Some Like It Hot: Montezuma, Emperor

With the war against France concluded, I paused to assess the larger game situation on Turn 124:

I was now out to 8 cities in total, with France reduced to 3 cities and Spain still sitting unchanged at 5 cities. My cities were also significantly larger and more productive than their counterparts, no doubt due to the AI's poor management of the economic side of the gameplay. In terms of overall score, I've passed both of my neighbors in the rankings. However, much to my surprise I was not the game-wide leader, and in fact remained a good distance behind someone out in the fog. Wow! Quite a contrast from the laughable performance that the AI put on in the Adventure One game. Here on higher difficulty, with land more tightly constrained, I was getting a significantly better showing from the competition. From what I've been seeing, the AI still fails at expansion, and as a result the larger map sizes tend to be a lot easier, as the player simply out-expands their way to victory. On smaller maps though, where the land fills up quickly, things get a lot more interesting (and, for me, fun to play).

Research was underway into the critical Industrialization tech, which would enable factories all around. I had already picked up Feudalism on the civics tree, and if you look closely in the above image, you'll see some Feudalism-buffed farming triangles near Xochicalco. I was heading next for Merchant Republic and the second tier of governments, although I was slowed by a lack of shipbuilding-related boosts on the top side of the tree. I'll have to work on planning that better in the future, I always seem to forget to build galleys and privateers and whatnot until it's too late to claim those boosts. I also have to try an archipelago map sometime soon to explore that side of the gameplay.

Here's another example of trade routes buffing up little Atzcapotzalco, more than doubling local production. They allowed the city to construct an Industrial district in a reasonable amount of time, about 15 turns in total. This is how you get cities founded in the later portions of the game up to speed, by using trade routes and forest chops to knock out the cost of those first few districts. After that, the cities can typically take care of themselves. I especially wanted a factory and power plant here to boost the capital's production down the road. Now if Firaxis can provide a better interface to handle trade route spam, especially adding in an auto-renew option, that will remove a lot of the tedium of this process.

A series of quiet building turns followed. When I finally reached Merchant Republic, here were the new policies that I adopted:

As usual, I'm running the builder policy available (Serfdom) and Charismatic Leader in the diplomacy column for lack of other options. The real story here is the other economic policies, with Town Charters and Caravanasaries providing a lot of gold income to play around with. I was close to 100 gold/turn at this point, and that gave me a lot of spending options. I've especially found myself liking Town Charters (+100% Commercial district adjacency bonuses) because it tends to be pretty easy to place Commercial districts on rivers for the +2 gold bonus, and the benefit from Town Charters kicks in immediately. Unlike Caravanasaries, it does not require building a trader unit after the district is finished. Of course, better yet is to do what I did here and simply get both of them at once. I'm also running good old Conscription in the military category, which is my default choice when not fighting a war or using Professional Army policy to get the half cost unit upgrades. There's another excellent policy at Mercantilism civic in the form of Logistics, which provides +1 movement to all units starting the turn in your territory. The bonus also applies to builders, and it saves so much time in moving them (or any other unit) around. With three moves and Industrial or Modern roads, movement finally stops being such a chore in Civ6, and it almost starts to feel like Civ4 again.

Medina Quarter and its predecessor the Insulae policy (+1 housing in cities with 2 specialty districts) are also virtual must-have policies as Civ6 rounds into the midgame. The key limitation on growth in this game is not food or amenities - it's housing. You need as much housing as possible to keep your cities growing upwards, and all sources of housing are therefore inherently very strong. Neighborhood districts are great, but unfortunately they don't arrive until the game is nearing its end (barring Kongo of course). These policies are some of the few ways to get more housing, and by Turn 150, every core city should have 2-3 specialty districts constructed already. More housing = more population = more of everything, in classic Civilization style. Don't ignore these policies.

