This basic question about Eclipse: The Codex Persona, and how it was designed, has come up a couple of times – so I thought I’d answer it:
“Did you use some sort of formula to come up with the costs of the various classes magic levels? Eclipse has rules for buying individual spells and powers, and how much caster/manifestor levels cost, but only limited rules for how to price out the spells or power points per day – and trying to reverse engineer the math is giving conflicting results”
I did use formulas to come up with the costs of the manifesting / casting progressions – but to maintain compatibility with 3.0, 3.5, Modern, and various third-party publications, it had to be a retrofit. In writing Eclipse, I took it that – since 3.0 and 3.5 both worked pretty well for a great many people – the systems were passably “balanced”. There were plenty of arguments about what class was best in combat, but I’ve seen a lot of games where combat was rarely a good idea. In any case, those arguments tended to be over classes like the Cleric and Druid – where the fact that members of other classes could take similar duties and obligations, and so jack up their own powers a bit if necessary, provided a built-in counterbalance.
Ergo, I simply expressed each of the 3.0 and 3.5 SRD classes as an equation – so many skill points at cost unknown cost “x”, so many feats at unknown cost “y”, so many levels of spellcasting at unknown cost “z”, so many drawbacks in the form of duties and restrictions, and so on – and then solved them as a set of simultaneous equations. That took some tweaking, since the classes hadn’t been set up systematically to begin with, but it provided some usable starting values to use as a basis.
I could construct some complicated arguments – “Well, once you can cast eighth level spells, a few more per day is not a big deal and they do get them later” to cover things like the relatively small cost difference between the Wizard and Sorcerer spell progressions – but it would still be empirical. After all, a lot of the basics were legacies from earlier editions, right back to Chainmail.
For good or ill, however, when it came to spell and power progressions, empiricism was largely unavoidable. The structures of the SRD tables – how many spell slots or how much power a character got, how high a level of effects they gained access to, and how many effects they were capable of using – were interacting with a purely subjective measurement; how versatile or broad was the list of abilities they were gaining access to and how useful were the individual abilities?
There’s a discussion on the details of that over here:
Since there was no way to precisely price subjective notions like “versatility” and “breadth of theme”, there really wasn’t a good way to provide a formula.
You can build your own spellcasting system though; simply buy a Base Caster Level (page 10) and buy Mana (page 36) as Generic Spell Levels or Power. Then buy some Spells or Psionic Abilities (page 11), paying double if you want to be a spontaneous caster and triple if buying spontaneous powers that can be augmented.
For example, a sixteenth level sorcerer has 188 levels of spells available and knows 32.5 spells (counting level zero spells as 1/2, as usual). It would cost 226 points (on the average) to buy that many Generic Spell Levels, 96 points to buy 16 Caster Levels (these really can’t be Specialized or Corrupted easily, since our freestyle caster here has no restrictions on the kinds of spells he can choose), and 65 CP to buy the spell formula for a spontaneous caster – a total of 387 points versus the 256 it would cost to buy those Sorcerer levels.
Of course, our freestyle caster isn’t bound to a table or to a theme; if he wants to throw 188 first level spells today, or twenty-three eighth level ones, so be it. If he wants to learn Cure Moderate Wounds and Heal to go with Fireball and Teleport, so be it. Those extra points have bought him a lot of flexability.
If he does restrict himself to a theme, he can cut down on the cost of those Base Caster Levels – all the way down to a mere 48 points.
If he wants to be bound to a table, it will corrupt the cost of his generic spell levels – but the table will probably be fairly strict; he’ll be saving quite a chunk of points there. Enough, in fact, to bring the cost down to 151 points.
That will take his total down to 264 points – just a little bit more than the 256 that a standard sorcerer would be paying.
That’s actually somewhat coincidental; the main point is that building your own, freestyle, power pool or spellcasting progression is normally a bit more expensive than using an existing one. That’s intentional; characters who are building personalized progressions will invariably build them to support the character they have in mind as efficiently as possibly. Ergo, it will cost them a bit more – forcing them to trim back on the bits that they don’t need to match that concept.