Larian Studios Forums Baldur's Gate III Baldur's Gate III - Suggestions & Feedback Focused Feedback: Spells

Dialogue encounter in Rivington: The spell Seeming becomes undone if active without explanation, so that the dialogue makes sense. This feels pretty bad...

Full details below:

Hi, this is a great thread to add to for commentary on the way spells work in general, but for specific uses of spells or bugs then it's probably best to create a separate thread in either the Technical & Gameplay Problems subforum. Note that's to discuss potential bugs and workarounds with other forum members, as the way to report bugs to Larian is via their support website. The Report a Bug link up along the top of the game launcher will take you to the right place.

Wall of Stone has 30 hp.

It is supposed to have 180hp, AND 15 AC.

Any character can just spend one turn to get through it, which makes it utterly worthless for a 5th level slot.

This is getting really tiring. Either this is the devs' hate for control spells manifesting again, or they don't bother to read 5e or care enough for a spell like this to be actually useful.

Another cool tabletop perk is that it becomes permanent after full duration. You could make defensive walls or create a bottleneck against an invasion. Something a Tactician would do in a CRPG against unfair odds. This is the kind of tactics I would want rather than having to fight against enemies with extra Bombs in their pockets you don't know are there, or a flat +2 to enemy attack rolls.

Gust of Wind does not disperse clouds even when the tooltip says it will.

It did the very first time I cast it, but never again. Tried about 20 times on different clouds from different positions.

Darkness replaces Cloudkill.

Two clouds need to be able to overlap and co-exist. Being able to counter a powerful high level cloud with a low level one that is also beneficial for your build is a bit too easy.

Um...I just used Greater Invisibility for the first time. You know, the 4th level spell that's supposed to be appropriately powerful for a 4th level spell? I attacked one enemy and then my character automatically attempted a DC 15 stealth check (which they failed because they have 8 Dex and are in Plate Armor) and the spell immediately ended. Of particular note, I didn't get any bonuses to my stealth check from being invisible.

I then checked the description for Greater Invisibility in BG3

What? Just...what!??! Greater Invisibility only lasts 10 turns and can already be broken by loss of concentration or unusable because you're already concentrating on something else. It could maybe be argued that it's a bit too powerful compared to other 5e spells, but it's certainly nowhere near powerful enough to need this severe of a nerf. (Especially when you consider the prevalence of auto damaging effects and the overpoweredness of Haste, a lowly 3rd level spell, in BG3).

As is, Greater Invisibility in BG3 is practically useless to cast on any character except rogues with Stealth expertise. For most other characters, it's effectively the same as the 2nd level normal Invisibility spell and worse than a regular potion of invisibility because of the concentration requirement.

Yes Larian hate spells except direct damage spells. As a mentioned very often many spells you can get via potions and items so wizards are useless better to have a four fighter party.

I recently made a thread that covered a number of pain points I've experienced when playing the full release of BG3, many of which included suggestions for spells and abilities. For visibility, I'm also posting the ones specifically related to spells here since it's stickied.

Minor Illusion:
Bit of a shame that it's only useful as a distraction, but the real issue for me is that there's no way for me to end the effect early. I've had a number of occasions where I use it to benefit from it for a round or two, but then need it gone. As it stands, I have to just wait until it vanishes, which sometimes causes issues with whatever I'm trying to do.

Mage Hand:
Both the EA and full release versions of this spell bug me, for different reasons. In EA the fact that it could be cast as many times as I wanted as long as I could afford to concentrate on it meant I was able to use it in early levels since I didn't need concentration for most of my spells, and it helped in combat a bit. Problem was that once I started really needing concentration for other spells, I stopped using it almost altogether. In full release, it's a different but similar problem. Now it doesn't require concentration, but you can only cast it once every short rest. This caused me to never use the cantrip, even at low levels, because now I feel the need to keep it on hand (hah) in the event I need its utility to grab a trinket or something. Outside of those niche instances, it's basically entirely forgotten about. IMO, there's no need to balance how frequently a player can use it, as the power given to it by its implementation in BG3 is roughly equivalent to that of 5e. Essentially, it traded out-of-combat utility for in-combat utility. Even if I could cast it endlessly, I don't think I'd always want to use my action that way.

