Are you familiar with the moonbeam spell? There’s several like it: blade barrier, cloudkill, cloud of daggers, Evard’s black tentacles, forbiddance, sleet storm, spirit guardians. What do they all have in common? They all share tricky little mechanic, underlined in red below:
Do you think this spell takes effect when you cast it?
You might be surprised to find that it doesn’t!
D&D Sage Advice clarifies that spells with moonbeam’s “enter” wording are not intended to proc when they’re cast (or moved onto a creature):
For most spells, this distinction only matters on the first round, since their area can’t be voluntarily moved like moonbeam. Spirit guardians will follow you around, since they’re centered on self. Cloudkill will drift away from you at 10 feet per round. The rest are stuck in place. But moonbeam can be dropped directly on to a new location every turn by spending your action.
The problem with moonbeam not working when it’s cast/moved is twofold:
- Mechanically, it’s ambiguous whether the creature entering the moonbeam’s area and the moonbeam entering the creature’s area are the same.
- Narratively, there’s an incongruity between the spell not working at first, then working when the creature later enters the space.
Mechanically, you can understand how “enter” is meant to be a one-way street. It only accounts for the creature moving into the space of the moonbeam. It is not intended to cover the moonbeam entering the space of the target.
It would be completely fair to assume that the creature and the moonbeam entering each other’s space are one and the same. Often, in the case of a rules ambiguity, DMs like to favor players, and players are happy to argue for a favorable interpretation.
If you have a good feel for 5e’s game balance, it quickly becomes apparent that obliging this interpretation leaves the spell a bit overpowered, since you guarantee the spell will proc twice on the target. That sets Moonbeam’s base damage at 4d10 for each target within its initial area, plus damage on the remaining rounds of combat. Compare that with the Dungeon Master’s Guide’s recommended damage for a 2nd level spell:
While moonbeam’s 5′ radius cylinder can fit up to four creatures laterally, the Dungeon Master’s Guide suggests approximating one target based on the spell’s size. For the sake of analysis, we should use the single target damage. Even then, doubling up on the first turn makes moonbeam outperform a standard 2nd level spell, even if the caster immediately drops concentration on their next turn.
DMs shouldn’t have to do that amount of system analysis to explain to a player why they’re taking the disfavorable interpretation. Errata could clean this up with new language instead of leaving it to Sage Advice.
Narratively, the distinction doesn’t really make sense. If the creature is in the spell’s area of effect, the spell should work! After all, that’s how it works when its not being cast or moved.
If the moonbeam deals damage from continued exposure, that would explain why a creature makes their save at the start of their turn. But, targets also have to save when they move into the area. In fact, because the spell does not have a “voluntary movement” clause, a creature can be pushed or pulled through the area multiple times in a round, stacking damage like they’re being dragged through caltrops. If moonbeam deals damage because of prolonged exposure, it shouldn’t proc multiple times in a round.
This narrative disconnect owes in part to D&D simulating simultaneous actions in a turn-based fashion. The beginning-of-turn save leans on the simultaneous narrative, while the enters-the-space save leans on the turn-based mechanics. There’s nothing wrong with leveraging both these ideas in the same spell, so long at it matches up at every point. When the spell doesn’t work on cast, it highlights the discrepancy.
There’s a couple approaches you could take to this spell. By changing a verb, you can align the spell with the designers’ expectations. You can also fiddle with the numbers to balance a result with a more narrative consistency.
The Word Swap
We can alleviate the mechanical confusion by changing one word: “enter”.
Swap “enter” with “move”. It’s clearer and matches 5e’s system language.
There are only two situations in which moonbeam is meant to cause a save. Either a creature moves into the area, or they start their turn in the area. When you cast moonbeam on a creature, there may be some debate about whether that creature has entered the area of effect, but everyone would agree that the creature has not moved into the area of effect. Generally, the stronger word for rules text is one that has fewer possible interpretations.
“When a creature moves into the spell’s area for the first time on a turn or starts its turn there…it must make a… saving throw.”
