Dragna's Devblog: A Designer's Framework

One of the most frequent questions I get from readers about Re-Reloaded is: “Why did you remove Morgantha from the Village of Barovia?”

Tangentially related, two of the most frequent pieces of advice I wind up giving to other people are “Don’t turn Argynvost’s body into a shadow dracolich,” and “Don’t have Strahd give the PCs a bottle of wine outside of Death House.”

I often have lots of very good reasons (I think, anyway) when I answer the first question and argue for the two pieces of advice. But until now, I’ve never really laid out a coherent theory as to how I arrive at these conclusions, and why I think they’re important. So today, we’re going to take a look at (what I think are) the two critical components of creating a campaign: narrative design and game design.

What are Game Design & Narrative Design?

Let’s start with some definitions. Dungeons & Dragons is a tabletop roleplaying game, which, in the name alone, tells us the three skill sets that any DM needs to master:

• Social Design (“tabletop,” i.e., social engineering and group management)
• Narrative Design (“roleplaying,” i.e., storytelling and emotional satisfaction)
• Game Design (“game,” i.e., gameplay challenges and tactical experiences)

Let’s put aside social design for now, because that involves a lot of social skills that we don’t particularly need to get into now, and (more importantly) because it’s pretty much content-agnostic. Instead, like I said before, we’re going to focus today on narrative and game design.

Good “game design” is the art of creating a game that will engage and satisfy its players. Think Pac-Man, Risk, or Super Smash Brothers. A good game provides its players with a fair and meaningful challenge, provides them with tools or skills that can be used tactically and/or creatively to overcome that challenge, and provides the players with a sense of accomplishment and progression that makes the experience feel “worth it.”

Meanwhile, good “narrative design” is the art of creating a story that will engage and satisfy its audience. Think The Lord of the Rings or Star Wars. A good story encourages its audience to emotionally invest in its characters, introduces dramatic tension to deepen that emotional investment, and concludes with a climactic resolution that defuses the tension and provides the audience with a “takeaway” that makes the experience feel “worth it.”

In a roleplaying game, these two things fuel each other in a beautiful symbiosis: good narrative design creates emotional investment and tension; tension makes gameplay challenges meaningful; gameplay challenges are used to resolve the tension; and the senses of accomplishment, progression, and emotional satisfaction make the journey feel “worth it.”

Interrogating Good Game Design

Let’s dig a little deeper, starting with the more familiar idea for many people: game design. While different people can, of course, have differing opinions, in my personal experience, a good “game” has the following attributes:

• a default loss state that will occur if the player does nothing, an idealized victory state that the player wishes to reach, and any number of “partial victory” states in between the loss and victory states;
• a fair set-up, such that the player could achieve a perfect victory on the first try if they were simply smart enough, fast enough, creative enough, strong enough, or so on;
• an iterative and cumulative game state, such that no player can win or lose with a single lucky or unlucky decision;
• a set of multiple tools and/or strategies, such that the player always has a choice to make when acting within the game; and
• an opportunity for meaningful choice between those tools and strategies, such that multiple routes could plausibly lead to victory and/or defeat.

To put it another way, a good game has stakes, is fair, takes more than one round to complete, has more than one thing to do, and gives players more than one path to win.

Many games fall short of these goals. For example:

• At launch, No Man's Sky was heavily criticized for its lack of a clear objective or loss condition, failing #1 (loss and victory states).
• Meanwhile, "Rock-Paper-Scissors" is a one-round game, preventing players from engaging with the game over a meaningful time horizon and therefore failing #3 (iterative and cumulative game-state).
• Children's board games like "Candy Land" or "Snakes and Ladders" offer no actual choices to their players beyond rolling the dice each turn, failing #4 (multiple tools and/or strategies).
• While "Tic-Tac-Toe" ostensibly offers players a choice of how to play, that choice is an illusion—there is always an optimal and "correct" move to make, failing #5 (opportunities for meaningful choice)
• And although most games are replayable, imagine a game of Touhou or Dark Souls that self-destructed as soon as the player experienced their first death—the need for memorization (and the inability to memorize key details) before gameplay challenges would make those games fundamentally unfair, failing #2 (fair set-up).

You might already be starting to see how several of these factors tie into narrative design as well:

• The narrative of the game provides stakes, which delineate the loss state from the victory state. (The better the players play the game, the better the narrative outcome they can attain.)
• The narrative of the game provides players with exposition and resources, which ensures that they have a fair chance and multiple options in actual gameplay.