Arcane Trickster Rogue's Mage Hand Ledgerdemain:
This trait completely lies to the player, and does not reflect how it's supposed to feel in the TTRPG. For Arcane Tricksters, Mage Hand Ledgerdemain is a core aspect of the class. Not only does it really need to be able to be cast as much as the rogue wants, it also needs to perform as advertised and honestly stick to the source material better. Its invisibility type should be Greater Invisibility, so just as in the TTRPG it always stays invisible if the rogue chooses to make it so. It also doesn't currently do much of anything other than what the normal Mage Hand does, despite what the traits claims. The only extra utility I've seen from it is being able to use it to lockpick or disarm traps at a distance, but unlike in the TTRPG doing so uses the Mage Hand's (terrible) stats rather than the rogue's Dex and Sleight of Hand proficiency/expertise, making this utility next to worthless.

Arcane Lock:
This is already a niche spell in the TTRPG. The way it's implemented in BG3 makes it even more so, for not really any good reason. It needs to last much longer (e.g. 100 rounds), and also when initially cast prompt the user for what kind of allow list they want. At the very least the player should be able to choose between "nobody passes through" and "only party members can pass through".
I feel it's pretty telling that this is a terrible spell as-is when absolutely no NPCs ever make use of it, and the scroll is hard to come by (as is the same with Knock).

Knock:
This is another terribly implemented spell. In the TTRPG this is meant to be a counter to Arcane Lock, following the general trend that spells with persistent effects have an opposing spell that can directly counter them (e.g. any cloud spells vs wind spells). This needs to be able to counter Arcane Lock, and it should also be bad for stealth if it isn't already, as it's supposed to be very loud. As-is, it's never really worth a 2nd level slot to make use of this since individual locked containers don't commonly have loot that's worth the spell slot cost. The value of this spell also hits 0 if you have a Rogue or similar in the party, because they can basically do this for free all the time.

Find Familiar (all variants):
Something I feel is missing from the Find Familiar spells is the ability to have your familiars deliver items and throw lightweight stuff. Having a tiny inventory and very low carrying capacity but the ability to toss water bottles and healing potions short distances for certain forms would be a welcome addition, as would being able to fit through tiny holes (currently restricted to Druids) to grab items or whatever. These would also fit the flavor that they have in the TTRPG. I realize this would be a bit odd to balance, but it would go a long way to making familiars feel more involved. As they're currently implemented, familiars feels pretty useless as anything but a small meat shield in combat, and they have limited utility outside of combat as well.

Starving for Rituals:
There appears to be very few rituals in the game, which causes it to easily be an overlooked mechanic, and also turns Ritual Caster into a feat with questionable value. Would be nice to see more rituals implemented, or more existing utility spells converted into rituals.

Polymorph Effects Shouldn't Prevent Ability Activation
Circle of the Moon Druids are the biggest offenders here. In the TTRPG, the way you're generally meant to play a Moon Druid is to use a mixture of their Wild Shape and their magic simultaneously. A first turn for a Moon Druid generally looks like using their Action to cast a concentration spell (e.g. Flaming Sphere) and then use their Bonus Action to Wild Shape, allowing them to continue concentrating on that spell. BG3 sort-of allows this, but with a catch: if the concentration spell can be re-activated using an Action or Bonus Action, you cannot do so while in Wild Shape because for some reason that counts as casting a spell and in the beast form you are silenced. This prevents comboing Wild Shape with classic druid spells like Moonbeam, Heat Metal, and Call Lightning. It feels more like a developer oversight than an intentional design decision, honestly.

Transfuse Health + Warding Bond Interaction is Broken
Warding Bond is supposed to halve all of the player's incoming damage so it gets shared with the person who made the ward. Transfuse Health damages the player by half of their current health and then heals their target for the same amount. The problem here is that Warding Bond does not halve this incoming damage, but still responds to the damage in turn. So if I'm warded with 20 HP and use Transfer Health on another character, I would expect that both my warded character and the character that's warding them to take 5 damage (10 total, half of the HP) and then the target of Transfer Health to get healed for 10. Either that, or for Transfer Health's self-damage to not be eligible for Warding Bond at all, and act as though the effect isn't there. What happens instead is that in this same scenario, the warded character using Transfer Health takes 10 damage, the character who used the ward take 10 damage, and the target of Transfer Health heals 10 HP, effectively doubling the downside of Transfer Health and making it never worth using, which is rough because it's already pretty niche. As with my previous gripe concerning concentration spells and polymorph effects, this feels more like an oversight than intentional design.

