Patreon DevBlog #5: Grappling with the Impossible

We’re back on the DevBlog, albeit a bit belatedly—apologies for the delay! I’ve been both a bit busy and a bit sick, but thankfully, it looks like most of the clouds have now cleared from the horizon. In the past week, I’ve worked a bit on Re-Reloaded but I’ve also spent a bunch of time on my own IRL homebrew campaign (the one with the ten players), which gave me a good opportunity to crack my head against some elementary design problems.

Problem 1: How do I avoid killing players with the vampires in the coffinmaker’s shop?

Everyone knows and loves and hates the six vampire spawn in the coffinmaker’s shop in Curse of Strahd. Given the sheer number of early-game TPKs they’ve caused, it’s honestly a wonder why Chris Perkins put them here in the first place.

Of course, given my rearrangement of the module for Re-Reloaded, I knew that the St. Andral’s Feast quest had to be the first big quest line that the PCs were introduced to upon arriving in Vallaki—along with defeating Izek Strazni, restoring the bones were one of the two big milestones I expected players to hit on their way from 4th to 5th level.

With that said, these vampires are tough. The players have no access to sunlight yet, since I purposefully put the Holy Symbol of Ravenkind way off in the Krezk-ish area. Some playgroups might have clerics (and therefore turn undead), but many playgroups won’t, and there’s no real way to balance encounters around abilities like that when it’s not unknown for them to whiff.

Let’s start with the math. Under CR2.0, five 4th-level PCs against six CR 5 vampire spawn makes for…

Uh.

I am not joking when I say that the scale literally does not go that high. My DM’s Toolkit encounter builder says that this comes out to consume 1,531% of the party’s total maximum hit points.

Except it gets even worse—according to the CR calculator on the site-that-shall-not-be-named, a single vampire spawn is actually CR 6, not CR 5! (I have no idea how this happened, except I imagine that WotC factored in the Vampire Weaknesses—including sunlight hypersensitivity—when doing this calculation in the Monster Manual. I think that this was a Very Dumb and Ill-Advised Idea.)

Okay, so now we’ve got an encounter that consumes 1,793% of the party’s total maximum hit points instead. Great.

Now, here’s the thing. RAW, the players aren’t really supposed to fight these vampires at all. The bones are in an entirely separate room from them, the vampires are asleep and sealed in little wooden crates, the coffinmaker immediately tells the players where the bones are and where the vampires are, and there’s nothing stopping the players from just taking the bones and leaving. As far as the module is concerned, once the bones are gone, the vampires just…cease to exist.

But it’s my guide and I get to choose the climactic confrontations. And dammit, a quest like this deserves a better end than “the players stroll upstairs, grab the bones, and then yeet to the finish line.” Plus, I want to use this quest to introduce Strahd’s bride Volenta as the local boss, making clear to the players that Strahd and his forces are an active and antagonistic force in the valley.

So now I need to find some way to make this encounter survivable.

One of the little tricks about the way encounter math works in 5e is that it’s easier to kill six waves of one-monster encounters than a single six-monster wave. (Why? Because each additional monster you add to the encounter also absorbs some of the damage that would have otherwise killed one of the original monsters, making those monsters live longer, which lets them keep dealing damage for a longer period of time.)

An interesting thing about the architecture of the coffinmaker’s shop is that it’s a very small, cramped space. That means…chokepoints! Let’s take a look at what happens if the vampires come at the PCs one-at-a-time, so that the players can focus-fire each vampire spawn as it appears.

An encounter with a single CR 6 vampire spawn will consume 50% of the party’s total maximum hit points. Across six waves, that’s 300% of the party’s hit points consumed. That’s an enormous improvement!

…Of course, it’s still an incredibly impossible encounter. But this is still a work-in-progress!

Now, the best anti-vampire tech in the entirety of D&D, excepting sunlight, is the spell spirit guardians. It deals reliable radiant damage in a massive radius to every single vampire that is trying to attack you, which shuts off their regeneration. It also slows those vampires down, making it harder for them to escape and making them spend more time in the aura (thereby taking more damage) if they want to get to you in the first place.

