Mapping a River for pointcrawling

While tinkering further on my Rivercrawl idea, I cam up with this notation to map a huge river network.

First I made a Melan diagram of the main river branches for my river and marked the branches in different colors, which then looks like this.

I then turned the same information into a big table. Below you have a heavily cropped down version to show the principle of how it works. The real thing I made actually has 180 rows over five pages, but most are still completely empty at this point. The principle is basically the same as in Ultraviolet Grasslands, but without the illustrations. I find this easier for river that curves and fans out, compared to the more or less straight trade routes in UVG, and it also allows to make more notes without making a huge unreadable mess. As a tool for GMs to use at the table, I think this plain look isn’t a bad thing.

Getting the whole thing set up was a bit tricky, so here’s how I did it: Since I have only three main branches at any given point of the river, I made a table with seven columns. I think you could also do it with four branches and fit nine columns on one page, but more than that probably makes the thing more a nuisance to read than a help. I found that my river has seven different combinations of parallel running branches, so I made the table with seven rows as well. At this point you first set the column widths that you want, because this will be a total bitch if you try adjusting those later. After that you merge cells together as the river branches fork and meet, which in my case looked like this.

At this point, you can simple select and row and use “add rows below” or whatever your program calls it, and you should get an identical row to the one that you had selected. Then add mile markers to the leftmost row, and you’re done.

Now to the new neat feature that I actually came up with myself. The River Ratings. Each river section row has a little field on the left side that quickly shows the GM the water conditions the players are moving into. It’s fairly self-explaining when you look at the legend above. The letter says what size categories of ships can enter that section of the river. In case of my emerging setting, it’s galley size, junk size, dhow size, and canoe size. Ships larger than that will get stuck on the bottom of the river. (The width of the river or any obstacles in the water are not considered as a separate factor for the sake of convenience. Either your ship can continue on, or it gets stuck on something.) The number indicates the speed of the current. This number is added to your ship’s speed when you travel downriver, and subtracted when you travel upriver. If the speed of the current reduces your speed below 1 mile per hour, you can’t continue by water. I had been thinking to mark the type of terrain on the riverbanks as well to calculate overland speed, but that would mess the readability of this format with too much clutter. For the setting I want to make, it’s going to be “dense forest” pretty much everywhere anyway. I did a bit of looking around for average speeds of the boat types I listed, and the numbers I went with seem to be quite realistic. They are actually leaning to the lower end, as I suspect the original numbers were based on strong ocean breezes, so it would be slower further inland.

My plan for the campaign is that there is an adventuring season of 8 month, which is then interrupted by a flood season of 4 month, where the water speed is simply too much along the whole river to get upstream. I think it would be cool to make a roll at the start of each new adventuring season to see if water levels are exceptionally high or low this year. A high river increases the size rating for the whole river by one, while a year of low water levels reduces it by one. The players might find that the expedition they had planned either needs to be canceled or attempted with a much smaller boat as the river conditions make reaching the destination in a junk or dhow impossible. You could also have a randomly determined special event that changes the water level or speed halfway through the season, which can lead to very inconvenient complications hundreds of miles away from civilization.

Row, row, row your boat, bravely up the stream

So, if you have a setting idea that is not centered around kingdoms and cities, what other reference frames can you use to give structure to the peoples and societies of a vast wilderness setting? How about rivers? All the earliest civilizations of the Bronze Age first appeared along the largest rivers in the old world because big rivers are really really useful. They provide a steady source of water, which in the sub-tropical zones where you find these civilizations can otherwise be quite a problem. But they are also extremely useful for transportation. Rivers allow you to transport large quantities of cargo just as easily as by rail. Load all the stuff on a boat, add a sail or go with the current, and wait until you’ve reached your destination. If you have goods to move, rivers are the way to go. Or to float. While water isn’t as much a problem in Central Europe, the region between Germany and France has been constantly contested for many centuries because it’s the origin of the Rhine, the Seine, and the Rhone, having easy access to the North Sea, Atlantic, and Mediterranean.

