The Campaign Builder's Guild

So, I am sure many of you are familiar with Medieval Demographics Made Easy. It is a great article that gives some insight into the demographics of medieval Europe (largely based on information from France). Well, I emailed the writer and asked if I could do an addendum article on fantasy races. I'm not an anthropologist, though. I also don't care if it's scientifically accurate, as long as it's believable, unless some good research numbers pop up.

I'm looking to address how these demographics may be different for different types of countries, societies, and groups. If a race is carnivorous, how does their population density change? What about a vegan race? What about nomads, or hunter gatherers? How many elves can live in that forest? How many goblins in the hills? How many dwarves live in a citadel of X size?

The article says the human population density in medieval countries range between 30 to 120, based on arability and climate. But, that's looking at the entire kingdom; the population density of settled land is based on farmland, or 180 people per square mile of farm land; this means a maximum population country still has 1/3rd of it's land as "wilderness", which could then be populated by other races in a fantasy world.

Now, rather than dig through anthropology articles like I did my first time,[note=The first attempt]On my first attempt, I found a few numbers that could prove useful: Hunter gatherers average 1 person per 61 to 122 square miles (0.0164 to 0.0238 people per square mile)
the !Kung top off the hunter gatherers at 1 person per 4 square miles (0.25 people per square mile)

Populations in different terrains
Arctic: 0.0033 people/square mile
Subtropical savanna: 0.166 person/square mile
Grassland: 0.066 person/square mile

Nomadic Herders
Sahel averaged 3.09 people/square mile

Forests
Indonesia, 57.14 people/square mile
Sengah Temila, 13.9 people/square mile
Simpah Hulu, 2.7/square mile

Now, most of these numbers are drawn from modern times. I also don't know if they're measuring only the people's territories themselves, or the greater area included in their range.[/note] I want to tackle this with my strength: gaming, and see how the numbers look.

In the Civilizations game franchise, food production is determined by the terrain and improvements. The most productive terrain is usually "grasslands" and lakes at 2 food each. During the medieval era, Farms provide +2 food to tiles that have connection to fresh water, or +1 food to tiles that don't have fresh water (you need to be adjacent to a river or a lake to get fresh water).

Now, in this case, grasslands and plains are different terrain. Plains, deep ocean, tundra, and forest (usually ontop of grasslands) provide 1 food. Flood plains and oasis provide 3. Hills and deserts provide 0 food. Jungles and Marsh reduce the food in their tile by 1.

Here's some inferences I draw from this. Irrigated farmlands are providing 4 food, and that corresponds to 160 people per square mile. Luckily, that's a multiple of 4. Less advanced societies, or groups that don't have adequate fresh water, may have unirrigated farmlands at 3 food. 1 or 2 food per tile would then correspond to 40 or 80 people per square mile (which compares to one of those forest dwelling groups).

Perhaps these numbers could help with some fractional multipliers to alter the demographic numbers for different races.

--------------

Here's another angle of thought: meat takes more effort to produce than crops. For instance, it takes 15 pounds of feed to make 1 pound of beef, 6 pounds of feed to make 1 pound of pork, 5 pounds of feed for 1 pound of chicken, and 2 pounds of feed for 1 pound of catfish (these are Department of Agriculture of USA numbers). Pigs (and sheep, which I assume are fairly similar) are the chief food animal of the medieval times, as cattle are useful for labor and milk. Not knowing what the percentage of meat in the medieval diet of the masses was, I can't really put these numbers to use other than saying the population density for carnivorous or chiefly carnivorous races would probably be smaller (if we say meat is 20% of a human's diet, to just pick a number, than we could say a carnivorous race may have as much as a 1/4th lower population density). A vegetarian race's population density could be a bit larger, but not significantly (as i don't think medieval people (at least commoners) ate too much meat; that's probably a simple thing to google ...[note=I googled it ...]...and found little other than fad diet information and little numbers articles.[/note]

-----------

I'm opening this to the floor to discuss some of the factors that I should address in such an article expansion. I want to tackle the classic fantasy races (dwarf, elf, gnome, and halfling for the player side, and orcs and goblins for the monster side) as they are classically imagined (read as the way D&D depicts them). I think it would be really cool to also include some information on predator populations (like dragons), to give world builders an idea of the density of monsters in a "natural" world, and what would need to be tweaked to get the types of populations a world builder may want.

still coming out of work coma.
Yes, I used the heavily, though tweaked it quite a bit for some crops and livestock having been magically genetically atered.

