What city placement do you use???

akillias

Prince
Joined
Nov 7, 2001
Messages
353
I use the square city placement if im surrounded by people and cross like city placement if i am near the coast. The only problem with the square city placement is that the infrastructure needs a lot of work early in the game & it has a high corruption level later in the game.
 
I use map oriented city placement. Basically I just look at the map to see what a good place would be without resorting to math or geometric forms. If a tile further is near a river, then I go a tile further. And if there's a hill where I don't want to be, then I'll go a little closer. It depends a bit on the map size and the city size I expect to have. With the scenarios, one can build a bit more compact because no city will use all 21 squares anyway.
 
Settle near rivers, forests, grasland, and near hills( not surrunded by) and not minding much about resources.
 
I try to maximize rivers to benefit from the trade and growth without building a aquaduct. Mostly, I use a strategy intended to crimp the AI's ability to expand. If I find a choke point further out, I'll go for it in order to block AI settlers from moving toward the area I've staked out. I can then settle back areas with less of a frantic land rush.

When most of the map is settled, I may grab defensive positions in order to defend my territory. For these cities, I chose a bend in a river so that I benefit from 2 or 3 sides where the AI has to attack across the river... or a hill. These are fortresses for the most part, so I build walls quickly and some culture improvements. Later with artillery, I use these as firebases in order to blast enemy stacks before they enter my territory.
 
I'm working on a system to maximize country size while providing border cities with close neighbors. I start with one city 4 squares away directly NW, SE, SW, or NE of the capital. That gives it 12 culture spaces. Then I go 3 more in the same direction for the next city. That gives the two secondary cities each 9 squares with 3 up for grabs between them. If possible, I do that again for the fourth city. The 2nd and 4th city then become part of cities ringing the 3rd city (which will eventually get a palace)

X----X---X---X

The ring looks like this (half of it, anyway):
Code:
         X
      X -|- X
      |- | -|
      |  |  | 
O----X---O---X

The cities marked with O have the full number of squares to grow into; the X's are used as smaller 'fort' cities. In the early game they are pretty much all evenly sized. All of the X cities have two cities who can reinforce it in one turn, and the main city can reinforce it in two turns.

Build a similar ring in each feasible direction. This design allows me to capture land quickly, and on a large map corruption begins being felt around the area of the 4th city. If I build beyond that point I build 3x3-square cities. The ring design also leaves room for 3x3 cities at the corners, and allows them to be reinforced quickly. This provides a LOT of cultural coverage immediately. It allows me to meet the border of neighboring civs quickly, establishing barracks and a cultural improvement. I can backfill gaps in my coverage later. If I have the chance to expand into the corner between rings, I use 3x3 cities. Of course the placement varies a little with terrain.

My major goal is to have citizens working every single plot of land I own.
 
There is a city placement scheme called Optimal City Placement (OCP) but I couldn't find the article in the War Academy. Basically it places the cities wide enough apart so there is only 1 tile of overlap between cities and every city can work 20 out of a possible 21 tiles. It takes population 19 cities to do that. When I have used this scheme the AI will never settle between these cities even before the culture fills in. They will move between your cities but wont settle. Some people wont use this scheme because it is only effective late in the game when hospitals have been researched.

I have been using this scheme again but interlacing 2 grids that are both OCP until I can decide which one best fits the landscape. Once that happens I desiginate the wrong OCP grid of cities as extra so I only use them for settlers, workers and military units until late in the game when they interfer with the expansion of the permanent cities. Once the premanent cities need the space the extra cities are abandoned. Also any other extra city that gets placed after I determine which OCP grid is the permanent one can be put wherever it is the most effective.

OCP is rows of cities 4 tiles apart with a city every 5th tile. The cities in the rows are lined up with a 2 or 3 tile offset. So every time you start a new row you get a choice of 2 tiles for the first city in the row.

My interlace just sets the second OCP rows half way in between the first OCP rows. This arrangement gives twice the city density of OCP
 
OCP is amazing. The ability to get the most out of your territory with minimum overlapping.
 
I use a 'loose' OCP strategy, basically using OCP unless there is a reason for me to put the city in another place (eg get an extra bonus resource etc).