It’s been one of the design elements for Planet Kaendor very early on that all civilization has to be on rivers or the coast, with the deeper forests being more or less inaccessible for heavy cargos. When I was thinking of city states, I was mostly thinking about the coasts and large ports, but that’s not where the adventure is. Adventure is deep in the forests where the ancient ruins are hard to get to. I am now thinking about moving all the pieces on the map to correspond to three huge river systems and one archipelago of islands of the coast. I really like ocean beaches, but Planet Kaendor is meant to be a forest world foremost. While there won’t be any along the major rivers, there’s more than enough in the islands region. While I have plans for a sub-arctic and a tropical forest set in Kaendor, for practical reasons it makes the most sense to only go with the temperate-subtropical one for now. I think any single campaign is best served by being based entirely on a single river.

The Setting

Since this first river is located in what I used to call the Dainiva forest, I’m going to call it the Dainiva river here for convenience. And since nothing is a permanent as a temporary fix, that’s probably now going to be its name forever. The great Dainiva river has been the home to many great civilizations over the ages. Cyclopean castles of the giant Rock Carvers overlook the river from cliffs towering over the meandering courses of the upper rivers, with the lower river being home to many old Naga cities. Ruins of the sorcerous Tower Builders rise above the dense trees flanking the river banks, as well as the magnificent living citadels of the Tree Weavers. All these civilizations have long ago faded from history, and it was many centuries after the Naga retreated to the jungles of the south that mortal peoples began settling on the lower banks, gradually but cautiously moving into the abandoned palaces of the serpentmen. Among the ruins they discovered the arts of casting bronze and mastering the secrets of alchemy, leading to the rise of the first mortal civilization. Over many centuries and generations, explorers ventured further up the waters, but even a thousand miles upstream, there were still no signs of the headwaters of the major branches. Only more water and trees, and the wrecks of explorers who had gone before them. And more ruins and monsters.

The Map

A setting of this type is perfectly suited for pointcrawls. Since travel is basically linear along the river branches or their banks, and ruins have to be visible from the river for characters to have any chance to find them, using a hexmap would not provide any actual benefits. Instead, a map showing the various main branches can show the distance between any fork, settlement, and ruin right next to them, and you can also use color to mark different types of water. For example, the common speed for rowing a canoe with no current is given in most places as 3 miles per hour. Currents of 1 or 2 miles per hour also don’t appear to be anything unusual, and while many rivers are much faster, the current generally is slower when you stick close to the shores where it’s more shallow. So you can mark the river conditions in three colors. Dark blue for the slowest water, in which rowers go 2 mph upstream and 4 mph downstream; medium blue for faster water, in which rowers go 1 mph upstream and 5 mph downstream; and light blue for waters too rapid to paddle against, that require continuing on foot. But you could still build a single-use raft from trees and go downstream at 6 mph. If you want to, you can also convert straight from miles per hour to miles per day, if hourly precision isn’t desired, but if you don’t have to deal with things like traveling 2.33 hexes in a day, I think tracking distances by the mile isn’t really any nuisance. On the major branches of a river of this size, there is easily more than enough room to navigate large cargo ships like a junk. With a slightly more sophisticated sail than just a plain square cloth, it is possible to sail up a river against the current, even with quite moderate wind coming from the sides. Merchant ships like these would replace the trade caravans seen in many land-based settings.

Settlements are all either directly on the river or at least have an accessible pier that connects to the actual village by a short path. Since they would want to be visited by traders, such piers would be clearly visible. But you could also have lairs of rivers pirates or secret cults hiding in barely visible side branches much too small for larger merchant ships. With civilization being based along the lower river near the coast, settlements become more scattered and smaller in size as one travels upstream. This can be used as a great indicator for players about the dangers they can expect to encounter. In civilized areas on the lower river, big monsters have long been driven out, but all the best ruins have been picked completely clean generations ago. But on the upper river, few mortals have ever set foot and there are both more dangerous monsters and much greater treasures to be found.

Since traveling on water is relatively simple and allows for the transport of great loads with little effort, I think a campaign of this type works best if you make it really big. Make it a river as big as the Volga, the Mekong, or the Columbia, where characters can go exploring for months between the end of the spring floods and the onset of winter. With the help of rafts, parties will be able to return with huge hauls of treasure, so the journey back to civilization should be a long one to compensate. Bigger hauls should translate to fewer hauls.

