A comment in the Mike Mearls thread about class proliferation made me think about the broadness / narrowness of classes. Especially, should significantly different cultures have different classes or at least subclasses?

Back in the 1E days, Gygax's _Oriental Adventures_ created new classes for "bushi" and "samurai" rather than using fighter. But what if your setting spans cultures as different as China and India and others? That's easily possible in Forgotten Realms or Mystara, or in Pundit's Silk Road setting.

Even if a fighter is generic enough, other classes like cleric or wizard are liable to be different for significantly different cultures. Obviously one can go crazy with proliferation of classes, but conversely it isn't necessarily crazy to split fighters up, or especially magic using classes like wizards and clerics.

In my last D&D campaign, I committed to using the standard core classes even though the culture was quite different than standard. I had some rules changes - like substituting quipu for spellbook and had special rules for heavy armor - but I tried to keep them minor. But that was a specific choice for that game. For a different game, I might want more variant classes.

Thoughts?

I think the correct number of base classes for a game to eventually hit is about thirty to fifty.  If the game is stuffed full of subclasses, then the number is probably around fifteen, but each one needs at least four pretty strong variants.

Bushi and Samurai instead of fighter was a solid play.  IMO the only class the OA didn't really benefit from was Oriental Barbarian, given that just plain regular barbarian was honestly similar enough.  Perhaps someone in this thread will defend the honor of the Oriental Barbarian class and change my view.
But overall- paladins aren't clerics, fighters, or cleric/fighter multiclasses, and ninjas aren't rogues.  Every time a new edition comes out with a player's handbook trotting out the same sorry assed lines about how what is a barbarian except for a fighter with worse weapons or what is a mystic shadow assassin guy but a rogue who takes some unoptimal feats I just roll my eyes and wait for the other classes to drop before I consider buying into it.

If my friends really want to play the game, then I have to add extra classes in myself so I can run it, paying the homebrew price.

You can't answer that question in a vacuum. 

If classes are focused on archetypes and/or cultures, then you need more of them.  If classes are focused on mechanical niches, then you need less of them.  If you have a mix of focuses, then the answer moves again.

I prefer few classes (4-10, avoiding the extremes of that range when possible), but that's not because I have some idea that there should be that many classes.  It's because I prefer that the mechanics pull culture and archetype out or class (and race), so that class can focus on niche protection and give a base to the character that will always work in the system.

Whether culture and archetype then gets expressed as color and flavor or has some other mechanics supporting it, is yet another question.  But either way, I'd rather run a game with either of those strategies than one that just keeps gloaming on more classes.

Finally, part of my conviction on this is that I don't think any game mechanic can have massive, straight lists that work well.  Sure, plenty of spells organized by level and school or arcane/divine/druidic or whatever way you want to break them down, as long as any given list is relatively short.  This is because of the tyranny of lists.  If you have 30 things on a list, I guarantee that some of them are substandard to outright garbage while others are the first choices of almost everyone.  You can barely avoid that with a list of 8-12 if you try really hard.  Furthermore, this makes successive choices on the list steadily less valuable.  So after you make your first few picks, players don't get excited about additional picks.

The effect is more subtle with giant class lists, because you don't get another class every time you level (except in wacky multi-class options, which is yet another problem).  Still, when you have 30 classes, you don't really have 30 classes. Each player has around 5 that they like, and maybe another 5 they tolerate.  Given a varied enough player base, the whole group may enjoy as many as half of them.

When you have lists that long it is a symptom that you had an editor without a spine.

Quote from: Steven Mitchell on March 04, 2025, 06:42:34 PMIf classes are focused on archetypes and/or cultures, then you need more of them.

That is the answer, use the archetype system to plug-in the cultural spin on that archetype. Seems the system was designed that way but they never exploited it and instead created the Samurai class at some point and the opportunity seems to have slipped by.

