I've recently committed to playing in a Pathfinder game, and I've become somewhat intrigued with the idea of playing a rogue/sorcerer multi-class character. I'm concerned, though, that as the game progresses I won't be able to keep up with my straightclassed party members in terms of damage output and survivability, essentially becoming a novelty character with a few parlor tricks and no real staying power ("pick this lock and then stay out of the way until we need you to cast detect magic on the loot").

That got me thinking: is there an effective way to identify class combinations which seem appealing up front, but which will fall behind as the characters level up (without actually taking the character through its paces in a game)?

I'm making a couple of assumptions here (NB, I'm limiting the scope of the question to core classes): first, that there are class combinations that DON'T scale well together into the mid- to late-game (if every class combination is equally (in)viable, this question is moot).

Second, that the definition of "working well together" precludes the necessity of min-maxing and shoehorning a multi-classed character into one specific playstyle ("this build works great as long as you only use sneak attacks"), though some combination will of course suit a style of play particularly well (you probably won't play a wizard/rogue if you're looking to be on the front lines and soak up punishment, for instance, but you might try cleric/fighter).

And finally, that a multi-classing character will limit the number of classes taken to two. It seems that multi-classed characters who take more than one additional class start to suffer from lack of focus.

  • 5
    \$\begingroup\$ Not enough for a complete answer, but... As a rule of thumb, caster classes (especially those with 9 spell levels) tend still to fall behind if evenly multiclassed with another. That's because caster levels don't stack (as BAB and saves do). To overcoming the lack of caster level and spell power, you could focus on buff spells only, or adopt a prestige class that advances both your caster level and your other class strength (like Arcane Trickster, Eldricht Knight or Mystic Theurge). \$\endgroup\$ Commented Jun 20, 2012 at 22:02
  • 1
    \$\begingroup\$ Note, however, that dual-progression prestige classes are often quite a bit lower in power than their single-classed entry classes. Being two-to-three spellcasting levels behind puts a huge dent in your ability to make your spells worthwhile. That said, gish (fighter/mage) characters tend to handle lost levels better, so Arcane Trickster and Eldritch Knight end up all right. Mystic Theurge is mostly a trap, barring shenanigans. \$\endgroup\$
    – KRyan
    Commented Oct 20, 2012 at 5:06
  • \$\begingroup\$ Quick idea: It might be more effective to put ranks into 'Use magic device' and use wands. The DC is 20 to use a wand, but you can easily pump your skill to hit that without rolling (traits, feats, charisma bonus, etc). I know wands only go up to 3rd level spells, but a multiclassed caster will only have access to low level magic through most of the game anyway. \$\endgroup\$
    – Macona
    Commented Oct 31, 2012 at 12:47

3 Answers 3


Thou shalt not sacrifice caster levels.

As the golden rule of 3.5 multiclssing, it's hard to beat the above. It's quite possible to apply requirements gathering methods to 3.5 as well as 4e, and estimate approximate to-hit and damage at each level.

Plan out your character to 20.

To answer:

s there an effective way to identify class combinations which seem appealing up front, but which will fall behind as the characters level up (without actually taking the character through its paces in a game)?

is simple: plan out your character to level 20. Identify, at each level, what benefit the character is deriving from your choices. Test against the requirements given in your requirements step. This way, when your character shows up in game, you have an idea of your intent and the capabilities of the character.

To answer what classes synergize well:

Non primary-casting classes tend to synergize well.

Given that most aspects of a level up are cumulative with prior choices (BAB, feats, HP) you want to avoid class features which depend on your level in the class. Therefore, avoid class features which have level as a variable within the feature.

Rages from barbarian are fine. You get more as you level up:

a barbarian can rage for a number of rounds per day equal to 4 + her Constitution modifier. At each level after 1st, she can rage for 2 additional rounds.

but you don't have the benefits of the rage reduced (proportionally to your compatriots) as you see in spells.

For the casting classes it's certainly possible to combine them in "gishy" characters (spell-slinging fighters) but they require a specific focus into the combination, preferably enabled by appropriate choice of prestige classes.

So, if you fail to raise rogue, you don't get as much sneak attack bonus damage, but if you're increasing fighter as the other class, you get a higher BAB to compensate.

Whereas a rogue would effectively be writing off "improved uncanny dodge" due to multiclassing.

Therefore, a good combination can be fighter/rogue, because each class brings a separate thing to the table, without having a significant opportunity cost for choosing the other. Other good combinations are those that lead up to exceptional PrCs, which combine features from both classes in useful synergy.

