A Guide to Sid Meier's Civilization by Gus Smedstad, War/Strategy
This is a general guide to playing Sid Meier's Civilization, based
upon my experiences with the game. Keep in mind that while the strategies
laid out here work, they are not the only possible approaches to the game.
Food Under Despotism.....................................1
Roads and Irrigation Under Despotism.....................1
Building the First City..................................2
What to Research?........................................2
Advantages of Small and Large Cities.....................2
Expansion Under Despotism................................3
Talking to Other Civilizations...........................4
Libraries, Marketplaces, Banks and Universities..........5
Switching to Democracy..............................6
Keeping Order in a Democracy........................7
Wonders of the World.....................................7
The Knowledge City..................................8
Converting Woods to Plains...............................9
The Parallelia Strategy.............................10
The Mongol Strategy.................................10
Food Under Despotism
At the start of the game your government is Despotism, and it's likely
to stay that for some time. It's actually a good government type, but it
has one major limitation: food production.
Under Despotism, any square which produces three of an item (food,
trade, or resources) produces one less item than normal. This is a minor
effect for trade and resources, but for food the results are far-reaching.
Every citizen in a city requires two food. Any extra food beyond that
is stored away. If your city is to continue growing, you need at least one
more food than you need to feed your citizens.
The basic source of extra food for any city is the city square, since
you work that square for free. It's also automatically irrigated when you
found the city.
Until you get railroads or change to another government type, all
squares produce at most two food, with the exception of an irrigated Oasis.
This means that working a square is at most a break-even proposition for
food, and you get at most two extra food from your city square.
Working a square that produces one food, such as ocean or woods, or
unirrigated plains or hills costs you one of your extra food. You can work
at most two such squares, or one if you wish to continue growth.
Supporting a settler also costs one food.
Irrigated plains squares therefore are very important, since they
produce two food and a resource. Grasslands and rivers with resources also
produce two food and a resource.
Other government types aren't as hungry since irrigated river and
grassland squares produce three food, adding one extra food. They still
must watch their food budget when few river or grassland squares are
Roads and Irrigation Under Despotism
It's likely that you will need to irrigate squares early in the game
unless you start near a river or many grassland squares with the dimple
that means resources.
Though it's not spelled out in the manual, you can only irrigate a
square that has a source of water. Initially, this means squares adjacent
to an ocean or river square. However, irrigated squares also count as a
source of water, so you can steadily irrigate inland. The "free"
irrigation you get in a city square doesn't count for this purpose.
It is not necessary to maintain the chain of irrigated squares. If,
for example, you irrigate a hill, irrigate an adjacent plains square, and
then change the hill into a mine, you will not lose the irrigation in the
Irrigation adds one to the food in a square. However, under Despotism
this has no effect on actual production in grassland and river squares.
Under Despotism only plains benefit from irrigation, but it is very
important to irrigate plains under any government.
While the movement benefits of roads are nice, the economic benefits
are more important. Roads add one trade to a grassland or plains square.
Roads and rivers are the primary sources of trade for most cities.
The ideal square under Despotism produces two food, one resource, and
one trade. To achieve this, a dimpled grassland needs a road, and a plains
square needs irrigation and a road. Some river squares start this way, but
you cannot improve those that do not.
Building the First City
The first pick for any new city is an unimproved plains square. Since
you get free road and irrigation improvements in the city square, building
on a plains square saves you the effort of improving the square later.
Second choice is a dimpled grassland square. Resources are important
for a small city, and starting on a dimpled grassland or plains square
guarantees that the city square will have a resource.
Final choice is a river square. This is gambling, since there is no
way to tell beforehand whether the square will have any resources.
Near a river is a good spot to found any city. Many of the river
squares will need no improvement at all, and the river will make it easy to
irrigate plains squares.
If there is no river nearby, look for at least two dimpled grasslands.
Plains squares are also nice, but you need to improve them before they will
feed your city. Ordinary grasslands produce food, but they're empty
calories, since they don't produce any resources.
In any case, don't take overly long finding a spot. Every turn you
spend wandering in search of a location other civilizations are collecting
taxes and building items.
What to Research?
Which path to take in research is a difficult question. Almost all of
the improvements are useful. Early on, the only advances to avoid are
probably Horseback Riding, Monarchy, and Feudalism. None of these provide
much in the way of benefits except being prerequisites for Chivalry, which
gives you knights. Knights are nice, but not very important, particularly
since chariots are almost as good and much easier to research.
