Whats the deal with prestige classes?

[Phawksin]

This is not a thread to talk about how much we hate prestige classes (we have at least two of those already). This IS a thread to talk about the history of prestige classes, how they've been implemented and how we might use those lessons going forward.

So! If my history/memory serves, AD&D introduced class "Kits": class specific customizations that would change or rework class features to focus on a specific concept. If my understanding of Kits is correct a more modern analogy would be class archetypes in Pathfinder that change a few specific features of a class to more accurately fulfil a concept. However, like so much of D&D's history, Kits came with their heavy dose of power creep and before long playing a class without a Kit was mechanical suicide.

Hindsight: unique character identity inside a class is not only cool but important as many concepts can't be reached with a limited number of classes. However, the idea of mechanical balance and focus (IE a through line of design expectations) are essential. Power creep is bad, mmkay?

Then, in 3.x, we introduced Prestige Classes that used multiclassing. Prestige classes followed a specific concept but came with the (theoretical) drawback of needing to meet certain mechanical and story prerequisites. While the idea of being able to allow your advancement and achievements grant you access to restricted features was objectively awesome, it came with an impressive set of problems. By needing to meet arbitrary base attack bonus and skill requirements it was difficult if not impossible to get access to the class organically so you needed to plan ahead, often multicasting in a variety of other base or prestige classes. This came with two major drawbacks: 1) by spreading your resources (class levels and skill points) thin to meet a prestige class's prerequisites as early as possible a character could be permanently nerfed to the point of unplayable. 2) with enough (read: marginal) system mastery it was possible to make a character that took advantage of every resource expenditure feature making them far more powerful than other characters of their level. And these two major problems don't even begin to touch on the impressively vast number of prestige classes released, upwards of five times the number of base classes.

Hindsight: prestige classes directly rewarded character advancement in a way that normal class progression, even normal multiclassing, couldn't. Reaching a prestige class was something akin to completing a personal quest. Still, their quantity, complexity and required system master was staggering. And, as always, power creep is bad, mmkay?

In the 4e era we had Paragon Paths and Epic Destinies. These were tied to specific tiers of gameplay (levels 11 and 21, respectively). The system for these drew a balance between Kits and Prestige Classes as a specific paragon path might be accessible only by a member of a specific base class or one of many base classes or even based on your power set (as in any Martial character) depending on the role of the path in both a mechanics and story. With paths and destinies being built into the progression of the system suffered far less from the power creep issue in editions of yore at the expense of losing the unique quality associated with Kits and Prestige classes.

Hindsight: Concept specific class concepts can be balanced in their design. This is a surprisingly big deal as it took us 30 years to figure it out. One might say that Paragon Paths and Epic Destinies were the single best part of the entire edition. But saying that might start an edition war so I wouldn't recommend it. Regardless, by removing the uniqueness associated with prestige classes some element of the wonder was taken out. They became just another part of the progression, not something special or spectacular.

My Conclusions: I think prestige classes could work in 5e if we remember some things: 1) don't make billions of them 2) maintain a focused design philosophy 3) reduce system mastery by removing many variables in the base game design (good job!) 4) make strong breakpoints at when they can be accessed (either level 6 or 12) with no exceptions 5) make strong mechanical and story prerequisites that are essential to the identity of the class and 6) each class' identity cannot be replicated with spell selection or traditional multicasting.

TL;DR: Call them Paragon Paths.

[Naanomi]

My thought is to go through the process in each step, and only pursue prestige classes if I say 'no' to all of these:
-is it a distinct concept that needs mechanical representation (i.e. Not just 'summoner wizard but stronger, or 'elf fighter/wizard')
-can I reasonably make it with existing classes and backgrounds?
-can it just be a feat or boon?
-can it be a subclass?
-is it complex enough that it needs to be a base class?

