Collections: Teaching Paradox, Victoria II, Part I: Mechanics and Gears

This is the first post in a three-part series that will be examining the historical assumption of Paradox Interactive’s grand strategy computer game set in the 19th and early 20th century, Victoria II. Readers will find a number of references here to our previous discussion of one of Paradox’s other games, Europa Universalis IV, but I think this discussion will be mostly readable without having to rush back and read (or reread) the previous posts; that said much of what we discuss here will, I suspect, be more interesting if one has read the previous entries. This week, we’re going to be looking in particular at how Victoria II treats what is arguably its central game system: economics and the industrial revolution.

As before, the reason I chose Paradox’s series of titles is not to subject them to withering critique, but because I think they are a particularly rich area to discuss precisely because they are somewhat more historically aware than many similar games. Paradox games, by and large, arrive at their subject matter with a theory of the history behind them and that theory is then expressed in interesting ways through the game mechanics. Moreover, the very presentation of Paradox’s games as historical simulations as much as games both encourages players to think about them as exercises in history and also lends their theory of history tremendous persuasive power.

Finally, I should note that, when I began this series with EU4, I noted that we would do VickyII next as, “I want to bully Paradox into green-lighting Victoria III.” Paradox almost immediately announced Victoria III so I want to note right away that I gracious accept Paradox’s surrender on this point and will happily take credit for ending the long drought of Victoria content. I humbly accept all of your accolades for my heroic service. But I think that the announcement of the next game in the series makes this look at VickyII more valuable, to get a sense of the ways that it succeeded and failed as a historical exercise, in the hope that the best parts of it will be preserved into the next version.

But seriously, if you want to give me accolades, please share my writing. If you really want to throw me laurels, you can support me on Patreon. And if you want updates whenever a new post appears, you can click below for email updates or follow me on twitter (@BretDevereaux) for updates as to new posts as well as my occasional ancient history, foreign policy or military history musings.

One of Victoria II’s loading screens, showing its namesake being crowned in 1837.

Introducing Victoria

Before we dive in, we should set some of the basics of the game. Victoria II (the sequel to Victoria: Revolutions) is a grand strategy game made by Paradox Interactive. In it, players take control of a single state and guide that state’s policy decisions, including military and diplomatic strategy, economics and politics from 1836 (the game’s start date) to 1936. One of the notable quirks of Paradox games is that rather than having just a few ‘playable’ countries, effort has gone in to putting nearly every historical state in the period on the map and they are all playable, so while the designers clearly expect many players to stick to the most common states (the United States, France, Prussia, Austria, and of course Britain, inter alia), if a player wants to play as less economically or militarily powerful developed powers, they can (note that we’ll mostly deal with those countries in the third post in this series as they have quite a different experience in the game).

Victoria II’s primary map screen at game start, playing as Great Britain. There’s a LOT of data in the top left, but of particular note is the small 43.3% next to the book; that’s Britain’s literacy rate at game start. It’s quite high and will help propel Britain to a dominant technological and economic position for the early game.
Note also, rightof the small British flag in the frame, the star, factory and crossed swords. These are the prestige, economic and military strength of the country; next to them are the rank. Britain at game start has the most prestige, the strongest economy and the third strongest military. That makes it the top Great Power, noted with the small ‘1’ to the bottom right of the flag.

The game is primarily played on a map of the world, though Victoria II, much more than more modern Paradox games, has much of its gameplay hidden in menus and detail screens. The world map is divided into provinces; each province is controlled by one country in its entirety. Provinces are then grouped into ‘states’ and certain things (like factories) are managed at the ‘state’ level. That distinction won’t actually matter much for us here, so I am going to talk almost entirely about provinces, in part because I will be using ‘state’ as a technical term to prefer to independent polities (or ‘countries’) and I don’t want to be confusing.

While Victoria II does not have an explicit win-state where the game ends, it does rank all countries by three metrics: economic and military power, and prestige. The top eight countries by these metrics averaged together are marked as the ‘Great Powers’ and the game clearly intends it as an informal goal for players to seek to become and remain one of these great powers by the end of the game. That said, players absolutely can choose to ignore this and focus instead on simply surviving as smaller, weaker states or focus on improving the quality of life for their people (represented as ‘pops’) and eschewing the ‘score’ entirely.

Generational Paradox

Before we go further into the particular mechanics of Victoria II and what they say about its vision of history, I want to return to a distinction I made briefly in our discussion of Europa Universalis IV between trends in different periods of Paradox’s development history, because VickyII is very much an older Paradox game and that shapes how we should understand some of its mechanics. It’s possible to divide the product of Paradox Development Studio, the core studio at Paradox Interactive, into something like three ‘generations.’ The first generation (Gen1) Paradox games are basically all of the games that came before Paradox rolled out the Clausewitz Engine (the core software engine they use in all of their subsequent games) in 2007: Europa Universalis I and II, Hearts of Iron I and II, the original Victoria and the original Crusader Kings. While the introduction of that engine (which brings with it a very distinctive Paradox ‘look’ and ‘feel’) is an easy break point, the next is more subjective. Generation II and Generation III (Gen2 and Gen3) differ from each other in important elements of design philosophy; Gen2 Paradox games (EU3, HoI3, and Victoria II) tend to designed so that major historical events are forced to happen more or less the way they did historically, often by hard-coding certain events to only happen to the countries that did them historically. There is an inflexibility in this design which is quite noticeable as you press the game’s systems. By contrast, the Gen3 Paradox games (starting in 2012, CKII, EU4, HoI4, Stellaris, and Imperator) are structured to allow a lot more flexibility. The games feature fewer forced events and instead more systems which guide the development of play towards historically plausible(-ish) outcomes.

It’s not yet clear to me if Crusader Kings III should be understood as continuing in that Gen3 tradition or if it will mark the start of a Gen4 of Paradox games; we probably won’t really know until we get at least Victoria III and probably the nearly inevitable EU5. But for what we’re doing now, what is important to note is that Victoria II is very much on the older side of the Gen2/Gen3 divide. This is a second generation Paradox title, unlike nearly all of the other current third generation games.

I didn’t have time to start a new Victoria II playthrough for this series, but I have a few old saved games I can use as examples. We’re going to look at this one, playing as Austria. This is Austria’s starting position. On the one hand, we look pretty strong – 7th best army, fourth in prestige. But in practice, our neighbors, Prussia, France, and Russia are all much stronger.
The larger problem is that literacy number, which you’ll note in just 16.1%, much lower than France, Prussia or Britain. Consequently, Austria is going to unavoidably fall behind technologically until we can get our literacy up. This means our position – already vulnerable – will begin to slip fairly soon and we’ll be playing ‘catch up’ for much of the game.

Let me make all of that design philosophy concrete with an example. Let’s say in your game you want to form a big cultural union, a large state which encompasses a bunch of related cultural groups and is marked in the game mechanics as their national ‘home.’ In EU4, a Gen3 game, you have multiple routes to this; many cultural unions have specific ‘formables’ where if you meet the requirements, you can form that new state on your current territory. There are a lot of these, from the historical to the fanciful. But if you don’t have one of these formables of your cultural union, worry not because there is a game system for you – if you become large and prestigious enough your state can advance to the ’empire’ tier, which automatically makes your state the ‘cultural union’ for your larger culture group. In short, there is a general system for cultural unions which any state can utilize.

By contrast, forming new states in Victoria II is restricted to ‘decisions’ – events with very specific triggers. Even then there are just 14 of these (I am counting all of the steps towards forming Germany as one); most of these decisions are in turn restricted to just a handful of states. Many countries have no cultural union they can form because they lack a relevant decision; there is for instance, I think, no formable cultural union anywhere in Africa in Victoria II. Want to create a post-colonial pan-African national state? Well, you can’t.

Migration in Victoria II is another example of a very Gen2 system. When a ‘pop’ (a group of population; we’ll come back to this) decides to migrate to a new country, there is a complex set of factors which determine where they will go; the presence of other pops in their culture, the power of the target country, its level of political reforms all factor in. Except not really because all of those factors collectively max out at around +170% whereas the last factor, ‘is a country in North America, South America or Oceania’ is worth +300% and utterly swamps the others. In practice, every other factor merely determines which country in the Americas (or Oceania) they go to; the mechanics are hard-coded to make it functionally impossible for an an Afroeurasian country to make itself a meaningful target of in-migration no matter what it does. You could build Wakanda in Victoria II and still no one would want to move there because the game is designed to produce the mass migration to the United States in particular (which gets a number of other bonuses to this effect) and other American countries more generally and there is effectively no way to alter that process.

Consequently, the older ‘Gen2’ Paradox games have a very ‘railroaded’ feel to them: player agency is heavily constrained to the channels that the developers have planned for and provided events and mechanics for. I bring this up here because coming from the Gen3 Paradox games, I think it is easy to mistake the Gen2 railroading for very crass Eurocentrism when I think it is instead borne out of a design philosophy that held producing historical outcomes above providing players a high degree of flexibility. Since the game is set in an era where the historical outcome was the greatest degree of European colonial dominance ever, that focus on producing historical outcomes produces that historical outcome. I will be fascinated to see how Victoria III, built (I assume) with a Gen3 sensibility, provides options for ahistorical outcomes which are less Europe-dominated. Judging from the trend towards a more meaningfully global historical perspective we discussed in EU4 (and also quite visible in Crusader Kings; just contrast the on-release map of CK2 and CK3), my suspicion is that Victoria III will tend to push harder against a deterministic viewpoint.

That said, Victoria II has several sets of systemic interactions which I think work extremely well. While I am sure you are all chomping at the bit to hear about war, peace and diplomacy, I think we actually need to start with a system that is even more fundamental to Victoria II: industrialization.

Pop-ulating the Countryside

One thing I want to be clear about from the beginning with Victoria II: this is a game about pops. It doesn’t entirely shake the Paradox focus on states – pops end up tied to one state or another, after all, and the player still plays as a state, but the game is fundamentally about ‘pops.’ ‘Pops’ in the game are discrete units of population simulated as a group, typically numbering a few thousand, reflecting an amalgam of similar households. A single ‘pop’ is simulated as a homogeneous block of people in a single province who have a religion, a culture, an occupation and so on.

From our Austria game, here are some pops in Vienna. At the lower right, you can scroll through every pop, it’s type (occupation), size (number of people), nationality, religion, militancy (the flame), consciousness (the head), party affiliation, issue-ideology, how much money it has, and the meeting of its needs represented by the water, beer and wine cups and so on. Note that the game is simulating Vienna with 276,320people at game start. That is a bit low; Vienna was around this size in 1790, but by 1830 was around 400,000. Still, the effort here is for in game population to mirror actual demography. It’s an impressive effort at simulation.
EDIT: Some commenters have pointed out this is because pops listed here reflect only adult males (presumably simulating heads of household); national population figures are multiplied to reflect women and children. That’s an unfortunate simplification and I hope that Victoria III instead goes the distance to simulate everyone.

On one level, we have to of course note that this is an example of state legibility taken to an extreme: absolutely every human [edit: technically the game simulates adult males and statistically assumes women and children, as some commenters have pointed out; that’s an awkward choice and I hope it is changed in Vicky3] in VickyII has been numbered, categorized and grouped with the other local humans who are identical in all of the ways the game cares about so that they can be treated a single block of people: a ‘pop.’ But I don’t want to take this criticism too far: attempting to simulate the buying habits, political views and occupations of several billion humans would melt everyone’s computers. This is a necessary mechanical fudge.

More to the point, it offers more detail than almost any other game at this scale out there; wildly more (just look at that screen!). In most games – take the Total War series, Civilization or indeed, EU4 – a city or a province might collectively have a religion, a culture and a population figure. In VickyII, that same province or city is broken down; its pops have many different cultures, several different religions. They have different literacy rates, political beliefs, occupations, militancy levels, degrees of political awareness and so on. They’re also different sizes, so the Jewish ‘laborer’ pop in a European city might be only a few thousand while the ‘North German Protestant laborers’ might occupy multiple pops sitting next to them of many thousands. Individuals can transition between pops too, as they change job, or culture, or religion, causing one pop to get bigger and the other to get smaller. They can also just leave, moving to another province (or another state!).

This is actually an enormous mechanical change from EU4. In EU4 the main resource was land; all of the value was in the provinces. But in VickyII, provinces are little more than containers for pops (provinces also contain factories and an ‘RGO’ – resource generating operation – which provide jobs for those pops): all of the value is produced by the pops. They work the jobs, produce the resources, pay the taxes. Crucially armies are recruited from pops; each regiment in your army is directly tied back to the pop that supports and reinforces it. If the pop is radicalized, so is the unit. If the pop is depleted by losses (or job changes, emigration, etc.) the unit stops reinforcing. I want to put a pin in that mechanic, we’re going to come back to it next week because it is very important. But overall I want to stress here how this changes the game: the most important, most valuable resource in VickyII is your population itself. A province with no pops in it is basically worthless. A low value province (different provinces produce different raw resources and some are much less valuable than others) with a huge population of high literacy pops, on the other hand, is tremendously valuable because those pops will produce soldiers, technology, taxes and more pops; once you have factories, such a province can also be an economic mega-center.

Another set of pops, this time in Thrace in a late-game Ottoman Empire. Notice the wide range of nationalities here, as well as the split in ideologies (especially the pie chart on the top right). This is late-game, so all of the ideologies are in place and as you can see, Thrace here has a distressing number of budding fascists (7.5% of the pops), though for complex reasons, the actual fascist party under-performs its support in the population.
The level of granularity here is a spectacular and I love it. While both Stellaris and Imperator notionally also use pops, Victoria II is the only one that really efforts to use pops as a means to simulate an entire population, down to individuals.

