increase is different when looking at 1st to 3rd hop vs. 6th to 8th hop.
And I haven't found an exact mimicking to Mass'ing spells via feats.





Basing a function on two points of data is too little. I cannot establish a valid relation that way.
Taking Heroism for example, one spell is an hour long and another is half that time. But most DM's I know won't track that long. The easy play principle (derived from K.I.S) means that only the distinction of number-time-use, concentration, couple-rounds or all-encounter is accounted for.



Let's pick a specific venue to focus on. Just randomly choosing Armor.

Mage Armor - 4 bonus to AC with no penalties.
there's Ironbeard that gives 1 bonus along with other armor worn. Can be used as weapon.
Aspect of Bear is 2nd level spell and it grants 2 natural armor and other bonuses.
Barkskin as 2nd level gives natural armor bonus of (caster level/3)+1
Magic Vestment increase armor bonus by (caster level/4)
Protection from ##### gives +2 to AC against specific something at 1st level and other bonuses.
There's also Glorious Raiment and Halo of Sand.

I'm using the following search options to find those spells:
http://www.d20pfsrd.com/magic/tools/...d-spell-search


Now, if in addition, I took a look at the spheres of power thing:
http://spheresofpower.wikidot.com/protection#toc3


It says -
You grant the target a +1 Deflection bonus to AC, +1 per 5 caster levels.
or -
granting them either a +3 armor bonus or a +1 shield bonus to AC (your choice). This does not stack with other armor or shield bonuses, but does apply against attacks made by incorporeal creatures. These bonuses increase by +1 for every 5 caster levels you possess.

So it is also a clue.
Now we need to see what function can it create. How does AC improves for rogue or warrior? Only through magical equipment?

BTW, the metamagic feats give a nice idea for the energy types.
Like this one
http://www.d20pfsrd.com/feats/metama...pell-metamagic

[aimlessPolymath]

Search up Chain Spell for conversions to mass effects. The feat is generally considered to be somewhat strong, but it's a comparison point.

Permanent (or equivalently, encounter-long) AC improvement, as far as I know, is primarily through the following:
-Equipment improvements:
--Upgrade to new kinds of equipment (janky rate depending on armor proficiencies, constant after that point)
--Upgrade the enchantment bonus of your equipment (To +1 per 3 levels per source of AC, roughly).
-Feats and abilities. No constant rate.
-Statistical improvements. Nearly insignificant outside magic items.

Of these, the +1/3 levels rate is the most significant effect, and the most comparable to scaling spells.

However, we actually have a very useful resource for this, which I forgot we had: The tables of NPC statistics, starting on page 110 of the Dungeon Master's Guide. While not fully equal to PCs, I think this provides a useful proxy statistic.

Unfortunately, I can't find these tables in spreadsheet form anywhere, so performing data analysis on these values isn't something I'm likely to do soon. Eyeballing, it looks like fighters, paladins, and barbarians go from about 17 AC to 30 or 34 (depending on whether they bought rings of protection). Monks also cap out at 34 AC, but start much lower, around 13 AC.

Rogues and rangers go from about AC 15 to 25ish from level 1 to 20.

What does this mean for design? It seems to me that we more or less two or three categories:
-Pretty Good (Three sources of "magic bonuses", i.e. AC is about 16 + level)
-Not That Great (Two sources, armor isn't great, AC is about 14 + level/2)
-Wizard (so low as to be irrelevant; resources invested elsewhere)
There's a fourth category, which I'd call Too High, where AC becomes irrelevant.

There are probably equivalent categories within other statistics, as well.

I think a reasonable goal should be that our system should (in the case of armor) be able to buff people up to the next category within the first three, but not more than one without significant effort. Preventing it from reaching Too High is a matter of controlling bonus stacking.

Spoiler: Monster statistics- not very helpful

Show

Analysis on creature AC values (also not quite the same as PC AC values, but related) shows that the standard deviation of monster AC slowly doubles with level, at a rate of around 1 per 10 levels from a baseline of about 2.2. So, it roughly doubles.
I'm not really sure how to say what that means for design, though.