Basing a sandbox around a river system is also really convenient for a GM. By its nature, its close to a fractal, allowing you to just keep expanding it with more and more side branches as the party continues exploring upstream. A river map does not have to bother with mountains or elevation, and generally there’s no need to be exact about the width and depth of the water. And if you should end up with a branch that gotten too narrow and shallow to continue on, the party can always go back downstream a couple of miles and go up another branch. Now for the lower river, I think the players should have a map of the main branches and major side branches, as those are areas frequented by river merchants making their regular round. But once you leave civilization behind, there’s no limit for how far you can continue.

Similarly, it’s very easy to create villages and ruins in a vacuum and just plop them down on the map wherever the players decide to go. That goes a bit against the common ethos that players should have control over where they go by making informed choices, but I think in a setting like this, there really are not a lot of choices to make. Check it out or continue up the river? And given how many branches a river system of this size has, I don’t think working with fixed locations would actually be feasible. You’d end up with a lot of “this branch gets too narrow to continue and you’ve not seen any signs of a ruin”. That’s not player agency either. You could very well establish some facts about a ruin when the party stops at a village or trade post and gets a tip from the locals. But there wouldn’t be any need to establish any of this before the party arrives at this part of the river.

Encounters and Sites

I think for a campaign of this type, random encounters might actually the bread and butter of many adventures. Ruins are cool, but when slowly travel up a river for hundreds of miles, you’ll be doing a lot of encounter checks.

In a world with river merchants, you’d also get river pirates. Those pirates would know not to bother explorers going up the river in the spring, unless they are desperate for supplies, but be waiting to pounce at any explorers coming back down the river in the fall with their big hauls of loot. Merchants might invite the party to get a free ride with no paddling on their ships in exchange for protection against pirates while they have the same route. On the upper river, you can have encounters with aquatic and semi-aquatic humanoids, who could either be friendly or hostile to rare visitors from downriver with goods to trade. There can be the wrecks of failed expeditions, which might even be salvageable and be sold for a huge profit if floated down the river without sinking. Or repaired and used for further expeditions the next year. Or there could be ancient crumbling dams from the old civilization that threaten settlements downstream with disastrous floods, allowing for some variation between dungeon crawls.

And then there’s of course the river creatures. Obviously crocodiles and big snakes, but I’m really giddy at the idea of giving players a paralyzing phobia of hippos. Someone suggested to me adding dire beaver dams to block of some rivers and require hauling boats over land to continue. I also really like the idea of creatures in the trees following the players in their boats from shore, waiting for an opportunity to attack.

It really is a fairly simple concept for a sandbox setting, but one I think has huge potential, while looking very manageable at the same time.

Call of the Planes

When I wrapped up my 5th edition campaign last year, I was pretty fed up with the system for it just not being the kind of game that works for the kind of campaigns that I had created my setting for. It also made me throw a small pile of notes into the corner that I had scribbled down for a Planescape campaign. Last week I came across a discussions about planned campaigns we never got around to run, which reminded me of those ideas. And here I am now, picking up right where I left a year ago.

Blade Runner
Dark Souls 3
Dark Souls 3
Dark Souls 3
Fury Road
Fury Road
Hellblade
Hellblade
Hellblade
Metro: Exodus
Metro: Exodus
Metro: Exodus

And yes, I am super pumped for Carceri. (But also Ysgard and Pandemonium.) And the factions I want to include in important roles are the Bleak Cabal, the Doomguard, the Dustmen, and the Revolutionary League.

With references like these, keeping the thing from going all grimmdark will be one of the priorities, but I think with the quirkiness of Planescape it shouldn’t be too hard to find a good balance. If everything breaks, a Solaire and Siegmeyer duo should always be able to save the say. With jolly cooperation

Planescape is a setting designed around the rules of AD&D 2nd edition, and no way in the Nine Hells am I going to try learning that mess of a game. I think 5th edition will do the job just fine. Adjusting my plans to what the system can provide might work out much better. I also think aiming to have the players reach a new level every four or five games was too fast a pace. Decoupling advancement from fighting enemies and aiming more for a level every 6 to 8 games should play out better and not feel as rushed and overloaded.

Buried Ruins

Hey look, I got a useful (and very short) video to share. People don’t seem to this very much these days.

I knew about ruins getting buried, but didn’t think the effect would be this pronounced. There’s some really good examples at the end, which is something I now think I really should include in most dungeons I make.

