What are your multiclassing house rules?

[Afrodyte]

What house rules do you have for multiclassing? What prompted you to use them? What effect have they had on your games?

 

[Celebrim]

What house rules do you have for multiclassing? What prompted you to use them? What effect have they had on your games?

a) PrC's do count against favored class limitations. This was an early attempt to proactively prevent a player from taking multiple PrC's, which to me violated the entire concept of PrC's (to the extent that they could be justified at all). This is made somewhat null and mute by the fact I have no PrC's, having banned them within months based on my experience running weekly open dungeon crawls at a local gaming store.

b) Each class has a minimum ability score entry requirement - for example, you must have strength 9 to be fighter. In order to multi-class, you must meat all classes minimum entry score requires. Each class you take beyond the first increases the minimum entry requirements by 2. In practice this tends to limit you to 3 classes without severely straining your resources. For example, a fighter/wizard needs only have an 11 strength and an 11 intelligence, which is likely of anyone taking such a combination. But a fighter/fanatic/wizard/rogue would need a 15 strength, 15 intelligence, 15 dexterity, and 13 constitution. This at the very least insures that multiple attribute dependency limits how much you can abuse taking multiple classes for their front loaded benefits. It also means that 5 classes is probably impossible to pull off, eliminating the 'class ala cart' problem seen in standard 3.X.

c) Lacking PrC's, I have a unique way to deal with the problem of multi-classing spellcasters (via a feat chain). It's more elegant than the typical 3.x method of having a PrC for each possible combination of classes, but on the other hand is somewhat more abusable. It tends to make multi-classing quite attractive, particularly 'dips' of 4 levels or so into either a non-spellcasting class (if you are primarily a spellcaster) or a spellcasting class (if you are primarily not a spellcaster). On the whole, it hasn't been a problem in terms of balance, but if you are optimizing its definitely an attractive option.

 

[DammitVictor]

Spent years trying to "fix" multiclassing in 3.5 and PF, and eventually settled on a Gestalt-with-LA system, I found... mostly tolerable.

I think I've got a handle on it for 5e now, but I've heard rumors they're going to have an alternate multiclassing method in the DMG and I'm waiting for that before I go on too much about it.

 

[Sadrik]

In 3e I had a feat that allowed you to add virtual class levels to your character equal to half the other classes levels (but no more than double). So if you were a 4/4 you would have class features of each class as a level 6/6 but in all other aspects would be a level 8 character. If you were 6/2 your class features would be counted at 7/4.

 

[nonsi256]

1. No favored classes.
2. No PrCs
3. No multiclassing restrictions.
4. Spellcasters advance spellcasting by advancing noncasting classes at a rate of 1/2.
5. Special feats allow you to also sum other spellcasters to the lot.

Haven't had a chance for playtesting, but you can find all the details and rationale here.
And the though exchange that has lead to them can be found here.

 

[Celebrim]

1. No favored classes.
3. No multiclassing restrictions.

We clearly have different goals here. I'm trying to make races much more differentiated. Likewise, I consider the fact that humans have relatively few multi-classing restrictions one of their main advantages, and if you remove it you more strongly move the game toward a non-human world.

2. No PrCs

On the other hand, we agree on this. Very few people inclined to write house rules for 3e disagree. The PrC stands out as the single biggest mistake in the 3e rules.

4. Spellcasters advance spellcasting by advancing noncasting classes at a rate of 1/2.
5. Special feats allow you to also sum other spellcasters to the lot.

I first read #4 and I was like, "What? No way. Clearly he hasn't play tested at all.", but then I read your more detailed explanation and was like, "Yeah, ok. That works." The rule actually reads:

"4. Spellcasters advance spellcasting by advancing noncasting classes at a rate of 1/2 to a maximum of 1/2 their class levels in the spellcasting class."

