Wizards just released the gist of multiclassing in 4E.
And my unhappiness with 4E continues to grow.
First off, we get a smarmy lecture on how multiclassing in 3E was a bully that picked on the single classes. It is very clear that the priority in game design for 4E was to create core classes and then graft on multiclassing that "plays nice" with those classes.
Note that "plays nice" means "restricts your choices compared to 3E."
Here's Mike Mearls describing multiclassing under 3e:
3rd Edition gave us a simpler, elegant, and intuitive solution that worked wonderfully … for characters who didn’t cast spells. The system also forced the core classes to delay abilities after 1st level to avoid cherry picking, where “clever” players simply took one level of as many classes as possible (or layered single levels on to a primary class) to reap the benefits of ungodly saving throws and bizarre but ultimately frightening combinations of class abilities that — like chocolate and pickle relish — were never meant to be combined by men and women of good taste.
The use of "clever" in quotes tells you that Mearls does not think it is clever at all; the implication is that multiclassers were abusers. And certain combinations are bizarre and in poor taste -- and of course it is Mearls who gets to decide what is in good and poor taste.
We learn that in 4E you "multiclass" by taking a feat from another class, and then at 11th level you can specialize in the second class. So there is no true multiclass at all until 11th level; until then you can add the flavor of another class to your base class.
And "once you take a multiclass feat, you can’t take a class-specific feat for a different class. You can dabble in a second class but not a third."
So multiclassing is really "dabbling," and you can only take two classes, max.
Excuse me! 3E was about adding more options for players, not fewer. These rules are a step backward. They limit what you can do. You have less flexibility, and you cannot make the same characters you could under 3E. So-called multiclassing in 4E gives you fewer options and takes away part of the game that many people enjoyed.
Let me return to something even Mearls admits: "3rd Edition gave us a simpler, elegant, and intuitive solution that worked wonderfully … for characters who didn’t cast spells." That is exactly right! I agree. But instead of fixing spell casting so that it did work with this simple, elegant, and intuitive system, they decided to trash it all and bolt on an inelegantly kluged fix to their beloved 4E core class system.
Let me talk a little more about the advantages of 3E multiclassing, which Mike Mearls mentions ...
Simple -- 3E multiclassing is easy. When you advance a level, you pick a class add a level from that class. That's it. The mechanic is very easy to grasp.
Intuitive -- The old system is readily grasped by new players. It requires very little explanation. it is conceptually easy to figure out.
Elegant -- In this context, we are not talking about white tablecloths and crystal vases. This is the mathematical sense of the word. Per the Wikipedia, "the proof of a mathematical theorem is considered to have mathematical elegance if it is surprisingly simple yet effective and constructive." The 3E multiclass system is elegant because it reflects an underlying design philosophy and applies that philosophy in a way that suggests larger answers to game mechanic issues.
Let me talk a little more about that. The underlying core philosophy of 3E is that the quantum unit, the chunk that cannot be broken down further, is the character level. Moreover, you can therefore combine character levels freely and the game will remain balanced.
The promise of this philosophy is that a multiclassed character can take as many levels of different classes as he or she wants because every level is equal. It should not matter that a character is a wizard5/cleric5, or a fighter2/rogue3/paladin2/monk3, or a bard 10, for that matter -- because all levels are equal, and any combination is equally valid.
Now, we know of course that it does not work that way in reality. As Mearls points out, the first level of a class is often the best level. Some levels are notoriously underpowered, like Fighter 5. And combining spellcaster classes is an especially bad idea, albeit somewhat patched in 3.5 with prestige classes like the Mystic Theurge, the Arcane Trickster, and the Eldritch Knight.
If I were to revamp 3E into a new edition, I would change the classes so that the promise of the philosophy was fulfilled -- that is, so that all levels of all classes are equal, and can be combined freely.
To me, the value of multiclassing is not to boost power. The idea is to be able to create interesting character concepts. If you are limited to core classes only, how do you make the priest of a god of thieves who is himself a thief? How do you make a battle mage? Or a singing knight?
The problem with the early editions of the game is that they were all about limits. Only humans can be paladins. Halfling clerics cannot advance beyond 7th level. Certain magic items can only be used by some classes. There are no druid/monks.
But 3E changed that. In 3E, you can be anything. I just love that flexibility.
Yes, I am the first to admit that my liontaur rogue/sorcerer/fighter/avenger/grim is a power-gamer's creation, although I would argue until the new moon that she is also a very fully developed and interesting character, not a hodgepodge or a portmanteau.
But I will also point to my liontaur bard/druid/seer as an example of multiclassing that is designed to maximize roleplay over power. In a few circumstances, Zeoll is a game changer, but usually he is just an underpowered summoner/buffer without any healing. The point is that I was able to use multiclassing to make a unique character not as a munchkin, but as an RP option.
One of the best things about 3E is the underlying philosophy that players should be able to make their own choices. What skills do you want? What feats? What classes? What PrCs? None of those were under player control in 1E after first level.
The problem is that if you want to offer a wide ranges of choices, it is practically impossible to make all choices equal. For example, a newbie makes a human fighter, a nomad of the steppes who is a tough guy. He takes Toughness, Great Fortitude, and Running. Put him up against the veteran gamer who also makes a human fighter, a bowman. He takes Weapon Focus (longbow), Point Blank Shot, and Precision. These are not equal choices.
So in any choice-focussed system, some choices will be more powerful than others. The trick is to make sure that standard options -- like core classes -- are comparable to the best other options. I think 3E does a decent job of this. Stack a high level pure wizard or cleric up against any core rules multiclass option, and the core caster will be reasonably close in power.
I'm not talking about splat books, which are often way out of whack.
The trick in revising D&D is not to throw the baby of player choice out with the bathwater of possible exploits. The trick is to keep the good while getting rid of the bad.