Look at my Works, you Mighty, …

When I wrote my first post outlining the purpose and goal of my latest worldbuilding efforts, I got a lot of replies, including this interesting link from Oliver Simon to a great academic article about the role of the environment in Conan: Exiles. It gave me a couple of great ideas for Planet Kaendor and I made some notes for a new post, and then completely forgot about it for the next six weeks while writing other posts I had already planned. Of my three notes, I only remembered the meaning of the first one, and I had to reread both the entire article and my post that led to it being recommended to me to figure out the second. And I still have no clue what idea the third note was supposed to be about. Don’t be me! Take better notes when you get great ideas for your own work.

The article takes a look at the open world survival game Conan: Exiles as a horrifying analogy of the cruel exploitation within human economic activity. While many of these survival games have the killing of other player characters and the looting of their equipment and resources as a key gameplay element, this game uses the well established and accepted norms of the Hyborian Age to take it to a much more grotesque level. Not only can you loot the possessions of your slain foes, you can also butcher their dismembered corpses for meat and crafting materials. Nonplayer characters can be taken alive and made slaves, that are a hugely important resource for expanding your own base. And some magical powers require human sacrifice to attain. Other people are reduced to “human resources” in the most literal sense. They are commodified as tools to be exploited for your own quest for wealth and power.

Being basically a sociological paper, it’s not an easy read, and the first part is crammed full of attributed quotations of other writers that don’t really add to the topic and mostly seem to be there to pad out the references list at the end to boost academic credibility. But after that it goes into how the environment with its ruined buildings, abandoned weapons and tools, and human remains also tells a story of how economic exploitation build the fallen civilizations as well.

I didn’t expect Marx and Derrida to contribute to my worldbuilding, but I guess stranger things have happened. One passage in particular really got me thinking about how I can give the ruins that fill the environment of Kaendor a more meaningful presence in the actual game instead of being irrelevant pieces of lore in some file.

Faithful to Howard’s original Conan stories, the landscape is one which Derrida would have recognised as being distinctly hauntological; it is a world scarred by its past. This environment is shaped by forces which still have agency but no agent, generating effects which exert great power over the player’s experience of and interaction with their surroundings.

Agency without an agent. Now that’s an expression that really appeals to me. In this case, we have to treat agency as different from the concept of player agency in roleplaying games, which is the ability of players to make choices about the actions of their characters that meaningfully affect the outcome of the developing story of the game. Without an agent, there are no choices that are being made or actions that are being taken. Instead we have the idea here of choices that were made and actions that were taken long ago by the long dead builders of the ruins that led to the creation of the current environment. The choices were made many centuries ago, but their consequences still affect and constrain the options that are open to the player characters in the present. The idea is that the ruins are not simply featureless and inert stones that litter the surroundings, but active entities that challenge and threaten the intruding explorers.

(At this point I want to apologise for falling into academia speech and getting abstractly philosophical. Four years in cultural studies do this to you. It comes automatically and leaves it marks on you forever.)

As a simple example, take an ordinary arrow trap that shots an arrow when someone steps on a certain part of the floor. To quote the endlessly poetically and quotable Darkest Dungeon:

Curious is the trap-maker’s art… his efficacy unwitnessed by his own eyes.

A trap is not simply a feature of the environment, even though many games treat them like that. An actual trap is not simply just there. It is there because someone made a choice and took action to put it there. With an intent to kill. The builder of the trap does not know who will fall victim to it, and the people it injures will most likely have no idea who caused them harm. In a ruined dungeon, the builder will have died centuries before the victims were born. But still, one person exercised agency to cause serious harm to another person. The agent is long gone, but the effects of the agency are present in the present.

It is not just traps. Every artificial obstacle that characters encounter is there because someone put it there with intent. Every constructed tool or weapon they find is in its place because someone made it for a purpose and put it in its present location as a consequence of choices and actions. And this even carries over to much larger scales. Planet Kaendor is conceptualized as a world in which the natural environment is outside of the control of mortals. Whatever they do in their attempts to shape the world that surrounds them will quickly be negated once the wilderness returns. But things look very different when it comes to the marks left behind by unnatural sorcery. A charred wasteland of ash haunted by ghouls attacking careless travelers surrounding a ruined city does not exist randomly. Its existence is the consequence of choices made by a sorcerer long ago. A consequence that directly affects the characters in the present. And the natural world in Kaendor is often directly controlled by spirits, who can exercise their own agency as well. Though being essentially immortal, this does not fall under the agency without agent. They are merely agents with an invisible presence, but I think the overall effect is quite similar.