Back to the Narrative

Now, let’s take a closer look at narrative design. (I promise—we’ll get to the Curse of Strahd stuff soon.) A “good” narrative, in my experience, has the following attributes:

• a cast of characters that resonate with the audience
• a setting that lends emotional depth, weight, and context to the characters’ actions
• a source of tension that pushes the characters to act and change in interesting ways
• a set of stakes that tie the tension to the characters
• a resolution that satisfies the audience’s emotional needs.

Some examples of narratives that fail to meet these factors include:

• Suicide Squad (2016), a movie that was broadly criticized for its relatively flat and uninteresting cast of characters, failing #1;
• Eragon, a book-to-movie adaptation that some fans have criticized for more resembling a paint-by-numbers fantasy tropefest than a meaningful world, thereby failing #2;
• The Room, a movie criticized for featuring plot developments and tensions that often seem nonsensical or unconnected to the cast, failing #3;
• Justice League (2017), a film often criticized for its weak villain and lack of personal stakes, failing #4; and
• Game of Thrones, a television show whose fanbase fractured when the series finale was seen to have largely failed to deliver upon its thematic promises, failing #5.

If you’re familiar with the dramatic question reference sheet in the DM’s Toolkit, a lot of this might look familiar! Good narrative design requires good dramatic questions, which themselves require clear goals (resolution) tied to strong stakes (stakes) and opposed by some adversarial force or antagonist (tension). Good dramatic questions also can’t exist independently of the characters that experience them (cast and resonance) or the areas in which they take place (setting).

In short: good narrative design requires good dramatic questions. This might sound simple, but as we’re about to see in application, things quickly get a little more complicated.

The DM’s Dual Mandate

Every piece of content that a Dungeon Master adds to a D&D campaign should, ideally, have both good game and narrative design. To put it another way:

1. Content should always introduce, develop, answer, or enhance some part of a dramatic question.
2. Players’ efforts to answer some part of a dramatic question should always be fair, take more than one round to complete, and offer more than one path to win

(Here, content “develops” a dramatic question when it changes it somehow, and “enhances” a dramatic question when it makes it more emotionally resonant with the players, such as by exploring its themes or by giving them sympathetic NPCs to fight for.)

Content can have good game design, but bad narrative design, and vice-versa:

• For example, a random encounter on the road can offer players a fair and engaging combat with a chance to test their abilities and skills (good game design), but be completely divorced from the stakes, setting, characters, and tension of the broader campaign (bad narrative design).
• Meanwhile, a scene or dialogue involving the players and one or more NPCs can feature strong stakes, evocative characters, and a clear source of tension (good narrative design), but reduce player input to a single dice roll or, worse, provide players with no opportunity for meaningful input at all (bad game design).

To make things worse, content can also have harmful game or narrative design:

• For example, a beautiful narrative can have such frustrating gameplay that the players stop caring about the game at all, reducing their emotional investment to avoid future pain. (Think Charlie Brown giving up on kicking the football after Lucy pulls it away too many times.)
• Meanwhile, engaging gameplay can include narrative components that actively harm the strength of the game’s dramatic questions, such as by diminishing the stakes, creating thematic dissonance, or by making the villain’s victims unsympathetic.

As such, wheenver you’re thinking of adding, removing, or changing content in your game, it’s important to ask yourself:

1. How will this change impact and interact with the broader dramatic question(s)?
2. How will this change enhance or interfere with current or future gameplay?

As you ponder the answers to these questions, keep one Big Idea front and center in your mind: elegance beats complexity. As Gordon Ramsay of Kitchen Nightmares would tell you, it’s better to do three things well than thirty things poorly. I’ve run one-shots that have proven more memorable than entire campaigns, and run campaigns that were more forgettable than single scenes.

There’s more content in your brain—let alone on the Internet—than you and your table could play through across ten lifetimes. Accordingly, always start with the D&D form of the Null Hypothesis: presume that your Content Idea is bad or worthless, and force it to prove its worth by deliberately exploring its narrative and gameplay design.

And remember: Not all ideas are born equal. Good content can come from anywhere, but good content can’t come from everywhere. As experienced writers like to say, “Kill your darlings.” If you’ve got a Cool Idea in your head, but you just can’t figure out how to make it serve the broader narrative or create engaging gameplay, then the solution is clear: tear it up into tiny shreds, burn the ashes, and start again.

Okay, I get it, but how is this relevant to Curse of Strahd?

I’m glad you asked! With that exhaustive (and exhausting) prologue out of the way, let’s finally turn to the fun stuff: how All Of This Theoretical Stuff is actually applicable to Morgantha, Argynvost, and Strahd’s darned bottle of wine.

