Quote from: Pat on February 01, 2021, 08:14:31 PM

Quote from: jhkim on February 01, 2021, 08:08:09 PM

Quote from: Pat on February 01, 2021, 07:56:15 PM

Quote from: TJS on February 01, 2021, 04:57:42 PM
I think a lot of the issue with off class race picks is not so much the ability score (unless you're overly fixated on it) but that you gain nothing for the tradeoff.  This frustration will continue.  A half-orc wizard will get something out of Relentless Endurance, but it's not anything they really want (As they want to stay out of melee combat as much as possible), and they will get nothing at all out of Savage Attacks.)  In my experience, the players who fixate the most on having the right optimal combination are often more bothered and frustated by the features they can't use.

That's one advantage of race as class. You don't have to come up with some theoretical orc racial ability that's equally good for all classes. Instead, you can have orc druids and orc gedriht who express the nature of their race in different ways.

What games have multiple classes per race like this, though? As far as I know, there are vanishingly few examples.

You have to make your own. It's not really that hard, in B/X.

ACKS has it and the Player's companion codifies it so they are even easier to build. 
My old 2e group used to build our own with those class construction rules as well.

QuoteWhat games have multiple classes per race like this, though? As far as I know, there are vanishingly few examples.

I have to look for it - but I've read a game where you have 3 races vs 3 roles combo - and each race have it's own different role class - but I cannot remember the name one.

QuoteRace and class only really seems to have a purpose if it keeping the races focused on one thing is to meet the goal of humanocentrism and have the races basically be reasonably rare and unique setting wise.

It justs seems overly redundant to have Dwarf Fighters, and Dwarf Clerics and Dwarf Thieves all as separate classes.

It's not that you couldn't have more than one class as race for a single race, but they should ideally be really clear and very distinct setting archetypes.

Indeed. And that's TBH why I dislike them

Quote6e isn't limited by anything. But .... would it work?

First and foremost your assumptions what 6e shall be are almost certainly doomed to failure.

QuoteACKS has it and the Player's companion codifies it so they are even easier to build.
My old 2e group used to build our own with those class construction rules as well.

But then why even bother if you can play dwarven thief and just suit him to be very dwarvish by choose of skills

Nothing is ever this pure in a game model, but the purist view of "race in class" would be that every such class consists of at least 3 things:  The race, the class, and the thing that is special for that combination of race and class.  That is, a dwarven thief isn't simply the same as an elven thief for more than just the differences in race.  Presumably, since a big part of the purpose of "race in class" is to make it easier to get into play and embed something about the setting into the mechanics, that extra third thing in the combo would reinforce the setting without being terribly complicated.

Using that thought, I might replace the BEMCI/RC elf with a fighter/magic-user hybrid that has a different spell list--perhaps radically cut down version of the magic-user list with a few druid spells thrown in to compensate (and maybe even remove the need for elven clerics in the setting).  Though being able to cast in armor is already a distinction that is much easier to explain.  A dwarven thief might not be that great at stealth in return for increased toughness or other combat ability--something most players would intuitively get.

Without that third thing to make the combination distinctive, might as well split race and class into separate pieces. 

The other problem with race-as-class is that it forces a lot of world-building assumptions onto the game.

You want magical dwarves in your world? good luck in BECMI. In 3e no default race gets an INT bump, so a dwarf can be a wizard as easily as a human, elf or halfling.

Even the "idealized" race-as-class where each has options for various roles, doesn't guarantee an alignment with anything but the developer's ideas for the class; at which point I may as well play a CRPG where such things are commonplace, but the play is smoother and prettier to look at.

If I'm playing tabletop I at least want the freedom to explore my own concepts, not those the developer insists on.

Quote from: Chris24601 on February 02, 2021, 08:32:56 AM
If I'm playing tabletop I at least want the freedom to explore my own concepts, not those the developer insists on.

Rules are oppression! Only a blank sheet of paper is freedom!

Less facetiously, everything you just said applies as much to race and class separately, as it does to race as class. You've drawn an arbitrary line based on personal preference, nothing more. Which is fine, you can like what you like. But it's not a straightjacket.