Early important advances are Bronze Working, which gives you
phalanxes, and the Wheel, which gives you chariots. Later you'll want
Ceremonial Burial and Mysticism for temples.
The most important long term goal is Railroads. Railroads are an
overwhelming advantage. First, they cut your defense costs enormously,
because you can shuttle units anywhere along your rail net in one turn.
Second, any square with a railroad gets a 50% bonus to all food, trade, and
resource production. Once you have railroads, you should build one in
_every_ square you are using, just for this production bonus.
Advantages of Small and Large Cities
Though the player's notes claim that bigger cities are always better,
they are mistaken. Small cities have several advantages.
The most important one is the free use of the city square. A size ten
city can work eleven squares, including the city square. Ten size one
cities can work twenty squares, one for each citizen and one for each city
Small cities grow faster under Despotism. Every new citizen costs ten
stored food plus ten food for every existing citizen in the city, or half
that if the city has a granary. A size ten city spends 110 food to make a
new citizen, or 55 if it has a granary. A size one city spends 20 food to
make a new citizen, or 10 food if it has a granary. Under Despotism,
larger cities don't have more surplus food.
Large cities require expensive improvements to keep the populace
content. Once a city grows past a certain size, every new citizen is
unhappy instead of content.
If you have version 3.0 or later of Civilization, you will find that
this last advantage no longer exists. Once you build a certain number of
cities, every new citizen in _any_ city, regardless of size, will be
Larger cities do have definite advantages as well. While smaller
cities can work the land with fewer citizens, larger cities can get more
from that work. City improvements can double the effective trade and
resource production of the land. These improvements are impractical for
small cities, since they cost more to maintain than they produce unless the
city is large.
Larger cities can also absorb more unhappy citizens without going into
civil disorder. This is crucial to waging war in a Democracy or a
Expansion under Despotism
Early in the game, cities should be small, to get the maximum benefit
from small population and to keep costs down. However, when founding new
cities, you must keep in mind that eventually your expansion will stop.
Sooner or later you will run into another civilization, or fill up your
starting continent. With that in mind, you should found cities roughly 3
or 4 squares apart so that they won't overlap when they reach their maximum
Initially, set taxes to 0% and science to 100%. You don't have any
expenses so you don't need to collect taxes, and your people are content
Build a militia to explore the countryside. They're cheap and
expendable, and you need to know where to build your next city.
Next, build another unit to defend your city, a militia or a phalanx
if you have bronze working. Now you can start building settlers to found
Each time you build a new city, you should immediately assign or build
a militia or phalanx to defend it. Even relatively friendly civilizations
will take an undefended city, and even warlike civilizations will talk
before attacking a defended city.
Once you have cities in all the suitable locations near your home
city, you can begin improving your home city. Your home city should build
one more settler to build roads and irrigation and then shift to things
like a granary or a temple.
Every city should follow the same pattern. Once it is no longer on
the frontier, it should build a single settler and then shift to city
The most important early military units are chariots, catapults, and
phalanxes. Other units such as militia, cavalry, and legions are generally
Most early units have a very low defense strength (one), so often the
side that attacks first will win. Chariots and cavalry move two spaces
each turn, so they can advance and attack. Other units must move adjacent
to the enemy first, which means the enemy can attack first and likely win.
Chariots are better than cavalry because they're stronger and cost the same
Sometimes, though, the defender has the advantage. Hills double
defense strength, and mountains triple it. A fortified phalanx on a hill
has a defense strength of 6, which is enough to defeat a chariot's attack 3
times out of 5. Even in grassland, a fortified phalanx is fairly strong.
Catapults do better than chariots attacking positions like this, though
they both cost the same.
It's important to note that units stacked together only defend as well
as the best defensive unit in the stack. If a stack is attacked and loses,
the entire stack will be destroyed unless the units are in a city.
Sometimes it's worthwhile to stack a defensive unit such as a phalanx
with an offensive unit such as a catapult or a chariot, but it's usually a
bad idea to stack more units than that.
Talking to Other Civilizations
Unless you are a Democracy or a Republic, it's almost always
worthwhile to talk to an ambassador from another civilization. Republics
and Democracies must be more selective about whom they receive, since the
senate will accept any peace offer.
