I just read this answer to Can you Plane Shift to the plane that you're on? and it made me wonder: As a spellcaster, how do I know why spell failed? Is there any roll for it in either Pathfinder or D&D 3.5? Any rule?

It would seem natural to use Spellcraft for this, but neither Pathfinder nor any D&D edition have this use listed.

For a high-level spellcaster who knows his craft to be unable to identify the cause of failure seems weird, at least, so is this really so?

Spells that can fail without apparent reason:

  • Resurrection and Animate Dead cast on a body that is actually an immobilized undead

  • Plane Shift, as already mentioned

  • Any spell cast in an antimagic or dead magic zone when the character does not know of it

  • Necromancy cast on a disguised construct (not impossible with Warforged around)

  • Animal Trance when the target is actually an aberration

  • Control Water when the liquid is not actually water

  • Know Direction in chaos planes with no North

  • Scrying a creature that doesn't exist

  • Spells forbidden by a Mythal or similar powerful magic

  • Spells that got Counterspelled

  • All spells if Goddess of Magic dislikes you and decides to cut your access (as far as I remember Mystra had this option in Forgotten Realms)

And probably many more — I'm looking for general answer. Hints on specific cases would be nice, but I ask as a player, not DM, so I can't hope to always know which specific case it is in the first place.

For example, if Animate Dead fails, it might be due to the target not being a dead body, a Mythal effect, an NPC with silent means of necromancy counterspelling, or divine interference. (All those have happened to me or mine before.) And while there are means to get to know why and what a goddess thinks, a clue that the failure is her doing would save a lot of unneeded work (and gameplay time).

  • 12
    \$\begingroup\$ Not an answer, but it's totally common and reasonable in the real world for skilled people to not know why their thing didn't work. I can't count the number of stories I've heard from programmers where a program failed because of a minor mistake that took hours to find, and seconds to fix. \$\endgroup\$
    – DuckTapeAl
    Dec 7, 2014 at 2:19
  • 2
    \$\begingroup\$ @DuckTapeAl I know, that's why I was so surprised there is none in the rules. I always did it "by the feeling", setting DC roughly based on the level of spell and "sneakiness" of the reason. But for speed and story we did many things based on the feelings. Now, as a player, I expect to need this rule in next campaign and to my horror I discovered it probably never existed. \$\endgroup\$
    – Mołot
    Dec 7, 2014 at 2:22

2 Answers 2


The rules are silent on this.

In practice, I've seen it both ways, depending on whether the group takes a more gamist/metagamey approach or a more simulationist approach. "Open rules" games just have the GM tell you "that fails" if it's invalid on its face (only concealing the reason if it plays into the challenge, like "Charm Person fails... I wonder what that guy is if not a humanoid?" But in general you roll for SR, roll the save, you know why it failed. In our current Pathfinder campaign the GM runs it like this, and whenever it's in doubt a flurry of Knowledge checks fills in the gap.

In a more sim game those things are kept more "under the table" because it's not about the rules challenge it's about the in-game-world reasoning challenge. I ran a 5-year 2e campaign that worked like this.

This falls into the lacuna of other things like "do I know if I made a save vs. a spell" that is undefined so that different games can do it differently based on their desired feel.

  • 5
    \$\begingroup\$ You kind of assume that the “real” spellcasters you’re simulating would not be able to tell why, so a more simulationist game would also not tell you why. But I don’t see any reason to assume that a spellcaster couldn’t tell in-character, even if it’s a heavily-simulationist game. Fairly minor point overall and agreed otherwise (it can go either way), I just don’t see it as a gamist/simulationist issue (as either of those can go either way as I see it). \$\endgroup\$
    – KRyan
    Dec 7, 2014 at 7:20
  • 3
    \$\begingroup\$ If my code fails, I know from the way it failed what's wrong. Others see it failed, I see "invalid resource" or "resource unreachable", for example. So I see no reason for it to be different for spellcasters in heavy sim. Not when there are skills centred on knowledge and understanding of magic. \$\endgroup\$
    – Mołot
    Dec 7, 2014 at 13:43
  • \$\begingroup\$ It could, if you add a lot of stuff not usually considered part of the magic system (like a return code). You could certainly set up a sim magic system where you do know, but I'm addressing this as usually found implemented in stock D&D. \$\endgroup\$
    – mxyzplk
    Dec 7, 2014 at 14:49
  • 2
    \$\begingroup\$ The example in your last paragraph is kind of a terrible one, since the core rules quite explicitly point out that anyone who makes a save against a non-obvious effect (but not those who fail, presumably) feels a hostile tingling sensation but cannot discern the exact source from that alone. \$\endgroup\$
    – user17995
    Dec 8, 2014 at 7:26

It depends on how and why the spell fails. The caster is assumed to have a complete understanding of the spell, so he immediately knows the subject of the spell violates those rules or how the spell might be disrupted. Most cases, he'll assume something along they lines of wrong type, etc (like Charm Animal on a aberration).

