For those who know D&D 3rd Edition, the system has more issues than an insane asylum, with the biggest being the huge disparity in the power and utility of the classes. How to fix that?
In First and Second edition, each class had its own leveling table. As a character tended to only be one class at a time, the power of the class could be taxed through the leveling process. More powerful classes advanced slower than less powerful classes.
Can the same be done for Third Edition?
The short answer is HELL NO, because Third is such the incestuous system. You can't touch one rule without bumping into its entire family.
For example, we could just give each class its own advancement table, but that messes with the system for the freeform mixing and matching of classes. The math to track all the hopping back and forth becomes rather muddled rather quickly. You could write a system for that, and the system would work, but nobody would like it. The system would feel justifiably awkward. Add in the plethora of prestige classes, some of which are lousy, and others of which are lousy unless they're paired with the correct build, in which case they are way powerful. How do you even price something like that? The only way to make that sort of system work is by killing all multiclassing and all prestige classes. And all that is without spending XP on item creation.
The whole case seems hopeless, yet remains as the most powerful tool to rebalancing the game without having to redesign every class. Similar XP should yield similar power, and mucking about with the leveling cost can go boatloads to equaling the field.
As a rough trial o the idea, Tier 1 would be +100% cost, Tier 2 +50%, Tier 3 +0%, Tier 4 -20%, Tier 5 -40%, and Tier 6 -50%.
Glancing down the table, I see that increased cost would add quite the delay to the top spellcasters. Good, that's what we what. However, lower tier discounts just don't matter that much. So although this helps bring the top tiers down, it does little to bring the bottom tiers up.
A second idea that I have is to attach a level penalty to any class level that provides spells. As a guestimate, (sum spell levels) x 1000. So getting 2nd level spells would require a (1+2)x1000 or 3000 XP premium to unlock that spell level, where getting to 3rd level (1+2+3)x1000 would be 6,000. That should provide a brake to the overly rapid advancement of spellcasters. The advantage of this technique is that this is prestige class neutral. As for druids, any new form should cost the sum of the CRs of all his other forms. That makes shapechanging into anything that you want quite the dicey proposition. The more forms that you want, the cheaper in CR that they should be. Tier 2 would only pay 500 xp, and Tier 3 only 100 xp. Lower than Tier 3 would pay nothing for spell levels.
For those bottom tiered classes, I would just put in a first level bonus. Those classes would start at 2nd, 3rd, or 4th level, depending on how far down they began. Essentially, 1st level fighter would begin at the 3rd level fighter line. Curiously, this solves many of the fighter's power woes. The fighter wins for early advantages, such as massive hit points, than slowly go away.
This doesn't begin to fix all the issues, but the utility of fighters should get extended up a few levels, while the utility of spellcasters should get restricted down a few levels. High levels are still borked, but I don't care about high-level play. That's just broken beyond measure.