How important is multi-classing, and why?

[Crazy Jerome]

Is multi-classing something that needs to be included, and if so how flexible or powerful should it be?

Is it important in world-building/setting terms? (That is, should something like a fighter/wizard be a valid combination to imply how a character developed over time, versus some hybrid class that is effectively the same thing mechanically.) Or is the only purpose of multi-classing to keep the number of classes down to a managable number while still allowing a lot of character concept variety?

Do you prefer 3E or 4E versions of multiclassing, or perhaps something similar to AD&D multiclassing but cleaned up mechanically? (That is, structure the classes such that a Fighter 8 or Wizard 8 or Fighter 6/Wizard 6 was roughly equal and obtainable, probably via independent XP by class.) Or none of the above and/or something new?

What other concerns should be addressed in multiclassing?

 

[Ahnehnois]

3e's multiclassing system is one of its great triumphs.

The freedom to be a fighter 1/rogue 6/assassin 5 really allows you to make the character you want, as opposed to being shorhorned into 2e's fighter/thief, 4e's rogue with a few fighter abilities, or the even less flexible systems of other games.

The only problem was the math. Saves were screwy, and casters got screwed. That's fixable. It's important that multiclass characters be balanced.

That said, after having received the ability to make the character I want, change paths midway, and look at my stats and see class levels that truly describe my character, anything less flexible would be unacceptable.

 

[delericho]

Interesting. Do you have Fighter, Wizard, and Fighter/Wizard; or do you have Fighter, Wizard, and Duskblade?

My preference would be to have few, broad, easily customisable classes, matched up with a fairly easy multiclassing system. Something a good bit more flexible than the 4e model, but certainly not the mess than was 3e at times.

The problem with trying to have a distinct class for each individual concept is that you very rapidly need a lot of classes. Plus, people will want to blend the ingredients to different extents - is your character a FIGHTER/wizard, a fighter/WIZARD, or half and half?

 

[Umbran]

Is multi-classing something that needs to be included, and if so how flexible or powerful should it be?

it seems to me that how multiclassing should be handled depends on how classing is handled.

Multiclassing is a tool that allows you to build things not available in a specific class. If they intend to have tons of classes, they have less need for multiclassing. If they intend to go with a few very broad classes, multiclassing becomes a far more important tool.

Both approaches can work, thought the former approach requires either a long line of splatbooks, or class-construction tools in the GM's hands.

 

[Crazy Jerome]

Well, if you want fairly robust and expansive multiclassing, how much narrowing of the classes would you find acceptable to make it work, if necessary, and at what cost?

Or put another way, the more broad and capable you make any single class, the more likely it becomes something people want to play as a single class. But this also complicates multiclassing, because of stacking (either too much, or because restrictions to reign in the stacking become too draconian at times -- i.e. 3E caster levels at certain points in the design). Make the single classes fairly narrow, and you instantly remove a good chunk of the multiclassing issues, yet make the single class option somewhat 1-dimensional as a character. (Picture a "wizard" who only casts arcane spells and gets no weapons or boosts to attacks or saving throws and so forth, or a "fighter" who only knows weapons and armor--for a very extreme example of what this might be.)

So I see the single-class/multi-class dilemma as the proper path to thread between two extremes. I'm fairly convinced the extremes should be avoided (though not certain), but how much to veer either way, I change my mind almost daily at times.

 

[DEFCON 1]

The freedom to be a fighter 1/rogue 6/assassin 5 really allows you to make the character you want...

The question I would ask you in this case, is which part is "creating the character you want"? The abilities you gained as a result of those three classes, or the fact you have the name of the class as part of your class description? (Or both?)

In other words, was that one level of Fighter in there because you wanted the weapon and armor proficiencies and extra hit points that level provided you... or because you wanted to be able to say your character was partially a 'fighter'? (Or are both parts important?)

Reason I ask is to wonder if having different/better weapons/armor/skills to possibly take under the purview of "rogue" or "assassin" meant you wouldn't NEED to have a multi-class system? If you could build an assassin that could wear chainmail armor (for example), is having multi-classing still important anymore?

 

[Stalker0]

I like many things about 4e, but multiclassing was not one of them.

I feel for the developers on this one. 3e's multiclassing is wonderful for the players. It gives a monstrous amount of freedom and customization...and all with a minimal amount of rules. But everything that makes it so free and wonderful makes it a NIGHTMARE for balance.

A great number of broken builds in 3e (both over and under powered) can be traced back to multiclassing.



To answer the OPs question, if nothing else, good multiclassing is necessary because it allows the game to feel complete out of the box. One of my biggest issues with 4e was the game felt very reigned in when I first started playing. There were a number of archetypes that I did not feel I could adequately play in core 4e. 3e's edition multiclassing for all its faults never had this problem because I could mix and match to get exactly what I wanted.

Now I will say that my issue with 4e's style changed with time. Because 4e tried to put each archetype into its own class box, as time when on more classes and builds were produced, and slowly my options expanded to cover all of my needs. However, that took more time and more books....and many people don't want to wait around.

So for 5e, I either need a robust multiclassing system, or a whole lot of archetype options given to me right at the beginning.

 

[Ahnehnois]

The question I would ask you in this case, is which part is "creating the character you want"? The abilities you gained as a result of those three classes, or the fact you have the name of the class as part of your class description? (Or both?)

Both.

A fighter 5/rogue 3/assassin 4 is a different character. So is a fighter 1/rogue 10/assassin 1. Just looking at that heading on top of the statblock gives me a great deal fo information about that character's history and goals; it's probably the second most important thing next to the six ability scores.

They're also very different in play. The more roguish, the more likely the character uses noncombat skills to get what he wants. The more fightery, the more he becomes overtly involved in melee.

If I could, for example, take 1 level of rogue but get my evasion and sneak attack through feats after several fighter levels, there really isn't much point in having classes at all. So I say, if you're going to have classes, highly flexible multiclassing is a must.

 

[mkill]

I don't think any edition of D&D has found the right multiclassing formula yet. It's very hard to balance power and flexibility. 3rd edition multiclassing came to the point that 3rd level fighters were ridiculed. It rewarded single-level dips and created headaches with caster level (I never understood why they didn't just set character level = caster level).
4th edition had the much-too-restricted PHB multiclassing and the overcomplicated and wonky hybrids.

I'd have to see the basic class design first o form any opinion on 5th ed. In any case it will be a nightmare that will cost the designers many sleepless nights. But it's very important to get it right from the PHB on.

 

[tecnowraith]

When 4e was announced I thought they use 2e multiclassing paths and reinvent it for 4e. But instead they used feats system for multiclassing which to failed IMO. I'm wondering could something like that, mix the old with the new for 5e?