Let me tell you a hilarious story, then later get technical about why it happens.
The Civilization series of games gives you control of a civilization and asks you to “win” history. You can win by conquering the world, or by having your civilization elected supreme leader, or my researching enough technology to escape the cradle of earth and go out to colonize the galaxy.
But fundamentally civilization is about cities. Cities are where everything happens, you build your military in cities, you get money from cities, you get research from them, your civilization is nothing without its cities, and when your last city is lost, you are defeated.
It makes sense then that you want to always have *more* cities so you can have *more* stuff. Two cities give you twice as much of everything as just one, a third city upgrades you 50% from two and so forth. The Civ games have tried to put limits on “infinite city spam,” but generally *more* cities is always better than *less*.
That’s why the One City Challenge is such a challenge. The One City Challenge is a longstanding challenge for Civilization veterans, demanding you win the game using *only one city*. This means staying unconquered long enough to either diplomacy yourself into the World King, or research your way into galactic colonization.
But the One City Challenge is nearly impossible when you’re up against AIs building as many cities as they can. I’ve never beaten the One City Challenge, and most who do beat it do so on the lowest difficulties. Beating the One City Challenge on Deity (the hardest difficulty in the game) is only for Civ Masters with a *lot* of luck on their side.
But Civ VI introduced something new, wonderful, and stupid. Civ VI introduced the No City Challenge, and it’s doable on Deity.
See in Civ VI, the Maori civilization starts with the ability to sail the oceans, and their starting settler and warrior both begin in the ocean. It’s easy enough to send the settler and warrior way down to the artic ice caps and hide in the ocean forever, never meeting or even interacting with any other Civs (because who would explore the desolate ice caps in this game?). Now you’re playing the “No City Challenge,” an attempt to win the game while hiding in the ice caps and never even settling a city.
But how on earth would you *win* this challenge? No city means no research, no money, no production. You could never settle the galaxy OR be elected world leader this way, could you?
Well galaxy no, world leader yes, because Civ VI also has a hilariously broken victory condition.
In previous Civilization games, Diplomatic Victory required a majority of the world’s population to vote for you as leader. This meant you needed to make very good friends with a good number of the other Civs, becoming allies and trade partners, and being such good friends with them that they’d be willing to elect you leader, even though it meant giving up their sovereignty to you.
Civ VI doesn’t do this though, instead Diplomatic Victory means collecting “diplomatic points” until you have 20 of them, and 20 points means you win.
But how do you get diplomatic points? Some ways still rely on production and money, for example you can help out after natural disasters and build wonders of the world to gain diplomatic points.
Clearly those ways are unavailable if we’re hiding out in the ice caps, so the No City Challenge instead relies on the World Congress, which is hilariously broken in its own right.
The Civ VI World Congress starts up once enough time has passed for the game to reach the medieval era. At that point, every Civ will gain the opportunity to vote for random “world congress resolutions.” These resolutions are chosen at random, you have no control over them. And they’re binding on you, even if you’ve never met half (or all!) of the other nations in the World Congress.
And these resolutions make no sense when you think about that. For example, our real world has done a lot of work banning Ivory hunting, even though Ivory was considered a luxury centuries ago. The Civ VI world congress can also ban Ivory, but it does so even if the people voting on the resolution have never met each other. So you can have a situation where people you’ve never met, on the other side of the world, are now enforcing an ivory ban on you even though your own ruthless Civ sees nothing wrong with Ivory hunting.
Anyway, any time you vote for the winning “side” of a resolution, you earn a diplomatic point. Even if the vote wasn’t close, *even if you only casted a single vote*. If the world votes to ban Ivory and you also voted Yes, you get a diplomatic point.
You get votes according to how many cities you have *but you also always get 1 vote no matter what*, and here’s where we come back to the No City Challenge. Our Maori Civ hiding in the arctic still gets to vote in the World Congress, even though they don’t have any cities. It’s also *very* easy to predict how the AIs will vote, and very easy to know which World Congress resolutions will pass or not. So if our Maori Civ can just cast their 1 vote for the winning resolution each time, they can rack up Diplomatic Points until they have 20 and they win.
