City Specialization: Where I do it

If you haven’t already, check out iwas’s article on Specialization of Cities and Excl’s Guide to City Specialization and Land Improvements. A lot of what I talk about here is based on the information in those posts, so if you don’t have at least a passing knowledge of the material in their articles, mine probably won’t be too much help.
With that being said, I think their articles very adequately answer the questions of why and how to specialize cities but fail to answer the question of where to specialize cities (i.e., where to build them and how to identify what types of land best suit certain types of specialization).

So that’s what I’m hoping to answer with this article: how to identify a city site’s specialization propensity. Hopefully, this will help you place better cities when dot-mapping, identify which enemy cities are poorly placed and subject to razing and more quickly convert and bring online captured cities as specialized cities of your own.

See also in Deutsch: Stadtspezialisierung – Wo man’s am Besten macht. (Special thanks to JeWnS (ΙΕΝΣ of for performing and providing the translation.)

Feeding Your City (an ode to Excl)[/COLOR][/B][/SIZE]

[INDENT]In Excl’s article, there’s a section on feeding your city where you count the amount of positive and negative food tiles, which results in the amount of food the city lacks to support working all tiles in its BFC.

This is a very valuable number – but not for the reason most people use it. Whereas this number does indicate the amount of extra food you’ll need to reach population 20 (and thus the number of pre-biology farms the city needs), a city of that size is largely unattainable for a significant portion of the game. That is to say, even though you’ll eventually need those extra farms, there may be no justification for immediately building those farms as a priority.

As Excl put it, the amount of food lacking in your city determines how many tiles you’ll have to devote to feeding your city – and “everything else can go towards specialization.” So, the smaller the number, the fewer tiles the city can devote towards specialization. Thus, large (hopefully positive) numbers indicate a large number of tiles you can specialize and can be used as an indicator of how good the city site is – of how well the city is likely to accept specialization.

What that number does not indicate is what type of specialization the city should receive. So that’s what I’ll address here now.[/INDENT]

Types of Specialized Cities

[INDENT]Cities essentially specialize in one of three things: :food: (as a surplus), :hammers: or :commerce:. This specialization is what allows them to become GP Farms, Production Hubs, Commerce Centres, etc. Which of these three elements the city can specialize in is ultimately dependent upon the land and the game Era.

[INDENT]Production cities specialize in hammers.

Food cities specialize in [surplus] food.

Commerce cities specialize in commerce.[/INDENT]

How each city then uses its specialization is up to you, the builder. For example, a production city can focus on building only Wonders to become a very powerful GP farm; or a Food city can focus less on specialists and more on simply growing to attract substantial trade routes. But I digress … just don’t think you’re limited to what I present here is all I’m saying.

[INDENT]Because a late-game city’s specialization propensity isn’t strictly dependent upon its available tiles, this post is primarily concerned with cities during the Ancient to Renaissance Eras.[/INDENT]

* * * * * * *
* * * * * *
* * * * *
* * * *
* * *
* *

These are typically the steps I start with:
[*]Check the production. Using only hills and production resources (consider them both as improved), count up how much production the city can handle, how much food it costs to maintain it and how many citizens are required to work the production tiles used. Ideally, you want to see high production (15 – 20+) with few tiles (4 – 7).

[INDENT]For example, a city (1P) with 2 Grassland Hills (-1F, 3P each), 2 Plains Hills (-2F, 4P each) and a Plains Horse (-1F, 4P) would be -7F, 19P @ 4 – meaning the city can generate 19 hammers at the expense of 7 food using 4 Citizens. Even though the city center produces +2F, I often keep its food production with the food calculation.[/INDENT]

[*]Check the +food. Using only food resources (consider them as improved and as if the city has a Lighthouse), count up how much +food the city can maintain and how many citizens are required to work those tiles. Ideally, you want to see high +food (+8 – +12+) with few tiles (3 – 4)

[INDENT]For example, a city (+2F) with Grassland Hill Pigs (+3F) and Clams (+3F) would be +8F @ 2 – meaning the city can generate +8F using 2 Citizens. [/INDENT]

[*]Check the terrain features. Make note of how many unworkable tiles the city has (Peak, Desert, Ice, etc.) and terrain features such as Fresh Water, being coastal and/or riverside, how many non-resource water tiles the city has, how many grassland tiles are available and how many of those grassland tiles are riverside, how many riverside tiles there are, how many Flood Plains are in the city’s BFC. Ideally, you want to see no unworkable tiles, many riverside grassland tiles (6 – 10) and minimal water tiles if coastal (2 – 4).