Goodberry Needs to Remember Who Cast It
This mostly matters for gear interactions, such as with Wapira's Crown and The Whispering Promise. If my druid has these sort of items on, I'd really expect any character who consumes the Goodberry to trigger the effects of these gear items, because the druid cast a spell that healed an ally.

Wood Woad's Regeneration to Work as Advertised
Recently got ahold of Conjure Woodland Beings, and while experimenting with it I noticed that despite the Wood Woad supposedly healing 10 HP per turn that it's in a vine surface (provided it doesn't take fire damage), this doesn't actually seem to work at all. I hadn't tested it in combat, though, so perhaps it's combat-only. But, if it is combat-only, this needs to be messaged to the player.

Polymorph is Extremely Disappointing
With all of the other spells that offer options, why was Polymorph restricted to turning an enemy into a sheep? Why not just treat it like the Druid's Wild Shape, and give creature options that scale with the target's level? The main draw of Polymorph in the TTRPG is how extremely versatile it is, and how it can be used by arcane spellcasters as a "pseudo-heal" for allies in danger. The implementation in BG3 doesn't offer any of this, instead opting for easily the least-used application of Polymorph, which is uncommon to see use because in the TTRPG other spells do the same thing but more effectively and with a better stat target (e.g. Banishment targeting Charisma). Polymorph can be used to do this, but it's worse than a spell that specializes in it. Polymorph is meant to be a spell that's a multitool, and that's its main draw. By taking out the soul of the spell, it's been completely neutered and is basically never worth using.

But have you telekinesed an owlbear yet?

In my tested experience Fog Cloud does not work as intended. In several different ways this spell does not function as described or as is its known abilities in 5e. In fact, its quite exploitable. The first issue is that fog cloud SHOULD create an area that is heavily obscured. While this works isometrically it fails when it takes into account line of sight and the Z axis. Creatures hiding behind a fog cloud would also be obscured from creatures that cant see over the height of the cloud. Simple line of sight rules. Currently a gnome can shoot his crossbow right through it at another gnome on the other side with no penalty. This now brings a bigger issue into play which is that Fog Cloud in BG3 is currently being countered by darkvision. I am assuming this may not be intended as the debuff created is "too dark" and may just be a copy paste situation from the darkness spell. Fog Cloud should "obscure" not cause things to be "too dark" for obvious reasons. This turns Fog Cloud into a ridiculously over powered ranged offensive blind with no penalty when metagamed properly. I even tested with a potion of darkvision and sure enough the disadvantage went away. This is an even larger problem because fog cloud blinds everyone in the cloud, which in BG3 is giving advantage. So not only are you blinding and giving disadvantage for free, you are also getting advantage for free as well. I don't really see the case for giving advantage against blinded targets for ranged attacks since your chances are really changed or hindered by the targets ability to see? But that is a somewhat different issue that just makes this one so much worse.

Please un-nerf crowd control spells such as Hypnotic Pattern. It's supposed to have a duration of 10 rounds, not 2. Even if we could compromise on 5 it would be an improvement.

Concentration spells are already overbalanced by the AI's propensity to go after casters, especially on tactician.

Confusion gives its targets an extra saving throw. They get one saving throw to resist its status effect, and then on the start of their turn they get another saving throw to resist it again. This means they have to fail 2 throws before actually suffering any ill effect whatsoever.

They should not have that extra throw on the start of their first turn.

Misty Step should allow you to escape physical forms of restraining. I was ambushed by Blights in Act 2 and they used an ability that grows roots and "Constricts" you. I tried to Misty Step out and the roots TRAVELED WITH ME. This completely negates the point and purpose of Misty Step and other forms of teleportation. I haven't seen interactions with other examples or forms of teleportation, so I can't speak to that.


Yeah, in 5e they get their save at the END of their turn. Having it at the start just feels bad.

So all the way until just now, I thought there were simply no undead crowd controls in this game.

Just found out on reddit that oathbreakers have control undead, and it works on a lot of bosses too.

Like we have hold person / monster, dominate person / beast / celestial or whatever its called, but no halt and control undead for clerics / wizards / sorcs.

But if we did, every single thing in the game would be trivialized :x

I think Larian got spells as right as they can for a single player adventure. Folk have already found game-breaking character builds.

