Session 15A: The Labyrinth’s Machines

…and then arcs of purplish electricity began to leap between the bulb-tipped rods protruding from the machinery in the center chamber and the rods in the coffin chambers. From there, the arcs converged on the coffins themselves. The electrical bolts danced with a horrible beauty, filling the three chambers with harsh light that seemed barely dimmed by the smoky glass.

Since writing the Three Clue Rule, I’ve spent just over a decade preaching the methods you can use to design robust mystery scenarios for RPGs that can be reliably solved by your players.

Now, let me toss all of that out the door and talk about the mysteries that your players don’t solve. And that, in fact, you’re okay with them not solving!

Note that I did not say “that you WANT them not to solve.” And there’s an important distinction there: If there’s a mystery you don’t want the PCs to solve (for whatever reason), then you simply don’t include the clues necessary for them to solve it. That’s not what we’re talking about here: These are mysteries that the PCs can solve. They just don’t.

And that’s OK.

In fact, it can greatly enhance your campaign.

You can see examples of this in the current campaign journal with the strange machinery the PCs investigate and are, ultimately, unable to fathom.

“So what did it do?” Elestra wondered.

“Without occupants for the coffins, most likely nothing,” Ranthir said.

First, why is this OK?

Because the mystery is structurally nonessential. There is something to learn here, but the failure to learn it does not prevent the scenario from continuing.

When we, as game masters, create something cool for our campaigns, there is a natural yearning for the players to learn it or experience it. But that’s a yearning which, in my opinion, we have to learn how to resist. If we force these discoveries, then we systemically drain the sense of accomplishment from our games. Knowledge is a form of reward, and for rewards to be meaningful they must be earned.

Second, why can this be awesome?

The technique I’m suggesting is that your scenarios can be studded with literally dozens of these nonessential mysteries. At that point, it becomes an actuarial game. When including a bunch of these in a dungeon, for example, it becomes statistically quite likely that the PCs won’t solve all of them. (They might, but they probably won’t.) And the ones they don’t are going to create enigma; they’re going to create a sense of insoluble depths. Of a murky and mysterious reality that cannot be fully comprehended.

And that’s going to make your campaign world come alive. It’s going to draw them in and keep them engaged. It will frustrate them, but it will also tantalize them and motivate them.

It should be noted that one of the things that I think makes this work is that these mysteries are, in fact, soluble. I think that, on at least some level, the players recognize that. And the fact that there is, in fact, a solution has a meaningful impact on how you design and develop your campaign world.

(There’s also a place for the truly inexplicable. My 101 Curious Items is an example of that. But it’s a different technique and its effect is distinct.)

In the end, that’s really all there is to it, though: Spice your scenarios with cool, fragile mysteries that will reward the clever and the inquisitive, while forever shutting their secrets away from the bumbling or unobservant. When the PCs solve them, share in their excitement. When the PCs fail to solve them, school yourself to sit back and let the mystery taunt them.

If you’re wondering what one of these little mysteries looks like when the PCs solve it, check out Session 10 when they’re confronted with the gory remnants of an ancient tragedy:

When they approached the corpses they discovered that they were just two among many: Rounding the corner into the larger room they saw more than a dozen humanoid corpses strewn around the room. In the center of these corpses a massive, wolf-like corpse lay – it, too, had been horribly burned until its remains were nothing more than charcoal which had endured the ages.

Investigating the other chamber flanking the door of glass and bronze revealed an even stranger sight: Shards of iron had been driven with seemingly random abandon into the walls of the room. They recognized this iron: It was of the same type used in the cages they had seen before.

“Did something cause one of the cages to explode?” Elestra said.

“Or explode out of it?” Ranthir hypothesized.

“Like that creature in the other room?” Elestra said.

In this case, you can see how the PCs apply information they’ve gained from other encounters within the Labyrinth to shed light on the current forensic analysis. I hadn’t really planned that, per se. I just knew what the story was behind this room and how that story tied in with the rest of the complex. So when the players started asking questions about the metal fragments in the walls, I knew they were related to the cages they’d already encountered.

Note that this specifically works because these aren’t purely random bits of dungeon dressing. These nonessential mysteries aren’t just unrelated puzzles: They’re all part of the scenario, which inherently means that they’re also related to each other. Thus, as the PCs solve some of them and fail to solve others, what they’re left with is an evolving puzzle with some of it pieces missing. Trying to make that puzzle come together and glean some meaning from it despite the missing pieces becomes a challenge and a reward in itself.

