A Few Well-Developed Classes or Many Shallow Classes?

[Erik the Bard]

Basically what it says in the title. I am trying to make my own OSR game, and want info for how the classes should be structured. Any response would be greatly appreciated.

[Alcore]

depends really... on how you want the party dynamic to turn out.

Dungeons and Dragons sell the party as a party; small cogs trying to work together in getting things done that none could do alone. the wizard is frail but powerful. the fighter can take a hit. the cleric can heal. depending on edition they are well developed classes but that forced development into a niche limits characters without branching out into traits or flaws or feats... whatever it is that you can do to separate your character from every other character of that class.

D20 modern went is a different way with base classes. "[ability score] class" of which there was six varieties and you can kinda guess what they were. Tough heroes (those who had a good constitution) were simply tough; supernaturally so with elemental resist and damage reduction built in. the levels capped out at ten so you had to multi class into an "advanced class". each book had a number of new classes (always at least one for each focused hero and a few for the unfocused) and depending on your choices you can multi class into. nearly all requirements could be gained just from skills and feats so the beginning choices in class and ability score didn't matter so much as getting the class skills or just ranks.

the problem is a bit of sameness where no one really stands out both from each other and most human enemies which helped with the modern feel. But if pressed the game can advance with just one maybe two people as a team and build in weaknesses are simply not a thing; though you will miss the healing a cleric can provide.

[Anonymouswizard]

I tend to prefer narrower archetypes selected to be suitable to the intended experience.

However for an OSR system I'd say it's better to have a handful (three/four) solid core classes. Possibly even just making them the only classes and using a system like 2e's kits for customisation.

Possibly my favourite class set is in Romance of the Perilous Land, because I can see how every class bar Thief can have a clear tie to the setting, which can then be customised via background skills. But that's just me liking setting ties, not a bad set of classes in and of itself though.

[LibraryOgre]

I tend to prefer narrower archetypes selected to be suitable to the intended experience.

However for an OSR system I'd say it's better to have a handful (three/four) solid core classes. Possibly even just making them the only classes and using a system like 2e's kits for customisation.

Possibly my favourite class set is in Romance of the Perilous Land, because I can see how every class bar Thief can have a clear tie to the setting, which can then be customised via background skills. But that's just me liking setting ties, not a bad set of classes in and of itself though.

This tends to be my approach, when I have classes at all. Solid classes, with some customization options.

[Anonymouswizard]

This tends to be my approach, when I have classes at all. Solid classes, with some customization options.

I honestly love kits, but I'm kind of not willing to play 2e at this point. Might work on backporting then to BECM or a retroclone based on it.

But I think my absolute favourite way to do classes is how Cypher handles it. You get your broad archetype and your narrow specialty, and the two are completely independent of each other. Sure, some combinations might be less optimal, but there's nothing stopping you from being a bard who burns people's sins.

However that's really hard to do, and probably not very OSR. So I suggest sticking with kits or class groups/subclasses. Or going for a feat system of you want to have even more customisation, but you're again getting away from OSR.

[Duff]

I suggest really, there's a few choices to make:
Do you want only a small number of character types - If so, a small number of classes is fine
Do you want all the diversity decisions to be made in session 0? - Large number of classes or small number with kits
Do you want ongoing options to customise your character? - then you want feats or something similar.

For me, I like lots of options. My favourite version of D&D 3.5, so, put me down for all the options!

[thirdkingdom]

I play B/X-derived games with race-as-class and not much in the way of customization, so lots of classes. For non-humans I typically created 2-3 classes, with a bunch a bunch of human options. The games I run are also very humanocentric, with all starting players human, but the possibility to acquire non-human retainers.

[Balyano]

There is always the option of having a few dozen very specific classes that are only 3-5 levels each. Means every character is multiclass many times over.

Bob the Holy-Knight/Berzerker/Chef and his buddy Phil the Cryomancer/Huscarl/Templar and their traveling companion Betty the Pirate/Viking/Ninja.

[oxybe]

Depends on my mood.

I loves me 2e D&D. the classes are generally pretty light chassis for you to drape stuff on top.

I also loves me 4e D&D where each class is (usually) designed for a specific purpose within the "party" dynamic.

Sometimes I feel like a lite 2e sandwich and sometimes I want a hearty deep dish 4e.

But... looking at my current rules docs for 2e stuff, in one I've got 8 pages of specialty priests written up for a Grayhawk-come-Ravenloft game (and I do have some kits planned/ready to steal), and I'm probably going to have to do just as much for the 2e Eberron demake i've been banging out (which doesn't include the new spells, races, artificer class and dragonmark NWPs I already have written up/stolen from online sources) so YMMV as I might just be a weirdo who wants to make a very complicated 2e sandwich..

[Cluedrew]

What is this game about? What sort of feel are you going for? Do you want characters to be tied into the setting or wanderers? Is optimization something you are interested in promoting? Do you want the minimal solution or are you willing to put in a lot more for smaller returns? What do you want to accomplish with this system and the classes in it?

[Easy e]

Probably not relevant as you are trying to do OSR. I prefer no classes as those are just shoe-horning characters in to a role.

[LibraryOgre]

Probably not relevant as you are trying to do OSR. I prefer no classes as those are just shoe-horning characters in to a role.

Hello? Is it me you're looking for?

