• Civ7 is already available! Happy playing :).

small farms

kaskavel

Warlord
Joined
Sep 11, 2023
Messages
103
It occurs quite a lot in my present archipelago game. Small, very far away, very desperately corrupted island cities with water tiles as the basic terrain. They do have food bonuses from fish, irrigated grasslands, etc, up to a +6 including the city centre. Got granary, temple, harbor, wondering now about the aqueduct and I think...what is the point? The city can go up to size 23, but size 6 or size 23 both mean 3 scientists. What is best?
1. Destroy the temple (it expanded already), the granary (which I shouldnt have built if this is the correct option), leave the city at size 6 with 3 scientists and production set to wealth
2. Keep paying temple, granary plus aquedect, hospital, hoping that it will pay off because...what? Some commerce may occur by sheer size?
 
It depends on your goal(s) as to what will work out best. 1. is likely better for research. 2. is likely better for eventual research output from that city. 2. would also work better for score if you played the game out until 2050 AD.

1. works out better for producing research more quickly. It would produce 10 beakers as soon as set up (since you have rails). Wealth might be good, but even corrupt cities like that can often produce a worker or settler well before the game will end. Then the worker or settler can help set up another specialist farm.

2. In contrast would have commerce from 20 (not 21 from the sound of it which is the maximum number of worked tiles) worked tiles. If running a Republic or Democracy, that's at minimum 40 corrupt commerce. Since 12 total commerce always gives one uncorrupted commerce, that's 3 uncorrupted commerce at least, though since you have fish it sounds like some coast squares come as available also, so maybe you would have 4 uncorrupt commerce. About 10% could go uncorrupt from each relevant building. So, that's 3 + 4 + 4 = 10 uncorrupted commerce. 10 + 5 (library) + 5 (university) + 5 (research lab) = 25 commerce. At 50% research, that's at least 12 science + 9 from scientists for 21 total science given you somehow would manage to get a library, university, and research lab built.

So 2. can produce more science for that city, but with a settler or three coming from that city also for another similar specialist farm, then we can have 10-20 science eventually on top of the original 10 we had for 20-30 beakers (if not 40 beakers). Thus, 1. (with the city producing a settler or worker instead of wealth) better fits a plan to have more research overall (though not if you set the city to wealth and leave it at wealth) and produces research faster in any case.

But, if you seek to go for a high score at 2050 AD, then the number of citizens on worked tiles which are happy or content has more relevance. In which case, 1. ends up better eventually. Assuming you have the domination limit victory condition on, though keeping the temple could make one citizen content, keeping the temple could also have the drawback of eventually leading to land or coast tiles in one's cultural borders that no city can work. Citizens can get made content or even happy sometimes by use of entertainers or the luxury slider (especially if a city has a courthouse, police station, and/or commercial dock). So, selling off the temple before the city reachs 100 culture can make sense.
 
It occurs quite a lot in my present archipelago game. Small, very far away, very desperately corrupted island cities with water tiles as the basic terrain. They do have food bonuses from fish, irrigated grasslands, etc, up to a +6 including the city centre. Got granary, temple, harbor, wondering now about the aqueduct and I think...what is the point? The city can go up to size 23, but size 6 or size 23 both mean 3 scientists. What is best?
In the long run the best approach is to let it grow to use 20 tiles plus the city tile. If it is 5 commerce from the city tile of the metropolis and 50 from the 20 other tiles that is 55 commerce prior to corruption. Deducting 70% = 38.5 = 38 corruption this leaves with 17 commerce after corruption. That is a science output of up to 17, 25, 34 or 42(lib,uni and research lab) in addition to the 9 science from 3 scientists. A temple is no use, you do want the library instead. A marketplace with 8 luxuries is all the happy faces you need.

What is best in the long run may not be the best approach, though. The approach to build workers or settlers may better suit your strategy. It really depends on the context. Having to build courthouse, police station, aqueduct, hospital, harbour, commercial dock, oil plattform, marketplace, library, uni and research labs takes up some investment, possibly in form of disbanding expensive units built elsewhere. Also you may need a granary to grow, or some workers or settlers to join to get to size 23. All that investment takes some time to pay off. It like to go the long road, but that is not completely rational.
 
