A question inspired by a couple of threads going on. I'm curious how people would answer this; because I think how you answer the question says a lot about how you approach RPG design. As a note, this question would also apply to professions, archetypes etc., any system where you take your character's primary skills/abilities as a package instead of picking them a la carte.

My own answer is that "a class is there to make it easy to play an archetype". So if a game has a class called "ninja", then that class should provide a player with whatever skills you could reasonably expect a ninja to have. In simplest terms: a class needs to do what it says on the tin. 

There are many other possible answers, though. To list a few:

--to delineate mechanical roles (damage dealer, healer, skill-monkey, etc.)
--to provide game balance
--to encourage teamwork by making it so that no one can do everything
--to reinforce world-building (i.e., these are the occupations prevalent in this world)
--to add tactical depth

and so on.

I imagine for many people the answer will be some combination of the above. For me, there are certainly secondary requirements: a class needs to be engaging to play and not useless. But I'm curious what others regard as the primary purpose or sine qua non of character classes.

1)  To play well understood archetypes

2)  To reduce broken combos - there are are far fewer synergies to deal with if you're only looking for broken combos within a class's silo

3)  Ease of character creation, compared to skill based games

4)  Niche protection / enforcement of teamwork

For me the most important of these is #4, by far

To retain players by giving the players radically different play type experiences in the game they play.

To help embody a role, or character type that appeals to the player.  Probably influenced by fictional books they've read, or movies they've seen.

"Hey, I like dwarves.  How about a Dwarven ranger, of the mountains about us?  He's lived there all his life."  Etc.

I think character classes are a relic from the past (wargame heritage)
Later rpgs abandoned this concept
Modern ones took it back because it's easier (all previous comments explain why brillantly) and maybe millenial players are a bit lazy

I think you've pretty much got it nailed down. Archetype reinforcement and worldbuilding on the narrative side. Then balancing and differentiation on the player side. I'd add one more thing: ease of entry. Most players I know are pretty casual. Building characters in games like call of Cthulhu or savage worlds from zero can be daunting. Giving prepackaged archetypes makes it much easier for players to get started.

I think of classes as "skill bundles" where you can grab a bunch of things that you can do as a collective instead of having to pick them one at a time. (Which is why 5E having classes and skills always seemed like overkill to me.)

I like classes because they encourage folks to specialize -- be the fighter or the wizard or the rogue or whatever. In games that are skill-only, I have found far too many times where every player tried to build a character who can do everything, which makes those characters tend to seem alike.

I've always viewed them as job description shorthand for adventurers.

In my own design I break classes up into three parts; essentially an Archetype (berserker, ironclad, mastermind, skirmisher, mage, mechanist, or mystic) defining broadly your fighting style, a Path (brigand, captain, defender, disabler, ravager, sentinel, striker -or- abjurer, benedictor, empowered, interdictor, maledictor, manifester, summoner) which defines your combat role within the party, and a Background (arcanist, aristocrat, artisan, barbarian, commonener, entertainer, military, outlaw, religious, traveler) which defines your non-combat skills and abilities.I

While there's some overlap at the edges, each of those Archetypes, Paths, and Backgrounds has exclusive elements to choose from initially and while leveling up.

The reason I didn't just go classless is because in my experience classless results in a sort of "grey goo" effect where every PC starts look like each other... unless the GM imposes some restrictions everyone is going to pick up some fighting and some magic and basically end up a gish unless a mechanic is included to prevent a PC from eventually learning to do everything.

Grey Goo can be okay in some circumstances; a real-world military unit or spy type genre would expect a lot of overlap with things like base stats being the main distinction; the big guy, smart guy, fast guy, face, etc.

But, in general, players like having things that set them apart and give them something unique to shine with. Classes or a set of mutually exclusive chunks of classes allow that easily and make it easier for new players to just pick from a menu instead of create from scratch.

The reason I did three sub-chunks is because it allows for a broader range of concepts to be expressed with less overall things to design. There are 490 potential combos just from archetype, path, and background.

It also allows for differentiation within a more constricted theme... everyone's an aristocrat or in the army or an outlaw, or two players what to fight with a skirmishing style or both be mages.I

Few players with a concept are entirely locked in on all three aspects, so you could have a party of three aristocratic skirmishers, but they can still differentiate themselves by one being a defender, one being a captain, and the third being a striker (and also by picking different options from the aristocrat background).

So, there's my answer. Classes are useful as discrete data chunks that distinguish one PC from another. They aren't the ONLY way to get there; a lifepath system where each step is essentially a menu of "what I did at this point in my life" accomplishes largely the same thing in a more homogeneous genre (military, spies, space explorers, etc.); but they're among the most efficient for a genre as broad as fantasy.