  • \$\begingroup\$ Note that OP was asking about Pathfinder, not 3.5 or 4e. I agree with your basic points, however: especially not sacrificing caster levels in multiclassing. \$\endgroup\$
    – Tyri
    Commented Jun 22, 2012 at 11:07
  • 2
    \$\begingroup\$ Yes, but pathfinder has same caster level mechanics, answered from my personal expertise. \$\endgroup\$ Commented Jun 22, 2012 at 16:38
  • 2
    \$\begingroup\$ @Tyri: Ultimately, Pathfinder changed very little about 3.5. Lots of little changes here and there, but little was done to shift the "center" so to speak. Basic stuff like "thou shalt not lose caster levels" and "plan it out" remain very much the same. \$\endgroup\$
    – KRyan
    Commented Oct 20, 2012 at 5:03
  • \$\begingroup\$ +1. Disagreeing about the sacrifice on caster levels but agreeing on your other two points. It depends on what one is seeking, my gish having only five levels of spells would be still formidable if their base attack reached at least 16. It just shifts the focus away from spellcasting and more onto combat/skill enhancing. \$\endgroup\$
    – LitheOhm
    Commented Oct 22, 2012 at 3:54
  • 2
    \$\begingroup\$ @litheOhm: the standard for a gish is 9th-level spells and BAB +16. Since that can be done in core, there's not much reason to ever miss the 9s. Brian is 100% correct about spellcasting: optimizing means having as much of it as you can. \$\endgroup\$
    – KRyan
    Commented Oct 22, 2012 at 12:46

A useful philosophy when it comes to multiclassing with a spell casting class exists to enhance the non-spell casting class. Using the rogue/sorcere example you mentioned, wanting a "sneaky sorcerer" might really limit your effectiveness, but wanted a pretty good rogue who could enhance himself with lots of spells then the multiclassing might work for you. There are a lot of spells in the first four levels or so that might be fun for a rogue to work into his/her arsenal, ie true strike, invisibility, spider climb to name a few.

Where I am going with with this is if you compare a sorcerer to your sorcerer/rogue then you are probably going to find the combo lacking since you are comparing them in terms of a sorcerer. Now compare that to a rogue to a rogue/sorcerer. Now you can start to make arguments like my sneak attack damage is not as high, but using improved invisibility I can sneak attack more often or I may not hit as hard, but my survivability is up when I use such spells as shield etc.

To sum up my argument, you will probably have better luck multi classing into a spell caster class to enhance your existing class than simple diluting a spell casting class. This should be a good basis of comparison by looking at how individual spells can help enhance your base class. If you do find you do not lose too much power overall, then you'll also be happy to know your build will be much less cookie-cutter compared to all the straight rogues out there.

  • 1
    \$\begingroup\$ I totally misread what you were saying and downvoted you :( It's locked in now, but if you edit your post I can remove it. In general, I think you need to emphasize that one has to be quite careful about how one dips spellcasting classes; low-level spells can be potent, but you need to know what you're getting before going in. Also, I cannot recommend Cleric strongly enough for this: thanks to Domains, Devotion Feats, and Divine Feats, Cleric 1 is one of the best levels in the game, probably the very best dip there is. \$\endgroup\$
    – KRyan
    Commented Oct 20, 2012 at 5:29

Final analysis: The concept of prestige classes involving primary spell casting classes is severely broken, from D&D 3.0 to 3.5 to Pathfinder. The basic problem: spell casting depends upon level of caster and ANY diversion from the spell casting class affects this. There are no synergies to be had that will make up for lost spell caster level: a lower spell caster level affects the number of spells, the level of spells and the effectiveness of spells.

The most effective prestige classes combine fighter classes because base attack bonus and saves ALWAYS add. The least effective prestige classes combine primary spell casters with anything else because you LIMIT your spell caster level. Rogue-Fighter mashups can be ok, because while doing less sneak attack damage, they hit more often due to fighter additions to base attack bonus.

Fighter type (or rogue) - spell caster combinations suck when it comes to combat because they will ALWAYS be less effective then then straight fighter types, rogues or spell casters. I am familiar with the argument that the character can augment fighting or rogue ability with the proper spells, but the argument ASSUMES the time to cast those spells! How will the other players feel when your eldritch knight ducks out of the first 6 rounds of every combat in order to buff up? What happens when it's the third (or fourth) combat of the day (possibly in a dungeon crawl) and the character is fresh out of buffs? Yeah you can husband the buffs, but then the character is less effective ALL the time.

In my opinion, they should have kept something similar to the old dual-class rules from D&D 2.0. It had the desirable trait that a dual- (or multi-) class character was always only a level behind the straight up types. Also, there's a built in limitation in the gaming system: no matter how many options you have, you can only exercise one at a time.

The Unearthed Arcana for 3e systems has a character gestalt, where you completely mix two (or more) classes and everything is additive. The problem here is that there are NO tradeoffs for multi-classing, and a gestalt character ends up MUCH more powerful than a straight up character.

I've been toying with an idea lately that might address some of the prestige and multi-class drawbacks from 3e systems: Set the characters spell caster level equal to his character level, period. This makes all the classes additive in multi- and prestige classes, not just the fighter derived sub-classes. It also still contains a trade-off for the multi-class character: loss of higher level spells. But, at least the spells a character does get will be as powerful as his straight-up brethren's.

  • \$\begingroup\$ You are incorrect that multiclassed warrior-mage or thief-mage types will always be worse than warriors or thieves (the statement is basically true for mages). Buff time is definitely a relevant problem but one that can be handled. You also seem to ignore the existence of prestige classes that can be entered and completed without losing any spellcasting, which you do not address but seem to include in your statement about spellcasting prestige classes always weakening the character. Your statement is true for PrCs that lose spellcasting, but many don't. \$\endgroup\$
    – KRyan
    Commented Oct 28, 2012 at 22:58

You must log in to answer this question.

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