Early on, almost all civilizations are willing to make peace, even if
they make threats first. It costs nothing to accept peace if you are not a
Republic or Democracy, since you can always break the treaty at will.
On the other hand, demands for tribute almost never get any response
if you've already accepted a peace treaty. If you really wish to demand
tribute, and don't mind being at war, refuse the offer of peace. If the
other civilization is truly afraid of you they will offer money,
civilization advances, or both in exchange for peace.
Normally, Democracies can't start wars. However, you _can_ use the
"revolution" option, which turns you into an anarchy, and then attack.
When the government menu pops up, you can choose democracy again.
The problem is that even if you have the Pyramids, this costs you a
lot of food and cash, because anarchy produces far less of these items,
just like Despotism. The reduced food income may starve some cities.
Further, you get massive corruption, which costs you even more cash. Since
most democracies rely on some of that heavy cash flow to pacify citizens
with luxuries, you also get civil disorder, which halts production.
Democracies and Republics can, however, still use a diplomat to incite
a revolt in an enemy city. This will provoke a war immediately, without
the expense of revolution.
Another tactic is to take several squares away from an enemy city by
first placing a diplomat or caravan in the square and then moving another
unit into the square. This usually provokes a war fairly quickly.
Depending on the difficulty level, the first few citizens of a city
are automatically content. At Chieftain this is 6, at Warlord 5, at Prince
4, at King 3, and at Emperor 2. After that every new citizen will be
The short term solution to unhappy citizens is to remove some workers
from the land and make them entertainers. This solves the problem quickly,
but it's very expensive. The city may halt growth or even starve.
The cheapest way to pacify citizens for a Despotism is military force.
Extra phalanxes don't cost much to build, and don't cost anything to
maintain unless the city has already built several other military units.
For other governments, military force is expensive, since military
units cost 1 resource/turn, and each unit only pacifies one citizen.
Despotisms reach this limit too, once the city is supporting at least one
military unit per citizen.
A temple is the next cheapest way to keep the city content. A temple
costs $1/turn, and pacifies two unhappy citizens, so it costs $0.5/citizen.
A cathedral costs $0.75/citizen, and a colosseum $1.33/citizen.
Luxuries are more expensive. The first luxuries make content citizens
happy at $2/happy citizen. Each happy citizen offsets one unhappy citizen,
so the overall cost is $2/citizen. Later luxuries first make an unhappy
citizen content for $2, and then $2 more to make him happy, which is still
$2/citizen is still a bit cheaper than 1 resource/citizen. A "rush
job" costs anywhere from $4/resource to $2/resource, depending on how much
of the new item you purchase with cash.
Despotisms, Monarchies, or Communist governments generally don't spend
money on luxuries, since these governments are often short on cash.
Republics and Democracies generally _must_ spend money on luxuries, since
it's the only way to offset citizens unhappy about military units.
Libraries, Marketplaces, Banks and Universities
Libraries and universities increase your research, marketplaces and
banks increase your tax and luxuries. All of them cost money to maintain,
so you must balance the costs against the benefits.
If a city is producing at least 4 lightbulbs, consider a library, and
if it has 8 or more, definitely build one. Universities are slightly more
complicated, since you must account for the 50% library bonus that's
already there. Generally, if it has 12 or more lightbulbs, consider a
University, and if it has 24, definitely build one.
Libraries cost $1/turn, and Universities $3/turn. Both add 50% to
research. To obtain any advantage, you need to gain an extra 2 lightbulbs
from a library and 4 from a university. 2/50% = 4 lightbulbs. 4/50% = 8
lightbulbs, but the library is already adding 50%, so that means 12
Libraries and Universities cost resources to build, resources that you
would use for something else to improve your civilization. Once a city is
past the break-even point, consider what else you'd rather build.
Sometimes you just have spare resources, so the decision is easy. Other
times you'll be at war, so military units are more important, or you'll
want another improvement such as a cathedral.
You can use these same criteria for marketplaces and banks if you
substitute tax plus luxury for lightbulbs, or perhaps just tax. This is a
complicated question, since extra luxuries don't benefit you directly. If
all of your cities have extra luxuries, however, you can lower your luxury
rate, which will give you more money for tax and research. Extra luxuries
in a city also means a Democracy or a Republic can afford more military
units, since the extra happy people offset the ones unhappy about the
As a general rule, only Republics and Democracies generate enough
trade to get real advantages from Universities and Banks.
