Its going to be very difficult to retrofit the old style of multi-classing to PF2 IMO. And before you begin, you need to outline what your goals are.
Are you wanting to support a more 50/50 split?
Are you wanting to support the 'late in life, change of class'? (or the early in live one level start, before change).
Are you wanting to support starting as a multi-class (w/o leaning on ancient elf)?
Are you wanting to support faster tertiary/quad classing?
Pathfinder 2e classes are far more front-loaded than 1e. I suggest you first decide what level 1 features a character gets when they take a level in a different class.
Obviously they'll need to gain training in class DC for the new class. I would suggest leaving out skills, saves and perception entirely, since those can be obtained easily through feats. But there still remains weapon/armor proficiencies that are harder to get, as well level one class features. For example, fighters get attack of opportunity at level 1, an EXTREMELY useful ability that's hard to come by. If you decide to give it to characters who multiclass into fighter, you'll make fighter-dipping the best option for a huge number of martial builds, just like in Pathfinder 1e.
I also recommend you look into how class levels stack with each other. In 1e, they simply don't, and therefore a character that's wizard 10/fighter 10 has neither the spell levels and DCs to reliably cast nor the BAB to reliably hit. 2e characters have +level to everything trained and above, so even without stacking levels in different classes, their abilities would remain functional if we just look at bonuses and DCs. However, they would lack high level class feats and features like high level spells and weapon damage bonus from specialization. How competitive would this be with a level 20 wizard with a fighter archetype, or vice versa?
Yeah that's why I'm asking here as its honestly much easier when you can bounce ideas and get insight on something you aren't sure about.
There are some things that I believe should had been given to everyone (AoO is one of them) so I personally dont mind making it easier. But getting abilities at the right time, level bonuses and proficiency are very complex subjects that's require multiple people (or a lot of time) to truly make sense of the effects regardless what is choosen.
There's a really good reason why AoOs are less common this edition. The developers want a game that is more movement and positioning focused, so that characters are expected to move around the battlefield rather than stay in place and full attack every single round.
If we're talking verisimilitude, we're going to be here all day. Pathfinder is a turn-based game, which means verisimilitude will always have to give way to tactical play and balance.
Another issue I've just noticed in terms of initial proficiencies: all classes increase all their initial proficiencies at different levels. That means when deciding what initial proficiencies are ommited by multiclassing into each class, you also need to decide how class features that increase them later work.
I wanted (preferably) to keep the system as it was, while eliminating some of the more obvious traps (like 50/50 multiclassing) and rebalancing it to avoid a lot (but not all) of the front loading.
Honestly I think that keeping the system as it was, and eliminating the problems with multiclassing being a trap option are mutually exclusive.
Though, assuming it can be done, I think the following things would have to happen;
Proficiency bonus (specifically the level part) needs to be based on character level and not class level.
Ability score increases need to be removed from the class tables and place in their own separate table based on character level instead of class level.
Multi-classing spellcasters need to have their spell slots and what levels of spells they learn be based on a composite of all of their spellcasting classes (for example, a wizard 2/sorcerer 1 needs to have 3 1st level slots and 2 2nd level slots just like a wizard 3 would have).
I would also recommend a clause where multi-classing requires you to take your class levels in blocks of 2 (or multiples of 2) at a time (so you can go wizard, wizard, sorcerer, sorcerer, but not wizard, wizard, sorcerer, wizard) to make sure players are still getting the normal amount of feats.
Those are the four really big traps of old-style multiclassing (in pf1 it was base attack bonus, but that is rolled into proficiency bonus).
Basing proficiency on character level would be complicated since every class increases different proficiencies at different times. I would recommend keeping them separate. If you start trained in one thing from you first class, and multiclass into a second class that grants you trained again, you stay trained. If you multiclass too many times, you might end up being trained in a bunch of stuff that will never increase to expert. Consider if these characters lacking behind by 2, 4, or even 6 points is reasonable; if not, you might want to add a clause that after a certain level, all characters become expert in every non-skill ability they're trained in,
Again I heavily recommend you leave skills out of initial proficiencies when taking a new class. Saves and perception should be included on second thought.
Stacking levels in all spellcasting classes is absolutely necessary for multiclass casters to function. You'll have to decide how the spell slots are going to work. All prepared or all spontaneous is easy (just let them prepare any spell from any class into any level-appropriate slot, or cast any spell known with any level-appropriate slot). But if you're say, wizard 5/sorcerer 2, how do the spell slots work? Do you have fixed numbers of prepared vs spontaneous slots? Or can you treat any slot as prepared or spontaneous? Can you sacrifice a prepared slot to cast a spontaneous spell?
