Quick guide to irrigation versus mining

[Theseus]

I'm in a bit of a rush, so I'm just gonna dash off a few thoughts... please chime in.

DURING DESPOTISM

Mine everything except plains, floodplains, wheat, cattle, and wines (possibly also game on grassland after cutting down a forest)... this is because grasslands are not allowed to exceed two food unless there's a special resource on it during this era.

All plains and most desert should be irrigated.

An exception might be a killer site that is not on a river, as it's going to hit 6 pop anyway, so why waste the food... mine whatever you can.

AFTER DESPOTISM

In general, you want tile production to be balanced, with at least one shield, in preparation for RR. This may call for re-working many tiles.

What I've been focused on lately is not wasting food when at 12 pop. This may mean re-mining many tiles, or even planting forests on unshielded tiles... even food bonuses like wheat.

AFTER HOSPITALS

This will vary greatly by city... having balanced out tile production for RR, and then for max shield production, you now want to increase food production (with minimal decrease in shields) in order to grow the city to precisely the size needed to work all available tiles, including sea and ocean. In general, don't re-irrigate tiles with only one shield, as they will then have zero.

EXCEPTIONS...

are obviously the rule. A town surrounded by water, hills, mountains, and / or tundra will require special treatment.

Any contributors? Screen shots would probably be helpful for newer players.

[vmxa1]

I tend to mine only, even cattle, until force to irrigate for lack of growth. If workers have noting worth doing, I make roads in tiles that are not being worked at the time in anticipation of pollution and RR. I want to be able to get to all squares with no movement points eventually for clean up and defense. This type of work is not needed until I am in republic and some size 12 cities as a rule.

[ducki]

I'm actually having a lot of trouble with mining vs. irrigating.

I don't know if I'm just choosing my city sites poorly or what, but about 1/3 of all my cities in all my games are either so high-growth that I end up too big for happiness, or they are so low-growth that I hit 3 or 4 citizens, go Zero Growth, and have either very low Productivity or very high Corruption, at which point, I really start to wonder about the location.


I found a good analysis of starting locations and tile values somewhere - CivFanatics, maybe - that had some pretty in-depth "tutorials" based on the GOTMs.


What exactly did you mean by "not wasting food"?

[Arrian]

Theseus,

One note on wines: if they're on plains, don't irrigate them under despotism. The penalty eats the bonus. Mine them (you get 2/2 out of them).

Another note on a luxury resource: furs on forested grassland. Cut down that forest! It will give you 10 shields, and then if you mine the tile, you will get 2 food, 2 shields (due to the +1 production bonus of furs). Once you cut the forest, it's like having a bonus grassland w/extra trade. If you leave the forest, you get 1f, 2s.

My general rule of thumb in the early going: if it is possible to get a city 4 food/turn surplus, I will. For instance, if a city has a cow on grassland (but no other food specials) I will irrigate it, thus getting 4 f, 1 s. If, however, the city has 2 cows on grassland, I will either irrigate one, mine the other, or mine both.

I totally agree with your "no food wasteage" - which I would actually call "no food surplus" - comments regarding size 6 cities w/no river, lake or aqueduct and size 12 cities pre-hospitals.

If possible, I will typically set up my size 12 cities such that they have 0 surplus food, but are using a combo of irrigated tiles and hills or mts. Once a city like that has a hospital, I can just switch citizens off the hills/mts onto grassland tiles and the city will resume growth - no tile reworking needed.

I will do similar things for size 6 cities w/o aqueducts. I will utilize tiles that provide max production with 0 food surplus, but have tiles available to provide a food surplus as soon as the aqueduct is complete (if possible). These cities often get to keep their forest tiles longer than normal, only to have them cut down to help build the aqueduct, or the next improvement.

Once I get out of despotism, I try to remember to irrigate bonus (shielded) grasslands and mine regular grasslands, due to the GA or Mobilization effects (remember that thread?). I don't spend much time re-working tiles that are already done, but I will try and pay attention from that point onward.

Though time is often short, I will try to mine hills next to productive cities even under despotism, in preparation for the switch to monarchy or republic. This is a helluva lot easier if I'm playing an industrious civ. Note that I'm not doing this right away - only as I approach discovery of a better government.

-Arrian

p.s. Consider how important this stuff is, consider how badly the AI does it, and then consider the power of buying the AI's workers early in the game. Depending on how early you manage to do it, it can cripple an AI civ.

[itsagreenday]

What exactly did you mean by "not wasting food"?

When at size 12, prior to building a hospital, the city cannot grow anymore. It would be a waster to produce extra food, so you should mine squares that were once irrigated or work lower food-producing squares (but hopefully better shield squares) such as mountains or hills.

[Theseus]

Thanks for the great reply, Arrian.

