Frey Thorvaldsson

Europa Universalis Musings

Rambling by Frey Thorvaldsson

I’ve been playing Europa Universalis 3 lately.

It’s a grand strategy game where you can pick any country between the years 1453 and 1789 to rule. Thus the game takes us from the fall of Constantinople to the initial phase of the French Revolution, leaving Louis XVI still with his head - but not for long.

A scenario I’ve played through a few times is starting with the Ottoman Empire in 1453. I usually start with annexing what remains of the Byzantine empire, consolidating Anatolia under Ottoman rule and then do my best to bring Europe under the heel.

For history buffs like me it’s amazing.

poland in eu3As you might be able to tell, EU3 was released in 2009.

What makes this game great is that war is almost an afterthought. The bulk of your time is spent managing resources, holding down inflation, making alliances and keeping your various minorities in check.

interfaceYou are now head of the Austrian central bank, enjoy

You need to balance all this and more as you attempt to outcompete other states. For example, if you're playing as the Ottomans you should try to modernise quickly to avoid falling behind the west. However, as you push your reforms your country might become less stable. This means you'll get fewer taxes and there’s a higher risk of revolt. The key is to time things well and not overextend.

You have metrics and dashboards for nearly everything so you can see how your actions influence your metrics in real time.

You're essentially acting as a monarch would have, making trade offs, taking risks and hoping for the best.

What’s missing, however, is that European rulers were basically flying blind.

Odds are that in 1453, a European ruler couldn’t tell you with any degree of certainty how many people lived in their country [1], let alone how much manpower they had to draw on in case of war. In all likelihood, they would only have had reliable numbers on the state income and expenditure.

To put it another way, if Louis XVI knew that his revolt risk in Paris had risen to 30% then he might have considered adjusting his “autocratic” slider a bit and stationed a few more troops about around to defuse the situation.


[1] Even today countries like the UK and US settle for estimating the total amount of residents.

Having a single source of truth for residents has subtle benefits. In Iceland there’s a national register which keeps track of all Icelandic residents. Built on top of this register is an authentication service (like google and facebook's social login) with which you can access medical files, apply for benefits, do your taxes and sign in to banks. You can also do relatively sophisticated stuff like sign contracts and give someone power of attorney. It’s really cool.

If you’re interested in knowing how some of this data is structured I recommend looking at this API which combines a few different government databases. The docs are in English and a pleasure to read!

[2] Today a CEO will have infinitely more data to work with than a late medival king. Where we lack meaningful data we try anyway, metrics like "engagement" and "customer satisifaction" to try to represent feelings in aggregate. At worst these metrics are actively harmful by creating skewed incetives. Attach someones KPI to increasing "engagement" and they'll eventually design a skinner's box. IMO it's fine to let sleeping dogs lie and leave these metrics out of data pipelines.
← Back to articles

Let me spam you!

Say goodbye to inbox zero! E-stalk me via email! Follow along with 5+ other subscribers including my mom and my girlfriend as I struggle to keep up a consistent writing habit.