In some cases a Spellcraft or Knowledge (Arcana) roll might be needed if there is a question about it. In the case of Plane Shift this is only an issue if the character thinks he's on another plane, since any wizard would know it doesn't work as a basic teleport.

For some of your other cases:

A dozen of other effects are suppressed by an Antimagic Field, if somehow he doesn't realize this, he will as soon as he attempts the spell. The spell doesn't really fail, it never starts.

Likely, the caster would assume the attempted Resurrection failed due to the soul not wishing to return (at least until he smelled the corpse).

There is a very short list of what Necromancy doesn't hurt or heal. That list is Construct.

Scrying on something that doesn't exist would result in the same failure as if a real person succeeded in their Will Save - which would be highly likely since they don't have any familiarily or likeness of the person

Counterspelling is obvious. Not only does the caster know it happened, he probably knows who did it (unless they're invisible or otherwise concealed)

If Mystra cuts you off from the Weave, the caster would know, similarly to if they were in an Antimagic field.

While this is a good question in certain cases - it most cases it would be the result of the caster being inept at his job. What cleric tries to revive a random corpse before doing a Heal check? What Wizard doesn't know what Plane they're on? What Diviner attempts to scry on a someone with knowing a great deal about them? When it comes up, either it should be clear what the reason is, or the DM could be giving a reasonable non-effect that leads you to the correct assumption.

For clarification, Saving Throws are noticeable by the victim which allows them a Knowledge check to discover the source. This implies the reverse is true and in any case they aren't, the caster should be allowed the same check to discover if the spell worked (which would be the same roll to determine if there was any active spell in the area). Not only that, but the bard spell Saving Finale claims that the caster can perceive one of his party members failing a save (the spell requires the bard to use an immediate action when the save is failed) as, apparently, a free action when the bard is neither the victim nor the caster.

  • \$\begingroup\$ Are you saying that the spellcaster should be able to tell if a spell failed due to a successful saving throw or a failed SR check? You don't really mention either possibility, just assume the failure was due to invalid targeting, so I’m not sure if that’s what you claim. Also, what is your basis for these statements? Are these rules, or houserules, or something you’ve seen in descriptions of 3.PF characters? \$\endgroup\$
    – KRyan
    Dec 7, 2014 at 7:21
  • \$\begingroup\$ I believe it does mention in the rule books that you know if they made a save (they also know they made a save in the case of mind affecting spells). Very few spell have no visual component and in most cases the complete lack of damage reveals the reason for the failure. Spell Resistance can be studied so the caster can make their spells better able to penetrate it, which implies some way to examine and study it in action (this would be one of the cases where the appropriate skill check helps - is it fire immune or is its SR protecting it from that Wall of Fire) \$\endgroup\$
    – Xander
    Dec 7, 2014 at 7:37
  • 1
    \$\begingroup\$ -1: I'm unaware of rules that support any of this stuff. I can't find anything in the magic overview in either system that makes it clear that you know why a spell would fail, or that a caster "is assumed to have a complete understanding of the spell". You can certainly run it that way, but you can just as easily run it the other way. If you can cite rules text that agrees with your interpretation, I'd reverse my downvote. \$\endgroup\$
    – DuckTapeAl
    Dec 7, 2014 at 8:36
  • 1
    \$\begingroup\$ The caster knows how the spell works, certainly, but not what the specific effects of this casting are. I don't know of any rules that say that a caster knows that the spell failed due to the target passing their save, for example. \$\endgroup\$
    – DuckTapeAl
    Dec 7, 2014 at 12:39
  • 3
    \$\begingroup\$ Just fyi, in Pathfinder at least, a caster is aware if a targeted spell is saved against (from the Magic Chapter, under Saving Throws paizo.com/pathfinderRPG/prd/magic.html#saving-throw) - Likewise, if a creature's saving throw succeeds against a targeted spell, you sense that the spell has failed. \$\endgroup\$
    – YogoZuno
    Dec 7, 2014 at 23:19

You must log in to answer this question.

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