This advice doesn’t just apply to dungeon scenarios, either. Any scenario can have these nonessential background mysteries woven around its essential structure: What’s the true story of how the Twelve Vampires came to rule Jerusalem? How did the Spear of Destiny end up in that vault in Argentina? What exactly was Ingen doing on Isla Nublar in Jurassic Park III? Who left these cryptic clues painted on the walls of the Facility?

In general, though, don’t think of this as something “extra” that you’re adding to the scenario. This stuff will usually arise organically out of the scenario as you’re designing it: You need to figure out what happened to the Isla Nublar facility, for example, to build your dinosaur island hexcrawl around its ruins.

This technique, therefore, can also help you avoid one of my pet peeves in scenario design: The incredibly awesome background story that the players have no way of ever learning about. Look at the background you’ve developed for the scenario: If there are big chunks of it which are not expressed in a way which will allow the players to organically learn about it, figure out how the elements of that background can be made manifest in the form of nonessential mysteries.

In the case of the Laboratory of the Beast, I knew that this section of Ghul’s Labyrinth had been used to work on several distinct research projects. Every lab, therefore, could be a separate little mini-mystery revealing exactly what Ghul’s arcanists had been working on down here and, thus, collectively also tell the story of these laboratories.

Session 14D: In the Beast Pits

As she moved to search it for hidden compartments or the like, the illusion screening it suddenly dropped away, revealing a wolf lying on the table. Its back had been carefully cut open and the flaps of flesh carefully pinioned to the table’s surface.

This moment in the campaign – with the illusion dropping away to reveal a splayed open wolf – was, in practice, a pretty good example of an “RPG jump scare”.

Generally speaking, without special circumstances, you don’t get actual jump scares in tabletop RPGs. Vocal narration simply isn’t a medium in which true jump scares can be properly performed. But there are broadly similar moments in which unexpected twists can be abruptly presented and provoke “oh shit!” responses from the table.

The illusion falling away, in this particular context, created immediate fear that the group was about to be hit with a trap. (A fear carefully cultivated throughout the Labyrinth of the Beast by a variety of immersive, disturbingly thematic traps.) The gruesome visage of the wolf itself capitalized on this fear, riding the emotional wave and using it as a channel for emphasizing the creepy imagery (and the even more horrific implications).

(Note the importance of the moment’s interactivity was also important here: If the PCs had just seen a dead wolf, the description might have creeped the players out a little bit. But the stasis field – and the implicit decision of whether to leave the stasis field intact or turn it off – forces the players to engage with the moment. That makes the moment more “real” and more meaningful than a simple description.)

But if you give this scene a degree of thought, you might be struck by a question: Why was there an illusion spell? Despite the moment playing out rather successfully, it seems a little odd, right?

The reason for that is simple: This isn’t how I prepped the scene.

WHOOPS…

What you’re reading here is actually the result of a mistake. When the PCs entered this room, I misread the room key and didn’t describe the wolf’s corpse lying on the table.

Doh.

Once I realized my mistake, I had several options:

First, the wolf no longer exists. I didn’t describe it. The environment has been interacted with as if it wasn’t there. So… it’s not there. Never was. The notes never made the jump from prep to the “reality” of what actually transpires at the table.

If you’re dealing with something nonessential, this can often be the easiest course to take. In the case of the wolf, it was, in any larger sense, nonessential. But it was very cool (if I do say so myself), and it would have been a shame to lose it.

Second, the simple retcon. “Whoops, I forgot to mention that there’s a dead wolf on the operating table.”

This approach is fairly straightforward, obviously. The drawback is that the open retcon inherently disrupts the natural flow of the game world’s presentation. Often this disruption is not so significant as to cause problems, but sometimes it is. One common example is if the PCs have already taken an action which they wouldn’t have taken if they knew the information you forgot to tell them. (Handling this specific example is something I discuss at greater length in GM Don’t List #1: Morphing Reality.)

In this case the PCs had not taken such an action, but I knew that the “retcon disruption” would blunt the impact of the imagery. (The players would be cognitively focused on processing the retcon instead of fully focused on the description.) And since the entire function of the corpse was its creepy imagery, blunting the impact of that imagery would defeat the purpose.

Third, swap rooms around. This technique works particularly well if there are multiple similar rooms in a particular area. For example, the PCs are supposed to find a dress with some weird stains on it. You goof up and forget to give them the clue in the Master Bedroom. I guess the dress is in the closet of a different room. (Or maybe it’s in the laundry room downstairs.)