The key idea that is presented here (indirectly) to worldbuilders is to create environments for adventures that are not simply passive and interchangeable backdrops that maybe have a couple of unusual but random backgrounds in them. Instead, they should give indications that obstacles and useful finds are the result of someone’s deliberate exercising of agency. Bad things don’t exist in the present at random. They exist because someone in the past wanted it or did a careless mistake. There was a purpose behind it and it was someone’s fault. And that someone’s presence should still be felt as a malevolent force seeking the destruction of any intruders, or a shade lingering among the ruins of its crumbled dreams. Of course this might not be a universal requirement for all fantasy environments. But I had written about looking for ways to give ruins and magical places are more active role in my world, and this article provided great insights on how I could be moving closer to that goal.

All the mentions of cannibalism also got me some interesting idea for ghouls, which have been a favorite of mine since I encountered a new take on them in Dragon Age, and have held a very important position in my deliberations about the nature of the supernatural in Kaendor basically from the start. But those deserve an entire post of their own.

The MEGACHURCH!

I am not usually a fan of megadungeons. Or to be more precise, I am not a fan of the concept of mega dungeons.

Dungeons that are on a colossal scale are a different story.

In a moment of inspiration, I got an idea for an awesome buildings to put somewhere in the forest in the Green Sun campaign and while I was doodling around with possible floor plans, it turned into something increasingly more concret and fun. It’s an absolutely enormous temple that looks similar to this.

But scaled up to this.

And on the inside, I think it would look something like this.

The floorplan consists of 29 vaults that are 30 x 30 meters at the base and between 120 to 180 meters in height. To not go insane with the maps, the platforms and bridges are arranged into 5 levels.  To hold up the roof, there are also 4 solid “towers” of 30 x 30 meters in the corners, which have 10 internal levels of rooms, and 4 in the center of the building that have 12 levels.

That’s just shy of 200.000 square meters of floor space. Or 5,350 5-foot-squares. However, in practice most of that will either be open air for the vaults or solid stone for the towers. That’s still something like 400 “rooms”.

This should be fun.

Pointcrawling in a Dungeon!

Formulating my thoughts on a subject always helps my putting my ideas into focus. And so after writing yesterday about the challenges of using a pointcrawl map for a large dungeon I came up with some new ideas for how this problem might be solved.

Pointcrawling really is all about navigation by landmarks or travel along marked paths. During overland travel this is usually the only way there is to pick where you are going. This makes is very easy to take a hex map and convert it into a point map by simply removing all the hexes that don’t have keyed sites or encounters or lie along a path that connects two of the keyed locations. And you can easily convert a point map back into a hexmap by simply overlaying a hex grid on the chart of sites and paths.

But with dungeons this is not the same. As people discovered very early on in the history of RPGs, dungeons are very different environments from outdoor landscapes. In a dungeon you don’t chose your path or destination, you can simply go forwards and backwards in a corridor and pick from a fixed number of other corridors at junctions. When you want to go dungeon crawling with a point map, you can’t just take any grid map dungeon and convert it to a point map. This doesn’t work and you only end up with nonsense that has the same complexity as you started with. Instead, you have to design the entire dungeon from the ground up with a point map in mind.

A dungeon that would be very easy to do with a point map is a ruined castle with an open courtyard. In fact, Chris Kutalik’s ruincrawl is really an attempt to explore cities, not buildings. With an open castle, you can have the first area be the gatehouse that has one entrance and one exit. Once the players reach the exit of this area they can overlook the courtyard and see the various buildings that make up the interior. There may be a keep, a chapple, stables, a tavern, a warehouse, an inner gatehouse, and lots of uninteresting small sheds. The sheds and the patrol corridors in the walls have nothing interesting in them so they can be ignored and are not getting mapped. The other buildings each get their own small map and are marked on the point map as distinct areas. From their position at the exit of the outer gatehouse the players have a clear path to the stables, warehouse, tavern, and inner gatehouse, and if they go through the inner gatehouse they can also walk over to the chapple and the keep. What the paths in the courtyards look like is irrelevant. That’s a basic and simple pointcrawl.

