Early generic cities

[noto]

So I read a lot of crap on these Civ fora and somehow arrived at the conclusion that all cities must be specialized all the time if I'm going to be a pro Civ player and take on the higher difficulty levels. After many odd tests with fairly pure economy types (food, commerce, or hammers) I have decided that specialization is not as good as many people make it out to be. What I have discovered is this:
-all cities need food to grow
-all cities need hammer to build their infrastructure unless you are running universal suffrage and can buy it, or have high food and slavery and can whip it out.
-not all cities need commerce

Therefore, what I do at the beginning of the game, regardless of leader traits is this: I settle cities near good food and hammer tiles. I improve those tiles. These cities will then be able to build/whip the early buildings like granaries and libraries, barracks, etc.
The first specialization decision I make is where to build commerce tiles (cottages). Some cities will not get any. These cities will start to focus on military production (or maybe wonders). Many cities will get cottages to get my economy going.
As the game goes on, most cities probably get some cottages, only a minority of cities are unit pumps.
The second stage of specialization starts when I get access to more national wonders. One city will be my best military city, with the HE. In the cities with the Oxford and/or wall street, I might remove a mine or two and replace it with a cottage.
Finally, if I choose to run universal suffrage, specialization takes a third stage. While running US I can cottage over hammer improvements in my commerce cities. Also, if I am settling cities this late in the game I don't need to build any hammer improvements for them.
Alternatively, if i decide to run state property I will probably remove farms and replace them with watermills. I don't usually put workshops over cottages, and never over towns, but I will start to spam workshops, possibly by cutting down forests.

So, to conclude, specialization is something that is best achieved gradually, for me. All cities start out with their food and hammers and then only select ones will get commerce. Even then, the so called 'commerce cities' will still have farms and mines. Later in the game certain civics, like US or SP might incline me to tweak with improvements and further the specializatino of commerce cities...and national wonders call for increased specialization as well. A final note - I try to plan in such a way that I might remove a farm or mine and replace it with a cottage...not the other way around.

 

[OTAKUjbski]

When you really get down to it, you only need to specialize 3 cities:

Ironworks: Production
Oxford University: Food/Commerce
Wall Street: Food/Commerce/Shrine/HQ

Depending on the map type and your own preferences, you may also want and be able to specialize for the rest of the National Wonders, but it's not uncommon to see some of them combined with the cities above:

National Epic: Production/Food
Heroic Epic: Production
Globe Theatre: Food
West Point: Production
National Park: Food/Forests

Outside the National Wonder cities, you can pretty much do whatever you want. Most cities should have at least 8-12 available to them for infrastructure builds, and many will be purely hybrid in nature.

-- my 2

 

[Rameau's Nephew]

I find that the hardest type of city site to find are the early-mid game production powerhouses. (Late in the game, you can make pretty much any halfway decent site into a good production center.) So those get first priority when I'm scouting out city locations.

Other than that, I'll try to make a GP pump, which is usually a site with lots of food, but nothing else really going for it.

With the exception of an occasional secondary production center, everything else usually becomes a mostly-commerce city by default.

 

[ese-aSH]

the point in early city specification is to avoid building a library in a future military city. or building a barrack in a commerce city.

 

[mercury529]

Yeah, as ese-aSH was getting at, specializing a city does not necessarily mean every tile in the city is dedicated to the purpose of the city (food, commerce, or hammers). Commerce cities grow much faster with access to a food resource. They can also build infrastructure faster with access to food (in slavery) and hammers.

Specializing essentially comes down to looking at a favorable city site, and deciding how to maximize its potential and meet your short and long term needs. It also comes down to avoiding building costly infrastructure that does not serve the purpose of the city.

It sounds to me like you kind of took the advice given here to mean focus exclusively on food/hammers/commerce in every specialized city. I don't think anyone was really advocating that. But it also sounds to me like you learned on your own that the advice given here is solid when implemented properly (i.e. making sure food and hammers are available for infrastructure even in cottage cities).

 

[Winston Hughes]

In the early game, you often have little choice but to use most of your cities for multiple different purposes. So, yeah, generic cities are fine early on. But I really can't see the benefit in leaving them unspecialised in the longer term, unless the land forces you into it somehow.

My experience is that I often regret failing to specialise a city when I had the chance. Comparing the total output (in terms of beakers/gold/units/wonders/etc.), the unspecialised cities almost always sit low down the table. Even if they have good sites and can produce lots of raw hammers or commerce, the actual output is usually much less impressive. And while I sometimes realise that I've given a city the wrong role, I don't think I've ever regretted specialising one.

