Refining the Trade System

I'm hard at work with classes as well as implementing Alexis' trade system right now and I wanted to document some of the ways I've needed to change the process in order to create the outcome I want.

The trade system depends upon references of goods bound to specific commercial locations (markets) where goods of that type are either produced or gathered from nearby, noncommercial areas (outlying farms, smaller towns, shipped downriver, etc.).  Following Alexis' suggestions, I plotted out regions of influence for each city, ranging in size from 4ish hexes for the smaller cities to 15 or 20 for the larger ones.  I initially used Welshpiper's medieval settlements calculator to determine how many cities I ought to have in regions of a given size and population density, which lead to more cities than I had reasonable room to place.  I did what I could - and I think the map has benefited from the density of locations in both the Southern Kingdoms and Confederacy (I scaled back the city counts intentionally for Arein because of its relatively young age (60 years)).  Obviously, the more closely packed the cities are, the fewer hexes each controls within a given area.

Each hex controlled by a market might produce resources, based upon its hydrology, elevation, and level of civilization.  While I haven't implemented Alexis' idea of infrastructure numbers yet, it is a future project.  As a rule of thumb, I decided that any hex more than two hexes away from the closest city was undeveloped and thus (for my two heavily-treed areas) still jungle/forest.  Hexes only 40 miles (2 hexes) from the closest city were also undeveloped, but if there was another city within 2 hexes from the are in question, the hex is settled instead.

It may be helpful to take a gander at the map I uploaded recently.

One of the early difficulties I encountered with the random determination is that Alexis' regions are based upon the kinds of terrain found in, say, Hexographer, rather than rooted in any kind of elevation or topological scheme, which is how I built my world.  While I absolutely understand the desire to be accessible, I had a little bit of work to do to adjust it.  My first approach was to classify any hex adjacent to another hex with an elevation difference of 400 or more counted as a hill (which graduated to mountain as soon as the hex's height breached 2000' above sea level).  This gave me no hills because the elevation change was smoother than I had anticipated.  My second approach banded hexes by color: hexes over 500' but under 2000' were hills, hexes over 2000' were mountains, and hexes between sea level and 500' were flat, either plains or scrublands (hexes between -50' and 0', found only at the coast, became marshes).  Alexis' plains designation really applies to land used for large-scale farming, whereas scrubland does not.  Thus, hexes 1 hex or closer to a city are plains, while those that I've ruled undeveloped are either scrub or forest/jungle.  Now, deciding between the two choices was tricky.  They are entirely different vegetation systems, and it was important that I get the resources from both.  I still don't have a satisfactory answer, but the answer I used had to do with distance from the forest's center, something for which I had a rough location.

The biggest change I needed to make was further differentiating the very generic labels of building stone, cereals, fruits (which replace grapes, since I'm in a tropical environment), gems, and vegetables.

My world's central distinguishing characteristic is the differences in both outlook and material culture between each of my societies.  Therefore, I need to be able to distinguish between the kinds of resources available to each people, as well as what specific foods is grown to determine diet and food culture.  I have a sedimentary ridge that separates the Confederacy and Southern Kingdoms, and so I used that to distinguish between the kinds of stone found: south of the ridge is sandstone, compressed sediment from the sea, while north of the ridge is limestone because of the karst topology (a choice I made to provide sinkholes to the area).  Across the Sea of Shadows to the south, I have granite mountains.

However, karst topology includes important things like dolomite and gypsum, which I wanted to produce as well, and this led me to the solution I employed for my other reference specifications.  I assigned probabilities to each type and rolled a die: 1 dolomite, 2 gypsum, 3-4 limestone.  The important thing is that I get different numbers for each resource, scattered throughout the regions.

I then specified different cereal categories, millet, rice, and wheat, fruit categories, bananas, coconuts, mangos, and pineapples, gem categories, ornamental, fancy, semiprecious, lesser precious, and greater precious, and vegetables, cassava, cotton, edibles (tubers and the like), jute, palm oil, and soybeans.  I determined these the same way, giving each a probability and then rolling for each individual reference - I tried applying a strict percentage on the overall reference to give me numbers, but the results were too clean.

The last way I've changed the references is adding a new one: warcraft, which governs the making of arms and armor for a given culture.  The Tarluskani use a khanda while the Southerners use a longsword.  While they are similar in many ways, the techniques to use it are different, as are the concerns and problems when forging it.  This also lets me price weapons and armor from different cultures like the exotic or commonplace items they are, even though the skills involved are the same.  I get my Warcraft value for each market from the sum of the leathercraft, metalcraft, and woodcraft references, since they are highly related industries.

My last change to the overall system was in using the reference numbers to populate information about the cities themselves.  It follows that cities with higher references are more economically powerful than those with a lower count -  they have more resources and can bring more to bear to any economic dispute.  So, since the urban population of the late Middle Ages was something like 25% of the total population, I have calculated the total population for each area based on population density (usually between 10-20 people per square mile), multiplied it by .25 and the multiplied it by the ratio of the market's reference total compared with the region's reference total to get the final population.

For example, Reyjadin is the capital city of the Southern Kingdoms.  Being a capital city, I've doubled its population because of that added importance.  The total population for the Southern Kingdoms is approximately 683,000 people.  Reyjadin controls 6 references while the Southern Kingdoms as a whole controls 291 references.  So, Reyjadin's popullation is 683,000*.25*2*(6/291)=7,041 people.  Population then defines a city's physical size (population/38850 square miles) and the strength of its guard (population/150 employed guards).  While obviously these latter two measures are adjustable, it gives me a systematic way to approach any city.

While there's a lot to refine about this process (like using the total local references rather than references controlled by the market), it's definitely progress.

Also, a helpful resource to intrepid worldbuilders.