Only then would I consider a prestige class. Most 3.x PrCs would not pass

Note that the idea of classes like this predate 3.x... Bards, Druids, Monks (I think); and Dark Sun Dragons/Avagnions... Ran on 'prestige class' type systems

[pwykersotz]

3.5 was where I started. I've since learned about earlier editions, but it formed my understanding of the prestige class system. Never played PF or 4e, so I'm missing those experiences as well.

Prestige classes served as awesome modular enhancements for characters at first. I was super excited to take my first one. Then I started to realize that the options they gave were incredibly specific and not of general use in a game. THEN I discovered these forums and how you could combine/swap them for great power. Then I realized that by doing so I was abusing a system which was not designed to be used this way and which skewed any and all RP options that were supposed to be ingrained in them. Then I discovered the Tier list, and found that some PRC's were straight power gains.

It was a heck of a ride.

In essence, the lesson that I've taken away from prestige classes are the following:

PRC's are fun addons, special ways to further distinguish characters. But including them in the leveling process makes them inherently subject to min/maxing. If I'm to get a niche and flavorful benefit, I'd rather have it as a boon of sorts, an addon that the DM allows.

RP restrictions are next to useless in the type of games I run/play. We seldom use default fluff, and my groups tend to focus on playing a campaign as opposed to a sandbox, so RP restrictions get clumsily added in or just ignored. In my case, I think that when paired with the lesson above, that one to three PRC's in total should be available as extras if special goals are met in the campaign and be entirely optional. Adding flavor like Magic items, entirely optional and providing a nice little boost in a specific way.

One PRC per character. If you have a goal that splits the difference, make a new PRC that embodies it. This is big for me when I both play and DM, specifically because of inter-table interactions. When I'm DM'ing, I don't have the mental processing power to remember you're a x/y/z/1/2/3, especially when combined with feats and the like. I like to remember these things because it helps me integrate the players into the action. When I have to be reminded after the fact that the player had X ability that would have changed the RP scene, it's frustrating. And when I play, I want the same thing. If I'm not interesting to the DM, it's not worth it because he runs the world, which means the world doesn't care who I am either. There's no point except combat to being a "Rune Scribe" or a "Swiftblade" if the NPC's don't care. It's a class that supposed to be Prestigious, after all. This point rambled a bit. I'm sorry about that.


If these things were adhered to, I would be much more open to Prestige classes in my game. Also Phawksin, interesting points and ideas. And Naanomi, good list of qualifiers. :smallsmile:

[Dimolyth]

My thought is to go through the process in each step, and only pursue prestige classes if I say 'no' to all of these:
-is it a distinct concept that needs mechanical representation (i.e. Not just 'summoner wizard but stronger, or 'elf fighter/wizard')
-can I reasonably make it with existing classes and backgrounds?
-can it just be a feat?
-can it be a subclass?
-is it complex enough that it needs to be a base class?

Only then would I consider a prestige class. Most 3.x PrCs would not pass

Add to the list:
- can it be done by organisation membership?
- can it be a magical item?
- can it be done by simply refluff?
- can it be done by creating one new spell or battle manneuvre?

Here I think more than 99% of 3.X PrCs would not pass.

[Naanomi]

Looking down the PrC index I don't see any I'd really feel the need to keep that couldn't be just fine using an existing system. Runecaster could acceptably be a wizard (or artificer) subclass or feat.

What concept do people feel can't be well represented by other methods besides PrCs? Or is it just to add mechanical 'complexity and options' (which have their place) for their own sake?

[NNescio]

Looking down the PrC index I don't see any I'd really feel the need to keep that couldn't be just fine using an existing system. Runecaster could acceptably be a wizard (or artificer) subclass or feat.

What concept do people feel can't be well represented by other methods besides PrCs? Or is it just to add mechanical 'complexity and options' (which have their place) for their own sake?

Well, it lets people who don't want to take Wizard (because they dislike the flavor, or the spellcasting mechanic) dip into this class.

