Farms vs. Cottages: a simple simulation

[RemoWilliams]

This topic comes up so much in these forums (especially the Strategy and Tips section), I thought I'd run a quick simulation in worldbuilder to give some semblance of a scientific answer.

Edit Note, this test is only reasonably applicable to noble difficulty, as we can easily get a large population (up to 9-10) in the BC era. On higher difficulties, happiness limits are very low in the early game, and therefore the simulation bears little resemblence to reality on higher settings.

This is a pretty contrived test, as most tests in worldbuilder are, but I think the results are striking enough to have some small meaning.

The question is, is it better to build farms for growth, and then work cottages, or to just spam cottages from the beginning, so that the cottages can grow into hamlets/villages/towns faster?

I've got a bias here, I've always intuitively thought that it was better to focus on growth first, and then commerce/science. Hopefully my experiment was not too heavily influenced by this bias, but I thought I should mention it.

Without further ado, here are the experimental procedures and results:

Settings

Difficulty: Noble
Victories: All disabled
Opponents: None
Civ: Inca

Both cities in 4000ad

Science Slider: 0%
City builds: 1 warrior, then nothing

- Globe theater
- Granary
- Aqueduct
- Hospital
- Recycling Center
- Automated citizens
- All grassland
- Every tile next to a river

My reasoning here was that I was too lazy to set up routes to the basic happy/health stuff that I typically have by 1000BC (enough for 9/9 happiness, and 8/9 health).

Farm/Cottage City

- 10 farms, and 10 cottages to start
- Max growth until 6 population
- Max commerce after that
- Turn on limit growth at 9 population (I think this is not unreasonable for the early game)

Cottage City

- 20 Cottages
- Max commerce

Results

I think this is a pretty good test, but if someone can think of a fairer test that is relatively easy to perform, I'll be happy to take it on.

1990 BC

Farm/Cottage city: 8 pop, 3321 gold
Cottage city: 6 pop, 2809 gold

990 BC

Farm/Cottage city: 9 pop, 6652 gold
Cottage city: 9 pop, 5522 gold

I hesitate to call these results conclusive, but I think they go a ways to bearing out my intuition.

[RemoWilliams]

I had to leave the start files off of the original post, because of the 5 file limit

[VirusMonster]

Great idea to test it out!

Can you extend the experiment for 500BC-1AD-500AD-1000AD? And can you explain why you switch to max commerce after 6 city size, but not 10 or some other number? What makes 6 the magic number in your eyes?

I think there is a sweet spot between, farming and cottages, but I am not sure whether it is 10farms/10 cottages. Can you try with 8 farms/12 cottages? 6 farms/14 cottages? 12 farms/8 cottages? Or even ifyou will not try, why do you think 10/10 ratio should work best?

Also, is there any difference for a Financial leader for this test? If you make all squares near rivers, you get +3 commerce with cottage and thus you might prefer cottage first instead of growing first.

[DaveMcW]

So you proved that given infinite workers and grassland, it's better to farm before working the cottages.

The situation is quite different at 4000BC in a real game.

[RemoWilliams]

Originally Posted by VirusMonster

Great idea to test it out!

Can you extend the experiment for 500BC-1AD-500AD-1000AD? And can you explain why you switch to max commerce after 6 city size, but not 10 or some other number? What makes 6 the magic number in your eyes?

I think there is a sweet spot between, farming and cottages, but I am not sure whether it is 10farms/10 cottages. Can you try with 8 farms/12 cottages? 6 farms/14 cottages? 12 farms/8 cottages? Or even ifyou will not try, why do you think 10/10 ratio should work best?

Also, is there any difference for a Financial leader for this test? If you make all squares near rivers, you get +3 commerce with cottage and thus you might prefer cottage first instead of growing first.

I chose 6 because on Noble difficulty, 6 is the max happiness cap for a while, so in a typical real game, that's around the point where I would stop maxing growth.