(Their mean increases by about 1.2 per level, roughly alongside Good BAB; this is significantly faster than )

Attack rolls have their data polluted by a significant amount of misread data, based on the existence of "+1 swords" and such, making the Google Sheets interpreter expect a numeral and throwing an error. As a result, its standard deviation is not meaningfully useful in my opinion. (Its mean has a slope of about 1.3). I found the original spreadsheet, and it got around this somehow, but pulling the data from it myself is a bit irritating.

[Indigo Knight]

Chain Spell takes spells with number of targets (or a target area) and adds secondary targets, with better chances to save against the spell, up to caster level. All this at the cost of 4 more levels to the spell level (meaning additive 4 cost).
It's not perfect, but it's a close approximation. Seems to me that this is the closest we'll get without playtesting.


I've looked to see if there is a mention for NPC statistics online, but couldn't find any as well. If I'll ever get my hands on another DMG then I might be able to upload the table.
We could use the tables from before:
https://docs.google.com/spreadsheets...8F8/edit#gid=3 (It's monster's statistics, but that's what we have).
https://docs.google.com/spreadsheets...#gid=724085674 (the second sheet displays AC. but I can't tell what else is involved in the calculation).
I've seen this, but I don't know if we can base on it
https://rpg.stackexchange.com/questi...sses-per-level


Regardless, I agree that the value goes from one time boost based on proficiency feat at the start and afterwords it scale up by gear wealth that buys enchantments.


So we need to imitate 2 separate cases:

• characters with good armor (fighters, paladins, even clerics sometimes. char's who rely on heavy armor and high defense).
• characters with lower ratings of armor (almost all others, whether it's a limitation imposed on them by the rules, like wizards, imposed by roleplay, druids, monks, or imposed by role in combat, like rogues).

That reminded me of the Defense bonus varient. You think it fits?

Anyway, we know that the way AC is improved comes from several sources:

• Natural
• Equipment + Enchantment
• Dodge, Deflection, Dexterity

I'm going to relegate Natural armor bonus to polymorph, so that's that.
The equipment and dodging oriented bonuses kinda compete with each other.
So we're left with only one slope.
As you already written - y = ax + b where a = 1/3
So the spell level divided by 3 will give an armor bonus. Question remains is what the starting point, b?

If we want to replicate Mage Armor then b have to be 4. But that seems too high for me.
Maybe if we change the slope to be 1/4 and then the initial +4 is justified.
Also;

• Does the spell stack with worn armor?
• Does the basic spell (single application of the unit) is limited? such as effects wear off after receiving hit or few rounds?
• Should the average wizard, casting a spell comprised only of this unit (armor + armor + armor... up to max') will be able to duplicate a fighter's AC? Or reaching this power only available to specializers who "pay" some sort of cost (banned school, flaw feat)?

[Indigo Knight]

These is just a sidenote, a side question, a thought I just had, And I'm sure we'll get back to this later.

Assuming I'm casting heal + heal + heal + heal
Is it better to:
Have the spell grant hp healing (determined by 1d6 per level of spell level) and additional heal units allowing for more complicated healing (regenerating limbs, removing afflictions, negating attribute drain).
OR
Having the spell heal 4d6 (as 4 units of heal) and requiring the spell to cost more if I want to heal more complicated ailments?

[aimlessPolymath]

I'm writing the second spreadsheet, so yes, I'm aware of them as a source. I'm copy+pasting from the first to get my data, then performing analysis.

One thing to keep in mind is that each source of AC has that +1/3 level progression. There are typically two to three:
-Ring of protection/amulet of natural armor
-Armor or Bracers of Armor
-A shield. Rogues and rangers don't have this.'

Also note: equipment and dodge bonuses don't always compete. The rogue I saw has Bracers of Armor in addition to his base statistics. Deflection bonuses stack with practically everything.

First of all, an observation about Mage Armor is that it's self only! It's almost exclusively intended to bring the wizard up from wizard-tier AC to rogue levels. Commensurately, it's also very long-lasting, on the order of multiple encounters.
Magic Vestment stacks with everything. In this case, the spell only applies the mx part, and the b is already there.
Shield of Faith stacks with worn armor, and provides a +2 bonus to AC, +1 per 6 levels (in Pathfinder, at least). It can be cast on other people, too.