Useful for people who want to say, build martial characters based on Irish and Celtic heroes who don't use magic other than Runic-based ones.

Like, say, Cú Chulainn? If someone wants to build Lancer, a Rune Scribe dip is probably quite up his alley.

[Naanomi]

If your conception of runes is that they don't take spellcasting to use, let's look at other options:
-can they just be magic items?
-could you have a runemaster feat or boon that covers what you need?

Having finished my review of PrCs I only saw three that I really don't know how to implement in 5e...

-racial PrCs that seem to complex for a feat but should allow entry by many classes (war-forged juggernaut)
-PrCs intended to expand on a feat but not tied specifically to any class (fortune's friend)
-PrCs meant to take a class wildly into a hyper-Specialized path(Master of the Unseen Hand)

[Daishain]

I love the idea of prestige classes from a fluff perspective. In game organizations and elite training leading to at least relatively unique abilities. Done right, it represents a notable roleplay achievement, not just power gaming.

For reasons that others have detailed in length, they got out of hand in 3.x. What could have been a great RP tool became just another facet of the great minmaxing game.

I am strongly of the opinion that prestige classes, or something that achieves the same effect (subclasses are a nice mechanic but don't quite cut it), can be a positive influence on the game. But only so long as they are carefully balanced to be on par with existing single or multiclass character options in terms of power, admittedly not an easy thing to achieve.

[MrStabby]

Personally I think that prestige classes would be good. There have been a few things I was struggling with.

Firstly is the problem that mechanically so many characters feel the same. Pick a paladin, pick an oath and you are pretty much only distinguished from other paladins by fighting style. Characters all end up being mechanically indistinguishable.

Secondly there are a whole lot of concepts that I cannot manage to build with either multiclasses or an existing archetype.

Finally there are a whole bunch of concepts that CAN be built with multiclasses, but are just terrible and sufficiently ineffective to not be much fun.

I think prestige classes can solve some of these problems.

[CNagy]

This is not a thread to talk about how much we hate prestige classes (we have at least two of those already). This IS a thread to talk about the history of prestige classes, how they've been implemented and how we might use those lessons going forward.

So! If my history/memory serves, AD&D introduced class "Kits": class specific customizations that would change or rework class features to focus on a specific concept. If my understanding of Kits is correct a more modern analogy would be class archetypes in Pathfinder that change a few specific features of a class to more accurately fulfil a concept. However, like so much of D&D's history, Kits came with their heavy dose of power creep and before long playing a class without a Kit was mechanical suicide.

Hindsight: unique character identity inside a class is not only cool but important as many concepts can't be reached with a limited number of classes. However, the idea of mechanical balance and focus (IE a through line of design expectations) are essential. Power creep is bad, mmkay?

Kits were incidental to the power creep. In AD&D, the class kits were like 5e Backgrounds on steroids. They had stat requirements, could affect your weapon/nonweapon proficiencies, your starting wealth, and your equipment. They could be unavailable to certain races or certain classes. And they all provided both a benefit and a background. The bonus provided was small but some were undoubtedly useful (the Gladiator got a +1 to attack rolls with a chosen weapon type, a Swashbuckler gets +2 AC when wearing Studded Leather or lighter armor), others were situational in their usefulness (Amazons get +2 to hit and damage on their first hit against a male opponent, Pirates got a +1 to attack when fighting on ships), to limited usefulness (Barbarians are received either very well or very poorly, +2 on reaction checks where the d20 roll was above X or -2 if the roll was below Y, Nobles get a +2 reaction roll bonus when interacting with other nobles, Assassins have an X% chance to identifying a poison.) And however useful their benefit was, their disadvantage was usually just slightly less disadvantageous, but not always. Kits were certainly not balanced, but their effect on the game was minimal.