Well, part of this discussion is the "simple game" versus "game toolkit" conundrum, right?  If you are making a toolkit, one of the things you really ought to do (despite so many designers that don't) is have lots of examples.  Preferably, examples that show different ways of doing the same thing and different slants on similar things.  It would, for example, be a whole lot easier to get a BEMCI magical dwarf class close to good enough on the first pass if there were 3 dwarf options in the game instead of 1.  You could infer some boundaries that way.  Not that it is all that difficult as is, but easier is relative.  Whether or not that would be worth the added complexity in the BEMCI rules is another question.

Quote from: Pat on February 02, 2021, 08:46:45 AM
Less facetiously, everything you just said applies as much to race and class separately, as it does to race as class. You've drawn an arbitrary line based on personal preference, nothing more. Which is fine, you can like what you like. But it's not a straightjacket.

Eh, my own system has species (with options), class (with a focus and a path to further differentiate them) and background (with options) precisely because I do find D&D-style classes to be too restrictive... particularly in connecting combat to non-combat features.

A prime example of it that got me started down that path was actually a recurring issue that came up with the 4E Fighter and Ranger. Specifically, someone would want to play an archer type character, which in 3e could be done easily with a fighter, but the only archer concept in 4E was the ranger, which carried with it built-in background non-combat abilities with a flavor that only fits with wilderness themes.

That was the primary impetus for splitting off combat abilities from background elements. Doing so just made it so much easier for players to model their characters without going full GURPS/HERO style point buy (because there's a sweet spot in options vs. complexity and full point buy is a little too far into complexity for most).

That also led me to a decision to nest various choices to keep it from being nothing but a pile of building blocks (you pick a species, then one or more options for that species; you pick a class, then a focus and path for the class; you pick a background, then particular boons from that background).

This was something 5e learned too with its focus on sub-classes instead of adding countless new classes like 3/4e did. Choose a class, then choose a sub-from a smaller list (whereas in 3/4e each of those would be a separate class from a massive list).

You can see it too in the character creation of the only other game that was once a serious competitor to D&D; Vampire the Masquerade. While technically "classless" it used nested choices to keep character creation on track; prioritize attribute categories, then assign points in those categories; the same for abilities; select a clan and split three dots between the clan's three disciplines. Only 15 points at the end are available to be freely assigned.

Now, you could easily reverse engineer the freebie point costs to the nested options and have the build system just be "distribute 184 freebie points as desired" but, as I've seen in practice... players in general had a much harder time building in that way (it was much easier to completely miss certain sections without realizing it) than having those nested and guided options.

Research backed by playtesting confirmed this for me with my own system; a number of nested choices is the best of both worlds on the options vs. complexity scale. You get plenty of choices but are only picking from a list of less than ten things for most of your individual choices (10 species, 6 classes, 3-4 class focuses, 6-8 class paths, 10 backgrounds, etc.) to keep the potential options manageable.

This lets you define your character as a unique figure in a way that ensures you're not overlooking some critical part of the PC that will be important for the game (i.e. you've assured of having both combat and non-combat mechanical abilities and the combination of species and background fluff will give you a lot of hooks to build your character's personality and backstory from) in the way that pure point might allow and without the straightjacket of more rigidly defined classes or race-as-class.

QuoteUsing that thought, I might replace the BEMCI/RC elf with a fighter/magic-user hybrid that has a different spell list--perhaps radically cut down version of the magic-user list with a few druid spells thrown in to compensate (and maybe even remove the need for elven clerics in the setting).  Though being able to cast in armor is already a distinction that is much easier to explain.  A dwarven thief might not be that great at stealth in return for increased toughness or other combat ability--something most players would intuitively get.

I think Dwarf Thief - should be some Engineer or Inventor guy, tbh

QuoteIf I'm playing tabletop I at least want the freedom to explore my own concepts, not those the developer insists on.

While I'm not gonna defend developers using some weird limits for limits sake, then well it's basically developer job to make clear what concepts are playable in his game, and which are - are not, and in many ways limits of choice shapes game as much or even more than multiplicity of options.

Quote from: Chris24601 on February 02, 2021, 10:29:09 AM
Eh, my own system has species (with options), class (with a focus and a path to further differentiate them) and background (with options) precisely because I do find D&D-style classes to be too restrictive... particularly in connecting combat to non-combat features.