I feel like Shield of Faith makes the most sense as a comparison point for AC bonuses that stack with armor.
Magic Vestment feels very low-impact at 1st level.

On structuring bonuses:
I'm not sure. I keep writing and deleting this section.

My general feeling, irrespective of actual comparison points, is that:
-A persistent, i.e. with a duration, bonus to any given major combat statistic, in a way that it is fairly easily stackable, should go from about +2 to +6. This is a general feeling based off the sort of bonuses that bards can grant, and the scaling on effects like Greater Magic Weapon.
-A single-use bonus, one which is expended, should go from about +3 to +10 or +11. I feel that this large a bonus is fairly justified- if you spend two actions(one for the buff, one for the buffed action) to do one thing, that one thing should be pretty consistent.

I would actually place a +4 armor bonus as a base as pretty okay, if it's explicitly an armor bonus! Since armor bonuses don't stack, this is really just a way to take people from Wizard-Tier to Rogue-Tier AC levels. There are some unwanted interactions with monk-type characters and such, though...
...
Idea: Armor effects can emulate any kind of armor, including light, medium, and heavy armor- including maximum Dexterity bonus and armor check penalty, but with reduced spell failure rate. Then, it applies an enhancement bonus to the resulting projected armor, based on the amount of extra "oomph" added (i.e. extra spell seeds or caster level- refer to my general feelings for scaling). If used on existing armor, it just applies the enhancement bonus.

Stacking with existing armor: No. That would take it into Way Too High territory- 20+ armor at level 1 if cast on another.
Limit the basic spell: See my general feelings section.
Emulate fighter AC: Not quite. There are additional sources of AC via extra seeds, which could be stacked. In effect, each magic item is a persistent "spell" buffing the fighter. So, you'd need to stack a variety of spells to reach full power- which presumably has costs to do so. Plus, armor proficiency is still necessary to reach full parity, due to the way I structured the Armor effect. Rogue-tier is a bit easier, since they don't rely on heavy armor as much.


On Healing: I'm not sure I understand what you're asking. Could you clarify the options you're discussing?
I'm not sure what your second choice entails as far as "costs more"- I thought the cost of a spell was determined by the number of units & targeting costs?

[Indigo Knight]

Sorry.
I've got so many open tabs and saved bookmarks for this...


Anyway,
The units system as is - doesn't give room for a spell that's only Self. There's no penalty or forbidding against pumping it to natural or hostile target. And if we do add this - how will it manage so? A long list of spells or combinations that are a valid self-spells? A long list of spells that are banned?

One thing that comes to mind is maybe changing the numbers so Self only is lower priced. Like

Target:	Self	Natural	Hostile
Cost:	-4	1	2

But then it needs to be an addition to the cost. Otherwise we get a spell that cost minus point.


+1/6 levels seems like too low for higher level adventures. My gut tells me to use +1/4 levels at the very least.
Oh! And it seems I don't understand the rules for stacking bonus for armor in 3.5 - I keep mixing them up.



I want to keep everything quite simple and understandable. The monk scenario you mention is especially unwanted. So instead of adding a lot of clauses and what ifs, I would go with something very straightforward like this:

Armor:
Grant the spells' target an AC bonus of +1 per four levels of the spell.
If the target has no armor, it is imbued with additional bonus of 2 AC.
Cost: 1

-A persistent, i.e. with a duration, bonus to any given major combat statistic, in a way that it is fairly easily stackable, should go from about +2 to +6. This is a general feeling based off the sort of bonuses that bards can grant, and the scaling on effects like Greater Magic Weapon.
-A single-use bonus, one which is expended, should go from about +3 to +10 or +11. I feel that this large a bonus is fairly justified- if you spend two actions(one for the buff, one for the buffed action) to do one thing, that one thing should be pretty consistent.

I feel like I need someway of
A - grants +2 bonus for a long long time.
B - grants a massive bonus for a really short time.
and that Cost(A) == Cost(B)

On Healing: I'm not sure I understand what you're asking. Could you clarify the options you're discussing?
I'm not sure what your second choice entails as far as "costs more"- I thought the cost of a spell was determined by the number of units & targeting costs?