Multiclassing in 2-level chunks doesn't seem quite necessary as many class features are as strong as class feats if not more -- a new proficiency tier in weapons or armor, or a new level of spells, for instance.
You should divorce skill increases and skill feats from class tables, as well as general and ancestry feats. And ability score increases too as mentioned before. Any time you take a level of rogue you get an extra skill feat if your new level is odd, and a skill increase if your new level is even.
Basing proficiency on character level would be complicated since every class increases different proficiencies at different times. I would recommend keeping them separate. If you start trained in one thing from you first class, and multiclass into a second class that grants you trained again, you stay trained. If you multiclass too many times, you might end up being trained in a bunch of stuff that will never increase to expert. Consider if these characters lacking behind by 2, 4, or even 6 points is reasonable; if not, you might want to add a clause that after a certain level, all characters become expert in every non-skill ability they're trained in,
Again I heavily recommend you leave skills out of initial proficiencies when taking a new class. Saves and perception should be included on second thought.
Stacking levels in all spellcasting classes is absolutely necessary for multiclass casters to function. You'll have to decide how the spell slots are going to work. All prepared or all spontaneous is easy (just let them prepare any spell from any class into any level-appropriate slot, or cast any spell known with any level-appropriate slot). But if you're say, wizard 5/sorcerer 2, how do the spell slots work? Do you have fixed numbers of prepared vs spontaneous slots? Or can you treat any slot as prepared or spontaneous? Can you sacrifice a prepared slot to cast a spontaneous spell?
Multiclassing in 2-level chunks doesn't seem quite necessary as many class features are as strong as class feats if not more -- a new proficiency tier in weapons or armor, or a new level of spells, for instance.
You should divorce skill increases and skill feats from class tables, as well as general and ancestry feats. And ability score increases too as mentioned before. Any time you take a level of rogue you get an extra skill feat if your new level is odd, and a skill increase if your new level is even.
With proficiency I specifically mentioned it being the part of the equation that references your level in a class that needs to be specified as character level (so every level 20 character gets the +20 part of proficiency) but not the gaining of trained/expert/etc.
Its going to be very difficult to retrofit the old style of multi-classing to PF2 IMO. And before you begin, you need to outline what your goals are.
Are you wanting to support a more 50/50 split?
Are you wanting to support the 'late in life, change of class'? (or the early in live one level start, before change).
Are you wanting to support starting as a multi-class (w/o leaning on ancient elf)?
Are you wanting to support faster tertiary/quad classing?I'm not the OP, but the first thing I would want is:
Giving the player the power to control which weapon, armor and save proficiencies his character gets.
That's 'create your own class' not multiclassing though. Just about any attempt to to go class-less, or build-a-class via tweaking multiclass rules (either PF1 style of PF2 style) is going to be either extremely unbalanced, or extremely non-satisfying.
Skills I dont think interact with regular multiclassing, except for initial proficiency: Which I feel should work but not increase proficiency, much like making a skill into a class skill twice had no effect.
What I mean is that multiclassing doesn't grant you new trained skills at all, and skill progression follows total character level, not class level. No matter what your class combination is, you always have a skill feat at odd levels and a skill increase at even levels. Whenever you take a level of rogue, count the skill increase at odd levels and skill feat at even levels to be a rogue class feature instead of part of the regular skill advancement.
Maybe that's exactly what you're saying, I'm just clarifying my intent.
Yeah spellcasting should probably be based on character level (simpler) or combined spellcasting level (future proof for 6th lv casting). I would probably increase the save according to character lv and not mix spell slots or give higher level spells; The reason being that the new save mechanics actually does help lower level spells compared to PF1. Then introduce a multiclass spellcasting feat such as (I know it's a tax, but still better than PF1):
* Combined Spellcasting: Choose one spellcasting class you have. Increase the number of spell slots and spell levels of this class by the combined level of all other spellcasting classes you possess -1. Ex: A Wizard 4/Sorcerer 4 that chose Wizard would be treated as being Wizard 7/Sorcerer 4 for determining spell level and spell slots.
I wouldn't combine spell casting style since that seems something for Theurges and Arcanists.
Avoid feat taxes altogether. Stack total spellcaster levels for spell slots, but limit the highest level of spells known to individual class levels. If you end up with higher level slots than the highest level spell you know, you can use them to prepare heightened lower level spells, or cast heightened spontaneous signature spells.