It's worth noting, however, that specialising a city doesn't mean you never use it for anything else. Some commerce cities will have enough hammers to produce a few troops (and can be whipped/drafted in an emergency). And most production cities will generate some commerce, and, when necessary, can be switched to produce extra beakers, gold and/or GPs via specialists.

And, when we're talking about specialisation, I would take the fact that (nearly) every city needs a certain amount of hammers and surplus food as a given. Perhaps this isn't always made clear, but I guess that's because it seems so obvious to some people that they forget to mention it.

 

[VoiceOfUnreason]

And, when we're talking about specialisation, I would take the fact that (nearly) every city needs a certain amount of hammers and surplus food as a given. Perhaps this isn't always made clear, but I guess that's because it seems so obvious to some people that they forget to mention it.

Actually, I think there's some disagreement on that score, especially when considering cities oriented toward commerce.

Better players than me have dropped some cities into some real junk land (ice, coastal tiles, and a single iced hill to mine? yucko). Horses for courses - I'm just not a good enough player to make that sort of move.

On the production side, the question to challenge yourself with (in my mind) is "how much production do you really need in a non production city?". Yes, if you don't have production, it takes "forever" to get any infrastructure built. So what?

After all, there is no "most infrastructure" victory condition; the development of your cities needs to be contributing to one of your other strategic goals.

However, I don't think that one can reasonably argue that a city with no infrastructure at all is specialized for anything. After all, part of the point is to be juggling with the imbalances among the various forms of income, and that really needs multipliers of one form or another - most of which come from having the right infrastructure in place.


I wouldn't quite say that generic cities in the early game "is fine". Two cities sharing two jobs is not nearly as effective as one doing each job. In the early game, the chores are mostly training units, so the improvements around the city will likely be similar; but the ones training units will take the time to build a barracks, and the cities training settlers and workers won't. Furthermore, you "should" see separation of the capital relatively early in any game where early bureaucracy is a strategic pillar.

 

[Catan_Settler]

It's usually not till mid to late game that all 20 tiles in a cities cross can be totally specialized the way you want them. Many of the production improvements won't be available yet for production cities, and cottages that are on plains are operating at a food deficit until biology. In the meantime you can use those tiles to use the city as a semi-hybrid-- put a mine on that hill, and have someone working it when you need to build the market and bank there. You probably won't have your cities up to 20 pop any time in the classical age even because of happy/health caps, so use some tiles for production to get the infrastructure in place. Later on new techs and civics will make it so that yes, you can cottage over everything and not be in a food deficit, but you ideally want to have all crucial buildings in place before you actually get rid of all the hammers around that city. For production cities it's not so much about that, but more that you probably want to save some forests for eventual lumber mills. And a good early production city probably has lots of hills which provide tons of hammers when you put mines on them, however they are almost always at a food deficit (any tile with less than 2 food is at a deficit that has to be paid for from somewhere else), and for these mine-heavy sites you absolutely must have some way to feed the miners... Be it a food resource, or a couple floodplains with farms on them. An early floodplain farm will feed itself, and a mined plains hill for instance. A nice irrigated wheat on the other hand, could pay for a mined plains hill and mined grassland hill. So basically, it's kind of a balancing act until later in the game when you get stuff like state property, biology, replaceable parts, etc.

 

[OTAKUjbski]

On the production side, the question to challenge yourself with (in my mind) is "how much production do you really need in a non production city?". Yes, if you don't have production, it takes "forever" to get any infrastructure built. So what?

I think low production really compounds over the long haul. Eventually, a city will reach its maximum population, and all of its Cottages will be Towns. At that point, it becomes impossible to improve the city (singularly) without infrastructure.

8 base while still growing is my minimum. This is how I arrived at that figure:

It takes 10 + 20 + 40 = 70 turns to turn grow a Cottage into a Town.

The 'basic' infrastructure of a Commerce City (to me) is: Granary (60) + Courthouse (120) + Library (90) + University (200) = 470

470 / 70 = 6.71 /Turn.

I typically want a Forge for the extra , so my minimum is 8, in order to also get +2 (instead of just +1) from the Forge (bringing me up to 10 production).

I figure if I can get those basic buildings in place by the time my first (or all, depending upon how I'm growing & working the city) Cottage becomes a Town, then I'm doing alright.


Now, obviously, that 'minimum' doesn't have to come from Mines, Lumbermills, etc. Whether they come from Towns or rush-buying (under US), Food (under Slavery) or Corporations (not under SP) doesn't matter. As long as I have 8 base , I seldom find myself in wanting for more.