A prime example of it that got me started down that path was actually a recurring issue that came up with the 4E Fighter and Ranger. Specifically, someone would want to play an archer type character, which in 3e could be done easily with a fighter, but the only archer concept in 4E was the ranger, which carried with it built-in background non-combat abilities with a flavor that only fits with wilderness themes.

That was the primary impetus for splitting off combat abilities from background elements. Doing so just made it so much easier for players to model their characters without going full GURPS/HERO style point buy (because there's a sweet spot in options vs. complexity and full point buy is a little too far into complexity for most).

That also led me to a decision to nest various choices to keep it from being nothing but a pile of building blocks (you pick a species, then one or more options for that species; you pick a class, then a focus and path for the class; you pick a background, then particular boons from that background).

This was something 5e learned too with its focus on sub-classes instead of adding countless new classes like 3/4e did. Choose a class, then choose a sub-from a smaller list (whereas in 3/4e each of those would be a separate class from a massive list).

You can see it too in the character creation of the only other game that was once a serious competitor to D&D; Vampire the Masquerade. While technically "classless" it used nested choices to keep character creation on track; prioritize attribute categories, then assign points in those categories; the same for abilities; select a clan and split three dots between the clan's three disciplines. Only 15 points at the end are available to be freely assigned.

Now, you could easily reverse engineer the freebie point costs to the nested options and have the build system just be "distribute 184 freebie points as desired" but, as I've seen in practice... players in general had a much harder time building in that way (it was much easier to completely miss certain sections without realizing it) than having those nested and guided options.

Research backed by playtesting confirmed this for me with my own system; a number of nested choices is the best of both worlds on the options vs. complexity scale. You get plenty of choices but are only picking from a list of less than ten things for most of your individual choices (10 species, 6 classes, 3-4 class focuses, 6-8 class paths, 10 backgrounds, etc.) to keep the potential options manageable.

This lets you define your character as a unique figure in a way that ensures you're not overlooking some critical part of the PC that will be important for the game (i.e. you've assured of having both combat and non-combat mechanical abilities and the combination of species and background fluff will give you a lot of hooks to build your character's personality and backstory from) in the way that pure point might allow and without the straightjacket of more rigidly defined classes or race-as-class.

That's all pretty reasonable. Sounds like you've put some thought into what you like. It looks like you want a relatively complex character creation system that allows you define a lot of aspects of your character. You don't want completely free-form character creation, like a pure point buy system, but you want more flexibility than you'd get if everything is hard-coded into a single, complex package. So you're considering various ways that have been used to split up the hegemony of that single package, and which allow you to reassemble it in different ways that provide an interesting level of variety without too much effort.

But almost all of that is orthogonal to race as class, because when people use that term, they're usually thinking of B/X or BECMI. Which are fundamentally different games than the version you cited (3e, 4e, and 5e). There are a lot less moving bits, especially in character creation options. You could literally roll 3d6 six times, pick a class, roll for hp and gp, buy some equipment, and enter the dungeon. Furthermore, look at the design of the classes: They're much simpler as well. When I said earlier in the thread that it's not hard to make your own, I wasn't kidding. The hardest part is coming up with a concept. There just aren't a lot of moving bits.

That's what race as class is intended for. Having a fixed set of races (r) and profession/classes (c) gives you r x c options. Many of which will be less than optimal, because the racial bonuses or traits won't always align as well with every possible class. So you gain O(n^2) options, but many will be avoided, and they can end up feeling a bit samey, because all elves will be elves in the same way. There are ways to break up the sameyness, like adding roles (not sure if I'm using the term correctly), so your elf can pick from different ways of expressing their elfyness. But that adds more complexity, more design overhead, more work to make sure all the pieces are balanced so nothing breaks, and you'll still end up a usable set of practical options that are far small than all the theoretical possibilities.

With race as class, on the other hand, you don't have to create all the pieces and make sure they all work together. Instead, you just create one thing: The class you want. This is a bit like point buy, but with judgment calls instead of bean counting. You don't have to design all these little pieces and make sure they're all inter-operable and hope the players find one that they like. Instead, you can just design directly for each player. It's bespoke character creation.