As an aside, can I complain about the 5e samurai subclass that knows how to use shields, doesn't have any reason to ever wield long and short (if you take the dual wield feat you should use two katanas (longswords), if you don't, you cannot wield a katana and a wakizashi, and in any event dual wielding is for turkeys in all of 5.0), and generally should specialize in the same janky weapons as every other 5.0 fighter (polearms, spear+shield, hand crossbow plus stacked feats), or better yet, be an elf, take the elf supremacist feat that stacks great with the samurai's core power, and then shoot a bow or a gun?

That thing really ticks me off.  The first samurai themed power is really the high level thing that gives you a free turn right before you die, and that's not a low level thing.

As others have said, it depends on a lot of other factors. My general rule is that the more involved your classes are, probably the fewer you should have.

In general, I like the idea that a "class" represents occupational training rather than broad archetype. I like culture-specific classes, but I also like games like Warlock! where all a "class" determines is which skills you get off of a relatively short universal list. Even a game like By This Axe I Hack, which does give out class powers, can afford to have dozens of classes because each class only gets a handful and many of them appear in several classes.

In a game where every class gets 10+ unique powers, having dozens of classes is almost inevitably going to turn into a cluster. Even if you somehow manage to invent that many powers without half of them being redundant or fucking the game balance through a garden hose, you're still expecting your DM to be able to account for too many potential options. Part of the reason why games like that acquire a reputation for overpowered PCs, even though numerically DMs could always just sic harder monsters on them, is that it's such a pain in the ass for a DM to even keep track of all the things their players can do. If MMO developers struggle to balance large lists of class powers with the benefit of multimillion dollar budgets, weekly patches and thousands of players they can collect real-time data and feedback from, tabletop designers don't have a hope in hell of pulling it off.

Quote from: Venka on March 04, 2025, 07:45:50 PMAs an aside, can I complain about the 5e samurai subclass that knows how to use shields, doesn't have any reason to ever wield long and short (if you take the dual wield feat you should use two katanas (longswords), if you don't, you cannot wield a katana and a wakizashi, and in any event dual wielding is for turkeys in all of 5.0), and generally should specialize in the same janky weapons as every other 5.0 fighter (polearms, spear+shield, hand crossbow plus stacked feats), or better yet, be an elf, take the elf supremacist feat that stacks great with the samurai's core power, and then shoot a bow or a gun?

That thing really ticks me off.  The first samurai themed power is really the high level thing that gives you a free turn right before you die, and that's not a low level thing.

Not to be too pedantic since i know we are talking about the fantasy archetype of samurai rather than the reality, but samurai, especially in the very early stages of their history, did have handheld shields. When we get to the later, more iconic eras, like the Sengoku Jidai, samurai still kinda had shields in the form of their heavy pauldrons (sode) and the movable shield walls that acted a lot like European pavaise shields.

On the subject of dual wielding, its documented, famously by Musashi Miyamoto, that dueling with two blades is usually worse than just leveraging a single katana with two hands. I think i remember the reason he gave for using two blades at once is basically just to flex how much more skilled you are than your opponent.

But, thematically, yeah, the 5e samurai subclass didn't really feel like it succeeded in fulfilling any samurai fantasy.

I see no need for distinct classes for fighter and bushi.  They are functionally the same thing by different names.  In D&D terms, there really isn't much difference between a pseudo-Englishman shooting a longbow and a pseudo-Japanese man shooting a differently shaped longbow.  There are differences but D&D doesn't model the differences in fighting techniques in enough detail for it to matter.  Both are just a fighter shooting a longbow.  There are cultural differences but D&D doesn't really model that sort of thing much.  That's mainly defined by the way the character is played.  There is no need for a difference class for a fighter that follows a particular code.  I don't think this is a difference that needs to be accounted for mechanically.   

