Quote from: beejazz;581604If they break down these little class-components into feats for those of us that want to tinker, they'll be feat dips rather than a third of your class.

The feats actually have to provide something worth a feat (or several feats in the case of "Feat Tax" prerequisites which I fully endorse but 4e shunned), which is where 4e went wrong.  You pay a feat that lets you lose a class ability and get another, essentially making you pay twice.  I don't understand why 4e doesn't also have a Feats like: "Benefit: You 'gain' HP as a Wizard, unless you already are a Wizard in which case you lose 2 HP per level; Normal: You normally have better HP", because that is how the MC feat chain reads, and why there really aren't any MC'd characters in 4e, and why they released 3 PHBs + Essentials books chock full of character classes (every concept needs it's own class or it doesn't exist).

A feat-based multiclassing system that doesn't take opportunity cost into account isn't going to cut it for me.  There are already feats in 3.5/PF that replicate the class features of other classes, which are fine and in many ways do create a character that feels mutliclassed, but you don't give up your primary classes abilities to gain them.  You pay once.  You give up what could have been a feat to gain greater specialization in order to gain more versatility instead.  

-

Advocating that multiclassing be deliberately subpar is exactly the same thing as advocating for 'trap options', and is a 100% gamist position.  It's absolutely no different than the much maligned Forgist 'play the game the way I think it should be played' mindset and is exactly what 4e already had to offer and flopped with.  

If these Backgrounds and Specialties do create characters that are conceptually different than the core classes, then by the same logic they should also all be inherently worse as well for the purposes of discouraging people from playing things that don't fit into the conception of what the classes are supposed to mean in the setting (despite there being basically no agreement on what that meaning is).  

That is to say that Knights should be worse in all ways to "Fighters", whatever a "Fighter" is, unless Knight becomes a class at which point it becomes 'sanctioned' as acceptable to play in it's own conceptual space and can then be un-nerfed mechanically.  Likewise all Thieves are criminals who steal property (and presumably murder their victims justifying Backstab).  If you wish to play a member of the Thief Class who is not a thief, but instead an assassin you should necessarily be worse at combat, stealth and taking people unaware than a "real' Thief.  

Not because it makes sense, is 'organic' or 'logical' but because you can't have people just going around willy nilly deciding they should be able to play things the game designers subjectively don't like and the game mechanics should harshly reflect these tastes just like how getting into a fight is a mechanically unsound option in Dogs in the Vineyard (which to be clear is a game about playing armed thugs that travel the land to enforce religious orthodoxy).

That's pretty much a recipe for disaster imo.  I think that 4e with it's 30+ classes and what amounts to 5+  different versions of the Fighter demonstrates why this design philosophy is bad.

Well, I was assuming you wouldn't need to pay twice the way you did in 4e. I was just thinking along the lines of class features *as* feats, leaving them fully swappable if the GM goes for it.

Quote from: MGuy;581608Most of the time it is for the extra options, to fill out a character concept, for optimization

The whole extent of any issue with multiclassing in 3.x is front-loaded features.    There isn't really any reason to be higher than 10th level as a Paladin.  There is very little reason to even be 10th level for that matter, but after 10th there is virtually no trade-off in taking levels in another class.

Other d20 games that use identical multiclassing rules don't have the same issue because you actually have to lose something to gain something, which is the way it should be.  In 3.x there is either an insignificant trade-off or a way to bypass it entirely (PrC's that offer full casting, et cetera), meaning it is a bad idea not to multiclass.  

A single class Druid is 'optimized' because their Wildshape ability keeps improving all the way to high levels, and they are full casters.  Druid is one of the only well designed 3.x classes that actually works with the mutliclass rules.  This isn't the case with most classes that either have no abilities that improve continuously, or that are class-specific, which is just poor design and completely avoidable.

I had no issues with the 2e multiclassing/dual classing/kits rules in terms of balance, and character concept because in most cases there were trade-offs that flowed from the class restrictions (ie Fighter/Mages can't wear shields or armor and cast, Fighter/Thief can either sneak well or have heavy armor, but not both at the same time, et cetera).

4e makes you bad at everything as a cost for playing most non-sanctioned character concepts, and 3.x expects no trade-off for stopping advancement in a class in most instances, neither of which are things I particularly care for.

Sounds alot like Fantasy Craft you use backgrounds to get perks.  Classic multiclassing isn't optimal.  Way worse than Pathfinder even.  You can do it but it's seriously a objectively a bad choice.  Much better to do pure class like a Magus or Swordmage if you want a GISH for example.

Quote from: deadDMwalking;581592I love options.  Not every character I've come across in fiction or seen on the silver screen, or even depicted in art is easily defined by a single 'archetypal class' in traditional D&D.

And of course some combination of fighter/wizard is archetypal as well.  While that could be represented with a base class, I haven't seen that done well in any version of D&D.  