Since I only grew to 9 pop, anything beyond 10 farms/cottages wasn't tested.

I'm absolutely certain that you are right that there is a sweet spot between farms and cottages, this was a simple test. I've seen numerous postings where people assert without evidence that it is better to work cottages first, and even people who say they never get to their happiness cap, as though towns are a magical panacea that will make them win the game.

One point of this test was to show that even without working any cottages at all until many turns had passed, you end up with more money if you have a faster growing population.

10/10 is for sure not a good ratio. The point wasn't to show the optimal path, that's highly situational. You'll never have a city with all grassland and every tile a river tile. In fact, there are some who would say that if you did encounter such a city, the best thing to do would be to put no farms at all, because you have the ability to grow to 20 without them. This guide is a good example: http://forums.civfanatics.com/showthread.php?t=158482

That's a great guide, and I learned a lot from it, but just enough farms to grow to 20 is a simplistic habit you should grow out of once you get more of a feel for the game.

I think stopping at 1000 BC is a good idea, because, flawed though the simulation is, up to 1000 BC it's not as much of a stretch. Past that point and we definitlely start to drift into fantasy land.

[RemoWilliams]

Originally Posted by DaveMcW

So you proved that given infinite workers and grassland, it's better to farm before working the cottages.

The situation is quite different at 4000BC in a real game.

I totally agree. But I don't think the sim is as flawed as you make it out to be.

The point is that if you have a lot of grassland and flood plains, it's better to cap your population before you work the cottages. In the farm/cottage city, I didn't work a single cottage until I got to 6 pop, yet by the end, I had 1000 more gold.

In a real city, you'd likely have better tiles than grassland for farming, so in that sense, I think the sim was actually a little biased against my point of view. If you had 2 specials, and 2-3 flood plains for farming, as you would in a typical prime city location, I contend it's better to get those up, at least until you max pop, than to spam cottages first and make sure they grow at the expense of pop.

I think for most advanced players this is understood, even a no-brainer. But a lot of newer players question this concept constantly, and I thought I'd run a sim.

Edit: and of course, I forgot to mention the slavery benefit, which is the real kicker, IMO. Having the food resources up first allows you to whip the slaves more often, and grow quickly back to the max. If the 1.61 nerf to tree chopping didn't turn you into a slave-driver, nothing will.

[DaviddesJ]

You obviously should have at least a couple of farms if you have zero food resources, so this isn't much of a comparison. On the other hand, if you have two food resources, plus some floodplains, then I'm very dubious that you want to farm the floodplains as opposed to cottages. In a real game (at significant difficulty, say Emperor) you're just going to end up with more food than you can use, and the lost commerce is significant.

[MestreLion]

Originally Posted by DaviddesJ

You obviously should have at least a couple of farms if you have zero food resources, so this isn't much of a comparison. On the other hand, if you have two food resources, plus some floodplains, then I'm very dubious that you want to farm the floodplains as opposed to cottages. In a real game (at significant difficulty, say Emperor) you're just going to end up with more food than you can use, and the lost commerce is significant.

I think the point of the OP is... if you have food resources and/or foodplains, that would only change the 10/10 ratio between farms and cottages. You would build less farms so you DONT end up with more food than you can use. That way you would have more cottages sooner, getting even more gold.

So his point of growning first (only till cap, no waste), then working the cottages is still valid, even reinforced by having foodplains and resoures.


As a side note: a city with all grasslands and rivers? Possibly with resources and foodplains? WOW, if its past 1AD, i wouldnt even bother to make cottages at all... i would farm everywhere, get 30+ pop, caste system civics, tons of specialists, thus turning the city in the sweetest GP farm ever!

[DaviddesJ]

Originally Posted by MestreLion

So his point of growning first (only till cap, no waste), then working the cottages is still valid, even reinforced by having foodplains and resoures.

The problem is that he's comparing something reasonable (lots of farms) with something unreasonable (zero farms and zero food resources). Obviously the former is going to come out better.

