I am aware of our go to question about spellcasting and multiclassing, and I am aware of the rules. I am nonetheless confused about something:

If we use the multi-class rules, in particular

You determine your available spell slots by adding together all your levels in the bard, cleric, druid, sorcerer, and wizard classes, half your levels (rounded down) in the paladin and ranger classes [...]

So, a 5th level paladin contributes as a 2nd level caster (5/2 = 2.5, rounded down 2). A 2nd level caster would have 3 first level spell slots, and that's all. However, a 5th level Paladin has a spell slot table that clearly reflects a 3rd level spellcaster. In general, it looks like the Paladin level is being rounded up when defining the class' spellcasting table. The same can be said to the Ranger and to the "1/3 casters" (Arcane Trickster and Fighter).

So, when you multiclass, the half-casters and 1/3 casters contribute less to the spellcasting levels than their own progression suggests. This means, for example, a 3rd level Paladin and 2nd level Ranger, although "jointly" count as a 5th level "half-caster" (if you sum them before dividing and rounding1), when multi-classing the character is only a 2nd level multi-class caster and doesn't have 2nd level spell slots, even though a single-class 5th level ranger or a 5th level paladin would both have 2nd level spell slots.

This is weird to me, so, to start with, is the math right? Am I missing something? Or multiclassing is, in fact, inherently harming the spell progression provided by the not-full-spellcasters, in a way that 5 levels of different half-casters is simply worse than 5 levels of the same half-caster?

1 I just realized there are three answers. I only read the accepted, but the most upvoted contradicts it. Anyway, dividing and rounding before makes it even worse.

  • \$\begingroup\$ @Medix2 There are three answers to that question and they're all different, so any method would be contradicted by at least one of them. \$\endgroup\$
    – Mark Wells
    Commented Jul 6, 2020 at 3:43
  • 1
    \$\begingroup\$ @Medix2 Oh, I read (only) the accepted answer. And yes, rounding down before summing makes it even worse. \$\endgroup\$
    – HellSaint
    Commented Jul 6, 2020 at 5:33
  • 2
    \$\begingroup\$ To clarify: I wanted to ask this because, if I'm right here, my next question is going to be about house-ruling differently. I just didn't want to ask about house ruling based on a false assumption. \$\endgroup\$
    – HellSaint
    Commented Jul 6, 2020 at 5:50
  • \$\begingroup\$ I would note that this is still a huge buff to mutliclassing compared to previous editions, where different spellcasting classes didn't add together at all and tracked their level and spells known and spell slots entirely separately, such that four casting classes at 5 was caster level 5 in all of them and cast like a 5th level caster. \$\endgroup\$ Commented Jul 6, 2020 at 7:56
  • \$\begingroup\$ I've added an answer to the question How do paladin and ranger class levels add up for multiclass spellcasting? that objects to the assumption made in my answer. \$\endgroup\$ Commented Jul 6, 2020 at 14:13

1 Answer 1


Your math is correct (assuming sum first, then round down).

The rule for rounding down sort of gimps the progression of multiclassing half casters. When the sum of all half caster levels is even, the caster contributes slots as expected, when it is odd we see the problem you have outlined here: it seems to be a level behind. It takes two levels in the half-caster class to contribute one level to the multiclass spell casting table, and you cannot contribute half levels.

To your specific question, is multiclassing inherently harming the spell progression provided by the not-full-spellcasters, in a way that 5 levels of different half-casters is simply worse than 5 levels of the same half-caster?


The overall progression remains the same, but it does change arrangement of the individual level progression, delaying slot gains at every other level. You are correct to observe that 5 levels of different half-casters is worse than 5 levels of the one half-caster. But 6 levels of different half-casters contributes the same as 6 levels of one half-caster.

Here's a graph. Notice at even character levels, the total spell slots is equal, and at odd levels, the multiclassed half caster is a level behind the pur half caster.

enter image description here

  • \$\begingroup\$ Would multiclassing with Artificer make the paths identical? Also perhaps a chart showing the results when you round each separately is worth adding but that... gets a bit more complicated because 3+3 would be different from 2+4 \$\endgroup\$ Commented Jul 6, 2020 at 13:40
  • 2
    \$\begingroup\$ The wording on artificer is making me think "add first then divide and round" isn't actually the correct method. The wording there is just incompatible with that assumption. \$\endgroup\$ Commented Jul 6, 2020 at 13:43
  • \$\begingroup\$ The wording is compatible, because the progression is every-so-slightly different - Artificers get spell casting at first level, paladins and rangers do not. They're not the same and therefore should not be expected to be identical when multi-classing. \$\endgroup\$
    – T.J.L.
    Commented Jul 6, 2020 at 18:50
  • \$\begingroup\$ Doesn't this means that rounding up makes it the same? I guess the only difference is lvl 0-2 \$\endgroup\$
    – 3C273
    Commented Jul 8, 2020 at 20:15
  • \$\begingroup\$ @3C273 That is correct. \$\endgroup\$ Commented Jul 8, 2020 at 20:16

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .