Multiclassing: it wears me out just thinking about all the twists this can cause for game mechanics, play balance, and immersion. Even in the early editions of Dungeons & Dragons there was a process for acquiring multiple classes. Older versions of the game have pretty brutal restrictions on multiclassing. You’d have to be a certain race to access classes, only certain class combinations were allowed, you’d suffer XP penalties, you’d need very high core attributes, you could only ever switch once (and you couldn’t switch back!), etc. I’ve already started tackling the issue of multiclassing with Prestige Classes in an earlier post, and I suppose the mere existence of Prestige Classes suggests a pretty free-form multiclassing system. Still, it’s worth thinking about again to make sure I’m happy with the way it’ll work and to make sure I consider the potential for abuse in multiclassing before I start really designing the various classes and races.
I think, in a role-playing game like Dungeons & Dragons, a class is The Role that you’re taking on, more than any other aspect of a character. It is a major defining thing. To begin the game as a cleric, fighter, wizard, or thief, it’s assumed you’ve had years of practice in your class to acquire the level of skill that comes with playing a hero, even a 1st level hero. With that in mind, it doesn’t make a lot of sense for a character to switch classes much, if at all, as they level – especially to classes that are far removed from their background training.
On the other hand, liberal multiclass rules can provide a great deal of customization options for a character. Maybe you had in mind a con-man turned religious, so you take two levels of rogue and then multiclass to cleric. One could even see gaining religious powers spontaneously without years of study in the mysteries of the religious order – since the actual powers could simply be granted by divine fiat.
When you get to this level of customization, the Optimization Problem begins to rear its ugly head again. If you can freely multiclass, you could find synergies between classes that produce unexpected results. The nice thing about multiclassing though, from a play balance perspective, is that it generally doesn’t provide more power to a character – instead it usually gives a character a wider array of choices. When you multiclass, you’re trading off gaining higher levels in a class to start over and gain powers in a different class that were balanced for 1st level. Powers across classes may syngergize well, but you often have to trade higher-level powers in order to get them. In fact, multiclassing can severely set a character back on the power curve. Take an old favorite, the Wizard/Fighter multiclass. As a wizard, you have to sacrifice a level of spell-casting to gain one level of combat prowess? By the time you’re a fifth level character, that extra level of combat prowess isn’t exactly making you much more effective on the front lines of combat – but you’ve delayed your access to third level spells, meaning while you do have a few more hit points and an extra +1 bonus to hit, you do NOT have a fireball at your disposal.
A nice solution to the problem for the Pathfinder game is to take the class archetypes, created to provide different build options within classes, and hang the multiclass mechanic on these archetypes. That way, instead of gaining powers from the archetype, you would gain some class features of another class. This strikes me as an elegant solution for Pathfinder. For Lost Worlds though, I’ve kind of done away with class archetypes already. This nice thing about this solution for Pathfinder is that you don’t have to sacrifice anything from your standard class in order to multiclass. You still get all the normal class features that you’d get anyway – instead, you sacrifice the bonus powers you’d get from your particular class archetype for bonus powers resembling the core powers of another class.
Another thing that 3e and Pathfinder don’t handle especially well without custom rules (although this can be handled with the archetype multiclassing strategy described above) is starting at first level with training in two classes. Maybe you were raised as a warrior, but then apprenticed as a wizard. How can you start the game with powers from both classes? In 3e, you just can’t. In older editions, you could actually start as multiple classes, but experience gained would be shared between them and you’d level up more slowly than your counterparts. Not really an elegant solution in my opinion.
I might be going back to the well here, but I can’t help but think that the solution here is pretty much the same as the solution to feats and prestige classes, and the optimization problem in general. If a person wants to play a ‘spell-wielding’ fighter, say we call it a “sword mage” – let’s create a class for that theme and work to balance it against the other classes. Maybe it’s helpful to have a similar class built already, so that it’s easier to do an adequate job of play balance, but if the system accepts easily constructed custom classes, then you can really handle any kind of theme a character might want – you don’t have to build the theme from pieces that may or may-not really fit. For example, if you’re really wanting to play a samurai, should you have to do that by taking classes in both monk and fighter? Why not just have a class that represents it, work with the GM to agree on some thematic powers, and if something ends up being too powerful, or not powerful enough, tweak it as you play? The main barrier to this is trust, but I think if the player can’t trust the GM, or the GM can’t trust the player, you’re likely to have a failing game anyway. If time is an issue, there’s no need to build out the whole class ahead of time, you only need to build the levels you need now.
So here are my thoughts on multiclassing. I dislike the idea of freely multiclassing between classes from a thematic standpoint. The question is: why are you wanting to multiclass? Is it because there isn’t a core class that meets your theme, or because there’s a mechanical benefit to doing so, or because your character has consciously decided to switch careers entirely? If you want a different theme, let’s come up with a class that meets your thematic vision. I think borrowing powers from core classes would work well here if desired, and I wouldn’t have a major issue with a character that decided from the outset to trade-off levels between two classes for a thematic purpose (in fact, depending on how the classes are built, doing so would likely end up in a somewhat weaker character, but one with a wider array of options). If you are dead-set on multiclassing your character from the get-go, take a character creation feat that allows you to level up as in Pathfinder or 3.5 (any class you want each level). I’d still encourage you to limit the number of classes you’re taking to two or three though!
If it’s because there is a mechanical benefit for doing so, then I don’t really feel the need to support it with the game rules – I’m not really interested in having people take from the buffet of classes (some fighter and ranger with a side of cleric) simply because it takes advantage of powers that synergize well (or were unevenly granted to a class at an early level with an expected sacrifice at a later level) and step them ahead in the power curve over the other players.
If it’s because your character has consciously decided to switch classes, then I think a mechanical answer provided by the game system is in order. The answer may be as simple as a “career switch” feat. I don’t want the tax for switching to be so severe that you can’t keep up, but I do want the decision to switch classes to be meaningful. So, in order to multiclass, you’d need to spend something like a ‘feat’ to allow your character the amazing ability to have mastered a second set of apprenticeship skills, all while adventuring. While I’ll want to design the first level of classes to be thematic in themselves, I’ll also want to design them so that “cherry-picking” levels of classes doesn’t become an obvious mechanical benefit. This means that, even though level 1 of a class needs to set the framework for the class, it can’t bundle in so much more at the outset than any of the other levels of the class that players are going to want to “grab a level” of various classes for the mechanical benefit. I don’t think this will be as hard as it would be for 3e or pathfinder, because I’m going to have a lot more levels early on with which to provide thematic class features.
The trouble with all of this is that prestige classes have to be somehow balanced with the loss of your core class, meaning early levels of prestige classes must be significantly better than early levels of core classes. This means that prestige classes almost HAVE to have a prerequisite other than simply role-playing because otherwise, accessing it too early or too late could mean the class doesn’t work in a balanced way with someone sticking to a core class. 4e handles this by having prestige classes gained at a specific level (10th). I’ll have to readdress this or re-think a bit how I plan to handle prestige classes to balance with a core class all the way through. Also, I’d want to encourage players to switch from a core class to a prestige class if they found one that worked well for them in the course of the story (as a story reward), so I’d need a solution for that. As a rough solution now, I’d say that if your prestige class falls into the same “category” as your original class (ex: martial, arcane, divine, stealth), you can begin leveling freely into a prestige class one time. After that, you can switch back to your old class for free, but to pick up a new class or prestige class would require the investment of a feat.