Ideal Number of Character Classes

[DungeonDevil]

For an old-school philosophy-based roleplaying game, what is your sweet-spot for the number of player-character classes?

In Ye Olde Dayes, the two seemed to be the case with a duo of warfare/combat-centric, i.e. the WARRIOR, or FIGHTING-MAN, and, on the other hand, the ENCHANTER/SPELL-SLINGER, i.e. the WIZARD. This seems to have been the case for CHAINMAIL and with In the Labyrinth/The Fantasy Trip. With the addition of an intermediate class, the ROGUE, in Tunnels and Trolls, three was the max. Functionally the Rogue was a mix of both, being the overlapping circles of the Venn Diagram, and, in that regard, seems to have been the same with D&D Basic's Elf race, AFAICT.

Not long thereafter the early players of D&D weren't happy with that dyad and came up with the Thief and Holy Man archetypes, and others followed in quick succession.

At what point were there too many classes to choose from, leading to, shall we say, class saturation?

[Vile Traveller]

One: the Thief. Everything else is just a henchman.

(I'm only a quarter joking, if that.)

[hamurai]

I'm toying around with making the Thief the "basic class" and adding other abilities, like better fighting capability or magic, on top.

[tombowings]

I'm a big fan of the main three: fighting-man, magic-user, cleric. But I also enjoy seeing a small smattering of the other AD&D classes. I'd like to see this party, for example:

Ajak, a fighting-man
Herald, a fighting-man
Borin, a fighting-dwarf
Vance, a magic-user
Brother Crispan, a cleric
Go Ji Wang, a monk

or

Sir Luther, a paladin
Garnet, a fighting-man
Thanbar, a fighting-man
Zelanzy, a magic-user
Brother Leo, a cleric
Alinsar Moonbow, an elf fighter/magic-user

or

Roin, a fighting-dwarf
Soren, a fighting-dwarf
Thorden, a fighting-dwarf
Kargen, a fighting-dwarf
Randalf, a magic-user
Bilblu, a halfling thief

What I don't really enjoy is a mashup of random disjointed misfits:

Kardek, a dwarf fighter/thief
Robin Wood, half-elf ranger
Tuck Greenfingers, an elf fighter/magic-user/thief
Elishein, a gnome fighter/illusionist/thief
Red Eye, a half-orc assassin
Bu Jo So, a monk

[Finarvyn]

I like a basic three or four classes. Fighter, Magic-user, Thief, with the optional one being the Cleric.

The thing is, from a player perspective it seems like the more classes to choose from, the better. 5E has (I think) 11 classes to choose from and each has (again, I think) at least 3 options therein, so 33 variants in the Player's Handbook. More if you add in Sword Coast and other tomes. (And rumor is that Artificer will become a core 12th class.) My players love choices.

From a design perspective I would go back to a minimal number of core classes and let choices be sub-classes again, so Paladin and Ranger are sub-classes of the Fighting Man and so on.

[geoffrey]

Dec 29, 2022 3:37:47 GMT -6 tombowings said:

I'm a big fan of the main three: fighting-man, magic-user, cleric. But I also enjoy seeing a small smattering of the other AD&D classes.

I basically agree.

When Dungeon Mastering AD&D, I enforce rolling abilities using 3d6-in-order but slightly tweaked so as to guarantee that a player can play his choice of the four basic character classes: cleric, fighter, magic-user, or thief. But what if a player wants to play one of the seven other classes from the Players Handbook (druid, paladin, ranger, illusionist, assassin, monk, or bard)? Then he has to roll high enough abilities using a strict 3d6-in-order. The chances of doing that are as follows:

assassin: 1 in 16
druid: 1 in 35
illusionist: 1 in 269
ranger: 1 in 623
paladin: 1 in 1,062
monk: 1 in 2,769
bard: 1 in 58,029

I apply this to my campaign's NPC demographics as well, resulting in about 90% of classed NPCs being of one of the four main classes, while the other seven classes are rare, very rare, or nearly non-existent (depending on the particular class).

I like that: A world in which clerics, fighters, magic-users, and thieves are the norm, while the other classes are the occasional spice. I'd regret the complete absence of the sub-classes, but (even more so!) I'd regret having them anywhere near as common as the four main classes.