Forget the Heal. Take the Armor from above.
The idea is that I can combine more then one effect. And nothing is saying I can't combine the same item. Unless we structure the system in that way. For now, we haven't.
So, what If I make an effect of Armor + Armor + Armor?
Would it:
1. grant +2 for each armor unit
2. grant bonus that is scaling with caster level or spell overall level, and adding same unit grant a special perk - like, in the case of armoring, protecting from criticals, defend from bull rushing, and on.

[aimlessPolymath]

Hm.

My comparison to shield of faith isn't 0+1/4 level, it's 2+1/6 level, scaling up to an eventual +5 (the same value); I feel that there should be a minimum level of noticeable effect. Does that change your opinion at all?
Furthermore, this is an AC bonus that stacks with existing armor; it's an enhancement bonus to an armor bonus, analogous to enchanted equipment.

In my write-up, there's an additional effect, which lets you "create armor", which explicitly does not stack with other worn armor.
How does this avoid the "monk problem"? Easy- the created armor penalizes monks, just as though they were actually wearing armor.

The following kinds of armor can be created:
-Leather armor (which becomes +2 armor, i.e. a +4 armor bonus)
-Breastplate
-I don't know, half plate? Banded mail?

All armor modes have their effective armor category for the purpose of spell failure chance negation reduced by one stage (see armored mage abilities like the Bard or Hexblade); in the case of leather armor, it's completely nullified.

So, Armor has two modes:
-Grant a +2 + 1/6 level enhancement bonus to existing armor.
-Create armor in one of three kinds. The created armor has a +2 + 1/6 level enhancement bonus.

I feel like I need someway of
A - grants +2 bonus for a long long time.
B - grants a massive bonus for a really short time.
and that Cost(A) == Cost(B)

Here's the trick.
What's the equivalent of Instantaneous for buff spells?
1 round duration.
You can build that equality right into the duration rules.

So, what If I make an effect of Armor + Armor + Armor?
Would it:
1. grant +2 for each armor unit
2. grant bonus that is scaling with caster level or spell overall level, and adding same unit grant a special perk - like, in the case of armoring, protecting from criticals, defend from bull rushing, and on.

Hm.
The scaling in 1) aside...
I think it's notable that we found no spells with more than about a +6 bonus to AC, anywhere (aside from one-shot effects like Bungle and Moment of Prescience).
However, there are plenty of spells which are qualitatively different from lower leveled effects- Iron Body grants DR, which seems to fall under the field of Armor, fortification effects like you mentioned, etc. Not all of these effects have "real" level 1 equivalents (speaking generally). However, we still want some way to represent these effects.
Given these examples, I feel that qualitative improvements are probably more consistent with existing spells than quantitative ones.

Does this make sense?

[Indigo Knight]

Yes, yes it does.

My comparison to shield of faith isn't 0+1/4 level, it's 2+1/6 level, scaling up to an eventual +5 (the same value); I feel that there should be a minimum level of noticeable effect. Does that change your opinion at all?

Partially. Having an initial value plays a huge part in the first level. But later levels are influenced more by how steep is the slop. Too much, and Wizard will trample all. Again. Too low, and the Wizard will feel too squishy. Which is why I didn't mention the b.

In my write-up, there's an additional effect, which lets you "create armor", which explicitly does not stack with other worn armor.
How does this avoid the "monk problem"? Easy- the created armor penalizes monks, just as though they were actually wearing armor.

True. But creating armor is, surprise, creating. i.e. conjuration. Conjure, Create, Fabricate, whatever. I'm not sure how to exactly describe what I'm thinking in my head, still it seems to me that if you want to add an actual armor piece then it lies in the scope of Conjuration. And if just want to emulate a piece of armor, why not simply duplicate the benefit, without all the penalties and numbers and limitations of whatever certain physical item have?

Here's the trick.
What's the equivalent of Instantaneous for buff spells?
1 round duration.
You can build that equality right into the duration rules.

Do you have an example you can chart down? Even if it's not fully ready.