I had a pair of caravels out exploring the map to find the other continent and the remaining AI civs. Norway had contacted me during the war with France, and can be seen in the screenshot at the top of this page. They were a mid-tier power, not the strong mystery civ on the scoreboard. That turned out to be Brazil, who had this impressive display appear on the trading screen when we finally made contact. Pedro had eight cities in total - that was the same as me! From the city names, he had already taken out a city state and captured Philadelphia from America, which proved to be the runt civ on the other continent. Pedro had also managed to accumulate five different Great Works of art, and he was making steady progress towards a Culture victory, with much more in the way of tourism than I had. Well, well, well. This was the best AI performance by far that I had seen to date in Civ6. Not that there was much fear of losing the game, this being only Emperor difficulty, but still not too shabby. I'm hoping I find more of this going forward.

This is an example of a mature city as the midgame of Civ6 starts to transititon into the lategame. Tlacopan has the two core districts (Commercial and Industrial), plus the third core district for a science victory in the form of a Campus. Yes, by this time I had decided that I would try to build the spaceship for the first time and get a chance to see the lategame toys at the end of the tech tree. Because this is a coastal city, I have also built a harbor here for the extra trade route, and the aqueduct helps provide more ever-needed housing. I also want to point out that the Commercial district itself was worth 8 gold/turn here, thanks to the doubling effect of Town Charters. That's part of the reason why I'm swimming in cash income in the above picture at 200 gold per turn.

Note that this city also has 48 production/turn in the picture above, and that's before power plants have arrived. One of my biggest complaints in my first game of Civ6 was that everything seemed to take forever to build. However, with good district placement and a strong emphasis on the production techs (Apprenticeship, Industrialism, Electricity), I've since found this to be a phantom fear. An average city can reach 50 production/turn in Civ6 without too much trouble, and that's enough to build even the lategame tier 3 buildings in about ten turns. Strong production cities can double or triple that mark. My capital had no trouble building new districts in just a few turns at the end of the game. The only things that are really expensive are the spaceship parts, and there are some ways of working around them, which I'll get to in a bit. Long story short: my initial complaint that everything takes forever to build has proven not to be true. That's another huge plus for me, as the slow pace of production in Civ5 was a major reason why I disliked that game.

As the player continues to build districts and advance down the civics tree, trade routes continue to get more and more powerful. I'm using the Triangular Trade policy at this point, which grants +4 gold and +1 faith for all internal domestic trade routes. With as many trade routes as I have in this game, it's worth a lot of money. The food and production aspect of the trade route comes from the specialty districts located at the destination city. By default, every internal trade route is 1 food / 1 production, and districts add on to that base value. I know that the Industrial district is worth +2 production for example; I'm still learning the trade route yield of all the rest. It may be worthwhile to pick districts specifically to customize the power of trade routes, for example having one city that goes Industrial/Harbor/Encampment to get the maximum production value possible on trade routes. This is something I'm still exploring in my games.

I also wanted to highlight one other important civic during this peaceful building era:

The Enlightenment has three useful economic policies. Free Market grants +100% gold from all Commercial district buildings (not adjacency bonuses), which effectively makes markets worth 6 gold/turn, banks worth 10 gold/turn, and stock exchanges worth 14 gold/turn. If you're building Commercial districts all over the place for trade routes - and why wouldn't you be doing so? - then this is an extremely powerful policy, probably the best wealth-generating one in the game for normal playing conditions. Rationalism does the exact same thing, only for Campus district buildings instead. Any game pursuing a Spaceship victory, or pretty much any game doing something with lategame teching, wants this policy. Since Rationalism often appears right around the time that the player is completing universities, there's a drastic increase in science as universities suddenly provide 8 beakers/turn apiece. Finally, Liberalism provides +1 amenities to all cities with at least 2 specialty districts, which will be almost every city by this point. It can be useful if your empire is facing a happiness crunch. I didn't realize how useful all these policies could be until I played this game, but now I view The Enlightenment as one of the more important civics to target on the tree.

Since the war with Catherine ended, she had remained a peaceful neighbor, denouncing me diplomatically but otherwise not raising a fuss. I had hoped that she would be stupid enough to declare war again, ah well. Guess we have to do this the hard way then. I declared a "formal war" myself, the first time that I had gone on the attack in Civ6, and went after the remaining French cities.

