I've had a lurking fear that since its inception, the "FOUR CLASSES AND NO MORE" proponents were going to eat all the classes in the name of "simplicity". They've already eaten the sorcerer, warlock, warlord, assassin, and specialist wizard/specialty priest, and I'm suspecting the "mage/wizardry/illusionist" method of constructing classes might not be a test balloon for a fighter/paladin/cavalier or a rogue/thief/burglar.
I really wonder if we won't lose paladin, druid, bard, ranger, monk, and barbarian as separate classes after all...
It could have been done, but IMHO it's quite too late now to totally revise 10 classes into 4 only. To me it sounds like they have frozen the current classes breakdown to 10.
I think this notion of "class" is a pretty good compromise between the legacy classes and the stated desire to have classes be defined by story. Each class defines what power source or sources a character has.
Fighter: weaponry
Cleric: prayer
Mage: (arcane) magic
Rogue: skill
Druid: nature
Paladin: weaponry/religion
Barbarian: weaponry/nature
Ranger: weaponry/skill/nature
Bard: magic/skill
Monk: weaponry/skill
Classes beyond the core four (five if you count druids) are essentially hybrids in the 4e sense.
I agree, although I don't like the concept of "power source" so I see the distinction based more on the
how than the
why. OTOH "power source" still makes some sense to me as an explanation for different spellcasting.
I am not a fan of the current design choice where 9 classes are archetypes of (let's say) intermediate breadth, while 1 class is an archetype of large breadth. I would prefer more equality, so that we either had Psion, Sorcerer, Warlock as separate classes in addition to the 10 above, or otherwise have 4 core classes and everything else as subclasses, then have subsubclasses (with a better name of course).
From multiclassing point of view, I actually feel like a Wizard/Sorcerer or a Wizard/Warlock is slightly more legit than a Fighter/Barbarian or Fighter/Paladin, because in terms of core archetypes, a Barbarian and a Paladin are already a Fighter. Choosing to play a Fighter/Barbarian or Fighter/Paladin IMXP is more a matter of mechanical optimization, or otherwise used for dual-classing (i.e.
first being one class
then advancing in the second class, to reflect a change in character history). IMHO there is a slight more need for mixing two different narrative "power sources".
If they had gone the 3e/4e route of plenty of classes, the multiclassing default would allow mixing them freely. The Mage case is an odd exception, because otherwise the rest of the classes do work like that.
But since I'm not a fan of multiclassing for optimization, I could have actually liked the opposite approach of making 4 core classes only. No Fighter/Barbarian or Fighter/Paladin, because a Barbarian or Paladin is already a hybrid or a more specific Fighter archetype, but multiclassing only between Fighter/Rogue/Cleric/Wizard. That would have been fine for me...
[note on the Druid: in a game with plenty of classes, the Druid is an absolute essential for my tastes! I could not easily tolerate an edition where Druid isn't in the first PHB. BUT in a core-4 game, Druid could definitely be a type of Cleric. In fact, in such a game I would actually like to see Clerics that were as different from each other as a Cleric and a Druid normally are]
Think that the first two levels are the basic training your shadowdancer must have inside the organization before the true secrets are teach to him.
That's ok, but the problem I referred to was the opposite one, i.e. when you already are a Rogue of level 4 or higher, and the shadowdancers tell you "sorry Joe, we can take you in but can't teach you our tricks, you're already a Thief...".
The rules for mixing up subclasses will come in handy here. Still, for representing elite groups, feat chains might be overall a better choice.
BTW notice how feats are single features that each PC get 4-7 starting from level 4, while subclasses are made of single features that each PC get 4-6 starting at level 3. This is the strongest similarity between the 2 mechanics. It does suggest that a lot of character concepts designed as subclasses can be turned into feat chains or viceversa without too many problems!