[jamesmishler]

I prefer the core four, and even at times have thought of getting rid of the cleric and folding the spells into the magic-user. Easy enough to do, even turn undead works as a spell. Then if you want a traditional fighting cleric or paladin, play a fighter/magic-user focused on those spells.

But as Geoffrey notes, most players, even old school players, prefer variety. Heck, additional classes were among the first developments in the game, so it is kind of hard wired into the system.

Like Geoffrey, most of my NPCs are from the core four; most are fighters or thieves. I use 1 in 4 NPCs as having class levels, 3 in 4 are 0-level Normals or Men@Arms.

I think of course that the reason Geoffrey and I hold similar values is due to the influence of the Wilderlands, where the core four predominated, and demi-humans we're relatively rare...

[tombowings]

I've been revamping my house rules for the new year (work in progress). I've decided to include all the AD&D classes, but have put them in an appendix. That way the standard classes to presented in the forefront without completely shying away from the other fan favorites.

drive.google.com/file/d/1hkS2tEAkCmNqnuJ72BSoDZM9_KhtyNUt/view?usp=sharing

[tetramorph]

I enjoy playing in campaigns with no more than the core four.

I could imagine playing in a campaign with only two: fighting types and magic types. That would be very S&S.

In my Perilous Realms campaign, I stick with the core 3 but allow thieves. The players have to beg for it though.

The first time they ask I always say, "you are all thieves."

In my experience folks that play thieves (then hobbits, and then finally some who play elves) wind up dominating exploration. Because I prefer exporation as role play over mechanics, I like running the game where exploration is what levels all the players because it is a player skill, not a character ability.

[talysman]

I think four classes are the ideal number to start, but not necessarily the same four all the time, or for every area.

First, because I think it's better for players, especially new players, to have a limited number of options so they aren't overwhelmed. Rather than giving players a huge list of classes and skills and feats and races to choose from, just give them four classes and three or four races and an equipment list. It's easier to get going.

A side effect of this is that if a player wishes they could be something else, not giving them that option makes them immediately aware of what they want... So instead of giving them decision paralysis, wait until they ask "I wish I could be a druid instead of a cleric" and then let them be a druid.

Second, though, I'm a big believer that outside of the core four, which I define as Fighter, Magic-User, Hybrid F/MU, and Other Talent (Thief,) all other classes should be variants of the core four. You shouldn't even really need class descriptions for these: just find the core class it most resembles, swap one or two abilities for reskinned abilities related to the new class concept, and go with it. So "Druid" can be just a Cleric with a custom spell list and some druid abilities that replace the turn undead ability, "Witch" can be an MU with a familiar instead of a spellbook and make potions or amulets instead of memorizing spells.

As for the GM side of things, I think a good way to sum up what's different about the starting area or any new area the players enter is to pick its four classes. So, one area might be Gladiator, Illusionist, Mystic, Tinkerer instead of the standard four. It gives the region a different character, and signals to players what's different here.

Because as GM you only have to deal with four classes at any time, plus any quick "class hacks" for players who wanted something different, you as the GM aren't overwhelmed, either, and it's easier to improvise new details when you need them unexpectedly.

[thegreyelf]

I think this is a great question and it really, interestingly, comes down to the VERSION of D&D you are playing. In short: the fewer rules you have in play, the less need there is for additional character classes.

I have been running an Age of Conan RPG for about 10 years now, on and off, using my OD&D hack. I do, in fact, have players running an assassin, a skald (magic-less Strategic Review bard), and a ranger (again, from Strategic Review and magic-less). The rest, however are fighters and a sorcerer. Arguably, I could easily eliminate the ranger and assassin and have them play a fighter and thief, respectively. The reason they chose those sub-classes is simple taste; they'd be just as happy playing the base classes, and the reason I offered them was sheer experimentation. The skald, in fairness, brings much-needed healing to the table since I do not have clerics (the sorcerer is a Howardian-style mage that can be clerical or arcane, but doesn't have healing magic).

This WORKS in OD&D because of just HOW abstracted everything is. The fighter wielding a rapier is just as effective in combat as the fighter wielding the bastard sward is just as effective as the one wielding the 2-handed battle axe, because all weapons deal 1d6 damage on a hit.

