The Great Antarctic *crawl pt. 10 - Clearing The Steppe

25 Aug, 2024

Though I've only posted on it in a fragmentary fashion, I have been working on keying my Antarctic pointcrawl/hexcrawl(??) over the last few months. But one area of dissatisfaction that has arisen is the density of POIs in the interior steppe (see here for more on how biomes were assigned).

This can be traced back to very early phases of this project where I had little real idea of where I was going with it and envisioned chopping the whole map up into mini-hexcrawls. This resulted in weird contortions like having one home-base city per prefecture, meaning cities were distributed semi-uniformly across the landmass. This, in turn, influenced the geographic weighting I used to randomly place other POIs.

The thing is, I can't really see an upside to having the steppe be this populated. It's lots of work to create interesting content for regions that are geographically flat and boring. It's unrealistic to place so many settlements in land that isn't arable without modern farming techniques, and results in a homogenous macro-scale geography, far less interesting than the realistic option.

An episode of Weird Studies recently got me thinking about Deleuze & Guattari's distinction between smooth and striated spaces, and I want the Antiborean steppe to be a smooth space of nomadism that plays by different rules than the striated coastal States. Still not sure exactly how this will work, but it's a worthy aim, I think.

So I'm working on re-running the POI placement code, using the cursed new topography and with stricter rules about keeping settlements on arable land, or by rivers or coasts. Since some rivers extend quite far inland, this doesn't mean that settlements will be exclusively coastal, but it should lend an intelligible logic to their placement and should make the steppe an obstacle that shapes how players explore rather than a vague zone where the POIs get farther apart.

Arable Land

To get a rough, unscientific idea of what areas could support settlements, I ran the following analysis. First, I mapped a base score onto the map of biomes:

'Arable' should be interpreted loosely as 'ability to support settled (i.e. non-nomadic) populations in a fantasy milieu'.

Then:

• Anywhere within 8km of the ocean gets +4 to its score, as communities can support themselves by fishing.
• Anywhere near a river gets +6 points. Rivers provide freshwater, food, and sewage treatment for free.
• Anywhere within 5km of a freshwater lake gets +3
• Any site higher than 2,500 meters of elevation gets a score of 0 (based on cursory reading about amounts of arable land at different elevations), and all scores are scaled linearly by where they fall on an elevation range of 0 to 2500 meters. This step is slightly redundant as it tends to prioritize coasts and river basins further, but I find it adds some nuance to the scores further inland.
• All scores were smoothed by a mean filter to soften sharp transitions

(Darker green = more habitable)

To place 5,555 settlement POIs, I raised all values to the third power and did a random weighted selection. And to place 8,888 'landmarks' I used the map raised to the second power, as landmarks can and should be clustered near settlements but generally more dispersed.

To place 4,444 lairs, I did a similar analysis but changed the biome scores so that forests scored higher than open terrain types, and privileged high elevations significantly to make mountains and hills more dangerous.

Finally, I placed 1001 ruins with full randomness.

Here's the result, zoomed in to the Knorthern half of the continent. There's still stuff in the steppe, but it's much less peopled. And lairs are now much more concentrated in the mountains and in the Xel-shogoth waste in the Psoutheste (this one's ludicrous, but I kind of like having a horrible Lovecraftian Mordor on the map). This does leave some arguably boring populated areas with few lairs, but I'd argue that that's fine -- I'd seed those areas with political intrigues if players ever wanted to explore them.