Most of the time, it's pretty straightforward. Dwarf mage? Ask the player what they think makes a dwarf mage different from other mages. You might tweak the spell list, and then figure out what kind of racial traits would fun for a dwarf wizard, and then you're done. There are certainly other concepts that can be much harder to express, like a vampire or balrog class (it's much trickier to balance), but those concepts would be even more difficult to express in system like those you favor. But those are exceptions. Most of the time, the core is fairly obvious, and then you just tweak a few bits here and there, and run with it.

You only think of race as class as a straightjacket because you're trying to apply it to a very different set of systems than it's intended for. It's not constraining at all, it's liberating. You just see if that way because you're importing your preconceptions from other systems. It's a sandbox method of character creation.

Just to emphasize how easy class building is in B/X, here's the fighter:

Prime requisite: Strength
XP: As fighter
HD: d8
Armor: Any
Weapons: Any
Saves: As fighter
Attacks: As fighter

That's literally it, the entire class in the Basic set. And here are the changes needed to convert that fighter into a dwarf:

XP: As dwarf
Ability score minimum: Con 9+
Maximum level: 12th
Weapons: Any of normal or small size (no long bows or two-handed swords)
Saves: As dwarf
Special abilities
* Infravision 60'
* Expert miner 2 in 6
* Bonus languages: Dwarvish, gnome, halfling, goblin

A dwarf is just a fighter with the additional restrictions of an ability score min, a maximum level, and minor weapon limitations based on size. In exchange, the dwarf is more resistant to magic and other attacks, has special abilities related to being a miner adapted to the underground, and speaks a ton of extra languages, for whatever reason (are humans assumed to be Americans?).

There are some additional abilities and limits that appear in the Expert set (setting spears, using lances, and at name level the bonus hp and how the XP/save/attack progressions change), but those aren't a concern until higher levels, so we can ignore them for a starting class.

That's worth reiterating as a general principle: Don't design too much ahead. It's a variation on the same principle often espoused for world and adventure creation, in a sandbox setting.

Magic-using classes require a little more effort because of the spell lists, but if we adhere to the principle I just espoused, we only need to worry about a couple spell levels for a starting character. The Basic set only has 1st spells for clerics (8 ), and 1st and 2nd level spells for magic-users (12 and 12). (Yes, there are a handful of 3rd level spells for MUs and 2nd level spells for clerics in the Basic set, but those are explicitly for the DM to help running NPCs, not for PCs.) Since variant classes tend to have less spells per level, a dwarf mage, for instance, might only need 8 and 8. You could even drop that to 4 and 4, and expand the spell list later. And most of the spells will be repeats or minor variations on existing spells, and others can be easily adapted from AD&D1e, the SRD, the 2e spell compendiums, or wherever. That's still pretty simple.

More unique concepts, like the balrog or dragon of OD&D, the monk of 1e, the marshal or warlock of 3e, and so on can be more work, but it's still simpler than in later editions.

Quote from: Pat on February 02, 2021, 01:36:23 PM
With race as class, on the other hand, you don't have to create all the pieces and make sure they all work together. Instead, you just create one thing: The class you want. This is a bit like point buy, but with judgment calls instead of bean counting. You don't have to design all these little pieces and make sure they're all inter-operable and hope the players find one that they like. Instead, you can just design directly for each player. It's bespoke character creation.

Quote from: Pat on February 02, 2021, 02:50:13 PM
Just to emphasize how easy class building is in B/X, here's the fighter:
...
A dwarf is just a fighter with the additional restrictions of an ability score min, a maximum level, and minor weapon limitations based on size. In exchange, the dwarf is more resistant to magic and other attacks, has special abilities related to being a miner adapted to the underground, and speaks a ton of extra languages, for whatever reason (are humans assumed to be Americans?).

Thanks, that clarifies a little more some previous comments. I'd note that bespoke is possible within point systems as well -- the GM can just make rulings about how to balance the PCs rather than insisting on exact point totals. I used to do this frequently in Champions. Newbie players would just say what they wanted and I'd hand them a character sheet.

Apropos the topic, the question would be -- how does the GM handle atypical requests like a dwarf wizard or a halfling mounted fighter? That answer would of course vary from GM to GM, but it is something to consider.

