Multiclassing, Part 4: History of the Multiclasses (5th edition)

Last time, I bitched and moaned about how 4e (that is, D&D’s 4th Edition) eviscerated my beloved multiclassing.  To be fair, some people were overjoyed at this move, primarily because they believed (and possibly still believe) that multiclassing is only for powergamers.  But I saw it as a crushing blow to my ability to be creative during character building.  If you don’t want to play one of the dozen or so classes we’ve predefined for you, that decision said to me, then we don’t want you to play at all.  So I didn’t.

And so Pathfinder came along, and I already talked about that in conjuction with my discussion of multiclassing in 3e, because it wasn’t significantly different from what 3e offered.  And I was perfectly happy with Pathfinder ... honestly, in many ways I still am, especially in terms of character building.  But, eventually, D&D came back to answer the upstart Pathfinder (which, many sources claimed, was outselling the game that had spawned it) with a fifth edition, which of course we call “5e.”

Now, the vast majority of what makes 5e stand out above Pathfinder is in the gameplay: faster combat resolution, simpler rules for many systems, the elimination of dozens and dozens of little bonuses all over your character sheet, saner numbers even at higher play levels, and so forth.  In terms of character creation, 5e can’t really compete with Pathfinder, and I suppose it doesn’t really aim to.  But let’s give them this: they learned from the debacle of 4e in many ways, and (for purposes of this discussion at least) one of the most important of those ways was the restoration of proper multiclassing.

Which is not to say that they put back all the crazy flexibility of 3e’s multiclassing, because they had learned from that as well.  They wanted something which gave the same amount of flexibility, but allowed for less abuse.  Again, we’re not yet talking about whether that’s a necessary thing or a good idea or any of that.  Let’s just look at the rules.

Multiclassing in 5e is a fascinating collection of rules that make it simultaneously very attractive and remarkably inconvenient.  Presumably this paradox exists so that they could both satisfy the people like me who think multiclassing is a necessary part of character customization, and the people who think that multiclassing only exists for powergamers to try to gain every power in the game.  This leads to a series of “yay!” moments which are promptly counterbalanced by “aww!” momeents.  So, for instance, you can take a level in any class when you level up, just like in 3e (yay!), but you can only multiclass if you have a 13 in the “prime ability scores” of both the class you’re coming from and the class you’re switching to (aww!).¹  Your proficiency bonus is based on your character level, and experience point penalties for multicass characters are gone (yay!), but now you don’t receive all the starting proficiencies of your second (or more) class (aww!).  Best of all, the horror of trying to be a multiclass spellcaster is utterly corrected: all spellcasting classes have a single table for spells per level, and all your spellcasting classes are added together to determine how many spells you can cast.²  On top of that, you can now cast a low-level spell in a high-level slot and have it actually be much cooler than using its original low-level slot, which completely eliminates the problem of multiclassed spellcasters only having pointless low-level spells and no way to use them on the high-level enemies they’re more likely to be facing.  Yay!  But worst of all: ability score increases are no longer tied to your character level.  They instead come as class features.  The smaller benefit of this is that some classes (fighter and rogue, specifically) can have more ability score increases than others.  The larger “benefit,” of course, is that multiclass can easily fall behind in ability score increases, unless they take at least 4 levels at a time in each class.  Aww.

(We could also note that 5e, like Pathfinder, adopts the concept of “capstones”: powerful abilities gained at 20th level in a class, making them simply unavailable to multiclassed characters.  However, as I mentioned when discussing multiclassing in 3e, nobody ever gets to 20th level anyway, so this is rarely relevant in a real-world game.  Still, it’s a carrot rather than a stick, so I generally approve of it as a practice.)

Now, some of these new rules are fine—better than fine, even.  The fact that taking your first level of fighter after already having several levels in another class doesn’t automatically make you proficient at wearing heavy armor actually makes sense, from a thematic standpoint, and it’s a perfectly reasonable mechanical restriction.³  Even missing out on the skill and saving throw proficiencies that you normally get at Fighter 1 can easily be explained by not having the extensive training that most apprentice fighters get, and it solves a lot of the issues that 3e had with dipping.⁴  As a restriction, limiting proficiencies that multiclass characters get for their first level in another class makes a lot of sense.

The fact that multiclass spellcasters are somewhat “fixed” from how badly they lagged behind in 3e is pretty awesome too.  Sure, as a sorcerer 5/bard 5 you don’t have access to 5th level spells (nor even 4th level spells), but you do have access to 5th level spell slots.  And there are plenty of 1st-through-3rd level spells (for both sorcerer and bard) that will do cool things if cast in a 5th-level slot: burning hands will do an awesome 7d6 of fire damage in a 5th-level slot, charm person will charm up to 5 people with a 5th-level investment, or just do a 5th-level healing word for 5d4 + Cha points of healing goodness for your compatriot, and you don’t even need to touch them.  And those are just your 1st level spell options; many of your 2nd and 3rd level spells have similar scaling properties.  And your cantrips?  Those now scale by character level, meaning even spellcasters multiclassing with non-spellcasting classes can get something out of them.

