D&D General Why is the multi-classing spell slot math so weird?

[Distracted DM]

In 3.5e, spells “upcast” automatically. Your caster level was a factor in the damage calculations, etc. Requiring you to spend a higher level spell slot to get the upscaled effect in 5e was a nerf, explicitly to tone address the “linear fighter, quadratic wizard” problem.

Yup! It probably took the nerf a little too far? Especially since cantrips were made so potent as to keep pace with martials' attacks... it seems a little counter to the mission statement.

 

[doctorbadwolf]

I tend to say that if you want to multiclass, you need to be able to accept the problems.

Not a great response to criticism of a system. Like...recognizing that something about multiclassing doesn't make sense doesn't indicate anything resembling an unwillingness to "accept the problems" of multiclassing.

 

[tetrasodium]

Yup! It probably took the nerf a little too far? Especially since cantrips were made so potent as to keep pace with martials' attacks... it seems a little counter to the mission statement.

Although I'd say itsv100% true that they took it too far and created new problems, it's totally misleading to claim it was done for anything to do with "LFQW" . That theoretical "problem" was caused by a failure in proper magic item churn when a gm doesn't follow the dmg guidelines. Instead it was obviously done in service of this silliness:

Spoiler

Trouble is that pretending d&d isn't a game they had been swimming in magic items from the start creates a ton of problems when a lot of classes§ are built around the expectation of multiplicative gains from them. The é.5Fighters were so solid in damage that you can find 3.5 wizard guides like the god wizard one praising that martial party role and slandering the very idea of glass cannon/blaster casters.

§fighter rouge barbarian ranger paladin and their various flavors of PrC/subclass.

 

[NotAYakk]

The math is weird to avoid exploits.

A simple version would be to give everyone an ECL. Half casters get (1/2 class level, rounded up), Subclass casters get (1/3 class level, rounded up), and full casters get class level. You just add up ECL.

This leads to an "exploit": 1 level splash of Paladin, then Ranger, then Artificer grants 3 full caster levels for a Wizard or Druid or Cleric or other full caster.

Another solution would be to add up your half-caster levels, then halve it and round up. Or 1/2 half caster, 1/3 subclass caster, add that up and round up, then add to full caster levels. This math gets complex to describe (no, really) and matters only in a few cases. So... they went with simpler math.

You can optimize around the simpler math, and for people who are doing complex multiclass builds that optimization becomes their problem instead of having to write elegant rules.

 

[Tessarael]

Rounding up after adding the fractions seems reasonable to me.

 

[Charlaquin]

The math is weird to avoid exploits.

A simple version would be to give everyone an ECL. Half casters get (1/2 class level, rounded up), Subclass casters get (1/3 class level, rounded up), and full casters get class level. You just add up ECL.

This leads to an "exploit": 1 level splash of Paladin, then Ranger, then Artificer grants 3 full caster levels for a Wizard or Druid or Cleric or other full caster.

This is possible in the 2024 rules as written , because all of those all half-casters round up when multiclassing now. It isn’t really an “exploit” though, because all of those classes also get spellcasting starting from level 1 in the 2024 rules. That is to say, they round up when single-classing and when multiclassing.

It only gets weird with 1/3 casters. They still round down, which makes sense because all 1/3 casters gain Spellcasting from a subclass, which kicks in at 3rd level. There’s no way for a 1/3 caster to “round up” because they need 3 levels in the class to gain spells at all. Their multiclassing should work the same way. You need 3 levels in the appropriate class to increase your effective caster level by 1.

 

[Mustrum_Ridcully]

In 3.5e, spells “upcast” automatically. Your caster level was a factor in the damage calculations, etc. Requiring you to spend a higher level spell slot to get the upscaled effect in 5e was a nerf, explicitly to tone address the “linear fighter, quadratic wizard” problem.

Wasn't it a bit more complicated than that - yes, a spell scaled with caster level -but the DC was based on spell level and caster ability modifier, so a 3rd level Fireball was easier to resist than a 5th Level Cone of Cold. Both deal the same amount of damage dice (at least until 10th level, Fireball ranged from 5d6 to 10d6, Cone of Could from 7d6 to 15d6, IIRC).
Requiring a whole feat to Heighten a Spell just for the DC however was rarely seen as worth it.

 

[Silam]

In 3.5e, spells “upcast” automatically. Your caster level was a factor in the damage calculations, etc. Requiring you to spend a higher level spell slot to get the upscaled effect in 5e was a nerf, explicitly to tone address the “linear fighter, quadratic wizard” problem.