Quote from: Hzilong on March 05, 2025, 12:03:03 AMNot to be too pedantic since i know we are talking about the fantasy archetype of samurai rather than the reality, but samurai, especially in the very early stages of their history, did have handheld shields. When we get to the later, more iconic eras, like the Sengoku Jidai, samurai still kinda had shields in the form of their heavy pauldrons (sode) and the movable shield walls that acted a lot like European pavaise shields.

5e definitely doesn't allow for many distinctions between shield types- and there's precious little design space for such things if you were looking to houserule something in.  The samurai in question are not, as you note, meant to stand in for literally all samurai and every weapon they used.  The fighter class (and the weapons table in general) is broadly meant to support the following styles in 5e: one handed weapon with a shield, two handed strength melee weapon, two handed dexterity ranged weapon, two one handed weapons.  By being a fighter, the samurai is overly drawn to the first of these styles, and of course, a samurai class is supposed to be rewarded by going for literally any of the three.

QuoteOn the subject of dual wielding, its documented, famously by Musashi Miyamoto, that dueling with two blades is usually worse than just leveraging a single katana with two hands.

I'm sure there's some quote that you're thinking of, but this is a guy who founded a school famous for teaching dual wield swords, still in existence today.  I'm sure you know all this; the purpose of this wasn't because it was worse.

Leaving that (possible weeaboo ragebait?) aside, the two ways katana were notably wielded were with two hands (5e considers the katana to fall into the "longsword" rules, so it's a "versatile" weapon that deals 1d8 when used in one hand and 1d10 when used in both), and in one hand, with the second hand holding another weapon.  That first way?  Also badly supported in the rules.  If you have two hands free to use a weapon, you should be using some two handed weapon which has more feats and better stats baseline.

This means that samurai in 5e D&D are definitely pushed away from any of the ways of using a katana that a samurai class would normally incentivize.  The mediocre two-weapon fighting should be the ahistoric two-katana fighting style, the one katana case is best done with a shield instead of with two hands, and if you do go all in on the two-handed katana it's just an odachi with less damage or a naginata with less reach.

It's just a weakness of using a subclass to do what a class should do; or perhaps, having classes that are just generally too complete to glue pieces on.  You could glue on a bunch of bonuses to make the class good at their historical weapons, but it would use up pretty much all the power budget because a fighter subclass just isn't a lot.  More relevantly, by making a class you can just give them the flavorful stuff, be it weaker or stronger, then balance them versus the rest of the game with that. 

This requires the class actually being made to be specific to culture, and the big secret is, the base classes always are specific to European culture- they are all sold to us as generic until its time to sell you another class, be it tied to a real world culture, a popular iconic presentation or myth, or even just some magical thing someone thought up for their world.  Then suddenly, the need to add embelishments arises, and, if mechanically supported, the need to subtract things as well.  And the base classes, even after all these years, are never really designed that well for it. 

So you end up needing another class to make something specific to culture.

I can also demonstrate that making them NOT specific to culture is a bit silly.  The fighter and barbarian don't really line up to stand in for all the various tribal warriors throughout human history- their lists of weapons and armor is too broad.  A version of a fighting-man that was truly generic would start without much armor proficiency or weapon training, would have some pool of learnable stuff, and the culture he grew up in would dictate how he could spend that pool.  That way your knight, samurai, and tribal warrior could all live under the "fighting-man" title, each with realistic skills and such. 

Literally nothing has gotten anywhere close to that.  The base classes have never even remotely had this as a goal; instead, their goal is to work for a specific fantasy vision, typically north eastern Europe, and there's some aside telling us to just treat everything as these things, as these things are playtested and you don't need to do all the homebrew work for them.

So basically I think you make stuff specific to culture either way- but one way, you're just much more honest.  If you go the way recent D&D has gone, you end up making things that are culturally specific (and if the designer knows this, they might try to file off some otherwise very valuable edges), then telling everyone to just jam it into place wherever as a best-fit.

