I’m plowing back into Chapter Three: Spells of Xanathar’s Guide to Everything this week. I got through the Fs last week, so let’s see what I can get done before I have to click “Publish” this week. Also, many thanks to the alert reader who pointed out my error on danse macabre – it does indeed play nice with the Necromancy feature Undead Thralls.
Barbarians and Bards | Clerics, Druids, and Fighters | Monks and Paladins | Rangers and Rogues | Sorcerers, Warlocks, and Wizards | Backstory and Feats | Rules Clarifications and Encounter Building | Traps, Downtime, and Magic Items | Spells A-Fa | Spells Fi–N
Oh man, am I ever excited about this one. We are about to hit the most controversial spell of XGTE. I am so excited.
Find Greater Steed (Paladin): This is not the controversy. This is a direct upgrade of find steed, which amounts to two class features stored in spells. Of course, storing them in spells means they can be “borrowed” by bards. Anyway, this does exactly what it says on the tin, and adds a whole list of new mount options. The Chronicles of Prydain fan in me is endlessly grateful to see saber-toothed tiger on the list. This spell also emphasizes how great paladin spell prep is – prepare this and cast it, and you can prep something else tomorrow – compared to the ranger, who would be stuck with it. You can’t ever afford to drop the spell completely, after all.
Also, I wish warlocks had a way to steal find steed and find greater steed. Maybe it’s my long-dormant WoW fan speaking, but I like warlocks getting cool mounts. Should probably be an Invocation, though.
Flame Arrows (Druid, Ranger, Sorcerer, Wizard): This is a damage kicker that druids, sorcerers, and wizards can hand out to archers, and it lasts for just twelve arrows (more at higher levels). It requires concentration, so for most rangers in most situations, it’s a very bad way to cast hunter’s mark using a 3rd-level slot instead of a 1st-level slot. It doesn’t need a bonus action to transfer targets, but that’s about all that this has in its favor. Dishing out 12d6 extra damage with a 3rd-level slot isn’t bad for druids, sorcerers, or wizards, but it’s got a lot of failure states as well. Probably give this one a pass.
Frostbite (Druid, Sorcerer, Warlock, Wizard): This is a great attack + debuff cantrip. It’s good enough to at least give warlocks something to think about in place of eldritch blast spam, but they’re sacrificing a lot of damage output for only the potential of avoiding damage. For druids, sorcerers, and wizards, this is a great choice, though targeting Con doesn’t necessarily play well at high levels.
Guardian of Nature (Druid, Ranger): This is a strange kind of shapeshifting spell that hearkens back to 4e’s druid and warden classes quite strongly. I love what it offers in flavor, and I loved the 4e druid and warden. Buuuut… using this well takes a lot of thought. The short version is:
- Primal Beast is for Strength rangers and Circle of the Moon druids who are about to go into a Strength-using beast form. Strength rangers are, how do you say, super rare. Of course this uses Concentration, too. Anyway, extra speed and advantage on all of your attack rolls is great, but you’ve got to think of the damage kicker as nothing more than a replacement for hunter’s mark, from a ranger’s point of view. For the druid, of course, there aren’t a lot of other melee damage kickers lying around.
- The misleading part about this is that Dex rangers in particular are likely to think of speed boosts as a high-priority choice, but the rest of this is not great. The visual of “what kind of character should want to be a Primal Beast?” would naturally point me toward a two-weapon fighter, for virtue of being the most like claws. What do I know, though?
- Great Tree is for Dex rangers (very close to all of them), Circle of the Moon druids who are about to go into a Dex-using beast form, and druids who plan to cast primal savagery, produce flame, or shillelagh a lot. That third category is a pretty dangerous place to live, because you don’t have the damage output or durability of most front-liners. Great Tree form helps with that a bit, thanks to temporary hit points, but still. One of the main effects of advantage on Con saves is that you’re less likely to lose this spell as you soak up damage.
- A 15-foot radius of enemies-only difficult terrain is huge for being hard to approach and hard to escape. Of course, the most archetypal Dex rangers aren’t defenders in the first place, but strikers, looking for damage boosts and movement speed. A defender ranger is probably stacking Strength, the Defense fighting style, medium armor, and a shield or two-handed weapon.
