How important is multi-classing, and why?

[Crazy Jerome]

My biggest problem with mutli-classing is that it is susceptible - especially of the 3e variety - of being exceptionally gamist. It becomes about picking up particular benefits to min-max the character. Sure you are building your character, but why should multi-classing be necessary? Conversely, the problem with a rigid system like 4e is that the class and multiclassing is not necessarily flexible enough to create the character you want. But is the problem though with multiclassing or is it with the way that classes and character options (i.e. skills, feats, etc.) are designed?

I think the problem is both with multiclassing and class/character options, in both editions, then coupled with the hard limits of what people are willing to manage. It seems to me that multiclassing, even more than most D&D mechanics, tends to pick up lots of little fiddly exceptions, interactions, unintended synergies, etc. Even 4E, as sharply limited as it is, is overly fiddly for what it delivers. So that raises the question in my mind if there can be a form of multiclassing that delivers more benefits than it costs. (At least to a widespread audience. Of course some people don't mind particular cost/benefit trades in a given version.)

 

[Celebrim]

Forget multiclassing.

Think outside the box and let people build their own classes.

Sure have some basic classes laid out in a kit based fashion, but you don't need multiclassing when each player gets to build their own class.

Just my thought.

In practice, a qausi-point buy system like this tends to defeat the purpose of having classes. You tend to end up with a lot of abuse of the system to create hyper narrow characters for the sake of mechanical advantage.

 

[Celebrim]

Single-class characters should allow some flexibility for differentiation. 3ed had the perfect system at hand: Feats.

I wouldn't go so far as to call it perfect, but yeah, I'm a big big fan of feats.

In my personal experience, multiclassing is not really needed to create a character concept: if you have enough flexibility in feats/class features and also in skills and spells, YOU CAN create every character concept you can think about, eventually with the help of a make-up of flavor.

I disagree. Lets say a player wants to play a ninja like character (with or without oriental flavor) and in the players mind being a ninja means not just merely normal roguish ability but some sort of supernatural ability as well. Either I have to create a specific class that captures 'ninja' or I have to have a multiclassing system that lets you effectively marry a generic roguish class with a generic innately magical class to end up with a 'ki wielding master of stealth and evasion'. The big problem with going the 'ninja' route is that you tend to end up carrying too much flavor and fluff bagage into the class implementation, and then it becomes a poor fit for the elven pickpocket who supplments his mundane slight of hand with magical tricks. And if you end up adding another 'arcane trickster' class to cover that concept, you now not only have lots and lots of classes, but probably have mechanical diversity in that each class has slightly different mechanics for trying to accomplish a lot of the same things.

Suppose you have a character who is the martial champion of a god of song. Either you need a specific class for this deity and this role - 3e's Prestige Class - or you need a really good way to marry a generic paladin class with a generic bard class. And so forth.

I don't choose characters primarily based on the game role I intend to have in the party. I don't look at RPG's primarily from the focus of winning a minatures/exploration/tactical subgame. That subgame may be an important part of the game, but when designing a character I start with a concept and then look around to see what I can fit to it with the character construction toolkit that the system provides.

Third edition multiclassing is not perfect, in the sense that when it would be most useful its least available - character creation. But it is one of the most powerful character development systems ever implemented, and 4e suffers from its lack.

They want to be good fighters AND wizards at the same time. I am quite old-fashioned and I think a roleplay game is a game of roles, so pick one at a time.

Even granting you are old fashioned, 'a good fighter and a good wizard at the same time' is one of the most old fashioned roles and tropes in the book. It is the BD&D 'Elf' class. And the very fact that the race was a class is I think a fairly strong illustration for what I explained above for why multi-classing just works better in the long run. Talk about carrying too much specific fluff into a class implementation.

Of course 5e should cater to everyone, so multiclassing will certainly be included. Also, after all it's better to have a single decent multiclassing rule than dozens of additional classes which essentially offer nothing new but a mix of the core classes. Both of these for me are inferior design choices than having a small bunch of core classes with good flexibility.

Ok, nm then. I guess we do agree after all.

I've seen it all the time in 3ed players writing up multiclass PCs (except beginners and spellcasters) and then complain that the game didn't offer good reasons to stay single class. Ban multiclassing from the PHB and you have your reason

This is I think do to some poor specific implementation details and choices rather than a general problem with multiclassing.

 

[Aldarc]

I don't choose characters primarily based on the game role I intend to have in the party. I don't look at RPG's primarily from the focus of winning a minatures/exploration/tactical subgame. That subgame may be an important part of the game, but when designing a character I start with a concept and then look around to see what I can fit to it with the character construction toolkit that the system provides.

Then why do you need classes for that? How about I just hand you the LEGO pieces you need to create your character rather than force you to take apart different LEGO sets to make what you want?