Quote from: Venka on March 05, 2025, 02:16:35 AM

QuoteOn the subject of dual wielding, its documented, famously by Musashi Miyamoto, that dueling with two blades is usually worse than just leveraging a single katana with two hands.

I'm sure there's some quote that you're thinking of, but this is a guy who founded a school famous for teaching dual wield swords, still in existence today.  I'm sure you know all this; the purpose of this wasn't because it was worse.

No, this is correct. If you read Miyamoto, it's clear. The reason for dual-wielding katanas was pure intimidation, and he was clear that it was tactically inferior.

That a school he founded teaches the technique doesn't change this -- a school can teach intimidation as well as tactics.

And of course, schools teach all sorts of things contrary to their founders' beliefs. Just look at Christianity.

Caveat, I don't recall Miyamoto extolling the benefits of wielding a single katana with two hands as suggested. Rather, he taught dual-wielding katana/wakasashi and spoke frequently about the advantages of being (1) fully armed and (2) having a variety of sword lengths to work with.

If you want (more) differences in weapons, fighting styles, and the like--put them in the equipment specs, the weapon abilities, and the combat rules.  Only the middle one of those three belongs anywhere near a class, and then only to the extent that the mechanics cannot or will not support them outside the class. 

For example, if you want a relatively simple system that approximates the reality that weapon training is an involved process, then yeah, you can have no weapon training rules but limit classes to specific weapons.  That means that you might have variant classes with the same structure and abilities but a different weapon list. Naturally, culture stats to get into that too.  Though again, it would be a lot more efficient in that case to have one or maybe a handful of related classes, supplemented by a culture element that provided the lists of valid options.  Which one is better depends on how many classes you intend to use in the same campaign.  The longer the list gets, the more the slight overhead of the culture element begins to pay off in reduced class complexity.

There's no free lunch.  If you want detail in the game, it has to be either mere color or description provided by the participants or it has to be in a rules detail somewhere. 

Quote from: Venka on March 05, 2025, 02:16:35 AMThe base classes have never even remotely had this as a goal; instead, their goal is to work for a specific fantasy vision, typically north eastern Europe, and there's some aside telling us to just treat everything as these things, as these things are playtested and you don't need to do all the homebrew work for them.

So basically I think you make stuff specific to culture either way- but one way, you're just much more honest.  If you go the way recent D&D has gone, you end up making things that are culturally specific (and if the designer knows this, they might try to file off some otherwise very valuable edges), then telling everyone to just jam it into place wherever as a best-fit.

I tend to agree that there will be at least some culture-specific stuff either way.

If you want your game-world to be broad, though, covering a range of cultures - how should it be handled? Does one just have many more classes, adding Kensai and Bushi and such to the class options?

It really depends on the mechanics of the class system. If there are no rules for abilities beyond the assigned ones of the class, then different classes are a must, IMHO.

On the other hand, if it's something like 3e or d20 Modern, where classes have feats and talent trees, then you just need culture appropriate feats/talents

If you were creating a set of rules or a game where... say there were three major cultures/regions in the setting/game.  Maybe you'd have one "fighter" class, and I would have a subclass for each culture if it was warranted where it would apply some cultural restriction and bonus from the start at level 1,  then as advancement is made the bonuses would be mostly the same, but you'd have whatever variation you wanted to make them unique.

Then keep up this pattern as needed, possibly with something similar to a loose feat system  if you feel like you want other abilities/advancements that could help make two characters from the same culture and class feel different, as long as the feats don't feel like there is only one optimal path at least.

It largely depends upon how customizable classes are.

If there is no customization within a class (sub-classes/feats/skills/whatever) then the system can stand to have a ton of different classes since they will overlap less.

If there is already a ton of customization within each class, then you should probably max out at a lower number since they will overlap faster.

Another factor is total system crunch. The more character options available the more different classes you can have without overlap. But despite 3.5 being pretty crunchy, by 2007-8ish they definitely had too many prestige classes. Though part of that was that so many of them obviously sucked.