How to predict city maintenance costs?

[Olavi]

Hi everyone. I've got a question about city maintenance costs and optimal expansion pace.

As we all know, building and/or conquering cities too fast can slow down scientific research quite badly, as the city maintenance spending goes up too quickly. But is there any way to predict the maintenance costs in advance? If I have, say, two cities and I want to get one more, can I somehow predict the financial effect of the third city before actually building it?

I assume the maintenance spending depends on the total number of cities, their distance from the capital, and the difficulty level. But what is the actual mechanism here?

[Rowick Hepan]

LOL, I got bored at work one day and actually crunched the number on this one.

To put it simply the optimum number of city you should have to keep your science from going down the toilet is based on the average population of your cities. The bigger your cities the more you can and should have.

For the first 4000-4500 years the numbers below work well.


For a non-financial non-organised civ:

Best number of cities = Average City POP +1


For a financial civ:

Best number of cities = Average City POP x 3/2


For a organised civ:

Best number of cities = Average City POP x 2


For a financial and organised civ:

Best number of cities = Average City POP x 3


NOTE: This assumes that MOST of your population are working coastal tiles or tiles with cottages.

I put together a spreadsheet that shows a nice graph of commerce vs number of cities for various city POP's and leader attributes.

If anyone is interested in the spreadsheet or more detailed formula's etc feel free to email me at "nbbuiz at hotmail dot com"

[Strategist83]

See this thread: http://forums.civfanatics.com/showthread.php?t=138473

[rjmatsleepers]

Originally posted by Rowick Hepan
LOL, I got bored at work one day and actually crunched the number on this one.

To put it simply the optimum number of city you should have to keep your science from going down the toilet is based on the average population of your cities. The bigger your cities the more you can and should have.

...

NOTE: This assumes that MOST of your population are working coastal tiles or tiles with cottages.

This is useful, but I suspect I'll need to adapt it for my style of play. Your comment about coastal tiles and cottages suggests that the crucial factor is average commerce rather than average population - is that correct? My problem is that I concentrate on building commerce in my science city, so my newly formed cities will not have cottages for quite a while.

Also, I would have expected the game level to be a factor since this has a big impact on the cost. What level did you do the analysis for?

RJM at Sleeper's

[Rowick Hepan]

This is useful, but I suspect I'll need to adapt it for my style of play. Your comment about coastal tiles and cottages suggests that the crucial factor is average commerce rather than average population - is that correct? My problem is that I concentrate on building commerce in my science city, so my newly formed cities will not have cottages for quite a while.

Also, I would have expected the game level to be a factor since this has a big impact on the cost. What level did you do the analysis for?

I did the analysis at Monarch level - I hadn't thought about game level difference but from a few other posts I've now read it does appear to make a small difference.

The crucial factor IS commerce rather than population - population is just much simpler to keep an eye on.

My prefered style of play is to build about 1- 1 1/2 workers per city in the early stages of the game to chop-rush and build tile improvements. In the early game I usually have one tile per city working a resource tile such as wheat, cattle etc. for the food but otherwise concentrate on commerce.

This post assumes that non-financial civs will be producing two commerce from most tiles and that financial civs will be producing three.

More commerce / tile will increase the number of cities you can support whilst less commerce / tile will decrease the number you can support.

[rjmatsleepers]

I had a look at an old save at monarch, playing as Lizzie (financial and philosophical)

I had 4 cities with populations 8, 6, 3 and 1. Their respective commerce were 61, 9, 5 and 8. The largest includes 50% boost from bureaucracy - I suspect I should discount this. The 3 smaller cities were each getting 4 commerce from trade; 2 of them were getting 3 more commerce by working a water special.

I can work out an "adjusted population" (ie what population would give the same commerce if I got 3 commerce per tile. This would be roughly 14, 3, 2 and 3. The average is grossly distorted by my high commerce city but comes out at just over 5. Either way, the "optimal" number of cities would be in the range 6 - 8. ( (Should I round my adjusted population up, down or not at all; what should I do about trade?)

However, I am already running a deficit of 5 per turn at 100% science. 2 more cities would give me about 10 more commerce, but force me to reduce my science rate. On the face of it, this is OK since the additional commerce is greater than I would lose by a 10% cut in the science rate. However, a large chunk of the reduction would come in my science city which had good science multipliers and no gold multipliers.

As I said at the beginning, I suspect that style of play makes a lot of difference. In particular, if you build a couple of production cities, this hits your cost, but doesn't give you much benefit in commerce.

I think I'll try to use worldbuilder to experiment a little to see what the equivalent formulae are for my style of play.

RJM at Sleeper's

[Paddy]

Most interesting indeed

[Shaka II]

Very interesting an nice work. What we more or less suspected but its nice to have this quantified. I just happen to be playing as Washington in my present game.

[fed1943]

Yes,very interesting and nice work.
I also tried it with Washington.
Hepan,in your first post you "most of pop.working cottages or coastal".Do you want elaborate a little more replacing the "most"to a number,please?So that it would be the number of cities,the number of pop. and the number of tiles that give more than 1 commerce.
Best regards,

[Rowick Hepan]

Hepan,in your first post you "most of pop.working cottages or coastal".Do you want elaborate a little more replacing the "most"to a number,please?So that it would be the number of cities,the number of pop. and the number of tiles that give more than 1 commerce.

Some numbers I have used in my analysis:

Difficulty Level: Monarch
Number of Commerce / tile: 2 (non- financial Civ) or 3 (financial Civ)
Number Tiles / City not producing average Commerce as above: 1
Average Population: Works for any.

So for a rough example:

1) Lets say a non-financial non-organised civ has 4 cities with an average POP of 3.
2) On average each city has three of it’s four POP working commerce tiles producing an average of 6 commerce per non-capital city and 14 commerce for the capital (due to +8 commerce for palace.)
3) Total Commerce = approx 32.

4) Maintenance for number of cities is approx 4 x 3 =12.
5) Maintenance for distance from capital is approx 3 x 1 =3.
6) Total maintenance = approx 15.

7) Net Commerce gain = approx 17.

There are a lot of additional variables that I havn’t accounted for as yet such as:

- Difficulty level modifiers to maintenance costs.
- Civic or Wonder effects that alter the amount of commerce produced.
- Banks, marketplaces etc later in the game - that tend to offset maintenance costs.
- The increased value of villages / towns later in the game.

[fed1943]

Thank you,Hepan.I will try in that ways:
a)Cross it with food to see the possibility;
b)Cross it with hammer to see,well,the prodution.
Example:2 flood plains(cottaged)and a metal(mined)-everything all right;proved.
But,without the metal...I don´t know,must try.
Do you want to try it,too?
Best regards,

[spammurabi]

I think an intersting related question is how the computer calculates the maintenance cost for City Number. Does this affect all cities equally? When does it increase? Every city, every 3 cities? Does population affect it? I think this is probably at the heart of the origin of maintenance.

Distance from Palace is self explainatory, but City Number is a little confusing. Anybody have a formula for calculating it and how it applies to empire-wide mainenance as a whole?

[fed1943]

There is a thread in CivFanatics explaining about maintenance.
Best regards,