Sidestepping Multiclass Requests

[Username17]

Apparently the previous thread got swallowed by a change in Google Policy, so let's get back to it. Here are some facts that I find important:

• "Open Multiclassing" was an experiment in 3e that was interesting and exciting, but ultimately cannot be made to work. You simply cannot give out an ability without knowing what level you will be when you get it and expect that ability to function in a level appropriate fashion.
• Multiclassing at all is basically bad for the game. At its best it is a way to make playable classes that are procedurally generated and therefore take longer in chargen and probably aren't playtested as well as other classes.
• When there are explicit, functional classes that cover a character concept, they are massively and understandably more popular than multiclass setups that cover the same concept.
• Nonetheless, if you tell people that there isn't going to be "multiclassing", people will lose their shit. Even if they acknowledge all the previous points.

The goal then is to get people who want multiclassing to calm down without letting Multiclass rules ruin the game. That is difficult, but basically involves making multiclassing as tiny as possible in its real effects on the game while still being able to say with a straight face that "There is Multiclassing". As I see it, there are basically three ways to do that:

• Make actual classes. Announce that some of them are "multiclass" combinations. The Ranger is now a "Multiclass Fighter/Druid", the Paladin is now a "Multiclass Fighter/Cleric", and so on. The main class names are more popular than the multiclass combos, so you'd probably write "Paladin (Fighter/Cleric)". You would probably do something involving the "multiclass" characters being able to pick their skill lists off of the nominally parent classes or something to make it less obvious you were trolling people.
• Sub Classes. Announce that at some level or another you get a power upgrade that is chosen off a list that happens to be a list of themed upgrades that are themed off the available classes. And then you'd get an upgrade that was themed "Wizard" or "Assassin", and you'd announce that because you could get an Assassin Theme for your Samurai that you had "multiclassing" covered.
• Short available Multiclass list. In any edition of AD&D or OD&D only a small list of classes were actually multiclassable. You could be a Fighter/Thief, but you couldn't be a Cleric/Illusionist or Paladin/Thief. So in this model, you allow a small number of classes to be multiclassed, and then you playtest those combos exactly like they were a normal class. And you let people partially procedurally generate these classes so that players get the "joy" of having to make their own classes.

Right now, I'm gravitating towards the third option. For 10KF, the classes of Hero, Rogue, and Wizard actually mix fairly well with other classes because their in-combat abilities are all at-wills. And while Cleric is a cool-down mechanic and really doesn't mix well in abstract, it could probably be made to work with specifically the other three. And I think it is not implausible to call those the "base classes" considering that they were just that in AD&D and 2nd Edition AD&D. And the rule "you can only multiclass a base class" doesn't sound nearly as imperious as it actually is.

-Username17

[Avoraciopoctules]

Option 3 sounds pretty appealing to me. Option 1 reminds me of some aspects of 4e I rather disliked. Option 2 is something that could be nice, but makes it difficult to have somebody who is thematically balanced between their different power sources.

[Ice9]

Option #2 sounds the best in theory, but possibly a PITA in practice, since you have to make and test a whole other set of things.

I remember on the other thread, someone had an idea to call the distinction out in-setting, call the main classes "devotions" (gives you the awesome stuff, with the various resource systems) and the sub classes "professions" (gives you basic backup attacks and gap-filling abilities). So you could be a Necromancer / Adept (shoot bolts of fire when you're not animating the dead) or a Necromancer / Soldier (stab people when you're not animating the dead). That sounded pretty cool, but also a significant amount of work.

[hogarth]

[*] "Open Multiclassing" was an experiment in 3e that was interesting and exciting, but ultimately cannot be made to work. You simply cannot give out an ability without knowing what level you will be when you get it and expect that ability to function in a level appropriate fashion.

This is only a problem if your goal is to create a "dartboard RPG" (i.e., if you throw darts at the rulebook, then whatever options you hit are supposed to make a workable character). I think the number of RPGs that aim for that as a goal is quite small.

[Username17]

[*] "Open Multiclassing" was an experiment in 3e that was interesting and exciting, but ultimately cannot be made to work. You simply cannot give out an ability without knowing what level you will be when you get it and expect that ability to function in a level appropriate fashion.

