Trade, Economy, and Supply

Trade and economy are difficult things to model in a game accurately. Especially so if you want to model trade across a long historical timespan. The reason for this is because before the 19th century and the rise of nation-states, trade was more of a regional affair in goods passed through multiple middle-men until it reached the end consumer. As nations became more powerful they were able to control how goods were imported and exported from their borders. The silk road passing across Asia from China to Europe is an excellent example of how trade diffused across continents.

The easiest implementation of trade would akin to how EU4 models trade: A graph of hard-coded trade regions with Europe being the end-node of money flowing into the coffers of the burghers. By using naval and controlling strategic regions you can direct the flow of wealth into your own home province.

However hard-coded trade node directions does not work for procedural maps, because it’s not obvious where the end-nodes should be. In an alternate history world, it cannot be obvious where the flow of trade will occur.

National markets based on the borders the player controls is more dynamic but has other issues. The biggest issue is that the concept nations did not even exist in early history, and so markets were more limited to small regions. National markets also do not model how small city-states like Venice and Lubeck in our own history were able to become trading powers even though they did not have a large country.

In the early development of my game, instead of national markets the economic simulation had regional nodes like in EU4. The total demand and supply of a given good would be aggregated in a region, and the price of that good would then be constant throughout the trade node. This allowed for a computationally simple way of calculating prices. However this leads to weird price gradients where two provinces bordering each other but belonging to different trade regions would have very different price for a good. It is both un-intuitive and can lead to weird trading behavior between cities that border each other.

Left: original deprecated colored trade nodes. Center: the deprecated price salt in each trade node. Right: The new price salt calculated using a gradient.

I’m confident I’ve settled on an economic model. Now instead of regional markets the price for goods is instead calculated locally but then diffused across provinces. Now there is less likelihood for sudden price jumps across provinces, and the player can intuitively visualize where there is high supply and high demand without being distracted by the arbitrary market regions.

Provinces can only produce one trade good. Limiting to one allows the player to more easily visualize where goods are produced in the map. Provinces can also produce a secondary strategic resource. This forces countries to capture strategically important provinces.

Another decision that needed to be made was how trade should be connected. Do kingdoms / countries import export directly with each other? This doesn’t seem appropriate if your kingdom is split up into multiple different areas. Why would anyone want to trade with venice if it was only a single province?

The way to properly model trade then is to use the concept of Trade Cities. Trade Cities are provinces with high trade power based on proximity to ocean, river, as well as having a high urban population. Based on their relative power, Trade Cities will control an area around them in which they have access to the produced goods of surrounding provinces. Those goods can then be imported from the Trade City. The trade power of Trade Cities also extends the range of their own import routes. Ultimately Trade Cities control an economic region around them, even if it’s crossing borders, since pre-modern kingdoms did not have significant border control when it came to economic activity. This allows for modeling entities like the Holy Roman Empire in which free cities had economic power but not necessarily political power of the provinces around them.

When a player wants to create an import route, they will see all neighboring Trade Cities that are available to connect to, along with the potential monetary value the Trade City has. The greater the power of the Trade City, the more goods it has the potential to access, the greater tempting target it is for import routes.

Left: visualizing provinces’ production value. Right: highlighting trade cities with > 0 trade power.

Controlling Trade Cities or having a trade embassy in a foreign Trade City allows for the player to extend their trade routes. Farther away Trade Cities are usually more profitable because they have goods that are usually not found in the player’s own continent or climate. If the player has a trade embassy in a city, they will need to pay a portion of their profit to the host country that actually controls the city. This means even small trade republics with good relations in large empires can have long trade routes even if they don’t control any of the provinces. However sea-going countries that can build their own trade ports can eventually bypass competing empires and have their own more profitable trade route.

Note: Each country is limited by the number of trade routes they can create, but trade routes that connect multiple cities in a row will still only count as only one 1 route, which promotes having long single-branching routes instead of fanning out from a central location. These sorts of long routes are more consistent with the sort of routes we see in our own historical timeline.

Owning or having access to trade cities allows for the trade route to extend much farther. This route connects 4 cities but still only counts as 1 route. The route can be automatically generated too based on max profitability.

Ultimate these many routes lead to the same sort of behavior in the Europa Universalis games where there is a flow of goods imported into the regions of high industry. However this flow is dynamic, it follows the natural chokepoints and islands found in the procedural map. Below we can see the highlighted trade traffic across the glob, and see how trade forms natural lanes through the internal seas, rivers, and coasts.

Highlighting land and oceans that have high levels of trade traffic.

With trade mostly settled in the grand scheme of things, the shape of the economy is starting to take form. The economy model is meant to be much simpler than in the Victoria game series, since we aren’t trying to model industrialization. Instead we want to force the player to think about how to connect to far away regions and increase production. We want to avoid micro-managing the internal economy or buildings. Simplifying the internal economy is done a number of ways.

1. Manufactured goods / industry is all grouped up together as “manufactured” trade good. So long as raw goods are flowing to high population regions with good education then industry will increase
2. There are no complex goods dependencies. If a certain trade good is no longer imported then that might affect a single manufacturing building… but it won’t cause a cascade of failures.
3. The demand of goods is mainly tied to population, wealth, and industry of a region. Building armies, infrastructure, or buildings may get bonuses from imported goods but building stuff isn’t limited by any sort of goods availability. I.e. a wood shortage will not stop ships from being built.
4. Prices should change very slowly over time and should be stable, which means not having to worry about sudden price shifts breaking trade. The pre-modern economy was slow to change and so too should be the game’s economy.

Addendum

Supply has also been implemented. There are several rules I’ve come up with for it:

1. The player’s supply is distributed from the capital in the lowest cost path. Rivers and ports can improve the distribution of the supply.
2. Armies and units in provinces with low supply will quickly lose organization and morale and have high attrition.
3. Armies in provinces with low supply can still be resupplied if they have supply wagon units in their army, but those supply wagon units are expensive and will quickly be consumed.
4. Non-friendly provinces will not have any supply and must be sieged down. Enemy forts will block any neighboring supply routes as well. This means several things: Armies are not restricted in movement around enemy forts, but supply propagation is limited. Therefore marching past enemy forts is possible but supply will not get past, meaning high rates of attrition. Sieging-down other enemy provinces won’t help with supply so long as the border forts blocking supply remain, which means effectively armies must still siege down enemy forts one-by-one into enemy territory.
5. Even if many enemy provinces are taken, the supply situation may still be poor if you are far away from the home capital.
6. Supply limited to capital connection means that large empires can benefit from releasing distant vassal states, since vassal states can better supply distant armies and will have more efficient administration.

The potential supply visualizes the player’s potential supply if they were to control all of the provinces around them. To improve the supply further than visualized here, the player must improve infrastructure routes.

On a final note I’m happy to say nation/province text placement is working decently well with the procedural map. We’re able to fit text in a way that looks visually appealing even with unpredictable province shapes.

Left: Kingdom/nation text works relatively well. Right: The map even further zoomed out

Zooming also works well now with diferent features being visible at different heights.