Does the primal warden Unstable spellcasting require material components?

Question

The primal warden archetype of shaman has the unstable casting feature

A primal warden’s spirit is ever shifting, and the spells it grants the primal warden change from moment to moment.

When the primal warden would spontaneously cast a 1st- through 8th-level spell using the spirit magic class feature, she produces a random spell from the list of shaman spells of 1 spell level higher than the spell slot expended; use an appropriate die or another randomization method to determine the spell. When she would spontaneously cast a 9th-level spell using spirit magic, she instead casts a random 9th-level spell from the shaman spell list, treating her caster level as 2 higher.

There is no way to predict the spell before the shaman begins casting it using spirit magic, and she cannot enhance the spell using metamagic feats. The spell uses the shaman’s caster level, even if that caster level would normally be too low to cast the spell. There are 47 potential 2nd-level spells, 46 potential 3rd-level spells, 41 potential 4th-level spells, 30 potential 5th-level spells, 22 potential 6th-level spells, 22 potential 7th-level spells, 18 potential 8th-level spells, and 14 potential 9th-level spells.

At 8th level, after determining the spell that she would cast using spirit magic, once per day as a free action, the primal warden shaman can choose to randomly select another spell instead, using the same method of randomization as she used before (if she obtains the same result a second time, she must keep that result). At 12th level and every 4 shaman levels thereafter, she can choose to randomly select a different spirit magic spell one additional time per day, up to a maximum of four times per day at 20th level. She cannot use this ability more than once in the same round.

This alters spirit magic and replaces the hex gained at 8th level.

Given that it is a random spell one could conceivably randomly cast a spell with costly material components like restoration or reincarnation.

would these random spells still require the costly material components?

Answer 1

Nothing there says that you can cast the spell without the components, so you need them. If you haven’t got them handy, you can’t complete the spellcasting and the spell fails.

Which itself makes a pretty good case for the primal warden being able to choose to let a spell fizzle, by refusing to do the somatic or vocal components. The description doesn’t explicitly address this, either.

But ultimately, unstable spellcasting is an atrociously poor design, so it’s unsurprising that the basic ramifications of this feature are simply left unconsidered. There are ways to design random—even very random—systems, that nonetheless can be used constructively and fit well within the game. This design, however, doesn’t bother to do any of that—it just says “roll on the whole list, that’s your spell!” and fails to even consider what that would mean in practice, how to handle corner-cases like this question or how such a character actually functions. It’s not just bad design, it’s also lazy design. It’s a complete abdication of the responsibilities of a game designer to provide quality content to customers. I would never allow this in my games, and I strongly recommend all GMs ban it.

Better designs are certainly possible here. Off the top of my head, a system that allows the character to bank a random spell for a period of time—at least allowing them time to figure out how best to use it, get any components they need ready, etc.—would be a start. It doesn’t have to be long, but even the ability to delay the spell by 1 round would alleviate a ton of problems, including those asked about in the question as well as @HeyICanChan’s really good comment asking about how long casting times would work under the official system.

The other thing that’s just not going to work here, ever, is rolling off the entire list. The numbers given in the description—what sources does that include? I don’t know, from reading it, but it definitely concerns me. What happens when someone buys a new sourcebook? Is the primal warden required to start including those spells in his random chances? What if he doesn’t like any of those spells, and now his chances of getting a spell he actually wants has dropped significantly? I mean, you can imagine a “book of fun and silly spells” that may well be fun and silly, but are largely difficult or impossible to use in a lot of situations. Fun if you can get them to work, a lot less fun if that book doubles your odds of getting a useless spell. Especially if someone else brought the book to the table and you never wanted it there to begin with.

The solution here is to allow the primal warden to set up smaller lists, subsets of their full spell list that, while they are still receiving randomly, they can at least be sure that all of the options on the list are reasonably useful. They can also avoid putting in things with expensive material components. Easiest way to do it is just one list of spells per level, maybe changeable daily. But you could allow the primal warden to make multiple lists per spell level, so they can make maybe a set of damaging spells, a set of buffing spells, a set of summoning spells—they still don’t know what they’re going to get, but they’ll have some kind of idea what they’ll get.

There are a lot of options. Even these are just really quick ideas off the top of my head. There are also other corner-cases, which options like these can either remove as serious concerns, or handle appropriately. Figuring out what works best, what corner-cases still need to be handled, and how to handle those, those are all part of the designer’s job. The designer of this class feature, in my opinion, did not do his job. It saddens me to think that people are spending money on his dereliction of duty.