This is only a problem if your goal is to create a "dartboard RPG" (i.e., if you throw darts at the rulebook, then whatever options you hit are supposed to make a workable character). I think the number of RPGs that aim for that as a goal is quite small.

No U.

Seriously, I can't even parse that statement. Are you saying that having 1st level Paladin Abilities being in no way a good tradeoff for not having the next level of Wizard spells when the next level of Wizard spells happen to contain the things that make you tall enough to meet the "you must be this tall" limits of the challenges of your level is somehow OK if you aren't drunk enough to want to play darts? If you're saying that, you're stupid. If you weren't saying that, your word salad needed more carrots.

-Username17

[Rawbeard]

I never quite understood the appeal of multiclassing. I also liked the halfassed attempt 4e made to make it not break everything everywhere, so... what do I know.

One problem I see with poeple losing their shit is classes being percieved as who you are, instead of what you can do, so while being a fighter/rogue/sorcerer is really fucking retarded, just picking a bard totally breaks the charakter concept. Or something.

Option 1 and 2 might be worth considering for me. 3 feels to much like a throwback and I feel like someone is already trying to reintroduce racial class lists. Dwarves finally managed to break free from being limited to 3 out of 4 most basic fantasy classes, don't let that happen again!

I'm going to take my meds now. I'll be fine, move along.

[echoVanguard]

If you weren't saying that, your word salad needed more carrots.

I believe what Hogarth is saying is that he doesn't believe it's necessary to support every combination because he asserts that people will pick their class combinations nonrandomly.

I personally don't think A follows from B at all, but that's what I got out of his post.

echo

[echoVanguard]

Incidentally, I'm personally in favor of Option 1, because that sells more books. But the other developers on the NA team don't agree with me, so we have limited multiclassing.

echo

[hogarth]

[*] "Open Multiclassing" was an experiment in 3e that was interesting and exciting, but ultimately cannot be made to work. You simply cannot give out an ability without knowing what level you will be when you get it and expect that ability to function in a level appropriate fashion.

This is only a problem if your goal is to create a "dartboard RPG" (i.e., if you throw darts at the rulebook, then whatever options you hit are supposed to make a workable character). I think the number of RPGs that aim for that as a goal is quite small.

No U.

Seriously, I can't even parse that statement.

And yet echoVanguard seems to have figured it out easily. Go figure.

It is not clear to me that banning all ability choices that don't "function in a level appropriate fashion" is a good thing. It certainly isn't true that every other system "cannot be made to work", you drama queen.

[codeGlaze]

Did I miss something? What's 10KF?

Aside from that, Option 3 had the least mental backlash from me as I read it.

Although, Option 1 played along better with how I break down classes in my head.

example:

	Source	Heavy (full plate/chain)	Medium (chain / leather)	Light (leather / cloth)
	Arcane	Barbarian (Rage is... arcane?)	Ninja (Shadow Rogue, etc)	Mage
	Divine	Paladin (Battle Cleric)	Cleric (Nerfed Cleric)	Priest (Caster Cleric)

edit: ...I don't know where that phantom row is coming from. >_>

[crasskris]

I actually liked the Legend (of www.ruleofcool.com heritage) concept of each class having three paths available, and each character being able to switch a path with that of another class. The concept, mind you, not necessarily the execution.

In my experience the main reason for multiclassing is gaining a single schtick that is 'abnormal' for the character. Taking wizard as an example base class, I've seen players take levels in ranger for the hiding/tracking/wilderness skills (for a wandering mage concept), levels in rogue for the stealth/streetwise/sleight-of-hand skills (for a underworld mage concept) and levels in (again) ranger for the double-wield/swording skill that comes in handy when emulating film Gandalf.

I think a lot of players would be happy to have a set of core abilities that then can be balanced against each other class in all relevant, complex core elements of game-play, and a set of secondary or flavor abilities that affect either only simple, secluded rules, or the complex ones in only cosmetic or trivial ways.

[TarkisFlux]

Frank - how does advancement of option 3 stack up with advancement of a single classed character? The other ones seem pretty straightforward, but it's not clear what sort of cost would be paid by the guy who multiclassed there that kept him in line with the single classed guys.

