Steering the Ship: What I Learned from Changing Course Mid-Campaign

 Whenever I start a new campaign, I usually have a personal goal as a GM—some way to step out of my comfort zone and grow. Maybe I want to try a new system, experiment with homebrew mechanics, or explore a type of setting I haven’t mastered yet. Earlier this year, that led me to His Majesty the Worm, and with it, a challenge I had previously failed: running a satisfying megadungeon.

I had tried before during the pandemic, running Barrowmaze for a group of friends over several intense days. It quickly became repetitive, and after a series of unfortunate deaths involving bottomless pits, we all gave up on it. This time, I wanted to make it more engaging. I designed the dungeon as a modular space that rearranges itself whenever the players return to the surface. Thematically, this made sense—the megadungeon was a dreamlike realm shaped by a colossal worm that feeds on the desires and dreams of adventurers. A shifting layout reinforced the dream logic.

In His Majesty the Worm, players descend into the dungeon with a clear quest in mind. This keeps them from kicking down every door just to see what’s inside. I thought that, with this structure and my modular twist, I’d solved my old megadungeon problems.

Several months and many biweekly sessions later, the party was still on the first floor. That floor has over 120 rooms—they’ve seen about 80% of them. They’ve been carefully mapping each module, helped occasionally by a few dream-linked gnome NPCs who can sense the ever-shifting layout. But despite their enthusiasm and strong roleplaying, something wasn’t working for me. They were having fun discovering strange traps and NPCs (like their archenemy Jimmy Sushi, a sword-obsessed nerd with four geisha bodyguards), but I, as the GM, was beginning to feel burned out.

The dungeon felt directionless. The players could only orient themselves with help from the gnomes, and every time they left to resupply, the map reset. Their progress felt ephemeral. I found myself considering wrapping up the campaign with a final boss fight and moving on.

Then I remembered: I’m the GM. The god of this world. So I made a sharp turn.

The party faced the Serpent’s Child, a massive, brutal enemy on the rooftop of a temple dedicated to the Worm God. It was an epic, memorable fight. And it was a test. The god declared that one among them might be chosen as its true herald—but first, they must prove themselves. With a flash of divine power, the entire party was teleported to a previously unseen level of the dungeon.

This floor doesn’t shift. It’s under quarantine. A dream-plague has taken root, and the worm god has sealed the area until the infection is purged. The twist? The god has told many adventuring parties that they are the chosen ones, all of them sent to do the dirty work. My players don’t know this yet—they remain loyal followers of the Worm God—but the revelation will come, and it will hurt.

---

Lessons Learned

This whole experience taught me a lot, both about megadungeons and myself as a GM. I’m also an industrial designer by profession, so I’m always thinking about mechanics, systems, and user experience. That lens has shaped how I look at campaign design—and this one gave me a lot to think about.

1. You can always change course—gracefully.
Just because a campaign isn’t going the way you imagined doesn’t mean it’s a failure. Sometimes it just needs a hard turn. And if you do it with care and style, your players will come along for the ride.

2. Players need informed choices.
This was my biggest realization. I now understand why I enjoy running hexcrawls more than megadungeons. In a hexcrawl, players have roads, rumors, landmarks, and cities. They know where they're going, even if surprises lie ahead. In my megadungeon, despite efforts to telegraph rooms with smells, sounds, and temperatures, it often came down to opening doors at random.

3. The dungeon needs navigational tools.
In His Majesty the Worm, the default rule is to give players a numbered map with no room details. I ignored that, because I love when players draw their own maps. But in a dungeon this big, that decision may have backfired. Next time, I’ll include clearer directions, treasure maps, or NPC guidance. Some things just need to be tested before you homebrew them.

4. Quests matter—even in the dungeon.
The quest system in His Majesty the Worm helps a lot. It gives players purpose beyond just looting tombs. That narrative drive is key in a space as disorienting as a megadungeon.

5. The players’ fun ≠ the GM’s fun.
Everyone at the table might be enjoying themselves, and that’s great—but if you’re starting to dread prep or feeling uninspired as a GM, that’s worth addressing. Your energy and excitement are part of what keeps a campaign alive.

6. Justifying a mechanic in the fiction doesn’t automatically make it fun.

One of the coolest ideas in my campaign was that the dungeon shifts every time the players return to the surface. It made perfect sense in the world: a dream-consuming worm reconfigures reality as it digests desires—of course the dungeon reshuffles. It’s thematically rich, surreal, and eerie.

But after months of play, I realized something important: even when a mechanic makes sense, it still has to be enjoyable at the table. The shifting modules created a sense of mystery, sure—but they also erased the players’ progress, undermined their sense of place, and made it hard to feel like they were advancing. It was fun at first, but over time it wore thin.

A good narrative justification can support a mechanic—but it can’t rescue a mechanic that’s frustrating or draining. As GMs, we sometimes fall in love with clever ideas that make sense in the lore, but the real test is whether they generate satisfying play. If not, maybe the fiction should bend, not the fun.

---

I’m glad I didn’t end the campaign. Changing direction brought it back to life and reminded me that being a GM isn’t just about building worlds—it’s about adapting them too. When something feels off, it’s worth stepping back and asking why. Because in the end, a campaign should be fun for everyone at the table—including the one behind the screen.