Yeah, both editions had various mechanics for scaling effects:

3.5e:

• Caster level (which were mostly equal to the class level the spell belonged to, barring minor exceptions like magic items that gave +1 CL or similar).
• Attack bonus, based on the BAB you’d get from various classes (they didn’t all give the same), typically adjusted by Dex (not the spellcasting ability).
• Save DCs, adjusted by slot level + spellcasting ability. Slot levels were increasable via Heightened Spell but that was a Metamagic feat, and often not a very good one.
• Interestingly, psionics had augmentation mechanics very similar to 5e spells.

5e:

• Proficiency bonus (based on character level, not class level) affects both attack and save DCs.
• Spellcasting ability bonus likewise affects both attack and save DCs.
• Character level affects cantrip scaling.
• Augmentation via higher slots, where slots come from the aggregation of all caster classes (and half-classes, etc.).

EDIT: @Mustrum_Ridcully beat me to it

 

[Tessarael]

Correct. Arcane Trickster and Eldritch Knight don't get to add to the spell caster levels until Rogue level 3 and Fighter level 3. Beyond that, rounding up after adding the fractions is not unreasonable.

Multi-classed examples:

• Arcane Trickster 3/Eldritch Knight 3 => level 2 spell caster by the rules as written. They get 3x 1st level spell slots.
• Arcane Trickster 4/Eldritch Knight 4 => level 2 spell caster by the rules as written. They get 3x 1st level spell slots.
  • Dividing by 3, adding fractions, and rounding up, they would be a level 3 spell caster, with 4x 1st and 2x 2nd level spell slots, which is the same as a level 7 Arcane Trickster, or level 7 Eldritch Knight gets, so not unreasonable.
• Arcane Trickster 5/Eldritch Knight 5 => level 2 spell caster by the rules as written. They get 3x 1st level spell slots.
  • Dividing by 3, adding fractions, and rounding up, they would be a level 4 spell caster, with 4x 1st and 3x 2nd level spell slots, which is the same as a level 10 Arcane Trickster, or level 10 Eldritch Knight gets, so not unreasonable.
• Arcane Trickster 7/Eldritch Knight 7 => level 4 spell caster by the rules as written. They get 4x 1st and 3x 2nd level spell slots.
  • Dividing by 3, adding fractions, and rounding up, they would be a level 5 spell caster, with 4x 1st, 3x 2nd, and 2x 3rd level spell slots, which is the same as a level 13 Arcane Trickster, or level 13 Eldritch Knight gets, so not unreasonable.
• Arcane Trickster 8/Eldritch Knight 8 => level 4 spell caster by the rules as written. They get 4x 1st and 3x 2nd level spell slots.
  • Dividing by 3, adding fractions, and rounding up, they would be a level 6 spell caster, with 4x 1st, 3x 2nd, and 3x 3rd level spell slots, which is the same as a level 16 Arcane Trickster, or level 16 Eldritch Knight gets, so not unreasonable.

The point here is that the rules as written (RAW) round down rule for 1/3 casters is quite crappy for certain builds, unnecessarily so. Using the finer-grained "add fractions, then round up" house rule makes an Arcane Trickster/Eldritch Knight more viable.

I had a multi-classed Fighter/Rogue/Arcane Trickster/Spellsword in 3E D&D. (Yes, it gets silly with prestige classes.) Since then, it has been quite difficult to put together a similar build in 4E and 5E D&D. About the closest viable similar build would be a Rogue multi-classed with a spell caster that get extra attack (1) Bard with College of Swords or Valor; (2) Wizard with Blade Singer; or (3) Warlock with Thirsting Blade. Something like Rogue Assassin 8/Warlock Hexblade 12 with Pact of the Blade, Thirsting Blade, and Devouring Blade would roughly match what I had in 3E D&D.

 

[Silam]

Arcane Trickster 5/Eldritch Knight 5 => level 2 spell caster by the rules as written. They get 3x 1st level spell slots.

• Dividing by 3, adding fractions, and rounding up, they would be a level 4 spell caster, with 4x 1st and 3x 2nd level spell slots, which is the same as a level 10 Arcane Trickster, or level 10 Eldritch Knight gets, so not unreasonable.

I thought the RAW were to divide by 3, add the fractions, and then round down. If that’s true, then an AT5/EK5 would get a CL of 3, not 2. If we divide, then round down, then add up, it yields a CL of 2 (as you said). Which is it (by RAW)?