The Power Creep came from the Player's Option rules. Stuff like subability scores made it so that, in practice, every character had powerful stats. Basically, it's like this: say you have a 16 Strength, and I say to you: "Hey, I'm giving you the option to treat your strength as 14 for the purposes of carrying allowance, but 18 for the purposes of combat related stuff--do you want it?" Or treating your Dexterity as 2 lower for the purposes of ranged attacks, in order to treat it as 2 higher for the purposes of AC and Initiative? That was the tip of the iceberg on the power creep of the Player's Option books. It was essentially min-max heaven, munchkin Nirvana, and at the same time it was about the most customizable that D&D has ever been, because it turned every class into a Point Buy for class features. When you combined that with Multiclassing or Dual Classing... it got crazy.

[Naanomi]

Secondly there are a whole lot of concepts that I cannot manage to build with either multiclasses or an existing archetype.

Finally there are a whole bunch of concepts that CAN be built with multiclasses, but are just terrible and sufficiently ineffective to not be much fun Can you give some examples of these sorts of things that PrCs solved in the past?

If we do get PrCs I hope they are as a whole weaker/less versitile than existing classes...
Yes you should get some unique tools to portray a concept that you otherwise can't in the rules, but never should one decide to enter one for the power boost

Also I think PrCs works best in one of two ways:
-10/15 level progressions that once you start you cannot leave
-3 level dipable affairs that to get tools to explore a concept and then get out to a more 'normal' progression

[Ouranos]

If we do get PrCs I hope they are as a whole weaker/less versitile than existing classes...
Yes you should get some unique tools to portray a concept that you otherwise can't in the rules, but never should one decide to enter one for the power boost

Also I think PrCs works best in one of two ways:
-10/15 level progressions that once you start you cannot leave
-3 level dipable affairs that to get tools to explore a concept and then get out to a more 'normal' progression

Then very few people will take them. Making them directly weaker than existing classes will directly prevent people from wanting them, no matter how "cool" it may look. It's one thing to be given two options, say Shield master vs. Sentinel, and be able to say "Hey, I can use both of these, but which one will benefit me more?" It's quite another to say "Hey, want this cool unique flavor thing? Well, you have to intentionally nerf yourself for it, not just for a level or two but completely."

I'd rather see them put in more feats as opposed to PrCs, or perhaps build upon each of the subclasses to make them more distinct. Another idea might be building upon some of the more unique Feats and having the Feat as the requirement, i.e. a PrC all about fancy things to do with a shield requiring, you guessed it, Shield Master. Seeing as how we've gotten rid of BAB, there's a nice, easy, flat measure we can use a prerequisite: Proficiency. Requiring a Proficiency bonus of 3, for example, sets a distinct point at which you can gain access to a PrC, and yet it's the same for everybody. A simple set of requirements like Proficiency being +3, a specific Feat, and maybe a weapon or skill proficiency, would be AMPLE gating for a PrC. Make the PrC's small, maybe 5 levels, and you have more levels than most people dip into a second class anyway. Make the capstone where the REALLY cool thing is and you can have a worthwhile 5 level "dip" that adds to the flavor, or enhances your concept, may even make a non-viable concept viable, all without (if done right) making currently standard builds OP as hell.

[Naanomi]

I'd rather have a weak but cool PrC than a strong one people seek out for the power boost; but I'm ok as work is made to make them limited enough that they don't eclipse base classes... 'Alienist' not 'Master Specialist: Summoner' for a 3.5 example

[Doof]

I am also against Prestige Classes in general and most thematic alterations should be handled case-by-case with the DM and the player working to earn those titles in-game (or by character creation), not by multiclassing into one.

Keep it 3-5 levels max too, if we're going to have some anyway.

edit: thinking back on 3.5e, Walker in the Waste PrC was such a great prestige class because the whole class was a pathway to becoming an entirely unique character- a Dry Lich without paying the LA cost (since, well, you already paid for the new abilities as class features).

I'd love to see a spellcaster PrC that takes me down the pathway to become a lich!

[CNagy]

Five levels is perfect of Prestige Classes.