I think it's notable that we found no spells with more than about a +6 bonus to AC, anywhere (aside from one-shot effects like Bungle and Moment of Prescience).
However, there are plenty of spells which are qualitatively different from lower leveled effects- Iron Body grants DR, which seems to fall under the field of Armor, fortification effects like you mentioned, etc. Not all of these effects have "real" level 1 equivalents (speaking generally). However, we still want some way to represent these effects.
Given these examples, I feel that qualitative improvements are probably more consistent with existing spells than quantitative ones.

Let's see...

How about this:

Armor:
Grant the spells' target one benefit of your choosing, from the following -

• Add +4 armor bonus.
• Add +1 enchantment AC bonus per 4 levels.
• Add 2/- DR.
• Add the fortification enchatment.

Notice that this intensify the difference between:
Add m amount of bonus per armor unit in spell
VS.
Add m per level bonus. Additional uses of armor unit gives different bonuses.

[aimlessPolymath]

Partially. Having an initial value plays a huge part in the first level. But later levels are influenced more by how steep is the slop. Too much, and Wizard will trample all. Again. Too low, and the Wizard will feel too squishy. Which is why I didn't mention the b.

I'm not sure I agree about what you're saying about scaling. It seems to me that a bonus that scales from +2 to +5 is pretty similar to one that scales from +0 to +5 at high levels, but is more useful at low levels.

True. But creating armor is, surprise, creating. i.e. conjuration. Conjure, Create, Fabricate, whatever. I'm not sure how to exactly describe what I'm thinking in my head, still it seems to me that if you want to add an actual armor piece then it lies in the scope of Conjuration. And if just want to emulate a piece of armor, why not simply duplicate the benefit, without all the penalties and numbers and limitations of whatever certain physical item have?

True. The whole point of Mage Armor is that it's protective, but without the associated downsides.
The issue remains, though, that granting armor bonuses to classes with ACs balanced by not normally having access to armor bonuses is a good way to end up with problems.

The only good way I thought of to prevent it is to reinstate some the limitations of those classes, which runs into the issue you brought up- it doesn't make intuitive sense.

Even if you don't apply all the side effects of armor, is a maximum Dexterity bonus (say, +3) too much? I think it would do a lot to shut down shenanigans.

Armor:
Grant the spells' target one benefit of your choosing, from the following -
Add +4 armor bonus.
Add +1 enchantment AC bonus per 4 levels.
Add 2/- DR.
Add the fortification enchatment.

I think this works! I have a number of changes that I would make before the final version(minimum number of Armor seeds before the last two are "unlocked", damage reduction limits analogous to Stoneskin and Protection from Arrows, explicitly can't use the second mode on the first except in the same spell, bonus scaling still under debate), but this makes sense as a set of abilities for the seed.

Do you have an example you can chart down? Even if it's not fully ready.

Yes.
Fly is a 3rd level spell, affecting one target.
Swift Fly is a 2nd level spell, affecting the caster only(-?), castable as a swift action(+4), and lasting one round (+?).
Similarly, Swift Invisibility is a swift-action spell which affects the caster only, and lasts one round. Same modifiers, but off-by-one for some reason (not sure why) compared to Fly.
Swift Expeditious Retreat lasts one round and is castable as a swift action(+4). It's 1st level, like Expeditious Retreat, but the original spell is already Personal only.

This suggests that converting a spell with a duration in minutes/level to 1 round gives you a spell approximately 4 levels lower, and a conversion from casting on an ally to self-only is worth about 1 level (likely not true in the other direction).

Rough estimations, and of course converting from an additive to a multiplicative system will give you issues, but those are some examples.

Comparing with my Mass Spell equivalence, the difference between a spell that lasts 1 round and one which lasts 1 minute/level is about the same difference, more or less, as the difference between a single-target spell and one which affects one target per level. Furthermore, there's a symmetry between the two that feels right to me in terms of how much effect the spell has.

More generally, consider the conversion from an Instantaneous damaging spell to one with a duration. Suppose the spell has an area.
What is an area damaging spell with a duration?
Something like Incendiary Cloud, where the spell deals damage each round.
What is the instantaneous version of that?
Something like Fireball, where damage is dealt only once.
If Fireball had a duration, what would it be, then?
One round, since damage occurs once.
QED: Instantaneous buffs last one round.

[Indigo Knight]

I'm not sure I agree about what you're saying about scaling. It seems to me that a bonus that scales from +2 to +5 is pretty similar to one that scales from +0 to +5 at high levels, but is more useful at low levels.