As we’re going to see in all three posts in this series, that pop-focus works in a number of ways to actually make VickyII a more nuanced and compassionate game than Paradox’s other offerings (including Imperator and Stellaris, which both use a much more simplified pop-system which tends to commodify pops rather more). The reason here is that, with the exception of some edge-case strategies (which I suspect were truly unintended) the player is almost always incentivized to try to be a benevolent shepherd to their pops. Literate pops are more productive. Pops that are getting all of their needs met (their jobs provide them enough money to buy sufficient goods which are in turn available on the market) are happier (‘militancy’ is reduced), more numerous (assuming the very basic needs are met), less likely to emigrate out of your country and more likely to assimilate happily into your culture. And of course pops that are cared for multiply and since pops are the most important resource, encouraging population growth is one of the most important things you can do.

That system, as we’re going to see, again and again radically realigns the incentives of the player. I really hope that the mechanics of Victoria III continue this trend (from the developer diaries so far, it sounds like they will).

But I promised factories! But first, let’s get…


The industrial revolution, more than any other set of mechanics (including war, colonialism, etc.), absolutely dominates play in Victoria II; perhaps only the personal relationships in Crusader Kings are so prominent a mechanic in a Paradox game. More than the scramble for Africa, the rise of nationalism or the First World War (which I’d argue are the other three major historical pillars here), the industrial revolution is the core process of the game. So let’s go over the mechanics first and then begin asking what Victoria II has to say about the industrial revolution.

Production in Victoria II is not entirely abstract; instead of some single resource, the game has 48 (+1 but we’re going to ignore ‘precious metals’ which just instantly convert to money) different goods all of which are physical products (there are other systems to represent non-physical goods) like steel, concrete, clothes, paper, cars, merchant ships and so on. Of these, 19 of these goods are raw resources. Each province has an ‘RGO’ (‘resource gathering operation’) which functions as a single giant employer which produces a single output good. The amount produced is dependent on production technology and the number of pops employed (with the latter being the most influential). The RGO goods are all fairly basic things: coal, iron, cotton, grain, wool, timber, etc. Lower class pops (including those who work RGOs, ‘farmers’ and ‘laborers’) can live on just these raw products, but even these fellows want some processed goods; Middle and Upper Class pops will demand such goods.

The trade screen showing all of the goods available at game start. Victoria II is full of wonderful, bewildering screens like this, jam-packed with details. For the most part, players can ignore a lot of these numbers, but the simulation is constantly running them, with each good having its price recalculated on the global market every turn (that is, every day).

Processed goods are produced in two ways. The old fashioned way, which makes up the vast majority of production at game start, is production by artisans. Artisan pops will – without any player input or control – pick a finished good to produce, buy an amount of raw materials for it, process them into that finished good and then sell that good. So for instance, artisans might turn timber into lumber (and then lumber into furniture), or dye and cotton into fabric and so on. Some goods have to go through multiple steps before hitting a finished good. Individual artisan pops are not very productive, but at game start there are a lot of them worldwide and so in the aggregate they produce most of the finished goods sloshing around the market in 1836 when the game starts.

Those goods then flow into two markets: first the ‘national’ market, where they can be bought either by the state (typically to maintain the navy or the army which absorb large amounts both of weapons but also food, clothes, etc.) or by the pops living in the state, then to the world market where they are available to other states and other state’s pops. The one quirk here we’ll come back to is that countries which are in the ‘sphere of influence’ of a great power all count as part of that great power’s national market, creating a big insulated pool of production and market demand. The game then dynamically simulates prices: goods that are scarce have their prices increase while goods that are overproduced drop in prices. Artisans, in theory (see below), shift in response to price signals, balancing the whole system. Pops that buy goods consume them and require more and so the system starts again.

The trade screen from a late-game Ottoman Empire playthrough, showing a number of additional goods invented since game start. Also do note how the value of nearly every kind of raw material and basic food has declined, sometimes by as much as a quarter. This is the general tendency over time (I had to let both of these games run to let them hit price equilibrium, by the by; prices always freak out when you first load a save) which we’ll discuss in a moment.

And that’s not a terrible model of pre-industrial production. What players will immediately note at the game’s start is that the relatively low productivity of the artisans makes just about any processed good scarce, while the relatively low productivity of the farmers and laborers in the RGOs means that the artisan class remains relatively small. Consequently, at the very early game, it is very expensive to get even modest quantities of the things one needs to maintain an army or navy, which leaves most states feeling long on people and short on goods because they can’t even mobilize a fraction of their total population. Which is exactly correct as a description of the problems of pre-industrial mobilization too; the problem is that per capita productivity is low, leaving societies long on people (even at relatively low populations) but short on stuff.

This status quo doesn’t exist very long in the game (indeed, for Britain and Belgium, which began their industrial revolutions first, it is already eroding rapidly at game start), but it provides a really useful baseline for players to contextualize what’s about to happen. Because that world of artisans and farmers is going to be up-ended by…


But first, we have to lay out the technology system. Technological development is player directed, but rather than being a product of the monarch’s abilities (as in EU4 or CK3), the main thing that generates technological progress is the average literacy of your population, along with the proportion of the two main educated pop types, ‘clergy’ (who also represent state-employed or private-employed teachers) and ‘clerks’ (accountants, middle managers, etc.). Pops have to be literate to be either, and both also produce literacy among other pops, so the key determinant is literacy. To get ahead technologically, you must educate your country (and the player may choose to invest government funds in doing so).

Technology comes, I should note, in two forms – the actual technology the player chooses to research (which has its benefits and effects) and then a number of attached ‘inventions’ which have a random chance to happen (and apply their effects, which can be positive or negative) once the associated technology is researched. This will matter subsequently but for now I am going to refer to both as ‘technology’ generally even though that isn’t quite the game’s terminology.

The earliest technologies enable the construction of a new kind of production: factories. Factories turn raw materials into processed goods, just like artisans. Unlike artisans, they employ ‘craftsmen’ (a lower class pop; artisans are a middle class pop), produce only one kind of finished good, require lots of capital and resources to set up, but have the potential to be wildly more productive, especially as more technologies related to them are researched. Initially the advantage in productivity for factories is very modest and unlike artisans, they require expensive things like machine parts to maintain, but early factories producing the most basic goods (especially steel and concrete) can be modestly profitable if they can find a large workforce. Even the most basic factories employ ten thousand workers; each upgrade adds ten thousand more job slots.

Great Britain’s starting factories. Britain has the most of any country in the game at start. I won’t bother to show you Austria’s starting factories, as Austria has none.

Where do you get these workers? The technology screen has the answer to this too: a number of technologies substantially increase per-worker RGO output. As countries research those technologies and RGO production rises, the prices of raw goods fall, which causes RGOs (which split profits between workers and upper class ‘aristocrat’ pops) to pay workers less, which causes those workers to shift to being craftsmen if factory jobs are available. The segmentation of national and world market means that this process doesn’t happen quite evenly: the shift from farm to factory happens more rapidly in countries that are already industrializing because their home market floods first. That said, the other big impact here is literacy: literate peasants and miners are more aware of changing economic winds and so shift jobs more readily.

Over the course of the game, as the player moves down the technology tree, factory production receives more bonuses; artisans do not. These bonuses improve not only throughput (enabling a set number of craftsmen to process more raw materials into more final products) but also input and output – efficiency bonuses which enable the factory to use fewer inputs to produce more outputs. Moreover, a factory can employ a second pop-type, clerks (who require a robust education system to encourage) who can further improve factory efficiency. Consequently, as a country’s technology rises, factories can profitably operate at lower sale prices, eventually undercutting artisans and driving them out of producing certain commodities. Alternately, of course, even if a country isn’t industrializing, if other countries are, they’ll inflict this on everyone’s artisans via the world market.

The technology screen showing the starting industrial technologies for Austria. This is actually a decent set of starting technologies,only modestly behind the game-start leaders, but Austria’s awful literacy means that the first few decades will see it falling further and further behind. Attempting to industrialize Austria immediately in 1836 is often unwise, as the factories will mostly be money-losers.

As a teaching tool, the way technologies are structured in Victoria II has some real advantages over the way they function in EU4. A number of the major innovations of the industrial revolution are represented in the main technology categories. The ‘power’ technologies, for instance, go neatly from ‘Practical Steam Engine’ (which is clearly early atmospheric steam engines because they lead into), ‘High and Lower Pressure Steam Engines’ to ‘Steam Turbines’ and then in the very late 1800s ‘Combustion Engines.’ The ‘Commerce technology’ section includes a number of economic theories (Classical, Collectivist, Neoclassical, Keynesian) which are likely to make for useful wiki-walks.

But perhaps the neatest mechanic are inventions, bonuses which have a weighted chance to trigger after a core technology is developed. What makes these so interesting is that they are often weighted to multiple technologies, reflecting the ways that these different technologies and systems of social organization interact with each other. For instance the ‘Rifled Guns’ invention for ship artillery also relies on ‘Precision Work’ reflecting precision boring techniques for making barrels, which in turn relies on both the Mechanical Precision Saw and having researched Early Railroads along with the Mechanical Production technology. And that makes sense! One of the major technological interactions of the early industrial revolution was that artillery-makers wanted precision-boring machines to make uniform diameter barrels (that is, barrels of a consistent caliber). The thing is, a machine that can bore a precise artillery bore can also make very large, very precise cylinders for pistons. And indeed, one such steam engine designer, James Watt ended up working with John Wilkinson, a maker of artillery, to produce more efficient steam engines using a method of barrel boring with a lathe developed for cannon; one of Watt’s improved engines ended up power Wilkinson’s lathe. This all happened in Britain a bit before the start-date for Victoria II, but it provides a great example of the interrelation of these technologies. The industrial revolution wasn’t about one technology, but a complex of interacting technologies, each magnifying the impacts of the others. VickyII let’s the player actually see this happen as developing one new technology often reveals through ‘inventions’ sometimes surprising synergies with other technologies.

Now of course all of this is nifty but as mentioned, in VickyII as in EU4, the player plays as a state and, as we’ve discussed, the tendency of the player is thus to assess everything through the impacts those things have on the state itself. Here, the immediate, direct impacts of the industrial revolution are clear. At the beginning of any game, the actions the player (as a state) can take are generally very limited due to lack of funds and resources. State revenues generally start low, which makes ideal funding of administration, education and the military difficult (or impossible) to balance; even where funds exist, the goods may not and the early game is generally marked by a crunch in the availability of the basic goods used to fuel military activity (particularly small arms, canned food and ammunition). States buy from the world market in rank order and so if there aren’t enough goods to go around, not only do the prices go up, but lower ranked states may be forced to go without (which leads in turn to those states, generally smaller, non-western states, having to rely on the inferior ‘irregulars’ unit which requires only basic goods, over the superior ‘infantry’ or later ‘guards’ units which demand processed goods. We’ll talk more about the western/non-western dichotomy in the game in the last post). And raising taxes to pry those goods out of the world market can also backfire, since you are often competing with your own pops for those scarce goods; taxing them more (which reduces their money available to buy their own needs) while driving up the price of staples can tip them into insufficiency (and from there into either angry militancy or emigration, both of which are quite bad). Early on, for instance, building programs are heavily limited not by available cash, but by just getting sufficient steel, coal and concrete to actually build anything since there is only so much.

The industrial revolution radically alters these constraints. States that succeed in pioneering the industrial revolution rapidly find that successful factories produce a lot of revenue which can be quite heavily taxed without triggering revolt or mass emigration, while at the same time the availability of goods both increases and the prices drop as production methods improve. Things like nation-wide railroad networks, which would have seemed absurd in the first decade of the game, are quite affordable by midgame because of this. Assuming the player is careful in their management, they’ll also see living standards rise, even as their population increases, since production in industrial countries should more than outpace population growth (but see next week on military spending!).

The player thus experiences industrialization, assuming it is done successfully (see below) as a transition from poverty to riches which radically increases their freedom of action. Letting the player do more and see more things happening provides a fairly immediately satisfying feedback. That said…

Again, our late game Ottoman playthrough with the facotry screen. Note only are there many more factories, but they have been upgraded several times. That concrete factory in Aydin, for instance (second row, second from the left) can employ up to 70,000 workers on its own. There are more factory jobs here in Aydin – several times more – than in all of Britain at game start.

Broken Gears

All of this sounds great (and it is) but it comes with the relatively massive caveat which is that almost nothing in Victoria II quite works right. The gaming term for this is ‘jank’ – literally meaning ‘low quality’ but also describing games where their systems mostly work (a game that has ‘jank’ is generally not ‘unplayable’ – it runs and its systems can be navigated) but occasionally go wrong in non-game-ending ways. Victoria II‘s ambition is grand and its systems complex, but as a result it has a lot more jank than your average Paradox game (and Paradox games, especially in the Gen2 era, had a reputation for considerable jank). In this case, the most complex game systems regularly go a little wrong; since the economic model (particularly the World Market) is by far the most complex system, it goes wrong the most.

This is most notable in the fit the market pitches when a new game is started or a saved game loaded. I am not a software engineer, so I can’t really say what the problem here is, but on starting a new session, the game has to ‘find’ equilibrium prices for all of the market goods, which typically takes several ‘turns’ (each turn is a day). For large states with big budgets and revenues, these bumps are minor, but small states can swing wildly between economic bursts and crashes in just those few days as prices wildly recalibrate (and as artisans wildly shift production in the background).

The broader problem is that the world market’s system of price discovery also never quite works smoothly, which complicates another one of the game’s aspects: the choice of economic systems. Depending on which party is in control of the government of the player’s state (which in more autocratic states can be selected by the player; in democracies, the player is left to influence the will of the voters in a system that seems to be intentionally opaque and difficult to chart outcomes based on actions, so that the drawback of unpredictability balances democracy’s substantial upsides in happiness and migrant attraction), the player might have one of four different economic systems.