Let’s start with Morgantha in the village of Barovia and do a full design analysis:

• Background: The players can meet Morgantha randomly while traversing the streets of Barovia, but she tries to avoid them. If they follow her, they can see her steal a child from a Barovian home, though the parents make a halfhearted attempt to stop her. If the players confront her, she gives the child up without a fight. Eventually, the players can meet Morgantha again at Old Bonegrinder, which they are already incentivized to explore after obtaining the deed in Death House.

(I’m going to skip the whole “if-you-threaten-Morgantha-she-gives-you-exposition” thing because that’s an even deeper layer of Bad Design.)

Let’s start with a game design analysis:

• Victory/Loss State. There’s a clear loss state (“Morgantha takes the child.”). However, although there seems to be a clear victory state (“Morgantha doesn’t take the child.”), this is a lie: the module explicitly says that Morgantha secretly plans to come back later. This encounter fails Game Design Factor #1 (+0 points).
• Fair Set-Up. Given that there’s not really any way for the players to lose, it’s impossible to complain that this specific encounter is unfair. However, there’s also no way for the players to figure out that Morgantha plans to come back later or to figure out how she might even do so (i.e., because she’s a night hag). Moreover, this encounter is actively harmful to future fairness: it explicitly sets up Morgantha as an antagonist despite the fact that the players will have no hope of defeating her coven when they next meet her at Old Bonegrinder. As such, this encounter is net harmful to the fair game design of future areas. That’s a big loss for Game Design Factor #2 (-1 point).
• Iterative & Cumulative Game State: There’s only one thing for the players to do: demand that Morgantha give up the child. This scene fails Game Design Factor #3 (+0 points).
• Multiple Tools & Strategies: The module doesn’t expressly rule out the players’ ability to use creative solutions like sneaking the child out of Morgantha’s cart or bargaining with Morgantha to “buy” the child’s freedom. As such, this scene succeeds in Game Design Factor #4 (+0 points).
• Meaningful Choice: Because demanding the child’s release always wins, there’s not really a chance for meaningful choice here; it’s strictly better than any competing option. This scene fails Game Design Factor #5 (+0 points).

Oof; negative points out of five! Let’s take a look at the narrative design and see if Morgs fares any better there:

• Resonant Cast of Characters. Neither the parents nor the child are particularly interesting. Morgantha herself pretends to be an old lady, but it’s unclear how the players are supposed to feel about her other than “she’s evil.” Even worse, Morgantha’s presence pollutes other narrative arcs as well—her unfettered access to deal drugs in Barovia makes Ismark look malicious and/or incompetent, which in turn makes Ismark unsympathetic, which in turn makes Ireena unsympathetic, which makes the players less likely to want to help her. This isn’t just a failure on Narrative Design Factor #1—it’s actively harmful! (-1 points).
• Emotional Setting. As presented in the module, the village of Barovia is a sort of sad, hollow place. It’s not entirely clear what’s caused this, however, or how Morgantha’s presence ties into the broader context of Kolyan’s death, Ireena’s fear of Strahd, and Doru’s participation in the rebellion. The fact that a substantial fraction of Barovia is just addicted to dream drugs goes completely ignored in the rest of the chapter. There’s no depth, weight, or context here—just a hollow conflict taking place in a vacuum. This is especially egregious because the players have no reason to want to follow Morgantha at all—unless she’s the first (and only) thing they see in the village, they’re liable to completely ignore her. That’s a failure for Narrative Design Factor #2 (+0 points).
• Source of Tension. There’s some small amount of tension here, for sure. It doesn’t last for very long, but it technically exists (“Will the players save the child?”). However, this is actually one of the rare cases where adding additional tension is bad. The village of Barovia should be the players’ main place of respite between the horrors of Death House and the many stressors of Vallaki. (Remember that good storytelling requires high-tension periods mixed with low-tension periods, to allow the audience and characters to decompress and assimilate what’s happened so far; Barovia should be one of the latter.) Morgantha’s inclusion takes away that sense of quiet sanctuary, and instead turns the “sad, but safe village” into a place that’s fundamentally unsafe. Bad for Narrative Design Factor #3 (-1 points).
• Meaningful Stakes. Nobody cares about this child. Nobody cares about their parents. Nobody really even cares about Morgantha, either—given how divorced the night hags are from Strahd’s storyline, you could even meaningfully argue that the coven is a distraction from the stakes of the main adventure, and therefore net harmful. Although the hags’ arc could be reforged to more strongly explore the themes of hubris, obsession, sacrifice, tragedy, and longing that suffuse the broader campaign, its refusal to devote any time to the actual addicted parents makes that a nonstarter. Failure of Narrative Design Factor #4. (-1 points).
• Satisfying Resolution. The resolution to this scene is…the drug-addicted parents (who are presumably still addicted to drugs) get their child back? And if the players eventually defeat the hags thirty sessions later, the parents…are still really depressed and guilty of being addicted to dream pastries and half-heartedly allowing Morgantha to take their children? That’s not just disatisfying—it’s downright harmful to the satisfying resolution of the campaign, which requires the players to feel good about the people they’re saving from Strahd. If you want to make your players the knight in shining armor, you need to give them a damsel worth fighting for, and Morgantha’s customers are not worth the effort. Failure of Narrative Design Factor #5. (-1 points).