-- my 2

 

[VoiceOfUnreason]

The 'basic' infrastructure of a Commerce City (to me) is: Granary (60) + Courthouse (120) + Library (90) + University (200) = 470

Which rather exactly illustrates the point - a 200 hammer university? really? And if that university is completed on turn 70 instead of turn 100, are you really making a profit? And then you want to toss another 120 hammers worth of forge on top of that?


Put another way, suppose we were playing a variant game, where in our commerce cities we were restricted to 4 (base) prior to US. What would then follow?

 

[OTAKUjbski]

Which rather exactly illustrates the point - a 200 hammer university? really? And if that university is completed on turn 70 instead of turn 100, are you really making a profit? And then you want to toss another 120 hammers worth of forge on top of that?

As with all things in Civ4, it depends ... primarily on the size of the city and the amount of Cottage-Towns being worked. With cities working a lot of Cottages, losing 1 to work a Mine is often worth it, and the loss is easily made up for over the course of the game.

In the example of a 30 turn disparity over 100 turns -- which is actually quite severe, I think you'd have to be working 10 Cottages to make building a University faster more profitable than working Cottages alone (math done off the top of my head).

So in the case of small cities, building that expensive University faster at the expense of working a Cottage isn't usually worth it.

If that University is holding up Oxford, then it's worth every penny.

Put another way, suppose we were playing a variant game, where in our commerce cities we were restricted to 4 (base) prior to US. What would then follow?

That seems a little one-sided. But I guess that doesn't make the answer any less obvious: You'd work every Cottage you could to work them up to Towns by Democracy (or the Pyramids).


There's definitely a Cost-Benefit curve to this. Too little production, and the buildings basically never get built. Too much production, and the city sacrifices valuable cottage-able tiles.

Prior to Printing Press and Free Speech, it's seldom worth it to invest in a University, because that +25% just doesn't add much to the baseline -- making it considerably harder to 'catch up' to where you would've been in terms of had you just worked Cottages the whole time.


Where I typically get screwed up is in the midgame (ish) when I start running into and problems. In cities I've afforded too little potential production (I don't work it when I don't need it), they are always noticeably below the other cities' caps. In those cases, I'm sometimes working 2 or 3 less tiles than I could be -- which I'm certain affects the bottom line.

 

[VoiceOfUnreason]

In the example of a 30 turn disparity over 100 turns -- which is actually quite severe, I think you'd have to be working 10 Cottages to make building a University faster more profitable than working Cottages alone (math done off the top of my head).

I'd probably be more worried about the slider.

So if you are constructing 470 hammers worth of buildings in 70 turns, and it takes me about 100 turns to do the same, then you've generated about 140 extra hammers during that time, which is basically 70 turns working a mine instead of a cottage. So you need a little bit more than two cottages per turn benefit during that 30 turn sprint? And the university is only multiplying what it gets after the slider?

That seems a little one-sided. But I guess that doesn't make the answer any less obvious: You'd work every Cottage you could to work them up to Towns by Democracy (or the Pyramids).

You'd also be stingy about what you spent hammers on until then.

There's definitely a Cost-Benefit curve to this. Too little production, and the buildings basically never get built. Too much production, and the city sacrifices valuable cottage-able tiles.

Right, but there's also the problem that the hammers go in the denominator. Going from 2 to 4 is a lot more important than going from 5 to 7, even before considering the 25% bonus.

And I think also is the key notion that you need to be thinking in terms or margins - not the yield delta, but rather the delta of the marginal values (also true for the commerce of course - how important is it to work the 10th cottage now when you are already working 9 of them).


One of the problems I have with cottage cities is working out how to manage the surplus food. For the sake of this example, let's assume that the whip is not an option.

So imagine a City with irrigated rice (+5F), and a current population cap of 11 (completely arbitrary choice). At size 1 we work the rice. At size 2, we add a grassland mine to the mix (following the principle that we want our production in multiples of 4). As we continue to grow, we add grassland cottages... so at size 9, we have the city center (+2), the Rice (+5), the mine (+1), 7 cottages (+14), for a total of 22 food. And then? keep working grassland and accept the agree people as our due? change which cottages we are working? move people from the marginal cottages to the mines? Should people be working the mines earlier in those cases where infrastructure will allow us to lift the cap sooner? Stop working the rice (hopefully, some neighbor can work it)?

