Patreon DevBlog #15: A Return to Form

It’s been a while since my last developer blog, an issue for which I sincerely apologize. It’s been a busy few weeks, but I should’ve taken the time to write these anyway. As a make-up gift, please enjoy this four-point devblog—one for each post I owe ya :)

Problem #1: How do I improvise a session entirely from scratch?

A few weeks ago, a few players were missing from my main IRL homebrew campaign, and so we decided to run a one-shot session instead. I offered the remaining players a choice: They could pick one of the homebrew one-shots I’d previously published for the Patreon, or they could choose the “mystery box”: a 100% improvised session that I would make up as I went along.

They chose the mystery box. And so, I found myself sitting there at the head of the table, mind completely blank, and utterly unsure which direction to take.

I’ve said it before, and I’ll say it again: I am not a creative person. I am, however, fairly adept at finding and exploring patterns—which is a skill that I decided to test out.

I asked my players to each vote for two of the four elements: water, air, earth, and fire. I then asked them to each vote for two of the following musical genres: rock, pop, classical, electric. They chose Air/Fire and Classical/Electric.

I spent about a minute or so pondering these two combinations. Air/Fire, to me, meant ash and soot, which immediately reminded me of the Dark Souls series. Classical/Electric suggested a combination of elegance and innovation, which meant steampunk.

I had one problem: My players wanted to play their normal PCs in the oneshot, and they wanted the oneshot to be canon to our current adventure—which meant it had to happen in the past while also having minimal impact on the actual campaign.

That gave me an idea. Ash isn’t so far from sand—and the Sandman is the purveyor of dreams. What if the players awoke in an ashen dreamscape?

In the spirit of character-driven, tension-driven storytelling, I then spent fifteen seconds glancing over my Patreon reference sheet on making dramatic questions. I decided to make the core dramatic question: “Can the players defeat the Lord of the Volcano”—the thing putting out all the ash—“and escape the dream before their minds are lost forever?”

This meant I needed an antagonist—and to make sure I had enough antagonists to keep things interesting, I made three antagonists instead: an incidental antagonist (the Lord of the Pit, a dream parasite consuming the players’ mindscapes), a consequential antagonist (the Shadow, a dream denizen that wants the dream to never end), and a procedural antagonist (the Duke of Ash, whom the players must convince to aid them in entering the volcano).

I spent another thirty seconds using the reference sheet on running NPCs to make ‘em each distinct: the Lord would be disgusting (self-obsessed); the Duke would be flattering yet disgusting (apathetic); and the Shadow would be endearing (fearful of death) yet irritating (unwilling to cooperate) and insulting (rude). To give the players an emotional connection, I decided to toss in an NPC ally: a lil’ half-clockwork goblin who would be amusing, endearing, and useful.

Finally, I spent an additional thirty seconds jotting down some basic encounters: a combat encounter at the start, then a social encounter with the Duke and Shadow, an obstacle course in the volcano, and a combat encounter with the Lord. I had no idea what they’d look like or include, mind you; I just knew that they existed.

I also jotted down a simple phrase between the combat encounter and the Duke: “the sludge pits.” No clue what they’d be. Just seemed like a cool thing to write.

And that was it. In under five minutes, I had the outline of a one-shot session, ready to run.

It worked beautifully.

I kicked off with a monologue describing the players slipping into sleep and awakening into shallow graves in the desolate, ash-choked wasteland. As they took in their surroundings, I had the goblin approach them, pleading for protection from a group of enemies. I punched in the math for a group of seven 2nd-level players, and the CR2.0 encounter builder spat out exactly how many skeletons and dust mephits (two monsters I pulled out of the top of my head) I’d need.

After the combat, I had the goblin set up the stakes: journey to the volcano and defeat the Lord of the Pit before he consumes your minds, or else the dream will last forever. The goblin pointed out the volcano on the horizon; when the players asked if there were any settlements below it, I described the small village at the base of the slopes that clustered around the “sludge pits.”