[*]Check the rest of your empire. Make note of all the cities in your empire – specifically what kinds of cities you already have and what kinds you don’t have.[/LIST]
[INDENT]Knowing what your empire needs and the terrain features available can often help ‘break the tie’ when a city has multiple specialization options. Hopefully after this next section, you’ll know which of these features helps each type of city.[/INDENT]

* *
* * *
* * * *
* * * * *
* * * * * *
* * * * * * *
The Production City

In my experience, production cities are the toughest to found. Not only do they need access to production-rich tiles like hills and strategic resources but because those tiles are usually food-deprived, the site also needs food-rich tiles to maintain the balance. Furthermore, because +production% buildings tend to substantially increase :yuck:, the city often needs even more food and/or access to Fresh Water.

[INDENT]How much production do I want? I like to see at least 15 to 20 hammers per turn – preferably using 6 tiles or less.

How much food do I want? At least enough to support the production tiles, preferably using 3 tiles or less. Feeding the production tiles with +2F or more left over is ideal.

What kinds of terrain features do I want? The best production cities usually have some or all of the following:
[*]Access to Fresh Water to counteract :yuck:.
[*]Riverside for Fresh Water and to build a Levee, Hydro Plant and/or Three Gorges Dam.
[*]Typically few Flood Plains (4 or less) because of the unhealth they cause.
[*]Few non-resource water tiles if coastal (2 is the fewest water tiles you can have without sacrificing coastal).
[*]Riverside tiles (at least 6 — especially Grassland) for the Levee bonus and for building Watermills.
[*]Forests for :health: and for building Lumbermills.
The Food City

Though high +food cities are typically used to run specialists and as GP Farms, food serves many purposes. As we see in the production city, good food is necessary to support the production tiles. In coastal cities, +food can be used to support immense growth to increase intercontinental trade route yield (special thanks to Diamondeye for reminding me “International” doesn’t always equate to Intercontinental). Or, it can be used to support rapid growth for whipping expensive buildings and/or units. Don’t get locked into rigid ideas on how food should be utilized in any given city.

I find food cities to be fairly easy to locate, because they essentially only need access to a couple food resources. The difficult art of building a great food city is grabbing up multiple food resources without hurting the cities around it.

Because this post deals with early era cities, it’s worth noting before Civil Service, lakeside and riverside tiles (especially Flood Plains) are vital for continued growth of a food city. Ironically, riverside tiles are also valued very highly by commerce cities, so especially consider the needs of your empire when choosing between food or commerce specializations.

[INDENT]How much production do I want? For a non-production city, I like to see at least 8-12 hammers per turn in order to build the city’s vital buildings (like Temples, Markets, etc.). Keep in mind that with the Universal Suffrage civic, buildings can be rush-bought.

How much +food do I want? For steady growth (and light whipping), I like to see at least +6F to +8F. For running specialists (and heavy whipping), I like to see at least +8F to +12F, preferably using 4 tiles or less. For a GP Farm, more than +10F using 3 tiles or less is ideal. In every case, the most +food using the fewest tiles is key.

What kinds of terrain features do I want? The best food cities usually have some or all of the following:
[*]Access to Fresh Water to combat :yuck: from high population.
[*]Riverside Grassland tiles and/or Flood Plains (6 total or more) for farming pre-Civil Service.
[*]Non-riverside Grassland tiles (6 or more) for farming post-Civil Service/Biology.
[*]Coastal to reap the benefits of larger populations in the form of intercontinental trade routes.
[*]Few non-resource water tiles if coastal, since they can’t increase growth.
The Commerce City

I find commerce cities to be the simplest cities to find, because nearly any city with any amount of +food has the potential to become a commerce city (not necessarily a good one, but a commerce city nonetheless). All you really need is land to cottage. However, don’t let this oversimplification lead you to believe only cities with low +food should be used as commerce cities. One particular strength of a high food commerce city is the ability to rapidly grow to maximum population in order to work all available commerce tiles.