There will be more patches to further refine and balance spells, and I think that they should really optimize Spells for levels 5-9.

Once past level 10 meaningful encounters are hard to make. Level 10 D&D characters are VERY powerful. If I have two attacks I can cast Conjure Elemental and then Cone of Cold for 8-64(!) damage. Or Cone of Cold twice in a turn is taking almost all Act one bosses/minibosses in one turn.

As a dev you run VERY quickly into making interesting things to challenge these New Godlings. The cheap route is enemy leveling with you. Bleh. Or you can have more "standard" mobs. Large battles! Well it turns ouit waiting for twenty enemies and conjures and summons to take thier turn is about the limit to the Engine and my patience.

It CAN be done, and I only trust Larian to make quality high level adventures.

I do not think we will see it though. Not without nerfing the higher level spells to the point where many would rather they didn't. Seventh level is where the "high concept" spells like Reverse Gravity and Plane Shift start. Such things can only live in our imaginations.

For now.

Simply the best CRPG yet.

I'm trying to understand why Chain Lightning and Dethrone have the same damage range but one is a single target spell while the other chains to three more targets, or why Art of War has a 100% hit chance and is on short rest one cast while Dethrone is one cast per long rest and like 2 higher damage only while having no 100% hit chance.

Where is the logic where multitarget spells are doing the same or more damage as single target spells? Sure, you can claim Dethrone is listed as lv5, but that's because it can only be learned through a scroll, for some reason unavailable to necromancers and spore druids at lv6. But even taking a spell like Blight upcasted to lv6, it basically goes to the 80 max damage range, is single target, and is subject to a save for half damage.

Sun Beam, same lv6 spell, narrow line aoe with friendly fire, and it does like 40 damage max a cast, really pitiful for a spell, especially considering that the idea that the recast factor will compensate is just a farce because a single cast of chain lightning or two is enough to fry at least half the room of mobs and reduce a boss to half health, whereas that sun beam will need to be cast six times to do the same damage, and it won't even hit the same amount of mobs.

Same goes for the lv6 Otiluke Orb ice spell, no way it can compete with chain lightning. Chain Lightning, Magic Missiles, and Art of War are just on a league of their own while the other spells are really underbudget for their spell slot cost. Necromancy and druid spells got done really dirty in the damage department, although Disintegrate is equally terrible, with a higher damage potential than chain lightning single target, but shorter range, it has to pass a DEX saving throw or it does zero damage, and even if it hits, it does less damage than Magic Missiles with clown gloves or Art of War.

Each spell of the same level being just as good as every other spell of that level, I don't think this has ever existed in any DnD game, or even in PnP.

Players end up creating games where some spells are banned, or you must use the spells the DM says, or just playing weaker characters for fun - this works in the context of PnP because you can make the game as challenging as the builds you are using.

In a video game, it is always the same content, so if one spell is clearly superior to another, no one will use the weaker one unless again their trying a self imposed hardcore run or something.

Regarding chain lightning, in BG2 it was a bottom tier spell, one of the lowest damage spells by far, and every damage spell was outclassed by level 3 skull trap. Same with Delayed Blast Fireball, simply a waste of a level 7 spell slot. Meanwhile in BG3, chain lightning is mega OP top spell to have, meanwhile enchantment and illusion spells beyond the hold spells are just trash.

Well it doesn't help that enchantment and illusion spells, like many CC spells including most druid CC spells and utility spells like Dancing Light are crippled by a nonsensical concentration mechanic, and they consume spell slots regardless of whether they pass a save or not, and unfortunately most saves in this game are DEX or CON saves, of which mobs are high on, plus all the bugs people have mentioned like checks taking place both at the beginning and end of a turn, and how drastically the duration of CC was reduced on top of being crippled with the concentration requirement.

Then there's the Jump garbage making so many of the static environmental CC's like vines useless. You can drop a spike growth or vines or a Darkness, and mobs can just trivially jump out of it in one turn, or if an aoe like fireball goes off, it burns ut the vines or spike growth without at least some combination explosion bonus to account for the fact that a spell you wasted a spell slot on and was bound by concentration was erased from the battlefield.