Here’s where the players did something really neat. Leaning into the nature of dreams, one of them asked if they could imagine a door connected to the village and then just step through it, like a lucid dreamer. Going with the improvisational flow, I said, “Yes, and”—and they went nuts.

That set the tone for the entire rest of the session. As the players ventured near the sludge pits, ascended to the Duke’s court, delved into the volcanic caverns, and ultimately confronted the Lord of the Pit, I improvised and riffed as many details as I could to gently nudge the players forward and add resonance to the scenes, while the players continued to poke and prod at anything that looked interesting.

Whenever they tried to do something novel or interesting, I usually went with a “Yes, and,” or a “Yes, but,” or a “No, but” to try and make something work—and on more than one occasion it went swimmingly. At the Duke’s court, for example, I had no idea how they’d manage to convince him to hand over the key to the volcano’s mighty stone doors—but before they even got there, we wound up exploring a whole scene at the sludge pits at which the players decided that they could use the pits to draw forth memories of good and hopeful things. One of those memories, an emerald embodying the courage and hope of their party’s first adventure, wound up being the token they gave to the Duke to win his favor.

Of course, we mustn’t forget the emotional stakes. At the end, as the players celebrated their victory over the Lord of the Pit, some of them noticed that their little goblin friend had keeled over on the ground, hands across her stomach. She, after all, was a creation of the dream as much as the Duke or the Shadow had been—and now, with the dream ending, she too would fade into ash.

They held her, comforting her, as she dissipated into nothingness—and as Hans Zimmer’s Time played in the background, asked them to promise that they would never give up on their dreams.

Tears were shed—and, as the characters awoke to the morning light, the session ended.

Now, it’s worth asking: was I entirely honest in calling this session “improvised”? After all, I spent a whole five minutes or so prepping it! Maybe “improvised” isn’t the right word, after all.

And yet, I think this experience underscores the power of understanding the fundamentals of game and narrative design—the basic concepts of dramatic questions, NPC resonance, and gameplay challenges. With little more than the fundamentals and a spark of inspiration, our party told a tale that’s sure to stick with us as long as any “real” adventure—and, at least for a non-creative person like me, that’s powerful stuff.

Problem #2: How do I fix the Missing Vistana questline?

It’s probably no surprise to any of you that the Missing Vistana questline in Curse of Strahd: Re-Reloaded—specifically, the merger of the RAW Arabelle questline with the RAW Mordenkainen questline—has given me no end of headaches.

So far, I’d managed to replace Mordenkainen (both as the leader of the Barovian rebellion and Arabelle’s kidnapped) with the vampire hunter Dr. Rudolph van Richten. The plan from there? The players learn about Arabelle’s disappearance at the Vistani encampment, follow the clues to Van Richten’s Tower, learn that Rictavio = Van Richten = Arabelle’s kidnapper, and follow the clues all the way back to Vallaki, where they have some sort of confrontation.

The problem: I couldn’t figure out how to craft a satisfying trail for the PCs to follow that would lead them from the Vistani camp to Van Richten’s Tower. In principle, this shouldn’t be hard: this is a standard puzzle encounter, which means the PCs need to spend some time exploring a point-and-click adventure that ultimately lets them deduce a solution from the evidence. But there was a big flaw that I hadn’t originally seen.

As one member of the Patreon Discord helpfully reminded me, the RAW module specifically states that the dusk elves are “skilled trackers.” Given that these “skilled trackers” were currently doing all that they could to help the Vistani find Arabelle, it seemed ludicrous that the PCs would be asked to do anything that the dusk elves hadn’t already done. (Even if one of the players has a tidy +6 to Survival checks at this point, the dusk elf scouts have +6 too!)

The players’ task, therefore, couldn’t be anything that the dusk elves could have done themselves. That meant “investigate the crime scene and follow the clues” was out. That led me to another problem: Why on earth was Luvash asking the PCs to find Arabelle?