Because I was the one initiating combat this time, I could set up a siege unit in position to fire on the first turn of the war. Catherine had resettled the old Rennes spot with Rouen, and now I was forced to recapture the same city. (Should have kept that one, learned my lesson.) Fortunately the bombard unit that I had on hand was up to the task. 140 damage in one shot, sheesh. I had three of the bombards prepared for this war, and a whole bunch of former archers/crossbows that had now been upgraded to field guns. As the picture demonstrates, Catherine was still running around with spears and chariots. The resulting combat was once again a bit, umm, one sided. (Firaxis needs to do a serious rethinking of the strategic resources and unit upgrade path portions of the gameplay. The AI does reasonably well at building units when they have the appropriate resources, but way too often they lack them, and wind up with warriors and chariots and whatnot far too late in the game. This needs a fix.)

Catherine and I had been rivals for more or less the whole game, and France had sneak-attacked me earlier. How would the rest of the world react to my declaration of war?

With a ridiculous and totally over-the-top "warmongering" penalty, naturally. This screenshot is a good example of how silly the numbers were in the release version of Civ6. I've declared exactly one war in this whole game, against someone who attacked me earlier, and that results in a penalty of -32 diplo points?! Come on. It's even more laughable given that the AI civs on the other continent have been warring amongst themselves constantly throughout this game, and Harald is currently at war with Teddy in this very picture. Those hypocrites! They're all warmongers too!

Two additional things to note on this issue. One is that Firaxis significantly toned down this warmongering penalty in the first patch, to the point where a single war declaration no longer destroys your reputation forever. I still think the screws are on too tightly even after the patch, but it's a vast improvement over the release version. Second, even here in the pre-patched version, I was able to keep Philip as a firm ally and in "Friendship" status for the rest of this game. We had enough positive modifiers from Open Borders, trade, delegation exchanges, and a series of other stuff that he stayed loyal to me. I've been able to pull off lasting friendships in some of my other games too, including a pair of them in an offline Immortal game. So while diplomacy in Civ6 remains very much a work in progress, it's not quite as broken as I thought at first glance, and it does seem to be getting (slowly) better in patches. Let's hope so anyway.

One thing that does suck about Civ6: missionary/apostle spam:

I'm about halfway through the war with France here, having taken Bourdeaux and Paris, and with Lyon as the next target. Unfortunately I can't even reach the French city because there are so many freaking religious units in the way! There are 7 Viking missionaries and 9 (!) Spanish apostles in this one screenshot, gumming up the roads and blocking my path forward. When the AI gets production discounts on the higher difficulties, that discount apparently also applies to religious units, and it results in a flood of missionaries all over the place. What should we call this, the carpet of prayer? Argh.

This would be so much better if Firaxis would change religious units so that they don't occupy the same strategic layer as combat units. I'm hopeful that they will make that change eventually, and remove the possibility of religious units blocking military units from moving. This is another place where One Unit Per Tile continues to bedevil Civ6's gameplay. Firaxis has made one helpful change in the first patch, making missionaries/apostles cost 50% more faith to purchase but also have 3 default spread charges instead of 2 charges. This makes each individual unit more valuable and helps reduce the spam of units on the map. It's an awkward kludge though, and doesn't solve the real problem. I'm hoping to see more development here in the future.

I do need to thank Philip though. With Spain spamming Taoist missionaries and the Vikings spamming Jewish missionaries, their faiths each canceled each other out, and I had no fear of losing to a Religious victory. If Spain hadn't been in this game, or if I had conquered them, I would have faced a legitimate threat of a loss via religion. It's something you need to keep in mind while playing.

After about two dozen turns of fighting, France was finally eliminated from the game. That ridiculous missionary spam had slowed my path to the last French cities significantly, and some rough terrain over by Lyon hadn't helped either. It was finally done now, and I was left as the undisputed master of my continent, with Philip happily tagging along in the sidekick role. I finished with a little over a dozen cities, and there would be no more fighting left to do in this game.

Check out the minimap in the above picture. One other thing that I like about Civ6 is that the land does fill up with cities, just as it used to do in past Civilization games. I've never set foot on the other continent, and about 80-90% of it has been claimed by AI cities, with the remaining empty regions largely consisting of deserts that the AI was correct to avoid. I don't mean to keep hammering on the same point endlessly, but the contrast with Civ5's empty, unsettled maps is again instructive.