Given that vampires are obligate melee combatants who couldn’t use a crossbow if their life depended on it, this is a pretty good situation for players to be in!

Except—this is a 4th-level quest. It is explicitly meant to be a 4th-level quest. And spirit guardians is a 3rd-level spell, which means no spirit guardians for the players.

Fortunately, all hope is not lost. Because there’s someone in Vallaki who does know spirit guardians: Father Lucian, who has a priest stat block. And, fortuitously, Father Lucian has a very obvious interest in wanting to see St. Andral’s bones recovered safely.

So let’s assume that Father Lucian comes along on this trip as a temporary NPC companion and casts spirit guardians as soon as possible. That lets us do two things: (1) eliminate the vampire spawns’ regeneration feature when calculating their CR (thereby bringing them down to CR 5), and (2) add a CR 2 NPC ally to the encounter-builder. (We’ll assume, for the sake of efficiency, that he also pulls the ol’ spiritual-weapon-bonus-action synergy for maximum damage output and action economy.)

Each individual vampire spawn will now cost 28% of the party’s total maximum hit points (including Father Lucian). That’s 168% over the course of the entire encounter. Still impossible, but hey—progress!

Father Lucian also has a very interesting ability called divine eminence. This lets him expend spell slots as a bonus action to deal extra radiant damage whenever he hits with a melee attack. The problem is that he’s got far better things to spend his actions on—guiding bolt, for example—and, furthermore, that if we ever let him get into melee range of a vampire spawn, he will lose his throat very shortly thereafter.

This is the part where we start cheating. What if instead of a bonus action to buff his own attacks, he could use a reaction to buff an ally’s attacks? According to my very rough back-of-napkin calculations, this brings him up to a lovely CR 3. Now, each individual vampire spawn costs 24% of the party’s total maximum hit points, for a total of 144%.

Still impossible. You can probably tell that we’re basically hitting a wall at this point.

That doesn’t mean that there aren’t other possible ideas…but not every idea is a good one. For example—I briefly entertained the idea of making St. Andral’s bones serve as a protective amulet; the person holding them (and/or everyone in a five-foot radius) has the effects of a protection from evil and good spell. While that would’ve done wonders to nerf the vampire spawn, it would’ve had two big deleterious effects:

• it would have made Volenta and her vampire spawn seem like bumbling buffoons who can’t hit an attack for their lives in what is, functionally, the PCs’ first fight with vampires; and
• players who don’t give a shit about the church might probably just taken the bones (as loot) and run without giving them back to Lucian.

So that idea went out the window.

It gets worse, too. I really really want Volenta to be the big capstone boss in this fight, but I’ve decided to make all of the brides (and basically every enemy NPC that might ever be encountered solo) into two-stage bosses. Assuming Volenta has two CR 5 phases, that means that the party basically has seven waves of vampires to kill, which puts us back up at 168%.

Sigh. Nothing can ever be easy.

Except when it is!

Nothing in the rules of D&D says that this encounter has to include six vampire spawn. In fact, the only reason there are six vampire spawn here RAW is to make sure that (at least in theory) the players don’t dare to fight them!

But I want the players to fight them. In fact, I plan to force them to fight them to get the bones. So why not just…remove some vampire spawn?

Look, I can even justify it in-universe. Volenta is one of Strahd’s brides, and basically all of Strahd’s servants are currently competing for his affections. Everyone’s got a different idea of what’ll please Daddy Strahd, and everyone is jealous of everyone else, which means that nobody is sharing resources. This whole St. Andral’s attack is Volenta’s brainchild, which means that she’ll necessarily have very limited resources—henchmen included.