---

And because I'm curious what the response will be (and I didn't see it in the other thread), here's a potential 4th multiclassing model:

[*]The Late-Term Gestalt - This is largely the model put forth by AD&D 2 multiclassing, but updated for modern "level appropriate" sensibilities and doesn't lock you in at 1. You have some primary class and when you gain a level you can choose to gain another level in it or jump into another class. The benefits of jumping into a second class are scaled to your character level, to prevent weird nonsense, . So if you're a character of level X your options are to be a :Class A X; Class A (X-1) / Class B 1 (that grants you X/2 levels worth of power); or Class A (X-2) / Class B 2 (that grants you X-2 levels worth of power).

In such a setup, you never take more than 2 levels of a secondary class because your secondary class scales to your primary class automatically based on the 1 or 2 levels you put into it. But you also never fall more than 2 levels behind the level appropriate curve in exchange for branching out. It doesn't deal with self synergy at all though, which is a potentially large mark against it.

[Saxony]

I feel option 3 is a bit too complex. I believe the term is exception based design (You can do this, except with that, that and that, and this and that don't mix, while that and this do mix, except....)

It could be made simple and easy to understand, but my first impression is "too complex".

Option 1 seems like shit. I do not trust game designers to make the classes well. That is why I like 3.5's multiclassing.... so I can make whatever I want. But mostly to help out the mundane classes keep even in terms of power/versatility with the spellcasting classes (which, non-coincidentally I don't multiclass as much because they don't need it to be powerful and nuanced).

I like number 2. People understand a character described as "Main class is fighter, sub class is mind knight" and it sounds good. Allow me an attempt to restate the idea. It sounds like at pre-set intervals, characters will get to take a level in from the sub-class category as gestalt for free. I say an entire level, rather than one power, because a level implies multiple powers. Getting a single power from an alternate list every 5 levels is not what I'd call satisfying multiclassing.

[ishy]

I think multi-classing works for non-casters in 3.5 because they never get anything but low level abilities. Thus you can just as well pick up the abilities of another class.

But for a better game, I'd prefer multi-classing to work more like pathfinder archtypes (3.5 racial class options) / prestige classes. That way the same class can also feel different and you can bridge some gaps with prestige classes.
Edit - Note: prestige classes should obviously have no requirements other than class lvl x or something like that.

[mean_liar]

Option 2, for me. I don't like classes that much, feel all abilities should scale, whatever. If it's classes, the standard PC wargame/Langrisser/Westnoth/FFT "start basic, get funky later" system feels the best.

It's the same as DnD3e with forced PrCing, and I like it.

[OgreBattle]

I'd like to see if multiclassing could be done like mana in M:tG

That is, a certain ability would be "2 Fighter levels, 2 anything else levels"

So if you are Fighter2/Wizard2 you have access to that attack.

Specific Class level= Colored mana
Total class levels= colorless mana

Mono color/classed characters would be able to spend more mana to pump up their shivan dragon strike compared to a multiclassed character. That sort of balancing point.

This is all vague theory though and there are lots of differences between drawing from a deck of cards and having everything on a character sheet of course.

[Username17]

I'd like to see if multiclassing could be done like mana in M:tG

That is, a certain ability would be "2 Fighter levels, 2 anything else levels"

That's easy: it can't.

It's been tried. It really has. It even looks good at 2nd level, where you have the option of taking a second level of your first class and opening up the FF and 1F options or a first level in a second class and opening up the 1F and 1W options. I mean, that's kind of a balance nightmare, but it sounds workable.

But advance that to 10th level. Now, to distinguish each level of multiclassing you need 1FFFFFFFFF options and 4FFFFFF options and so on. Just to get to 10th level, you need 55 sets of options per class. If you want to take things to 20th level, that is two hundred and ten options per class. Even for five classes (RWBGU?), you're talking about having to write over a thousand class level options. And that's before we get into the absolute balance nightmare that a 4F level ability could be very plausibly mixed with abilities from four other classes at the same time. Before we get into the issue where mixing resource management systems is really hard, and not having multiple resource management systems is boring and stupid. Just when get to the "can we possibly write all this?" stage, the answer is "no".

It is not clear to me that banning all ability choices that don't "function in a level appropriate fashion" is a good thing. It certainly isn't true that every other system "cannot be made to work", you drama queen.

