The Digital Brittleness of Victoria 3

In 1994, John Meriweather founded Long Term Capital Management on the basis of a computer program that looked for discrepancies in expected value of various speculative and financial instruments. In that first year, they made a 21% return, rocketing up to the low 40s for the next two years. Their secret sauce was that their trades were entirely guided by a computer program performing quantitative analysis on the market. They had rendered the chaotic swirl of the stock market down to hard physical bits in a computer program; bet on this, hedge that. Certainty crystalized. But, it was brittle. The fund, bolstered by such success, borrowed increasing amounts of money in order to make bigger investments. After all, they were nearly always right, so more risk maximized the gains. They had started with $1 billion, but four years later in 1998, the fund had $129 billion in assets, $125 billion of which was borrowed. Much of this was tied up in trades in Asian and Russian markets, a seemingly savvy bet by the program that bolstered Meriweather’s confidence. While the algorithm might have been right, it could not account for human factors: fear, uncertainty, and doubt. When dual economic crises hit Russia and Asia in 1998 and spooked investors, LTCM lost all their investments and would end up getting recapitalized by the Federal Reserve, setting a precedent for the bailouts of the Great Recession eight years later. Meriweather’s fantasy was having a computer program that could make the market perfectly understandable, but LTCM’s downfall was their own human factor. They were so confident in their program’s predictions that they borrowed way more money than they should have, in order to maximize their gains. When the bet was wrong, they lost it all. In an economic system which is ever more biased toward those who accumulate most of its wealth, which requires us to exchange ever more of our limited time on this planet for things which were once thought to be middle class table stakes, like home ownership and parenthood, the idea of making a hedge against conventional wisdom and profiting so massively we unshackle ourselves from financial dependence is a seductive one. The impulse to escape the Matrix is one on which a hundred crypto scams prey. I too sometimes wish to live in this fantasy world of information perfection, but I am too aware of the instability of the real world to think it exists here. Fortunately, I can visit it in a video game.

I have a lot of hours in Paradox Interactive’s suite of grand strategy games, going back to playing Hearts of Iron 2 in high school. Crammed into my friend Boris’ bedroom in his family’s 1-bed apartment (his parents slept in the common room), we played an unintended, non-diegetic minigame we called the “connection minigame”; restarting our game clients until the checksums matched so we could connect to the same local multiplayer server. It was 2005, and our toolkit for getting into a game together included Hamachi VPNs and a confusing tangle of Ethernet cables. Paradox’s multiplayer netcode has come a long way since then, and today we seamlessly join games of Stellaris and Crusader Kings 3 together from across the country instead of huddled around a LAN switch in a bedroom. So, like many grand strategy aficionados, I gave Victoria 2 a try years ago and, like many others, failed to wrap my head around its simulation of the Industrial Revolution, third-wave colonialism, and Great War realpolitik. I found the mechanics alternatively too nuanced, opaque, or fiddly to enjoy tinkering with. Paradox games get slightly slicker graphics over time but look basically the same as their earlier renditions, and playing them reliably transports me back to that apartment and the smell of strong Russian Caravan tea.

Victoria 3, upon its release in 2022– a lengthy 12 years after its predecessor–was clearly geared toward making the most notoriously complex member of its repertoire more accessible. The Victoria series has always been touted as the least warlike of the Paradox grand strategy games, its focus instead being on political and market control over physical conquest of territory. Luckily for the warlords among us, It is still possible to conquer the world; even trivial once you find the right strategy.

Playing Victoria 3 is like playing in the world that LTCM’s computer program thought existed. Victoria 3, like all “grand strategy” games of its ilk commonly published by Paradox Interactive, renders the uncertainty and ambiguity of human things–massive economic and political trends such as industrialization, revolution, and colonialism–into the concrete physics of bits in a computer system, thus converting them into game mechanics with crystalline transparency and deterministic brittleness. Because of the complexity and quantity of interlinking mechanics in this genre of game, there are often ways to supercharge your playstyle and far exceed the AI’s ability to compete with you. Often, this means hyperspecializing to gain a large advantage in one area, then leveraging that advantage to outcompete across all areas. In most of these games, going all-in on research is a strong path to success, as tanks will generally win against calvary (though don’t tell the Polish cavalry, who dispersed German tanks at Mokra).

However, as Victoria is an economics and politics simulator foremost, and war is diplomacy by other means, optimization also means building the strongest economic base. In the Industrial Revolution period depicted in Victoria, that means moving from agrarian systems where land is the main store of value, to industrialized societies where capital is the store of value. A savvy player therefore knows to focus on developing their country’s wood, coal, and iron, even at the expense of domestic food production–it can always be imported, after all. Developing these industries also empowers capitalists, who will eventually supplant the conservative landowning aristocracy, and allow more progressive trade laws to be passed. Indeed, it is almost always best to launch into no-holds-barred laissez-faire capitalism, robber barony and all, just to strip the landowners of power as quickly as possible. Pesky things like worker’s rights, minimum wage, and child labor laws can wait until later, and you won’t have the political clout to get those laws passed anyway. The reason to rapidly capitalize on your natural resources is not to build furniture or clothes factories to give your digital citizens cheap goods, though that may feel like what a benevolent leader would do, but rather to engage the ouroboros of the construction sector. Each construction sector increases your ability to build. Build more tool and steel factories, so that you may supply more construction sectors. Use the increased construction to build more factories to supply more construction. Spend as much money as you can to fund more construction, so you can continue to construct new construction sectors. Even if you have to go into debt, so long as your gross domestic product grows faster than your interest rate, you can keep borrowing money. This is the dream of LTCM’s quantitative analysis, a little white line that never, ever has to stop going up.

