How WotC Already Found the Future of Spellcasting Monsters
The latest spellcasting monsters still have design issues. Let’s look at an alternate approach.
Summarizing the Issues with Spellcasting Monsters
As we saw last time, while Monsters of the Multiverse is overall an improvement, even the latest design has issues:
- Many spellcasting monsters are still too complex to run for any DM, let alone new and casual DMs.
- The optimal choice between non-spell actions, new magical attack actions, and Spellcasting actions is not always clear.
- The choice of spells in the Spellcasting block is usually unclear and forces us to look up spell details in other books.
These issues aren’t trivial. They create work for a DM, and may intimidate them to the point where they decide not to DM (as it did for me with 3rd edition). And, we should ask… what are we gaining from this design?
We can review our goals from last time with regards to spells on monsters:
- We want to capture the wondrous, chaotic, and dangerous nature of magic.
- We want magic to reflect the mechanics of the game.
- We need to weigh the above against how easy and fun it is to run spellcasting monsters, especially for a new DM.
- We want spellcasting monsters to achieve the desired challenge level, based on their CR.
Why Spell Lists?
That second bullet above? It drove 5E spellcasting monster design. Most monsters essentially have levels in a spellcasting class. Take a look at the Transmuter from Volo’s:
We can clearly see that this monster is a 9th level Wizard with the Transmuter subclass (it even has a Transmuter’s Stone feature, and the number of spell slots are exactly those of a 9th level wizard). The whole point of this design is verisimilitude, to make the monster feel like the actual PC class. That’s a neat goal. But here is an interesting thing. In 5E, a 9th level Wizard and a 9th level Cleric should be equals. And yet, the 9th level Transmuter is CR 5, while the 9th level War Priest cleric is CR 9. Why? It goes back to how WotC has that secret spreadsheet and how it calculates spell damage. The different spells drive different CRs. But with huge spell lists, the DM has trouble choosing the right spells that drove that CR!
That’s one of the reasons why WotC has changed the design in Monsters of the Multiverse to use smaller spell lists and shorten the spell lists. This is the new Transmuter:
We no longer have the same spell slots as a level 9 wizard. And we now have the ability to make three “Arcane Burst” attacks. They aren’t spells. You can’t counterspell them, and there is no such spell that a PC wizard could choose.
So… I have to ask… if that verisimilitude no longer matters, why bother with spell lists at all? If they are no longer actually accurate, and they create so much work… what is the spell list achieving?
Forged from Fire and Ice
Two designers I was discussing this with pointed something out to me. The design in Multiverse and even in Witchlight isn’t new! We can actually find it in Baldur’s Gate: Descent into Avernus and the adventure in the Essentials Kit, Dragon of Icespire Peak. Both of these were published in 2019… and I actually like aspects of that approach better. (Note: sadly, none of the monsters I discuss below were reproduced in Multiverse.)
In Icespire we can find the Rock Gnome Recluse. While he has a spell list, his actions actually state what two of those spells do. These are actually reproducing the spell text, so we get that verisimilitude (Magic Missile even expends the spell slot, though I don’t think that’s necessary design).
Baldur’s Gate: Descent into Avernus also features this design. The monster Black Gauntlet of Bane has the action Guiding Bolt, and the text contains the critical wording from the Player’s Handbook needed to run the spell. The same is true of the Skull Lasher of Myrkul, who has the Ray of Sickness action and faithfully reproduces that spell’s text.
The best example in the adventure is the Master of Souls. While it still has a Spellcasting block, we can practically ignore it. Take a look at the complete stat block:
That’s a relatively big stat block, but it appears larger than it actually is. There is the Grave Magic feature to turn spell damage into necrotic, plus the Action for its Flail attack. And then we have three spells. Each is an actual Player’s Handbook spell, with the text needed to run it.
Do we need the Spellcasting section? According to WotC designers, a monster is expected to last around 3 rounds on average, sometimes 4. I see 3 spells in that Action block. What if we remove the others?
Here is that stat block with the Spellcasting section removed:
I like that! Yes, we lose the flexibility of the 11 spells and spell slots it once had. But we also pare down our spellcaster to just the actions that were intended to be used anyway. This is still a capable monster with a variety of options. It isn’t overwhelming. And, it becomes really easy to correct the damage and turn this into a fearsome monster (it should deal 27-32 damage per round, so Chill Touch and Ray of Sickness need to deal higher damage). If we want to prevent a spell from being used too often, we can up the damage and add a recharge. This emulates having limited spell slots.
By having fewer options, we can easily design them to all be useful and relevant, removing trap choices and confusion. By using abbreviated but faithful spell text, we gain that verisimilitude back. Players aren’t tracking spell slots. In the three or four rounds this monster will be alive, it will seem like a true caster, slinging up to three different spells. It wasn’t going to do more than that anyway! Because the spell text is accurate, players will recognize this monster as a sorcerer or wizard.