If you ignore what comes after the +5 bonus then you are correct. In a theoretical mathematical view you have two lines that cross each other and the one which started lower is going to overtake the other.

Even if you don't apply all the side effects of armor, is a maximum Dexterity bonus (say, +3) too much? I think it would do a lot to shut down shenanigans.

The spell's bonus stands somewhere between light armor and medium armor. This place Dex limit at +3, even +4 at a stretch. But isn't this already happening? Just buying bracers of armor - they don't have dex limit. I'm not sure I'm seeing what your pointing at. And if I did, I wouldn't reiterate that tiny font under equipment-armor for it. I'll try to limit it in other ways, such as stating armor bonuses don't stack, that the spell only works if you wear no armor, or that the spell is pricey. 5th ed' uses "change base AC to 14 + Dex" as a way to override other similar benefits.

QED: Instantaneous buffs last one round.

Took me a few good reads, but I got there in the end.

Let's see.
Cost(t) => Target(t) ? Duration(t) ? Range(t) ? Effect(A)

Therefore
Cost(A) == Target(A) ? Duration(A) ? Range(A) ? Effect(A) == Target(B) ? Duration(B) ? Range(B) ? Effect(B) == Cost(B)

Target is the same, Range is the same. Assuming this doesn't **** up the calculation;
Duration(A) ? Effect(A) == Duration(B) ? Effect(B)

B's duration is 0 if additive or 1 if multiplicative, so we can remove it
Duration(A) ? Effect(A) == Effect(B)

Same for A's effect?
Duration(A) == Effect(B)

So we need a huge buff effect to have the same cost for an above encounter length duration.
Ok, that's clearer now to me after I've written it down. It still doesn't give me numbers and operators to work with. But all in due time, I think.

[aimlessPolymath]

If you ignore what comes after the +5 bonus then you are correct.

I do, because I'm not concerned with epic level scaling. The lines intersect around 20th level. I mean, at that point, you're dealing with stuff like Wish-equivalent stuff.

The spell's bonus stands somewhere between light armor and medium armor. This place Dex limit at +3, even +4 at a stretch. But isn't this already happening?

My concern isn't with people wearing armor and getting the spell- armor bonuses already don't stack. That's just how bonuses work already.

My concern is that people who normally don't wear armor, especially monks, gain the full benefit of armor regardless.
In effect, classes that normally gain "unarmored bonuses" gain the simultaneous benefit of their unarmored bonus as well as the benefits of armor.

Bracers of Armor are not balanced against actually wearing armor- they're balanced against generic AC bonuses like amulets of natural armor and rings of protection, with a cost reduction factor. They're mathematically very similar to wearing +x clothing, or gaining only the enhancement bonus effect of the Armor seed.

[Indigo Knight]

I'm terribly sorry for my long absence. Was not my intention.

I mean, at that point, you're dealing with stuff like Wish-equivalent stuff.

I see wish and other power-house spell as a core problem with the system as it is now. And if I must have them then a major edit would have to take place.

Bracers of Armor are not balanced against actually wearing armor- they're balanced against generic AC bonuses like amulets of natural armor and rings of protection, with a cost reduction factor. They're mathematically very similar to wearing +x clothing, or gaining only the enhancement bonus effect of the Armor seed.

5th edition has a very nice approach in wording the effect in the follows:
"Target's base AC becomes 13 + Dexterity modifier". This turns any other duplicate bonus to override the previous one instead of stacking together.


I tried coming up with more of the keys:

Figment:
Create or conceal one sensation of your choosing, from the following -

• Sight.
• Sound.
• Odor (and taste).
• Touch.
• Magical aura.

Haste:
Target creature gain one benefit of your choosing, from the following -

• Gain extra move action.
• Use move action to act a standard action.

Morph:
Target creature gain one benefit of your choosing, from the following -

• +1 natural armor.
• +10 feet for move speed.
• 1d6 bite attack.
• Two 1d4 claw attack.
• Scent sense.
• Low-light vision.
• +2 to one ability related rolls (chosen on casting).