So if there had been another convenient examination room nearby, I might have just moved the wolf corpse in there.

Fourth, create a reason why your screw-up wasn’t a screw-up. This is what I did with the wolf. Why didn’t they notice the wolf as soon as they walked into the room? Because there was an illusion spell preventing them from seeing it.

This basically moved the wolf from something noticed with passive observation (which is automatically triggered) to something that required a declaration of action from the players (i.e., interacting with the illusion). But you can apply the same technique in other ways, too: That NPC really should have told them about the Duke’s relationship with Countess Lovelace. Why didn’t they? Come up with an explanation. Blackmail? A hidden agenda that creates a conflict of interest?

The interesting thing about this technique is how often it actually creates additional interest: The RPG jump scare of the illusion dropping was effective. An NPC with a secret agenda is probably more interesting (and the scenario more dynamic) as a result.

The difficulty is that there was probably a reason why this additional layer of interest didn’t exist in the first place. (Or maybe not. Good ideas develop through play all the time.) It can be difficult to make sure that the continuity tracks on your hidden retcon.

For example, what if the players want to know why there’s an illusion spell covering the operating table?

First, you should have some rough idea of why the retcon makes sense, even if it doesn’t necessarily track 100% right out of the gate. In this case, my rationalization was that the wolf, in mid-surgery, is super gross. Nobody wants to look at that. Might as well throw up an illusion spell so you don’t have to look at it right?

Second, you’ll benefit from the fact that continuity problems that seem glaring to you behind the screen will be significantly less so to the players. The wolf-masking illusion, for example, ends up being pretty deep into fridge logic territory.

Third, you can just smile enigmatically. You are not obliged to pull back the curtain on your campaign and explain its inner workings. If something seems mysterious to the players and they want to figure out the why and wherefore of it, the obligation lies on them to take actions in character and figure it out. (In the process of which, you’ll probably be able to flesh out your initial rationalization to the point where it actually does make complete sense.)

CODA

It should be noted that none of these techniques are ideal. In an ideal world, you don’t screw up the room description in the first place, right?

Session 14C: The Temple of Ebony

They decided to start with the hall to the east: The iron doors opened onto a large room paved in glistening ebony. It rose in three tiers. Two horrific statues stood in the corners of the second tier. To the left a twisting pillar of coiled tendrils and to the right a squat monstrosity. In the center of the third tier, at the top of the room, there was a slab of black ebony…

One of the cool things about a well-designed dungeon is that you can never be entirely sure about what sequence the PCs will encounter content in. It’s like tossing all of your scenes into a blender: Stuff will end up juxtaposed in ways you never anticipated, prompting all manner of creative interactions and dynamic consequences.

For example, the very first room of the Laboratories of the Beast features three exits. Two of these expand outwards in ever-growing, heavily xandered loops. The third exit – which PCs heading straight ahead through the initial chamber would pass through – led to a single room: The temple of the dark gods described near the beginning of this week’s journal entry.

In designing the dungeon this way, I had lightly assumed that this temple would be encountered relatively quickly: It was right at the entrance! As you can obviously see, this ended up not being the case and the PCs actually spent prodigious amounts of time exploring this dungeon complex over the course of several expeditions before finally entering the temple.

As a result, they entered the temple dedicated to the dark gods after they’d already encountered Morbion, who worshipped one of those dark gods. The intended foreshadowing, therefore, was completely inverted: Rather than, “Oh shit, here’s one of those dark gods!” it became, “Oh shit! That dark god is a bigger and older deal than we realized!”

… except the PCs didn’t actually connect the dark gods of the temple to the dark god worshipped by Morbion, so the whole thing actually became a lingering mystery rather than being thematically connected.

With all that being said, of course, the design of a dungeon DOES impart some degree of sequencing, some or most of which will stick. So this is something you want to give some thought to when you’re designing things.

CHOKEPOINTS: There are points, even in a xandered design, where certain areas will lie beyond a chokepoint. For example, it was basically impossible for the PCs to reach the Laboratory of the Beast without first passing through the Complex of Zombies. Or, “If they go to Shardworld, it’s because they definitely passed through the Portal of Shattered Obsidian.” That sort of thing.

If there’s something that the PCs should know or experience before being in a given location, place it in a chokepoint they have to pass through in order to reach that location.

