Something you may have noticed if you’ve been following third edition since the beginning is just how refined the balance of the game has grown over time. Yes, there a few glitches, with the Book of Vile Darkness widely perceived as overpowered and a stretch in 2005 where Wizards started to skimp on the editing, but the quality of both official and third-party material has definitely risen since the game’s inception as developers get a better feel for the rules and how to benchmark power.
It’s tough for writers to admit that their old stuff was just plain broken, but I congratulate Wizards of the Coast for shelving their pride when it comes to arcane caster prestige classes. Back in the day, most arcane PrCs fell squarely into one of three categories: too good, mediocre, and awful. Let me conjure up some examples:
The Embermage (Book of Eldritch Might), hailed from the Tome and Blood era of 5/10 arcane PrCs – that is to say, prestige classes which granted only five levels of wizard casting. While thematically cool, was awful in terms of power because it forgot an important rule of PrCs: the benefit must be equal to the cost. Lets call this the law of Equivalent Exchange. The abilities gained – for example, shooting fire rays or burning a monster to ash in an instant – could be easily emulated by the spells granted by those missing five levels of wizard. The modern quick-fix for 5/10 prestige classes, which involves upping them to a cleric’s hit dice and wizard, ignores the rule of dependency wherein each class complements the other’s weaknesses – a fighter and a mage together are superior (and more fun) than a pair of fighter/mage multiclasses of the same level.
Also somewhat rubbish was the transformational arcane prestige class, as per Dragon Disciple and Acolyte of the Skin. Cool prestige classes, the problem was that they were absolutely useless for serious arcane spellcasters because the abilities weren’t worth losing the power and versatility of the spellcaster levels you would give up.
The Loremaster (Dungeon Master’s Guide was the original arcane caster PrC, and like the sorcerer it erred on the side of mediocrity. Granting 10/10 casting progression, the cost of the class was a heavy investment of feats and (as with most PrCs) loss of familiar progression and two bonus feats. The benefits were somewhat weak and linked to Intelligence, but were not entirely bad. Thus this class was not incredibly powerful or special, and merely mediocre. As an early class, however, we might cut it a little slack.
Another artefact of the Tome and Blood era was the 10/10 arcane PrC which also granted some ridiculous amount of special abilities at nearly no cost. You’d have the nominal prestige class entry requirements, lose familiar progression and two bonus feats, and gain yourself some excellent benefits. The problem with these classes was that it raised the bar for every PrC and created a situation where it was ludicrous to go straight wizard when a prestige class could give you great bonuses almost for free.
So what’s the correct way to build an caster PrC? I imagine that writers must simply understand the flaws of early material and accept that some stuff was overpowered, underpowered or too dull. Modern arcane prestige classes typically cost one or two levels of spellcasting, such as one at entry and one at the last level when the ultimate class ability is gained. Any more than this and the spellcaster is set more than one spell level behind, which compromises their important role as a primary spellcaster.
The reason why 5/10 casting was initially considered a fair trade was in part that it was assumed that a wizard would lose only one quarter of his overall power. In reality, lost spellcaster levels are taken from your most powerful spell levels first, so when you trade in two levels you’re really losing ninth level spells in the long run and your ability has to be worth that. In retrospect, a particularly poor prestige class was one that offered a single spell-like ability in exchange for a spell level which would eventually be more powerful than the ability.
Naturally, being limited to 8/10 casting progression or better, sets an cap on how powerful a prestige class’ own abilities can be. However, another option is to take the Archmage route and trade powers for spells per day. Spell slots are very easy to judge and assign a rough gold piece value to.
Finally, you can always offer more powerful abilities than usual if you require a penalty