You may notice something else. I didn’t remove the original text saying “1st-level Spell,” “Cantrip,” and so on. This will make the Action a spell, and allow counterspell, dispel magic, antimagic shell, and similar spells and rules activate normally against them. (I’m not a fan of counterspell, but many folks are!)
Already in Playtesting
At the start of the pandemic, my son and I began writing the adventure The Clockwork Tower. In it, there is a spellcasting monster. Long before we had pondered anything in this blog post, my son and I looked at that monster and decided to simplify. I don’t want to spoil the foe, but they have three actions which are spells, plus a reaction that is a spell. We have been playtesting the adventure, and that foe has been very easy to run effectively. Not once did a player fail to see this monster as a spellcaster.
When a monster lasts three rounds, spellcasting actions are really all we need. With them, we can achieve all of our design goals. And, it isn’t that different from the design that WotC has employed in the past. I hope we might see something like this in 2024, when 5E is revised.
22 comments on “How WotC Already Found the Future of Spellcasting Monsters”
Leave a Reply to Alphastream Cancel reply
Support This Site
Mastodon
BlueSky
A minor note here. In recent 5E design, there is guidance that the Spellcasting Action should have no relevance to a creature’s CR. This reflects the idea we mentioned last time, that all monster options should be valid. If it attacks with a sword and casts spells, then spells don’t make it stronger or weaker.
In the path I’m recommending above, that’s not necessarily true, because these spells are now truly treated as actions. They therefore reflect the action (for example, if they have recharge or limited use, they can deal higher damage).
I think the approach you’re suggesting works really well for monsters & NPC spellcasters that are only going to stick around for one battle. I do think you’ve made the Master of Souls a bit less effective by taking the poor guy’s shield and misty step swap away though!
I’d like to see the Spellcasting section retained for more powerful NPCs but kept to utility spells. As a DM and designer I often need a central villain for an adventure who can appear several times and also get up to stuff in-between battles with the PCs. These characters need to have more options.
I don’t really mind a utility-only spellcasting section, which is what was first discussed on the D&D Studio blog. But I would argue that it should be placed in the monster description rather than the stat block. I don’t need scry, alter self and similar spells in a stat block.
It’s interesting to note that 5E monster design guidelines direct us to have no more than 6 actions, and to generally use far less than that. And yet… we can see 30 spells on a spell list. I think if we have a melee option, a cantrip ranged option, and 4-5 spells… that really captures all we need. Shield as a reaction and Misty Step as a bonus are still plausible features (though 5E guidelines state that Bonus Actions and Reactions should appear only rarely on monsters).
5E’s big villains tend to have huge stat blocks. We can see that on Zariel or the demon lords. But then we have to ask ourselves what we will really use. Zariel has that flail, the longsword, and the Horrid Touch… and Spellcasting with 9 spells. Can DMs really tell the difference as to when to use one of those spells vs the stated actions? I would rather drop one of the melee options and stat out 4 really useful spells. The stat block might be slightly larger, but we now have a foe that is far easier to run and I doubt we lose much. It will also mean the damage will all line up with the CR, regardless of the actions taken.
Yep, maybe these could utility/non-combat spells could go in a sidebar? I think it would be good to have them stand out from the lore section.
Great article, Teos. One of the things I find frustrating about 5E is that, in its desire to distance itself from 4E, it unlearned some of the lessons from 4E that worked. Monster stat blocks were so much easier to use as a DM in 4E for exactly the reasons you described in this article – everything you needed as a DM was right there in front of you. I also loved how the stat blocks were presented right with the encounter so you didn’t have to flip back and forth between the map and terrain features and the monsters (like you have to for 5E).
I really liked 4E, though I recognize that the 4E approach could be a step too far. A lich might have “Shadow Ray” and “Frostburn” as a sort of necrotic version of spells, rather than a spell list. I like that, but I think we can stay closer to 5E design if we generally replicate spells as actions. The 5E lich has 26 spells, but we really only need about 4 of them to capture what a DM would use in a typical battle.
I liked having the stat blocks with the encounter too. Where this didn’t work so well was when the encounter itself was separated from the rest of the room descriptions. They did fix this in later Dungeon adventures but it led to a lot of page flicking prior to this.
Oh, yeah! Wow, that was so rough when they did that!
I really appreciate the thought you’re putting into these analyses. I hope the folks at WOTC are taking notes as they finalize the new 2024 designs!
Thank you!
This is exactly what we did in A5e Monstrous Menagerie. Significant attack spells are in the stat block, along with their slot level; utility spells (and there are fewer of them!) are relegated to Spellcasting. I hate having to refer, not just to a different chapter, but a different BOOK to run a single monster.
In MoMe we also include the spell’s components. Nothing worse than getting hit by a Silence spell and then having to look up 19 spells to find the ones without a V component.