[INDENT]How much production do I want? For a non-production city, I like to see at least 8-12 hammers per turn in order to build the city’s vital buildings (like Temples, Markets, etc.). Keep in mind that with the Universal Suffrage civic, Towns generate 1:hammers: and buildings can be rush-bought.

How much +food do I want? I like to see at least 1 food resource — ideally generating +4F or more total. (Again, high +food is a bonus.) A ‘soft’ requirement is enough food to support production when necessary.

What kinds of terrain features do I want? The best commerce cities usually have some or all of the following:
[*]Coastal to reap the benefits of intercontinental trade routes.
[*]Few non-resource water tiles if coastal, since they yield little :commerce:.
An exception might be a Financial Civ with the Colossus.
[*]Riverside tiles (at least 6 — especially Grassland) for cottaging.
[*]Grassland tiles (at least 6 — preferably Riverside) for cottaging.
[*]High-commerce resources like Dyes, Gems and Gold.

An Example

[INDENT]Here’s a city screenshot from a recent game (special thanks to DiamondEye for supplying me with this game) with its improvements stripped:

[*]Check the production. Using Plains Hill City Center (+2F, 2P), 2 Plains Hills (-4F, 8P), 1 Grassland Hill (-1F, 3P) and a Grassland Hill Pig (+3F, 1P), Paris can generate 0F, 14P with 4 Citizens.

[INDENT]Good production, but probably not enough to make it a contender.[/INDENT]

[*]Check the +food. Using City Center (+2F), 3 Clams (+9F) and 1 Grassland Hill Pig (+3F), Paris can generate +14F with 4 Citizens.

[INDENT] Just working the food resources, I can support 7 specialists at population 11. Not at all shabby.[/INDENT]

[*]Check the terrain features. “+8” Riverside & Coastal city with no unworkable tiles. 13 non-city Riverside tiles (7 Grassland, 2 Flood Plains). 5 Water tiles (2 non-resource). Good forestry.

[INDENT]+8 riverside and coastal … wow! That’s an indicator this city is going to easily become a megalopolis. Coastal and riverside means better trade routes and ample +:health: to support a large population. This city could easily be a superb GP Farm or Commerce City.[/INDENT]

[*]Check the rest of your empire.

[INDENT]This was pretty much the best non-production city in the empire, so it was a tough decision. There was another good coastal riverside candidate for GP/Commerce with fewer food resources and less production, so I made Paris the GP Farm (I wanted the production available for nabbing good GP-making Wonders). I was also playing a warmonger game not expected to even see Liberalism, so I wanted short-term specialist payout instead of the long-term cottage ‘investment’.[/INDENT][/LIST]

If you’d like to see more examples, here’s the other 7 cities in that game:


[*]Check the production. -10F, 24P using 6 citizens.

[*]Check the +food. +6F using 1 citizen (+12F using 4 citizens on farmed flood plains).

[*]Check the terrain features. “-2” lakeside city with a few riverside tiles and not too many grasslands. Farmed Flood Plains can support production.

[*]Check the rest of your empire.

[INDENT]This was the best production city on the map, so it clearly became the Production City.[/INDENT][/LIST]

[*]Check the production. -8F, 18P using 6 citizens.

[*]Check the +food. +9F using 2 citizens .

[*]Check the terrain features. “0” riverside city with 14 workable Riverside tiles (4 Grassland, 3 Flood Plains).

[*]Check the rest of your empire.

[INDENT]This city has the potential to be anything. It was #2 in production and #3 in GP/Commerce. Being at war, I made it a secondary production city — primarily because it already had production infrastructure in place.

EDIT: As Quechua points out in a response post, Turfan is more ideally suited to becoming a commerce city (Gold and Desert Hill Mines aren’t particularly good production tiles). Keeping it a production city was a poor long term decision but a good short term decision. Had this game not ended early in a Conquest Victory, Turfan would’ve eventually been converted into a Commerce City.[/INDENT][/LIST]

[*]Check the production. -6F, 15P using 4 citizens.

[*]Check the +food. +4 using 1 citizen.

[*]Check the terrain features. “-8” Coastal & Riverside city with 1 unworkable tile. 10 Riverside tiles (4 Grassland, 1 Flood Plains). Shares 2 tiles with Paris.

[*]Check the rest of your empire.

[INDENT]This city was probably the worst. I farmed the Flood Plains and 1 Grassland, Cottaged the rest and just let it work them.[/INDENT][/LIST]

[*]Check the production. -7F, 22P using 8 citizens.