Here's another peek inside Tlacopan from later in the game, this time with the Free Market and Rationalism policies running. Note that the initial adjacency bonuses from the Commercial and Campus districts have been dwarfed by the building output now that the tier 3 structures are in place. That's not to say that placing the districts wisely doesn't matter, because it does, especially in the early game where getting a couple of +2 or +3 adjacency bonuses could double your entire civ's science output. However, here in the lategame it's more important to get the districts down anywhere, as the buildings that go into the districts are where the meat of the benefit is coming from. The whole system overall is quite elegant, and it's keeping my interest in the game high right now.

One interface feature that the game could use is a list of what districts are present in each city. The closest thing that the game has is this Espionage screen:

It includes most of the specialty districts, but puzzlingly leaves out Encampments and Holy Sites for some reason. I guess that's because this screen is supposed to be used for placing spies, not serving as a district overview. I could really use one of them though, exactly like this (with the little icons next to the city names) just with all of the districts present. Hopefully someday.

You might notice that science jumped up by a huge margin in the ten turns between the past two screenshots. That's because I was completing lots of research labs in this period, with each one increasing science by 10 beakers/turn. Finishing a half dozen of them did wonders for my research rate. Now science was far out in front of culture, which was amusing since the opposite had taken place in my Adventure One game. I was pursuing Rocketry tech at this point, and getting ready to start the space race in earnest.

Everything associated with the spaceship is extremely expensive to build in Civ6. Spaceports are their own district (make sure you have population enough to put them down!) that cost a cool 2000 production to build. My capital had Ruhr Valley and was getting 140 production/turn, and the spaceport still took over 15 turns to build. Better get started early... Then the first spaceship component costs 1500 production, the second costs 2000 production, and the last three cost 3000 production. Each! I have no idea why the designers wanted to make the parts so insanely difficult to build, when seemingly everything else in the game outside the wonders is very manageable. They must think of the spaceship victory condition as a tech + production victory condition, which is not the right way to think of it. The spaceship victory is the tech victory; reaching the end of the tech tree is how you win the game. (The military victory condition is the production victory condition, or at least it used to be in pre-One Unit Per Tile eras.) Making the parts crazily expensive doesn't make the space race interesting, it just makes it tedious, having to sit around with all the techs already completed hitting next turn and waiting on production. If I were in charge, I would probably chop all of the spaceship costs in half to avoid wasting people's time. Fortunately there are ways of speeding up the spaceship using Great People, and we'll get to those in a minute.

Another note: the zoomed out interface at Tenochtitlan has an icon indicating that the city needs amenities. However, that's actually false, and the detailed city interface reveals that the capital actually has +2 amenities. Pretty confusing.

Some of the Great People in the lategame are vastly more useful than others. Take Albert Einstein for example. He provides +4 science to all universities - wow! That's awesome for anyone pursuing a spaceship victory condition. But I could have ended up with Alan Turing instead, who provides the eureka boost for Computers tech and one other random Modern era technology. That's significantly less useful (sorry Alan). I don't mind the fact that the Great People are random in each game, but what I really do want to know is which Great People will be coming up NEXT on the list, so that I know which ones to focus on and which ones to pass. Civ6 should provide a list of all Great People available in each game, since the same Great People are not always available, and that list should be visible from the outset. Too much of the Great Person system is random right now, and this simple change would make it much easier to do real planning and targeting of the desired Great People. If there is a way to know what Great People will be coming down the pipeline ahead of time, someone please let me know about it.

Einstein ended up being worth about 50 beakers/turn, a not inconsiderable increase to my empire's beaker output. Science had now climbed over 400 beakers/turn and I was tearing through the remaining techs at a pretty good rate. A lot of the posters at CivFanatics are claiming that science moves too fast in Civ6, and I simply don't agree. I want to be able to research the entire tech tree in under 300 turns. What would be the point of dragging things out longer? If they want a slower game, then they should just turn on Epic or Marathon speeds. I like the current pace of the gameplay. Anyway, Tenochtitlan is about to start the first spaceship part (the earth satellite) in the picture above. Check out the build times of all the non-spaceship parts as well: 3 turns or less across the board. Definitely not a case of a game where everything takes forever to construct!