Despotism and Democracy are the strongest government types in the
game. Despotism has an overwhelming production advantage early in the
game, because a Despotism pays little or no maintenance costs for military
Monarchy and Communism produce more food than Despotism, but the extra
food is not worth the resource production lost to maintenance costs. On
the other hand, Monarchy and Communism are more forgiving of mistakes than
the other governments. Despots must worry about food production, and
Republics and Democracies must worry about citizens unhappy over military
Republic and Democracy nearly double the trade of Despotism and the
other governments, which _is_ enough to offset the maintenance costs.
These governments cannot use military force to pacify citizens, but this is
not a serious problem, since luxuries are cheaper than military force under
Monarchy or Communism.
In a Republic, military units generate fewer unhappy citizens than in
a Democracy. If your civilization is fairly widespread, though, the money
you gain by eliminating corruption with Democracy will more than pay for
the luxuries you must buy to offset those unhappy citizens. It's also much
easier to keep the outermost cities happy.
Switching to Democracy
Switching to Democracy can be a painful process, and sometimes it
seems it's impossible to make everyone happy afterward. Once you've
actually made the switch, inspect every city to make sure that all units in
the city actually belong to that city. If any don't, select them and
change their ownership with the Home command. Next, dump _every_ military
unit not in a city. It's hard wasting units, but the unrest at home may
otherwise be too hard to control.
Your income has just increased radically, so reduce your tax rate
until your income is the same as your expenses, or perhaps a bit more. Now
increase your luxury rate to the maximum, and then reduce it one step at a
time until a glance at the Attitude report shows a city with more unhappy
citizens than happy ones. Now raise it back a notch so that city won't go
into civil disorder.
Keeping Order in a Democracy
Keeping the citizens happy under Democracy can be difficult, since you
can't fall back on military force as you can under more repressive forms,
and every military unit outside a city generates two unhappy citizens.
Generally you get 2 trade/citizen with Democracy. A 30% luxury rate
and a temple will keep your city happy to about 5 citizens (2 basic content
citizens, 2 from the temple, 1 pacified with luxuries), at which point
you're getting 10 trade. If you add 30% taxes to that figure, you've got 6
luxuries and taxes, and a marketplace looks attractive. Certainly a
cathedral or a colosseum would be overkill.
A temple, marketplace, and 30% will hold you to 7 citizens (14 trade x
30% = 4 x 1.5 marketplace = 6 lux = 2 happy citizens, 1 content, 4 unhappy,
which the temple makes 2 happy/3 content/2 unhappy). Your choices are a
bank, a colosseum, or a cathedral. The colosseum is cheapest to build, but
very expensive to maintain. A cathedral costs more, and is cheaper to run.
A bank is a bit more than the colosseum, and doesn't increase your limit as
much, but is effectively very cheap to run since it increases taxes as
A marketplace/bank, temple, and 30% luxuries brings you to 10 citizens
(20 trade x 30% x 2 market/bank, +2 temple = 4 happy/2 content/4 unhappy).
At 7 citizens, the net cost is $3/turn - 14x50%x30% tax = $1/turn, and at
10+ citizens it's $0/turn.
The next step is a cathedral. It's more expensive to build than a
colosseum, but operating costs are becoming a real concern - a size 10+
city probably has a granary, an aqueduct, and a library as well, so the
costs are $1+$1+$3+$1+$2+$1 = $10/turn. Taxes are 20x30%x2 = $12/turn, so
you're just breaking even.
Temple, marketplace, bank, cathedral, 30% luxury gives us 20 citizens
(40x30%x2 +2 temple +4 cathedral = 7 happy/6 content/7 unhappy). This is
the absolute population limit - any more citizens MUST be specialists. If
need be, they can be Elvises, though that's strictly break-even.
Wonders of the World
The various Wonders of the World vary greatly in quality, though by
the end of the game you'll want all of them under your control.
The best Wonder is without a doubt Women's Suffrage. With Women's
Suffrage, a Republic can wage war without any unhappiness, and a Democracy
can wage war with severely reduced unhappiness.
Next in quality are the Colossus and Copernicus's Observatory (if used
together), J.S. Bach's Cathedral, the Cure for Cancer, and the Hoover Dam.