Quote from: Pat on February 02, 2021, 01:36:23 PM

Quote from: Chris24601 on February 02, 2021, 10:29:09 AM
Eh, my own system has species (with options), class (with a focus and a path to further differentiate them) and background (with options) precisely because I do find D&D-style classes to be too restrictive... particularly in connecting combat to non-combat features.

A prime example of it that got me started down that path was actually a recurring issue that came up with the 4E Fighter and Ranger. Specifically, someone would want to play an archer type character, which in 3e could be done easily with a fighter, but the only archer concept in 4E was the ranger, which carried with it built-in background non-combat abilities with a flavor that only fits with wilderness themes.

That was the primary impetus for splitting off combat abilities from background elements. Doing so just made it so much easier for players to model their characters without going full GURPS/HERO style point buy (because there's a sweet spot in options vs. complexity and full point buy is a little too far into complexity for most).

Having a fixed set of races (r) and profession/classes (c) gives you r x c options. Many of which will be less than optimal, because the racial bonuses or traits won't always align as well with every possible class. So you gain O(n^2) options, but many will be avoided, and they can end up feeling a bit samey, because all elves will be elves in the same way. There are ways to break up the sameyness, like adding roles (not sure if I'm using the term correctly), so your elf can pick from different ways of expressing their elfyness. But that adds more complexity, more design overhead, more work to make sure all the pieces are balanced so nothing breaks, and you'll still end up a usable set of practical options that are far small than all the theoretical possibilities.

Regarding mathematical balance of a character design system (non-random-roll)...

In general, I think adding more complexity makes it *harder* to balance all of the pieces so that nothing breaks. Usually it's easier to improve balance by making things simpler. One way is by separating concerns. In Chris24601's system, he separates combat function from non-combat -- so this allows a fighting archer without the wilderness skills of a ranger, for example. That increases the variation possible with fewer problems of suboptimal combinations, because combat balance is mostly independent of non-combat balance. The more the same concerns are touched on with the same choices, the more likely that there will be over-optimal or sub-optimal combinations.

Within D&D, if the goal is to make more of the (R x C) options closer to optimal, the simple solution is to have all races use the same generic attribute generation -- like how it's done in Savage Worlds (among other systems). The posters who oppose this mostly *want* a halfling fighter or half-orc wizard is suboptimal. i.e. It's not an accident, it's that they intentionally want those combinations to be unbalanced.

Of course, a character design system isn't just a mathematical exercise -- it's also about flavor and other priorities. I think the better way to get at this is to separate out what we want character creation to accomplish versus how to design for that.

Quote from: jhkim on February 02, 2021, 03:39:13 PM
I'd note that bespoke is possible within point systems as well -- the GM can just make rulings about how to balance the PCs rather than insisting on exact point totals. I used to do this frequently in Champions. Newbie players would just say what they wanted and I'd hand them a character sheet.

Yep. In fact, some degree is necessary even within point buy systems. With games like GURPS, depending on what's allowed and the focus of the game, it's often fairly easy to make a hyper-specialized character who's far too effective, even without explicit searching for exploits or synergies. And if you don't know what you're doing, it's easy to miss essential components that will end up hurting you later. Having a GM who can look things over and tell certain players to tone things down or make suggestions to other players is essential. HERO is effects-based, so they built in some guidelines and rules to prevent certain bonuses from getting out of hand, but even there I imagine some level of tweaking is necessary (it's not a game I've played). And in general, bespoke is my favorite way to make super hero characters. That comes up naturally in the Marvel RPGs, where it's more fun to start with a concept, and then only later worry about balance between the PCs.

And balance isn't about perfectly matching positives and negatives so everyone comes out equal based on subjective point assignments. It's more about spotlight time and effectiveness. Does everyone have something they're good at? Can everyone contribute in different ways? It's okay if characters aren't equally good at combat, but if combat is really important, then all characters should be able to contribute in significant ways, and nobody should be able to completely outshine the rest. It's a group activity; ensure everyone gets to play. And if social or investigation rolls are important, then everyone should have a niche or a way to help, because otherwise large portions of the game will be spent with half the players twiddling their thumbs while the rest are engaged.