This does the same job and ends out working rather similarly to my system in terms of the actual caster levels you can actually obtain through multi-classing, except that in my case the sort of advantage gained in #4 itself comes to you by way of special feats. Your system on the other hand offers a bit more protection from primarily non-casters dipping a level or two of caster in order to gain spells, or primary casters dipping for a level or two of non-caster with no loss in spell casting ability. You also provide more support for low level gishing below 3rd level - ei, a Ftr2/Wiz1 or Ftr2/Wiz3 is probably more viable/balanced in your system than mine, though mine can demonstrably work and gets a much bigger boost than yours at Ftr2/Wiz4 when the additional general feat becomes available. Likewise, your Ftr4/Wiz16 is giving up more than mine (but in practice perhaps enough you'd never see it actually happen), and my Ftr16/Wiz4 is gaining more at less cost (again, yours gains so little that I'm not sure it would ever happen). I'm not sure how I feel about all that, as the problem in my system is mostly theoretical, the actual realized power level differences are small, and I'm hesitant to change in my system what isn't clearly broken. However, the fact that you are producing these results naturally to me suggests that your system is likely to be at least as robust when playtested as mine has been, and in general I'm less worried about your system getting busted in the edge cases than I am about mine getting busted. Your idea to introduce things as capped ratios seems solid.

 

[nonsi256]

We clearly have different goals here. I'm trying to make races much more differentiated. Likewise, I consider the fact that humans have relatively few multi-classing restrictions one of their main advantages, and if you remove it you more strongly move the game toward a non-human world.

I'm all for races making more difference, but one of my design goals was to minimize shoehorning races to classes as much as possible.
As for humans taking a back seat - I'm not worried about that one bit.
Human is already considered the top race, alongside Dwarf. In my settings, I've replaced skill-point bonus for +2 to ability score of choice (if you choose Int, the skill-point bonus is an add-on).

On the other hand, we agree on this. Very few people inclined to write house rules for 3e disagree. The PrC stands out as the single biggest mistake in the 3e rules.

Sure thing, but they never came up with a decent alternative that would enable them to throw PrCs out the window without losing character concept flexibility.

I first read #4 and I was like, "What? No way. Clearly he hasn't play tested at all.", but then I read your more detailed explanation and was like, "Yeah, ok. That works." The rule actually reads:

"4. Spellcasters advance spellcasting by advancing noncasting classes at a rate of 1/2 to a maximum of 1/2 their class levels in the spellcasting class."

Yes. That was a definite oops on my side.

This does the same job and ends out working rather similarly to my system in terms of the actual caster levels you can actually obtain through multi-classing, except that in my case the sort of advantage gained in #4 itself comes to you by way of special feats. Your system on the other hand offers a bit more protection from primarily non-casters dipping a level or two of caster in order to gain spells, or primary casters dipping for a level or two of non-caster with no loss in spell casting ability. You also provide more support for low level gishing below 3rd level - ei, a Ftr2/Wiz1 or Ftr2/Wiz3 is probably more viable/balanced in your system than mine, though mine can demonstrably work and gets a much bigger boost than yours at Ftr2/Wiz4 when the additional general feat becomes available. Likewise, your Ftr4/Wiz16 is giving up more than mine (but in practice perhaps enough you'd never see it actually happen), and my Ftr16/Wiz4 is gaining more at less cost (again, yours gains so little that I'm not sure it would ever happen). I'm not sure how I feel about all that, as the problem in my system is mostly theoretical, the actual realized power level differences are small, and I'm hesitant to change in my system what isn't clearly broken. However, the fact that you are producing these results naturally to me suggests that your system is likely to be at least as robust when playtested as mine has been, and in general I'm less worried about your system getting busted in the edge cases than I am about mine getting busted. Your idea to introduce things as capped ratios seems solid.

I've been busting my head on this for years.
I then ran into rule #4 somewhere and used it as a baseline.
It's not until the exchange on my GiantITP codex thread that I managed to build the layout around it and finally nail this one right.
I haven't play-tested this, but I did have others' feedback and their concerns expressed during the exchange, so I know we've managed to cook something solid.
I'm always keeping an open mind for changes, but for now I'm more than content with the result.

