Ideal tile improvements

[Feyd]

Do you have any hard and fast rules on what tile improvements you place on specific terrains? Do you think long term or change them regularly as new techs and abilities become available?

I'm not really intending this to be a discussion on chopping, but what are your impressions of lumbermills on a map where you only chop moderately?

Finally, how do you allocate your cottages? Do you get a ton of them started from the get go or leave open spots for workshops, irrigation-spread farms, etc.? Do you maybe cover half of the available ground with the cottages or do you go all out?

Here are a few TIs that I have come to appreciate over my play experience and will place there for the long term...

Windmill on Plains/Hills... (2/2/1) and more with upgrades

Mines on Grassland/Hills... Food and production in one nice package.

Floodplains and watermills... Mmm mmm good (irrigation until you get the mill)

Any others?

[kobo1d]

My method is simple but effective.

Resource ---> Resource specific improvements

Hills ---> Mines

Other ---> Cottage

Forest ---> Chop if beneficial

Occasionally farm if food is way too low in a city.

If you still have forests when you get lumbermills, build them.

Windmills, workshops, watermills. Ha! If you are financial, none can top the mighty cottage.

[fed1943]

My rule of thumb:for floodplains,cottages;for hills without forest,windmill;for flatlands with river,watermill;for flat grass without river,lumbermill;for hills with forest,windmill or lumbermill;for plains,farm;for resources,always the specific.
Best regards,

[Feyd]

Thanks for the responses.

I normally play financial so I'm very "cottage heavy", but have encountered production problems in the past.

Right now I'm playing a game as Washington and even though I just discovered industrialism, my capitol is only sitting on 30 or so production! Railroads have been helping as have the lumbermills I've finally been able to build, but I'm low on forests there because I needed to choprush a few things over the years.

Overall it's a quasi problem everywhere, but I've got to hand it to my civ when it comes to gold. I've got towns everywhere because, like I said, I went a bit overboard. 9 gold towns are very, very nice

Anyway, keep the comments coming. Any statistical folks out there have some numbers?

Happy New Year!

[Proteus_MST]

Forest --> Lumbermill or Chop (I normally try to have at least some woods around every city )
Floodplains --> most of the time Cottages, sometimes watermills
Plains --> Farm
Hills --> Mines most of the time, sometimes windmills

[Sparta]

I like to mix the improvements up, depending on my needs at the time and short term goals.

Early on I focus on farms to produce food which speeds up the rate my city grows. As a side note, I also focus on Granaries early on.

Resources are pretty self explanatory

I really focus on Mines for extra production. Really helps when I want a city to churn out miliary units.

[Diadem]

Am I the only one who does not always improve resources with the resource-specific improvement?

Especially plantation-resources. Take for example sugar. Building a plantation yields +1 food and +1 commerce. Well, then I'd rather have a watermill, for +1 food, +2 hammers and +3 commerce. Often I build cottages on those as well.

Of course, a plantation also give the happiness. So the first sugar I get I always build a plantation. But if I have multiple... I'd rather have watermills or towns on them. Plantations come pretty late in the game as well, so by that time your cottages will already be villages at least. Same story for spices, silk or dye. In fact only for bananas and incense are plantations worth it - and even then not always.

For other resources, the upgrades are usually worth it. They will already be available early game, and normal improvements will not usually give more then. Though for some resources, like fur and ivory, changing them into something else late game is usually better.

As for normal tiles. The best upgrade is of course the watermill. The worst is the farm. All other upgrades are about equal. So build as many watermills as possible, and then try to minimize your farms. Unless of course when you are making a GPP city.

[Diadem]

When building improvements, you should not first look at what tile is underneath. You should first look at what city will be using it.

A production city wants hammers. I don't care much about commerce in such a city. So no cottages, period. Hills get mines, of course. River tiles get watermills. Usually you need a few farms to be able to work all the hills around your city (of which there will be a lot in any good production city :P).

If you want to produce GPP then you build farms almost everywhere. A few mines in hills, a few watermills perhaps, to get some basic production. But mostly farms.

The commerce city? That's a no brainer. Towns, towns and more towns.

Mixed cities? Those are the hard ones. Then I take a close look at what tiles the city has and build my improvements accordingly.

[Feyd]

Thanks for the comments Diadem. My problem is that I'm not too thrilled about much other than "mixed" cities. I somewhat specialize later in the game, but as I'm a production junkie it is very hard for me to say okay... let's clearcut this grassland city and put all farms so I can get GPs. How though do I build the National Epic and other wonders that will aid in my GP quest?

[gentle]

Originally posted by Diadem
Am I the only one who does not always improve resources with the resource-specific improvement?

Especially plantation-resources. Take for example sugar. Building a plantation yields +1 food and +1 commerce. Well, then I'd rather have a watermill, for +1 food, +2 hammers and +3 commerce. Often I build cottages on those as well.

Of course, a plantation also give the happiness. So the first sugar I get I always build a plantation. But if I have multiple... I'd rather have watermills or towns on them. Plantations come pretty late in the game as well, so by that time your cottages will already be villages at least. Same story for spices, silk or dye. In fact only for bananas and incense are plantations worth it - and even then not always.