Pulling this off requires a deep understanding of the many mathematical modifiers present in Paradox games, which are additive and which are multiplicative, and which to prioritize for a given strategy. The goal is to stack marginal gains for maximum utility. Because increasing construction capacity is always the best use of your existing construction capacity, any bonus that benefits construction or its input goods is preferred. Generalist Gaming on YouTube has done exhaustive research on these subjects vis-a-vis Victoria 3, crafting spreadsheets of the most optimal goods to produce, and the most optimal places to conquer. Players often refer to this kind of optimized play as “breaking” the game, because while it usually doesn’t require exploiting actual bugs in the game code, once the player is sufficiently powerful the AI can no longer provide any meaningful resistance to whatever the player’s whims may be. Thus, the game has been “broken,” and when I use the word I am referring to this state of player dominance over the computational world.

The third major DLC which further expands the Victoria world, Spheres of Influence, came out in the summer of last year, followed by the India-centric Pivot of Empire. I haven’t explored this latest expansion yet, which means I’m about to dabble again in what will certainly feel familiar to the game I previously knew as Victoria 3, and yet substantially different. This is how Paradox likes to develop and monetize its grand strategy games: release a game featuring a complex series of interlinking mechanics out of the box, then, through DLCs, add additional major mechanics to the mix. At a certain point, it’s impossible to balance all of these features against each other, so there almost certainly arises ways to unduly break the game in the player’s favor. Nevertheless, a steady stream of patches concurrently works to rebalance the game and close loopholes. Step away for a lengthy period of time, and returning to any of these games is an exercise in returning to the planet of the apes. Achingly familiar, but everything works a little differently than you remember. Old strategies will not be as effective, but the game is not any less breakable than before–the locks have just been changed.

Victoria 3 almost takes a masochistic pleasure in being broken by the player. There are achievements for accomplishments that would be impossible without a deep understanding of how to exploit the game mechanics, such as the minutely complex series of decisions required to take Poland from the tiny rump state of Krakow and re-establishing the sprawling Polish-Lithuanian Commonwealth; or subjugating the United States after starting as its underpowered subject, the Indian Territory, which requires tricking the game’s AI into fighting itself. YouTube and Reddit are full of threads and videos identifying key strategies for maximizing early game power, and soon after launch this collaboration caused a series of key moves to cohere, regardless of what country you’re playing: subjugate or conquer territories in South Africa and Borneo–where gold is sure to be discovered–for low infamy costs due to low population. If strong enough, take Beijing from China for its workforce, unique regional coal and rail company, and the Forbidden Palace, which grants its holder increased government legitimacy. Use an exploit (since patched) by way of the Opium Wars to force Great Britain to grant independence to the East India Company, thereby destabilizing the British Empire. After following this strategy, the player is perfectly set up to accomplish any desired goals essentially unopposed.

In the last major patch which brought new content to the Indian subcontinent, a number of powerful opening moves have been rebalanced. The previous addition of Agitators made it possible to turn any serf-based nation on the periphery into a free trade paradise within a few years of the opening date. Now, the new Political Movements mechanic adds a speedbump. An earlier patch made it less attractive to outsource all your manufacturing industry to an unincorporated outpost in a high population Indian or Chinese state. But, Power Blocs were added which grant powerful additional bonuses based on their type. Whenever something changes, a tooltip is happy to appear and explain in precise detail how the inner workings of the world now operate so it can be understood, then utilized for maximum advantage.

The push and pull between min-maxing players and busy developers reveals an internal tension in the development cycles of Paradox’s grand strategy games. The developers know their most passionate players are going to pick apart their game and figure out the best strategy, and they don’t want to alienate those loyalists. However, repetitively executing the same best strategies detracts from their vision of making a game where you can play as any nation that existed in 1836–since what would the point be if they all played the same?

And yet, I know that there are 12 gold mines seeded in Transvaal, 16 in neighboring Vrystaat, and they will randomly appear within the first decades of game time, providing a reliable injection of funds into my economy. I know that Beijing’s Forbidden Palace grants a precious +20 Legitimacy to my government; I can hover my mouse over it and get a tooltip saying so. The fog of war that blankets the real world and obscures perfect knowledge, here, is stripped away. Presented with such information, it is hard to resist the satisfaction of playing the game “the best”, executing the optimal strategy based on perfect information, like a computer program making trades in a hedge fund office in 1994. When Paradox inevitably changes something, there will still be a tooltip under my hovering mouse to tell me exactly how it works, now.

To look upon the world as a math problem that can be solved is the fantasy that Victoria 3 provides, just as John Meriweather once hoped the real world market could be solved by an algorithm. In an uncertain world, this is what keeps me coming back to the 1836 start date again and again, to perform a reassuring ritual of the opening moves in a world of known quantities and perfect information, a timeline which has already played out one way in our world but might play out differently this time.

Since the last time I played through a game of Victoria 3, I have become a father. Of course, becoming a parent recontextualises all lived experience in the way that having a little bit of puke on you all the time only can, and being a father is uncertainty incarnate. The way my child might take a midnight bottle peacefully one time will be met with flailing limbs and tears the next. Just as I think I’ve figured out the crescendo of their flatulence that indicates it’s safe to change their diaper, they’ll remind me of the accompanying water fountain show. It’s humbling, and funny. Harry Baker captured it best and most lovingly in a poem for his child:

Babies cannot be optimized, though that does not stop everyone from having an opinion on how best to raise your child. Often, they have a product to sell you. But for me, who is soothed by rote repetition, playing this game is a way to bring the comfort of certainty, something that can be done perfectly, with precision, to recapture a little island of control in the eye of the storm. So I’ll return reliably to my maps, charts, and spreadsheets after the next DLC is released, to see what has been changed. I’ll read Reddit and watch YouTube videos explaining the minutiae of a world that looks like ours but is remarkably different because I can always, always, place the right bet.