(like, every commune or major farm has paid for this and maybe this or others)

I love the fact you are taking the time to check for it and give some level of baseline internal consistency.  Will try to look in more soon.

Quote from: XeviatHere's some inferences I draw from this. Irrigated farmlands are providing 4 food, and that corresponds to 160 people per square mile. Luckily, that's a multiple of 4. Less advanced societies, or groups that don't have adequate fresh water, may have unirrigated farmlands at 3 food. 1 or 2 food per tile would then correspond to 40 or 80 people per square mile (which compares to one of those forest dwelling groups).

As a "veteran" Civ player myself, I certainly understand the appeal of using their system in a general sense, though there are a number of ways in which it will lead you astray if you take it too seriously.  For one thing, human agriculture has evolved to many different geographies and circumstances; mountains and hills might be poor food sources in the franchise, but Incan terraced gardens managed to make those areas quite productive.  Likewise, the crop matters as well - a marshy area wholly unsuited to wheat might become much more productive with the introduction of rice.

The Civ model's other big problem, though, is that geography is not reducible down to "tiles."  There is a big problem in Africa right now in which rivers critical to local ecosystems are drying up - not because of water overuse, but because upstream people are clearing forests to make more farmland, and that forested area is critical to collecting rainwater that feeds the river.  Interactions between terrains don't exist in a model in which each mono-terrain tile is treated as wholly independent of and distinct from every other one.  A "mountain" in the middle of a desert is not the same as a "mountain" above a rainforest that catches the moist wind coming off the coast.

QuoteNot knowing what the percentage of meat in the medieval diet of the masses was, I can't really put these numbers to use other than saying the population density for carnivorous or chiefly carnivorous races would probably be smaller (if we say meat is 20% of a human's diet, to just pick a number, than we could say a carnivorous race may have as much as a 1/4th lower population density).

Well, watch out here.  Consider a German peasant.  He eats very little meat, but he does eat some, and this comes in the form of a pig.  The peasant takes the pig to walk around in the forest, eating acorns and tubers and things.  The pig gets fat over the course of the year, and when winter comes, the pig is slaughtered.  The peasant gets to give his family a little something special for their Christmas feast, and the meat helps tide them over during a lean time of year.

Notice, however, that no additional land was required for this pig.  No pasture was set aside for it - the only land it "used" was the wilderness, which was there anyway, and would have been there whether or not the peasant decided to raise a pig.  The same could be said for meat acquired by hunting or fishing.  Since one third of the land (at a minimum) is wilderness, products of that wilderness do not necessarily come at the expense of cropland.

In the middle ages, at least, the greatest devourer of farmland was not a demand for meat, but a demand for wool.  Lords discovered that wool offered much better profit than grain, and one could get away with paying a very small number of shepherds to cover a very large territory that once held far more peasants and their families.  (Coincidentally, sheep were also much less likely to rebel than peasants.)  A civilization with plenty of arable land need not necessarily use it all for food - non-edible products like cotton, tobacco, wool, opium, flax, and so on might have great value for trade and be crucial for local culture but contribute nothing to maintaining a larger population.  I question whether the Medieval Demographics guy really gave that part as much thought as it deserves, because unless the peasants are running around naked, presumably someone, somewhere must be producing the fabric they wear.

The point is that simply because a race is vegetarian doesn't mean they use all their land for agriculture, and just because one race eats more meat than another doesn't necessarily mean this translates into a loss of available cropland.

QuoteI'm opening this to the floor to discuss some of the factors that I should address in such an article expansion. I want to tackle the classic fantasy races (dwarf, elf, gnome, and halfling for the player side, and orcs and goblins for the monster side) as they are classically imagined (read as the way D&D depicts them). I think it would be really cool to also include some information on predator populations (like dragons), to give world builders an idea of the density of monsters in a "natural" world, and what would need to be tweaked to get the types of populations a world builder may want.