The open multiclassing system's entire argument for existing is "There are 20^11 ways to get to 20th level!" If you can't guaranty that more than half of those are actually playable, that argument can be rewritten as "We have thousands of trap options for every man, woman, and child on Earth, and thousands more trap options for every human who has ever existed in the entire history of the planet!" Which is not remotely a positive thing. And that's the argument in favor. The argument in favor is "more trap options than you could count in a thousand thousand lifetimes!" Which is insane.

And let's be honest here: claiming that even half those options are even a little bit viable is unmitigated horse shit. The truth is that the system has no way to ration out high level abilities to multiclass characters at all, meaning that once you get to high level, the entire system breaks down. Which is why of course, wizards don't multiclass in 3e. The only reason it even apparently works for anyone at any level is because as Ishy points out: Fighters don't get high level abilities at all anyway, so the fact that the multiclassing system would bone them out of their high level abilities doesn't hurt them any more than they've already been hurt.

-Username17

[norms29]

Make actual classes. Announce that some of them are "multiclass" combinations. The Ranger is now a "Multiclass Fighter/Druid", the Paladin is now a "Multiclass Fighter/Cleric", and so on. The main class names are more popular than the multiclass combos, so you'd probably write "Paladin (Fighter/Cleric)". You would probably do something involving the "multiclass" characters being able to pick their skill lists off of the nominally parent classes or something to make it less obvious you were trolling people.

not sure if it's a good thing, but this option has some tradition behind it. several of AD&D's first "kits" were essentially just redoing multiclassing in this format.

also this was how RoleMaster handled the concept, although "rolemaster did it this way" is more or less the exact opposite of praise, if the goal is to minimize trap options in character creation.

[Aryxbez]

[*] Sub Classes. Announce that at some level or another you get a power upgrade that is chosen off a list that happens to be a list of themed upgrades that are themed off the available classes. And then you'd get an upgrade that was themed "Wizard" or "Assassin", and you'd announce that because you could get an Assassin Theme for your Samurai that you had "multiclassing" covered.

First off, is 10KF, referring to Warhammer 40K, its Fantasy equivalent, or otherwise what is being referenced here?

The second option or "sub Classes" sounds what I'd be into, grabbing power(s) from another class, and matching them together for a cool combo. Though, I did like the idea of some suggestion ye had, about background powers, like being a farmer let you tame monsters or whatever (need to find that post again).

Some point, if someone doesn't beat me to it, I'll poll together what options were most preferred on here to say the least, though probably wait till it gets to 2 pages before doing so.

[tussock]

FACTS

OK to straw-man a little, many apologies:

[*]Per-level additive multiclassing is shitty for casters because the spell progressions weren't made additive in any way, you have to rely on low-level self-buffs and that's not what wins D&D. OTOH, it works so well for the Fighter/Rogue type classes that they had to nerf them all at low levels in 3.5 to stop people using it.

Suggestion: Building unique non-caster combos in 3.0 didn't suck, at low to mid levels. Still, go back to overlapping like AD&D, where multi-class casters and single-class casters get along quite well, and multi-class Thieves and Fighters don't suck at all.

[*]Per-level options are explosive in their complexity, which is harder on game design processes than most game designers realise. Even trying to copy a known balance point via per-level choices is really hard.

Suggestion: Overlapping choices, limited like AD&D to the basic classes in each group only, effectively gives you a tightly controllable number of total options to cover.

[*]People love classplosion. Having to work hard to make something happen is not what people want, explosive per-level permutations suck the fun out of character-building. A new class that does the work for you makes people happy.

But: Classplosion is very bad for the game. You can only support so many classes without annoying the fuck out of people buying your books to only find half a page or less for their favourite class. At some point, strictly limited building-blocks provide more options for everyone in a way you can actually support.

[*]People also like combining options, no matter how many of them you make. As much as many players hate the work needed to make complex combinations work, other people love working with combinations.

Thus: Simple combinations satisfy everyone. AD&D again.


Which is me supporting your option 3. Warrior, Mage, Priest, maybe Rogue if you can fill out the class a bit (or maybe you have to multi-class Rogue). It's what I've seen work.