I really hope they do a spell balance pass and look at Haste and how it works with actions so it doesn't just duplicate martial extra attacks. They could even do a half damage fall off on chain lightning, but let's be honest it's a two-three spell cast per long rest if you got the legendary staff only or the very rare Act 3 staves, otherwise it's one spell cast per long rest. And it would only nerf the single niche you would want to bring a spell caster over a martial class.

I also don't think spells should be gutted just because they made metamagic as a sorc mechanic so incredibly stronger than the other casters. Blanket nerfs to spells affecting already garbage classes like caster warlock make no sense, so if you gotta nerf something, you could just apply a damage fall off to the metamagic mechanic so if you duplicate the spell or cast another spell as a bonus action, they do half or a third of the damage of the original spell. Then balance the spells from there.

And give druids a proper spellbook with Bless and Chain lightning as a lv6 spell while Spore Druid should get Dethrone alongside Necromancer. Some spells like Cure Wounds and Healing Word also need buffing, because they're really bad unless you got a Life Domain cleric passive propping them up. It's crazy how much better tossing a potion is for heal quantity than a freaking spell that takes up a spell slot plus an action/bonus action. The Spores spells for spore druid also need numerical buffs, they fall off in Act 3 really hard, a 2d8 in Act 3 is just pitiful, and being unable to turn off spore clouds while these clouds affect NPC's and can turn them hostile is dumb, so make them not affect neutral NPC's if you can't turn them off.

It can be argued that "Dethrone is listed as lv5"? Dethrone flat out is a level 5 spell. Chain lightning is level 6.
Dethrone deals 10d6+20 (30-80, average 55) necrotic damage, CON save for half, single target, 30m range.
Chain Lightning deals 10d8 (10-80, average 44) lightning damage, DEX for half, chains to 3 others within 18m of original target, 18m range.
Necrotic is ostensibly a rarer resistance than lightning. CON vs DEX is not an advantage either way; depends on targets (both CON and DEX are common saves). Before regain, free cast, etc. you only get the one level 6 slot, whereas you get 2 level 5 slots. That seems to make sense, yeah; the 6th level spell has better multi-target performance, while Dethrone has better single target and range.

Sun Beam deals 6d8 (6-48, average 27) radiant damage, CON save for half, 18 meter line. It also blinds those caught within.
Twice that, as it allows to be recast once, and your per-slot damage is comparable to Dethrone, while having an area of effect, and it lets you blind people for two rounds. Sure, the damage is comparable to a level 3 Fireball, but two rounds of blinding AoE is nothing to scoff at, especially since it is unconditional (making the save just halves the damage).

Otiluke's Freezing Sphere is cool to give to a support caster before a fight (casting it gives you the Sphere item, that you can toss). Damage is not impressive, but the ginormous area of effect makes for a nice opener. 5E expects fire and cold damage to be roughly "on par" in potential and resistance popularity, but this game features a relatively small number of enemies that have cold resistance.
The clown gloves are broken; no sense in commenting about an item that doesn't work as intended.

Dethrone is only available via scroll learning for wizard from the Ramazith Tower with a rule of one per long rest cast, and is once per long rest. It's not subject to lv5 spell slot rules. And even then, with a wizard's arcane recovery and the legendary staff you can get out 3 chain lightnings vs. 5 lv5 spells, and necrotic resistance in this game is actually WAY more prevalent than lightning resistance. And since it's not available as a regular lv5 spell, it is a single use affair that can't even be used with arcane recovery or the legendary staff arcane battery passive.

What you also forget is that the legendary staff gives Kerestra's blessing, which adds a proficiency bonus to your lightning damage, so add that to your calculation, and add the fact keresta's also grants you an extra chain lightning cast for free plus a lightning bolt. Kerestra has no equivalent blessing for a necrotic spell pool. Even then, a spell that can affect 4 targets, and most encounters will have at least that amount of targets in this game, makes Chain Lightning practically better in all of the game than Dethrone and the entire necromancy spell pool. Especially if you add the bonus with wet surfaces, while necrotic damage gets no such bonus.

The blind from Sunbeam is underbudget and doesn't even work on some boss enemies, and it's not like you're using a lv6 spell slot to blind a target. 27 average radiant damage on 90+ HP enemy health pools in Act 3 is pathetic.

The area of effect is of hardly any practical value for Otiluke. If I want a large prone area of effect, I'll cast Ice Storm to open and save my precious lv6 spell slot for a follow up chain lightning.