Thank you both. So, I guess we generally tend to build them up if we got them early in the game.
 
in really nearly any circumstances: NO! ;)

justanick formulates a hypothetical optimum, which is practically never to reach, ore even to be desired, in any normal game. you simply won´t ever get a return on that investment.

the usual optimum goes: have your farms max either at size 6 (or below, if not enough food for working citizens + specialists); or at size 7+ (if fresh water is available, for unit support!). you don´t want ANY buildings in those farm towns, only a respective optimum number of specialists (=usually: scientist).

t_x
 
If I get to a point where mass-farming has become viable, my general rule (since I never play for 100K or Histo wins, and have come to hate CxCxC city-pox!) is to Settle farms at CxxC, and strictly limit buildings.

Circumstantial:
Walls on borders (maybe)
Harbour near Coastal Fish

Assuming Pop7+ is reachable, and terrain will allow ~1/2 of those citizens to Geek
Courthouse to help build + pay for any other buildings
'Duct if no freshwater
Market for extra Happies (+ Tax-income from decorrupted Commerce, when the Sci%-slider is reduced on the last 1-2 turns of research)

Otherwise, where farms can't reach Pop2+ without starving (e.g. Tundra without Game/Fish, or Desert without Oases or SteamPowered-AGRI):
None

Any buildings I do put up will be built usually (but not necessarily) in the above order, using whatever combo of chops, disbands, cash-rushes or (post RepParts) CE shields is most convenient, and gets those buildings up reasonably quickly without bankrupting me -- or frazzling me with endless micromanagement.

I usually don't build any Culture unless I 'need' a border-pop -- and even then, usually only if SCI or REL. For continent-wide border-expansion, I'd prefer to capture Artemis in the early game, or build/capture The Internet in the late game.
 
Thank you both. So, I guess we generally tend to build them up if we got them early in the game.
Not quite. As templar_x correctly points out it is unlikely to pay off before winning the game. Having many expensive units ready to disband usually only happens in later stages of the game. Getting 1360 shields from disbanding takes up 5440 shields from building those units. That could alternatively be spent on wealth resulting in up to 2720 gold. It takes about 100 turns to break even in that regard.

If however you spend those 5440 shields or 68 cavalry to conquer more territory you winning the game will be expedited. Essentially the long road only pays off by deciding to delay winning.
 
Thank you both. So, I guess we generally tend to build them up if we got them early in the game.

Not generally, no, I haven't. I think I've talked to people elsewhere who do. It's probably more consistent with the designers intent, since even CxxC specialist farms were not intended as optimal (Soren Johnson called something significantly less strict than CxC spacing ICS... and they didn't want ICS as optimal for this game).

It may seem tempting to think that it will pay off to eventually have a better research rate in say the modern age. Maybe you believe that might make winning a spaceship launch game easier? But, even then a whole cluster of science specailist farms can outdo wider spaced cities with a whole lot of corruption. And as a reminder, research has a strict limit of 4 turns at minimum, thus more science only helps with later research, but that stops applying at some point in the modern age. The timing of when more infrastructure actually can pay off for research in highly corrupt areas is close to when there is very little, if anything, left to research of any consequence.

I found a size 23 city from my recent Maya game:

23 size.png


I adjusted the science slider from the final save so we have 70% of commerce going to research. Though Sid level makes extreme corruption easier to find, it still often can get found on lower levels. The city produces 15 science from specialists, but only 11 from the city! At 100% science it would produce 16 from the city. The citizen's research barely outdoes the scientists without any multiplier buildings!

Now, of course, one can point out that if it has a commercial dock, it would produce more science. But, the cost of the courthouse (80), police station (160) (how did one get Communism... was it worth it?), and commercial dock combined (160) equals 400 shields. That's 5 specialists civil engineering those buildings in 38 turns (8 for the courthouse, 15 for the other two) AFTER the hospital and aqueduct get built.

More cities getting up scientists earlier likely ends up better than two. Two cities at size six with 3 scientists each produced at least 20 science. And if we had a third, because of 2 earlier settlers from the city, that city producing settlers instead of infrastructure early could work out better for research purposes even if late game.

Science specialist farms are very simple to setup (no infrastructure, just get them food, and make scientists) and I think @BlackBetsy would agree with them as fun to run too. The only issue might be that one might end up with "too many of them", but the game thankfully is hard coded to never allow for more than 512 cities also.
 
Science specialist farms are very simple to setup (no infrastructure, just get them food, and make scientists) and I think @BlackBetsy would agree with them as fun to run too. The only issue might be that one might end up with "too many of them", but the game thankfully is hard coded to never allow for more than 512 cities also.
Yes.