The most common of these is ‘Interventionism,’ typically favored by ‘conservative’ parties (each state has its own unique blend of political parties which might have different ideological alignments) under that system the state must rely on capitalist pops to build factories, but can freely expand or subsidize factories that exist. Laissez Faire systems, favored by ‘liberal’ parties blocks essentially all government intervention in the economy (and thus nearly all player control of the economy), but gives an important 5% output bonus to factories (note that 5% output bonus increases output by 5% without increasing inputs by the same, so this is an efficiency bonus); under this system, capitalists can build factories much more cheaply, so the private sector acts as a huge force multiplier in terms of expanding industry. On the other end, State Capitalism, favored by reactionary and fascist parties allows the player to do everything interventionism does, plus building the factories themselves; capitalist pops can still build factories on this setting, but they are substantially more expensive. Finally, under ‘Planned Economy,’ favored by socialist and communist parties, completely removes capitalists from the system, making the state responsible for all of the costs of running the economy (and gives a 5% throughput bonus which is strictly inferior to Laissez Faire’s bonus).

Our Austria game, now in 1847, after our first big war. I’ve started industrializing and you’ll note that our literacy is coming up now, 24.5%. The goal of this game was to form Germany as Austria, which is tricky. The way the decision is hard-coded, one either has to destroy Prussia, or ‘sphere’ it (force it into your sphere of influence). The latter is only possible if Prussia is not a Great Power, but Prussia’s formidable industrial base makes pushing it out of the top 8 very hard. I adopted the strategy here of allying with France and relentlessly picking on Prussia to push it out, while playing nice with the Ottomans and Russia to diplomatically isolate the Prussians.

The clear intent here is for Laissez Faire economies to be relatively more flexible and efficient than their state-managed counterparts, with the downside that the state cannot direct production towards things the state wants but the market doesn’t, the obvious example being surplus weapon production (being a major arm’s producer is a huge help when the outbreak of war causes the need for arms to spike upward as everyone mobilizes). However, because Laissez Faire doesn’t allow industries to be state supported, it will only work if you actually are competitive which – as we’ll see in a moment – is only going to be true of well industrialized, high-literacy states. So for late-comers to the industrial revolution, the more planned economic systems can be beneficial; state capitalism and planned economy can both allow a government to ‘boot-strap’ its industries through early, unprofitable stages and to protect industry against market fluctuations.

The problem is that the combination of the janky world market prices and the often baffling failures of the capitalist AI in selecting what to produce where means that an actual player is wildly better at planning factories than capitalists under Laissez Faire. This is, to put it bluntly, not what happened historically. I won’t argue here that planned economies did not have some upsides, but the fairly clearly established drawback they had was an inability to rapidly adapt to changing market conditions.

(Note that systems of ‘social democracy’ are absolutely possible inside this system, through the ‘social reforms.’ You can absolutely have a Denmark-style social democratic state by having fairly high taxes and generous government social programs running on top of a ‘Laissez Faire’ or ‘Interventionist’ economy.)

What the game clearly seems to want is a system where planned economies make sense to jump-start early industrialization, but that successful states will gravitate to Laissez Faire, which ought to outperform the others in raw production once it gets rolling. Instead, skilled players will learn to manage sudden, intentionally triggered periodic shifts from liberal, laissez faire parties (because the bonuses are still powerful) to brief stints where reactionary or socialist parties are in power so that the player can clean up the incompetent mess of unprofitable factories the capitalists left and build a lot of very obvious factories (like steel mills in regions that produce both coal and iron, etc.). That seems fairly clearly an unintended outcome. As we’re going to see in just a moment, the game absolutely has its critique of capitalism, but assuming that capitalists are just generally worse at business than governments isn’t it. For one, when the game was actively being patched, repeated efforts were made to improve capitalist decision-making (with only limited success). Moreover, it would be an a-historical outcome, given that capitalist market economies (notably the USA and Britain) are quite well represented among the historical economic ‘winners’ of the period (1836-1936).

But I said this game has a critique of capitalism, so let’s get to…

Being Industrialized

Perhaps the area where VickyII sets itself most apart is in its willingness to embed unforseen (at least to the first time player) consequences in all of these systems. So far, after all, I have been presenting industrialization in the game as a completely positive thing: more goods, more money, more population, more security, more everything. And in a great many other games featuring building and industrialization that is precisely what you get. Compare factory-games like Factorio and Satisfactory, where you plant assembly lines in what is effectively empty green-field settlement (plus or minus hostile alien wildlife): at most negative externalities are limited to ‘polution’ in some abstract form. There is no human cost because there are no humans. Or, alternately, look at how the same transition is handled by the Civilization series: researching ‘industrialization’ enables various kinds of factories and power plants which create production (‘gears’) with no negative impacts at all. You just get more stuff. Building enough factories opens the choice to adopt communism through the ‘class struggle’ civic, but you don’t have to and there aren’t any negative impacts for doing so. It’s all benefits.

By contrast, VickyII‘s mechanics and systems attempt to simulate both the social and political disruption of industrialization. While industrialization as a whole is still presented positively (a view with which I concur! I rather like modern industrial technology which makes things like mass-produced electronics possible) the game wraps all of these systems in unintended consequences. That’s actually an extraordinarily interesting game design decision which runs against quite a lot of industry consensus: consequences in game design are often supposed to be directly foreseeable based on player action. But consequences in history are often unintended and VickyII in some ways relishes in this fact. Let me get specific here and talk about what I mean, starting with the implications for industrialization on pops.

Remember? This is a game about pops. And that’s where this system starts to shine because it can create situations where major changes have positive and negative impacts on different people because those different groups are simulated, which can in turn cause these changes to have both positive and negative effects on states. Let’s walk through some of them.

As countries research new industrial technologies, as I mentioned, the productive capacity of the RGOs that each province has increases, without a matching demand for more workers. Some of these increases are very large (one innovation – representing tractors – increases farm production by 50% on its own), as they spread, they cause the value of RGO products to drop compared to the final products (you can actually see this in the trade tabs I screen-shotted above). You, the industrialist, hope this will push those farmers into factories, but that only works if there are factories to go which can hire, especially because while the demand for, say, farmers is decreasing, the population is increasing. Moreover, those shifts are disruptive; as price of basic products drops the income of farmers and laborers drops with it, which makes it harder for them to get their basic needs (some of which are processed goods whose prices are not dropping at the same rate). Moreover, their own basic needs increase with technology too – slower than production, but remember while their production is going up, their wages are going down or stagnant. Meanwhile, ‘aristocrat’ pops, which own the RGOs, likewise may see their income fall or stagnate as a group (even while capitalist pops, fat off of their factories, are driving up the prices of goods both groups compete for). Meanwhile, of course the biggest ‘losers’ from industrialization are your artisans, who are being steadily competed out of their jobs and become stuck in situations where they cannot buy the raw materials to produce finished products. Initially they’ll retreat into complex goods you don’t have factories for, but one by one those options will be lost as technology advances. Eventually they’ll demote to a different class (typically into the lower class, though if they are literate they may move laterally into being clerks), but obviously they are going to be upset by this process of impoverishment.

As pops get upset they gain ‘militancy’ – a statistic that reflects how bothered they are by the status quo. Militancy is, for the most part, a bad thing and this directs the player’s eye back down to the pops. Unlike in EU4, where the player, implementing policies that were good for the state but bad for people often had little feedback on this harsh reality, in VickyII, militancy is the direct feedback. The game’s population windows directly note pops who are not having their needs met. High militancy, among other effects, pushes out-migration which, because that means losing pops and pops are the most important resource, is to be avoided at all costs. Consequently, the player is made very aware of the negative impacts industrialization is having on their people; those impacts have real consequences for the state.

Meanwhile, you are pushing education too, trying to get a large block of literate citizens, because literate citizens more readily switch jobs to being craftsmen or clerks, who you need for your factories and they also produce faster technological advancement. But literate pops also gain ‘consciousness’ – a statistic reflecting how politically aware pops are. All pops have a set of ideological views (which can change over time), but low consciousness pops are generally both less bothered by a government that isn’t doing what they want (because they aren’t aware of it) and less active in pushing the party that matches their ideological preferences (because they’re not politically active). But pushing literacy unavoidably means raising consciousness. And, to be fair, consciousness can be a good thing for you – if your pops want to implement the same reforms you do, consciousness helps you overcome the old aristocracy which is invariably reactionary and already very politically engaged.

But the impact of consciousness is heavily dependent on what the pops in question want (and how able you are to deliver it)! High consciousness pops who find that the government doesn’t reflect their views begin to develop militancy if they cannot effect reform. This can be very tricky if the view that those pops have is that they are a subject people in an empire who deserve to be free! Consequently, for large, multi-ethnic empires, high literacy can actually be dangerous. On the flip side, consciousness among the new industrial classes – craftsmen, clerks, capitalists – will push them to demand liberal or socialist governments that accord with their views (generally), which is fine if you want to go that way, but if you intended to maintain an old fashioned style of government (or a more extreme one!), those pops will start to get angry. On the flip, flip side, remember that industrialization has produced a whole lot of angry farmers and laborers whose jobs have vanished, alongside a smaller number of upset aristocrats. These can easily build in militancy and coalesce into a reactionary movement if their consciousness is allowed to rise.

Even more than the low literacy, this screen is Austria’s problem. Each of these potential releasable nations (and this bar scrolls even further) is going to have its own nationalist movement if consciousness is allowed to rise. But we need literacy to keep up, which means we will end up with high consciousness anyway, which means movements of national liberation. Austria is trapped on the horns of a dilemma.

Consequently, the game presents a situation where, at game-start, much of Europe (and the world) still has a fairly stable social system: high consciousness conservative aristocrats and low-consciousness conservative peasants. But as a state industrializes (and it must, for reasons we’ll get into in just a moment) it disrupts that stable balance, leading to rapid social change. The good news is that productive, wealthy states can manage this transition, enacting reforms (either political reforms to deliver actual freedom or a Bismarck-compromise of social programs without political reform) to transition into efficient, industrial countries where the steady flow of goods and state benefits (combined with fairly low taxation because factories throw off so much money) keep everyone happy through the transition.

But that’s not going to happen for everyone.

So far we’ve been looking at this from the perspective of a state at the forward leading edge of industrialization, but what about everyone else? For states that begin with lower literacy rates and less favorable economic situations, the way forward is a mine-field. Simply refusing to industrialize can maintain the stable social fabric in the short term, but in the long term the decline of raw material prices will immiserate your pops, while stagnant state revenues will fall behind militarily with disasterous effects (recall the interstate anarchy of our EU4 series). Leaping forward through industrialization can work, but only if you actually get to the pot of gold on the other side of the rianbow. But what can equally happen is a country becomes trapped; it has partially industrialized and built the initial, foundational factories making first stage raw goods (mainly steel and concrete, the demand for which remains very high throughout the game). But moving to more complex goods means competing with other economic powers who (due to their higher literacy) have more economic tech than you do, which means they have more efficient factories than you do. Your entry into the market depresses the price and you promptly price yourself out; the factory closes and the people riot. Obviously you must improve literacy, but that means increasing consciousness, with all of its perils.

The result is a ‘middle income trap‘ of sorts. But the stagnation often pushes a state into more violent politics as militancy rises. High militancy can push pops to migrate out of a country, but it can also push them to revolt. While these revolts are often easily managed militarily (an event chain causes the 1848 ‘Springtime of Nations’ to occur on schedule in each game, but badly managed countries often see this as the beginning of long periods of instability), crackdowns damage the very resources (pops and factories, the latter close when a province is occupied, for instance by a revolutionary army) that you need to advance. Meanwhile, high militancy pops that aren’t of the core culture of the state might, instead of demanding political reforms, demand freedom and begin petitioning outside powers to try to get it, either in a war or a conference. Consequently, while industrialization is good for the ‘winners,’ it can also trap less successful states in the abyss of perpetual violence and revolution that gapes between the stagnant, stable pre-modern social structure and the modern industrial societies. As we’ll see next time, the game is perfectly willing to measure that abyss of revolutionary violence in blood and every dead pop weakens a country. Meanwhile, countries that don’t industrialize at all remain internally stable, but become weaker, and weaker, and weaker.

Those unintended cross-pollination of technology now has its impacts too. One category of technology is ‘political thought’ – each level of this increases the number of national foci (which you can use to encourage a certain kind of industry, or job type, or push colonization, etc.) you can have, reflecting a government that is able to do more. But the later levels have some surprises hidden in their inventions. While ‘Nationalism and Imperialism’ will improve your troop morale and reduce war exhaustion (and so is necessary to keep pace militarily), it also substantially boosts separatism and nationalist rebel organization. The subsequent technology, ‘Revolution and Counterrevolution’ opens up much more dangerous revolutionary movements. Suddenly militant pops aren’t turning into relatively harmless (or even modestly helpful) liberals and socialists, but into positively dangerous fascists, communists and violent anarchists (the last bizarrely named ‘anarcho-liberals’ but who clearly represent the anarchist bomb-throwers of the period). Just as there are hidden bombs lurking in the economic technologies, so too there are hidden bombs in the cultural technologies – the unexpected consequences of trying to harness popular movements for state ends.

So while on the one hand, VickyII clearly presents industrialization as good on the whole in the long term, in the short term, it does not shy away from the idea that these processes hurt some people, and break some nations. This is a welcome level of nuance compared to EU4‘s ‘state power good’ framework and it comes out of the game’s focus on pops. A player that relentlessly focuses on short-term state power at the expense of their pops will ruin their nation in the long run. Put bluntly, Victoria II presents industrialization much like the proverbial ‘powerful servant, dangerous master‘ even before we get to its impact on conflict (which we’ll get to next week).