That certainly would be cool.  I think the most important step would be to establish what they eat and how much of it - if a halfling's calorie needs are only a fraction of a human's, they have a significant advantage in terms of food production, and any disadvantage from their inferior strength might be compensated for by draft animals and farming technology.

I'm impressed that you're tackling this.  That's a great resource, and I'd be interested to see just what you come up with.  You definitely do need to begin with determining how much each race really need to consume compared to a human.  Also, you'd want a baseline for food production (hunter-gatherer vs. pastroal or aggricultural society).  It looks like you have the lattermost.

As far as fantasy monster density goes, that'd be interesting, but I'm not sure it'll produce the numbers required to keep adventurers' bloodlust sated.  A large, apex predator requires a LOT of land.  That said, it'd be fascinating ot make a world where it all makes sense.

Poly, my thoughts of looking at Civ were merely to have a baseline to reference since the game "makes sense" to its players. Again, I don't need this article expansion to be 100% scientifically correct; I just want it to make sense. Notice I didn't even bring up food resources, which aren't done realistically in Civ (someone who finds a source of grain is going to spread that grain across the world). You're right about the limited meat production, or meat gained through hunting, but in that case I'm not sure we'd need to even account for that. Commenting on a specific thing, though, mountains in Civ V are impassible, so I'm assuming these are only steep peaks and that hills could represent other mountainous terrain (the Himalayan plateau is made with hills after all). Also, farms can be made on hills that are adjacent to rivers.

In the initial article, I'm under the assumption that fabric production is probably lumped in with the farm land, as his numbers are largely drawn from the demographics of medieval France.

So, I'm assuming we use the 3E racial stats to get an idea of their caloric needs; 3E's nice since so much of it is open source. We could be easy and just gauge them by size, but I want to try out being a smidge more particular. Here's weights:

Humans: 5 to 6 feet, 125 to 250 lbs
Dwarves: 4 to 4.5 feet, (almost as heavy as humans)
Elves: 4.5 to 5.5 feet, 95 to 135 lbs
Gnomes: 3 to 3.5 feet, 40 to 45 lbs
Orcs (based on half-orc): 6 to 7 feet, 180 to 250 lbs (odd that they aren't heavier)
Halflings: 3 feet, 30 to 35 lbs

Now, none of these races are described as carnivorous, nor are any of the races described as vegetarian. The main things we need to address is whether the race should be described as having a fast or slow metabolism, and whether the race has any thematic constraints to farmland.

For example, dwarves build down, making underground living spaces. Without magic, though, they'll need to use the surface for farms. Should they have artificial sunlight down below?

Elves live in the forest, and it wouldn't fit the themes if elves were cutting down trees to make farm lands. They could, though, replace fallen trees with fruit and nut bearing trees, creating vast orchards instead of things like grain farms. They can hunt the wild animals, but if their numbers grew too large they'd exhaust that food source. They could very well use magic to up this, but I now think I'd like to establish a "magic multiplier" to population from a square mile of "farmland", rather than getting too fiddly with it.

Gnomes live in wooded hills, so they face some of the same issues that dwarves and elves will. Halflings in 3E, though, don't have their own lands and just live with the other races, so the article could assume you choose a percentage of Halfling population and they end up being something like 4x the number of medium race they displace.

For Orcs, and goblins (who can benefit form the same thought), we can explore what non-farming communities would be like. How much less food could they produce? Can we assume raiding can account for anything more than subsistance? (you can't raid 100% of the time, after all, or you'll run out of people to raid).

As for monster densities, I'm more interested in figuring out what the world would have to look like to get monsters in the numbers like D&D. It would be cool, though, to know how many square miles it takes to support a great wyrm dragon (though D&D dragons can eat rocks).