Phantasm: (or phantom, or dream)
Target creature perceive (=suffer) a mental sensation. Apply one from the following -

• Deal 1d6 psychic pain damage.
• Becomes Shaken.
• Becomes Frightened if was Shaken already.
• Becomes Panicked if was Frightened already.

[Indigo Knight]

I'm trying to look at fire vs. cold;

Other names are like frost or freeze, ice.
And also heat, or blaze.


But I can only think about 1d6 damage.



And are they opposites? Does casting frost is the counter to fire? Lessening it? Like a linear coordinate.

[aimlessPolymath]

Leaving the discussion of damage types aside, I'd like if you could give a solid answer(for discussion) on how much damage the Harm seed does in your notes, based on the scaling of damage that was observed and discussed.
Does a seed deal 1d6 damage, and you can stack (2+L/2)(1+L/3) of them in one spell?
Does it deal (2+L/2)d6, and you can stack (1+L/3) of them?

Related to this question is the question of how much you penalize more generalized seeds. For example, your Phantasm seed deals damage, but also does other things, while the Harm seed just hurts people. How much more damage does the Harm seed to in exchange for this specialization?

Haste granting extra actions is a bit of a thorny road to go down, because its power scales with level in a somewhat dangerous way- it goes from net action disadvantage, to net action advantage- an important threshold

[Indigo Knight]

My notes say that it's around an extra die for 2 levels. 1 for 1st, 2 for 3rd and so on.
I don't remember why I wrote it, so this can be changed.

The next thing to note is that there is a difference between damage inflicted because of source:
Resistance, Save, combination with other spells, environment, etc'.
One way is to have damage coming from a 'damage' seed. So, if your spell is damagy it has to contain that seed. Additional effects are coming from other seeds.
Different way is to have multiple seeds to have the ability to do damage. Like force, electricity or necrotic. Each contributing it's own flavor.
Naturally, these seeds will differ by the amount of damage they present and none would be more then the 'pure' damage seed.

I can see it goes both ways and I'm not sure which is the more appropriate. A sole damage seed is more in alignment with the divisive nature of the system. But the other way explains the way that damage is manifested.

Similarly, Morph grants +1 natural armor, which is analogous to the Armor seed's +1/4 levels mode, but also does other things, too.

This, in essence, mirrors the question about damage.
There's the numeral application. And then there's the origin of said effect. In this case - AC.
Morph seed represents the hardening of the creature's outer layer.
Armor seed represents the adding of a protective layer.
In a general case they both up the number for AC. But in certain situations, like touch attacks, they are different

Haste granting extra actions is a bit of a thorny road to go down, because its power scales with level in a somewhat dangerous way- it goes from net action disadvantage, to net action advantage- an important threshold

It's the best way I found to express this casting. I'm unsure how to portray this otherwise. Where do you think the crossing point, from loss to gain, is?

[aimlessPolymath]

It's the best way I found to express this casting. I'm unsure how to portray this otherwise. Where do you think the crossing point, from loss to gain, is?

When you go beyond a 1:1 ratio of actions spent/gained. At that point, the ability of a player to "go nova" takes a giant leap, similarly to how Celerity and similar effects are so powerful.

My notes say that it's around an extra die for 2 levels. 1 for 1st, 2 for 3rd and so on.
I don't remember why I wrote it, so this can be changed.

The next thing to note is that there is a difference between damage inflicted because of source:
Resistance, Save, combination with other spells, environment, etc'.
One way is to have damage coming from a 'damage' seed. So, if your spell is damagy it has to contain that seed. Additional effects are coming from other seeds.
Different way is to have multiple seeds to have the ability to do damage. Like force, electricity or necrotic. Each contributing it's own flavor.
Naturally, these seeds will differ by the amount of damage they present and none would be more then the 'pure' damage seed.

I can see it goes both ways and I'm not sure which is the more appropriate. A sole damage seed is more in alignment with the divisive nature of the system. But the other way explains the way that damage is manifested.

My personal preference is for somewhat more versatile seeds. While dividing effects more strictly makes for better balance and control over character versatility, I feel that (particularly in the case of damage, one of the most universal effects) makes for freer build decisions, and greater nuance- an illusionist whose illusions can harm is still not necessarily good at dealing damage generally.