[*]Check the +food. +5F using 1 citizen after Calendar.

[*]Check the terrain features. “-4” Riverside city with 2 unworkable tiles.

[*]Check the rest of your empire.

[INDENT]Having very little food means this city can’t support specialists or any serious production without either being huge or having Biology-boosted farms, so I Cottage-spammed this city.

EDIT: As Quechua points out in a response post, Medina is also suited to become a production city. However, the circumstances in my game prevented this primarily because I didn’t yet have Calendar or any infrastructure in the city when I conquered it. I also prefer to afford my production cities +2F while working its production tiles, which wasn’t immediately plausible. So when presented the choice of bringing it immediately online as a growing commerce city or stagnating it for mediocre production, I chose the more immediately necessary economic support. Like my decision in Turfan, this may not have been the best long term decision but was good in the short term and didn’t appear to detract from my Conquest Victory.[/INDENT][/LIST]

[*]Check the production. -5F, 11P using 3 citizens.

[*]Check the +food. +6F using 1 citizen.

[*]Check the terrain features. “-4” riverside city with only 5 riverside tiles (3 Grassland, 2 Flood Plains). Sharing 2 tiles with Beshbalik.

[*]Check the rest of your empire.

[INDENT]With only moderate production and food potential and since this was my capital, (which I didn’t feel like moving), I cottage spammed Karakorum and ran Bureaucracy.[/INDENT][/LIST]

[*]Check the production. -3F, 10P using 3 citizens.

[*]Check the +food. +6F using 2 citizens.

[*]Check the terrain features. “+3” Coastal & Riverside city with 10 workable riverside tiles (3 Grassland, 4 Flood Plains). Shares 1 Flood Plain with Uruk (who needs it).

[*]Check the rest of your empire.

[INDENT]Being #2 in GP/Commerce potential, I turned this into a secondary specialist city to run Merchants (I wanted a Great Merchant) with the intent of cottage-spamming it if the game ever made it to Democracy (it didn’t).[/INDENT][/LIST]

[*]Check the production. -3F, 12P using 3 citizens.

[*]Check the +food. +4F using 1 citizen.

[*]Check the terrain features. “-4” landlocked city with 3 unworkable tiles and no riverside tiles. Sharing 2 tiles with Karakorum.

[*]Check the rest of your empire.

[INDENT]This city is hurting for food but has Grassland Horses to redeem it. Since I cottage-spammed Karakorum, I took the Plains Hill and turned this city into a “defense-only” military production city making nothing but cheap Archers and Spearmen.[/INDENT][/LIST]

I had really good land in that game, so the choices were actually pretty simple. The main things I noticed from that land is that 50% of my cities could crank production pretty high, and 25% had great production. That in itself lends itself to warmongering and Wonder-hogging. I chose warmongering, so it was easy for me to lean towards an early-ending Specialist-heavy economy, which my city specializations mirrored.[/INDENT]

A Few Remarks

[INDENT]I don’t purport this ‘method’ as the end-all, be-all of analyzing city potential. It’s simply the method I use (BTW, I play at Monarch, so I can’t vouch for this method at Emperor+ yet.) If you use a different or better method, I’d love to hear about it and try it out.

I prefer to run the Caste System civic for the vast majority of my games, so my empires usually lean towards a Hybrid Specialist Economy. In these games, my commerce cities typically have little food.

That being said, I don’t restrict my economy based on my own preference. If I see the land immediately around me favours cottaging, I’m not at all averse to running a Commerce/Cottage Economy and usually will.

I use this method for determining what I feel is the best dot-map for the land around me as well as a guide to whether captured cities should be kept or moved (as well as what to do with them if they are kept).

This method of analyzing cities and terrain also helps me determine the most viable strategies. For example, if my land is weak in production, I know war and/or Wonder-hogging may not be the best game plan … so I know I should prioritize only the Wonders I need and avoid war unless it is especially quick and decisive or will gain me a good production city. Similarly, if my land is especially heavy in production, I may prioritize warmongering and Wonder-hogging.[/INDENT]

In Closing …

[INDENT]I consider this post largely “under construction”, so I’m open to any suggestions on how to improve it …[/INDENT]

— my 2 :commerce:

Discuss this article on the forum