This is actually, I think, a brilliant bit of design. Students often find the halting, two steps forward, one-step back industrial and political policies of places like Austria, Russia and China in this period utterly baffling. Looking from the safety of modernity they ask, “why couldn’t these fellows have seen that the factory and the ballot were the way of the future and go that way?” One of the great virtues of the simulation here – much as EU4 presents the same problems for war and peace in a system of interstate anarchy – is that VickyII provides the answers to that question with the power of a simulation. One merely need to play a game as Austria or Russia (especially as a new player who doesn’t have mastery of all of the systems) and watch as your state shreds itself under the pressures of nationalism and industrialization (as both states did historically) to immediately understand why so many of their leaders attempted to ‘stand athwart history, yelling stop!’ You see, as they saw, the abyss – and also feel the same pressures they felt which impelled them towards it.

National Success Sequence

There is, I think, one more thing Victoria II wants to say about the industrial revolution, which is that the game mechanics clearly suggest a ‘success sequence’ of sorts, a set of tactics and processes that a player can use (assuming they manage militancy well) to pull off the transition from pre-industrial to industrial country. Countries that succeed avoid ruinous wars, build literacy aggressively from the beginning, wait to industrialize until they have a literate population which can compete globally. While state funds may then jumpstart industrialization, excessive subsidies are to be avoided – you need factories which can actually compete globally. Finally, you want to either be a great power with your own Sphere of Influence (which functions as a jumbo-sized national market) or a secondary power inside of a large Sphere of Influence. Either will actually do. Use political and social reforms (whatever you can get your political class to tolerate) to keep militancy low. Finally, anything and everything that encourages pop growth and discourages emigration is important; your people are the most important resource. Since social reforms to the former and political reforms the latter, these are valuable as ends unto themselves.

The success of this sort of open-societies, open-markets approach to fostering national economic growth is of course not without its critics, but it certainly does accord with prominent strains in the literature. As noted, the game is also open to a more state capitalist approach where the state takes an active role in the economy and provides social services to keep the peace (roughly mirroring the trajectory of Germany in the late 1800s) as well as a full planned economy (roughly matching the Soviet Union’s rapid industrialization in the last decade the game covers), but these are mostly minor variations on the standard success sequence; literacy, factories, etc. are still required.

Of course that success sequence also embeds some values. VickyII‘s mechanics fairly clearly prize dynamic, industrial and free societies over static, pre-industrial, closed societies. On this last point it is worth noting that the political reforms mostly lack explicit drawbacks for a state intending to move towards liberal democracy (social or otherwise). The social reforms are more of a mixed group; health care, unemployment subsidies, public school system are all clearly more than worth the cost is nearly all situations. Minimum wages and safety regulations have to be introduced with care as they can push factories into the red, but in the long run benefits outweigh costs. Maximum work hours, with its factory throughput malus is the only truly ‘situational’ reform from a mechanics perspective. Consequently, I’d argue the game also presents shifting towards liberal social democracy as the desirable path.

Those values are baked into the mechanics but they represent a theory of history which views the movement towards modern post-industrial economic systems along with modern liberal social democracy as ‘progress’ – a clear movement from systems which were ‘outdated’ (that is, both ‘older’ but also ‘worse’) to systems which are ‘modern.’ The general term for that historical view is ‘whig history,’ a vision of history which imagines it as a progression from a bad, dark past to a current, enlightened, superior (and perhaps perfected) present. The phrase is often pejorative, a critique that the historian is assuming the end-point of a process whose final end is as yet unknown (and indeed, a process which may not have a single final destination as such) and moreover accepting as obviously good systems of value and organization on the value of which people may and do disagree. On the other hand, there are few historians who, if pressed would admit to pining for serfdom or the return of polio.

And here is our Austria game, now in 1867. The strange pattern of territorial acquisition here is intentional – I am aiming for Prussia’s highest literacy and production provinces because my goal is not to destroy Prussia, but just to demote it. Note that it is only now that my literacy rate matches what Britain has at game start (the research total is much higher because there are technologies that act as multipliers for it). This is after 40 years of maximum education spending and using national foci to encourage the training of ‘clergy’ pops (which also represent teachers and increase literacy.
I’ve also used one of the bespoke decisions (a bit of that Gen2 railroading) to cause a number of my ‘spherelings’ (states in my sphere of influence) in southern Germany to merge into the South German Confederation. This is mainly useful because it reduces the diplomatic upkeep of keeping them in my sphere, letting me focus elsewhere.
Also, a Wild Italy has appeared.

The deeper problem with ‘whig’ formulations of history is the general sense of determinism; not that conditions are shown to be better now than they were then, but rather the assertion that history must move inevitably towards this current point. The neat phrase for this trap of thinking is that it is ‘teleological’ from the Greek telos (τέλος) meaning an decisive or final point. It is the trap of assuming that because this is where we ended up, this is where we must have ended up. And VickyII‘s vision of industrial change is, in fact, teleological in that sense – there are only so many technologies, you will research most of them and they lead this set of systems, quite inevitably to this particular conclusion. As with EU4, the sense of inevitability is strengthened by the persuasive power of the simulation. And precisely because VickyII is a Gen2 Paradox game, with a greater degree of historical railroading, the end-point is relatively more fixed, making it feel like the historical end-point was the only possible end-point, which it wasn’t.

Though, to be frank, I have trouble imagining how this particular game could avoid some of those traps, given its time frame. It’s not clear to me how, by 1836, the genie-in-the-steam-engine could be put back in its cylinder, short of some sort of vast catastrophe consuming much of the globe. Consequently, I am inclined to pardon VickyII a lot of its teleological nature (far more, for instance, than EU4, but EU4 is mercifully less railroaded) given how short its time-scale is. By 1836, the broad outlines of what was to come (that is, industrialization and the new imperialism, but not particular events more narrow than that) seem to me to have been substantially baked in by the combination of the establishment of Europe’s colonial empires over the previous three centuries, the economic developments in Britain going back decades and finally the legacy of the American and French revolutions. There was a lot of space for variation within that framework (and there is in the game, just look at the Austria screenshots), but many of the factors which shaped the rest of the century were already active. I hope that Victoria III will be a bit more open and a bit less teleological in this regard, but that seems a tall order for a game such as this.

At the same time, I appreciate that VickyII is willing to demand the player think about what these momentous changes meant for actual people and to grapple with the idea that these changes did not always leave people better off.

Next time, we’re going to turn and look at what all of this industrialization means for how Victoria II treats war, which I think even more than the industrialization system itself is perhaps the crowing achievement of the game’s historical vision.

144 thoughts on “Collections: Teaching Paradox, Victoria II, Part I: Mechanics and Gears

  1. Typo:
    “Attempting to industrialize Austria immediately in 1936 is often unwise, as the factories will mostly be money-losers.” I assume you mean 1836 here.

    This game sounds more interesting to me than EU4, which sounded to me like a game where you have no choice except to be an amoral bastard, which probably won’t work either. Note I know nothing about either of these games except what I have read here,

    1. It’s definitely more interesting to talk about than EU4, though in play many of its mechanics are very much one-solution things. Once you understand the game, for all the underlying systems and simulations, the player inputs are actually pretty simplistic. It’s also quite easy.

      Part of that is Vicky 2 does not compromise simulation so much for what we might call ‘game value’, but part of it is also Vicky 2 is just particularly designed in the first place.

    2. I hesitate to say this given that I have played neither game, but based on my experience with other Paradox titles (Hearts of Iron IV and Stellaris) once you have mastered the game’s systems you are such a vastly superior player to the AI that you can make some highly “sub-optimal” macro level choices (such as playing nice) and still succeed by whatever metric you choose.

      1. Why is the typical player so much smarter than the AI in these games? For a few bucks, you can buy a chess program that will beat anyone reading this (unless maybe there is a grandmaster lurking). Are these games so complicated in a way that’s categorically different from the complexities of chess? Or is it a feature, to assure that the player wins and gets to feel good? I’d really be interested in what someone who knows something about AI, programming, or game design has to say on that topic, if such a person is reading.

        1. It’s basically all of the above.

          Firstly, there’s just a lot more time and effort put into chess engines than the AI for most commercial computer games. Hundreds or thousands of researchers and programmers have been working on chess engines for 50+ years. A game like Vicky II might be lucky to have a team of a handful of AI programmers for a couple years (and if it’s not lucky it might only have a single AI programmer for a couple months), and those programmers are aiming at a moving target: they’re developing the AI at the same time that the designers are tweaking the systems that the AI is supposed to interact with. With chess, you can be confident that the AI you write today won’t be irrelevant next week because the designers decided that castling should work slightly differently.

          Secondly, a game like Vicky II is orders of magnitude harder for a computer to play well than something like Chess. A complicated midgame chess position might have 30-50 legal moves that the computer has to consider. In Vicky II, a player has 30 different options for how to use a single National Focus slot (even leaving aside the choice of which state to target with that National Focus). A midgame, medium-size state with 10 states and three focus slots would have a little less than 27,000,000 possible ways of using them, and that’s just one mechanic among many interlocking mechanics that the computer would have to master. In practice AI programmers either write very simple rules to handle this complexity (so, the AI might always apply a specific focus to its three largest states) or they just cheat (the AI might not use the focus system at all, and just get passive bonuses to make up for the weakness).

          And ultimately, commercial game AI just isn’t trying to do the same thing. Chess engines are designed to beat humans in a fair fight (and in recent years beat other AIs, since it’s been ~10 years since even the best human players could hope to win even an occasional game against the best AIs). In some ways creating a commercial game AI is more like designing a puzzle than it is like creating a chess engine. For one thing, there is no such thing as a fair fight. The programmer controls the rules of the game, so making the AI just win is trivially easy, in the same way that designing a puzzle that has no solution (or an impossibly obscure solution) is trivially easy. The goal isn’t to win, it’s to give the player the right amount of challenge and the illusion of playing against an intelligent opponent.

        2. There are accusations that it is a feature, that Paradox could deploy a ‘Very Very Hard AI’ that was actually smarter, not just more cheaty than the Very Hard AI; the most conspiratorial part of these rumors is the idea that it’s a business decision, that having the hardest setting be too hard would somehow drive players away. Paradox simultaneously has a reputation among those same disgruntled fans for extremely poor long-term business planning, so I don’t credit the theory much.
          In terms of the chess analogy, there are multiple systems of inter-player competition, not just military tactics, but even though they do try to keep each one simplified a bit, each still effects the others. So it’s like you have multiple boards stacked on top of each other, with moves on one board slightly changing the shape of the other boards. The key difference is size, though; calculating all possible moves from a static state on an 8-by-8 plane takes a certain amount of computing power, and could even be done by a pre-electronic computer with infinite time. As I understand it, each ‘board’ is more like a 40-by-40 plane, or bigger. Current improvements in chess AI are all about making better choices from a high information position. Even getting a Clausewitz engine game to that level of information in the first place would take a lot of processing power, and consequently slow things down immensely. Stellaris actually builds most of its ‘board’ complexity during gameplay, so an absolutely typical player experience for Stellaris is to have a late-game that progresses through turns (days) so slowly that playing at the top speed setting will move through days at a rate that is slower than what the slowest speed setting will give you ate the start of the game.

        3. In adition to what others have already pointed out, there’s also the consideration that the AI for Victoria II (or Europa Universalis IV, etc.) is running over a hundred countries at the same time. Or if you prefer, the player’s computer has to run over a hundred instances of the AI at the same time, in parallel. Clearly, that dramatically reduces the AI’s potential capabilities.

          Furthermore, the player has complete control over the speed that the game runs at. The player can pause the game in tricky situations when there’s a lot to think about (or a lot to click on) but can go up to “speed 5” (i.e. time passes extremely quickly) when things are uneventful. Except, of course, that an uneventful time for the player doesn’t necessarily equate to an uneventful time for all the AI countries. There might be peace in Europe but war in the Americas, for instance. The AI has to be able to conduct all of their affairs successfuly at speed 5, even though most human players would struggle to do so. Or rather, they have to be able to do everything over a hundred times faster than speed 5, since each country only gets a fraction of the AI’s total processing time.

          Finally, consider that the game developers want the game to be able to run on low-end hardware, since that maximises the size of their potential audience. So this means that we’re essentially left with an AI that must be able to operate at lightning fast speeds, a hundred times at once, while running on a toaster. It’s hardly surprising that it isn’t making great decisions.

          It would be interesting to see what would happen if someone were to create an AI that was allowed to play in the same way as a player (only running one country, being able to pause at will, etc.). My guess is that it would be relatively easy to create an AI that could outperform human players, but there isn’t really any demand for this so no incentive for anyone to actually work on it.

    3. I’m not sure if you’re less of an “amoral bastard” if you implement stuff like minimum wages and unemployment insurance not for their own sake, but because they make your state more powerful. You might feel better about it as a player, but both games presume a set of international realist ethics in which the security/wealth/power of the state you’re playing is an end in itself, not a means to the ends of life, liberty, or the pursuit of happiness. “Amoral” is a pretty good word for someone who is equally willing to help or harm others in order to achieve their own goals.

      You could make the argument that Vicky is the more cynical game of the two, because its mechanics imply that powerful reformers aren’t motivated by compassion or idealism, but by expediency.

    4. Victoria 2 is a much more interesting game, indeed. Though I was surprised by how much this blog could talk about EU4, even that game’s fans often recognize Vic 2 has more going on on a conceptual level. The “jank” talked in the article is the real problem, though. EU4 is just a more player friendly game and far less broken.

      However, it should be noted that Vic 2 still pushes you into being an “amoral bastard”. I imagine this will be talked more in a later part, but imperialism is, mechanically, mostly a positive thing in game. Literacy, consciousness and militancy matters far less in “colonial” territories which means acquiring more of them provides you with resources and population for army without you feeling the impact on the colonized pops. It is one of the ways I hope the sequel will improve.