Absolutely! Specific to what was done in Avernus and Icespire, it is super interesting to me that the current monster design guidance from WotC says something like: we experimented with reproducing spells as Actions, don’t do that. I am curious why they made that decision, because it’s to me much more useful than a spell that sort of feels like a PC spell but not quite, or using a spell list.
I also really liked that idea on the D&D Studio Blog of spell lists being just utility. But in that case, I would put it back in the Features section rather than Attacks.
Since the example of Arcane Burst is that of a non-dispell-able “Spell Attack”, would the text be better off without “Spell”?
“`
Arcane Burst: Melee or Ranged Attack: +6 to hit, reach 5ft or ranged 120ft., one target. Hit: 19 (3d10+3) acid damage.
“`
For the Transmuter, I wish the official version had a Spell designation and level (or Cantrip) entry, as I drew up for the hand-0edited Master of Souls. So, it would look like:
Arcane Burst (Cantrip): Melee or Ranged Attack: +6 to hit, reach 5ft or ranged 120ft., one target. Hit: 19 (3d10+3) acid damage.
This would allow DMs to still see this as the spell that it is, and it would interact properly with all of the 5E rules (Oath of the Ancients paladin, Counterspell, Dispel Magic, Abjurer, etc.). The entry would tell us what this is (which may also be important so we can roleplay what the monster is doing) and then if a DM wants to houserule that counterspell or something else doesn’t work, they can choose to do so.
“are still too complex to run for any DM” You state this as a fact but make no effort to explain why. Frankly I don’t see it. I’ve been a DM for almost 35 years and have never had a problem with spell casting monsters. In fact it’s never been easier. In 2e I’d run dragons that would have access to nearly 2 dozen spells, each of which you’d have to track individually. It was no more difficult than having a small stack of index cards prepared ahead of time. If you’re looking up spells at the table you haven’t spent any time prepping for the encounter. It’s pretty simple to just have a selection of index or actual spell cards just stuck in the book ready to go.
The problem that I want a solution for is the player who takes 15 minutes deciding which spell they’re going to cast because they haven’t bothered to learn what each one does.
“we experimented with reproducing spells as Actions, don’t do that. I am curious why they made that decision, because it’s to me much more useful than a spell that sort of feels like a PC spell but not quite, or using a spell list.”
Shapechange is the reason you don’t replace spells with actions that mimic spells. High level Wizards don’t need any buffs…
I see a lot of players and DMs, from conventions across the US to online games, from store visits when I travel around the country, to online forums and Discord channels. I see more DMs who are trying to find their confidence than those that have it.
Telling them something is “too complex for any DM” probably doesn’t help with that. Especially when it’s easier than it’s ever been.
The solution from BG:DiA was perfect. Have the full spell list for people who want to know all the monster’s options while including its most likely (read: spammable or optimized) spell choices included in the action section for people who don’t want to study the whole list. Full stop.
What WotC has done since this perfect solution has been nothing short of a backslide toward the terrible design that caused the community to reject fourth edition. The changes are not even one step removed from the return of the doombolt-spamming lich (https://pbs.twimg.com/media/FGmL5R_UcAITUEl?format=jpg&name=4096×4096). Yuck.
I will never, ever run spellcasting monsters the way they appear in the new stat blocks. If I ever run adventures with them, I will fix the stat blocks and publish them to DMsGuild, provided some noble soul hasn’t already done so. I refuse to stand by and let 5e revert to 4e without a fight.
I am surprised by this. I ran that 4E lich many times (including in a scenario I authored, though I didn’t design the lich) and really liked how it played. I never once had a player say, “that doesn’t feel like a lich” (as I slowly drained their life away).
They also experimented with variations on this style in 2020 and early 2021 Rime of the Frostmaiden, Candlekeep, and Van Richten’s. They pretty clearly made a conscious decision to abandon this approach, but they’ve never really explained their reasoning.
I am very curious as well why they decided on that. Or why the Studio Blog said the spells would only be utility spells… and yet Witchlight was already out with non-utility spells on the spell list, and Multiverse had to already be completed by then.
I loved running most monsters in 4e, and have no problem with monsters using different mechanics than players (Slaad crystals or Beholder Eye Rays, etc). But when classed NPCs that have the same name as a PC subclass (Transmuter) have abilities that a PC can never learn, it’s offputting.
It makes loads of design sense from a “this is designed for 3 rounds of combat” perspective, but it’s frustrating, as a player to encounter that, at least for me. As a DM, saying “you can’t learn that, because…uh, it’s a monster and you’re a PC” does not feel great. That’s as meta as it gets, and feels ungrounded and arbitrary.
Similarly, worldbuiling a reason that *these* Transmuters have access to something the PCs do not — or, worse, the DM gets pressured to homebrew a version of Arcane Burst (for example) for PCs — is a higher tax on DM prep than the original spellblock, I think. And also is likely to catch more DMs flat-footed, in the sense that a PC will often say: “Hey, I want to learn THAT” when they see something awesome they feel like they should be able to do, which is not something the DM is primed to prepare for.