Excellent point about the additive effects of a GA and of mobilization to shield production... they make it that much more CRITICAL that every worked tile be producing at least one shield.

[vulture]

Originally posted by Arrian
Once I get out of despotism, I try to remember to irrigate bonus (shielded) grasslands and mine regular grasslands, due to the GA or Mobilization effects (remember that thread?). I don't spend much time re-working tiles that are already done, but I will try and pay attention from that point onward.

I'm a little wary of the 'irrigate bonus grassland' rule of thumb - at least in principle. I suspect that in practice we all do something similar to each other.

For small cities I want to get as much production as I can out of a tile, and will usually mine bonus grassland, since 2 production, 2 food tiles are valuable. And I always irrigate plains, pre RR, at least until I want to cut down on growth in that city. But once I need to start irrigating grassland tiles, it is the bonus ones that get irrigated first. I'd prefer to leave them mined if possible. So rather than 'irrigate bonus grassland', I'd say 'if you have to irrigate any grassland, irrigate bonus grassland, but irrigate the minimum for your needs'. Not as catchy, and as I said, probably what everyone does anyway.

p.s. Consider how important this stuff is, consider how badly the AI does it, and then consider the power of buying the AI's workers early in the game. Depending on how early you manage to do it, it can cripple an AI civ.

Bien sur. Catching a civ early, with one worker and one city, and being able to buy that worker (or kidnap it if you are willing to have a minor war) does horrbile things to the AI civ in my experience. It'd be even more crippling (relatively) if the AI used its workers more effectively.

One other, less important consideration: try to have tiles worked in less corrupt cities. If you have bonus tiles producing extra trade, that could be worked by either of two cities, then work it from the city with the lowest corruption. And in the later game (or pre-hospitals if you pack cities in a bit more) you should never have entertainers or other specialists in a city while there are tiles in that city's radius being worked bu cities with higher corruption. Have as many tiles as possible worked by low corruption cities. This becomes increasingly more important with higher corruption levels (higher difficulty, small world sizes, earlier governments).

[alexman]

Originally posted by Arrian
Consider how important this stuff is, consider how badly the AI does it, and then consider the power of buying the AI's workers early in the game. Depending on how early you manage to do it, it can cripple an AI civ.

This is very true. I wish Soren would improve the AI in this respect. Right now, the AI irrigates too much, so its cities grow very fast and it has to assign entertainers to deal with unhappiness. It ends up with very large but unproductive cities.

I had an idea for the AU mod to get around this AI weakness. If we mod the game to give two luxuries per entertainer instead of one, the AI would not need as many specialists to keep its cities happy, so it would become more competitive.

[Arrian]

Alexman,

I know what you mean. In my last game, I established an embassy in Paris after making contact across the ocean, and this is what I saw:

Size 9, 2 luxuries, republic (at war): 2 happy, 1 content, 2 unhappy, 4 entertainers! It was producing 4 shields.

I do think the AI worker orders need work. I'd break it up by government type: worker orders that would provide no benifit under despotism would be disabled (such as mining a cow on plains or irrigating grassland). Further, a city which hits one of the growth "walls" (size 6 no aqueduct, size 12 no hospital) should be ordered to maximize production.

The problem with this is getting the AI to re-work tiles later on. Once hospitals are built, some more irrigation is desirable... if only for growth to approx. size 20.

-Arrian

[Theseus]

Hmmm...

As a warmonger, this is another point in favor of Industrious. I find that as I take over an AI civs territory, I need to re-work at least half the tiles, as well as do the forest/jungle clearing that the AI invariably screws up. Often, at first at least, this is being done with slaves.

In AU 106, the Greek / English "clean up" was thus made significantly easier, even at half-speed.

But you guys are definitely right... both the mine / irrigate algorithms and the AI civ's willingness to sell workers are critical problems.

[MyOlde]

Theseus,
In the first posting, you wrote:


"AFTER DESPOTISM

"In general, you want tile production to be balanced, with at least one shield, in preparation for RR. This may call for re-working many tiles."

What is the connection between the tiles and RR? How does one affect the other? And what is the best way to deal with them?

BTW: how do you guys do those blue/purple quotes?

[Theseus]

In this refard, RR increase shield production, so you want to make sure every tile being used has at least one shield.

To do quotes, use the "reply with quote" button at the top right of the post (it took me months to figure that out ).

[Harovan]

BTW: how do you guys do those blue/purple quotes?

Or, if you want to use the quick reply feature, you can copy/paste the text to quote and enclose it in [QUOTE] ... (text) ... [/QUOTE] tags. For instance, the above quote will look like this:

[QUOTE]BTW: how do you guys do those blue/purple quotes?[/QUOTE]

[Arrian]

But you guys are definitely right... both the mine / irrigate algorithms and the AI civ's willingness to sell workers are critical problems.