If we go back to the original D&D game, a character class represented who your character was rather than a mere job or profession. If a player rolled fighting man, then that told you everything you needed to know about who that character was. Archetypes were strong and well identified. Even elves needed to choose an identity before each adventure. Players natural desire to be able to do whater they wanted led to multiclassing, specializations, and a la carte add ons that diluted the original concept of strong classes as an identity. Once all of that is thrown into the mix there really isn't any point of holding onto classes. 

Quote from: Mishihari on December 07, 2024, 01:03:41 AM1)  To play well understood archetypes

2)  To reduce broken combos - there are are far fewer synergies to deal with if you're only looking for broken combos within a class's silo

3)  Ease of character creation, compared to skill based games

4)  Niche protection / enforcement of teamwork

For me the most important of these is #4, by far

Historically, I agree with this take.  For me, designing a system, however, I drop the archetype part, and put protecting against broken combos ahead of ease of character creation.  So it's niche protection and party-based mindset first, then protection of casual players from themselves, then whatever ease of character creation you can leverage out of the remaining design space last.

I take that stance, because my contention is that chasing archetypes with classes hurts the other three.  I can easily see a design where archetypes really are first, then everything else fits in where it can.  That's basically the early D&D pattern.  However, that works best if the GM (or at least setting designer) customizes the classes for that setting.  The AD&D ranger and paladin are not great fits from a cohesive design perspective.  They are excellent examples of "someone wanted to play Aragorn or Holger Carlsen knock-off in the game, so here's the GM's way to make that happen."

But then when I'm doing my own design, I'm not trying to replicate what D&D already did well.  If I want that, I'll just play D&D.  I'm hitting it from a different angle.

Embedding archetype into class is akin to embedding culture into race.  Yep, you get a lot of great punch with minimal complexity going that route, as long as you don't layer more complexity on top of it.  Once you separate them out, that inevitably ramps up the minimum complexity a bit, but not the extent of having way too many classes, sub classes, prestige classes, multi-class, backgrounds, etc craziness.  One of the reasons it mostly works in AD&D 1E is all the restrictions on race/class combinations.

Finally, I'm heavily influenced by the design patterns of Dragon Quest, where the misnamed "professional skills" are things like multi-level ranger, assassin, navigator, merchant, etc.  These are smaller/narrower than classes, but bigger than skills or even "feats".  They are, in fact, what 3E feats should have been.  That's where I want my archetypes to reside.

Classes are useful for introducing newbies to roleplaying or enforcing particular play styles, but otherwise they're kinda stupid. D&D's class bloat is a perfect example of why. There's way more classes than you need and most of them would be more appropriate as sub-classes.

Quote from: Exploderwizard on December 07, 2024, 11:21:38 AMIf we go back to the original D&D game, a character class represented who your character was rather than a mere job or profession. If a player rolled fighting man, then that told you everything you needed to know about who that character was. Archetypes were strong and well identified. Even elves needed to choose an identity before each adventure. Players natural desire to be able to do whater they wanted led to multiclassing, specializations, and a la carte add ons that diluted the original concept of strong classes as an identity. Once all of that is thrown into the mix there really isn't any point of holding onto classes. 

Yeah, there's a whole continuum between pure class and pure skills.  My experience is that when you get near the middle you lose the benefits of both without gaining much.

I prefer skill-based but the "gray goo" issue is a real thing.  It's useful to promote differentiation between characters in skill based games.  One way to do this is to have some skills give bonuses to related skills, which works but gets complex pretty quickly.  The one I'm going with now is to have checks for each skill depend strongly on a single ability score, those being attack, defense, stealth, perception, magic, and athletics in this particular game.  It thus makes sense to have skills related to your best ability score, which gives six different types of characters from the get-go.  Sounds good in theory, I'm still waiting to see how it works out in practice.

Quote from: Mishihari on December 07, 2024, 02:09:57 PMI prefer skill-based but the "gray goo" issue is a real thing.  It's useful to promote differentiation between characters in skill based games.  One way to do this is to have some skills give bonuses to related skills, which works but gets complex pretty quickly.  The one I'm going with now is to have checks for each skill depend strongly on a single ability score, those being attack, defense, stealth, perception, magic, and athletics in this particular game.  It thus makes sense to have skills related to your best ability score, which gives six different types of characters from the get-go.  Sounds good in theory, I'm still waiting to see how it works out in practice.

An alternate approach to this might be along the lines of Exalted 3e's caste/favored skills. At creation you pick 5 skills from your caste list and any 5 other skills as favored skills for a total of 10 out of 26 abilities.

Those skill ranks and powers linked to them get a xp and training time price break on purchasing them.

You can improve other things, it just costs more, so it somewhat channels your growth towards certain abilities and not others. A sort of really soft class system if you will.