Waging War in Civilization VI
This is the counterpoint to our earlier piece about how to survive an early invasion. In this one: Information about how to go on the offensive.
First, let's review a couple changes from previous iterations of Civilization.
Like in V, military units can't be stacked, and the map is a hexagonal grid of tiles with rivers being lines between tiles. Unlike in V, units can no longer move into a tile unless they can pay its full movement cost. Since most infantry units have two movement points, this means they can't move next to a hill and then get on the hill in the same turn; they have to start the turn next to the hill to climb it. It's much more common now to end a unit's turn without spending all of its movement points. Crossing a river also ends a unit's turn, so even fast cavalry units will be slowed down by rivers. Roads remove terrain movement penalties (and, in later ages, speed up unit movement) as long as you are moving along the road, from a road tile to another road tile; otherwise, they do nothing for you.
These much-maligned changes to unit movement mean that terrain matters a lot more in unit positioning, and they also that retreating is slightly favored in a lot of situations. An archer can now escape to safety from an infantry unit even though it's not faster than the infantry, simply because the infantry can't attack it as it is pursuing, as long as the archer can get on difficult terrain.
The other major changes are to zones of control. Now, units that enter an enemy zone of control with their movement have to attack an enemy or end their turn; units that start the turn in a zone of control can move freely (but if they move into another ZoC tile, they once again have to attack or end their turn). City centers and encampments exert ZoC all around them; melee units exert it all around them too, but not across rivers; ranged units don't exert ZoC at all. Overall, this has made the battlefield feel more fluid, with more opportunities for positioning around enemies. They also make it easier in some situations to break through with cavalry units to pillage improvements and districts behind enemy lines.
Speaking of pillaging, it's become much better than it was in previous iterations. You can now steal significant amounts of gold, science, culture, and faith from enemies; and the damage you cause is generally harder to recover from. Pillaging districts (which you can do multiple times, as long as there are buildings left in that district) is better than pillaging improvements, as you might expect, and it's also a major setback: the city has to spend turns of its own production to repair the district afterwards, while you get to reap some of its productivity. In effect, you're stealing from your opponent future productivity, to get immediate gains now.
Thoroughly pillaging a city before conquering it seems counter-intuitive, but remember what you actually plan to accomplish in the end. Sometimes, it's not worth wasting units trying to assault a city center you can't really take, and grabbing what you can and getting out is better. Also, any warmonger penalties you get for taking a city will be "refunded" later if you return the city to its former owner in peace. If you want to avoid warmonger penalties, but find yourself with the upper hand in a war anyway, a good strategy is to simply ravage the enemy's closest cities, take them, then return them in exchange for a punitive peace (say, everything they have). This is particularly useful if you're not going for a conquest victory but you need to slow down an AI opponent that looks set to win before you do. Since Great Works can now be taken in deals, you can even do this to completely deny someone’s cultural victory strategy by cutting them off tourism and getting a bunch of tourism-generating works of your own.
Fight smarter, not harder
Against the AI, one thing is for sure: You'll be outnumbered. Unless you are playing on a lower difficulty setting, the AI's production bonuses will give it many more units than you will have access to under most circumstances. This means you'll have to take other advantages when you get them.
By far, the biggest advantage there is is technological superiority. Civilization has always vacillated on how much tanks should beat spears, but VI falls squarely in the camp that tanks are much, much better than spears. This is particularly true of modern, atomic, and information era units, which tend to swamp all other advantages by being so much more powerful than everything that came before them. Even low-powered Modern era units like the AT crew are dangerous to armies from earlier eras.
But it's not the only advantage. Terrain matters, usually by giving bonuses to defenders (though marshes give defending units a penalty), flanking (attacking a unit that is in multiple units' zone of control) and support (defending when the attacker is in a friendly unit's zone of control). Inherent bonuses given by governments, religious beliefs, and policies (mainly Oligarchy, Defender of the Faith, Crusade, and Wars of Religion). And of course, Great Generals/Admirals. Those tend to matter more in the early game, because they're all flat bonuses; +5 from a Great General is great early on, but it's proportionally smaller later as base unit power gets more significant. In part, this creates an imperative to get ahead economically or technologically if you want to win late game wars; eventually, better tactics won't be enough to beat the AI's economic advantages and you'll need to bring a bigger or better army, too.
Two other combat advantages are in the game which are somewhat more involved include unit counters and unit promotions.
Unit counters are key to a lot of strategy games, particularly in the RTS genre, but in Civilization they were never foregrounded and in VI they seem even weaker. Yes, spearmen counter cavalry, but they don't have such an overwhelming inherent advantage over cavalry that you absolutely need to make them. Ironically, the most noticeable unit counter bonus in the game is the +10 bonus that melee combat units (the Warrior line) get against anti-cavalry units (the spearman line). If I seem down on spearmen, I am: They are no longer a slight step up over the Warrior; pikemen are no longer a good unit to make in bulk that doesn't take strategic resources.