In the RAW module, Luvash invites the PCs to find Arabelle on a whim—they come to him seeking escape from Barovia, and he idly says, “sure, I’ll help you if you do this for me first.” But in Re-Reloaded, the PCs aren’t prioritizing escape from Barovia, and Luvash sure as hell won’t offer it to them (let alone give it). So, from his point of view, why would he ask a bunch of random strangers/outsiders who just waltzed up to his caravan to go find his daughter when his trusted friends and neighbors are already on the hunt?

I realized that I needed the PCs’ mission to be qualitatively different from the dusk elves’ (i.e., a completely separate kind of task), rather than quantitatively different (i.e., something that the dusk elves failed that the players will do better). This meant that I needed to look at qualitative differences between the PCs and the dusk elves—and the biggest item that jumped off my list was “able to enter Vallaki without pissing off the Baron and his guards.”

But what could the PCs find in Vallaki? My first thought went to Szoldar Szoldarovich and Yevgeni Krushkin, the Vallakian wolf-hunters. They’d been all over Barovia—perhaps Van Richten had dropped some distinct item he’d taken the tower, and they would immediately recognize it and lead the PCs there.

I was pretty happy with this for a moment—until a patron helpfully pointed out that the dusk elves have lived in the valley for over four hundred years, which means that these “skilled trackers” almost certainly know about the tower and would recognize stuff from it. I was sour about this for a bit, and then ChatGPT agreed that the dusk elves would know about it, so I threw the idea out and started over.

Back to square one: the PCs need to get or do something in Vallaki. But what?

I did some thinking, and remembered that Van Richten’s Tower was originally Khazan’s tower—and that I’d already decided that Victor Vallakovich, the Baron’s son, had stumbled upon Khazan’s spellbook to use as his own. I also remembered that Baron Vallakovich had a library in his manor—which, while small by modern standards, would surely be well-known in a feudal settlement.

The dusk elves would recognize the outside of the tower, but they’d have no knowledge of the inside—which might include, for example, Khazan’s personal sigil. What if Van Richten had taken something with Khazan’s sigil—say, a signet ring—and dropped it on the grass?

Luvash, assuming it to be a noble’s sigil, would naturally ask the PCs to investigate a lead by researching it in the Baron’s library. Their efforts there would turn up nothing—but as Victor stuck his head in to see what was going on, they’d see Khazan’s sigil on the front of his spellbook, and would be off to the races.

Instead of a puzzle encounter (follow the clues to the tower), I wound up with two social encounters (persuade the Baron to let you look at his library / persuade Victor to let you look at his spellbook). As an added bonus, this would also tie the PCs into the Victor/Stella questline, which is always a plus.

Designer Commentary: The Role of Random Encounters

I was rewatching the first season of One-Punch Man when something occurred to me: a solid number of the fights in the first few episodes are, functionally, random encounters.

What do I mean by this? They don’t advance the plot in any way. They’re just stuff for Saitama (the protagonist) to beat up and pad out the episode—or so it seems. But these episodes reveal a fundamental, but oft-ignored role that random encounters can and should play in our D&D campaigns: as character development.

In the immortal words of Rick and Morty’s Dr. Wong, most DMs treat random encounters like work: “It’s not an adventure. There’s no way to do it so wrong you might die.” But while these encounters shouldn’t be deadly, they should have real emotional stakes—and even when those stakes aren’t tied into the plot directly, they should still tie, at least indirectly, into the characters’ own arcs.

In the first episode of One-Punch Man, the protagonist fights an assortment of random monsters: an embodiment of pollution, a titanic giant, a car-man fusion, and the Subterranean People. None of these foes are particularly important—they’re all, again, functionally random encounters—but they each highlight or explore important aspects of the protagonist’s character: his power, his loneliness, his apathy, and his yearning for a true challenge.

We can apply the same concept in our D&D random encounters by creating more complex dramatic questions. The players could fight off a band of goblin bandits…or they could fight a band of goblin bandits that surrenders as soon as the players show teeth, and whose members’ cowardice is a dark reflection of the druid’s.