Here's one of the Great People that I really wanted to help contribute to the space race. Robert Goddard's 20% production bonus towards spaceship projects would go a long way towards churning out those 3000 production parts. There was just one problem: I had claimed a previous (useless) Great Engineer just a few turns previously, and this one was going to get taken by Brazil. Argh, if only I had known that Goddard was going to come up next, I would have passed on the last guy! Fortunately, I had enough money in my bank account to purchase the Great Person via Patronage, at a cool cost of 4593 gold. Very pricey but worth it. This is why it's a good idea to save up some gold and faith in the lategame, as you never know when one of these Great People will be the one that you absolutely have to recruit. Remember, if you pass on a Great Person you must wait on the AI to claim him or her before being able to take the next one, and that can sometimes be a very long time. (I still feel like I need to learn more about how this whole system works with the Great People. It's making more sense now but I don't have it all figured out yet.)

I was also trying to hunt down the remaining boosts on the tech tree to keep speeding along the space race. One of the last boosts comes from mining an aluminum resource, and of course I had no aluminum anywhere in my territory. (Strategic resource placement in this game is bizarre. There was no land-based oil anywhere in the entire world, and the only sea-based oil was a pair of offshore resources clustered together next to Rouen, just offscreen in that picture. Two oil next to each other at one city and none anywhere else. Huh?) There were two aluminum resources at the extreme northern end of my territory, however, and so I built a settler and rushed it up here to found a city for that tech boost. I could have done a better job planning for some of the other tech boosts, not getting most of the naval or air power ones, which would have been doable with better foresight. This was my first time playing out the Civ6 lategame though, and I'm sure I'll get better with time.

My capital was going to build most of the spaceship parts. It was the best production city by far, thanks to Ruhr Valley, and the largest city thanks to its strong starting terrain. I wanted another city to share the load, however, and decided to set up a second spaceport in Cempoala:

This city had nearly perfect terrain for a production powerhouse. The land to the north was all hill tiles, and I managed to place the Industrial district in a spot where all six of its adjacent tiles were mines. A perfect circle surrounding the district for the +6 adjacency bonus! I should have gotten some kind of extra bonus for that, heh. The southeast was then full of flatland terrain perfect for establishing farms, which could all buff one another with the Feudalism/Replaceable Parts bonus. The only issue is that as a coastal city in the deep south, Cempoala did not get the factory/power plant bonus production from too many of my other cities. Nevertheless, it was quite the city. I would have the capital build the first of the 3000 cost spaceship parts, Cempoala do the second part, and then the capital complete the third and final part.

Note as well that I've rebased all of my trade routes to Tenochtitlan and Cempoala here at the end of the game, to help speed them along with the spaceship part construction. That meant a lot of shields, but also a lot of food; the city is getting 105 food/turn in the above screenshot, thanks to the lategame civic that adds +4 food to all domestic trade routes. (In a very ironic twist, this is the "Collectivization" policy. Needless to say, in the real world collectivization of agriculture did not lead to increased food output.) The surplus of +67 food was a wee bit more than I needed here, and resulted in endless growth despite the housing cap. Kind of amusing how that worked out.

I finally spotted another highly useful Great Person: Sergei Korolev was worth 1500 production towards a space race project. I took that in a heartbeat and prepared to use him to speed along the last part. Korolev isn't even the best option here; Carl Sagan provides 3000 production towards a space race project, instantly completing a part for free, and Stephanie Kwolek and Werhner von Braun are both worth +100% production towards spaceship parts. Sagan and Kwolek are both Great Scientists while von Braun is a Great Engineer; all of the good spaceship stuff seems to be located under either Great Scientists or Great Engineers. Fortunately you'll probably have lots of Scientist and Engineer points if going for a spaceship victory, so this does work pretty well. Definitely keep an eye on the Great Person screen if going for a space victory in Civ6.