Gust (Druid, Sorcerer, Wizard): It’s forced movement from a cantrip, though it’s way way less than the passive benefit of Repelling Blast. It also offers less-controlled object movement than mage hand. It’s… okay? But in a world of a limited number of cantrips, it’s not a great pick unless you’re super committed to being air-themed. We can certainly wish that a spell with the same concept as gust of wind didn’t also have so similar a name – this can only cause confusion at the table.
Healing Spirit (Druid, Ranger): Yeah, good odds that you already knew this, but this is the controversial one. I don’t think Jeremy Crawford has offered unofficial alternate rules for any other XGTE spells. It’s going to take a few extra paragraphs. The core issues of the spell:
- Okay, even the superficial: 10d6 of healing output, an average of 35 points of healing off a 2nd-level slot.
- One layer deeper: Yeah, you don’t have to work real hard to go way higher than 10d6. 15, 20 dice spread across the party is juuust above trivially easy. As long as you can move, it’s a lighter touch on the action economy than the 3rd-level spell aura of vitality (hard cap of 20d6 healing output). There’s even handling to make sure your healing zone doesn’t accidentally heal bad guys.
- As a follow-on to 2, if you cast it with a 3rd-level slot, its base healing becomes 2d6, so it’s probably more than twice as good as aura of vitality. (If you’re willing to take the hit to your action economy, its minimum effect is as good with a 3rd-level slot as aura of vitality, so it can only go up from there).
- Out-of-combat, it’s wildly better than other out-of-combat healing options, where the only reason it might not generate 50d6 (assuming a 5-person party) of healing is that you might not have 175 damage for it to heal. Jeremy Crawford defends this with “we assume you always start fights at full health,” but if that’s true, why does the game assign any value at all to Hit Dice, Song of Rest, prayer of healing…? Out-of-combat healing is still part of the attrition consequences of the game’s dynamic.
- Multiclass into Life cleric and watch this just get more insane. Disciple of Life means all of those d6s become d6+4. That’s, um, slightly better than doubling its average output. Whee!
So, this is a mess. I’ll quibble with James Haeck’s recently-offered definition of “broken” by offering my own. We can reasonably call something broken when it fully eclipses or is eclipsed by other options, so that a player is making a substantially suboptimal choice (outside of doing things for characterization, which is legitimate; I am not addressing that) to use or not use a feature. This is a subset of James’s “not fun to play (with),” but I hope it is useful in setting aside the nebulousness of “fun.” What I hope you see here is that with this spell, a druid or ranger is an absolutely dominating healer, compared to any other way of generating healing.
Here’s Jeremy Crawford’s unofficial variant, on Twitter. It has its own set of serious problems, but let’s look at what it does.
- Prayer of healing and aura of vitality are still our benchmarks here. PoH is out-of-combat-only, while AoV is fine whenever. It scales with party size, from a probable minimum of 3 party members (6d8 + 3*Wis, so with 20 Wis, that’s an average of 42) to a hard upper limit of 6 party members (12d8 + 6*Wis, or an average of 84). PoH also has a pretty generous scaling function – 1d8 per target. AoV, we’re looking at an average of 70 (20d6) healing output, with no scaling function (but it also doesn’t care how many people are in your party).
- Variant healing spirit (sorry, gotta be clear here) probably kicks out 6d6 (avg 21) or 8d6 (avg 28) when you first get it (you likely don’t have 20 Wis as a 3rd-level druid), or 2d6-4d6 (avg 7-14) if you’re a 5th-level ranger. My point here is that linking healing output to ability scores both punishes the ranger unduly for being a multiple-attribute-dependent class, and further pushes the payoff for taking that one level of Life cleric. Now, you mostly wouldn’t multiclass to benefit one spell, no matter how good, so let’s not get bogged down there, but one level of cleric offers a lot of other fringe benefits.
- The good part about #2 is that 5th-level ranger is no longer a top out-of-combat healer. The bad part about it is that it’s another ranger spell that turns near-worthless because of how the ranger’s ability scores work out. A deep study and rework of ranger spell options is on my to-do list, probably for Harbinger of Doom. Anyway.
My takeaway from the whole thing is that healing spirit is not a good candidate for addition to the game, but if you want to improve druid, paladin, and ranger healing (a worthy goal! Share the load!), you can do a hell of a lot worse than breaking prayer of healing out of being cleric-only. Healing spirit wants to be a WoW-style Lightwell (pre-Lightspring), and has a lot of the problems that would obviously follow from that. (PS. I’m aware of some apples-to-oranges math comparisons in my points above; I hope my points are still clear.)