Oof! Once again, we’re at a negative score. Though this scene offers some ostensible amount of tension, it does so in an unsuitable place and it offers no resonant characters, no meaningful context, no meaningful stakes, and no satisfying resolution. And to make matters worse, it actively makes it harder for other, unrelated arcs to achieve the same!

To look at this scene through the lens of dramatic questions, this is a single scene that has no meaningful stakes and an undesirable goal, and which diminishes the stakes, goals, and fairness of other dramatic questions. That’s…not great!

And what about the other two popular community content ideas I talked about at the start of this blog?

• Turning Argynvost’s body or spirit into a dracolich or shadow dragon: Seems like a really cool idea, but ultimately pollutes the broader narrative because Argynvost, as a symbol of hope and forgotten valor, needs to be symbolically pure for the Beacon of Argynvostholt storyline to work. Allowing Strahd or the Mists to corrupt him destroys that, thereby diminishing the satisfaction of the arc (Narrative Design Factor #5). Bad idea.
• Allowing Strahd to place a bottle of wine outside of Death House for when the players escape: Seems like a cool idea, but ultimately destroys future gameplay by indicating that Strahd is omniscient, which makes the players (wrongfully) believe that any efforts to evade Strahd’s spies and attention will be fruitless and therefore unfair (Game Design Factor #2). Bad idea.

Content has consequences. The hard part of prepping or planning a campaign isn’t coming up with ideas (especially in the age of Reddit, Discord, and ChatGPT).

The hard part is coming up with ideas that don’t break or otherwise detract from the rest of your campaign.

It’s a Mindset, Not a Secret Sauce

There's a good part and a bad part to taking this view of campaign design.

• The good: Better game and narrative design mean (hopefully) that your players should have more fun, remain more engaged, and make better memories.
• The bad: This is not one of those "one weird tricks" where you just run through a checklist and everything magically turns out okay.

I've boiled things down to their essentials here, but the actual process of figuring out which things are relevant to the analysis is very hard! That's a big reason why I tend to urge people so strongly to focus on elegance before complexity: It's technically possible to do a fantastic job of designing a sprawling, ambitious, and complicated campaign, but it's very very difficult to do so!

You're probably going to be better off working at a more manageable level and refining and improving the components that you have. The fewer balls in the air you have to juggle, the fewer things you have to keep track of, and the easier your job will be.

Over time, as you work within this framework, you'll start to develop a natural instinct for recognizing potential design flaws; rather than viewing each scene or encounter as a separate nugget of "content" in your mind, you'll start to see each individual component as a part of a much broader and richer tapestry. You'll start to see connections where you hadn't before, and start learning how to avoid any connections that may be more trouble than they're worth.

D&D campaigns can be deep, intricate webs of story and gameplay. Managing them can be stressful. It can be challenging.

But it can also be beautiful.

Campaign Advice Roundup | June 25, 2023

• Remember that players shouldn't be rewarded for failure. (For example, if a player character dies and comes back to life, the resurrection process should feel, at best, net-zero, not net positive.)
• Don't view dungeons as dungeons. Instead, view them as narrative arcs with lots of little pre-chosen branching paths. Try to figure out what incentives your players might have to take one path over another, and experiment with placing intentional incentives to lure them one way or push them away from another.
• Not all Bad Guys are (or should be) fight-to-the-death antagonists. Many Bad Guys are best used as social antagonists, antivillain allies, or even unsavory brokers that the players are forced to deal with despite their best efforts.
• The best combat hazards are those that force or encourage players and monsters to move around the battlefield.
• Remember that consequences should always be fair—players should always have a fair chance to predict what might happen as a result of their actions and to avoid it. If you ever find yourself saying "Are you sure you want to do that?" OOC, your game design has already gone off the rails.