Paladin 1, smite evil (deflection):

In addition, while smite evil is in effect, the paladin gains a deflection bonus equal to her Charisma modifier (if any) to her AC against attacks made by the target of the smite. If the paladin targets a creature that is not evil, the smite is wasted with no effect.

Sword-devil 4, untouchable (untyped):

Untouchable (Ex): At 4th level. the sword-devil's confidence and personality distract her foes in combat, making her more difficult to hit. When unarmored and unencumbered. the sword-devil may add her Charisma bonus (ifany) to her AC and her CMD.

Osyluth Guile feat (dodge):

Benefit: While you are fighting defensively or using the total defense action, select one opponent. Add your Charisma bonus to your AC as a dodge bonus against that opponent’s melee attacks until your next turn. You cannot use this feat if you cannot see the selected opponent.

According to Paizo, bonuses of the same type (except dodge) don't stack. There you have three different bonuses so, apparently, they should stack, which looks pretty strong.

Am I missing something?

  • \$\begingroup\$ Several other questions ask about a similar situation, like here, here, and here. These probably indirectly answer your question, so there's no need to close it or anything, but you may want to use them to inform a self-answer. \$\endgroup\$ Dec 14, 2020 at 19:11
  • \$\begingroup\$ So, according to the faq, they do stack. I'll use that faq for the answer. Thank you. \$\endgroup\$
    – Daniel
    Dec 14, 2020 at 19:23

1 Answer 1


You are correct, you would add 3×Cha to AC. This FAQ addresses the situation, and explicitly allows these to stack:

Do ability modifiers from the same ability stack? For instance, can you add the same ability bonus on the same roll twice using two different effects that each add that same ability modifier?

No. An ability bonus, such as "Strength bonus", is considered to be the same source for the purpose of bonuses from the same source not stacking. However, you can still add, for instance “a deflection bonus equal to your Charisma modifier” and your Charisma modifier. For this purpose, however, the paladin's untyped “bonus equal to her Charisma bonus (if any) on all saving throws” from divine grace is considered to be the same as “Charisma bonus (if any),” and the same would be true for any other untyped “bonus equal to her [ability score] bonus” constructions.

(Emphasis mine. Note that this FAQ’s claim that untyped bonuses equal to an ability modifier “[are] considered to be the same source” was utterly untrue prior to the publication of this FAQ, and as such this is a “stealth errata” of the type that Paizo so frequently engages in. The reasoning contained in this FAQ is basically hogwash, but the ruling itself is official. Just don’t try to pretend it actually makes sense based on the original text—it’s just errata that they’re pretending isn’t, as they so often do.)

You are also correct that this could be fairly strong.

On the other hand, smite evil is only for a few specific targets per day. A paladin does not get a lot of uses of that ability. The sword-devil ranger’s untouchable ability replaces spellcasting—and while ranger spellcasting isn’t great, it’s still spellcasting. It’s still very good, and very painful to give up. Moreover, combining paladin and sword-devil requires multiclassing, which is painful, and in this case directly limits how much smite evil you have available.

And then there’s Osyluth’s Guile, which costs two feats (Dodge is awful and should be avoided if possible, so being forced to take it here is very much a cost), and only applies to one target, and only if you use fighting defensively or total defense. Fighting defensively and total defense are both very weak options; any time you are using those, you are losing the fight.

So while 3×Cha is quite a lot, the restrictions on these greatly limit how powerful it actually is. And keep in mind that AC is one of the worst defenses in the game—no matter how good your AC is, there’s a lot of stuff out there that will just ignore it (because they target saving throws or what have you). If this was all the defense someone had, then they aren’t actually very strong defensively.

Were it me, I feel like a 4th-level sword-devil/16th-level paladin isn’t a terrible idea. That’s a lot of stuff you’re using Charisma for, and that’s a good thing. But I’d probably skip Osyluth’s Guile. And this doesn’t strike me as a particularly strong character; decent, not amazing. For example, a basic “oradin” built off of a 2nd-level paladin/18th-level oracle with one of the several Cha-to-AC mysteries is far superior.

  • \$\begingroup\$ As long as you're talking about costs, one of the three sources of +Cha prevents you from wearing armor; the incremental benefit of +3xCha over +2xCha when it means giving up high-level Paladin standard magic full plate doesn't seem like a huge benefit. Sure, the extra +Cha works against everything, where full plate (assuming Pathfinder followed D&D's lead) doesn't work against touch attacks, but most of the time, the full plate is giving you higher AC for the non-touch attacks, so it's kind of a wash, right? Not being a PF player (and stale on D&D 3/3.5) I could be missing something. \$\endgroup\$ Dec 14, 2020 at 19:43
  • \$\begingroup\$ Should probably address the FAQ just to be clear that you are allowed the Dodge Bonus(es) with the straight Charisma bonus. \$\endgroup\$
    – Ifusaso
    Dec 14, 2020 at 19:45
  • \$\begingroup\$ @ShadowRanger You are correct, but if one is doing this, they should probably be pushing Charisma as hard as they possibly can, and thus can easily outstrip even +1 full-plate. Getting higher than +1 is a waste of money, and in any event being effective against touch attacks easily makes the Charisma bonus worth twice or more as much armor bonus. So that much isn’t that big a deal that you have to be unarmored. \$\endgroup\$
    – KRyan
    Dec 14, 2020 at 19:49
  • \$\begingroup\$ @Ifusaso Sure, for some reason I’d thought the question had already linked that but you’re right, it makes sense here. \$\endgroup\$
    – KRyan
    Dec 14, 2020 at 19:50

You must log in to answer this question.

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