But the corresponding downside I have mixed feelings about.  Every class gets an ability score increase at level 4, meaning that a strict alternation of levels is not always the best way to go.  For instance, a sorcerer 1/bard 1 is fine, and a sorcerer 2/bard 2 is okay, but a sorcerer 3/bard 3 is just stupid.  And then when you get to sorcerer 4/bard 4, you get two ability score increases in a row, which is weird.  You’re really much better off going single-classed up to sorcerer 4, then sorcerer 4/bard 1 and so forth until you hit sorcerer 4/bard 4.  And, even then, there’s a weird breakpoint at 5th level, where classes change from “tier 1 play” to “tier 2 play”: martial classes get their extra attack, spellcasters hit 3rd level spells, etc.  So you can make a strong argument that even sorcerer 4/bard 4 is inferior to sorcerer 5/bard 3 (or vice versa).  And I see what they’re going for here: make it painful to jump around a lot, use the progression itself to strongly encourage sticking with one class as long as possible.  And it’s clever design, and I really respect that.  But what I don’t like is how it forces you into these artificial channels.  If you believe multiclassing is only for powergamers, this is brilliant: it’s like a carrot that weirdly turns into a stick when you try to turn up your nose at it.  On the other hand, if you believe multiclassing helps you build more interesting characters, being punished for strictly alternating between two different classes, which is as close to a smooth progression as you can come and therefore the most realistic expression of a “blended” class, just seems bizarre and unnecessary.

And then you have the ability score requirements.  This is just generally bad.  First of all, it’s a blatant stick, which I don’t care for.  Thematically, it’s on shaky ground; the Player’s Handbook says:

Without the full training that a beginning character receives, you must be a quick study in your new class, having a natural aptitude that is reflected by higher-than-average ability scores.

Which almost makes sense, if you squint just right ... but only for the second class.  Except that it also applies to the first class, which is not explained by this logic at all.  I don’t need to have a 13 wisdom to be a cleric in the first place; why do I need it to study anything other than divine magic?  But the worst problem is with something that D&D players often call “multiple attribute depedency,” or just plain “MAD.”  Certain classes are considered MAD all by themselves: monks, for instance, need strength for their attacks, dexterity for their acrobatic and defensive maneuvers, and wisdom for many of their powers.  Many multiclass combinations are inherently self-limiting because of MAD issues.  For instance, you may find people railing on the Internet about how “broken” certain combos are, like sorcerer/monk or barbarian/wizard, but the truth is, no powergamer in their right minds would build such a character, because you need to maximize 4 out of the 6 ability scores to do them right, and, if all you care about is having the best numbers on your character sheet, there are far better ways to go about it.  So being MAD is already a problem for some single-classed character concepts; forcing an artificial MAD-ness by demanding high ability scores in order to multiclass just exacerbates the problem.  And the hardest combinations to achieve by these restrictions—paladin/ranger and paladin/monk, both of which require four 13s—are generally considered to be two of the worst possible combinations by hardcore optimizers anyway.  So who exactly are we punishing here?  The only people who would even want to play a paladin/monk are going to be the players who have a crazy vision and have already said “who cares that my character will be suboptimal?”  We want to discourage those players from achieving their vision?

So 5e makes a huge amount of progress, but perhaps still doesn’t completely hit the mark that I’d like to see.  It’s one of the best expressions of the multiclassing concept that we’ve seen yet, but it still has a few warts and wrinkles that I think can be improved on.


Next time, we’ll step back from the history of what has been done, and start looking and what we want to see in the future.

__________

1 For the most part, the “prime ability score” for a class is exactly what you expect it to be: intelligence for wizards, wisdom for clerics, etc.  Fighters get to pick either strength or dexterity, making it the easiest candidate for multiclassing, which is probably sane.  The more “gonzo” classes—that is, those which are already a bit like multiclassing compacted into a single class—have two prime ability scores (those would be monk, ranger, and paladin).  Also, “prime ability score” is not what they call it in 5e, but that’s what it was traditionally called, going all the way back to first edition.

2 With the half-casters only contributing one every other level, and the spell-dipping archetypes (like eldritch knight and arcane trickster) adding only one every three levels.

3 You might still magically become proficient at medium armor, of course, but that’s okay.

4 Both single-dipping and double-dipping.  Of course, to be fair, 5e has different issues with dipping—dipping two levels of warlock or moon druid, or one level of rogues, for instance, can still make powergamers lick their lips.  But there are fewer issues overall.