[RemoWilliams]

Originally Posted by DaviddesJ

The problem is that he's comparing something reasonable (lots of farms) with something unreasonable (zero farms and zero food resources). Obviously the former is going to come out better.

It's obvious to some, but definitely not obvious to everyone.

Remember that I didn't work a single cottage in the farms city until I reached 6 pop, whereas I was working cottages constantly in the cottage only city.

A lot of people think that the best thing to do is work cottages first so that they'll become towns more quickly.

Furthermore, according to some guides and articles, the best way to farm is to build only exactly enough farms to reach size 20, and no more. In the cottage city, the number of farms needed for size 20 was zero. However, we saw a large benefit from the farms in the farms/cottages city.

If I proved nothing else, I at least proved that the "just enough for size 20 pop" rule is not a hard and fast rule. Hopefully we can at least agree on that. I get the feeling you think I've drawn a whole bunch of conclusions that I didn't draw. I was very careful in the article to point out that the results were merely interesting rather than being conclusive proof of anything.

I think one of the issues here is that you play emperor whereas the article applies to noble difficulty. Noble starts with a 6 happiness cap in the capital, and that probably makes a significant difference in terms of opening strategy when it comes to which tiles you improve and how. I wouldn't know, I don't play emperor.

Stay tuned, I'm going to try to devise a different test that will be more realistic.

[DaviddesJ]

Originally Posted by RemoWilliams

A lot of people think that the best thing to do is work cottages first so that they'll become towns more quickly.

I think that is usually correct. But that's because most cities (in the early game) have one or two food resources.

Furthermore, according to some guides and articles, the best way to farm is to build only exactly enough farms to reach size 20, and no more.

If there's one thing that's absolutely clear, it's that what happens when your city reaches size 20 is completely irrelevant to the early game. You can always modify your improvements later, when your cities start to get that big. (The midgame generally sees plenty of workers with not much to do.)

[Zombie69]

Originally Posted by RemoWilliams

Stay tuned, I'm going to try to devise a different test that will be more realistic.

If you want to see something realistic, try the following examples :

1. Capital limited to 4 pop at high difficulty level, which has 2 food resources making it have +8 food (+2 base and +3 from each resource).

2. Non-capital city limited to 3 pop at high difficulty level, which has 1 food resource, making it have +5 food (+2 base and +3 from the food resource).

3. and 4. Try the examples above, but at +1 happiness thanks to a happiness resource or a religion.

I think you'll find that in each case, it's extremely dumb to build even a single farm.

These cases encompass 90% of all cities in a typical game of mine.

[Zombie69]

Also, building a granary makes the case for farms even worse, and granaries are usually the first building i make in every city, because of how overpowered pop rushing is.

[Tech Step]

you forget that having a couple of farms worked and using slavery kills all the strats that you have posted.

whipping your pop is the only way to win in the early game.

[RemoWilliams]

Originally Posted by Zombie69

If you want to see something realistic, try the following examples :

1. Capital limited to 4 pop at high difficulty level, which has 2 food resources making it have +8 food (+2 base and +3 from each resource).

2. Non-capital city limited to 3 pop at high difficulty level, which has 1 food resource, making it have +5 food (+2 base and +3 from the food resource).

3. and 4. Try the examples above, but at +1 happiness thanks to a happiness resource or a religion.

I think you'll find that in each case, it's extremely dumb to build even a single farm.

These cases encompass 90% of all cities in a typical game of mine.

Not everyone plays deity with the 4 pop cap, myself included. I prefer to play noble, and as the settings section of my article indicates, that's what this article is about. Why deity players feel the need to pop up in every article about lower difficulties and explain how it doesn't apply to higher difficulties is perplexing to me. You already know that it doesn't, and I never claimed it did.

On Noble you can easily have 10 happiness in the BCs, and that's what this article is about.

