Hmmm.
I always had the impression that most broken builds out were mainly made of non-core material (not counting the basic mechanisms like having hit points and saving throws, for instance). Just check out the GOD guide or other handbooks here.
From the handbook in question:
0: As level 5 example
1: Wall of Smoke, Grease, Expeditious Retreat (swift)x2, Targeting Ray, Silent Image, Blockade x2
2: Web, Glitterdust x2, Cloud of Bewilderment (sculpt metamagic for free), Rope Trick, Invisibility, Create Magic Tattoo, Fog Cloud
3: Haste x2, Corpse Candle, Stinking Cloud, Dimension step, Mage Armor (greater), Phantom Steed x2
4: Dimension Door, Solid Fog, Bloodstar, Wall of Sand, Assay Spell Resistance x2, Celerity x2
5: Teleport, Cloudkill, Wall of Stone, Wall of Good, Friend to Foe, Evacuation Rune, Shadow Evocation, Transmute Rock to Mud
6: Freezing Fog x2 (One enhanced with sculpt spell), Tunnel Swallow, True Seeing, Antimagic Field, Dispel Magic (Greater)
7: Stun Ray (Extend for free), Choking Cobwebs (CS) (Sculpt Spell for free), Summon Monster VII, Brilliant Aura, Reverse Gravity
8: Maze, Plane Shift (Greater), Deadly Lahar (CS - huge cone slow effect), Chain Dispel
26 out of 55, less than half of the spells are non-Core. Of them, Chain Dispel, Expeditious Retreat (Swift), Cloud of Bewilderment, and Deadly Lahar can be replaced with Core spells with no noticeable decrease in power. Furthermore, the reason the handbooks list non-Core builds (builds that utilize non-Core PrCs and feats) do so because they have no real choice in the matter:
Core PrCs and feats suck.
However, this does not mean non-Core material is broken. It means it is either better designed or worse designed than Core material.
- I guess they are hardly playtested at all in their interaction with the rest of the core and the other non-core rules, in particular not at the higher levels. See, for instance, the persistent spell/metamagic/nightstick example brought up above.
The Core classes were playtested for years, but were not playtested
properly. The non-Core classes, starting around the XPH-era, were playtested more appropriately, despite only a year's worth of playtesting actually happening.
There are some sections that were not given playtesting because the playtesters did not find those sections interesting (which is how we ended up with stuff like Planar Shepherd, Shadowcraft Mage, and Tainted Scholar). However, by and large the non-Core sources were given better playtesting.
- they favour core caster classes, since I guess all the new material adds new spells and thus way more options for spellcasters (as opposed to more feats for, say, fighters, which are much rarer than new spells, and also will have to fit in a more limited number of feat options available).
This I agree with. The amount of support casters received is unfair, including spells that either replicate entire classes or negate entire splatbooks (Soulmeld Disjunction will forever be the one spell in all of D&D that I refuse to admit exists).
- they introduce many new classes and prestige classes which, of course, to appear more attractive to players to play them (and buy the books), will have a certain power creep (example the ToB warblade)
While power creep exists, only a handful of sources provide bad examples of power creep. Classes that played off of Core support often resulted in significant amounts of power creep. For example, the Archivist is powerful because it has universal access to the Cleric and Druid spell lists, a pair of spell lists that have been expanded upon hundreds of times over. Had the Archivist been restricted to Core Cleric Domains only, it wouldn't have been nearly as much of a game breaker as it is now simply because it's options would not have expanded as dramatically.
Classes that received little to no support are significantly more balanced. As much as I feel Incarnum could have used support, the MoI and DrM soulmelds are more than enough to make a balanced class.
- they introduce completely new game mechanisms which are more powerful than the mere core rulesset like intermediate spell actions (removing the low-level vulnerability of wizards, for instance, a typical core balancing factor) or new kinds of magic (like Psionics)
Immediate actions existed in Core, they just weren't codified properly. It was expanding those options
without those expansions being universally applicable that caused the problems. Immediate actions for Spellcasters were usually new spells, whereas the precious few Immediate actions a Fighter could use were restricted to
feats.
Whenever a system receives partial support like that, it causes problems. The Swift and Immediate actions themselves were not bad ideas, it was how narrow the support was that caused problems.