 

[Olgar Shiverstone]

You need multiclassing, at least as an optional rule-module, because it provides players more flexibility to realize their character concepts than the designers will be able to provide by single class concepts -- even if some of the single class concepts are more elegant than a multi-class solution.

Edit: And let's face it, though this can be accomplished more elegantly by a building block system rather than classes, D&D has always been a class-based system, and if WotC cooks that particular sacred cow they'll never put the fan base back together with 5E. So we're left with multiclassing, feats-as-multiclass, or dedicated single classes as solutions.

 

[Celebrim]

Then why do you need classes for that? How about I just hand you the LEGO pieces you need to create your character than force you to take apart different LEGO sets to make what you want?

Good question.

Two reasons.

First, not every player has the same primary goals I have. Typically when you put the LEGO peices in the hands of a gamer who is focused on winning that subgame, you get a character which is basically a well honed weapon and only a well honed weapon. Every superflous peice is stripped off the character sheet in order to get the optimally honed high performance character you can have. Now, there is nothing inherently wrong with that, but what it tends to do is make groups with mixed gaming styles more incompatible than they would be if you had them work within the constraints of a somewhat flexible set of peices - a 'class' if you would. One of the great advantages of D&D in my experience is that it supports not just a variaty of styles of play, but a variaty of styles of play coexisting at the same table.

When you have a class you are ideally ensuring that the resulting character has a bit of breadth to it. You can attach of minor abilities to the class which rarely come up but which make any character that is a member of the class more well rounded. Enforcing versitility is hard to do in a point buy framework. You are also ensuring a bit of archetypal power at the level of starting character. You can have concepts built right into the game. You'll note that even point buy systems like M&M go out of there way to make an initial presentation of the system in terms of archetypes - the speedster, the vigilantte, the brick, the paragon, etc. And you also speed up character creation a bit. Really in depth point buy systems can take hours to put a single character together. That's not such a big deal from the stand point of the player, but its a huge advantage from the standpoint of the DM.

Secondly, point buy systems are notoriously hard to balance. Brokenness is usually around every corner. With a class based system that has a reasonably small number of classes (say under 20), it's a lot easier to play test and tweak the game until they all are roughly balanced. Also, in practice, point buy doesn't necessarily ensure more diversity than a class based system precisely because of the lack of balance. Pretty quickly in a point buy system there is a gravitation towards a small number of 'best builds'. This is kind of equivalent of the feat tax issues you sometimes see in 3e and especially 4e. Everyone has to do certain things to make the build work well, and so you end up with relatively few points to actually differentiate your character from someone else.

I've been both places and depending on the game I might go with point buy or classes. But classes are the right choice for D&D, and frankly even if a good point buy system could be devised despite the tradeoffs, the game wouldn't feel much like D&D if it didn't have them.

 

[Crazy Jerome]

In support of Celebrim's points, it should also be noted that more than a few Fantasy Hero groups over the years have managed campaigns by putting together packages of abilities that had to be bought together, and had prerequisites. In effect, they took the Lego blocks and built classes with it before the game started, and then stuck to those classes unless a major problem arose. (I never went that far with Fantasy Hero, but I approached it a couple of times.)

I don't know how much it changed in GURPS 4th ed., but in GURPS 3rd, in the default game some pieces remained individual blocks while others had defacto ties. In particular, the default GURPS spells came in groups with prerequisites.

The advantage of this approach is that if you know the system well, you'll get probably very close to the exact classes you want. And if you don't get it right the first time, since you put them together, you also know how to take them back apart and fix them. The disadvantages should be obvious, but note that they are far worse for people who aren't used to thinking in terms of building/modeling things.

Now, of course one way you could square that circle theoretically is to have the designers go the Lego route, not release that into the wild at first, and build the initial stuff they do release out of the pieces. Later, they then release the underlying framework as a supplement, and it is assumed as part of the design that you never go straight to the Lego pieces, but must first use them to build something acceptable which can then be used. That might even work. My biggest reservation for that is that Hero and GURPS have spent a long time honing there systems, yet they are still niche.

 

[Desdichado]

It is vitally important, because to a lot of players, classes are just mechanics bundles, not archetypes. Being able to multiclass enables the creation of a lot of characters that otherwise can't be created. It allows--as long as it's not terribly clunky or crocked, both problems which have frequently plagued the notion in the past--for gamers who would otherwise be happier playing a more point-buy type system to make peace with the notion of having to use classes in the first place.

 

[Wormwood]

I don't see how it could be accomplished using a rigid class system, but man do I love Skyrim where my greataxe-wielding orc in heavy dragonplate has two signature spells that he's practiced for specific situations.

Feats or themes could work to emulate that, as could a mechanism where one could swap class features.

 

[Desdichado]

Then why do you need classes for that? How about I just hand you the LEGO pieces you need to create your character rather than force you to take apart different LEGO sets to make what you want?

Because many D&D players like classes. Multiclassing is a necessity, because it's the only way to bring class-likers and point-buy-likers together under the same tent.