[DaviddesJ]

Originally Posted by RemoWilliams

I prefer to play noble, and as the settings section of my article indicates, that's what this article is about.

I don't think that's really clear from what you wrote. I thought the Noble difficulty level was one of the simplifying assumptions for a "contrived test". Not the level at which you were actually giving strategy.

[RemoWilliams]

Originally Posted by DaviddesJ

I don't think that's really clear from what you wrote. I thought the Noble difficulty level was one of the simplifying assumptions for a "contrived test". Not the level at which you were actually giving strategy.

Fair enough. I thought it was clear from the difficulty setting, and from the fact that I said 9 pop was reasonable for the early game, and that we were limited to 6 happiness in the early game, etc. But I've added an edit to be sure there's no further confusion.

[ownedbyakorat]

Originally Posted by RemoWilliams

Not everyone plays deity with the 4 pop cap, myself included. I prefer to play noble, and as the settings section of my article indicates, that's what this article is about. Why deity players feel the need to pop up in every article about lower difficulties and explain how it doesn't apply to higher difficulties is perplexing to me. You already know that it doesn't, and I never claimed it did.

On Noble you can easily have 10 happiness in the BCs, and that's what this article is about.

Well, if we're discussing Civ strategy, then the viability of different approaches at different levels is fair game. I'm no diety player (trying emperor now, getting butt kicked) but I understand that what works on Diety will work on lower levels, but the reverse is not true. So for a player that wants to get better, it's not good to invest in learning a strategy that is not scalable with levels.

[malekithe]

Originally Posted by RemoWilliams

Not everyone plays deity with the 4 pop cap, myself included.

Just wanted to point out that the 4 pop cap starts at emperor, not deity. It's very applicable to most of the games I play. In my opinion, writing strategy guides for noble-level play is a bit short-sighted, unless it's some sort of variant style of play (eg. how to win without building a single unit more advanced than a warrior). I think the majority of players who read these articles are doing so in an attempt to move up to a higher difficulty level. The players who read these just to become more dominant at lower levels, I believe, are in the minority. As such, teaching players tips and tricks that will not be applicable once they move up in difficulty is not terribly beneficial.

The thinking that every city should be setup optimally to reach size 20 was born out of people's dabblings at lower difficulty levels. Your article plans to reach size 10. That's very reasonable for noble level play. But, just at you scoff at those who give advice on how to best reach size 20, those who play above noble think it's inadvisable to give advise on how to reach size 10 from the beginning. If you want to plan for higher levels of play, consider an initial max pop of 4-6.

For the same reasons that you shouldn't heed (or give) advice along the vein of, "Build stonehenge, then the oracle, then the pyramids, followed up by the parthenon..." You shouldn't heed advice that goes like, "Do this while growing to 10 population in the beginning." You're building a crutch into your game. Once you move up in difficulty, you won't have the luxury of either.

[Naismith]

Very interesting thread. I'm kind of a combination of stuck in a rut and lazy - I have never *once* used slavery to whip anything. I played with it a bit in Civ3 and thought it was overrated. I assumed the same was true for Civ4. I'm very much aware that I'm revealing my own ignorance here.

Instead, I am a cottage-spammer. Well, I almost always build cottages on flood plains and green squares, unless a city is seriously food-challenged. I build farms on plains. I tend to build a lot of windwills, once they become available. Mines where the food surplus allows it. I like to cities with a surplus of at least 2 food, 3 is more normal. I try to get pottery as one of the first 4 techs or so - BW first of course.

It's pretty simple-minded, but it works well for me at Prince level. So, I have my own bias. My intuition tells me that I couldn't do as well using farms to grow my pop to the cap, and then replacing them with cottages. Here's my thoughts:
1) Using farms early means less money in the very early part of the game, which is the most critical part of the game.
2) There are times in the game when workers don't have much to do, and other times when they are busy and you wish you had more. I suspect that workers might be at a premium when you needed to convert farms to cottages.