I don't think that is an argument in favour of "we must have multi-class rules in our class-based system", as much as "we probably shouldn't be using a class-based system in the first place if it doesn't match what we want from our game".

Class-based systems do what they do really well, open-development systems do what they do really well, but they rarely mix well.

The description of how things work in 5E currently sounds OK. My main concern with it would I guess be how well 'theme' abilities scale up with level.
 
I suppose I'd also like to see a genuine multi-class system to represent characters who have backstories meaning that forsake their old class to pick up a new class (e.g. the thief who touches a Book of Exalted Deeds and becomes a 'born-again' cleric, the fallen cleric who becomes a fighter, the wizard who loses all their powers in a nasty Mordenkainen's Disjunction accident).

Quote from: Wolf, Richard;581613If these Backgrounds and Specialties do create characters that are conceptually different than the core classes, then by the same logic they should also all be inherently worse as well for the purposes of discouraging people from playing things that don't fit into the conception of what the classes are supposed to mean in the setting (despite there being basically no agreement on what that meaning is).  

That is to say that Knights should be worse in all ways to "Fighters", whatever a "Fighter" is, unless Knight becomes a class at which point it becomes 'sanctioned' as acceptable to play in it's own conceptual space and can then be un-nerfed mechanically.  Likewise all Thieves are criminals who steal property (and presumably murder their victims justifying Backstab).  If you wish to play a member of the Thief Class who is not a thief, but instead an assassin you should necessarily be worse at combat, stealth and taking people unaware than a "real' Thief.  

Not because it makes sense, is 'organic' or 'logical' but because you can't have people just going around willy nilly deciding they should be able to play things the game designers subjectively don't like and the game mechanics should harshly reflect these tastes just like how getting into a fight is a mechanically unsound option in Dogs in the Vineyard (which to be clear is a game about playing armed thugs that travel the land to enforce religious orthodoxy).

That's pretty much a recipe for disaster imo.  I think that 4e with it's 30+ classes and what amounts to 5+  different versions of the Fighter demonstrates why this design philosophy is bad.

Maybe I'm not following.  Backgrounds and specialties have nothing to do with the class. A knight fighter isn't any worse at being a fighter than a thief fighter is, they both get the same class attributes that comes with being a fighter.  They just get different base skills.

Sacro has the right of it on Themes I think. They are more like Kits than sub-classes. Meaning you create a new swashbuckler theme and you can apply it to Rogues, fighters even Clerics (Aramis was a clergyman after all) and their base abilties don't change just the trappings.
The important bit is to avoid adding themes that make genuine mechanical issues. This happened with Kits in 2e and then optimisation takes over from roleplaying and background.

As for multi-classing in general I tend to take a modified Skills and Powers approach. Skills and powers itself was a good idea poorly executed.
Players shoudl never be allowed to design classes. However as they progress I ahve no issues with a fighter trying to learn magic and rather than have them 'get a level in Wizard' I far prefer them to sacrifice some of their figther abilities in order to get some casting. This way it is possible but probably sub-optimal so you do it for roleplay in game reasons not for min-max reasons.
You can kind of do this in S&P or at least a version of it.
So your fighter cna not increase his Thaco for 3 levels and in return he gets the ability to cast 1st level spell a day. He still has to find a spell and go through the usual rigmarole he is still a fighter, just one that had diversified.
I find this much preferable to take a level as a MU so you can use the wand you found in that dragon horde.

Multi-classing has been horribly and terribly broken since inception. Themes seem to "fix" the issue with better core design while allowing players to mimic the multi-class option.

I tend to agree with Ladybird on this one.  Multi-classing rules in a class-based game are a sign of deep ambiguity in the design goals.  There shouldn't be rules for multi-classing; there should be rules for creating new classes as needed for the setting.

Quote from: Sacrosanct;581675Maybe I'm not following.  Backgrounds and specialties have nothing to do with the class. A knight fighter isn't any worse at being a fighter than a thief fighter is, they both get the same class attributes that comes with being a fighter.  They just get different base skills.

I was mostly responding to RPGPundit in saying that a Multiclass character should receive mechanical demerits for the purpose of protecting archetypal classes.

The same logic would dictate that anything that is not represented by a core class and it's explicit theme, such as the Thief class member that is a spy or assassin rather than an actual thief should likewise be discouraged through mechanical penalties.  That this should be done so that the player that wants to play the archetypal Thief as an actual burglar and mugger gains a mechanical benefit over non-compliant gamers.

If the goal is to protect the archetype then Backgrounds/Specialties don't accomplish this goal, since they give you the ability to create non-archetypal members of any class.  

I personally don't think that the archetypes actually need protection (or even representation) since most of them are absurdly broad to start with.  "The Fighter" is 'nothing', as so is "The Thief".  I can steal shit as any class and any member of any class that steals is a thief.  Thief status is not dictated by selecting the Thief class, and no mechanical benefits should be bestowed upon player's that select the Thief class that also use their class abilities to rob and murder.  