Other good Wonders are the Pyramids, Isaac Newton's College,
Magellan's Expedition, Michelangelo's Chapel, and the SETI program.
The Apollo Program is an oddball, since the benefit is nice but small,
but it's required to reach Alpha Centauri. If there are other advanced
civilizations about, building the Apollo Program will result in a space
race which the other civilizations may very well win.
The Great Wall and the United Nations aren't worth much, since other
civilizations will often offer peace and then break the treaty within a
turn or two.
Most of the early Wonders expire quite quickly, with the exception of
the Colossus and the Pyramids, so they are rarely worth building.
The Great Library can sometimes be used to good effect, but in general
it's just a way of playing catch-up, since you don't get an advance until
at least two other civilizations have it. This Wonder is worth more in
games with many civilizations.
At the higher difficulty levels in can be difficult to build the
Colossus or the Pyramids before another civilization does. One strategy to
counteract this is to follow the usual expansion strategy, except that once
you have built a second city, one city should start work on the Wonder.
The remaining city should continue to send out settlers to expand. It
should also periodically send a phalanx to the Wonder-building city, since
the Wonder city is not building troops or temples to keep its citizens
The Knowledge City
One very useful strategy is to place both the Colossus and
Copernicus's Observatory Wonders of the World in the same city, creating a
"knowledge city" which produces lightbulbs at a fantastic rate. Many
players overlook these two Wonders because they only affect one city.
Taken together, though, their effects multiply, making them twice as
effective as two separate cities each with just one of these Wonders.
The Knowledge city can shift the burden of research from many cities
with libraries and possibly universities to a single city with the
Colossus, Copernicus's Observatory, a library, and a university. A
library/university pair costs 240 resources. The Colossus/Copernicus
combination adds 500 resources to that cost, about as much as 2
library/university pairs, but improves the city's research by a factor of
4. Further, they don't require upkeep, and you get additional tax/luxury
income from the Colossus.
Even more important, you can place the Knowledge City to maximize
trade, which is harder to do with several other cities. Maximal food is
important too, since you want a large population. Generally this means
picking a spot with lots of grasslands, plains, and rivers and putting
roads in every plains and grassland square.
Gold would seem to be very nice for a Knowledge City, but Gold squares
don't produce any food. Under Despotism, this will halt your city growth
entirely, and it can stunt it badly under any government until you have a
large surplus from irrigated grassland and river squares. So if you can
pick a spot with plenty of grasslands near a gold square, do so, but hold
off on using the gold until you have left Despotism and have at least 4-5
extra food production.
Warfare remains much the same through the rest of the game.
Musketeers aren't much better than phalanxes, and cannons aren't much
better than catapults. Not until the introduction of battleships,
riflemen, and tanks does technology make much difference.
Until the introduction of Bombers and Artillery late in the game, city
walls can make a city very difficult to capture. If the enemy civilization
is not a Despotism, however, it must pay 1 resource/turn for every unit in
the city. If you cover every square in range of the city which can produce
resources with one of your units, the target city will have only the city
square's resources. All but one of the defenders will vanish because the
city can't pay for them.
Battleships are the best military unit in the game. Not only do they
have the highest attack and defense strength, they move very quickly,
particularly if you have Magellan's Expedition.
Bombers are too expensive for what they do. They do have the
advantage of ignoring city walls, but fighters can shoot them down very
easily, and they can only make one attack every other turn. Artillery is
superior in just about every way.
Nukes are something of an iffy proposition, since the massive
pollution they create can ruin land all over the world by triggering global
warming. At sea, though, they're perfectly safe. They're also good for
taking an enemy capitol that you must take quickly in order to force the
recall of a spaceship.
Building the Manhatten Project is a terrible idea if any of the other
civilizations have or will have shortly the ability to build nukes. Other
civilizations are generally quite lax about the use of nukes.
Diplomats are actually very good units, though they're easily
overlooked since they don't attack like other units. Diplomats don't have
any maintenance cost under any government, and they don't make republics or
The most important use of a diplomat is to establish an embassy. This
gives you a lot of information about the other civilization, and gives you
constant reports on the other civilization's progress for the rest of the
Inciting a revolt is the next most important ability, particularly if
the target city has city walls. Alternately, a wave of diplomats can
perform sabotage on the city until one destroys the walls.
