Cadfan: I almost forgot to respond to you, which would have been completely unfair because I think you gave the smartest strongest response:
Cadfan said:
1) Under the current game design, a character's class determines what skills he can access, and even more importantly it is the most important determinant of how many skills that class can access. If you want to make swashbucklers have more skills per level than a mounted knight, you have to have two classes.
I agree that this gets tricky. In theory, a swashbuckler could be trading his armor proficiencies for more skill points. But if that is the case, it suggests that the best base class for a swashbuckler isn't necessarily a fighter. The problem is right now there isn't a good alternative because barbarian and ranger are overloaded with unnecessary flavor and aristocrat wasn't really designed with PC's in mind. However, even within the system as is, there are several things to think about which suggest you might not need two classes.
First, you can just choose to put more points in intelligence. Viola, more skills. Why should a swashbuckler have more skills than a knight anyway? Can't you have smart, sophisticated, refined, chivilric, learned knights? Isn't that the ideal anyway? So you could have smart skillful knights or smart skillful swashbucklers. Everyone is happy. I should point out that knights and swashbucklers belong to the same social niche, just from different time periods in history. Maybe there isn't as much difference there as you think.
You might object that both knights and swashbucklers both have a broad range of skills that you might not want on the ordinary martial class skill list. There are several possiblities there. One, you could widen the skill lists. Two, you could create a feat called 'unusual background' that permenently added 3 non-restricted skills to a classes's class skills (this is what I currently do). For example, you could take a fighter and choose 'Unusual Background: Martial Artist' to add tumble, balance, and move silently to your class list. Now, you can play a eastern styled graceful fighter. Or you could add tumble, diplomacy, and bluff to play a sauve swashbuckler. Whatever. Three, you could give all or some classes a flexible skill package, where they can choose X number of additional skills to reflect thier particular flavor (I do this already in the case of my homebrew 'Fanatic' class, and have considered it for other classes.) And really, if you care about being a skillful fighter, can't you buy Skill Focus?
Now granted, if you force the spending of a general feat, you are losing raw combat power. But fighter's get lots of feats so its not a big hurt in the long run, and more importantly I don't think you should ever get something for nothing. So much of the time I see PrCs and even base classes were the implicit goal of the design was to squeeze in a few more bonus feats and class abilities than you'd get from a base class. I see people wanting to design a 'duelist' or 'swashbuckler' class that has the same combat power of a fighter, the same AC as a fighter, gains the equivalent of the same or more feats than a fighter (only they are preselected and then called class abilities), has more skills than a fighter, AND doesn't have to put up with all the drawbacks of wearing heavy armor like armor check penalties, the fact that its expensive, and the fact its a serious drowning hazard. I think we should agree to a power level and stick to it.
Finally, this at least can in part be solved by multiclassing. You want to play a skillful swashbuckler, then pick up a couple levels of rogue or another skillful class over the course of your career.
2) Classes provide an important balancing mechanism for preventing combo from breaking the power curve. If Power A is appropriate for level 4, and Power B is appropriate for level 5, but Powers A and B combined are broken at level 5, you have to have some way to stop a player from having both at level 5. Classes do this automatically by giving Power A to one class, and Power B to a different one. A new way would have to be invented to fix this problem.
I'm not quite sure I understand this one. Could you give an example? Most classes are front-loaded anyway, and the exception to that is spell-casting which already has such a good way to 'fix the problem' that its actually a problem.
Part of archetype driven design is what characters cannot do. If the archetype of the swashbuckler is that he's a fast talker, but you give all the same fast talking abilities to the barbarian, you've screwed up the swashbuckler archetype.
But have you screwed up the Barbarian archetype, and is the Barbarian archetype really an archetype? I don't think small access to skill problems are really the best example of what you are trying to get at. I think a better example would be, "Can I play a really skillful character that isn't good at sticking you in the back and twisting it, and if so, what would that character be like?" One of the few things that bothers me about the Rogue (which I think is very well designed) is that it is the only example of a playable high skill non-magical class in core. But I'm not sure that's a failing of the class, so much as a failing of the overall design to fill the right niches.
The most general answer to the "why can't that just be a feat?" question for 3.x, and now the "why can't that just be a talent tree?" question from 4e, is that if you make something a talent tree, you're stopping the player from picking something else.
I don't see this as a bad thing. I think too many of the non-core classes are expressedly 'something for nothing'. This is particularly true of full BAB progession PrC's and full spellcaster level progression PrC's, but generally true of most classes introduced since the core.
If the 3e ninja were transformed into a feat chain for rogues, it might work on a certain level, but then a ninja character would have a bunch of rogue abilities they didn't want, and few feats to spend to customize themselves in any other way than by becoming a ninja. This is sort of the opposite of point 2.
Yes, and it's a good one. However, alot of this depends on how you see this archetype of 'ninja'. You see, I tend to think 'ninja' is just Japanese for 'rogue', and anything else you want is particular options for the already flexible class. Just pile a few of the more powerful ones into the rogue special ability options that they can first take at 10th and make the rest general feats. Afterall, much of the class abilities for ninja are perfectly general feats which really don't need to be confined to the ninja. Why can't other classes 'great leap', 'speed climb', take 'atheletics', use poison and what not? Why can't non-ninja rogues have a 'ghost mind' or 'ki dodge'? So yes, I might make the ninja's that appear a little less flexible, but I'll make everyone else more flexible. Take a few levels in 'monk' if that is to your taste, since really the ninja class is more or less rog10/mon10 with some variaty thrown in. Optionally, add one or two additional bonus rogue abilities at high levels to reflect the problem 'mundanes' have staying relevant at high levels. But the main thing I see as a problem with something like 'ninja' is that often the archetype design is trying to get something for nothing. Eastern flavored martial classes in particular have always had a tendency to have more abilities than their western flavored counterpart. It's not hard to line up the two classes and see, 'Heh, ninja is getting just a bit more than rogue.' But 'Rogue' is just english for 'ninja'.