On the other hand, it does go against the intent of the system, as you mentioned.


Your notes are likely based on my spreadsheets, which means that Harm would deal (2+L/2)d6, and you can combine up to (1+L/3) seeds in one spell. I don't think you ever explicitly said "this is the value I'm using", so I needed to ask.

[Indigo Knight]

When you go beyond a 1:1 ratio of actions spent/gained. At that point, the ability of a player to "go nova" takes a giant leap, similarly to how Celerity and similar effects are so powerful.

I agree with you about that. So what's the alternative?
I tried something that I thought was being a bit conservative and not that overpowering. Are you able to show where and how the power spike goes for this?

My personal preference is for somewhat more versatile seeds. While dividing effects more strictly makes for better balance and control over character versatility, I feel that (particularly in the case of damage, one of the most universal effects) makes for freer build decisions, and greater nuance- an illusionist whose illusions can harm is still not necessarily good at dealing damage generally.

On the other hand, it does go against the intent of the system, as you mentioned.

Yeah. I kinda goes at it the other way - preferring everything to be organized neatly to logical boxes. Blue does this, red does that. The way I think about it, diminutive partition makes more 'sense'.
However, as you said, it means that every caster suddenly know how to damage and at the same capability, which is very homogenized. It also means that creating variations of damage (or other effects) are trickier to emulate. Plus, Every seed will be skinnier.

Your notes are likely based on my spreadsheets, which means that Harm would deal (2+L/2)d6, and you can combine up to (1+L/3) seeds in one spell. I don't think you ever explicitly said "this is the value I'm using", so I needed to ask.

That's probably right. For transparency sake - this is what's written:

Level Damage
1st 1d6
3rd 2d6
5th 3d6
7th 4d6
9th 5d6
11th 6d6
13th 7d6
15th 8d6
17th 9d6
19th 10d6

[aimlessPolymath]

Are you able to show where and how the power spike goes for this?

I can.

The point above 1:1 is when three seeds can be combined. At this point, you are able to give yourself a standard action, and also give yourself an additional "half action" (i. e get a move action, or buff a move to a standard). You then buff your move action to a standard action. You now have two standard actions. You then cast the spell twice; you now have three standard actions, and can effectively cast three different spells in one turn.

Level Damage
1st 1d6
3rd 2d6
5th 3d6
7th 4d6
9th 5d6
11th 6d6
13th 7d6
15th 8d6
17th 9d6
19th 10d6

Buff that by a small, constant amount (say +Int, or +1d6) and you have my standard progression numbers.

[Indigo Knight]

I can.

The point above 1:1 is when three seeds can be combined. At this point, you are able to give yourself a standard action, and also give yourself an additional "half action" (i. e get a move action, or buff a move to a standard). You then buff your move action to a standard action. You now have two standard actions. You then cast the spell twice; you now have three standard actions, and can effectively cast three different spells in one turn.

Basic draft is using a standard action (casting spell) to get extra move. This is kinda counter intuitive for the caster himself but could be beneficial if granted for an ally (thus complicating the spell creating and upping it's cost).
Doubling the haste seed gives you the ability to get an extra move action for the price of standard and then turn it into a standard action. i.e. breaking even.
Trippling the seed gives you the result of doubling the seed from before as well as additional extra move action. This isn't much but I think this is where the return outgrows the investment.
What's missing in this equation is that unless you cast each seed part separately (3 standard action for 3 seed effects) you increase the price of the spell. We do not have, yet, a mathematical formula for spell cost. But this is where the perpetum mobila can be countered. If it's exponential - 3^3 then the price for a bunch of turns (time stop) is very high. The cost could rise factorialally, logarithmically, geometrically. It's still up in the air. There also the option to inflate the price for those spell in such a way that paying an increase in casting time would have them more or less on par in function slope.

Just a thought. There are a lot of variables I haven't charted down yet. But what do you think?
How would you change that? (Keep in mind that the goal is to have a seed of expediency in some capability).

Buff that by a small, constant amount (say +Int, or +1d6) and you have my standard progression numbers.

Do you mean buff all seeds that grant damage? Or only the best damaging seed?

If it's all of them then I would go with extra damage from ability modifier.