 

[OTAKUjbski]

I'd probably be more worried about the slider.

I usually assume a 60 or 70% 'peacetime' slider.

So if you are constructing 470 hammers worth of buildings in 70 turns, and it takes me about 100 turns to do the same, then you've generated about 140 extra hammers during that time, which is basically 70 turns working a mine instead of a cottage.

Wouldn't it be more like 200 hammers? ( 470 / 70 = 6.714 * 100 = 671 - 470 = 201) I don't reckon it matters too much.

I only work Mine(s) when I need the production. After the build is complete, the Miners go back to their Cottage-Town(s).

So you need a little bit more than two cottages per turn benefit during that 30 turn sprint? And the university is only multiplying what it gets after the slider?

Grrr. (not at you; just in general: this is the one area of CivMath I hate)

You definitely need more than 2 Cottage-Towns. My rule-of-thumb is not to boost production except when 10 Cottage-Towns are worked, though I've never sat down to 'prove' it's a good rule. (I need Quechua -- he's good at this stuff.)

Where increasing the population cap is concerned, I'm not sure you can ever go wrong increasing the production a little.

Right, but there's also the problem that the hammers go in the denominator. Going from 2 to 4 is a lot more important than going from 5 to 7, even before considering the 25% bonus.

Exactly why there is such a thing as too much production in a Commerce City.

And I think also is the key notion that you need to be thinking in terms or margins - not the yield delta, but rather the delta of the marginal values (also true for the commerce of course - how important is it to work the 10th cottage now when you are already working 9 of them).

As [I think] I mentioned (maybe just not very well), I agree here. Shutting down one of your two Cottage-Towns represents a 50% loss of (assuming equality of both tiles). Whereas losing 1 of 10 is only a 10% loss.

One of the problems I have with cottage cities is working out how to manage the surplus food. For the sake of this example, let's assume that the whip is not an option.

This feels like it's about to go down the road of Question re: new commerce cities ( OMG ... holy math! )

So imagine a City with irrigated rice (+5F), and a current population cap of 11 (completely arbitrary choice). At size 1 we work the rice. At size 2, we add a grassland mine to the mix (following the principle that we want our production in multiples of 4). As we continue to grow, we add grassland cottages... so at size 9, we have the city center (+2), the Rice (+5), the mine (+1), 7 cottages (+14), for a total of 22 food. And then? keep working grassland and accept the agree people as our due? change which cottages we are working? move people from the marginal cottages to the mines? Should people be working the mines earlier in those cases where infrastructure will allow us to lift the cap sooner? Stop working the rice (hopefully, some neighbor can work it)?

I hate working out examples. It's always at 2 AM, and I always make some trivial mistake that makes it all invalid.

I assume you're already @ population 10 or 11 (with ) or time the building and growth to finish at the same turn (i.e., the 'extra' population is accessible the same turn the building finishes).

Also, as soon as the building is done, all Miners go back to Cottage-Towns.

If my 2 AM math serves me well, taking one (or even two) Cottage-Town citizens and moving them into Plains Mines (for 8 & 12 hammers respectively) equates to more baseline commerce.

This is made especially apparent if the building grants +2 to the population cap, since the extra citizen is grown into sooner than if the building is delayed.

Again, assuming my 2 AM math is good, I'd like to point out the difference between 8 and 11 (or 12) base hammers is neglible at best; and taking 4 Cottage-Town Citizens and putting them to work in Grassland Mines (for 16 HPT) actually hurt the baseline commerce -- though if some or all of the mines were riverside, it would likely bring the equation back to parity.

Spoiler my scratchpad ... a whole bunch of nonsense you can interpret yourself ;) :

150 / 4 = 38 Turns to build a Market

7 * 1 * 10 = 70 commerce over 10 turns as cottages
7 * 2 * 20 = 280 commerce over 20 turns as hamlets
7 * 3 * 8 = 168 commerce over 8 turns as villages
= 518 commerce total

150 / 8 = 19 Turns to build a Market

6 * 1 * 10 = 60
6 * 2 * 20 = 240
6 * 3 * 8 = 144
=444

last 19 turns

3 * 1 * 10 = 30
3 * 2 * 9 = 54
=84
;=528 (+10 commerce)

150 / 12 = 13 Turns to build a Market

5 * 1 * 10 = 50
5 * 2 * 20 = 200
5 * 3 * 8 = 120
=370

last 25:
4 * 1 * 10 = 40
4 * 2 * 15 = 120
= 160
;=530 (+12 commerce)

150 / 16 = 10 turns:

3 * 1 * 10 = 30
3 * 2 * 20 = 120
3 * 3 * 8 = 72
=222

last 28:
6 * 1 * 10 = 60
6 * 2 * 18 = 216
= 276
;=498 (-2)

150 / 1 = 150 turns

8 * 1 * 10 = 80
8 * 2 * 20 = 320
8 * 3 * 40 = 960
8 * 4 * 80 = 2560
= 3920

150 / 4 = 38 turns (-112 turns)

7 * 1 * 10 = 70
7 * 2 * 20 = 280
7 * 3 * 40 = 840
7 * 4 * 80 = 2240
=3430

2 * 1 * 10 = 20
2 * 2 * 20 = 80
2 * 3 * 40 = 240
2 * 4 * 42 = 336
=676

== 4106 (+186 .. hmm)

150 / 8 = 19 turns (-131 turns)

6 * 1 * 10 = 60
6 * 2 * 20 = 240
6 * 3 * 40 = 720
6 * 4 * 80 = 1920
=2940

3 * 1 * 10 = 30
3 * 2 * 20 = 120
3 * 3 * 40 = 360
3 * 4 * 61 = 732
=1242

== 4181 (+262)

160 / 16 = 10 ( -140 turns )

3 * 1 * 10 = 30
3 * 2 * 20 = 120
3 * 3 * 40 = 360
3 * 4 * 80 = 960
= 1470

6 * 1 * 10 = 60
6 * 2 * 20 = 240
6 * 3 * 40 = 720
6 * 4 * 70 = 1680
= 2700
== 4170 (+251 ... -11 by comparison to 8)

150 / 4 = 38 turns

7 * 4 * 38 = 1064

150 / 8 = 19 turns

6 * 4 * 38 = 912
2 * 4 * 19 = 152
=1064 (equal .. hmm)

------------

6 * 2 + 20 = 32 / 2 = 16 / 4 = 4.00
10 * 2 + 20 = 40 / 2 = 20 / 4 = 5.00

;@ +4 = 5 turns to grow from 7-10 by +1

150 / 1 = 150 turns

8 * 1 * 10 = 80
8 * 2 * 20 = 320
8 * 3 * 40 = 960
8 * 4 * 80 = 2560
=3920 (+/- 0 baseline)

150 / 4 = 38 turns (-112 turns)

7 * 1 * 10 = 70
7 * 2 * 20 = 280
7 * 3 * 40 = 840
7 * 4 * 80 = 2240
=3430

2 * 1 * 10 = 20
2 * 2 * 20 = 80
2 * 3 * 40 = 240
2 * 4 * 42 = 336
= 676

1 * 1 * 10 = 10
1 * 2 * 20 = 40
1 * 3 * 40 = 120
1 * 4 * 37 = 148
=318
;=4424 (+504)

150 / 8 = 19 turns (-131 turns)

6 * 1 * 10 = 60
6 * 2 * 20 = 240
6 * 3 * 40 = 720
6 * 4 * 80 = 1920
=2940

3 * 1 * 10 = 30
3 * 2 * 20 = 120
3 * 3 * 40 = 360
3 * 4 * 61 = 732
=1242

1 * 1 * 10 = 10
1 * 2 * 20 = 40
1 * 3 * 40 = 120
1 * 4 * 56 = 224
=394
;=4576 (+656 ... +152 vs 4)

150 / 12 = 14 turns (-136 turns)

5 * 1 * 10 = 50
5 * 2 * 20 = 200
5 * 3 * 40 = 600
5 * 4 * 80 = 1600
=2450

4 * 1 * 10 = 40
4 * 2 * 20 = 160
4 * 3 * 40 = 480
4 * 4 * 66 = 1056
=1736

1 * 1 * 10 = 10
1 * 2 * 20 = 40
1 * 3 * 40 = 120
1 * 4 * 61 = 244
=414
;=4600 (+680 .. +24 vs 8)

150 / 16 = 10 turns (-140 turns)

3 * 1 * 10 = 30
3 * 2 * 20 = 120
3 * 3 * 40 = 360
3 * 4 * 80 = 960
=1470

6 * 1 * 10 = 60
6 * 2 * 20 = 240
6 * 3 * 40 = 720
6 * 4 * 70 = 1680
=2700

1 * 1 * 10 = 10
1 * 2 * 20 = 40
1 * 3 * 40 = 120
1 * 4 * 65 = 260
=430
;=4600 (+680 .. +24 vs 8)

Geez .. that was too much at this hour.