I would also suggest you ONLY divorce the 3 spells per level progression from class. Wizard bonus school spells, cleric channel energy, and even the sorcerer's extra spontaneous slot per level are class features.
A wizard 7/sorcerer 4 will have the same number of "universal" spell slots as a level 11 wizard, that is, 3 spell slots each for spell levels 1 to 5, and 2 level 6 spell slots. In addition, they can prepare bonus school spells as a level 7 wizard, and has a bonus spontaneous slot each for level 1 and level 2 spells that can only be used to cast sorcerer spells.
The extra spells at low levels is intentional to balance out the inability to cast higher level spells.
I don't think arcanist casting is appropriate either. I recommend the following for mixed spontaneous/prepared casters:
You must choose during daily preparations whether or not to prepare a spell into each spell slot. If you prepare a spell into a slot, it becomes a prepared slot with which you can only cast the spell you prepared into it. Otherwise, it becomes a spontaneous spell slot and you can only use it to cast a spell from your spell repertoire. Bonus spontaneous slots gained from sorcerer levels can only be used to cast sorcerer spells.
Proficiency increases (outside of skills) are such a pain to balance given how no class has the same progression. I'm really tempted to just turn it into tracks with fractional increases (like fractional BAB and Saves). Effectively, a class that improves a proficiency at 6th level has 1/6 while one that increases it at 8th has 1/8; so a lv 3/lv 4 character with those classes would have 3/6+4/8=1 and increase the proficiency at lv 7.
That's extremely complicated and still unlikely to be universal.
Just don't let them stack, period. I would just add a clause such that whenever you get to a level where one of your classes improves weapon/armor proficiencies, ALL martial weapon and armor proficiencies gained from all of your classes improve to the same degree, and advanced weapon proficiencies improves to one step below unless that class feature specifically improves advanced weapons. For example, a champion 1/wizard 11 gains wizard weapon expertise, which gives you expert in only wizard weapons. However, since you are trained in all simple and martial weapons from your level in champion, you can increase all simple and martial weapons to expert instead.
You don't need to do the same for saves or perception, since saves that lag behind too much can be caught up with Canny Acumen.
The best way to implement traditional multiclassing without creating a bunch of traps is to start by divorcing a lot of mechanics from class level.
Ability score improvements, general feats, skill increases, skill feats and probably proficiency should scale independently of your class level, otherwise you're going into some really problematic scenarios where people multiclass their own stats to death and become terrible at everything.
Then you probably wanna take a note from 5e and curate a specific list of abilities that you gain when you multiclass into a class as opposed to when you take it at level 1. This helps address dipping, although it creates a new issue where a wizard 1/champion 1 is radically different than a champion 1/wizard 1, which might seem kind of weird.
@Frogliachi you are saying not to give the initial trained skill, and to give level up skills as when a class gives it to you.
I'm saying, do give the initial trained skill (not stacking) to mimic class skills in PF1. But now I think following the 5e rule of "you may choose a different skill" would work okay.
The alternative is to remove the initial trained skills from all classes and instead bring back class skills but only adding a +1 to a few skills. This would then allow multiclassing to add to the list without breaking anything (in my opinion).
If you're only giving the initial trained skill (eg Acrobatics and Athletics for fighter, Crafting for alchemist, etc), but not the x + int mod additional skills, then it's fine, as long as the new trained skills don't stack. I wouldn't allow a character to be trained in an additional skill of their choice either, because multiclassing already grants you tons more proficiencies than normal.
As I said in my first post, with stacking weapon/armor/save proficiency as you get them doesn't really work. A Fighter 10/Barbarian 10 ends up at Master of weapons at lv 15 2 levels after Barbarian, master reflex 4 lvs after Fighter, stays at trained proficiency (which is worse than a Wizard); And how you level up has a huge effect.
Not letting increases stack in some way would make that character unplayable getting at lv 15: Trained AC, Master Fort, Expert Reflex, and expert weapons: The character has literally worse martial stats than a full class Wizard.
I suggest the following then:
- If any of your classes start Expert in a save or in perception, and increases to Master or above, you gain Canny Acumen for that save or for perception if your total levels would not allow you to reach Master in it before level 17
- Whenever one of your class levels increases a weapon proficiency, it increases all your trained weapon proficiencies (except for advanced weapons, which stays one proficiency level below). If you don't progress to expert or above before level 11, all your trained weapon proficiencies (barring advanced weapons) increases to expert at level 11.
- Whenever one of your class levels increases an armor proficiency, it increases all your trained armor proficiencies. If you don't progress to expert or above before level 13, all your trained armor proficiencies increases to expert at level 13.