The moment you get into variable weapon damage, suddenly a fighter with a 2-handed sword is far superior in almost every practical way to the one wielding a rapier, and more mechanical differentiation becomes needed (at least, in the perception of many players). In AD&D and beyond this happened in the form of new character classes. Once you implement special abilities based on class, players tend to want more, newer, more creative...which gives you Arduin.

I kid.

But not really.

So in the end, I feel like the more heavily abstracted and freeform the game, the fewer character classes are needed. By extension, the more rules you have and the more mechanical differentiation WITHIN classes, the more classes you start to see (and even need).

Just my observations.

[barbaribunny]

Like many, I like like thief, fighter, magic user; but if it's sword & sorcery, and it usually is, I like adding the Crypts & Things barbarian too.

In pure game terms it's in roughly the same niche as the thief; but I love the archetype and C&T has a faithful, flavourful take on it.

I don't mind clerics; but I seldom want more than four or classes unless it's a one-shot. If it is a one-shot, I'm fairly open to random classes from zines.

[tombowings]

One important factor is the number of player at the table. At a table with 3-5 players, 4 types of character is probably enough. For my PbP game, 4 classes had worked out wonderfully. My home game, however, had 14 players, although I've never had more than 11 show up on any one afternoon. So large a roster really cries out for a bit more variety, at least in my opinion.

[Vile Traveller]

Dec 29, 2022 23:04:51 GMT -6 tombowings said:

So large a roster really cries out for a bit more variety, at least in my opinion.

Good point, though I've never even attempted to run a game with 11 players! This could be handled nicely by something like the single class supplements that exist for Delving Deeper - just pull one out if you need more variety. Though I think you can still get away with at least 2 of each without things seeming too 'samey'.

This has me thinking, following up on my not-facetious 'thieves only' comment above, and inspired by 5E's archetypes that focus you class (e.g. burglar, assassin, arcane tricjster). So, you could have only thieves, but with lots of different specialisms.

I also like the idea of just giving thieves' skills to everyone.

[Finarvyn]

Dec 29, 2022 9:21:07 GMT -6 tombowings said:

I've been revamping my house rules for the new year (work in progress). I've decided to include all the AD&D classes, but have put them in an appendix. That way the standard classes to presented in the forefront without completely shying away from the other fan favorites.

drive.google.com/file/d/1hkS2tEAkCmNqnuJ72BSoDZM9_KhtyNUt/view?usp=sharing

By the way, love your house rules. Great art and a lot of thought clearly put into this. I'm trying to collect my ideas into a single doc at some point, and yours is similar to what I hope to end up with.

[Greyharp]

Dec 29, 2022 11:27:54 GMT -6 talysman said:

I'm a big believer that outside of the core four, which I define as Fighter, Magic-User, Hybrid F/MU, and Other Talent (Thief,) all other classes should be variants of the core four. You shouldn't even really need class descriptions for these: just find the core class it most resembles, swap one or two abilities for reskinned abilities related to the new class concept, and go with it. So "Druid" can be just a Cleric with a custom spell list and some druid abilities that replace the turn undead ability, "Witch" can be an MU with a familiar instead of a spellbook and make potions or amulets instead of memorizing spells.

Because as GM you only have to deal with four classes at any time, plus any quick "class hacks" for players who wanted something different, you as the GM aren't overwhelmed, either, and it's easier to improvise new details when you need them unexpectedly.

This is my preference too - core four for me being the vanilla Fighter, Cleric, Magic-User, Thief - and then a light reskinning for players wanting alternate classes. 

[Parzival]

The core seven of B/X/E: Fighter, Cleric, Magic-user, Thief, Dwarf, Elf, Halfling. None others are needed.

The truth is that the various “sub-classes” and “prestige classes” added later are all basically over-powered cheats meant to allow you to have a “more powerful” character than the core options. It’s foolish to choose a Fighter when you can choose either a Paladin or a Ranger with all their assorted special powers? Why be a plain vanilla Cleric (or Magic-user) when you can be a druid and change into an animal at will, among other things? And so on. It’s a power trip which leads to munchkin gaming.
Far better are the basic classes, each with their own advantages and disadvantages, and none really superior to any other.