No! Don't take away my AI worker buying, you big meanie!

-Arrian

[MyOlde]

[SIZE=1] To do quotes, use the "reply with quote" button at the top right of the post (it took me months to figure that out ).

Thanks! (Yeah, those things are almost invisible, aren't they?)

[Killerdaffy]

Originally posted by Theseus
But you guys are definitely right... both the mine / irrigate algorithms and the AI civ's willingness to sell workers are critical problems.

How about making the price of a worker depend on the fraction of the total population it represents?

Price of worker = 20+(1/total pop) *400

E.g. Rome is a size 2 city, the Romans have two workers

Pop 4: 20+(1/4)*300 = 120

and:

Pop 3: 20+(1/3)*400 = 154
Pop 5: 20+(1/5)*400 = 100
Pop 6: 20+(1/6)*400 = 87
Pop 7: 20+(1/7)*400 = 77
Pop 30: 20+(1/30)*400 = 33

This shouldn't be too hard to implement for Soren (or does somebody see a way to do this in the editor??)

Edited: Calculating with small numbers can be difficult...

[Theseus]

Smart... unfortunately, not gonna happen anytime soon.

Arrian'll hate me for this, but I'm starting to think buying very early workers might be *gasp* an exploit.

[Arrian]

:glares at Theseus:

:sigh:

Yeah, maybe it is. It doesn't feel like one, because you *are* paying for them, but I guess they are too cheap.

Still, if it gets patched somehow, I'm gonna feel like a kid whose candy just got stolen.

-Arrian, who seeks to create vibrant, multicultural societies at the earliest possible opportunity by actively encouraging immigration in the ancient era.

[Theseus]

[laissez-faire]

Often, at first at least, this is being done with slaves.

Theseus, can you explain what you mean here? How does the work of slaves differ from citizens?

Also, from the sound of your discussion here, it looks like you don't ever turn the worker on full-auto. Is there true?

[Killerdaffy]

Originally posted by laissez-faire
Theseus, can you explain what you mean here? How does the work of slaves differ from citizens?

Slaves cost no support but also work only at half speed, compared to your own workers. Furthermore the civ whose citizens those slaves are wont like that at all. They'll be angry at you, which you can use e.g. if you want *them* to declare war.

Also, from the sound of your discussion here, it looks like you don't ever turn the worker on full-auto. Is there true?

Most people here will manage all of their workers themselves until at least the medieval age, since the AI worker rountines are ... - let's say we are having a discussion in this thread how to improve them When you do automate them, at least make sure to use shift+A since then they wont destroy any of the improvements *you* thought were important.

For lots of useful advice on strategy check here

[Tiberius]

Originally posted by Theseus
Mine everything except plains, floodplains, wheat, cattle, and wines

I respectfully disagree with that. I always mine cattle, because it has a +2 food bonus and a +1 production bonus, so if you mine them you'll have plenty of food plus production, too.

Mined Cattle on planes: 3 food, 3 production
Mined Cattle on grassland: 4 food, 2 production (on shielded grassland 3 production).
edited: without rairoad, of course.

Some people do the same thing with wheat, too. I don't, because wheat has only food bonus, so minig it is not a big difference in terms of production.

Obviously, all of this applies to "regular" cities. If the only irrigable land near my city is a cattle, naturally I won't mine it!

IMHO cattle is the best bonus resource, maybe except for whales, but you can't work that

[Theseus]

Having just captured a number of towns / cities, I will have also captured enemy workers (slaves), as well, and they are obviously first on the scene to re-work the land.

Slaves work at half the rate of your own workers.

I do turn workers on full-auto, but usually not until I have 1) irrigated or mined all tiles, 2) cleared all or most forest and jungle, and 3) built my RR connections.

[vmxa1]

Captured workers work at half speed compared to your workers. Auto is to be avoided at least until you get your RR built. Even after that it can be a problem as workers are poorly implemented and will undo some of your work if you do not tell them not to. They will build in areas that will nevr be worked and that cost you money, they will put themselves in harms way and probably even more bad things. I do let them go very late in the game when I do not care about all those issues, but still get annoyed to see them send 40 workers to a tile next to an enemies forces or tile or any tile for that matter.

[Theseus]

As an exception to the rule (and there are ALWAYS exceptions), I was looking at Cracker's Opening Plays thread on CFC, and in an example from their GOTM 8, there is a Russian starting position that is so strong from a food perspective (two wheat on floodplains tiles), that Cracker advises mining plains:

http://www.civfanatics.com/doc/civ3/...russ_start.htm

I think he's right... the important thing is to remember to balance shield and food production at both the tile level and overall.