As much as I’d like to delve more deeply into the idea here, it’s still a bit underbaked, and I’m sure the possibilities are literally endless for all of the different players and parties it might apply to. Still, it’s something I’m going to be considering as I work to figure out random encounters in my own game.

(Note: I’ve mentioned it before, but I’ll mention it again: it’s also totally valid to use random encounters to foreshadow upcoming arcs, to create a particular atmosphere, or to create particular themes. The “character arcs” application is just another addition to the pile of not-so-random-random-encounters.)

Designer Commentary: Three-Phase Strahd Playtest Postmertem

Last night, I finally had a chance to playtest my revised three-phase Strahd statblock—and boy did it deliver. It was two-and-a-half hours of epic finale action, and I learned a lot.

Let’s start with the good stuff, first: players loved the new dynamic monster templating, and praised the encounter as dynamic, engaging, and tactical. Strahd’s statblock was so simple that it seemed to run itself, but it still had enough emergent complexity to allow for a surprising depth of strategy. The players also loved the variety and flavor of his new abilities, and didn’t seem to mind playing through an uninterrupted fight that took more than two hours to complete.

(It didn’t hurt, of course, that we used my new combat rule of “I’ll-skip-your-turn-or-your-dice-roll-if-you’re-not-immediately-ready-but-we-can-come-back-to-you-later,” which sped things up significantly and ensured that there was always someone talking. It was also useful to assign control of the players’ NPC allies—Ezmerelda, Ireena, and Kasimir—as well as initiative and damage-tracking to the players themselves, which lightened my load considerably, kept players engaged, and kept the game moving fast.)

With that said, there were some issues that cropped up. One player pointed out that some of his lair actions were underwhelming compared to the others. The players also noted that his first phase seemed more threatening than the other two. Okay; easy enough to solve.

More interesting was the feedback I didn’t receive—but took note of anyway. More than anything else, the biggest issue that I noticed was that I kept forgetting that Strahd’s abilities existed.

I wasn’t forgetting, mind you, his actions, bonus actions, reactions, or lair actions; those were all easy. The tricky things to track were his passive features. I had enough trouble tracking his sunlight hypersensitivity at times, but my problems only got worse with the new features I’d added.

Take, for example, the chill shield and tactical riposte features I’d given his Mage and Soldier phases, respectively. In principle, these abilities were supposed to punish the players for attacking him from up close. In practice, however, since they didn’t require a reaction to use, I kept forgetting to use them because I was so laser-focused on his reaction triggers—and because I already had a lot to keep track of. (I don’t think I had a chance to use Strahd’s battlefield awareness ability—which gives him advantage on DEX saves against effects he can see or hear in his Soldier phase—but I forgot about its existence nonetheless.)

I also found that a few passive features I’d added for flavor wound up doing literally nothing. Most notable was the Vampire phase’s mistform mastery, which had been intended to let Strahd pass through hostile creatures’ spaces, but in practice was basically useless because he was already so damn mobile without it.

What did I learn from this? Well, other than that ringing endorsement of dynamic monster design, multi-phase bosses, and the importance of playtesting, I learned this: keep your monsters simple. That doesn’t mean that they should be boring, or that they shouldn’t have options! But I did learn that you should always aim to minimize the mental load, both on yourself and your players, and to avoid taking up space with abilities that don’t do anything. Better to have one impactful, memorable, unique ability than three that either won’t be remembered or will barely make a splash.

Campaign Advice Roundup | May 28, 2023

• Not every campaign makes a good home for individual character quests (to be contrasted with individual character development). For example, a campaign like Critical Role works because everyone at the table has explicitly signed up for a particular kind of character-driven campaign and because everyone at the table trusts everyone else to approach these storylines maturely, thoughtfully, and inclusively. In the average plot-driven campaign (e.g., Curse of Strahd), however, it’s far too easy for a character-specific storyline to prove alienating and boring for the other players, and distracting for the DM. If a character has unique narrative needs or goals, it’s almost always best to tie those into a storyline that the party would pursue anyway, rather than creating something special and unique for the party to pursue on that character’s behalf.