Lets see how my thread necromancy works. At the moment, I am working on 3 projects: modding D&D (either 3rd or 4th, still uncertain), establishing elements of my setting which could impact its system (like magic), and filling out my setting itself. This project here could directly impact my setting, and I still think it would be an invaluable resource.

I dug up some numbers online; I was bad about gathering my sources.

Hunter gatherers are estimated to have had a population density of 1 person per 61 to 122 square miles (0.0164 to 0.0238 per square mile). There is a group called the !Kung (modern I believe) that have a 0.25 per square mile density.

A book called "A concise history of world population; by Massimo Livi Bacci" gave these numbers, but I forget what era they were drawn from:

Arctic: 0.0033/square mile
Subtropical savanna: 0.166/square mile
Grassland: 0.066/square mile
Semidesert: 0.014/square mile

Another book said:

Nomadic Herders (the Sahel) came out to 3.09/square mile.

Another book studied the forest dwelling people in Indonesia. The populations ranged from 14 to 2.7 per square mile (and my notes show 57/square mile for something from that book).

--------

I'm going to be fussing with "Civilization V" again, to simply gain some rough ideas. I'll be making some assumptions like "for elves, forests supply +1 food", just to see how numbers look. I'd still like to hear more thoughts from the smart folks here.

Using a Basal Metabolic Rate calculator, assuming an 18 year old male, here are the calorie needs I got for the races:

Humans: 5 to 6 feet, 125 to 250 lbs (5 ft, 6 in; 187.5 lb)
   1950 calories
Dwarves: 4 to 4.5 feet, (almost as heavy as humans) (4 ft, 3 in; 187.5 lb)
   1760 calories
Elves: 4.5 to 5.5 feet, 95 to 135 lbs (5 ft, 115 lb)
   1422 calories
Gnomes: 3 to 3.5 feet, 40 to 45 lbs (3 ft, 3 in; 42.5 lb)
   704 calories
Orcs (based on half-orc): 6 to 7 feet, 180 to 250 lbs (odd that they aren't heavier) (6 ft, 6 in; 215 lb)
   2274 calories
Halflings: 3 feet, 30 to 35 lbs (3 ft, 32.5 lb)
   603 calories

Now, these are assuming these races abide by a fairly human metabolism. Note, that while the halfling is almost 1/6th the mass of the human, his BMR is only 1/3rd the human's. I don't know if the calculator was designed to handle such variances, but assuming this was true, I could chalk it up to smaller creatures have higher metabolisms; their hearts beat faster and all that.

This could make mixed race communities interesting/difficult to determine.

How believable are these numbers?

I don't really have the expertise or the math skills to comment on specific examples and how believable they are, but if you are interested in the relationship between mass and metabolic rate, you might read about Kleiber's Law, which holds that an animal's metabolic rate scales to the 3/4 power of an animal's mass.  Simply put, as a creature gets larger, its relative surface area decreases, making it harder for it to dissipate heat.  It must compensate by being more efficient with energy and maintaining a proportionally lower metabolism.  If an elephant's metabolism to body mass ratio was the same as ours, the elephant would burn up.  I don't know if it would literally burst into flames but it would certainly have one hell of a fever.  This goes towards explaining why your halfling would have a higher metabolic rate to body mass ratio than a human; he is small with a lot of surface area, easily dissipates heat, and does not need to use energy as efficiently to avoid overheating.

Thanks Poly, I knew I had read something like that at some time.

I need some thoughts on this: How do you think non-human aging rates will affect population density? Will it only effect how fast populations regrow after calamity? I would think that populations will only grow as large as they can be sustained, no? Am I wrong there?

QuoteI would think that populations will only grow as large as they can be sustained, no?

Not necessarily. While in the short-term a population may be sustainable while its numbers are of working age, a race like the elves (traditionally associated with extreme longevity and a low birthrate) may end up much like the modern day Japanese, who have a huge number of non-working elderly folk to support (something around 24% as of today and still increasing) and not enough children being born to replace the ageing workforce.

In other words, what may be sustainable in the short-term while the working population is strong and made up largely of younger workers will rapidly grow unsustainable when burdened (as elves and also dwarves would be) with a very slowly ageing population who just won't die, possibly resulting in massive overpopulation problems and a huge need for human resources to put towards age-care that due to birthrate just aren't there.