Holy Weapon (Cleric, Paladin): This spell grants a weapon a 2d8 radiant damage kicker for an hour, for a 5th-level slot, and costs the caster’s concentration. There’s a super involved set of considerations around cleric/paladin buffs and concentration, with degree of benefit, number of targets, and duration as competing sliders. This one pushes number of targets all the way down, so that it can push duration way up, and be fairly generous on degree of benefit. It’s a paladin buff that is good enough to make you think about not using any of the smite spells, trading the amazing burst-damage capacity for equally amazing sustained damage. It’s competing with destructive wave, and that’s a tall order. Anyway, this spell is amazing if you’re going to be in enough rounds of combat in an hour, or you pick your moment for its dismiss effect. I love that it has a dismiss effect – that’s a design concept I’m playing with heavily in my current work for Rite Publishing – and I would like to see similar mechanics show up elsewhere.
Ice Knife (Druid, Sorcerer, Wizard): This spell is probably about as classic a post-core addition to D&D as Aganazzar’s scorcher, though again I don’t have details on that. Anyway, this is an area effect that also has a primary target who takes some extra damage. It’s hail of thorns, but good because it doesn’t eat your concentration or require a hit, and deals slightly better damage. Its scaling is not quite as good as hail of thorns, and the primary target probably takes slightly less damage. Good 1st-level damaging AoEs are not easy to come by, so this belongs in every casting druid’s, sorcerer’s, and wizard’s lists.
Illusory Dragon (Wizard): Matters are dire for high-level Illusionists looking for spells in-school (that is, making Illusion Savant do anything), so I’m happy to see this one. Every edition of D&D that has had illusionists has had them slide toward semi-real shadow magic at higher levels. Up to this point, 5e only really does that with the Illusionist’s Illusory Reality feature. This spell accomplishes exactly what it says on the tin, with a combination of dragonfear and strong, sustained damage output. It targets Wisdom for the fear effect and Intelligence for the damaging effect; I’m super happy to see that XGTE continues the trend of bringing Str, Int, and Cha saves up to parity on frequency of use. There’s an expected up-to-70d6 AoE damage output from this spell. Looks great to me. Is it overpowered? It’s an 8th-level spell, so the question is almost irrelevant, but I would probably pick this over weird in most cases. It deals slightly more damage and has far more situational adaptability, in exchange for taking a bigger bite out of your action economy and requiring a lot of space for a Huge creature. Oh, and the incredibly common immunity to frightened doesn’t make you immune to illusory dragon the way it does with weird.
Immolation (Sorcerer, Wizard): Here’s another top candidate for addition to the warlock list. It’s a damage-over-time effect with extra initial damage and new saves each round to end the effect. Its damage, compared to the Spell Damage table, is about right if the creature fails its initial save and takes one further round of damage. It’s a great, stylish spell; my only quibble with it is targeting Dex is an odd choice.
Infernal Calling (Warlock, Wizard): This is a top candidate for my favorite spell in XGTE. It reflects exactly what you’d want from a D&D-ified interpretation of Goetic magic. There’s further benefit to capturing diabolic talismans and truenames as treasure. This spell oozes story potential, and it’s a much better spell-level-to-CR ratio than you’re going to find in other conjurations. We first saw this model of high-risk conjuration in UA: That Old Black Magic, though this is Infernal rather than Abyssal. Also, I think it’s good to keep in mind that a 9th-level spell plus a suitable talisman can net you a horned devil, ice devil, or pit fiend (but not an erinyes, for some reason).
Infestation (Druid, Sorcerer, Warlock, Wizard): This is a strange attack cantrip that deals poison damage and forces movement randomly. I’m not sure how you’d get a good, desirable effect out of random forced movement (without an implausible degree of setup), and the damage output is underwhelming. Pick anything else, for any reason.
Investiture of Flame (Druid, Sorcerer, Warlock, Wizard): This is the first of the four investiture spells, and I am a fan of none of them. It could be worse; at least these require a spell slot and concentration. This is a multi-purpose self-buff, with a mix of defense and offense, which is kind of what you need to do with high-level buff spells. It grants a new at-will attack for its duration: Aganazzar’s scorcher, but half-length and +1 spell slot level. I mean, getting to spam a 2nd or 3rd-level spell is amazing, to say nothing of the fire aura. If it weren’t for the immunity, I would probably like this a lot, and even so it’s a good benchmark for multi-effect buffs.