With the core seven, you’re not picking the best mix of powers— you’re picking a set of options and tactics you want to play. I much prefer that approach to the “awesome character build” which has overtaken the game since 2e on.

As for 5e and its equivalents, the complex potential combinations of classes, feats, character paths and what not are a headache to design adventures for as a DM. And “bounded accuracy” makes it even worse. IMHO, at least for old style play, 5e is broken.

Gimme those old time core classes,
Gimme those old time core classes,
Gimme those old time core classes,
That’s D&D to me.

Oh, the Fighters do the fighting
And the Clerics undead smiting
While the MU hurls the lightning
That’s D&D to me.

Get a Thief to pick their pockets
And open locks and lockets
Sneak right by their old eye sockets,
That’s D&D to me.

Got a dwarf for stone trap finding
And an elf for ghoul denying
While the halfling’s nimbly hiding,
That’s D&D to me.

So…
Gimme those old time core classes,
Gimme those old time core classes,
Gimme those old time core classes,
That’s D&D to me!

— To the tune of “That Old TIme Religion”

[dwayanu]

I don’t think there’s a ‘saturation’ point, as long as each new class enriches your campaign. It can also get pretty arbitrary whether something is called a new class or just another ‘build’ of Class X.

The thing is, the more options formally defined within Class X, the more it’s a smorgasbord, the less it looks and works like a classic class (at least what that means to me).

Professions in RoleMaster are more like packages of point costs. As Monte Cook was an RM guy, it’s perhaps not surprising that D&D 3 ended up with a roughly similar paradigm. That in turn shaped 4E and 5E.

[dwayanu]

I like the idea of folding Thief functions into the Fighting Man class. (I also like to simplify by using a single d6 factor for most of the ‘skills’, starting at 2 and improved by 1 in each 4-level bracket.)

One can distinguish Fafhrd and Grey Mouser by the former having better scores in strength, constitution and wisdom, and tending to wear a mail shirt (and bear a shield, at least outside the city). Mouser is better in dexterity, intelligence and charisma, and eschews armor (which would give a penalty to stealth). His Florentine style fencing may be as good as a shield hand to hand, but no benefit against missiles.

[dwayanu]

The Cleric has always irked me a little as a pure D&D-ism. Bearing the knightly arms of sword and lance it might serve as a Paladin, but I don’t recall Roland (or Gawain or Lancelot) working so many miracles. In the Mabinogion version of Arthur’s court, each of his men has a single outré gift — like the Chinese Ten Brothers tale, but perhaps more oddly colorful.

Even in its role as a D&D-ism, that got somewhat messed up to my thinking as people insisted on giving it offensive spells.

[aldarron]

Depends on what you mean by class. Ideally, IMHO there are three archytype classifications - the martial architype (fighter); the arcane archytype (magic) and the divine arcytype (priest).

So F, Mu, C., But under each archytype or Class, there can be an infinite number of subclasses.

[doublejig2]

Dec 29, 2022 3:37:47 GMT -6 tombowings said:

I'm a big fan of the main three: fighting-man, magic-user, cleric. But I also enjoy seeing a small smattering of the other AD&D classes. I'd like to see this party, for example:

What I don't really enjoy is a mashup of random disjointed misfits:

Kardek, a dwarf fighter/thief
Robin Wood, half-elf ranger
Tuck Greenfingers, an elf fighter/magic-user/thief
Elishein, a gnome fighter/illusionist/thief
Red Eye, a half-orc assassin
Bu Jo So, a monk

Unless they are led by a tall illusionist with impossibly long arms and hands like Otus' cover of Rogue's Gallery, I agree! Otherwise, I question how they would hold up against another more "traditional" adventuring party.

[sixdemonbag]

You really only need to two classes: those that can use magic, and those that can't.

So, for me, magic-users and not magic-users are the only essentials. All the other classes are just flavor variations of those two.

[howandwhy99]

I think it comes down to the breath of activities the simulation challenges the players at. Also how many people will be players on a team, on average. And how the different roles can designed to complement each other.

Three is good. I think it's a small and simple to cover amount. Two or four could also work. I don't think the game necessarily requires different classes to be picked, but it does help. I think the subclasses are different and make it more difficult to contribute without a core class character.