One other arguement in favour of resource specific improvements is the trade option.
Surplus resources can be traded to the AI.
It is not uncommon to get +4gold/turn for ones surplus sugar or gold.Have multiple trades running and it is equal to a shrine output and gives diplomatic bonus (rather than malus =P )

[Blake]

I also often pave over surplus resources if I have heaps of them. Some games I've had like 10 wines, even paving over half of them still leaves plenty to trade.

Value of Farms:
People diss farms but they are good post-biology.
Take for example Wine, if you build a vineyard you get +1 food +2 commerce. How does this compare with a biology farm?
For ease of calculation, take 2 wines in a city with mostly plains. Farming the wines results in a net change of +2 food, -4 commerce. +2 food is enough for a rep specialist (+6 commerce, +3GPP), it's also enough to work 2 extra plains tiles. If these are fully upgraded towns that's +4 hammer, +14 commerce. If they have State Property Workshops it's +8 hammers.

In any case, farms (especially biology farms) trump all (except SP Watermills) if they allow more tiles to be worked. In nearly all cases you want to work every tile in the city radius (or up to health/happy limit), building as many irrigated farms as nessecary to do so. But any extra food (over the minimum to work every tile) is very wasteful, as it can only be used for specialists, which are extremely inferior to towns, mines, lumbermills, workshops etc (with the exception of the GPP pump city and in some cases Rep specialists).

Altough there are some tiles which aren't worth working.
Take for example coastal tiles from a landlocked city (so no lighthouse, tile produces 1-0-2), also assume civ is running Rep.
A worker on a coastal tile nets -1 food, +2 commerce.
A rep specialist nets -2 food, +6 commerce, +3 GPP. One specialist is better than 2 coastal tiles in relative terms of +2 commerce, +3 GPP, +1 happy, +1 health. The specialist will still be better for financial civs (for the GPP and health/happy). Without the rep civic however, the coastal tiles are better than specialists, but it'll be better to work an improved plains tile.

Is it worth making biology-farms to work non-lighthouse coastal tiles?
For ease of calculation lets take a grassland and 2 coastal tiles.
If you farm the grassland you get +2 food, which gives enough food to work the 2 coastals. Total net change is +4 Commerce, -2 happy, -2 health.
You'll note a rep specialist will produce +6 commerce and only consume 1 happy and health, it is better in absolute terms.
Now what if you did something else with the grassland?
Workshop: +3 hammers.
Town: +1 hammer, +7 commerce.

In short, coastal tiles in a landlocked city should be considered dead tiles in the later game, altough for financial civs with Collosus they will be worth working for a period of the game, and if you don't care about hammers in the city and aren't running rep they are worth working.

Desert incense is actually a dead tile for rep civs, as specialists are better.

Any tile which produces 2 food plus something else is always worth working, as long as you can afford the health and happy hit from employing the worker. This means it's worth (non-bio) farming a plains, making it a 2-1 tile. There may however by more turn advantage in working something else or employing a specialist.

These calculations can be useful for getting the most out of your cities, and especially maximizing productivity at lower population levels.

[Diadem]

Very interesting calculations Blake. You make the point a lot clearer than I did. I think basicly your calculations come down to my recommendation: Minimize your farms. Build enough to reach your happy / health limit, or size 20, whichever is smaller, but no more. Unless you want a GPP pump city, or sometimes when you're running representation.

However, one comment about your calculations. You're stepping a bit too lightly over the effects of health. In small cities, every population point costs 2 food, true. But in bigger cities, you should count 1 population point as 3 food. This is more realistic, because you'll almost always lack health in these cities.

[Dominae]

As I mentioned in another thread, I will usually avoid putting Farms on Plains unless there's really no other good option. What a 2.1.0 tile gives is not 2 Food and a Hammer, but just a net 1 Hammer. I will almost always prefer splitting this amount of Food usage across 2 Cottages instead.

Obviously if I do not have access to a bonus Food resource I can only do this once, but this is rare. It's a good way to stop growth; the alternative is a Farm and a Specialist, which is only superior for the GPPs (I do not assume having the Pyramids for Representation - a luxury!).

On Emperor/Immortal, with the low H&H caps, Farrms are a lot less useful. I will build Farms around Hill-intensive cities and around my GP cities, but that's about it.

The thing about Biology is that it's too far into the tech tree to reasonably plan for; that's a long wait, during which you're tile usage is suboptimal. The same could be said for Lumbermills, the difference being, of course, that Forests provide a Health bonus in the interim, while all Farms do is spread irrigation (not exactly comparable benefits).

[fed1943]

Good points,Blake and Diadem.Nevertheless,I think the conclusion in big cities 1 population=3 food is too light;I don´t know the necessary math to make the calculation,but,at very least,the level of the game and the speed of research must be taken in consideration.
Best regards,

[ghen]

If you have two tiles, one grassland (with water) and the other floodplains.. whats the difference if you put a farm on one and cottage on the other? If you aren't financial there is none.

When I improve a city, I figure out what I need most and put it where it will be most beneficial at that time. That means if I need farms I'll put them on the floodplains first. If I still have floodplains left and have enough growth, I'll then put cottages on the floodplains. Since your cities won't be growing past 10 before you can re-issue worker improvements, placing improvements by terrain is a complete waste early on.

The only thing I don't do is farm plains if there's a grassland forest not being worked, as you get the same output. Use your forests as mini-improvements and you can have more freedom elsewhere with your workers early on... plus an extra health point or two usually.