Investiture of Ice (Druid, Sorcerer, Warlock, Wizard): Where flame offers a damage aura, this investiture radiates difficult terrain. Where flame offers a narrow blast of fire, this one is a broader cone of slightly less damage, plus a speed reduction. A speed reduction and difficult terrain means this could be a lot of movement lockdown. Also, free movement over ice.
Investiture of Stone (Druid, Sorcerer, Warlock, Wizard): Well, this one is a stoneskin without an expensive material component and a shorter duration, an AoE knockdown, and an earth glide form of movement. This is more defense + utility than the others, but it’s certainly potent. I guess I don’t mind this one.
Investiture of Wind (Druid, Sorcerer, Warlock, Wizard): As with so many elemental air spells, this one emphasizes disadvantage on ranged attacks against you. It grants a fly spell and a damage-and-forced-move at-will area effect. Flying certainly makes up for having no other defense against melee, and tossing off a whole lot of damaging whirlwinds sounds like fun. I guess I’m fine with this one.
Invulnerability (Wizard): From the department of doing what it says on the tin, we have a winner. Also, the shortest spell description in the game. Sure, you can still get screwed over by all kinds of conditions, and anything that incapacitates you also breaks your concentration, but total damage immunity is phenomenal. On the other hand… 9th-level spell. It’s probably fine.
Life Transference (Cleric, Wizard): I feel like this could have gone to the sorcerer and warlock too, without much problem. It’s healing with hit point transfer – in fact, healing twice what it costs you. Interestingly, it works less if you have necrotic resistance, and not at all if you’re somehow immune to necrotic damage. I’m surprised they went with that, instead of simply having it ignore resistance and immunity. The best thing about this spell is Necromancers getting a way to heal others that also gives them a driving need for Grim Harvest.
Maddening Darkness (Warlock, Wizard): That thing I said about weird and its issues? This is like that. Maddening darkness deals more damage, has a longer range, larger area, longer duration, immunity to frightened is irrelevant, the spell effect doesn’t end just because you pass one save, and a big cloud of darkness is probably more benefit than hindrance if you get to choose its placement. So yeah, this is a whole lot better than weird, at one spell level lower. Breaking away from that analysis for a moment, this spell is fine. Compared to Abi-dalzim’s notorious murder magic, you need a little more time to work, but this and a wall effect or some forced movement and you can just go nuts on the damage output. Very solid Mystic Arcanum for a warlock with the various forced-movement Eldritch Invocations.
Maelstrom (Druid): Another area damaging effect, with a duration and forced movement. As long as you can keep throwing enemies back in when they get out, this could be amazing. Difficult terrain combined with a 10-foot forced move? Oh yeah, that’s awesome. Nice big area and long range, too. It is no-damage-on-save, though, which is why you need something to keep pushing them back in. Your fighter buddy’s Shove is perfectly cromulent.
Magic Stone (Druid, Warlock): No, I don’t know why this is on the warlock list either. It’s the ranged match for shillelagh, and useful if you need to bring down a flying enemy with resistance or immunity to nonmagical attacks, but I’m not sure the effectiveness justifies spending a precious cantrip slot on it. It has no scaling effect with increased character levels.
Mass Polymorph (Bard, Sorcerer, Wizard): Alternate title: how to turn up to 10 commoner allies into rampaging murder machines in one 9th-level spell. As I’ve said for other things, it’s hard to say that a 9th-level spell is overpowered in some ways, but this spell is so wide open to various kinds of abuse as either a buff or a debuff that the mind boggles. All I can think of here is the Romance of the Three Kingdoms campaign I played in college, where our wu jen turned the otherwise unstoppable Lü Bu into a carp. This was remarkably effective crowd control. If you have the means to apply further awful conditions, mass polymorph is also a great debuff to all saving throws, since just hitting Lü Bu would free him from his fishy form in 5e. Or, you know, you can just quietly watch him suffocate, since that does not inflict hit point damage.