      1. One of the things I noticed in my earliest games of Victoria II was that the game incentives was effecting my behaviour. I was running a fairly “nice” game as a Canada released by the UK and was doing well with internal development and strong social reforms. But I got tempted into conquering small tropical states for their rubber because otherwise I wasn’t getting enough to support the high tech automotive and electronics industries I had so painstakingly created.

    5. Another typo:

      “I will be using ‘state’ as a technical term to \prefer\ to independent polities (or ‘countries’) and I don’t want to be confusing.”

      I think you mean refer, not prefer.

  2. Great post as always.

    Minor correction: you mention in screenshot #4 that the game models fewer people in Vienna than were actually there at game start. This is actually because the game does *not* count all the population in a country in its pops. Victoria 2 Pops represent only the *healthy adult males*. All other parts of the population – children, women, elderly, disabled, etc – are abstracted away. It is assumed that part of the purchases of Pops in the game goes to their dependents, hence why e.g. Pensions increases the money available to Pops. But everyone except for adult males are not actually simulated in the game, and that’s why pop numbers are systematically smaller than historical population numbers.

    Victoria 3 looks like it will actually count people other than adult males as part of Pops, as well as having laws (child labor laws, women’s suffrage, etc.) that can move parts of a pop between dependents and workforce, as well as between politically powerful or suppressed. So it will be more granular. But Vicky 2 abstracts them away entirely.

    Also I feel I should note that various entries in the Civilization series *do* attempt to model negative impacts of industrialization. That said, it is true that Civ in general is less interested in the internal dynamics of your civ, so these negative impacts are almost entirely pollution and climate change, which various games in the Civ series have modelled in different ways and with different levels of success.

    1. Defining “population” as equivalent to adult males may be politically incorrect, but for much economic analysis, the household, not the individual, is the appropriate analytic unit. And for most of the Western and Asian world in the period being considered, the number of households would equal the number of adult males.

      1. For most economic analysis, individuals are the appropriate unit unless households even out to approximately the same composition everywhere, in which case population and households are equivalent. Composition and/or the size of the household matters for food use and other goods use, the types of work people do, how much education is being provided, health care/retirement, etc.

        1. Well, someone should tell the census department to stop reporting median household income. And my family makes most of our major consumption decisions on a household basis. Indeed, the children weren’t even consulted when we bought our house, our single biggest expense.

          1. Household income is a useful piece of information, therefore it gets collected. Individual income, however, is the more useful measure.

            Decisions made by several people also are modelable as individual decisions. Assuming that an individual makes a decision, which is sometimes similar to other similar people, is a more useful way to model things than assuming household decisions, and trying to scale back when households don’t make sense, or when members act independently.

          2. I just want to note that someone tells the US Census that every damn year. This is America, we get what we get.

          3. “Decisions made by several people also are modelable as individual decisions.” True, but that introduces a lot of complexity which may not be analytically useful, or helpful in designing a game. Children have very little role in most household decisions. Even among the adult decision-makers, many households operate on a “separate spheres” or “division of labor” basis, whereby some decisions are made by the husband and some by the wife. I would guess that this method of allocating responsibility was even more prevalent in the Victorian period. For analytic purposes, therefore, it often makes more sense to treat the household as a unit, without regard to its components–just as, in many cases, it makes sense to treat an organism as a unit, without regard to the goals of its individual cells, or to treat a nation-state as a unit, without regard to the desires of its individual inhabitants, and so on.

  3. XD.

    I mean, being state capitalist is strictly better then letting capitalists build clipper factories in all your provinces forever. The player knows what will profit (and what has to be subsidized for the military) while capitalists are, well, dumb as rocks in the game. The power of decision overrides all other concerns of efficiency.

      1. The main flaw in the simulation is the existence of the player, an immortal super-minister who knows the future. As long as the player directs the state while an AI directs the capitalists, this problem is probably unfixable.

          1. Eh. There are two problems here: First, that Connecticut Yankee “knowing what I know, can I do better than Napoleon/Shaka/Meiji?” thing is part of the appeal of these sorts of games. So you don’t want to eliminate it entirely: a player should be able to identify a goal and work towards it. (The trouble with Vicky’s capitalists, IMO, is not that they sometimes turn down blind alleys, but simply that they look at any destination, check the map, and then turn right at every intersection. If they just thought clipper factories were good investments because steamships seemed farfetched, that’d be one thing, but they simply have no idea what a good investment would look like in any case.)

            Second, and more seriously, to make your historical simulation sufficiently unpredictable that the player’s knowledge isn’t helpful, you would have to create wildly unlikely outcomes. It would, necessarily, have to be possible that this whole “factory” idea was just a flash in the pan and by 1850 everyone realizes that cottage production was the only truth. Or that nationalism never really catches on outside of France and is relegated to the history books as a weird Jacobin excess, like the decimal calendar. Or that all-big-gun battleships just don’t work and are roundly defeated by pre-dreadnoughts.

            It’s possible to do this, somewhat! Lots of Eastern Front wargames have some kind of “randomized Rasputitsa” mechanic, so you can’t game out your knowledge of future weather. For a more substantial example, fans of janky historical gaming might know Rule The Waves, a sort of Admiral Jackie Fisher simulator. The sequel has an optional setting where the effectiveness of carrier-based aircraft is randomized and unknown to player and AI both. This means you play without knowing whether the “carrier admirals” or “battleship admirals” were right, and you must balance your known-quantity big-gun ships against enough carriers to give these new aircraft a try. (Even there, of course, the real question isn’t “do planes work?” so much as “is it time to build a navy centered around carriers, or should we keep them as vital support ships for a little longer?”)

            Trying to do that with EVERY major variable in a game of broad scope would lead to madness. The ahistorical outcomes would have to be fairly likely, at least 10-25%, because if the player knows X is going to be better than Y 99% of the time, we haven’t really changed anything. So if you’re randomizing 30 or 40 variables (which isn’t really very many for a game of this breadth) you can expect an average game to have half a dozen major deviations from reality. That could be quite interesting, but it would no longer be even vaguely historical.

          2. That, and also problem #3: knowing the future is only half the issue. A decent player will generally have a better grasp of game mechanics than even the best mass-market strategy AI. An excellent player with 100+ hours of experience will thoroughly outclass the AI, to the point of being able to produce wildly improbable outcomes. Central planning is actually a good idea when the planner is an omniscient super-genius.

            Maybe someday, AI will advance to the point where every game studio can build an AlphaStar or Deep Blue, and every gamer has a machine that can run it. Unless/until that day comes, Paradox games are doomed to be as much a sandbox as a simulator.

          3. As with randomizing the effectiveness of carrier-based aircraft, you could produce historical indeterminacy by varying the effectiveness of particular industrial processes. What if cotton spinning were much less efficient than it actually was, for example? (I don’t have the technical expertise to say why, but my recollection is that wool and flax spinning were much less effective in the 19th century, so the time and money invested in those enterprises was not well-spent.)

            Even more difficult would be randomizing the effectiveness of intellectual or cultural movements. What if nationalism simply never appealed to any large number of people? (As, for example, Buddhism has never developed into a mass movement in the US.) That would make for weirdly different history.

          4. The core problem with relying on greater divergence in these areas is that the things that need to be made divergent are based on physical realities,

            For example, if it turns out that the Bessemer process doesn’t work and you simply cannot make steel on an industrial scale, then this radically changes much of the history of the Industrial Revolution, in a way that makes player foreknowledge much less useful.

            But the Bessemer process is grounded heavily in the basic laws of physics and chemistry; it’s hard to imagine a world allegedly “ours” in which mass-production of steel is a simple impossibility given that all the prerequisite concepts and scientific knowledge to create such a process exist. The game occupies a point of tension between inflexibility and improbability.

            I think the better solution here is to just find a way to make the in-game capitalists… not as dumb as rocks. The problem isn’t that the capitalist pops are making decisions based on recognizable belief systems that happen not to line up with reality, but might conceivably have worked out in a plausible, self-consistent alternate universe. The problem is that the capitalist pops are effectively squandering all their investment capital building factories that a sapient being would tell them aren’t going to work out.

            If every capitalist in Britain simultaneously decided to blow all their money on building shipyards to make (say) sailing ships and factories luxury furniture, and stuck with that decision, it would inevitably have crippled Britain’s industrialization. No remotely realistic economics/politics/military affairs simulator could save Britain from that outcome. But then, realistically the British capitalist class wouldn’t have done something so stupid in the first place, not as a collective whole.

            Victoria 3 may well be able to largely resolve this problem just by finding a way to program capitalist pops so they’re not a bunch of useless ninnies.

          5. Paradox AI has gotten noticeably better in the past 15 years, but it is still frequently dumb as rocks. What players rarely acknowledge is that this is mostly a feature, not a bug, because if the AI *weren’t* dumb as rocks, they wouldn’t be able to conquer the world as Kolguyev or whatever.

            The “AI too dumb” complaint only pops up on the rare occasions when the AI needs to assist the player somehow. In CK2 (after years of complaints), PDX mostly solved this by grafting on a set of “ally order” buttons that the player could use to tell AI allies what to do. Now CK3 is out, and “my allies won’t help me correctly” is once again a common player complaint.

    1. But capitalists aren’t dumb as rocks, they go where the profit is. Demand equals profit. And they tend to support innovation. By and large they’re not going to insist on building clippers when steam is the coming thing.

          1. Right. And the general consensus is that the programmers didn’t want the capitalists to act as dumb as rocks, either- this wasn’t a deliberate attempt to denigrate capitalism by stacking the deck against laissez-faire economics in their simulation.

            They just couldn’t figure out a practical way for capitalists to make sensible decisions in the context of their situation… and eventually released the game without it.

          2. The problem with that is the developers ALSO know the future, just like the player. It’s probably a difficult but not impossible problem. (Whether it’s a feasible solution is another matter.)

          3. Huh? How is it impossible to make NPC have some jusgeme? Don’t they call it AI?
            Am I showing my age?

          4. @Roxana

            The problem here is making sure the capitalist pops “know” the right information and are programmed to use the right information. For instance, if your country already has a dozen giant steel mills churning out more steel than you can use, you don’t want the capitalists robotically starting more steel mills. You want factories making something else, but what else?

            Programming the capitalist pops so that they “know” to construct factories whose goods are in demand, appropriate to the overall technological level of development (no building an airplane factory in a country that lacks the industrial capacity to make its own internal combustion engines), drawing only on the precursor materials available… Well, that turns out to be hard, not least because there is no one “right” build order or prioritization.

            I’m not saying it’s impossible, just that it’s not a trivially simple task, and that Paradox didn’t do a great job of it with Victoria 2 back in 2007.

          5. Perhaps the capitalists could be simulated more like capitalists, and go bust if they make poor choices, like investing in production for items with high supply and low demand. That doesn’t even need AI.

          6. “Perhaps the capitalists could be simulated more like capitalists, and go bust if they make poor choices”

            That could weed out the dumb ones, but it doesn’t work if they’re all dumb.

          7. Honestly, the best suggestion I’ve seen for capitalists in Vicky II is to make them truly random, with no pattern of investment at all, and be much harsher on capitalists with bad investments. They’re still dumb, but they won’t mechanically, collectively all decide to invest in clipper factories until the end of days. Instead, once you have a sufficient amount of them, you just let the market d what the market does best: Sustain profitable businesses and make the rest go bankrupt.

  4. As a historian, not a game-player, I second the commendation of the game for showing how difficult many of the choices were for decision-makers of the time, and particularly for those in weak positions – China, the Ottomans, Persia, Russia.

    I would note that ‘laissez-faire’ was always a minority position politically – European (and US) governments were very interventionist in the period, even if the interventions did not take the form of direct control (think land-grant universities and railroads in the US, British manipulation of Indian tariffs, French subsidies for steel and railroads, deliberately weak patent recognition, telegraphs…). In the British case, industrialisation was the result of a century of very active industrial policy – something other states noticed but found difficult to emulate.

    1. Ottomans are indeed difficult, but here railroading helps you. Just send your egyptian armies into the desert so they die. Then when egypt breaks away you can just wreck them without any trouble.

      1. Not at all “weird”. They complain about the types which discourage initiative and encourage dependency, interfere with mutually beneficial exchanges to no rational purpose, and/or have been proven by the last 200 years of history to greatly reduce economic growth and hence human material comfort in the long run.

  5. Ah, military force from population! This was the era in which the Highland Clearances led to inability to raise Highland regiments — unless you manned them with sheep as was pointed out at the time.

    1. This was, naturally, mainly a problem because the British had a very odd fixation on the idea that certain populations (highlanders, gurkhas etc.) yield better recruits. While this may have been true in the 19th century, WWI showed very well that a boy from East End of London is able to fight and die for his country quite as well as any other Briton. The point was actually training to your recruits.

        1. But if the highlanders’ martial qualities are racial – that is, genetic – you’d expect their descendants to keep those qualities even if the family moves out of the highlands.

          1. They tended to move to Canada or the United States, which made recruiting a bit difficult. Indeed, I believe some of them were forcibly shipped over the Atlantic, which would also create some ill will.

          2. Mostly. Even in the ‘scientific racism’ era, all of these popular understandings of racial character were not rigorously scientific. So people recognized that the mountains made their resident families breed soldiers, but believed that his was hereditary to the Scottish race; no one thought you could starve Englishmen at the top of a hill for a few generations and get good soldiers that way, but everyone knew you could with the Scots (and the Swiss!). There was even a sort of half-directed effort to create a colonial ‘soldier pop’ by taking (primarily lowland-ancestry) Scots from Ireland and the Borders, and settling them in the Appalachians. Today, the descendants of those people will often tell you that it worked, too!
            And if families didn’t lose their ‘soldier-ness’ for a generation or two, people saw that, but it was just easier for people to understand that as ‘Scots need rough country to be who they were meant to be’ than understand that as ‘it can take a lifetime, or more, for economic refugees to find their place in a new economy, but they will find it.’