Ten is too many; it's basically a base class at that point, only without ASIs and Archetype features. Fifteen even more so; if WotC is going to spend design resources on making 15 levels of a Prestige Class, I'd rather they just go ahead and expand it to 20 levels and make it a base class.

And I'm all for Prestige Classes being more powerful than base classes, though that is a difficult thing to judge. Level requirements ensure that the power of the class doesn't come into the game too early. The stat and skill requirements can force all but a very few builds to accept some level of subpar optimization, in effect paying for the power boost.

But I think the best argument for Prestige Classes is the sheer library of them from previous editions. Already dreamt up ideas, right there for the official conversioning. See, the thing that made PrCs so frustrating and important in 3.X doesn't apply in 5e--with bounded accuracy, you don't need every odd numerical bonus you can dredge up from a menagerie of splatbooks (and so the 5e PrCs won't need to give such bonuses), and the "this DC is the new hard" effect on skill checks doesn't happen. With bounded accuracy, power scales upwards very slowly if at all so instead of going up, it spreads out creating new and ever more varied ways to reach the top of the power scale.

We can immediately toss out any PrC of old that could be summed up as "this base class but better," instead focusing on those PrCs with cool concepts. And then, from that cool concept and whatever useful parts it had in the previous edition, you craft a 5e PrC with 4 levels of cool abilities and 1 level that is basically ASI+. Boom. Done.

[MrStabby]

We can immediately toss out any PrC of old that could be summed up as "this base class but better," instead focusing on those PrCs with cool concepts. And then, from that cool concept and whatever useful parts it had in the previous edition, you craft a 5e PrC with 4 levels of cool abilities and 1 level that is basically ASI+. Boom. Done.

I agree with this. I think that a prestige class should be a difficult choice mechanically but a very easy choice thematically.

I do think that Prestige Classes need a bit of care as they are intended to be multiclassed. Things like the ASI you mentioned need to be balanced but also any abilities like casting - just getting higher level spell slots but no new spells is a bit rubbish for casters. Likewise a second attack is wasted for most martial characters.

To overcome lack of synergy between base class and prestige class the prestige class probably has to have a little more power per level than the base classes.

[Phawksin]

Racial classes and PrCs! I would totally love to see that idea return as either prestige classes or feats in 5e. I also think it would be a cool way to introduce some of the old templates or monster classes, though I do hate to think about the adolescent angst that would be available to a Werewolf class if you become a lycanthrope. Still, a class progression to get some monster abilities would be really rewarding and I think fits the proper mold of a PrC (to broad to fit into a spell/background/subclass to too broad to have its own class with 20 full levels and subclasses).

[Maxilian]

Add to the list:
- can it be done by organisation membership?
- can it be a magical item?
- can it be done by simply refluff?
- can it be done by creating one new spell or battle manneuvre?

Here I think more than 99% of 3.X PrCs would not pass.

I don't think it should be limitated to a magical item (i agree with everything else though)

[Ardantis]

I feel as though customization through subclasses removes the need for lots of PrCs, and keeps PrCs in the realm of 'special ability' and out of the realm of 'overarching character flavor,' and it makes PrCs feel like something your character grows into rather than something he already was.

Hats off to WotC for that.

[Mr.Moron]

3.5 was where I started. I've since learned about earlier editions, but it formed my understanding of the prestige class system. Never played PF or 4e, so I'm missing those experiences as well.

Prestige classes served as awesome modular enhancements for characters at first. I was super excited to take my first one. Then I started to realize that the options they gave were incredibly specific and not of general use in a game. THEN I discovered these forums and how you could combine/swap them for great power. Then I realized that by doing so I was abusing a system which was not designed to be used this way and which skewed any and all RP options that were supposed to be ingrained in them. Then I discovered the Tier list, and found that some PRC's were straight power gains.

It was a heck of a ride.