 

[Celebrim]

Sure thing, but they never came up with a decent alternative that would enable them to throw PrCs out the window without losing character concept flexibility.

PrCs ended up getting used for a large variety of disparate things.

1) Monte as he originally conceived them saw the PrC as a way to convey the flavor of major NPC driven organizations. The PrC here was a role-playing hook with mechanical rewards where the DM would introduce various elite, esoteric, or secret organizations with their own agendas in to the campaign and then at a certain point allow the player to join them. Thus, the PrC was a way to integrate the PC's into the campaign and provide direction to play. This idea may have some merit, but it was almost universally ignored, not only by DMs, but by WotC itself. Also in practice, the PrCs Monte created had the problem that every member of an organization was typically the same character. While in theory Monte intended to avoid this by making PrC's generic enough that they applied to many classes, in practice that never worked well.
2) They were an attempt to fix the multi-classing problems, particularly with multi-classing between casters and non-casters. This was one of several examples of WotC attempting to patch problems in the core rules. And for the longest time this was the one area that I felt a PrC was justified as an ugly but at least welcome kludge.
3) They were an attempt to patch the balance between casters and noncasters by giving noncasters more robust options. This however was almost completely undermined by the fact that the casters got even better PrCs.
4) They were an attempt to patch problems with the baggage carried by core classes to allow for choices that the core classes didn't allow for. For example, consider how the Blackgaurd is suggested by the fact that the Paladin must be good. What if you wanted a ranger that wasn't a spellcaster? Or a ranger that was suited to an urban environment? And so forth. I ran into this problem really early on with a need to make a Barbarian class that was Lawful rather than Chaotic. Of course, after having made the PrC (the only one I ever homebrewed), I immediately realized that the easier and more elegant solution was to strip unnecessary baggage like 'primitive wilderness warrior' and alignment restrictions from the notion of 'Barbarian'.
5) They were an attempt to patch balance problems with archetypes that were rather weak when implemented in core but were nonetheless important fantasy archetypes you'd expect a generic system to support.

Over time, I began to see all of these attempted solutions as not only ineffectually slapping a Band-Aid on the problem that actually needed curing - problems in the core balance and limitations in how much diversity the core could provide; but also as contributions to the problems of balance and narrow development options that they were designed to cure.

However, as bad as all that was, none of that was as bad as the core function that WotC began to see the PrC as fulfilling - marketing. Over time, most PrCs came to be justified as follows:

6) PrCs are attractive to players because they provide for power creep via at least specialization and sometimes strictly superior replacement of regular class levels. And they are attractive to developers because, as long as you don't have to play test them, they are very easy to create content. And they are attractive to marketing because the inclusion of PrCs allows a book to be marketed to both DM's and players when the content might otherwise only be interesting to or traditionally seen as the exclusive province of DMs.

WotC began to print PrCs that actively undermined its own game system. They were cheap to make as long as you didn't play test them (imagine, not play testing a class in a class based system), as witnessed by the hordes that appear in 3rd party supplements.

I think the decent alternatives are suggested by something like your multi-classing fix, plus...

a) Broader embracing of the feat and what a feat is allowed to do, particularly in a more 5e style. For me this realization occurred when I realized that every PrC was simply a feat chain generally geared to a particular class which had as its requirements the requirements of entering into the PrC.
b) General rebalancing of the core classes. For me a lot of that has involved a combination of thinking about class abilities in terms of 'feat equivalents' (if every feature of this class was bonus feats, how many feats would the class features be worth), toning down the power of spells/spell-casters while increasing the power of feats and skills (and those that rely on them).

 

[nonsi256]

Monte as he originally conceived them saw the PrC as a way to convey the flavor of major NPC driven organizations.

This is a bad approach, because RL experience has taught us that what drives organizations are interests (political / religious / economic / . . . )
Tying organizations to game mechanics is just asking for troubles (they just didn't know what kind of troubles in advance).

Also in practice, the PrCs Monte created had the problem that every member of an organization was typically the same character.

