Design Diary 11

"I want to play a rakshasa fighter/rogue/wizard/Assassin!"

(No, I haven't actually heard that said, thank god - I made it up.)

As I said, I've been working on a fix for multiclassing, and I think I've got a workable system now. I've decided to use the fractional BAB/save progressions, which work quite well and are very short:

Base Attack Bonus: Adding a level in a class with a good BAB (barbarian, fighter, ranger, paladin) adds +1; a level in a class with a medium BAB (bard, cleric, druid, monk, rogue) adds +0.66; a level in a class with a poor BAB (sorcerer, wizard) adds +0.5.

Fractions are rounded down when calculating BAB, though you still keep track of them. For example, a Rog 1/Ftr 1 has a BAB of +1.66 (rounded down to +1). If he gained a second level in rogue, it would become +2.33, or +2.

For saves, I made some modifications. I did an analysis of saves a few months back and came to the conclusion that the poor save was too low, and boosting it to 40% of level would make it more balanced. So, with that in mind…

Saves: Adding a level in a second class adds +0.5 to the class' good save(s) and +0.4 to the class's poor save(s). Exception: If the new class' high save(s) are different from the existing class's, add a one-time +2 bonus to the high save. For example, a Rog 1 would have Fort +0.4 (+0), Ref +2 (+2), Will +0.4 (+0). Adding a level of fighter would give him a +2 bonus to his Fort save and +0.4 to his Ref and Will saves, so: Fort +2.4 (+2), Ref +2.4 (+2), Will +0.8 (+0). If he took a level in ranger for his third character level, he wouldn't gain a +2 bonus to either Fort or Reflex, since he's already taken levels in classes with those as good saves.

This mainly fixes the problem of taking levels in multiple classes with the same saves, where the good save keeps stacking and the low saves never advance - Rog/SD, for instance, Wiz/Clr, or Ftr/Bbn. In all three cases, the character with 1 level in each class would have one save at +4 and the other two at +0; a L10/L10 PC would have +14 in the good save (which is 2 higher than a straight L20) and +6 in the low one (or +8, with the 40% low save). By this rule, the same PC (at L2) would have +2 in the good save and +0 in the poor one; a L20, he'd have +12 and +8.

It also fixes the potential abuse of dipping into multiple classes - a Clr 1/Ftr 3/Rog 3, for instance, would have base saves of Fort +6, Ref +4, Will +4. By the variant rule, it would be Fort +4.7 (+4), Ref +5.0, Will +4.4 (+4). (As an aside, the thing that finally convinced me to go to 40% for low saves was the fact that I had to add 0.33 for each low save using the fractional system; I had to use a calculator once the decimals got over 1, because they were so odd. Try adding 2 + 1.5 + 0.99 in your head to get the Fort save; I'm good with numbers and I could do it, but it would require too much thought and ends up with strange numbers. Adding multiples of 4, though - that's easy. 2 + 1.5 + 1.2 is a snap.)



Several folks on ENWorld (myself included) have put forth fixes for multiclassing spellcasters - that is, a way for spellcasters to multiclass without being hamstrung, power-wise. As most of you already know, a character with 10 levels in two spellcasting classes (say, cleric and wizard) is not nearly as powerful as one with a straight 20 levels in one or the other - they don't have the caster level to penetrate SR, their spells generally aren't powerful enough to do enough damage (those based on caster level), and their saves are pretty wimpy (did I mention I changed spell DCs to 10 + 1/2 caster level + spell level?), all of which means they can't hold their own in a party of characters of their level.

The solution I've come up with is based on the UA magic rating system. Initially, I had a system laid out based on this one, then after I tried out the fractional BAB/saves, I decided to apply it to magic rating. I took the UA table, added a 3/4 column for bards and adepts, and pushed everything further to the right (so half casters use column C and noncasters use column D).

What I ended up with is this:

All character classes have a magic rating, which increases by level much like base attack bonus. For a multiclass character, add up the character’s magic ratings from each of her classes to find the character’s total magic rating. This is applied to the character's effective caster level (the level at which she casts spells. Next, add the magic rating from the column to the right for each class to find the actual caster level (the level at which she gains spells). Class levels in the last column aren't counted toward actual caster level.

The following restrictions apply, however:

  • Only the highest bonus is used.
  • A character's effective caster level in any spellcasting class cannot be increased to more than three times the base value. For example, a Clr 15/Wiz 2 would normally gain 15 levels from his cleric side (so he would cast spells as a Wiz 17), but by this rule, he is limited to Wiz 6.
  • A magic rating gained from a class can't be added to that class - in this case, you use the next higher rating for the highest-level class. For example, a Clr 10/Wiz 8 would add 8 levels to the cleric, not 10.
  • A character's actual caster level in any spellcasting class cannot be increased to more than double the base value. For example, the Clr 15/Wiz 2 from above would gain two effective caster levels added to his cleric class and 0 actual levels; he would add 4 effective caster levels to his wizard class, and 2 of those levels would be actual levels. So, in effect, he'd be a Clr 17/Wiz 4 with access to 7th level cleric spells and 2nd level wizard spells (as a Wiz 4).
  • Each full-caster class (cleric, druid, sorcerer, wizard) beyond the first applies a -1 penalty to all levels applied (after said levels are determined). For example, the Clr 10/Wiz 8 would actually add 7 effective levels and 5 actual levels to the cleric, and 9 effective and 4 actual to the wizard. He would be treated as a Clr 17/Wiz 13 with access to 7th level cleric spells and 6th level wizard spells.

For example, a Wiz 6/Rog 4 has a +6 rating from her wizard levels and +1 from her rogue levels. The +1 is applied to her effective caster level - she is treated as a 7th-level caster for determining the range, duration, and other effects of her spells. For her actual level, her wizard levels are checked on the 3/4 column - so she would have +4.5 (+4, rounded down). Since the rogue is on the last column, it isn't counted, so she is treated as a Wiz 7, but gains access to spells as a Wiz 6 - she'd have 4/3/3/2.



As for ECLs/LAs… well. After playing around with some things, I found that you can't really use CRs in conjunction with ECLs without including LAs - they're like the a translator between the two. Think of CRs like pounds (the weight, not the monetary unit) and ECL like kilograms - they measure the same thing (weight) but use a different standard; LA is the conversion factor. Sort of. Anyway, v6 of Upper Krust's Challenging Challenge Ratings has revised rules for ECLs and their use, so I'm thinking of going to a straight ECL system - instead of CRs, all monsters will be rated by ECLs. This will be similar to the 4E level system, but not identical - ECL will simply take the place of CR in the statblock, and monsters won't be labelled as "Level xxx". I think this will make PC-ifying monsters a lot easier, as well as figuring out appropriate encounter levels. Unfortunately, v6 will be part of Godsend, UK's last 3.5 book, and he said he's got a few last little bits to do, then send it off to his editor, so I'm figuring somewhere in the neighborhood of 6-8 weeks. Which is all right, I suppose; I can still do conversion work on the monsters (skills, SLAs, abilities, etc.), and then go back and do the ECL work later.

Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-ShareAlike 3.0 License