I generally will NOT build any infrastructure in corrupted places. If you can, say, generate 3 beakers per turn, I will build (buy) you a library, especially if I am scientific and they are cheap. If they have an fresh water sources, then I will let them grow to 12 and run 6-7 scientists instead of 3. (2 food city center + 4 food from 5 tiles + maybe a food bonus from one of the 5 is enough for 7 scientists).

With ICS you can create as many of these 6 citizen towns with 3 scientists based on irrigated, railed grasslands (C is City, C1 is City 1, W is worked tile, W1 is tile worked by City 1):

1734106465665.png


If you were going to produce any infrastructure in these farms, you are just going to burn cash. In a 100K game, you are obviously going to build cultural stuff, hopefully by cash rushing or by disbanding units + using civil engineers to create shields.

ICS won't work for a Histographic game where you are trying to max citizens / happiness, since you at best can have 3 tiles worked and 3 specialists except where you hit the coasts. What you want is 6 happy citizens. Also, new cities consume 2 tiles, so for Histographic you want fewer cities with more citizens. Optimally, it's 20 tiles per city (the fat cross) which requires the C----C----C spacing vs. the ICS C-C-C spacing. That said, C----C----C spacing is something I've never done.
 
But, even then a whole cluster of science specialist farms can outdo wider spaced cities with a whole lot of corruption.
When we are talking about land that is sensible. But if a read kaskaval correctly, then we are talking about a lone island where you cannot found more towns. This shifts the balance of reasoning in favour of going for size 20+ compared to the more usual case you describe.
Now, of course, one can point out that if it has a commercial dock, it would produce more science. But, the cost of the courthouse (80), police station (160) (how did one get Communism... was it worth it?), and commercial dock combined (160) equals 400 shields. That's 5 specialists civil engineering those buildings in 38 turns (8 for the courthouse, 15 for the other two) AFTER the hospital and aqueduct get built.
Not having built a library is rather inconsistent. If you go the long road you should not stop in the middle of it. Only the combination of multiplyer buildings and anti-corruption building together makes truely sense.
ICS won't work for a Histographic game where you are trying to max citizens / happiness, since you at best can have 3 tiles worked and 3 specialists except where you hit the coasts. What you want is 6 happy citizens. Also, new cities consume 2 tiles, so for Histographic you want fewer cities with more citizens.
I cannot follow that reasoning. You want all tiles worked and as many happy citizens as possible. But distributing them over more towns should not hurt a Histographic game. The extra food from the city tile may actually help it.
 
I cannot follow that reasoning. You want all tiles worked and as many happy citizens as possible. But distributing them over more towns should not hurt a Histographic game. The extra food from the city tile may actually help it.
It's because city tiles take away a citizen. A size 1 town works 1 tile plus the city center, for 2 tiles consumed for 1 citizen. A size 2 town works 2 tiles plus the city tile, for 3 tiles consumed for 2 citizen, etc.

Imagine you have 42 tiles to work, arranged perfectly as two big fat 21 tile crosses.

If you put cities in the middle of the two big fat crosses, you can have 40 happy citizens (+80 per turn), plus as many specialists as can be produced per turn.

1734109005060.png


If you ICS that same territory, you can have 10 cities (colored for cities / tiles worked. Each city has a border, and the colors = cities those tiles are worked for).

1734109552982.png

Here, since you have 10 cities, there can be a maximum of 32 tiles worked (42-10 city tiles) and therefore 32 citizens * 2 happy citizens = max score of 64 + however many specialists the terrain affords you to create with excess food.

In reality, maps don't create territories just like that (and there are sea squares), but its true that for any given grid/ set of tiles that fewer city tiles consumed = more citizens.

And that doesn't get into the fact that if you use a grassland for a city tile (2 food for non-ag civs), you might be reducing food available by 2 per turn, because you could irrigate/ railroad that tile to produce 4 food. That difference is 1 specialist, which is +1 score per turn. Which is why in a histo game, you should found cities on hills or other areas (tundra) where the city tile's 2 food actually increases the overall level of food.
 
When we are talking about land that is sensible. But if a read kaskaval correctly, then we are talking about a lone island where you cannot found more towns.

Kaskaval did mention irrigated grasslands. Yes, that could mean no more cities as possible to build. However, he says that the maximum size of the city will eventually be 23. I think we'd need a picture to resolve this.

This shifts the balance of reasoning in favour of going for size 20+ compared to the more usual case you describe.