[ahyangyi]

Hello? Is it me you're looking for?

Complete Psionics-based D&D?

I wonder whether a single-class variant of Sphere of Power could also fit that.

For example, "everyone is a Hedgewitch", and that probably still leaves a decent amount of character customization options to everyone.

Hedgewitch is picked as a completely random but not-convoluted example.

[LibraryOgre]

Complete Psionics-based D&D?

Character Points, from the Player's Option books. There are rules for Complete Psionics stuff in there, though.

[Anonymouswizard]

Hello? Is it me you're looking for?

Honestly, while I'm sure it's amazing (and will have a look through once I finally get contacts which will let me read from a computer screen without staining), I kind of don't see the point of point buy D&D. I'd much rather just pay one of the many point buy games already out there.

[LibraryOgre]

Honestly, while I'm sure it's amazing (and will have a look through once I finally get contacts which will let me read from a computer screen without staining), I kind of don't see the point of point buy D&D. I'd much rather just pay one of the many point buy games already out there.

Basically, it comes down to people who are familiar with D&D's mechanics, but stifled by some of the choices. Thieves score out way cheaper than every other class.

Mostly, though, it's because I started this project 20 years ago and have been noodling with it ever since.

[Anonymouswizard]

Mostly, though, it's because I started this project 20 years ago and have been noodling with it ever since.

Makes complete sense. I've got two RPG passion projects that I've been working on intermittently for the past ten years. Both of which have gone through complete rewrites, at least partially due to forgetting to back the previous version up a couple of times.

I fully get that at stone point tinkering with it becomes it's own reward.

[thorr-kan]

Basically, it comes down to people who are familiar with D&D's mechanics, but stifled by some of the choices. Thieves score out way cheaper than every other class.

Mostly, though, it's because I started this project 20 years ago and have been noodling with it ever since.

I wonder how all the other rogue classes cost out compared to thieves?

[LibraryOgre]

I wonder how all the other rogue classes cost out compared to thieves?

Well, core only has 2 rogue classes.

I have not checked these numbers in a while (I've made a few changes, but I don't think any impacted Thief and Bard), but Thief is 45 points, and Bard is 172. Clerics, the most expensive, are 225. The mean value of a class in the PH is 159.25; the median is 172 (exactly; mage and bard are the same, and they're also 4 and 5 out of 8).

Thief is so wildly out of range that you could fit it entirely within the fighter or ranger classes they'd stay around the median, if not the current mean. It is less than 1/3 the value of either the median or the mean, and 1/5th... only 20%... the value of the most expensive class.

If you use the suggested XP tables from my initial post about it, the thief would get 2nd level at 450 XP, and 9th level at 63,000. Currently, those are 1250 and 110,000. A cleric, under this system, would be 2250 to 2nd level (far more in line with the second best fighter, who can wear armor while casting a wide variety of spells), and 9th at 315,000.

ETA: One aspect that I liked about AD&D was that character creation, for the most part, took place at 1st level. You picked your race, picked your class, and went from there. However, classes wound up as very big chunks. Don't like part of what you get? Well, then, there's not a lot you can do. There's always the option of kits, but those varied wildly in power and concept. S&P started letting you customize classes, but it didn't do a lot of cross-class customization; you couldn't make a Paladin with mage spells or move silently. Add to that the fact that some classes got abilities way too late to matter... a ranger or paladin get spellcasting so late that they might as well not, unless you're playing a game that will reach, or simply starts, at high level (where the game starts to get buggy, anyway).

My C&P D&D combines these. You make your class at 1st level, and then you leave it alone... but you make the class you want. I'm intending to throw in racial abilities, as well, so if you want to make an elven wizard, you can either take the higher XP cost of all the elven abilities you choose OR just say "I am an elf wizard" and not have any abilities that reflect the elven part.

[thorr-kan]

Thanks for he summary, Mark.

I've got a doc somewhere the indexes the various rogue subclasses across settings. I bet I could find a few even lower than the thief.

[LibraryOgre]

Because I was wondering:

If you remove the thief as an outlier, the mean goes from 159.25 to 175.57... a 16 point jump, moving the range from 13 points to 3.

Thieves, mechanically, suck.

[cfalcon]

Thieves, mechanically, suck.

I agree with this, but I always wondered if, at the time, the thief abilities were simply assumed to be worth a lot more than they were, or if maybe they were played differently in the 80s than the 90s. For instance, if your table treats a successful move silently and hide in shadows as largely undetectable by almost all means, and if the thief is backstabbing a decent amount of the time, the class will probably not feel as weak as it is if you play everything to the letter, and I suspect that a lot was not played to the letter at the start.

[Anonymouswizard]

I agree with this, but I always wondered if, at the time, the thief abilities were simply assumed to be worth a lot more than they were, or if maybe they were played differently in the 80s than the 90s. For instance, if your table treats a successful move silently and hide in shadows as largely undetectable by almost all means, and if the thief is backstabbing a decent amount of the time, the class will probably not feel as weak as it is if you play everything to the letter, and I suspect that a lot was not played to the letter at the start.

I think part of the issue with thieves is that their abilities, even their combat ones, are pretty situational. Plus their noncombat abilities had to compete with whatever noncombat skill rules you could use in that edition, which means that even at mundane levels they were likely much more useful in 0e than 2e.