Maximilian’s Earthen Grasp (Sorcerer, Wizard): Another golden oldie of D&D spells, I first saw this one in 2e’s Tome of Magic, but I have every confidence that it predates that venerable text. Anyway, this spell shapes the earth into a hand that you can use repeatedly over the next minute. During that time, it can grab creatures and restrain them (requiring a Strength save, not a grapple check, to escape), and while it has them, it can crush them. That additional crushing demands a lot from your action economy, but let’s keep our eyes on the prize here: restrained is incredibly good. The one problem I see with the spell is that it takes both an action to grab and an action to move the hand, so you could have a bunch of wasted rounds if your enemies scatter from it.
Melf’s Minute Meteors (Sorcerer, Wizard): Yet another of the classic post-core attack spells, Melf’s minute meteors in 5e is sort of an odd idea. You can cast this in advance and go into battle ready to fling these meteors, tacking on an extra bit of AoE damage to whatever you did with your action. Its 12d6 of small-area damage, distributed over 3-6 attacks, is solid. If you cast this before entering battle, you’re doing awesome things for your action economy and damage output in the critical opening rounds of a hard fight. Its concentration requirement is the big tradeoff. For a bunch of fruitful points of comparison, hold this up next to crown of stars and see what an additional four spell levels gets you (starting with “no concentration required”).
Mental Prison (Sorcerer, Warlock, Wizard): This is a particularly brutal illusion: damage and either a solitary prison cell or a lot more damage. 15d10 from a 6th-level spell is a kick in the teeth, and it even does respectable damage on a successful save. Highly recommended.
Mighty Fortress (Wizard): This is both the ultimate expression of Leomund’s tiny hut and a downtime action unto itself. It conjures a fortress, a huge number of unseen servants, and food. I love this spell for what it says about top-end wizards and the implicit connections to Daern’s instant fortress and Mordenkainen’s magnificent mansion. I think that clerics – maybe only some domains? – should have access to this spell too… until I see that they get their own version, one spell level lower. Right, then. Okay, to dig for critique a bit, I would like it if wizards could invest something extra to add more to the fortress that they conjure.
Mind Spike (Sorcerer, Warlock, Wizard): Bards should have gotten this one too, I would say. It seems appropriate for its psychic damage and divining the location of a person in a city. You don’t see a lot of divination attack spells, so even if the damage is a few points shy of the target for a single-target spell, its secondary effect is interesting (phenomenal if you’re targeting a villain who is likely to try for a getaway), and Diviners with the Expert Divination feature are going to fall in love with it. Get some hyper-efficient spell slot use by casting this with a 6th-level slot, then a 5th, then a 4th… five spells for one 6th-level slot. Good times. Also, note the errata on its scaling effect: d8 rather than d6.
Mold Earth (Druid, Sorcerer, Wizard): If you need to dig up loose earth in a hurry and move earth isn’t available, mold earth is here for you. Much like gust, giving such similar spells such similar names seems like asking for user confusion to me. This can also create difficult terrain and create art in the dirt, so if you have a few minutes to prepare an area and a mattock to loosen the earth, you can do a whole lot. It’s probably fine, but it makes me think very much of my friend Dan Head and his efforts to explain combat engineering to me. That is to say, this is terrifyingly dangerous in the hands of a motivated player who knows what they’re doing.
Negative Energy Flood (Warlock, Wizard): This spell really needs to go to Death clerics as well. It’s a great example of XGTE delivering on the core promise of necromancy, in a way that the Player’s Handbook often doesn’t. It’s a single-target damaging spell that also gives you a zombie if it deals the killing damage, or it’s a nice temporary hit point infusion for an undead. The damage is 11.5 points below the chart’s target for a 5th-level spell, so you need to be confident that you’re going to create the zombie and get your Grim Harvest kicker to really make this pay off. The downside is that, because PCs get death saves, it’s not clear that they “die” from this spell unless it causes their last failed death save. Animating dead PCs is good fun, so I’d much rather see this work if a PC dies after being reduced to 0 hit points by this spell.
On that note, I’m stopping here for the week, having covered another 29 spells. I love a lot of what I’ve covered in this article, as we’ve seen several significant patches for the holes in the Player’s Handbook spell lists, especially in terms of coverage for each wizard Tradition. The druid and sorcerer are also big winners here. The warlock hasn’t done badly, but by chance, a lot of the most compelling warlock spells fall a little later in the alphabet. Shadow of Moil and soul cage, we’re coming for you.