A lot of the advantages certain types of units have over other types of units are no longer inherent, but come with unit promotions. Units tend to get their first promotion pretty quickly this time around, and if you have a standing army that you're moving around to fight several wars, you will reliably end up with units that have three or more promotions, provided you're taking good care of your veteran units. Overall, units now have more meaningful specialization, and the risk-reward game of getting use out of experienced units without endangering them is a big part of the war game. It's also important to take XP into account when making attacks - units heal about half their health and end their turn when they take a promotion. In effect, there's a little reward for being aggressive or overextending, especially if you can set things up so that a unit gets a promotion and then has a guaranteed safe turn in which to heal.
Also: use your unique units. Every civilization has them and they are all good, sometimes much better than what they replace; except maybe the hoplite, which suffers from all the weaknesses of the spear line while not having a significant enough bonus to compensate for it.
Armies and Corps
Another big change is the possibility of "combining" two or three units into an army or corps. This is not really equivalent to stacking two units together and adding their power; the resulting combined unit is definitely more powerful than the original unit, but not twice as powerful. Armies and corps are still worthwhile, though.
A corps is essentially a tech era better than the original unit, and so they are in effect a way of compensating for shortfalls in your tech tree; an army is two tech eras better. They still fall significantly short of actual more advanced units, particularly modern/atomic era units that represent big leaps in power; a knight corps is not comparable to a tank, but it is significantly better than most medieval and renaissance units. They're useful if you have a huge preexisting army that's also technologically backwards; consolidating it into corps/armies is like upgrading your units by spending units instead of gold, which is sometimes a better trade-off.
The other thing armies/corps do well is push through choke points, heavily fortified cities, and other situations that favor the defender. With a bigger health bar and more combat power, they can take the incoming ranged fire well enough to be useful for a turn or two. Ancient walls, in the early game, would fold pretty quickly to a catapult; in the late game, a bombard or artillery corps can do the same for renaissance walls or modern city defenses (which all your cities get as soon as you research Civil Engineering).
Finally, corps allow you to make your elite veteran units even better by increasing their base power level while keeping their promotions. Remember that when you use the "create corps" button on a unit interface, the unit you have selected will "move over" the other unit, and it's the _other_ unit that will keep its promotions.
With a military academy, cities can build armies and corps, and this is significantly more efficient than combining preexisting units, enough so that building corps becomes preferable if you can afford the extra time. This gives encampments a lot of late-game utility; it's hard to gauge how much encampments help, given that it's an aggregate of a lot of small effects, but you will want some number of them for sure.
Cities: Defense and Attack
Finally, city sieges are probably the most tactically complicated aspect of warfare in Civilization VI. In practice, over the course of the game, you will see a series of stalls created by better defensive technology, which have to be broken by matching siege technology. Not all cities are created equal; terrain, encampments, and garrisons will change their combat capabilities. A city in the middle of the plain is much more vulnerable than one surrounded by jungle or nestled between mountain ranges. Bigger cities have more health than smaller cities, and so on.
To recap some of the new city defense mechanics: city centers and encampments now have two health bars— a green one (district health) and a blue one (outer defenses). Outer defenses take very reduced damage from attacks, except by Catapult line units. However, it can only be healed by spending production to repair it (a city project that can only be worked on if the city has not been attacked for at least 3 turns); and if that bar goes to zero, the city's walls are gone and have to be rebuilt.
City health takes much more damage from regular melee units, though it's still quite bulky. It also heals naturally every turn, unless the city is besieged. To besiege a city, you need zone of control around all of its surrounding tiles. Once again, zone of control doesn't cross rivers, so cities on rivers can complicate where you have to place your units to keep the city besieged. Attacking a city with the bare minimum units is possible, but it often means putting your attackers in vulnerable spots. Cities can only bombard if they have defenses (either from building walls, or from the Civil Engineering civic, which grants every city maximum defenses automatically and for free as soon as it is researched). Encampments, of course, can also bombard. The strength of the city bombard attack improves every age, too, so that late-game cities can overwhelm outdated units pretty easily.
Ancient walls are the first stall in the game, ending the brief early-game period where just warriors on their own can take out cities. You can still take out a walled city with just melee units, but it will take overwhelming numbers or extreme technological advantages; it's just not efficient. And while the battering ram is an effective counter to early walls, it doesn't really invalidate them like the siege tower or the catapult do.
Medieval walls really make the battering ram look weak, and they highlight the key weakness of the siege tower: It doesn't help you destroy city defenses that are going to bombard your units, and unlike a catapult, it can't attack a city without endangering your besieging units. Catapults can deal with a medieval walled city that isn't too defensible, but you will eventually need bombards to break some stalemates.
Renaissance walls, frankly, almost never get built. The cumulative production it takes to get all the way to them is such that, by the time you (or the AI) would be considering building them, you're right around the corner from Civil Engineering anyway. But they can be a real roadblock; a Renaissance walled city in a good defensive position is almost unassailable until the late-game stallbreakers show up.
Modern city defenses are even better than Renaissance walls, and they will grind city assaults to a halt in any minimally defensible city. If you're planning on an offensive war against a Modern-era enemy, you will absolutely need those late game stallbreakers: the observation balloon, air power, and battleships. The common thread in all of these is being able to do full-strength ranged attacks against district defenses without being exposed to their extremely powerful attacks.
There's not enough room to go over every nuance of the combat system in Civilization VI—it's very deep and complicated, and in a lot of ways a little too opaque and under-documented by this iteration of the Civilopedia. But it should give you a picture of the shape of warfare in the game.