Quote from: jhkim on February 02, 2021, 03:39:13 PM
Apropos the topic, the question would be -- how does the GM handle atypical requests like a dwarf wizard or a halfling mounted fighter? That answer would of course vary from GM to GM, but it is something to consider.

First off: It's a social issue. It should be dealt with at the social level. It really depends on the player and GM. It's fine if the GM has some specific ideas about the setting (dwarves can't use magic), but it's also fine if the player wants to play an atypical character, like a dwarf wizard. They should talk, and figure something out. That might be a hard line, no dwarf wizards. If the player really has a problem with that, that might be a sign the player and GM aren't compatible. Which is fine, there are other groups. Or the player could become a GM. Or the GM might work the idea in, somehow. Are dwarf wizards rare? Is the PC unique?

I throw a lot of this back on the player. What do they want? And then tunnel down from there, because the first answer is always pretty vague. Sure they want a dwarf wizard. But do they want to be the only dwarf wizard? What makes their character so unique? Answers to those kind of questions can start fleshing out not just the social context, but the class details. Maybe the dwarf wizard was born with an infirmity, and lacked the natural dwarven resistance to magic. So terrible saves, but they discovered they can cast spells. This becomes part of the trade off the class, having a significant limitation like losing the standard dwarven saves allows some room to add some new bonuses elsewhere. And the background, like who taught them spells, helps define other traits like the spell list.

Though one caveat: In old school D&D, death is common and characters are cheap. Tunneling too far down is wasted time, if the character dies in the first adventure. It's good to get a general idea, throw something out, tweak it as necessary, and then run with it. And while special snowflake classes are fine, making it a bit more general also makes it less likely the work will be wasted, because it allows Wizdwarf 2 to replace Wizdwarf 1.

Quote from: Pat on February 02, 2021, 02:50:13 PM
Just to emphasize how easy class building is in B/X, here's the fighter:

Yeah, I can't do rules-lite like that. I'd be immensely frustrated by having to rely almost entirely on "GM May I" to accomplish anything. I HATE "GM May I" with the fiery passion of a thousand suns.

I've told the story before, but the short version is that my initial experience with AD&D was via one of the most God-awful dick DM's to every sit behind a screen. If you weren't doing exactly what he wanted to happen in the story in his head the answer to "GM May I" was universal failure.

The residual loathing was sufficient to poison all things pre-WotC D&D for me pretty much forever (and of the D&D editions 4E is far and away my favorite, quite possibly because it so unlike all the other editions). If I hadn't discovered the Robotech RPG ad in the back of the Dragon Magazine subscription my parents got me, I would have given up on RPGs thirty years ago (combat with strikes, parry, dodge, roll with impact, called shots, burst fire and hit locations, plus an actual skill system... wow did that hit JUST the right spot for that point in my rpg development).

Basically, any system without clear "Yes, your character can do this (even if it requires a check to determine the outcome)" level rules for character actions is an immediate turn off for me. There will always be edge cases, obviously, but there's a LOT of actions you can predict being attempted given the genre of the game and I have better things to do with my time than deal with system that can't even be bothered to cover those basics.

It may not be rational, but the experience was so formative (coming in early teens) that its not even something I have an interest in getting over; particularly when there are so many options out there that DO cater to my preferences in terms of game design.

Quote from: Chris24601 on February 02, 2021, 04:12:34 PM

Quote from: Pat on February 02, 2021, 02:50:13 PM
Just to emphasize how easy class building is in B/X, here's the fighter:

Yeah, I can't do rules-lite like that. I'd be immensely frustrated by having to rely almost entirely on "GM May I" to accomplish anything. I HATE "GM May I" with the fiery passion of a thousand suns.

It's not mother may I, that's a bad mischaracterization of how it works. The rules are objective, and once set they're set. What you're doing is collaboratively designing the rules by which your character will be played. Which is no different than saying let's play OD&D, or let's play Traveller. Or saying no to magical flying pixies in a Star Trek game.

Sounds like you just had a bad DM.

Again if we are going back to race as class and where it fits...

No it doesn't let you play a Dwarf Wizard.  But odds are you wouldn't get to play one anyway.  You roll to see what you get, and you play something from the options available to you.

You don't design a unique character.

This is absolutely crucial to understand the shift in paradigm.