Option 1 also works though. There's a lot of flavour and joy to be had with customised things like Druids, Paladins, Bards, and Beguilers. You then have a flavour-free Fighter-Cleric that people can do all sorts of character customisation on, but the specific oath-bound Paladin of Law and Good can also take prayers and stances like a Cleric and Fighter, because she is.

Option 2 also has some merit, allowing mid-level switching into mid-level classes totally fills up splatbooks without excluding people or closing off basic options (even though 3e didn't do it like that). Everyone with a touch of Fighter should be able to join any elite Fighter-like organisation, including Fighter-Mages, Paladins, Warlords, and so on, and instantly be cured of whatever bad build they walked into.

I guess 4e's idea of capstone epic classes (or Mentzer's Immortal Paths) also works. At some point the game stops working and you have to start over with 10 "mega hit points" and 1st level "mega spells" in a small selection of "epic path to immortality" classes. Fighting aspects of Orcus in tiny chambers of neutrality, along with anything else that can do actual mega-damage and force mega-saves.

[Sigil]

I think it'd be fairly easy to sell people on option 3, especially if you presented the freely multiclassable classes as 'base' classes, and the others as 'advanced' classes. "You may take levels in any number of base classes, but may only have levels in one advanced class." I may even uses this setup for my wip fantasy heartbreaker.

[hogarth]

It is not clear to me that banning all ability choices that don't "function in a level appropriate fashion" is a good thing. It certainly isn't true that every other system "cannot be made to work", you drama queen.

The open multiclassing system's entire argument for existing is "There are 20^11 ways to get to 20th level!" If you can't guaranty that more than half of those are actually playable, that argument can be rewritten as "We have thousands of trap options for every man, woman, and child on Earth, and thousands more trap options for every human who has ever existed in the entire history of the planet!" Which is not remotely a positive thing.

On the contrary. In my experience, avoiding traps is half the fun . Hell, a whopping chunk of D&D revolves around avoiding traps (whether that involves picking good spells or choosing the right equipment or figuring out the appropriate tactics for a fight or literally bypassing boobytraps)!

I'm not going to defend the sheer number of crappy choices available in D&D, of course. You're actually lowballing the number of possible multiclass choices -- it's 11^20, not 20^11 (although in practice many of those would be very similar since a fighter 1/wizard 1/monk 1 is not much different from a fighter 1/monk 1/wizard 1, for instance).

The other argument against banning poor choices is a simulationist (oops, I used a GNS term) one: in real life, some people make shitty choices, so banning that possibility limits the verisimilitude of your game.

[ishy]

Allow me to disagree Hogarth. This is a game, all options you can choose should be playable and possibly fun. It shouldn't be possible that I don't know the rules well and then create a character that can't do jack.

Note: Things like a wizard with sub 10 int, thus can't cast any spells is obviously not intended and not what I mean.
edit - Note 2: This is talking about static options. I'm 100% okay with certain actions being terrible with certain circumstances.

[Voss]

Wait, you're not going to defend the pile of shit choices in D&D by saying that... the pile of shit choices is part of what makes D&D enjoyable?

Wouldn't the game be better if the huge page count of pure shit was half or even a quarter of the size but not shit?

[RadiantPhoenix]

Allow me to disagree Hogarth. This is a game, all options you can choose should be playable and possibly fun. It shouldn't be possible that I don't know the rules well and then create a character that can't do jack.

Note: Things like a wizard with sub 10 int, thus can't cast any spells is obviously not intended and not what I mean.
edit - Note 2: This is talking about static options. I'm 100% okay with certain actions being terrible with certain circumstances.

To elaborate on this, a one-year D&D campaign where you play for eight hours once a week fifty weeks a year is a thousand hours. Assuming that people figure out their mistakes at a random point during the campaign, you're sentencing people to an average of five hundred hours of living with those mistakes when they're supposed to be playing a game and having fun.

This is not reasonable, so if you allow people to make crippling mistakes in chargen, you have to allow them to rebuild their characters when they realize their mistakes.

Unlimited rebuilds damages character consistency (i.e., establishing a character and following their adventures, as opposed to each adventure being about a different set of characters), so we should try to minimize the amount of rebuilding that is necessary.

Thus, we should minimize the number of trap options in the parts of characters that are supposed to be consistent, such as, in 3e D&D, class, race, feats, and skills.