Here's a final screenshot of the capital with all of its production modifiers in place. I was able to hit 218 production/turn here at the end of the game, with about a third from worked tiles, about half from buildings (much of that from neighboring factories and power plants), and the remainder from trade routes. With all of this stuff in place, I was able to get even the monstrosity spaceship parts down to a build time of 10 turns, which is lengthy but not absurd. In the end, therefore, the spaceship isn't as bad as it looks at first glance. With proper planning and some luck with Great People, it can be completed in a pretty reasonable timeframe. As it turned out, I would complete the spaceship a mere 4 turns after finishing the final required tech on the research tree. Those horror stories I was hearing about players sitting around with the tech tree finished, taking another 50 turns to construct the spaceship parts... that situation can largely be avoided with some planning. Maybe it won't always line up as neatly as it did for me in this game, but it shouldn't take anywhere near that long.

I burned Korolev on Turn 259 to knock out half of the cost of the last spaceship part. That dropped the ETA from 10 turns to 4 turns in the capital, which was exactly the time remaining on the other spaceship part in Cempoala. Much as I'd like to say I planned that, it was a happy coincidence. As a result, the two spaceship parts completed together at the end of Turn 262:

And the result was a spaceship victory on the next turn, Turn 263. That was much faster than I had been expecting when I started pursuing the space race in earnest after my second war with France, and I was pleasantly surprised at how quickly it all moved. (Firaxis, please do not drag this out with more pointless busywork down the road.) I just barely ticked over 500 beakers/turn at the very end of the game, and that was a sweet little milestone to achieve at the end.

I enjoyed the victory cinematic; there's a unique one for each victory type. The spaceship parts themselves have a small little wonder movie when they are completed, which is another nice touch, even though I'll likely skip them after a few games. The bar graphs were noteworthy for the amount of competition put up by Brazil, which genuinely kept pace with my civ up until I completed my conquest of France. Then I pulled ahead in the lategame as the advantage of human planning plus a larger territorial base came to the fore. This particular graph details "total buildings constructed", which is a fairly good metric for overall empire strength. Note that I was behind Brazil for a long time in this game. In other words, while the AI has major problems, it's not a complete pushover in the way that a lot of our posters at Realms Beyond have been suggesting. On higher difficulties and with more cramped map conditions, it can be competitive.

As a final word here, I want to discuss the Civ6 AI a little bit further. We've had a lot of deserved criticism of how the AI plays this game on our forums, and I'm certainly not going to dispute that. The AI has no idea how to play by the One Unit Per Tile rules, and it makes dumb move after dumb move. The AI has no real ability to threaten the player with conquest after the early stages of the game, and it inevitably falls apart economically in the middle to late stages of the gameplay. All of this is completely true about Civ6.

However, I want to make one point: is this really THAT different from past Civilization games? The AI certainly isn't any worse than the one in Civ5, which was so dumb that it couldn't understand how to kill one of my scouts with a whole city full of military units. It's definitely not worse than the Civ3 AI, which would turn its entire military around if you pulled its "puppet strings" by threatening a city, and which would walk huge stacks around in aimless circles achieving nothing. Sirian wrote the classic study on the Civ3 AI, and there was nothing remotely intelligent about how it played the game. Even in our beloved Civ4, I'm not sure that the AI was much better. I've gone back and re-read a lot of the Civ4 reports on my website in recent weeks, and over and over again, in game after game, I see the Civ4 AI failing to mount any kind of credible military threat to the player. The only way that it could threaten military defeat was to turn up the difficulty to Deity and overload the AI with units and production bonuses... which is also the only way that the Civ6 AI can threaten the player with military defeat. Sensing a pattern?

I don't think it's unfair to call out the Civ6 AI on its shortcomings. The AI is unquestionably stupid in this game. I'm just not certain that there's much of a change here from previous Civilization games, which largely worked the same way. Because our community has been playing so much Multiplayer in recent years, I feel as though a lot of the posts are judging the AI in comparison to human vs human games, and yeah, of course the AI looks terrible in those comparisons. But all of the Civilization games look terrible by that comparison! Long story short: trying to judge the AI opposition in a Single Player game through the lens of human opponents in a Multiplayer game is never going to leave anyone satisfied. And it's perfectly fair to feel that way; however, we should be honest about what we're looking for, and it's not really a better AI. It's a human player to compete against.

Hopefully I'm not pontificating too much here at the end. This was my favorite game of Civ6 to date and left me much more encouraged than my first two writeups. I'm hoping this wasn't a case of fool's gold behind a false rainbow! Thanks as always for reading.