Timekeep. A ruined city. Destroyed in the recent Perfection Wars, as it stands, it is a cracked fortress of once-great technological wonders.

Its golden walls, set in the frozen mountains of the north, were created to resemble the faces of a clock from the air. Now they are shattered and leaning inward, yawning under their own weight, their finery in contrast to the rubble that lines most of the inner city. The great forges that littered the mountain have collapsed inward upon themselves and in the city depths, all manner of things now dwell amongst the ruins of steel laboratories, creating a vast sprawling dungeon of great inventions and unleashed horrors.
Deeper still, the implosion of the forges when the War was lost cracked down into deep, old, civilisations. Somewhere in the bottom of Timekeep, these elder nations now examine these interlopers to their underealms. Passages deep beneath the mountain lead to old drow colonies, abandoned fey halls and frozen glacial warrens.
Timekeep is a place of adventure and danger. Many places and people were lost amongst its ruins when the city collapsed. Adventurers could have many expeditions here, trying to find things they were looking for. Finding entrances and exits that riddle the upper streets where cracks have riven the mountain, diving into the forges for evidence of what was lost, gaining maps and resources, then returning to the city above to restock to explore once more. It’s a magnum opus of dungeon crawling. 900 or so rooms full of factions, personalities and treasures connected to the surface intrigues and greater world politics.
I really wish I’d made this clear to the group who I sent into Timekeep early in 2024. Because their experience was not this. And I think what happened tells us a lot about how D&D has split into differing branches of play, how we explain play and how not to run a ‘mega-dungeon-like-structure’ for any game. So let’s go through what happened and what we can learn.
Pre The Foray
Timekeep has sort of accidentally become a place most longer-running campaigns in my D&D world visit. It was ruined during my first real 5e campaign and was a focus for an important villain group in another game, with the group diving into the mountain to find and defeat a cult. It even featured as a brief port of call for my spelljamming group. So I had reason to think it would go well in my long running Crone Kingdom Rebellion game as a change of pace, more of a return to standard D&D expectations with a tie to two of the backstory elements of the group, one searching for a friend who was locked in the body of a contruct buried somewhere in the mountain collapse and another searching for the alchemical processes to help their wife, who had been transformed into a machine-assassin by a campaign villain’s old laboratories that were located in the mountain. So when the group arrived in the city, I was optimistic.
The Crawl Begins
After some intrigue in the city went well, the group moved into the mountain. And I think here’s where I made my first mistake. Up until this point, dungeons in the game were 5-15 rooms. The sort of thing you go into and explore most of and don’t come out until you’ve solved it. They’d infiltrated a patrolled dam, liberated a city of catfolk and such.
But they’d not met something that had lots of disconnected parts that required more effort to go in and just keep plugging away at until you find a way through, rather than resurfacing and discovering different routes. They’d also never come across dungeons that had lots and lots of factions in them. I see now that setting these expectations inside the fiction might have helped the group make different choices. Because when we think about modern D&D as opposed to older experiences, lots of parties never think about leaving the dungeon and coming back: This is a factor of older play where rests took longer and supplies were more exhaustible. In the modern game, the idea that you could get trapped under a place and would need to retreat or die is far rarer. I could see that as the game progressed, players weren’t going to ever take that option.
Mapped

Accompanied by a group who had gone in to clean out some of the forges and rescue trapped warforged, the group went straight down and fought through rooms, eventually finding themselves in a hub with maps printed on the walls.
I now see I should have had these maps as a handout. Instead, I explained it to them as a group at the end of a session. By the next session, the group had sort of a muddy, half-remembered understanding of the details. Even after re-explaining, they understood there was a way down to where they thought they needed to go, but had no way of physically seeing that it would be much easier to resurface. They had it in their subconscious that there must be a way through – if they just kept plugging.
And there was, but gods, it was a grinder. They ended up getting battered, separated, washed down a drain hole, forced through a drow spider pit, and weaving through a horrid part of the mountain that eventually brought them to a place where they could ride a sort of railcart to a part of the dungeon they needed to be. I think with maps, the vastness of the journey in front of them, if they chose this route, would be way more apparent. Or if they were making maps themselves, but I hadn’t fostered this culture.
We can start to see here how sometimes the worst thing you can do is to change playstyle mid-stream. I had moved so far into a different field of play that the players were a bit lost. Some of them were still enjoying themselves, and the dungeon held a lot of revelations about various backstories and stories. As the players began to align with where I originally expected them to be, they began to hit story beats I had placed closer to their eventual goal. But for a good few sessions, it was pure, gruelling, unending, dungeon.
Fractioning Factions

As mentioned before, the dungeon had factions. Many factions. The drow, who were actually two separate factions, a hidden fey court the players could unlock and talk to if they placed certain stones in the right place, a duregar group who had were working with the scientific marvels they knew where buried there, an enslaved small illithid colony, disciples of a entombed God and finally the villain who’s lab they were after was secretly still alive down there with minions. I’d planned for some of these factions to be allies, some to be enemies, some to add intrigue.
But I’d not really made this clear in my play expectations. Lots of the factions became sort of temporary set dressing, represented by the first member of the faction the group found. I’d also underestimated the group’s desire to ‘get the job done’. They felt like diverting to talk to groups or investigate things like fey stones was a distraction and cost them time and focus on their individual missions. So they powered through.
This is an exercise in both knowing your group and providing some signposting about the fact that some factions in a game can be helpful. With the drow warrior they rescued, I hinted at the idea that he might be open to helping them and be a little more flexible about outsiders than you’d expect from a drow. The group really hadn’t taken it. By the end of the dungeon, I had altered some encounters to make introductions far less subtle. The surviving member of a lost frost giant expedition, looking to free his friends, made his new, improved entrance by kicking in a wall and killing people the players needed to get past. An alliance here felt more easy for the players to see as a net positive.
When we’re thinking about factions outside of a place like a city, having a moment of the faction demonstrating itself is useful. And as I have said before, there is some expectation from the players that not every room is going to just be a challenge to get through. In a dungeon, there aren’t so many ways for a group to learn about a faction by word of mouth. You have to set up what they are about in a series of shown elements to the players, rather than just expecting them to do all that questioning.
As we can see by this burgeoning article, there was a lot to learn from this experience about how we signpost and how we need to be aware of the expectations of our table before we begin to plan. I hope these revelations from the table help you when running a mega-dungeon.
Image credits: The Gate by Emkun, Grim batol by Holma2208, and Rains of Castamere by Ertacaltinoz.