100% of the party’s total hit points divided by 24% of their hit points per vampire spawn gives me…four vampire spawn. Or, in other words, two vampire spawn plus both of Volenta’s phases. This is a little risky vis-a-vis turn undead, but given that I plan to have Volenta flee battle basically as soon as she’s “bloodied” (i.e., right after her second phase starts), I can toss in a third vampire spawn too without feeling too bad about it.

As for implementation—it seems a bit silly to have all of the vampires conga-ing their way through the door one-by-one. Still, vampire spawn do have the spider climb feature, so I might just have them waste a few rounds climbing over the roof, punching through Henrik’s iron-reinforced windows, and climbing in through those windows for some extra scare factor. This also keeps things nice and clean, since Volenta technically has some ranged abilities that would make things pretty messy pretty quick.

So now I’ve got two vampire spawn coming through the front door, then one spawn and Volenta coming through the window. Lucian has at least some common sense, so he’ll surely encourage the players to group up and force the vampires through chokepoints.

Will it work? Hopefully! Players are resourceful little buggers, and I’ve hammered the numbers enough that I’m confident that they should at least have a chance, so long as they’ve got Lucian with ‘em and they stick together.

And if they don’t?

Welcome to Barovia, kids.

Problem 2: How do I design the concept for my next adventure?

As I’ve mentioned in previous DevBlogs (I think), I’m currently running a homebrew adventure for ten IRL friends who have never played D&D before. One of the fun parts of this is that I get to have a lot of fun introducing them to various aspects of D&D.

One of the frustrating parts is that I have to figure out which aspects to introduce and in what order—which is made all the more complicated by the fact that I have ten flippin’ players.

(I love them all dearly.)

For context: My players had just finished a fairly linear adventure in which they fought some kobolds, defeated their evil half-dragon leader, and stopped them from un-petrifying a Rather Unpleasant adult red dragon. This brought them to 2nd level.

It just so happened that I’d received the D&D orc warband miniature set for Christmas, so I was very interested in the prospect of getting some mileage out of them. There was one problem: Due to recent controversy, as well as my own discomfort in portraying humanoid races monochromatically, I wasn’t particularly interested in running a “kill the evil raiding orcs” adventure.

That wasn’t a problem, at least in principle. Even so, I still wanted an excuse to use these miniatures, dangit, which meant they had to be enemies in some capacity. Fortunately, it also turned out that I’d recently introduced a jerk-ish xenophobic local knight, who the players naturally despised.

After some deliberating, I came up with a central dramatic question that went like this: “When skirmishes on the county border lead to rising tensions between the PCs’ town and a nomadic orcish tribe, can the PCs help a peace-seeking heir claim the leadership of the tribe and defeat their war-mongering leader?”

Again, in principle, this was totally fine! And yet, there was a problem. I wanted the players to eventually fight the war-mongering leader, but I also wanted the players to get a chance to meet him before things went south. And that meant I had to come up with a way to get the players into the orcs’ camp before the final fight, so that they could meet the peaceful heir, and then do all sorts of orc politicky things, and eventually have tensions rise until the final confrontation. (And oh, dear; I just realized I accidentally reinvented Pocahontas and Avatar, which probably doesn’t bode well for the ultimate fate of this quest.)

Now, this all works out great in a film, where everything works out very linearly. But this is D&D, and players are prone to act in very unpredictable ways. You can see my dilemma:

• in order to establish the proper relationships for the climax, I had to have the players’ interactions in the orc camp unfold just so, without breaking anything; but
• social interactions in D&D are naturally unpredictable, which is exponentially compounded by the fact that I have ten players and most of them, by virtue of their newness, are luxuriating in their newfound discovery of murderhobo-ness.

(They’re not actually murderhobos. But they’re also very very transactional, greedy little Gollums, and I didn’t want to risk anything that would deprive them of that agency.)

On top of that, we’d just finished two very very linear questlines. Now, there’s nothing wrong with linear adventures, but part of the fun of D&D is its unpredictability. I could, in theory, unleash my players onto the orc tribe and just watch what happens, but I wasn’t sure if enough of the group would be comfortable with that level of in-depth roleplay.