This doesn’t work in a dungeon that is entirely underground, though. Simply because you can’t see other areas from a distance. You can’t say “let’s go over there” when you can’t see “there” and don’t even know that “there” exists. Ina grid map dungeon this is not a problem since you simply follow the corridors and then see where they lead to. With navigation by landmark being out of the picture, following marked paths is the only remaining option. But the whole point of pointcrawling a dungeon is that you want to strip out all the empty halls and chambers that make up the majority of the place and only have the players interact with the most interesting areas. I think the solution to this conundrum is to design dungeons for a pointcrawl along a network of highways. Coridors that clearly stand out from the rest and serve as main connective routes between major prominent area of the dungeon. Dozens or hundreds of minor side passages branch off from these leading to living quarter and storerooms, but they are not included in the map and can not be actively explored by the players. To look at a part of the dungeon in detail there has to be something that visible stands out and draws the attention of the PCs to the fact that there might be something worth looking at.

This is of course a few steps removed from the complete hands off approach you can have with grid map dungeons where the GM really only tells the players what they see and what happens as consequence of their actions. By designing a dungeon like this the GM takes the previlege to tell the players that their characters are not interested into exploring certain corridors and rooms and that they simply continue on their exploration along the provided path, which might not sit completely well with everyone. But really, this is exactly what outdoor pointcrawls do too. It’s an aknowledging nod towards the fact that games are abstractions of adventures and not accurate simulations. It takes away player choice, but really only choices that the GM knows in advance to be completely irrelevant. I don’t think many players would feel sorry about not having to explore 80 empty rooms in a room before they find a room that has something in it. When you look at fiiction this happens all the time. If you take a book or movie at face value, characters are instantly teleporting around all the time between scenes. But everyone is in agreement that the characters did walk between locations and that they probably talked during that time and perhaps even had some encounters that are completely inconsequential to the story. This is the underlying idea here.

A secondary issue I had been pondering is how to indicate the border of an area when each area consists of multiple rooms and is surrounded on all sides by more or less indentically looking rooms. What I think might be the best approach is to construct the whole dungeon not as one continous cluser of rooms but instead as multiple self-contained cells, which each cells having only two or three exits that lead to other cells. This is one of the things that makes it necessary to design a dungeon as a pointcrawl from the ground up. Grid map dungeons usually don’t have such clear cell structure and a great number of possible connections between thematically linked areas.

A system of self contained grid mapped cells connected by highways running through otherwise nondiscript cells is my solution to having huge dungeons that don’t involve long stretches of tedious searching through empty rooms or being implausibly cramme with treasures and monsters. Or at least my current attempt at a solution.

What about sub-mega dungeons?

When I started doing research about good dungeon design, I made an interesting observation. Pretty much everyone who is writing about the design of dungeons is talking specifically of the rather special case of megadungeons. I can understand the fascination with these massive places that can remain not fully explored even after multiple campaigns set inside of them. But when you start looking into the details of what people are writing about the design, you find a lot of recommendations that I don’t think carry over well into games of exploring multiple different dungeons over 1 to 4 sessions each.

The whole logistics are really quite different and the journey to and from the dungeon takes up a much bigger role in the campaign when the party constantly has to relocate their base camp. Like most advice that is around for worldbuilding, megadungeon advice mostly concerns itself with the really big pictures. But what seemingly gets overlooked is the small scale design of encounters and individual rooms. These are problems that a lot more GMs are having to deal with, yet it seems that nobody really has anything smart to say about this element. As a GM poorly experienced in running dungeon crawls, I’ve been doing a lot of searching for such information, but it appears to be curiously absent.