It does shift the balance of reasoning for that city. And if we had a sufficient amount of strict islands where only one city is possible it could for the empire. Maybe if we have an 80% water map? I don't know... I definitely do find islands with only 1 or 2 squares, and I can imagine one with 3 and a mountain. But, we have to have enough those cities to shift the balance of the reasoning for research with respect to the empire.

Not having built a library is rather inconsistent. If you go the long road you should not stop in the middle of it. Only the combination of multiplyer buildings and anti-corruption building together makes truely sense.

My experience is that with enough irrigated tiles and rails and hospitals eventually 6-7 turn research of future tech becomes possible even on Sid. By all means load the 2050 AD save from the referenced link above to check the research rate. Of course some libraries could make increase the research rate. But future tech counts very little with respect to score, and more future techs barely add any score at all. Keeping too many around also could be dangerous for finishing a histographic game as @superslug 's recent 130K game showed ... he tried to finish in 2050 AD by histographic, but finished in 2045 AD. So, to what end would having the libraries serve?
 
My experience is that with enough irrigated tiles and rails and hospitals eventually 6-7 turn research of future tech becomes possible even on Sid.
Yes, depending on details it may even be closer to 4. The challenge there is not winning by any other relevant measure.
But future tech counts very little with respect to score, and more future techs barely add any score at all.
That is rather true. To have any significant effect on the total score you probably would need to have 10+ future techs 100+ turns before the game ends. Else the effect is hard to even notice. So for any realistic histographic game it is not suitable. Pumping up the luxury slider wins there.
So, to what end would having the libraries serve?
Having libraries, unis and research labs in fully developed metros will have a minor effect on space victories. Due to the 4-turn-limit this tends to only become relevant at Sid and there it will only save a few turns. Still, it pleases my mind to fully develop a civilization. So i guess that is the real merit.
 
All this ICS does not create problems with OCN?

OCN is determined by map size alone. The actually relevant figure is Nopt, which depends on the anti corruption buildings in the respective city. Let us first make a bunch of assumptions to assist discussing the problems related to rank corruption.

-Standard map size => OCN = 20
-Emperor difficulty level => L = 80
-Commercial Civ => c = 0.25
-Republic government => Gr = 0.1
-courthouse and police station built => Ni = 2
-FP built(and reasonably close to the capital) => Nwe = 1

This results in Nopt = 37,6. This means that in addition to the capital there are 37 cities with rank corruption below 50%. Those are the 37 cities closest to the capital. The next 4 more cities have a rank corruption below 60% and thus likely a total corruption of less than the corruption cealing of 70%. So in total there are 41+1=42 cities in the core and the semicore. Everything further away from the capital will have maximum corruption anyway. There in the periphery corruption is maxed out anyway, therefore ICS does no relevant damage to corruption.

But closer to the capital than the 70% threshold going for fewer but better cities or rather metros does make sense due to rank corruption and also due to buildings maintenance(up to about 30 gtp) and possibly also due to city tile yields and free unit support in some governments.

In the periphery buildings maintenance and city tile yields still matter. And also free unit support. But for a republic this may matter little, in fact cities instead of metros may make sense there if optimizing for free unit support alone.
 
The next 4 more cities have a rank corruption below 60% and thus likely a total corruption of less than the corruption cealing of 70%.
Where does the number 4 come from?
There in the periphery corruption is maxed out anyway, therefore ICS does no relevant damage to corruption
When DOES it make damage? Since in this setting you will always have 37 uncorrupted cities, only thing that matters is to keep those cities big and not found farms between them?
 
Where does the number 4 come from?
By checking the math. At R = Nopt rank corruption doubles its growth. A rank of 110% of Nopt equals 60% rank corruption. Of course R is an integer, while Nopt is not.
When DOES it make damage? Since in this setting you will always have 37 uncorrupted cities, only thing that matters is to keep those cities big and not found farms between them?
Almost. Not having lesser cities closer to the capital than the 41 ones that should be closest to the capital is the idea. Maybe even 42. That would be a rank corruption of 61.7% and thus a total corruption of probably slightly above 70%. But a few policemen can reduce that total corruption below the maximum corruption of 70%. So there is a grey area where there is still some potential damage remaining, but you may not realize this damage if you donnot use policemen. If however you need more than ~2 policemen to lower total corruption below maximum corruption, then it probably is not worth it.
 
You mean when the first policeman does nothing? If the first policeman does nothing, I typically consider the city to be beyond salvation. Max corrupted. I do not try to salvage production or commerce from that one anymore. Am I wrong?
 
Top Bottom