There are several good ways to eek out as much as you can from your cities in Civilization 3, although sometimes it's a bit of work. In this guide I'll discuss some ideas to "micromanage" your workers, citizens and economy for maximum benefit. I hope to revise this document as I learn more about the game and as I get feedback.

I. Production

Do you just tell the city to build something and then forget about it until it appears? If so, and if that's the way you like it, put down the guide now and save your time :P Production micromanagement is shifting citizen working locations between food emphasis and shield emphasis. Let's say you have a unit that takes 20 shields and you produce 6 shields per turn. It will take 4 turns to produce, and on that last turn you're wasting 4 shields, most of your production. Instead, over the course of one or two turns, switch a worker from (for example) a grassland with one shield to a mined hill with three. Presto! You can produce the unit in 3 turns. Cost: one food. Compare to having that worker STAY on the hill. It also takes three turns to complete, but you still waste four shields in the process, and lose three food.

Implications? i) Get in the habit of scanning the turns-remaining in the F1 advisor window for things with 2-3 turns left, see if by changing a worker you can get that to one turn. ii) The great majority of units take a multiple of ten shields. If you set your "steady-state" shield production to 10 (or 20) shields, chances are good you'll get zero shield waste, *without* having to hawk over things like this. One other thing to look out for: sometimes shifting a citizen to get an extra shield instead of a food doesn't work because you get an extra shield lost to corruption.

Ib. Whipping (Despotic Rushing)

[This one got nerfed quite a bit in the 1.17f patch, but it's still useful]
The one value to remember here is 39. For 1 to 39 shields remaining to rush a structure or unit, the cost is the same, one citizen. For each 20 shields after that it's an additional citizen. There are two implications to this: i. Target pop-rushing with as close to 39 shields as possible, and no more. ii. The next 'magic' number is 79. From 40-59 the cost is two citizens, and for 60-79 shields the cost to rush in one turn is three citizens. BUT, by taking one extra turn that cost can often be reduced to two citizens. How? If there is unit available that costs 39 shields, switch to this unit and rush it. Immediately (or at least before the turn is over), change back to the desired unit, which will have 40 shields left. Wait one more turn, for the number left to be 39 or less, and rush the unit to completion. This can be chained down if there are intermediate units to build, e.g. 119 -> 80, wait for 79, rush to 40, wait one, rush to complete. Three citizens instead of five. In games where I've been despot whipping like mad, I've done things like purposefully build NO library (40 shields for scientific civs) in my cities so that I would always have a handy 39 to rush. If you have NO 40 shield unit to build, it's impossible to get 79 complete in two turns and save a citizen. [Sirian labeled this the 'WWW' move, Whip-Wait-Whip]

The awkward 'no mans land' is between 40 and 59 shields-to-complete. Depending on how dire the need is they can be rushed immediately, or you might need to painfully wait until that number drops to 39. Using the Production tips above, you can often switch citizens to high shield locations for a turn or two to get it down to 39 quickly, then poprush and move the workers back.

Popular structures for this technique: Longbowmen, Cannon, Caravel, all 40. Also Barracks, Coastal Fortress, the Library for Sci civs, Harbor for Mil civs.

Example: Techniques I and II... You start a Temple costing 60 shields and would just love to see it come online asap. If you rush now it will cost 2 citizens (and you might even be too small to do that). Your production rate is currently 5. If you don't switch, you can't poprush for another *5* turns. Instead, move something to a mined hill to gain two more shields. At rate 7, you'll be at 39 (tada!!!) in just 3 turns - much quicker. On many maps I like to rush a temple in every city (often the only rush, or along with one more), and so I get used to asking, when I start a temple, how can I get this to JUST under 40 asap?

II. Workers and terrain improvements

Despotism. In the early periods, despotism means that any square that would produce 3 OR MORE of food or shields will produce ONE LESS. This completely rules out certain improvements, such as irrigating grassland (2 to 3 gets bumped back to 2) mining OR irrigating cattle in the plains (already 2 food and 2 shields.) Some confuse this to think that trying to get '3' of a bonus is useless, but this is not true. Wheat on Flood Plains gives 4 food, or 3 in despotism. Irrigating it will give one more food even under despotism, so it's really always a winning move. No brainers: put a mine on every hill and mountain, irrigate those wheat squares, do NOT irrigate grasslands (until entering Monarchy or Republic), feel free to either irrigate or mine cattle squares (which show up on Grassland and Plains).