Of course, there are exceptions. Sometimes simpler words become loaded with other implications because of how they are used elsewhere in the game design. Notably, move is not such a loaded term. In 5th Edition, move does not mean “voluntarily move” unless specified in the rule. Proccing on forced movement is an intended spell feature noted by Sage Advice that we want to preserve.
With the language cleaned up, you can preserve the balance and intent of the original design.
For narrative consistency, moonbeam should cause the first save when it’s cast, and successive saves when moved onto a creature. Instant effects are more fun for spellcasters, and they should get something for spending their action to move the area. We’ll need a little more work to tweak this language.
“Once per turn, when you cast or move the spell’s area onto a creature, a creature moves into the spell’s area, or a creature starts its turn there…it must make a… saving throw.”
As we covered above, doubling damage to the creatures you drop it on reaches outside the scale for a 2nd level spell. So, we also need to trim back the damage a bit.
You also get more consistent damage spreading it over more saves. Lower damage may be a fair tradeoff for consistency, which you get from spreading the damage over a greater number of saves. But, players will need a certain baseline to consider a spell competitive with its peers.
Halving the damage is tempting, because it’s clean. The current spell only gives us two dice to work with (2d10), and we can expect about twice as many procs. You can easily stop there, but the damage might fall a little short, since you’re also halving the damage for enemies who get pushed through the area. Rolling a single damage die also feels a little unsatisfying for a 2nd level spell.
You can also change the damage dice. This gives you a little more room to fine tune the damage. That can be useful, since we really want to be somewhere between 2d10 and 1d10 damage. Adding more dice also makes the spell feel more powerful.
Note: adding more saves can slow down the game. Though, if a spellcaster is not moving their moonbeam, they are likely to be doing something else with their action that will take a roll. One way to counteract this slowdown is to remove half damage on saves. Be careful with this, especially when attacking CON saves, which are the strongest for monsters in 5e, as it can make the spell ineffectual at higher levels.
Half on Save. Try setting moonbeam’s damage at 2d8. That’s an average damage of 9, shaving a couple off the original spell.
Flaming sphere clocks in at 2d6 (average 7). It can be moved as a bonus action, forces a save when moved, and covers a larger lateral area (9 squares vs. 4), sheds light, and sets things on fire. However, flaming sphere doesn’t go vertical and it only damages a creature who ends its turn within 5 feet. While that’s a big timing difference, I think the action economy ultimately justifies a small damage boost.
Cloud of daggers is another 2nd level spell functioning on moonbeam’s same mechanic, so it should give us an idea of a baseline. Without getting into how many squares a cloud of daggers covers, we can use the DMG assumption that tells us to divide the cube size by 5. That’s the same number of assumed targets, at the same spell level at moonbeam. Notably, cloud of daggers can’t be moved, meaning it can’t be used to exploit the double proc trick past the first turn. So, it’s damage should be higher, right? Nope! Cloud of daggers comes in at a paltry (but thematic) 4d4. At 10 average damage, that’s lower than the original moonbeam damage. Why so low? Because there’s no save! Cloud of daggers is like magic missile–no roll required. It’s hard to divine how much the designers adjust the damage for auto-hits. A level 1 magic missile does 10.5 average damage, which is just under the recommended damage for level 1 spells. Note that iconic D&D spells are often permitted a little damage boost, as in the case with the legendary fireball. These nuances make cloud of dagger squishy as a baseline.
No Damage on Save. If you want to cut the half-damage rider, the DMG suggests increasing the damage dice by 25%. Try setting moonbeam’s damage at 3d6. At 10.5 average damage, it’s only a half damage short of the original spell’s 11 damage. Remember that removing the half damage rider also pushes the effective damage down (to 8.4 average, backing out the DMG’s 25% rule).
Which fix do you like better? If you’re curious about damage adjustments for the other moonbeam-style spells, there’s a chart up at the ThinkDM Patreon.