      1. As our host pointed out in the post on speeches, a military tradition definitely has its advantages in fighting, especially when fighting less than total war.

      2. The ‘martial races’ notion is a post-facto justification that took on a life of its own. Sikhs, Rajputs and Gurkhas were recruited because they stayed loyal in the Mutiny, Highlanders because they were cheap and to bleed the clans and so on. As you say, anyone trained can fight well – but the East Enders came to the uniform with uncomfortable political ideas (as did Bengalis, Rhinelanders and others) – in this game mechanic they were highly conscious and militant.

        1. Prussia (and later Imperial Germany) were much more conscious of and explicit about this – they saw rural recruits from certain traditional areas as desirable more because of political reliability than because of fighting quality. A notable factor in pre-WWI German military planning was a hesitancy to extend the draft to include too many potentially-subversive urban workers.

          1. Indeed. In the early 1900s the Minister for War rejected a proposal to increase the army because it would require letting the urban middle classes become line officers (they could sign up for the engineers and artillery because educated, you know) and they might – horror – have social democratic leanings.

        2. Of course, the post-facto justification may well have reflected underlying reasons, if not the reasons the 19th century British would like to admit.

          For instance, suppose you are exercising colonial rule over a giant subcontinental landmass. A general uprising blows up in your face. Which groups of soldiers are those most likely to remain loyal? Among the answers to that:

          1) Soldiers from ethnic groups that are effectively outside the radius of your control and who are in effect mercenaries with no personal stake in the outcome of the uprising beyond “if my boss loses, I don’t get paid and may be beaten to death by angry rebels who resent my role in their oppression.” This is especially true if you’re trying to recruit troops with an intense death-or-glory approach to combat, and who will be very unlikely to surrender.

          2) Soldiers from ethnic groups you’re not exploiting very hard because they have nothing worth taking, typically because they live in terrain too harsh to support more than subsistence agriculture, with few or no known natural resources. This is especially useful if you’re trying to recruit light infantry, because such populations typically live adjacent to among larger patches of wilderness area, and are more dependent on that wilderness for the means of survival. As such, they are often somewhat more experienced with things like hunting and moving quickly across rough country than their urban or dense-river-valley counterparts- skills that can be taught, but not easily.

          3) Soldiers recruited from traditional hereditary warrior castes, who typically occupy a privileged position in their home society and are partly insulated from your oppression. This is especially useful if you’re trying to recruit cavalry, because hereditary warrior caste members are much more likely to have the essential basic skills a cavalryman requires, skills which cannot be quickly or easily taught.

          Now, I could be wrong, but I think most of the “martial races” the British ‘found’ in India fall into one or more of these three categories.

          1. I think there was always more emphasis on ‘natural fitness’ in general public attitudes than there was in actual recruitment, but it’s a consistent problem in human history that you can do something for the wrong reason that still succeeds.
            People like to bring up the Gurkhas a lot, but what happened there is that the British fought and won a limited victory against Nepal, and decided in the process that the Nepalese (the citizens of a multi-ethnic polity) were a ‘martial race.’ The kings of Nepal then spent the entire rest of the span of the British Empire trading recruitment rights in exchange for independence.
            Gurkha military prowess does appear to be real, but as a combination of elite unit cohesion and the competitiveness of it being a job that offers England-cost-of-living pay to people raised in Nepal-cost-of-living homes.

          2. 1) The Gurkhas were definitely one of the groups I was thinking of when I mentioned category #1. Nepal was far enough outside the ‘core’ of the British Indian system that for practical purposes the Gurkhas were highly paid foreign mercenaries within India’s own context.

            2) Hypothesis: the actual Victorian age was a period where a lot of people were debilitated or scarred by endemic diseases or bouts of child malnutrition. This may have fueled a greater fixation on the idea of “natural fitness” as an essential quality of good soldiers, due to the large numbers of prospective European recruits who were physically unsuitable or inadequate in some way, or who just tended to die in huge numbers to tropical diseases Europeans lacked resistance to.

          3. I also think that there’s a self-fulfilling prophecy element as well, because warfare is highly cultural and done better by those with a professional warfare culture.

            Which group got hired to do the job originally might be shear historical accident, but if that policy is held in place for a generation or two a military culture can build up and the next generation is acculturated to be good soldiers in a way that’s self-reinforcing with time.

      3. It could often work, just not for the reason the British writers thought.
        As Prof Devereaux and Mary point out, morale has a big impact on the battlefield. Highlanders or Gurkhas who’ve been told they are “elite” or “martial” will have higher morale, hence more likely to win.
        Doesn’t take much for a positive reinforcement loop to be established. As Simon_Jester points out, there are economic reasons for soldiering to become a high status occupation for Highlanders and Gurkhas, so recruiting can be more selective. Inside the Highlander and Gurkha regiments there is peer pressure to train hard, fight bravely, otherwise live up to expectations. On the battlefield this translates into more victories. As word gets around, opponents of Highlanders or Gurkhas have lower morale, bringing even more victories…
        So Highlander and Gurkha etc regiments often really are the best in the British Army. Which unfortunately just confirms the original bias about ‘martial races’ existing.

        1. It is also possible, amplifying Simon Jester’s comment above, that Highlander young men were in fact fitter and healthier than urban young men. Cities were disease cesspools in the 19th century, with high rates of childhood disease and often malnutrition. Life in the highlands may be hard, but there’s no cholera. I don’t know enough to even speculate as to whether Gurkhas might be healthier than other potential South Asian recruits

          1. I recall some story, but can’t find it again, that for some 19th century British war (possibly the Boer War) the recruiters found that the average height of urban recruits had gone *down* from the previous war. This lead to some of the first laws designed to make the lives of the urban poor slightly less horrible.

  6. Was playing Simisle recently (1990’s game, vaguely like Tropico, simulates a third world indonesia-ish island), and it’s descriptions describe that issues of modernization pretty cynically, including some of the problems that can occur. Though in game most of these don’t show up, apart from pollution which is very obvious and unavoidable if you go for industry and cities. (Lots of dying trees makes such pollution very, very obvious). Like a lot of such games, however, the “game” part ends up pretty straightforward once you figure out some standard tricks, I’ve noticed this is a problem with a lot of games recently.

    I had been mulling around how to represent different types of governments, social systems, et. as a way to represent internal politics in a 4x type games, but sure enough actual game designers have thought of similar things already. Is certainly fun to read about, and the capitalist/price system not quite working right is something I’d expect. (Was tying to think about similar mechanics for a medieval/ancient setting game, to get good trade going, and figuring out a good way to do this without lots of iterations is…tricky.)

  7. Fascinating article. I never really got into the victoria games, preferring to stick with the Europa Universalis line, but this is really quite interesting and makes me think I was missing out. Oh well.

    1. I’m into these mechanics and this philosophy, but I was deeply allergic to the railroading of HOI2/3. Definitely makes me interested in Vicky3, but does not make me regret skipping Vicky2.

      1. Honestly, Vicky 2 wasn’t THAT railroady; it was nothing like Victoria 1 or EU2 or the other early games. It has a few hardcoded things and country-specific decisions, but stuff like that is still in EUIV (i.e., powerful formables like Prussia or the Mughals require specific countries/cultures to form, or Banner Infantry for the Manchus, etc.), just proportionally less. Paradox games were already moving towards sandbox at that point (and there was much hollering and screaming about “dumbing it down for the casuals”, or removing history from the games, of course).

  8. Hurray! Viccy 2!

    I’m looking forward to next week’s critique and analysis of warfare, because Victoria 2 can masterfully show both the fanciful glorious side of the one-sided colonial (and sometimes European) conflicts that led to a rise in militarism in Europe, but also the tremendous and horrifying cost of war. Just a while ago when playing a multiplayer game with friends (it used an alternate-history mod, but all the mechanics were exactly the same) Europe was consumed in the brutal ‘War of Military Access’ in the post-machine-gun era of the 1880s, and fought in *horrifyingly* bloody conditions for about a decade. Much of my nation spent the war occupied, in a desperate holding-action against superior numbers until aid could arrive, and although in the end my alliance won the war its scars crippled Aquitania (my nation).
    Over the war, the Aquitanian population *declined* by more than half a million pops-2 million humans, according to the population counter’s 1:4 ratio suggestion-in an era where technology and industrialization allows unprecedented pop growth. With casualties (many of these wounded and thus not removed from their pops, but still many dead!) in the millions, and with the catastrophic emigration caused by high war-exhaustion, the game managed to excellently simulate the post-WWI-national implosions caused by damaged economies, knee-capped demographics, and restless populations.
    The postwar economic collapse and unrest actually knocked Aquitania-the fourth-highest Great Power worldwide before the war-out of Great Power status, down to rank 9. A lost generation indeed!

  9. Great read! I’ve been looking forward to this since your EU4 series.

    Small note-the population number for each pop is supposed to represent only the adult male members of that groups, not women or their dependents. This is noted somewhere in the tutorial iirc, but not really anywhere else. Anyway, the displayed population is supposed to be 1/4th to 1/3rd of the actual total population.

  10. I’m not a Vicky player, but the game is a bit notorious for simultaneously being the only Paradox title named after a woman and abstracting all women out of the population numbers. This may be the main cause of the population discrepancy you note in Vienna, that it’s actually estimating an adult population of over half a million plus a roughly equal number of children; only the adult males in the pops are actually numerically represented in the game’s data and mechanics.
    Note also that Paradox basically never models an urb/hinterlands divide in its population simulations, so what you probably have is a little over a million people who live in the general environs of Vienna, an unmodeled percentage of which are urban within the real-world legal boundaries of 19th century Vienna.

  11. I strongly disagree with a lot of what’s written here. Especially regarding the evolution of Paradox’ design philosophy. What you describe as the generation 2 philosophy is in fact the generation 1 philosophy. EU2 was famous for having your country go bankrupt with an overflowing treasure because historically it did go bankrupt in that year, or an Austria that had been massively reduced by war losses being given dominion over an ascendant Hungary because that’s when Mohacs happened in real life. EU3 saw a massive shift away from this with historical monarchs being replaced by random ones, and historical generals and admirals being replaced by ones based on Army/Naval tradition (so no more planning wars around when you would get a good king or general), replacing cores only being given by events (which always coincided with historical conquests) with cores being given from owning land for 75 years, and removing a vast number of events based on what happened in a country in a given year and replacing them with context dependent events. This was massively controversial in the Paradox community at the time with an EU3 favoring side arguing that simulating history is better than retelling it and an EU2 favoring side arguing that Paradox had ripped history out of the game.

    If anything the change from what you call generation 2 to what you call generation 3 has seen a movement away from simulation, with Paradox favoring game-play focused abstract elements over historical simulation. This has again been very controversial, in EU4 to some degree and massively more so in Imperator: Rome. There has been enough pushback against this change that both games have been patched to be far more simulationist. In release EU4 for instance the rebel risk from newly conquered territory disappeared as soon as you cored the province by spending admin points.

    1. “Generation 1 was even worse about the trends you ascribe to generation 2″ does not mean the transition from generation 2 to 3 saw them moving even further from those trends.

      1. No it doesn’t inherently mean that, but it’s still the case. If anything Gen 3 has seen mild regression on the railroading trend, because it’s a lot easier to gamify railroading than it is to simulate it.

      2. Hmm. No. In principle, yes, but in practice the gen 1 -> gen 2 transition was *dramatically* more radical than anything that came later. There is considerably more continuity between the release version of EU3 and the current version of EU4 than there was between the final version of EU2 and EU3.

        I’m not even certain that the abandonment of historical constraint for maximal freedom of expression properly belongs to generation 3. It was something that developed over the course of that generation, mostly towards the end of it. And I’d say the trend only reaches its full expression with the absolutely bugfuck insane “build your own religion” system they created for CK3.

        (I think the gen 2 -> gen 3 shift is actually characterised by an increased focus on legibility, both in UI and in systems design, and a transition to the continuous development model that’s been such a bane and boon to Paradox’s games over the past decade.

        I don’t really credit Kzickas’ “abstract element vs historical simulation” model- both because I think the shift away from simulation has been driven by other factors and because, for as much discussion as they’ve generated, the abstract currencies they’re referring to were only ever a dominant design element in EU and Imperator- they’re not characteristic of the company’s output as a whole.)

        Look at us though, inventing a Paradox Games Periodisation Controversy, lmao.

        1. I guess if we’re doing this I should also note that there was a major overhaul of the game engine that happened during Imperator’s development (“Jomini”), I think in part to make it easier for them to reuse code between games?

          Whatever CK3 and V3 do new is going to be building on that, so there’s probably a case to be made that Imperator is a fourth generation game, at least technologically, even if it’s fairly conservative in its design.

        2. I didn’t specify abstract currency though. I think the HOI4 focus tree system is very much an example of abstracted gameplay over historical simulation.

  12. I would say that CK3 and Vicky3 are constituting a new Gen4 of paradox games purely on the massive improvements they are making in UI and general presentation of information to the player. The new tooltips-within-tooltips system is sorely missed when playing the gen3 games, even the more recent Stellaris. It doesn’t so much change their historical viewpoint but I think its similar to the introduction of Clausewitz in terms of technical/gameplay evolution.

  13. Bret, I’m not angry, I’m disappointed. “absolutely every human in VickyII has been numbered, categorized and grouped.” This is only true if women aren’t human. I expect better from the pedant who seems to go out of his way to highlight the oft overlooked people in history.