The classes simply reflect a set of abilities, and have semi-arbitrary labels.  Most settings will not have hard coded social roles for members of every class as if they reflect a harsh caste system where being a member of "The Fighter" class means something socially.

It's impossible for me to not see original Dragonlance Magic-Users, or the various Forgotten Realms catastrophe's to allow for edition changes as not being absurdly gamist (setting justification for mechanical limitations of classes, et cetera) and I don't see mechanical 'archetype protection' being something outside of this vein.

I don't really want Wizards to be what they were in original Dragonlance, where you are basically handed a character complete with a mandatory background, religion, and even wardrobe and gave you all of the customization of choosing a name and hair color because that was lame as fuck.

I don't want "The Fighter" to 'mean' anything or be 'something'.  I don't want a player to try to make a Joan of Arc that has spontaneous, inexplicable talent and be told that they will receive mechanical demerits because she isn't an archetype represented by a core class, and that they should be penalized because they are playing an unauthorized concept.

I think that is an absolutely disastrous philosophy for anyone that wants to sell a TRPG (and I think it was a disaster with WotC with 4e, which already reflects this design philosophy).

Quote from: Wolf, Richard;581717I was mostly responding to RPGPundit in saying that a Multiclass character should receive mechanical demerits for the purpose of protecting archetypal classes.

The same logic would dictate that anything that is not represented by a core class and it's explicit theme, such as the Thief class member that is a spy or assassin rather than an actual thief should likewise be discouraged through mechanical penalties.  That this should be done so that the player that wants to play the archetypal Thief as an actual burglar and mugger gains a mechanical benefit over non-compliant gamers.

The way I'm understanding the rules in 5e, there are no mechanical bonuses or demerits for having an "out of place" (for lack of a better term) background or specialty.  Your class is your class, and it's what you do best.  I.e., fight, cast spells, skill monkey (thief), etc and is pretty broad.  A fighter with a knight background isn't any better at being a core fighter than a fighter with a thief background.  It just means they start with different background skills.  And if you have, say, a high elf (which gets a racial bonus of being able to cast minor arcane magic (read: cantrips) and choose a fighter class, having that character have an arcane dabbler specialty (the most notable benefit being able to get a familiar at level 3), I don't think you should be penalized in your core fighting skills because you didn't choose the archer specialty.  Both characters are equally good at taking and dealing out damage, the only difference is the archer gains extra bonuses and abilities with a bow, while the aforementioned elf gets a pet and cantrips.

This works for me because in an in-game setting, I can easily see an elf warrior who has a bit of arcane skill, just as easily as I can see a warrior who is master at the bow.  I have a hard time seeing how a character reverse engineers their dual-wielding bow demon firing cleric to fit the game setting.

I didn't interpret Pundit as advocating a penalty in all cases, but rather to have a way to mitigate otherwise natural class profession to move to a different, unrelated profession just to min/max your build.  The way 5e seems to be built, it seems to have done that.

QuoteI personally don't think that the archetypes actually need protection (or even representation) since most of them are absurdly broad to start with.  "The Fighter" is 'nothing', as so is "The Thief".  I can steal shit as any class and any member of any class that steals is a thief.  Thief status is not dictated by selecting the Thief class, and no mechanical benefits should be bestowed upon player's that select the Thief class that also use their class abilities to rob and murder.  

I think we're on the same page here.

Quote from: daniel_ream;581699I tend to agree with Ladybird on this one.  Multi-classing rules in a class-based game are a sign of deep ambiguity in the design goals.  There shouldn't be rules for multi-classing; there should be rules for creating new classes as needed for the setting.

I wouldn't go as far as to say there should be class generator type mechanics but I believe that feats should mean something. I believe that classes should be somewhat general (only protecting a certain theme) but allow for say a Druid who can handle a sword (transform into a monkey and wield like three at once). There should be a swordsman who has a knack for conjuring flames. What's more is the classes should be flexible enough that both mechanically and background wise you can have a number of people pick the same class but be very distinct from each other (again both mechanically and otherwise).

If you're going to have a system mutable enough to just generate any class you want anyway you might as well get rid of the whole idea of having classes at all.

Quote from: Ladybird;581654I don't think that is an argument in favour of "we must have multi-class rules in our class-based system", as much as "we probably shouldn't be using a class-based system in the first place if it doesn't match what we want from our game".

Class-based systems do what they do really well, open-development systems do what they do really well, but they rarely mix well.

Class/skill systems exist and work decently well. They just don't really feel similar to D&D style classes.

You have it right Sacrosanct at least mostly because it seems like a variation of Fantasy Craft's take and 2e kits. Could be wrong here but that's what I'm getting as a takeaway.