Read Bamspeedy's article on OCP and building up a large empire.



torrasque


(This image is easier to see how to place OCP-style cities, and is much easier to see than to look at a screenshot. Unfortunately, I can't remember the thread that I found it on, so apologies to the creator)
 

Attachments

  • civ3ocp.jpg
    civ3ocp.jpg
    93.1 KB · Views: 833
I would like to point out that the so-called OCP is far from the optimal city placement if you awant to play as effective as possible.

The main problem with OCP is that for most of your game, you're only using around half of your territory. Your cities takes quite a time before they reach size 12, and they certainly don't reach size 19 (or 21) before somewhere around the end of the inustrial era.

Another problem is that defensive units will need two turns to travel between cities by road. A city that falls to a surprise attck could have been saved if the defenders of the nearby cities could be able to reach there in one turn.

The third problem is that when those cities finally reach the necessary size of 20 or so, its quite hard to keep them happy, since there will be 18 or so that's born unhappy.

The solution to all of this is to build the cities closer and keep them smaller. If you let each city have around 13 squares each instead of 21, you'll have an empire that's about 80% more productive during the 2.5 first eras of the game.

If two otherwise equally skilled players played each other in MP, one with 21 squares for each city and the other with 13 (or less), the second player would easily beat the crap out of the first long before the first had built a single hospital.
 
well most of the time i try to expand my border quickly to keep the AI out of the expansion race
 
TheNiceOne has good points, but there are problems with that strategy as well- in the later two eras, your cities will have twice the maintenance costs (after all, 2 size-12 cities would have two marketplaces while 1 size-24 city would have 1), meaning that your empire will be hit with a much higher maintenance cost as well...

However, with C3C, smaller city placements can also be rewarded by the Feudalism government, which emphasizes lots of smaller cities.
 
Do I miss something, or are Hydarabad and Punjab misplaced in the above picture?
 
Originally posted by Corrado
TheNiceOne has good points, but there are problems with that strategy as well- in the later two eras, your cities will have twice the maintenance costs (after all, 2 size-12 cities would have two marketplaces while 1 size-24 city would have 1), meaning that your empire will be hit with a much higher maintenance cost as well...

However, with C3C, smaller city placements can also be rewarded by the Feudalism government, which emphasizes lots of smaller cities.
The extra costs for infrastructure are more than balanced by the extra production earlier.
Actually, it's not unusual for my games to be over before hospitals are even available.
 
But also- isn't unhappiness affected by the number of cities you control?
 
The early penalty is why I'm now interlacing 2 OCP layout schemes until hospitals become available. The average population for that arrangement is 9.5 to work every available tile. All you need to know is which OCP layout is the permanent one so you only build infrastructure and wonders in that one. The tempory OCP layout cities will be given a number rather than a name. I'm playing HOF games where I milk until 2050 so the number of cities is important. BTW I have cities in very corrupt areas with populations over 40 and all citizens are either happy or specialists. It is very easy to build infrastructure in cities like this with the use of civil engineers.
 
One question:

Ring City Placement - no longer works due to changes to rank corruption in C3C 1.15, right?

Now this is good.

I love a mix of OCP and G2RS. Go 2 Ressources, someone created this term some time ago, I love it... :)

Especially in ancient times a city grows much faster if directly besides wheat or another food giving ressource. You hardly reach the max pop number of a city, most of the time you are limited to 12 citizens - so NiceOne is right, it is a waste NOT to waste some tiles... :)

I still go for building very large cities and only "sacrifice" few tiles to overlap... ICS is a pain for my eyes, I cannot stand this closely packed strategic abomination... :p
 
Originally posted by Commander Bello
Do I miss something, or are Hydarabad and Punjab misplaced in the above picture?

They are misplaced. If I recall correctly, this graphic was made from an actual map and the terrain did not allow for those two cities to be placed in their "optimal" positions. :)

Herse
 
But you shouldn't mind much about resoures, what is the point of settling next to wheat when you could've had settled next to a river, unless you were thinking about a stettler factory.
 
I use OCP. Since my games go 'til modern times the larger cities are more worth it. If I attacked everyone like many people I probably wouldn't.
 
Top Bottom