Think about this, a Civ sitting in the arctic, never founding even a *single* city, has “won” because they voted for the winners in every election of the World Congress. The other Civs of the world have determined that the Maori (who they never knew existed until now, wait how did their votes even get cast?), the Maori who have zero cities mind, are truly the skilled diplomats the world needs to lead it to peace and prosperity. And these Civs (who again, *have never met the Maori*) will give up their spaceships and their weapons of war to let these Diplomats rule the world.
And this isn’t even a theoretical victory condition, it’s actually happened. Several times.
This insane “victory condition” comes about because the AIs in Civ VI are very bad at *winning* even if they’re pretty good at *not losing*. See, the World Congress is Weird and Broken, but even then, previous Civ games would never have seen this type of victory because an AI would have won some other victory before then. Previous AIs were pretty good about conquering each other, culturally dominating each other, or reaching Alpha Centauri alone, especially if the player wasn’t there to stop the strongest Civ from running away with the game. And that’s what the rest of this post is about, Civ VI AIs can’t easily *lose*, but they can never *win*
I recently got the Civ VI bug again and wanted to write about it. I made some posts long ago discussing how Civ VI is the only Civ game I’ve ever beaten on Deity (the hardest difficulty level). This isn’t really because I’m good at the game, it’s because the AI is bad at it.
See, there are really two sides to “winning” a game. One side has to lose, the other side has to win. This seems obvious, but let me be clear: the AI in Civ VI is *really really bad at winning*, so much so that if the player can even become *moderately good at not losing* then they are guaranteed to win eventually, even if they themselves are bad at winning.
Let me compare Civ VI to its predecessor, Civ V. I once played a very high-level game of Civ V with Polynesia. I settled islands, I built my navy, and since this was an “archipelago” map where there was lots of water everywhere, this made me undefeatable in war.
See Civ V made it so that land units traverse the water by just walking into it and conjuring up a boat for themselves (maybe they built their boat on the land). But these land units are completely powerless in water, they are instantly destroyed by any true naval unit. A roman trireme can attack a division of marines, and as long as the marines are on the water the trireme will win and take zero damage.
So in this Polynesia game, my main war strategy was to bait enemy land units into the water and slaughter them with my ancient, obsolete ships. I would repeatedly send triremes against marines and modern armies, and win with no casualties because the AI never build naval units to defend their sea-borne land units.
It was impossible for me to lose. But I was never going to win.
See although I had an impregnable military, my economy was in dire shape. High level AIs get obscene bonuses to production, research, and the economy. My enemies were in the Industrial Age while I languished in the Renaissance, and even if this didn’t matter militarily it would soon matter technologically.
Civ has always provided a number of ways to win, both through war *and* peace. You could conquer all your enemies, or you could build a spaceship to Alpha Centauri and say neener-neener as you colonize the galaxy, that also counts as winning. Well my enemies were clearly going to get to Alpha Centauri while I was still figuring out coal and oil. They were going to *win* even if it it didn’t feel like I would *lose*.
Militarily, I was unstoppable. Culturally, I was fine. Economically, I punched above my weight. But in the end, my enemies could always win through Technology, and win they did.
This story is meandering, but it proves an important point: winning isn’t just about *not losing*, it isn’t just about staying in the game and staying active. There are victory conditions that the AI can still meet, and they can use those to win even if they don’t knock you out of the game, even if it feels like you never “lose.”
Civ VI though, Civ VI AI’s don’t have this. Civ VI AIs are like me in that Polynesia game, they’re good at *not losing*, they’re terrible at *winning*. And in fact they’re so bad, that they are almost incapable of winning at all.
The Civ VI AIs are terrible at building a spaceship to go to Alpha Centauri. They are incapable of achieving cultural or religious domination. They will never conquer most of their neighbors. And with those being the main ways you can win, a player playing competently will *eventually* luck into one of those. So long as a player just *doesn’t lose* they can slowly crawl their way into *winning*, even though the AIs are strong enough that they *should have won long ago*.
So long as a player just doesn’t lose they can slowly crawl their way into winning, even though the AIs are strong enough that they should have won long ago.
This should be an Atlantic article.
Dana Blankenhorn — https://www.danablankenhorn.com/
https://www.facebook.com/dana.blankenhorn
https://danafblankenhorn.substack.com/ https://www.amazon.com/Dana-Blankenhorn/e/B001HP8HCK
LikeLike