Plus, I wasn’t sure if I could generate the level of resonance necessary to keep my players entertained for the whole quest. It’s easy to make players feel amused, or flattered, or endeared when the NPC’s personality is somewhat orthogonal to the quest as a whole. It’s somewhat more difficult when the NPC’s personality—and I would need to make a bunch of NPCs!—is intertwined with the ultimate goal of the quest.

Now, that’s not to say that I couldn’t have done it. But I am, fundamentally, a very lazy person, and I didn’t want to have to come up with a big cast of characters on the Friday night before the game. That doesn’t mean that I won’t try to resurrect this adventure idea eventually—but for now, it seemed a bit too ambitious.

Fortunately, when you’ve got new players, everything is new and incredible. So, as I reluctantly put my social-focused adventure back onto the shelf, I remembered an article from the Alexandrian about jaquaying the dungeon—that is, creating intentional nonlinearity in your adventures—and decided to create a brief, but meaningfully nonlinear, dungeon crawl instead.

So I did what I usually do when designing dungeons (and adventures in general): I filtered the monster list by monsters of an appropriate CR to serve as a final boss, and perused it until I found something that interested me.

In this case, I found two, the nothic and the berbalang. I briefly pondered my recent adventures, and decided that I wanted an opportunity to also introduce a type of D&D monster I hadn’t introduced yet—undead. I punched their descriptions into ChatGPT and asked it to generate “ideas for a D&D adventure suitable for 2nd-level players involving a berbalang, a nothic, and hostile undead.”

The first result it spit back out was “The Lost Library. The players are hired by a powerful mage to retrieve a rare spell book from a long-abandoned library. The library was once the home of a powerful wizard who was cursed by Vecna, turning him into a nothic. The players must navigate the dangerous maze of the library, avoiding the traps set by the nothic and the hostile undead, while searching for the spell book. Along the way, they encounter a berbalang that is searching for the same information. The players must decide whether to ally with the berbalang or fight against it to obtain the spell book.”

I dug this, so I tried asking ChatGPT to generate a bunch of encounter ideas for the dungeon. All of them, unfortunately, were pretty crap—while good in a vacuum, they didn’t really fit with the vision I was going for. So, instead, I asked it to write “the backstory of a lost library that holds a powerful artifact, hosts a large number of undead skeletons, and is of great interest to both a berbalang and a nothic.”

It gave me some great stuff, and some ideas began to take shape in my brain. From here, I started flashing back to adventures like The Sunless Citadel, whose dungeons are great particularly because they have intelligent NPC factions. (In other words, most nodes in the dungeon have Faces.) So I asked it to generate a few ideas for factions to include in the dungeon, and it suggested an army of skeletal warriors, a colony of spectral spiders, a horde of undead rats animated by the library's magic, a pack of shadows, and a group of wights.

I started winnowing this list a bit for CR balance purposes—maybe include only one phase spider and a single wight—but I was struck by the idea of a horde of rats who had obtained intelligence via the ambient magic of the library and decided to run with that instead of the undead rats that ChatGPT suggested.

Now that I knew what the dungeon was, roughly, I had to figure out what the adventure’s dramatic question was. I knew that I wanted the players to recover a powerful artifact that lay at the end of it, but—after the linear, fast-paced adventures they’d just finished, I wanted to give them a bit of a lazy adventure this time. So the dramatic question was, “When a client asks the PCs' adventuring guild to recover an ancestral heirloom from the depths of the necromancer's library, can the players use their tactics, influence, and reasoning to retrieve it and get paid?”

(I intentionally changed the stakes from “before X bad thing happens” to a simple “get paid” because I wanted to make this a low-tension, low-stakes romp through a dungeon.)

Now, because this group plays fairly infrequently, I’ve got a bunch more time to actually implement this idea, but for now, I’m actually pretty happy with it!

Designer Commentary: Campaign Module Layouts

I was chatting with a friend about Re-Reloaded recently, and noticed something interesting.