Mixed age communities are probably the answer (carefully designed and implemented ones, that is, not just the natural mixing of generations that would occur anyway) ensuring that for every group of elderly elves who are only so capable of physical aspects of food production (harvesting is probably not a problem though, in the most part - your elderly elves would make a good workforce there), there are enough younger elves living with them to deal with any of the aspects of ('food forest' style orchards, you said above) food production requiring strenuous manual labour and/or dangerous climbing.

On the note of food production, I think it far more likely that dwarves would farm fungi, underground fish and various species of giant insect/arthropod than construct surface farms (especially given that the surface-level property of any race of surface dwellers will be of extremely limited defensibility and an attractive prospect for thieves/colonisation).

QuoteI would think that populations will only grow as large as they can be sustained, no?

Populations constantly outstrip their environment's ability to sustain their numbers, most especially once something deleterious happens to the environment. Famine is an ever present companion of civilization.

Quote from: Xeviat
I need some thoughts on this: How do you think non-human aging rates will affect population density? Will it only effect how fast populations regrow after calamity? I would think that populations will only grow as large as they can be sustained, no? Am I wrong there?

Here are a few formulae that may or may not help.  From what I see in them, it's not about how long a race lives so much as how fast they reproduce.

Population growth at an instant:

r = b - d + m

r = growth rate
b = birth rate
d = death rate
m = net migration

Life expectancy is figurd into the death rate, while fertility gets figured into the birth rate.  So let's take a look at each of these.

The NRR (Net Reproductive Rate) is the number of female children an average female each newborn girl can expect to have.  Ignoring migration, an NRR of 1 means the population is replacing itself with 0 growth.

NRR = s x c x f

s = the fraction of newborn girls who survive to adulthood
c = the average number of children an adult woman can expect to have (including women who have none)
f = the fraction of newborn children who are female

We can estimate the grown rate over a generation from this:

r = m + NNR^(1/g) - 1

g = th elength of a generation (currently 28 years in the US)

Now on to the death rate.  Old age is only one factor that plays into the death rate (which lacks a spiffy formula like those above).  The average life expectancy (LE) depends on infant mortality, accidental deaths, old age, violent deaths, and disease.  One thing that factors into all of these in some way is the amount of resources available to a population.  Unfortunately, I have no formulae to share regarding that facet of demographics.  I can offer up this one, for what it's worth:

LE = 1/b = 1/d, if r = 0 and m = 0.

This would let you sort out the average age of death, if the population growth is 0, net migration is minimal.  I hope this all helps somehow.

Don't have much to add, other than this would make a sweet ezine article.

I am thinking of these world factors without outside interference right now. What I mean by a nation growing as far as the resources will take it to be that once it is gone over that point, people will either expand or die of starvation. On the grand scale, it seems like it will hover at some point near that resource cap. If some outside influence, like an abnormal weather pattern, effects the resources, things are going to end up having to correct themselves, no?

As for growth rates, we do have life expectancy for the 3E D&D races, sans anything that would infer death from anything other than old age. Perhaps we could average some "number of encounters" a commoner deals with (one a season?) and the odds of them surviving those encounters. Or we could just pull up stats on life expectancy in the medieval times and then adjust them for the different races (if halflings are psudo-nomads, they'd be more prone to leaving an area than staying and starving, even if that means they starve on the road).

Thanks for the support guildies. I'm still not sure exactly what angle to tackle from first.

I think i'd sort out life expectancies as a percentage of human expectencies.  Frex, a dwarf might live 150% as long as a human.  And apply those percentages to human life expectancies during whatever historical period you want this to represent.  No one is sure just how long a human can live before it just falls appart from wear and tear.  So far, we've pushed old age pretty far back, but people do keep living longer and longer.  Saying an elf dies at age 900 isn't as meaningful as saying "under these conditions, the average life expectancy of an elf is 900 years, counting accidents, nutrition, disease, violence, etc."  That's why I suggest using the percentage approach.  We have stats for that.