I honestly think that it's not a ride the majority of the real player base goes on. I'd say that of people I've played with, if I were to raise the idea of a "Build" reactions would break down roughly like this:

60%: "Huh? What exactly do you mean"
20%: "Oh yeah. I'm just kind picking up damage feats, know any good ones? "
10%: "Oh, like planning everything out? Too much work for me I just wanna play so I went the book suggestions."
5%: "Oh, yeah. I'm the best at that. Check out these sweet <suboptimal plans> maybe I'll share my secrets someday"
5%: "<some rough approximation of what places like this forum see as the process>"

The kind of intensive build exercises popular on this forum just seem exceedingly rare among the general player base. People prepping their character with a hyper-competitive RAW-focused tournament-style approach just doesn't seem to be the standard approach out in the wild.

[pwykersotz]

I honestly think that it's not a ride the majority of the real player base goes on. I'd say that of people I've played with, if I were to raise the idea of a "Build" reactions would break down roughly like this:

60%: "Huh? What exactly do you mean"
20%: "Oh yeah. I'm just kind picking up damage feats, know any good ones? "
10%: "Oh, like planning everything out? Too much work for me I just wanna play so I went the book suggestions."
5%: "Oh, yeah. I'm the best at that. Check out these sweet <suboptimal plans> maybe I'll share my secrets someday"
5%: "<some rough approximation of what places like this forum see as the process>"

The kind of intensive build exercises popular on this forum just seem exceedingly rare among the general player base. People prepping their character with a hyper-competitive RAW-focused tournament-style approach just doesn't seem to be the standard approach out in the wild.

That's probably a fair assessment. I had the rather unique opportunity to play among a mix of power gamers and super heavy RP'ers for my first game. I spent the first year or so basically impotent because the power gamers were better at numbers than me due to experience and the RP'ers had heavy DM favoritism. It was still tons of fun, but it might have been a rather special case.

[Naanomi]

Magic items shouldn't replace many PrCs but I could see a few... Acolyte of the Skin, Greenstar Adept, Walker in the Waste, Winterhaut of Imborghu... All involved finding or making magic 'thing' then specializing in its use

[numerek]

One of the ways I think 5e fails is best shown with the fighter archetypes a champion fighter has no good way to get battlemaster manuevers (that feat is an absolute joke and I don't know the legitimacy of the spell-less ranger) But anybody starting in any other class (whose dungeon master allows multiclassing) can get battlemaster manuevers with a 3 level dip and many classes would benefit from the other 2 level's abilities also. If there was a prestige class that gave battlemaster manuevers then that could fix this problem.

The same can be said for wizard schools, a 18 barbarian / 2 wizard can have an arcane ward, but a 20th level wizard that took some other school can't. It makes absolutely no sense that a barbarian can learn something about wizardry that a full wizard can't.

[CNagy]

One of the ways I think 5e fails is best shown with the fighter archetypes a champion fighter has no good way to get battlemaster manuevers (that feat is an absolute joke and I don't know the legitimacy of the spell-less ranger) But anybody starting in any other class (whose dungeon master allows multiclassing) can get battlemaster manuevers with a 3 level dip and many classes would benefit from the other 2 level's abilities also. If there was a prestige class that gave battlemaster manuevers then that could fix this problem.

The same can be said for wizard schools, a 18 barbarian / 2 wizard can have an arcane ward, but a 20th level wizard that took some other school can't. It makes absolutely no sense that a barbarian can learn something about wizardry that a full wizard can't.

...that's just plain old opportunity cost. It's present in almost every RPG in some respect. Without it, there's no reason to even have classes except as general guidelines, so it's hard to consider that a 5e fail.

[Naanomi]

...that's just plain old opportunity cost. It's present in almost every RPG in some respect. Without it, there's no reason to even have classes except as general guidelines, so it's hard to consider that a 5e fail.
I agree, did you feel cheated in 3.X if some Alternate Class Feature combos were unavailable because they both traded in the same feature to get?

