I still see a lot of conversation around the spell witch bolt, so I want to talk about why it’s bad in both straightforward math and in likely tactical situations. I’m also going to talk about easy fixes, possibly in multiple versions, just as I’ve done with a few other spells so far. I know from reading a lot of online conversations that this spell has its hardcore supporters. If you read what I have to say and still want to use this spell, do your thing – but informed choice is the best choice.
As a 1st-level damaging spell, its main competition comes from magic missile (which, in fairness, is a little too good – h/t to Samhaine for pointing how that it’s better than scorching ray unless the target AC is uncommonly low), chromatic orb (the best 1:1 comparison, because it’s single-target and uses a spell attack), and thunderwave. Cantrips are also a big deal for this comparison, especially poison spray and fire bolt.
Casting Time: 1 action
Range: 30 ft
Components: V, S, M
Duration: Concentration, up to 1 minute
The casting time is standard, of course. The range is noticeably short for a single-target spell; burning hands is deliberately high-risk, while thunderwave is specifically about getting people off you. Chromatic orb has an impressive 90-ft range, by comparison. The V/S/M components are pretty standard; the material component isn’t expensive, unlike chromatic orb. That duration is the spell’s defining element – it grants a new at-will action for the duration. Requiring Concentration is pretty rough – the other spells that grant new at-wills also require Concentration (moonbeam, call lightning), but don’t share witch bolt‘s further strengths or limitations. It’s important to keep in mind that a Concentration duration also means the target may be able to end the spell by punching you in the face.
A beam of crackling, blue energy lances out toward a creature within range, forming a sustained arc of lightning between you and the target. Make a ranged spell attack against that creature. On a hit, the target takes 1d12 lightning damage, and on each of your turns for the duration, you can use your action to deal 1d12 lightning damage to the target automatically. The spell ends if you use your action to do anything else. The spell also ends if the target is ever out of the spell’s range or if it has total cover from you.
At Higher Levels. When you cast this spell using a spell slot of 2nd level or higher, the initial damage increases by 1d12 for each slot level above 1st.
- The ranged spell attack means it does nothing at all on a miss – your spell slot is gone, and so is your action. Chromatic orb and inflict wounds are some of the only 1st-level damaging spells that include a spell attack; it’s considerably more common among cantrips, since you’re losing the action but not a spell slot. Since I’m going to keep coming back to poison spray, I’ll note that it uses a Con save. It does nothing on a miss, because it’s a cantrip, but given its short range, it’s relevant because it doesn’t suffer disadvantage if you’re casting it with an enemy adjacent to you.
- 1d12 lightning damage averages to 6.5 damage. Compare that to chromatic orb‘s 3d8 acid, cold, fire, lightning, poison, or thunder, which is a lot of adaptability (not that lightning resistance or immunity are common), and inflict wounds 3d10 necrotic damage. Poison spray deals 1d12 poison damage (one of the most common resistances and immunities). Since poison spray is a cantrip, its damage scales up to 2d12 at 5th level.
- The evocative image of the spell is that you can keep using your action to deal 1d12 lightning damage for the duration, without a new attack roll or save. It’s supposed to be Star Wars-style Force lightning, which is totally why this spell shows up on the warlock spell list. The problem is, dealing a guaranteed average of 6.5 damage per round is not nearly as good as it seems like it should be. For more on this, see below – it’s part of bringing all of the spell’s features together.
- The spell ending if you use your action to do anything else is completely unnecessary restriction. Thus, all of my fixes drop this clause real hard. Moonbeam and call lightning have no comparable restriction. If this is part of imagining the “sustained arc,” well, change that to imagining that you’ve got your targeting lock, maybe a thread of non-damaging light, that you can turn into damaging lightning when you exert your will (as an action). Sorcerers, warlocks, and wizards are all about different approaches to versatile problem-solving, and anything that gets you out of the mindset of creatively planning or reassessing the problems of each round of combat is just getting in the way. That’s more philosophical than material, though, so don’t get hung up on it as part of my argument.
- The spell also ends if the target is ever out of the spell’s range or if it has total cover from you. This gives the target creature two more ways to end the spell early. It can get 30 feet away from you, or it can move around a corner or duck behind cover. This might induce the creature to provoke attacks of opportunity, and/or spend its action doing something other than beating up your party for the round. That’s kind of a best-case scenario for you. Wizards in particular are not going to lead in damage output most of the time, so they need to focus on disrupting the enemy’s plan. The problem is that the enemy has three ways to respond effectively: breaking your Concentration, getting out of range, and getting behind cover. You can’t control which one they choose, and setting up a situation so that all three are good for you is really hard. If nothing else, getting punched in the face is really not what you do best.
- The damage scaling of a higher-level spell slot only applies to the initial damage. Obviously, an additional d12 per spell slot would be too much, but I suspect a different base die value could have created room for some kind of scaling after the first round. With no way to scale the damage after the first round, most cantrips surpass its damage output once you reach 5th level. They still have to hit with a spell attack or get an enemy to fail a saving throw, but it can be a lot more damage (cf. poison spray or, you know, eldritch blast with Agonizing Blast).