Later on, you mentioned feats.
A hard working DM could split feats into groups, where some are common knowledge and the rest are to be divided among elite-esoteric-secret organizations you mentioned.
The same approach could be applied to certain spells, skill-tricks, gear, substances etc.

I only wish I had the time and energy to implement this approach.
Thinking out dozens of organizations with their sets of feats / spells / skill-tricks is way too much work for me and I wouldn't know how to do it correctly without closing the door on a multitude of character concepts.

They were an attempt to patch the balance between casters and noncasters by giving noncasters more robust options.

Which I view as something that one should strive for from the get go.
I'm well aware that there's never 100% symmetry between the classes (no matter what those classes would be) and that's ok, as long as there's a healthy tradeoff between raw power and versatility.

They were an attempt to patch problems with the baggage carried by core classes to allow for choices that the core classes didn't allow for.

Actually, one of the more annoying excess weights for me was making Trapfinding a class feature. It's a baggage that goes all the way back to BECMI (didn't really examine Chainmail, so I couldn't go that far).
Using core only, it forces you to take Rogue at low levels out of game necessity rather that out of wanting the class for its awesomeness.
And, it makes no sense whatsoever.

What if you wanted a ranger that wasn't a spellcaster? Or a ranger that was suited to an urban environment?

I've been arguing for the longest time that Ranger and Barbarian should be Fighter variants or Fighter build-options.
If you think of it, in concept and spirit, Ranger and Barbarian are not so far apart.

WotC began to print PrCs that actively undermined its own game system. They were cheap to make as long as you didn't play test them (imagine, not play testing a class in a class based system), as witnessed by the hordes that appear in 3rd party supplements.

It's not just 3rd party supplements.
They didn't see the following monstrosities coming to life:
- Sublime-Ur-Lyrist
- Ur-Theurge
- Mindbending-Hellfire-Ur-Lock (via Eldritch Disciple)
- or even Incantatrix / IotSFV / Planar Shepherd

But yes, with a 5^th level entry point and a single Cleric level (with Able Learner), 3^rd party Dweomerkeeper is insane on its own. Just begging for completion with Archmage or Abjurant Champion

I think the decent alternatives are suggested by something like your multi-classing fix, plus...

a) Broader embracing of the feat and what a feat is allowed to do, particularly in a more 5e style. For me this realization occurred when I realized that every PrC was simply a feat chain generally geared to a particular class which had as its requirements the requirements of entering into the PrC.
b) General rebalancing of the core classes. For me a lot of that has involved a combination of thinking about class abilities in terms of 'feat equivalents' (if every feature of this class was bonus feats, how many feats would the class features be worth), toning down the power of spells/spell-casters while increasing the power of feats and skills (and those that rely on them).

I had all these criteria in mind when I set out to making the last incarnation of my codex, but I also realized that since...
1. feats are not (and could not be) made equal
2. feat-chains are not (and could not be) made equal
3. feat-combos are not (and could not be) made equal
then feats should not become the dominant factor or measuring tool for character power and versatility, and the assessment should be applied to whole packages.
Feats, however, should definitely remain a decisive customization tool, but I don't like the fact that you get so few of them in 5e and that you have to weigh them vs. advancement of ability scores.

 

[Quartz]

For 3E?

1. No Favoured Classes.
2. No restrictions.
3. Saves are level/2, round down. Classes & Prestige classes give a +2 class bonus (which does not stack).
4. Add levels on BAB tracks cumulatively (to L20) and add the three tracks in at the end to give actual BAB. So a Rogue 1 / Cleric 1 / Monk 1 / Fighter 2 has a BAB of +4.
5. Multi-classing is not only encouraged but expected.
6. The Epic Prowess feat grants +1 EAB.
7. Rework all feats that require levels in a class. E.g. Weapon Specialisation simply requires +4 BAB and the Item Creation feats require Spellcraft.
8. Skills for one of your classes are no longer cross-class even if you're taking a level in another class in which they are cross-class.