The Ancient Builders

Ruins play obviously a great part in the Ancient Lands and someone has to have build them in the first place. While I don’t want to have a generic post-apocalyptic world I decided to keep things simple and focused and have the majority of ruins be from an ancient time and build by fey races instead of being build by past humanoid civilizations. Ruined castles can have signs of past inhabitation by people, but people-build ruins are going to be almost all abandoned villages. Mortals just don’t have the ability to build castles like spirits can, and never did. It’s not that mortals lost knowledge and abilities they had in the past, but simply that spirits are just much more powerful in every way. Fortunately for mortals, most fey castles have long been abandoned and only few of them are currently inhabited. The fey races are not particularly attached to stability and leave lots of stuff behind when they are done with them, but these tend to remain around for a very long time after that and over thousands of years those abandoned castles add up.

Naga

Of all the old builders, the naga are the ones that still have a strong presence on the material world and inhabit and maintain many of their ancient castles, particularly in the southern lands. Around the Inner Sea and the northern lands these are all long abandoned, though. Naga ruins often consists of steep, tower-like ziggurats surrounded by smaller buildings with flat stone roofa that housed vast numbers of elven slaves. They can easily be identified by their lack of stairs, with levels being connected by ramps. They are also generally found near coasts.

Tower Builders

The tower builders were a civilization of shie that built a large number of citadels throughout the northern parts of the Ancient Lands. All their castles have a square base and smooth walls that slighty narrow inwards until reaching a flat top. These towers are build from huge stone blocks that are very accurately fitted together and each have a completely unique shape. Inside, the ceilings are held up by massive supports that start close to the walls and lean sharply inwards to form a kind of trapezoid arch and provide a lot of cover and hiding spaces.

Rock Carvers

Ruins of the Rock Carvers are mostly found in the area of the Inner Sea, particularly the Akai and Tavir Mountains, but also in various places in the steep rocky cliffs on the coast. Rock Carver ruins are often seemingly low and bulky constructions compared to the other styles of castles, but are mostly located underground, which hides their frequently massive proportions. Often the only parts visible above ground are a steep and extremely thick wall and an only slightly higher citadel that guard the main entrance to the rest of the stronghold. Often these have very few windows that are barely more than arrow slits, but castles in well defensible locations high up in the mountains sometimes have large numbers of great balconies and tarraces overlooking the landscape below. Smaller ruins are often completely invisible from the outside and appear only as easy to miss caves that reveal themselves to be spectacular underground castles. While the working of walls and tunnels is always extremely precise and without any irregularities, many ruins include large natural caverns that have remained entirely unworked and kept in their magnificent natural form.

The identity of the Rock Carvers is a complete mystery to everyone but they have often left behind many enchanted stones and gems.

Tree Weavers

The Tree Weavers were another group of shie who created numerous castles and palaces in the Ancient Lands. Their ruins are widely regarded as the most magnificent ones and appear to have been grown from gargantuan petrified trees and living rock rather than build by masons. Except for floors and stairs, which are covered in tiles of polished marble, there are no flat surfaces or straight angles to be found anywhere in these ruins, with all walls slightly bulging outwards and all ceilings being domed. Many ruins consist of tall towers that often have smaller side towers branching off from them in seemingly impossible angles like enormous stone trees. Other buildings have large marble roofs made from a single piece that look like the shells of colossals turtles.

Tree Weaver ruins are often home to a great amount of plant life, much of it being dangerous and even deadly to people getting too close.

Glass Makers

Ruins of the Glass Makers are the most mysterious ones and widely regarded to be by far the oldest. Many even believe that they were build by spirits that have long retreated into the eternal dark of the Underworld. Ruins of this type are made or a very hard dark material that greatly resembles glass an comes in various hues. Surfaces require a great deal of force to damage and tend to merely chip rather than fracturing. Rumor has it that at least in some ruins such damages slowly heal over the course of decades and centuries but they are so infrequently visited that nobody has ever really seen it happening so far.
The architecture of these steuctures, if it can even be called that, tends to be highly bizarre and often seems to serve no discernable purpose. However, they all share in common that they contain large underground sections of which generally much less information is available.

Glass ruins often have areas of high corruption and are sometimes haunted by trapped demons. Sorcerous relics can often be found in them and they are highly sought after by naga and sorcerers.

Three videos on level design from which adventure writers could learn something

I know, snappy title.

Bonus content: Three articles on dungeon design that I also found very eye opening.

Jaquaying the Dungeon

Four Things You’ve Never Heard of That Make Encounters Not Suck

Battlefields and BattleFEELS

If you’re interesting, bring some time. This might take a while to get through.