Given that last point, this spell isn’t worth the spell slots you’d spend on it after 5th level. Now, all 1st-level damaging spells eventually get outclassed by cantrips – but most of them don’t get so definitively beat until 11th or 17th level. Witch bolt still has guaranteed damage going for it, though, so let’s break it down a little further. This is going to get mathy. I’ll be comparing witch bolt to fire bolt, for being a cantrip open to sorcerers and wizards that just kicks out damage. If math makes your eyes bleed, skip to the next header and take my word for it.
Imagining an AC of 13 and a +3 ability score modifier – or any other arrangement that makes you hit on an 8 or better – fire bolt deals an average of 3.85 damage per round. (In case you’re not familiar with damage calculation, or in case I did it wrong: on 7 of the 20 attack roll outcomes, you deal 0 damage. On 12 of the outcomes, you deal your average of 5.5 damage. On 1 outcome – the natural 20 – you deal an average of 11 damage.) At 5th level, assuming defenses have scaled so that you still need to roll an 8+ to hit, fire bolt deals an average of 7.7 damage per round. (That’s (0 * 7 + 11 * 12 + 22 * 1)/20.) When it comes to average damage, 1.2 damage per round is enough to matter.
Just to show my work, a warlock with Agonizing Blast – the definitive damage-dealing warlock’s Invocation – deals an average of 5.8 damage per round with eldritch blast, assuming a +3 Cha modifier, 6.45 with a +4 Cha modifier, and 7.1 with a +5 Cha modifier. I… think we can all see that this is only going to get a lot worse at 5th level.
From Numbers to Tactical Reality
In the last section, I showed that witch bolt deals more average damage than fire bolt between 1st and 4th level, but falls off thereafter. Let’s take a step back and try to imagine realistic tactical situations in D&D. Fights almost never last 10 rounds. In the rare cases that they do, individual creatures really almost never last 10 rounds. Certainly not before 5th level! One can imagine niche cases to increase the chances of such a thing happening, like a chase or a cover-shooter situation… both of which witch bolt can’t handle, because of how it handles range and total cover. Now, that holds true for any kind of 1-min-duration combat spell, but witch bolt looks a lot less impressive at 3-4 actions, when you could come closer to affording to burn spell slots for instantaneous spells.
I think there are a bunch of good ways to fix this spell and give it a competitive niche. I’m going to suggest three. Anything I don’t change in the fixes below can be assumed to stay the same.
The Reliable Keyword
(In 4e’s format, powers could be Reliable, which meant they weren’t expended if they didn’t hit something.)
A beam of crackling, blue energy lances out toward a creature within range, forming a sustained arc of lightning between you and the target. Make a ranged spell attack against that creature. On a hit, the target takes 1d12 lightning damage, and on each of your turns for the duration, you can use your action to deal 1d12 lightning damage to the target automatically. On a miss, you can spend your action on following turns within the spell’s duration to make another ranged spell attack. The first time you hit within the spell’s duration, you deal the spell’s initial damage. You cannot deal damage on a round that your target is more than 30 feet away from you or has total cover from you.
A beam of crackling, blue energy lances out toward a creature within range, forming a sustained arc of lightning between you and the target. Make a ranged spell attack against that creature. On a hit, the target takes 1d12 lightning damage, and on each of your turns for the duration, you can use your action to deal 1d12 lightning damage to the target automatically. On any round in which your target takes damage from this spell, its speed is reduced by 10 feet until the start of your next turn. (Another option: reaction denial, as per shocking grasp.) You cannot deal damage on a round that your target is more than 30 feet away from you or has total cover from you.
A beam of crackling, blue energy lances out toward a creature within range, forming a sustained arc of lightning between you and the target. Make a ranged spell attack against that creature. On a hit, the target takes 1d12 lightning damage, and on each of your turns for the duration, you can use a bonus action to deal 1d6 lightning damage to the target automatically. You cannot deal damage on a round that your target is more than 30 feet away from you or has total cover from you. (Note: As this is an ongoing effect and not casting a new spell, you can still use your action to cast a cantrip or a spell of 1st level or higher.)
Stands-in-the-Fire also proposed a valid fix to this spell that converts it to something more similar to moonbeam and call lightning – a spell that chooses a 5-ft square to target, requires a saving throw, and “locks on” after a failed save. You can pick a new target at any time. I’m not writing that one out in formal rules language because I don’t want to implicitly take credit for his idea.
The Reliable Keyword fix is probably the simplest,of my approaches, and changes the spell’s overall application the least. It makes damn-near-sure you get that burst of initial damage, and thus improves the use of higher-level spell slots substantially.
The Force Lightning fix is less specifically about the speed debuff or reaction denial than about adding any kind of effect onto the damage, bringing it a little more in line with tactical control. Any effect that suits “target is wreathed in lightning and wracked with pain” would be appropriate here. Obviously, this can still lose out to a whole lot of ray of frost or shocking grasp, but with effects like speed debuffs or reaction denial, the automatic hit might be a bigger deal to you. Also, the speed debuff makes it a little harder to get out of the spell’s range, so you might be costing your enemy another action.
The Bonus Action fix is about using Concentration to manage a damage over time effect. It isn’t hex or hunter’s mark, but it has a not-entirely-dissimilar effect as an unavoidable damage kicker. This fix could plausibly be blended into one of the other two, pending some field testing.
In short… damn near anything would be better than witch bolt as written.