Note that chokepoints can also be non-physical: You need to find the gillweed gardens before you can extensively explore the underwater portion of the complex. Or you need to find the mending stone before you can use the damaged portal without ripping yourself into a thousand shards.

Chokepoints also don’t have to boil down to a specific chamber or door or the like. They can be more abstract than that: The PCs don’t have to pass through a specific room to get to the Lilac Caverns, but they do have to pass through the Golem’s Workshop complex. I sometimes think of a dungeon has having different “phases” or regions”.

How does the idea of establishing a particular piece of knowledge or experience sync up with this concept of a phased chokepoint? Well, think about the Three Clue Rule. You want them to know that the Golem was obsessed with cloning his creator before they find the creator’s body in the Lilac Caverns? Well, every facet of the workshop can (and should!) reflect that obsession in an almost fractal way, right?

In other words, rather than locking yourself down into a specific “this is the experience they need to have”, you can can kind of spew “experience DNA” everywhere and let the actions of the PCs slowly assemble a custom experience of their own out of it.

NON-SYMMETRICAL APPROACHES: Thinking of things strictly in an “experience A before B” way is a trap, though, and you should avoid it. When you’re looking at a key room (or suite of rooms) and thinking about how it’s structured, you can also embrace the fact that there are multiple paths to that area.

One of the things you can do to make your dungeon a richer and more interesting environment is to specifically make the experience of the area different depending on which direction you approach it from. A really easy example of this is entering a Goblin Princess’ bedchamber by fighting your way through the guards outside vs. slipping in through the forgotten secret door at the back of her boudoir.

DON’T THINK ABOUT IT: The other thing to keep in mind is that you don’t actually have to try to keep all of this stuff in your head while you’re designing the dungeon. In fact, it can become way too easy to start obsessing about this, at which point you can end up with a kind of hyperthyroidic version of My Precious Encounters™, except now you’re trying to carefully cultivate specific experiences by somehow conceptually corralling an incredibly complex and dynamic scenario.

You’ll generally be happier (and have better games) if you don’t rely on things playing out in a specific sequence. And, in reality, it’s not that hard: Use situation-based design as your primary design stance and things tend to work out. This is really the same thing as Don’t Prep Plots: Don’t prep pre-packaged experiences, prep situations.

So why bring up dungeon sequencing at all? Because it WILL have an effect on how a scenario plays out, so you shouldn’t blind yourself to that or simply leave it up to blind chance. I’m just saying it should be used at a broader, more conceptual level and primarily as a diagnostic tool rather than the primary motivation for your design.

For example, consider the Goblin Caverns of the Ooze Lord scenario that ran from Session 10 through Session 13. I designed the scenario like this:

Thus, the PCs encounter the goblins suffering the ooze infection before traveling to the ooze caverns from whence the ooze infection originates.

But I could have just as easily designed the scenario like this, with virtually no changes to the local topographies:

So that when the PCs come to the checkpoint manned by the (as yet unnamed) Itarek, as described in Session 10D, they would have the choice of turning left to the clan caverns or turning right to the ooze caverns.

(This would also change the logic of the world somewhat: In the scenario-as-designed, the goblins are – wittingly or otherwise – guarding the surface world against the Ones of Ooze. In this alternate version of the scenario, it’s far more likely that the PCs will encounter ooze-goblins coming up through the Laboratory of the Beast and infiltrating Ptolus. Which is what I mean by a situation-based design stance, right? But I digress.)

You could even go one step farther with:

So that the PCs enter the ooze caverns, probably fight Morbion, and only later discover the goblins who were being victimized by them.

(This wouldn’t work at all with the original seed for the adventure, which posited that the goblins in Grayson House were refugees who fled to the surface to escape the growing threat of the ooze-goblins. We could probably just swap it up, though, and have them instead be ooze-goblins who Morbion sent to the surface as initial scouts. But I digress again.)

It shouldn’t be too difficult to see how each of these macro-sequences would fundamentally change how the resulting scenario would play out at the table, right?

The specifics, of course, are all still up in the air: As I’ve discussed in the past, I didn’t anticipate the PCs forming a long-term alliance with the Clan of the Torn Ear. But that option wouldn’t have even been on the table if I’d sequenced the dungeon in a different way.

And that’s what you have to be mindful of: Not trying to pre-control outcomes, but figuring out what the most interesting chaotic bundle of unrealized potentials is and bringing that into play.