[TripleD]

The same can be said for wizard schools, a 18 barbarian / 2 wizard can have an arcane ward, but a 20th level wizard that took some other school can't. It makes absolutely no sense that a barbarian can learn something about wizardry that a full wizard can't.

The barbarian knows Abjuration because he studied Abjuration. The wizard studied something different, so he knows something different.

I took Computer Science for five years at university. I've worked in my field for years. I practically dream in code. But if you started asking me about the best way to control heat and air flow in a gaming rig, I could at best take an educated guess, because that's not something I study. The charted accountant who has never opened an IDE in his life, but regularly builds and tests rigs, is the guy to talk to.

[Malifice]

The 1E Bard was a prestige class.

You had to start as a human or half-elf fighter with a 15 in Strength, Wisdom, Dex and Charisma, a 12 Int and a 10 Con. You went as a normal fighter until at least 5th then switched over to thief until 5th. At that point, you switched over to Bard, gaining a bunch of music and druid stuff.

BECMI also had pseudo prestige classes for non land owning PC's starting at 15th level. Paladin, Avenger, Druid, knight etc. Although they were more like high level archetypes.

They more or less vanished in 2E (barring kits) and came back in force in 3E.

[numerek]

The barbarian knows Abjuration because he studied Abjuration. The wizard studied something different, so he knows something different.

I took Computer Science for five years at university. I've worked in my field for years. I practically dream in code. But if you started asking me about the best way to control heat and air flow in a gaming rig, I could at best take an educated guess, because that's not something I study. The charted accountant who has never opened an IDE in his life, but regularly builds and tests rigs, is the guy to talk to.

But you can study gaming rigs if you want to, in this game the evocation wizard can not study abjuration(except that he can know and cast every abjuration spell more than the barbarian can)

[Naanomi]

They more or less vanished in 2E (barring kits) and came back in force in 3E.
Except Darksun's 'epic prestige class', dragon and avangion

[CNagy]

But you can study gaming rigs if you want to, in this game the evocation wizard can not study abjuration(except that he can know and cast every abjuration spell more than the barbarian can)

Yeah, but you know what an evocation specialist can't do? Specialize in abjuration. He can know all the abjuration spells available to the Wizard, create handfuls of more abjuration spells the likes of which the world has never seen, but he can't give it the single-minded devotion necessary to unlock the secrets of the school because all of that single-minded devotion is already directed at something: evocation. But let's be real, if characters could dabble in a second specialization the world would be overflowing with amateur diviners.

It's a game, and at some point game logic breaks from life simulation. It remains internally consistent, because no class is allowed to go back and double up on the levels they've already passed through. That barbarian can go pick up Battle Master or dabble in Wizardry, but he can't go back and become a Totem barbarian if he already chose to be a Berserker. Heck, if I had the option, I'd retake Fighter level 1 a few more times; hello Defense style, hello Mariner style, hello Archery style, glad I could get you all to join GWF on one fighter without having to be a Paladin/Ranger/Fighter Champion mix.

[Malifice]

Except Darksun's 'epic prestige class', dragon and avangion

True. Knights of Solamnia and Mages of High Sorcery were PrCs in 1 and 2E Dragonlance too.

[Phawksin]

Subclass features as prestige classes is an interesting idea. For most of the subclasses in game it wouldn't be worth it (like the spellcasting subclasses), but Battlemaster has its own unique thing. Personally I don't think its a good idea to give more ways to access the unique parts of a class or subclass, though I can agree that I wish the feat (and the Battlemaster progression) worked a little better. It might be interesting to have something like a PrC that requires 11 levels of wizard that gives you access to a second school of magic, with class features that mimic or exactly replicate the wizard school features. It would probably be too powerful to also grant normal spellcasting progression, but maybe half caster progression. You would lose 9th level spells but all the subclass features from two schools.... Might be interesting.