Bribing single enemy units is usually too expensive, but it can be
useful in an emergency.
Aside from their other abilities, diplomats can be useful just leading
other units. Normally, a unit cannot move from one square adjacent to an
enemy unit directly to another square adjacent to an enemy unit. Diplomats
don't have this restriction. Other units can also ignore this restriction
if the destination square already has a friendly unit, even if that unit is
just a diplomat. Diplomats therefore can lead other units into squares
they couldn't otherwise reach.
Converting Woods to Plains
When you've made all other improvements you can, including mining all
the nearby hills, it's worthwhile to convert any woods squares to plains
for a Republic or a Democracy. Woods only produce 1 food, and your city
size is limited by food production. By converting to plains you make the
square self-sufficient, so you can spend that extra food on utilizing
another 1 food square, such as a mined hill or an ocean. Woods produce 2
resources; plains + hill produces 4 resources and 2 trade, and plains +
ocean produces 1 resource and 5 trade. Resources cost about $2 each in
rush jobs, so even if production is your only concern plains + ocean is
The Woods/Plains tradeoff becomes even more extreme once you have
railroads. Woods with RR produces 1 food and 3 resources. Plains with RR
produces 3 food, 1 resource, and 2 trade. That extra food means you can
now support 2 extra 1 food squares, or a specialist.
Specialists are generally a bad buy, since they produce only 2 trade,
and practically any square produces that under Republic/Democracy.
sometimes you have no choice, though, as when the city is unhappy or has
more citizens than available squares.
Still, even an extra specialist and 2 trade (total 4 trade) is worth
roughly 2 resources, which is what you lose by converting from woods.
Money from trade is more flexible than resource production, and more
important in the endgame.
The Parallelia and Mongol strategies are game-beating strategies that
exploit loopholes in versions 1.0 and 2.0 of Civilization. Both of them
are fairly sure wins at any difficulty level, but they remove much of the
enjoyment of playing Civilization, since they avoid two important facets of
play: city improvements and citizen happiness.
Both strategies start with the same early expansion outlined in
"Expansion Under Despotism."
The Parallelia Strategy
The Parallelia strategy is exploits the free square of each city to
Once you have run out of room to expand, plant a city in every plains,
grassland, or river square. Do not build any roads or irrigation, since
these come automatically with a new city.
Eventually you will carpet almost your entire territory with size 1
cities. Any city that is not working a hill, mountain, or ocean square
should support a Scientist instead. Only outer cities need have an
This strategy produces technology advances at an incredible rate under
any government. In effect, each square produces 3 trade under Despotism,
or 4 under Republic or Democracy, since the Scientist is worth 2 trade, and
there are no costs for city improvements.
With version 3.0 of Civilization, cities under 5 citizens cannot
support scientists or taxmen, so the Parallelia strategy is unworkable.
The Mongol Strategy
The mongol strategy is relies upon the production advantages of
Despotism and the efficiency of small cities. The basic premise of the
strategy is to concentrate all production on military units and overrun
enemy civilizations with your hordes.
Once you run out of room to expand, plant more cities in the spaces
between the original cities. Continue to do this
Maximum city size under the mongol strategy is 3, because a size 3
city requires no crowd control beyond the single unit you'd want for
defense. Once a city reaches size 3, you should switch citizens from 2-
food squares to 1-food squares until you have no excess food production. A
normal mongol city works 2 irrigated plains squares or grasslands with
resources, one of which is the city itself, and 2 woods or mined hills, for
a total of 6 resources.
Tax is initially 100% science. Your goals for research are the Wheel
for chariots, Mathematics for catapults, and Navigation for sails.
Absolute minimum is the Wheel and Map Making. Once you reach those goals,
you can either set tax to 100% for more units or keep doing research.
You should build no city improvements except for barracks, which cost
no money to maintain. You should not build Wonders of the World, though
you may capture a few from your enemies.
Version 3.0 of Civilization has several changes designed specifically
to thwart the mongol strategy. The most important is that citizens become
progressively more unhappy as you acquire more cities. Eventually it is no
longer worthwhile to build or capture more cities.
It is still possible to use the mongol strategy with version 3.0, but
more difficult. You must build temples to assist with the keeping the
peace, which was not necessary before, and often deliberately destroy or
starve captured cities. Barracks are also prohibitively expensive to
maintain, so you should not build them.