  14. Minor correction, you stated that the more prestige you have, the higher on the priority list your country is in purchasing goods on the world market. This is unfortunately incorrect, as the actual priority order comes from your nation standing as a power and what sphere you may belong to as a secondary or lower power

    1. Ah, I see that Podcat had to clarify this post-release. It was long assumed it was prestige order – I’ve been playing Vicky2 since release so I haven’t exactly checked the documentation in years. Will fix.

  15. One correction: Higher consciousness makes POPs more liberal and more socialist, not more likely to support their own interest. If you want to democratize a monarchy you want to raise the consciousness of your aristocrats since they will be the majority of your enfranchised population and raising their consciousness will make them more liberal and more supportive of extending franchise to the lower classes.

  16. I was wondering if if this series would convince me to re-download Vicky 2 and give it another shot. But it opened with a reminder that it’s from an earlier generation of Paradox games than the ones I enjoy, and in the middle it pointed out some of the jank in its core systems, so…um…I’ll end up sticking to Crusader Kings.

    I bring this up here because coming from the Gen3 Paradox games, I think it is easy to mistake the Gen2 railroading for very crass Eurocentrism when I think it is instead borne out of a design philosophy that held producing historical outcomes above providing players a high degree of flexibility.

    When it comes to analyzing a text, the only thing as important as the text is the con-text!

    …on starting a new session, the game has to ‘find’ equilibrium prices for all of the market goods, which typically takes several ‘turns’ (each turn is a day). For large states with big budgets and revenues, these bumps are minor, but small states can swing wildly between economic bursts and crashes in just those few days as prices wildly recalibrate (and as artisans wildly shift production in the background).

    That sounds quite unpleasant for anyone playing a small country, but it’s amusing to imagine what this would look like for people living in a Vicky2 world. Everything runs smoothly for years, and then for no reason prices go wild for maybe a week before returning to normal.

    Or, alternately, look at how the same transition is handled by the Civilization series: researching ‘industrialization’ enables various kinds of factories and power plants which create production (‘gears’) with no negative impacts at all.

    In some entries (like my personal favorite, Civ IV), you get a penalty to Health, but that basically just gives a penalty to population growth until you build enough hospitals or whatever to counteract it.

    1. Though on the “I don’t see this series convincing me that Vicky 2 is worth trying to get into again if it keeps going this way” note: Vicky’s pops are one of my favorite mechanics in any strategy game, period. For reasons that should be obvious if you read the blog.

    2. “Everything runs smoothly for years, and then for no reason prices go wild for maybe a week before returning to normal.” Substitute “for maybe a year or two” for a week and it looks very like the real world.

      1. Yeah, but that changing timescale makes all the difference.

        Victoria simulates events day by day throughout a roughly century-long historical timespan. Prices going crazy for a couple of weeks tends to cause automated systems in the game (like pops and AI-run states) to make decisions that don’t make sense even in the short term because they make irreversible binding choices like “build a factory” based on ephemeral nonsense.

        You get outcomes like, from what I’ve heard:

        “The price of luxury clothing just went through the roof and we don’t bother to remember what the price was a month ago, it’s high NOW dammit, so liquidate all my assets and plow them into a luxury clothing factory!”

        Real life recessions and wild price fluctuations usually happen slowly enough that people have sensible reasons to respond to them: “the market can stay irrational longer than you can stay solvent.” Here, the problem is purely artificial and not driven by market logic, and it would have no impact whatsoever if the AI didn’t have the attention span of a goldfish when it came to financial planning, investment, and stockpiling goods for a rainy day.

        1. True – but ‘the market’ is actually people making decisions. It’s not “the market can stay irrational longer than you can stay solvent” but “the irrational people in the market can persist in their irrationality longer than the rational can stay solvent”. There are plenty of people stuck with insane mortgages in Hong Kong or Japan, or who have never recovered from the various crashes (dot-com, GFC…).

  17. You mention problems in multicultural empires (like Austria) with rising nationalism leading to separatist uprisings. Is it possible within the game to create a more federal government structure in which, say, the UK creates separate parliaments for Scotland and Ireland while keeping them within the larger union?

    1. Austria can pass a decision to form austria-hungary, which reduces the militancy of the hungarians, but increases it of every non-hungarian minority. You can also release nations as puppets (vassals), these manage their own internal affairs, but you as their overlord handle international affairs. So they are forced to fight with you in wars, and cannot sign their own treaties. This is basically peaceful self-dissassembly though. POPs are power, and relinquishing (partial) control over them weakens you.

      1. It’s also possible to transition into a multi-ethnic democracy (constitutional monarch or no) as a way of easing separatist tensions. Political reforms tamp down on militancy, which in turn reduces the nationalist impulses of non-primary culture pops. This Austria game I’ve been using screenshots for, for instance, is going to end up with an absolutely mega-sized ‘Germany’ that includes all of Austria’s starting territory and has a bewildering array of cultures in it, but because I end up enacting a lot of social and political reforms, militancy remains very manageable.

        But that’s a tough tightrope to walk, especially for new players.

  18. > [edit: technically the game simulates adult males and statistically assumes women and children, as some commenters have pointed out; that’s an awkward choice and I hope it is changed in Vicky3]

    Good news! That’s exactly what they are planning per their dev diary on pops. Pops are now divided into “Workforce” and “Dependents”. Dependents are those people who have virtually no political power of their own and can only engage in limited economic activity, representing stuff like pensions and odd jobs (including child labor and minor home business stuff that women frequently did while their husbands were out) that they might engage in. But laws later on in the game can permit certain kinds of dependents, like women, to move from the dependent category to the workforce category (and presumably, children will do so well as they grow up).

  19. “the last bizarrely named ‘anarcho-liberals’ but who clearly represent the anarchist bomb-throwers of the period”

    Small correction about this one: Anarcho-liberals seem to be – for some anachronistic reason – modelled on anarcho-capitalists, not actual anarchist, as far I’m aware.

    1. I played Victoria and not Victoria II, but from that game I gathered that you had three political ideologies to choose from Conservative, Socialist, and Liberal. The three ideologies thus named are the ‘moderate’ versions of those ideologies. Then you have the radical versions of those ideologies; Reactionary, Communist, and Anarcho-Liberal.
      Reading the Vicky 2 wiki, it looks like the Anarcho-Liberal parties are supposed to get started around 1848, which is when you see lots of radicalized liberals in real life

      Besides, it’s the 19th century, someone has to be bomb-throwing anarchists. While your typical bomb-throwing anarchist would have been more socialist in real life, in Vicky the radicalized socialists already have the job of becoming communists, so we have to let the radicalized anti-government liberals try.

      This might be an example of jank, just in the political realm. The game needs radicalized liberals to go with radicalized socialists and radicalized conservatives.

      1. Gotta love Viccy 2, the game where an actual invention (under one of the Culture late-game techs; Revolution and Counterrevolution, if I recall correctly) is ‘Unlock Anarchist Bomb-throwers’

  20. Looking forward to this series. I got into Paradox games with EU 4 and have played all the “Gen 3” games (barring CK 3) to a greater or lesser extent, but decided to hold off on Victoria 2 despite it always sounding really great. Partly this was because of a vague feeling of what you’ve crystallized here that it was from a “different era” of game design choices, partly because it didn’t have a Linux native version, and partly because I was expecting Vicky 3 to come along a few years later…which turned into several years later…which turned into—well anyway I’m glad it’s finally coming and thank you for getting Paradox to finally release it! I’m looking forward to playing it on launch, and reading this series to get an idea of the themes and ideas that will probably underlie it even if the specific mechanics are a bit different. 🙂

  21. Speaking to the software bits:

    That instability on loading the simulation is very familiar to me from physics-simulating games. e.g. in Kerbal Space Program (KSP), g-forces spike and craft bounce all over the place on switching to a craft. They mitigate the effects by running the simulation in a sort of “safe mode” for the first several simulation ticks: limiting forces and accelerations, special-case code to place landed craft safely onto a planetary surface, turning off some damage effects, &c. This is a bit tricky because at some point you have to turn off safe mode, so your safe mode still has to bring the system closer to the regular-mode equilibrium, and KSP still sometimes has trouble on switching to a craft.

    Generally this results from a differences between the runtime representation of game state and the save file representation. In specific terms, this can be:

    1. Use of different precisions or coordinate systems.

    Example: in KSP, objects are represented in save files with their orbital elements for ease of background “inactive” simulation, while when they are loaded they are represented in a Euclidean 3D coordinate system.

    Problem: When you save the current state of the system and reload it, rounding errors and conversions mean that the loaded state is not the same as it was when you saved. This means that the system is every so slightly away from equilibrium. In the Vicky2 case, said equilibrium would be the short-term equilibrium prices determined by the “real” production factors like pops and factories (which Vicky2 presumably saves precisely).

    2. Dropping “unnecessary” information in the transition from runtime to save states that turns out to actually be important.

    Example: in KSP, the relative positions of parts to each other are dropped in the save files, as are rotation rates. Crafts’ structure is represented in the “ideal” form of part connections to each other.

    Problem: The system may not be stable with the placeholder “ideal” values that are inserted to substitute for the dropped state. Alternately, the dropped state may include some hacky hysteresis (path-dependence, for history/econ people) inserted by developers to stabilize a system that would otherwise be prone to oscillations.

    In both of these cases, you can consider a load event to produce a transition from a pseudo-equilibrium (the save state) to the actual equilibrium in a single simulation tick, causing change rates to go to infinity and all kinds of funky stuff to happen. In Vicky2 terms as you’ve described them above, the disruption is on the same scale as tons of pops all over the world suddenly starting work at their factories in a single time tick.

    1. Sounds like a computer version of Walras’ invisible auctioneer. When economists were trying to envisage the price system, the best they could imagine was an open auction for all conceivable goods conducted simultaneously.

      1. It’s not really a computer thing, more a maths thing (if we’re talking about computers in terms of the physical device, rather than a thing or person that calculates). If you want to calculate some quantity, you don’t have infinite precision (the universe is finite ;)), and similarly, you typically need to work in discrete rather than continuous terms. All of this falls out the maths (no assumptions needed), and while this is usually comes up when using computers, much of the basic theory pre-dates the transistor.

        1. In this case though the original system is simulated on a computer which has exactly the same precision problems, so it’s possible to save and restore the state without precision loss.

          The disruption of the model on loading just shows that they haven’t saved/restored enough state. It’s not a fundamental problem in this case, just an implementation bug.

      2. Yes, and that is probably a good model for this setting.

        Problems come from implementation details, in this case surrounding serialization/deserialization of state.

    2. I suspect the hysteresis is the problem, as I can see the game engine forcing the use of non-ideal numerical schemes (in the quality sense), especially if certain subsystems have been bolted onto the main engine.

      A third option (similar to the first, but with a different origin) is balancing changes modifying internal parameters of the simulation (e.g. the effectiveness of a certain technology going from 2x to 3x), and so loading a game that started with an older set of values would need some time to switch to the new equilibrium.

      1. Given that this is such a reproducible and common issue, I doubt it’s about changes in game version.

    3. I always wondered if Vic II economic hiccups could be read as practical demonstrations about numerical practicalities about dynamic system simulation.

      1. Sure! Physics simulations like KSP provide more visceral and intuitive and entertaining demonstrations, but a Vicky2-style simulation provides easier code for a student to understand.

  22. I really like these Paradox writeups! I’m a big fan of their games and their attempts to be realistic (or historical for the earlier games as you mention). That realism makes them a good springboard to talk about the historical factors that the gams are or are not modeling well. At the very least, it makes it more interesting than using, say, Risk and talking about how well it models Napoleonic Era warfare.

  23. Would it be possible to link to bigger versions of the screenshots? A lot of the stuff you mention is basically unreadable.

  24. I’m not a gamer but I do like to play with alternate histories so these sound interesting. But like most people with that hobby I tend to focus on personalities.
    The Great Man (Woman) theory of history is inherently flawed and rightly derided but the Blind Historical Forces and what Bret calls the Whig theory of Historical Inevitability is no better. The truth seems to be a combination of all three. Individuals impact history by their choices but what choices and how much impact is limited by the sociological and economic background.

    Napoleon I, a man with a massive historical imprint, himself stated he was a creature of his times. In the famous and cliched case of Hitler the economic and social conditions of Germany were ripe for a totalitarian ideologue and Hitler didn’t invent fascism, or anti-semitism. Another man might have been less expansionist and less genocidal granted but a second war could still have happened. Of course it might have been against the Soviet Union instead.

    1. Crusader Kings is the Paradox franchise that focuses on specific individuals driving historical change. Victoria seems to go more for a Marxist historiographical approach, where economic forces and class conflict are the cause of that change. And our host already established Europa Universalis’ state-centered and IR realist approach.

      I appreciate the diversity, but I haven’t actually been able to get into any Paradox franchise other than CK. Stories are what keep me engaged, and CK’s character-driven and better at emergent narratives than any other game I’ve come across. Victoria 3 might change that; time will tell.