Re-Reloaded is, of course, an unofficial D&D supplement. Because Reloaded is not produced by Wizards of the Coast, I can’t use it to reproduce, verbatim, the contents of copyrighted text from the original Curse of Strahd module.

Instead, whenever I need to refer back to something like NPC lore or location flavor text, what I reliably do is write, See Section X (p. YY) in the original module. On the one hand, this tends to produce a bunch of page-flipping—but on the other hand, it lets me organize the actual guide’s text in a way that keeps it laser-focused on what you actually need to have at the table.

Let me explain. Curse of Strahd, as a text, is largely organized as a setting book from which gameplay and story organically emerge. With a few exceptions—such as monsters that attack at particular times and special events that occur under particular conditions—the vast majority of the text is just, “Here’s what you need to know in order to use this campaign flexibly.”

In my experience, though, most DMs don’t want flexibility; they want order, understanding, and control. They want to be able to look down at the campaign module they just bought, do zero prep aside from a quick skim, then crack it open and run the adventure straight out of the book.

(This is not meant to be a slight—this accurately describes my preferences for campaign modules, too.)

Curse of Strahd is not unique in this respect—every single published WotC 5e module is organized around places rather than scenes. It’s no secret that many DMs find this objectionable; I’ve heard many times that official modules require far too much prep and understanding in order to run. At the same time, I think that this distinction between “locations” (information you need to react to player actions and adapt your plans) and “scenes” (your initial expectations for how a session or adventure will unfold) could hold the key to something meaningful.

Specifically: what if the ideal layout for a campaign module is to split it into two halves: a story half and a setting half? The story half contains the “critical path”—a list of all of the scenes that are expected to occur throughout the adventure, plus some basic information for different ways those scenes might resolve and how they could affect other, future scenes.

Every time a scene in the story half references a particular place or NPC, it references the appropriate entry in the setting half, so that the DM can refresh their memory and improve their adaptability both before and during the session.

Will this work? I’m not entirely sure, but it does seem to follow the workflow I’ve worked out for myself after years of DMing. And, when you think about it, this kind of shift does make a certain amount of sense.

D&D modules started out, originally, as wargaming dungeons—the kind of space-based environments that didn’t really have much emergent narrative gameplay and instead focused on the challenges or loot that the players would encounter in each “room.” But as time has gone on, this style of play has shifted and expanded, leading to a much more narrative approach that is less delineated by space than by narrative circumstances—i.e., scene.

Perhaps, then, it’s time that module design changed to match this changing play reality as well.

Campaign Advice Roundup | February 14, 2023

• The only way for a player to know they’re being scried on is if they can see invisible objects. To conceal that a player is being scried upon out-of-game, you can either ask them to make a saving throw (but not tell them what it’s for) or just make the saving throw yourself.
• When running travel sequences, the best way to create a sense of progression and time is to (1) run a number of random encounters that foreshadow future plot arcs, develop current plot arcs, and/or build out the world in a way that adds depth to the campaign’s dramatic questions; (2) use lots of landmark-specific flavor text that makes the players feel like they’re watching a travel montage; and (3) if the journey takes several days or more, create one or more ongoing mysteries and/or roadside dramas that slowly unfold over the course of the journey.
• The wish spell is not omnipotent—Wild Beyond the Witchlight makes this clear. It can give unexpected results, or even fail entirely. A useful metaphor might be like water on the top of a pockmarked hill. The players want the water to flow down into their bucket at the bottom, but the water will happily flow into the closest crevasse that will hold all of it, even if it's still most of the way up the peak. The taller the hill (i.e., the more ambitious the wish), the sooner the water will find a crevasse that the players didn't want or expect. The easy way to use wish is to use one of its listed properties or replicate a spell; that's like pouring water directly from a faucet into your bucket. But as your wish gets more ambitious, the water finds easier and easier ways to flow—like the spell description itself suggests, wishing for the villain to be dead may only transport you to a time period in which they’ve been dead for fifty years.