Tuesday, March 27, 2012

Population Density, Demographics, and the Dark Country

I have a really hard time placing villages and towns in my campaign settings, anyone whose read this blog with any sort of regularity will have probably picked up on that by now.  What it essentially boils down to is that whenever I look at a map I've made I have no idea how many villages feels "right."  My eyes just sort of glaze over and I don't get any settlements.  I have long searched for some sort of process that will spit out a number of villages that I could place on a map.

This problem has caused me to post a number of times about demographics.  Whenever I do so, someone inevitably suggests Medieval Demographics Made Easy or one of the many tools designed to make calculations based on that article.  It's very easy to see why.  First of all, if the same people who are concerned with demographics are usually concerned with a sort of bean counting realism that I find strangely attractive but could never wholly embrace.  More importantly, it does exactly what I want it to.  It spits out a number of villages that should be in a given area.

But there are a few problems.  The primary reason I had previously been uncomfortable with these tools is that they generate a massive amount of villages.  An area the size of a Wilderlands map -- like the Dark Country -- would have thousands of villages even on their lowest population density levels.  I want villages I can name and plop near a dark forest so they can be menaced by werewolves, not an endless blanket of them.

This used to be my only problem with the system.  I decided it was simply too "realistic" and that I was better off looking for solutions elsewhere.  However, in doing a bit of "research"* I found some scant information of the population of Tranysylvania in the 15th century.  Now that is a bit later than my assumed time period, which is somewhere around the late 14th century, but it seemed like a good place to start.

It was then that I discovered another problem with the Medieval Demographics Made Easy system.  When I plugged in the area of Transylvania (roughly 77,000 square miles) it told me that at the lowest population density it would have about 1.5 million people.  It turns out that the figures I found for Transylvania said it only contained about 500,000 individuals.  That's a margin of error of one million people.

As regular readers will remember, the Dark Country is loosely based on medieval Transylvania (mixed heavily with the Northern Crusades).  So it would seem that the Medieval Demographics system doesn't account for the realities for that part of the medieval world.**

What I'm currently thinking of doing is figuring out about how many people would live in the Dark Country based on the populations of Transylvania, Wallachia, and Moldavia.  Then I'll divide up the population among the cites and try to make enough farms for them to reasonably eat and raise armies and stuff.  I'll also need to leave some room for the lingering pagan tribes so that they can be a realistic military threat to the Seven Cities.***

If I've messed up somewhere in my calculations, please let me know.

*looking on wikipedia -- so grain of salt yadda yadda

**I will assume it represents the populations of England fairly well without actually checking.  Those more concerned with that area can do so if they're interested.

***Post on pagan/settler warfare coming shortly

6 comments:

  1. So it would seem that the Medieval Demographics system doesn't account for the realities for that part of the medieval world.

    I think you're spot on. IIRC, the original article noted the source calculations were based on England/France and that other European or Middle Eastern areas could have very different population figures.

    Checking the Medieval Demographics Online, a population density of '20' and an area of 25,000 square miles spits out a population of 500,000, which suggests that dividing the physical area by 3 is equivalent to reducing the population density by the same factor.

    So as a quick fix (meaning, until I can update the tool), you may want to consider a mental conversion such that 1 square mile of English countryside means 3 square miles of Transylvanian desolation. Perhaps as one moves west, the conversion improves to 1:2...

    That still gives you 989 villages, 6 towns, and 2 cities. Which is a lot. But perhaps more in tune with the area in question.

    ReplyDelete
  2. I took a quick look at the tool, thinking to simply add more options to the Population Density drop-down. But then Inspiration thwacked me upside the head:

    Thought #2: Maybe adding a drop-down for Region Type (Civilization | Frontier | Wilderness). The choice affects population density.

    Thought #3: Should Region Age be a factor in population density? Is it reasonable that a 1,000 year old region has a higher density than a 100 year old region?

    ReplyDelete
    Replies
    1. That still gives you 989 villages, 6 towns, and 2 cities. Which is a lot. But perhaps more in tune with the area in question.

      I might say that the existence of magic doubles or triples crop yields to get that village number down, but I might not.

      Using the demographics from the Ready Ref sheets means that the 280,000 persons living in the Dark Country occupy much much MUCH less land than the ones using your system. That's to be expected though.

      Maybe adding a drop-down for Region Type (Civilization | Frontier | Wilderness). The choice affects population density.

      That definitely seems like the best option from where I stand.

      Should Region Age be a factor in population density? Is it reasonable that a 1,000 year old region has a higher density than a 100 year old region?

      I don't think so. As the original article mentions, the cities of the Kievan Rus were some of the largest in Europe and their not nearly as old as London or Paris.

      Russia is another one of those areas that shows the flaws in the MDME system. It was highly urbanized and so the population density was really high, but they were concentrated around cities and there was a lot of forested hinterlands and wilderness.

      Delete
    2. Good catch on the Ready Ref sheets - I need to take a closer look at the JG method.

      I'm working on some sandboxy things, and it occurs to me that MDME, while sporadically accurate when creating an historic "snapshot," might be too beaucoup for more casual applications.

      I think MDME is supposed to account for density differences between urban and rural environs by accepting a region's total area, but figuring out the correct "average" is a guessing game--you might as well ignore it in favour of a desired population total.

      Delete
    3. Good catch on the Ready Ref sheets - I need to take a closer look at the JG method.

      I recommend checking out this article on Rob Conley's blog:

      http://batintheattic.blogspot.com/2009/03/wilderland-demographics.html

      Delete
    4. has anyone come up with a solution

      Delete