    2. The Paradox Grand Strategy games as a full series present an unbroken alternate history creator (for Europe at least) from the post-Alexander Greek wars to ~3000 CE. In theory, all incorrect models of how history happens have time periods where the model has explanatory power, that’s how we discover incorrect ideas in the first place. So Paradox tries to roughly match models to the portions of history where it looks like the models work, and changes to a different game with a different title and often very different mechanics, but the design question (when they create a successful game) is ‘can mechanics derived from this theory move our simulation from one actual historic shape to another?’
      One of their most recent games is also the earliest in the timeline, Imperator, is generally regarded as a failure, because it tries to isolate out the ‘correct’ bits of multiple theories to inspire its abstractions and mechanics; the resulting mess can only please people who don’t have a personal favorite theory of history (and that is an incredibly small slice of the strategy gamer pie!).
      The next game in the timeline is Crusader Kings. CK (2 and 3) start with some people in tribal societies and some people with relatively sophisticated imperial governments, but everywhere people are in some form of feudal or para-feudal relationship, and all of the general population is abstracted out. As such, it leans so hard into the Great Man theory that it’s primarily a human breeding simulator. There are mass battles and conquests, but everything that happens is heavily modified by the individual skills of leaders and their subleaders; the game owes far, far more to Star Stable than to Risk.
      One of the first events you can effect in Europa Universalis 4 is to hold the last jousting tournament, but you already know all about how the player makes that decision as a state not a character because I saw you in the comments on those posts…
      Victoria really does lean into Whig History, but the version of it where real-world North Atlantic laissez-faire representative democracy is the pinnacle everything is moving toward (China, yes China, begins the game as an “uncivilized nation” because it is so far away from that pinnacle).
      Hearts of Iron only covers a few years, from the end of the European-led Great Powers system to the establishment of Mutually Assured Destruction, but notionally we in the real world are still in an era best modeled by it’s late-game mechanics and will continue to be until the discovery of faster-than-light space travel. HoI presents as inevitable the consolidation of the world into a single multinational alliance, and that military conquest will be the most important element of that consolidation.
      Stellaris is mostly Star Trek DS9, really: it references a lot of sci-fi, but the text of it’s theory of history is Deep Space 9

  25. “It’s not clear to me how, by 1836, the genie-in-the-steam-engine could be put back in its cylinder, short of some sort of vast catastrophe consuming much of the globe. Consequently, I am inclined to pardon VickyII a lot of its teleological nature”

    While it’s very easy to look at our current world and believe this technological teleology, I believe it’s pretty clear that it’s false.

    My rationale? There is a parallel universe to ours where someone is writing “it’s not clear to me how, by 1980, the genie-in-the-nuclear-reactor could be put back in the bottle. Even ignoring the massive environmental benefits, the energy density is just so much higher that it makes no economic sense to do anything else.”

    Yet that teleology is clearly false. Our universe provides the counterexample. Except for France and maybe Iran, the world just sort of decided nuclear energy was icky and didn’t do it.

    There are other examples of this – dismantling the Chinese treasure fleets, for example.

    1. I’m not sure the Chinese treasure fleets are the best examples, since what I’ve read on the subject suggested that they were not economically sound, but rather showy and wasteful displays that brought little material benefit to an already wealthy and hegemonic empire. I think trying to look at them as somehow equivalent to European voyages of exploration is exactly backwards: the Chinese sent out treasure fleets because they were rich, while Europeans sent out caravels because they were (relatively) poor.

      1. Europe was trying to get to China, and for a long time didn’t have much to offer other than stolen American silver, or outright piracy (Portuguese in the Indian ocean.)

        1400s China was already China, already in a trade network with a majority of the world’s population and a supermajority of the world’s production.

    2. Nuclear power runs into issues because lots of other tricks exist to do almost the same thing, but cheaper, and if nuclear power develops more we still live in about the same world, just with possibly more radiation and less chemical pollution. (For nuclear power to be cheap enough to take over more electric production, some combination of less/different safety regulations and/or development of better reactor designs, needs to occur. Based on how such development is going today, it’s hard to say how likely this was in the mid 1900’s.)

      Steam engines are different in that they do lots of stuff that other techniques just can’t do. Trains are a big one, but factories/mills away from water ae also unique. This means that steam power will continue to be developed, and at some point will start to outcompete existing technology. It’s not certain that steam power is almost guaranteed to be used (maybe people react strongly to coal pollution and smoke, though even than filters might get common and the technology still develops), but steam and industrialization spreading is a safe guess.

      1. Nuclear power could have made under sea facilities feasible, and facilitated an expanded space program. Although even with more research it p still probably wasn’t likely.

        1. Nuclear power can do that now (since it physically works, just isn’t as cheap as other sources), it’s likely that undersea stuff and the space stuff that would use it just isn’t worth enough to use it. (More common nuclear power might mean more research, and better plants, that allow a few such applications, or maybe less resistance to it would do the same, but the cost of nuclear power is just one part of the cost of such facilities.)

      2. For a long time, steam power *was* in the situation nuclear power is now – it was considered Very Unsafe (because boilers had a nasty tendency to explode due to insufficient appreciation of the dangers involved and not having developed now-necessary safety features)

        We are more risk-adverse now, and the dangers of a nuclear power plant “exploding” are worse than that of a (conventional*) steam boiler. The main reason that nuclear plants are not cost-competitive with other sources of power is that we have chosen to over-engineer and over-regulate for safety while we figure out how to do it cheaper, instead of accepting the occasional blowup as the cost of doing business.

        “Any day now” we will get the cheaper-and-safer nuke plants (scare quotes because, while the safety features are subject to being engineered cheaper and more effective, the safety regs are subject to the political process, which is not rational.)

        (*: Conventional, because a nuclear power plant is, of course, a steam boiler, just one that gets its heat from nuclear reactions, not chemical ones)

        1. “we have chosen to over-engineer and over-regulate for safety while we figure out how to do it cheaper, instead of accepting the occasional blowup as the cost of doing business.”

          Also that we *don’t* hold fossil fuels to the same standard of “must contain all waste from dispersing into the environment”. If we did, they’d basically be unusable and nuclear would be dominant. We’d be poorer in market goods, but OTOH not facing imminent global warming crises.

          1. I’ve seen it argued (and been convinced) that, just during the period of nuclear power plants, coal plants/i> have released more radioactive waste into the atmosphere due to naturally-occurring radioactives in coal than all the nuclear incidents in the same time.

    3. I don’t think Bret is generalizing this to other periods; he explicitly says that technological determinism isn’t an enormous factor in EU4, and he didn’t criticize it much for that.

      The period Vicky2 models happens to be one where rapid technological change altered societies and politics more profoundly than any time before or since.

    4. Steam was powering river boats in the Mississippi and across the Atlantic by 1836. It was much more entrenched than nuclear power in 1980 which already was having pushback from the Three mile island accident.

      If battery technology developed further instead of steam power then different provinces might become important, and different states might become powerful but the dynamics of artisans being replaces by industry would very likely still take place.

    5. Well, the problem with nuclear power is that the risks outweigh the benefits in most situations. We’ve been using nuclear power for only about 75 years, and we already had one catastrophic misshap that almost destroyed Europe, were it not for the Soviet’s speedy crisis management. What we are beginning to realize now is that nuclear energy should have a big “no-touchy” sign written all over it. Looking at man-made climate change, we will probably soon have to face the truth that the steam- and combustion-engines are not worth the risk of using them any longer.

      1. On the contrary, nuclear is unquestionably the safest form of energy known to man. Solar kills more people merely from people who die in falls while installing panels. Furthermore, nuclear is carbon neutral.

          1. It’s an illusion of scale. If we were building and installing attic-sized nuclear reactors at the rate that we’re installing rooftop solar, you could expect to see a lot more nuclear construction deaths.
            It’s also the case that the union can much more easily establish safety standards on a nuclear construction site than for home contracting, for a variety of reasons.

          2. So your argument is the real death toll is an illusion? The safer one is the one that kills fewer people.

      2. “We’ve been using nuclear power for only about 75 years, and we already had one catastrophic misshap that almost destroyed Europe, were it not for the Soviet’s speedy crisis management.”

        More accurately, the Soviets’ terrible management practices created the problem in the first place, were largely responsible for why the problem was so bad as to be noticeable, and said problem was only really catastrophic for the area immediately surrounding the reactor.

        Sorry, but nuclear power is ridiculously safe.

    6. Steam engine gives you heat engines, a way to convert (some) heat into work. This is a brand new capability that has never existed before. Combined with abundant fossil fuels as a heat source, it gives a huge advantage. Even if someone is more sensitive to pollution problems or to Arrhenius’s 1896 prediction global warming, I don’t think competitive countries can afford to dispense with it. At best they could jump heavily into nuclear power ASAP… *which still uses steam engines*: a nuclear reactor is simply a steam engine that uses fission instead of burning coal to generate heat.

      Oh, and modified steam engines in reverse give you true refrigeration (rather than just harvesting ice), another huge leap in capability (meaning not just more comfort and chilled food, but industrial stuff like liquefied gases.)

      And with a start date of 1836, coal+steam is already in commercial use: steamships for some time, first intercity rail line in 1830.

      “Except for France and maybe Iran, the world just sort of decided nuclear energy was icky and didn’t do it.”

      And France doesn’t have some huge economic (dis)advantage from being nuclear. It’s cleaner, and the world would be better off if more of us had been like France, but it hasn’t been some killer difference, compared to a steam navy vs. a sail navy, or steam railroads vs. horse transport.

  26. Good morning, Bret! I’ve never played and am not likely to play any of the games you write about, butI am happy to read your posts!
    Here are a few proofreading corrections I noticed today:
    tend to designed so that -> tend to be
    Caption to Another set of pops: one that really efforts to use pops [not sure what word efforts was intended to be?]
    engines ended up power Wilkinson’s -> powering
    VickyII let’s the player -> [delete apostrophe]
    smaller, non-western states -> non-Western
    the western/non-western dichotomy -> Western/non-Western [Bret: capitalization of references to the West has been your style in previous posts]
    like nation-wide railroad -> nationwide
    Caption for late game Ottoman: screen. Note only -> Not
    under ‘Planned Economy’…completely removes -> [delete under?]
    major arm’s producer -> [delete apostrophe]
    well industrialized -> [insert hyphen]
    is a mine-field -> minefield
    cross-pollination of technology -> cross-pollinations
    reforms to the former and political reforms the latter -> [is there a missing to in this?]
    the cost is nearly all -> in nearly
    meaning an decisive -> a decisive
    short its time-scale is -> timescale
    the crowing achievement -> crowning

  27. Will you cover plurality and the other effects of consciousness in a later post? It’s not absolutely needed but I find the interplay between consciousness, plurality, and militarism to be one of the most fun things about the game.

    For the curious, as consciousness rises, it acts as a multiplier for the needs of the population. Therefore the demand gets higher and prices rise. This makes sense. A literate middle class man has higher QoL expectations than some some equally wealthy illiterate individual simply because he is aware of the wealth around him in his nation. This means that more liberalized and open nations will have more demand and higher prices than authoritarian ones. Very interesting.

    Plurality is essentially your population’s zeitgeist. It increases your consciousness if your people are content under your rule, otherwise it increases militancy. It basically is their response to how you dealt with this system in the past and makes middle of the road solutions more difficult.

    Again, awesome system.

  28. Hmm. That to me seems to be more an attempt to recreate the old system of specific historical events in a way that makes them more legible- no more trawling through the event files to find the precise sequence of magic gestures you need to perform to get a PU on Aragon, it’s all laid out for you in the tree.

    I think I can see your point though- they’re privileging gameplay over simulation by putting the player in charge of what happens when, rather than presenting them with situations to which they have to react.

  29. For those interested in computer game design and AI in particular, I recently found a discussion with two game designers/programmers from SSG, a company known for their good AI opponents. Interesting background on not just AI implementation, but meta issues surrounding game design and player reactions to AI.

    I’m never sure how to put URLs into a comment, so if there is no link, search for
    Space Game Junkie Podcast SSG Roger Keating

  30. Spheres of influence don’t actually share markets- what happens is that any excess goods are duplicated across the national markets of all sphere members. So if Belgium is in the UK sphere, produces 10 iron, and uses 5, then the UK and all other members of the UK sphere will have 5 iron added to their stockpile while the excess Belgium produces is deleted (rather than sold on the world market). Because of this spheres can massively increase the amount of raw resources available for anyone not top GP (industrialization late game often outpaces raw resource) but being a part of a sphere reduces income because you’re not getting export tariffs.

    Militancy is actually mostly a good thing, mechanically, if you know how to handle it. The healthcare and education reforms are incredibly powerful if you can get them early, and you need at least 5 but usually 6-7 militancy to get them before socialism shows up. Usually what you want to do is spike militancy as much as possible with events, holding elections, and changing ruling party to get a few levels of those reforms out early on. Militancy is only bad when you can’t get it above the threshold to get reforms or you start getting large rebellions. But if you can spike militancy fast enough you won’t get too many rebellions before reforms push it back down to lower levels.

    Anarcho-liberals don’t represent 19th century anarchist movements, they represent right-libertarianism from the 1950s onwards. It’s anachronistic, but when they take power they implement “bourgeois dictatorship” government so it’s pretty clear what the intent is. Vic2 doesn’t actually have period-appropriate anarchist movements.

    I don’t think Vic2 mechanically trends towards social democracies, fascism and communism are very powerful government types because of the military bonuses, ability to build optimal factories, and ability to pass social reforms (the AI also tends to collapse into repeated revolution and counterrevolution between those two and anarcho-liberals, the world doesn’t end up looking very democratic). Pure democracies are outright worse than constitutional monarchies. Late game there are a few viable directions to take a state.

  31. “It’s an illusion of scale. If we were building and installing attic-sized nuclear reactors at the rate that we’re installing rooftop solar, you could expect to see a lot more nuclear construction deaths.”

    No. The deaths are given per terawatt-hour, so it’s already scaled.

    Estimates vary: ourworldindata has 0.02 for solar and 0.07 for nuclear and just 24 for coal. But both sets agree that nuclear and renewables are far safer than any alternatives. Search “deaths per gigawatt” or stuff like that.

  32. “A player that relentlessly focuses on short-term state power at the expense of their pops will ruin their nation in the long run.”
    Depends on the country you play. If you play a country with lots of pops like China (I’ve played this way as China, ended up as great power #1 by a long distance), or an extremely powerful country like UK (I haven’t played UK but I assume), you can get away with killing millions of your own rebel pops because you have so many other strengths.

Leave a Reply