In general, past despotism, hills and mountains automatically get a mine, grasslands can see a mix of irrigation and shields depending on the need, plains are highly likely to be irrigated, and deserts can go either way (a good shield 'swap site' for food rich lands, and a needed irrigation spot in mountainous lands)

Railroads add +1 to irrigation *AND* mining, for all improvements where the bonus is already non-zero. A mined hill seems so useless in despotism, giving only 2 shields (base 1, +2 mining bonus, -1 for despotism). In any other government it jumps right up to 3 shields. Add a railroad and it's cranking out five shields. With coal, make that 7 shields :P The food bonus railroads give to irrigated squares is an unexpected bonus to Civ 2 fans. Basically, make it a goal to lay railroad on every square in a city radius.

III. Citizens

We already touched on moving around citizens to maximize production, or more precisely, to minimize waste. You set a production priority/goal then do it in such a you don't starve long term. Or you set a food/growth goal and kick over to shields for a turn or two to avoid shield waste.

Variety. Let's say you have a typical city area with some hills and mountains one plains, one desert, a forest and lots of grassland. Compare this for a size six city to one surrounded by grasslands, three with bonus shields. In the latter you'll get 12 food and 3 shields, and have no shifting that can alter that. Having only irrigation would make it 18+3 but still no variety, likewise all mines makes 12+9 continuous. The first city can go all the way from 18+3 to 4+16 as needed. One turn from finishing something that only needs 3 shields? Max food production. Need to eek out an extra 3 shields to finish something a turn early? No problem! Or even +7 per turn if you run max production and a food deficit.

Food deficit? Did you say food deficit? Why would you do that? *Short* term food deficit is helpful, and a KEY REASON for granaries! In the example above with a varied citiscape, we could go to 4 food + 16 shields. What if you need 6 food to avoid a shortage? With no granary, if this need comes at a time just after city growth with an empty food box, you can't go negative food for even one turn, or you'll shrink and probably lose the shield worker. With a granary that stores 12 pieces of food, even if you just expanded, you can burn at a min food rate for 6 turns, raking in an extra 42 shields! Then switch back when done and let the food box build back up by running at a food surplus.

Running at a food deficit briefly sometimes makes the difference between getting a wonder and missing it by one turn. When it comes down to a handful of turns left before a wonder is due, in a tight race, run food deficit when you can get a reduction in the time to complete the wonder but not quite starve the city down.

IV. How big will my city eventually be?

There's how big it can be, and how big it SHOULD be. The first asks what is the max size with maximum food, and the latter, with maximum shields and no growth.

I originally had a complicated discussion here, but have tried to simplify things. This question takes on very different meaning in different eras. Ancient- What size will it grow to before it runs out of food Middle- Size 6 before aqueduct, size 12 without. Industrial- size 12 pre-hospital Modern and late industrial - Railroads change the nature of the question.

Ancient cities in Despotism - It's a food game. You start with enough food to feed size one city, and can grow one more in size for each '2 food' square around, 0.5 in size for the remaining '1 food' squares. See that handsome hill mountain location with gems and gold in it? No grassland you say, but it should be great for production? Think again! The hills can supply 1 food, meaning 1/2 size. Two worked squares produce enough food for 1+0.5+0.5 or size 2. That's it!! This little runt of a town will never get above size two! Add ONE grassland - how does that help? In despotism you get two food from there and irrigation doesn't help. So one grass and all hills is stuck at size 3 for the next 4200 years! In the early expansion phase of the ancient era it's all about food! Oh and please, don't be caught irrigating a desert square instead of a cattle square! That's just crazy! If you have a square you'll work this millenium which can benefit from irrigation, do it.

Rivers. Starting a city on a river is huge. There is a commerce bonus but more importantly, it can grow above size 6 without an aqueduct. If that doesn't sound wonderful, since you can research and build those in ancient era, you need to play a few games where you DO have a river city with good food and see just how beneficial it is. In particular, an early city that can build a granary and rearch size 7 quickly will be of GREAT value to your civ as a settler or worker factory.

Modern era, or more accurately, railroad era. In Civ3 unlike Civ2, railroads increase the FOOD bonus of irrigation, which can lead to an explosion of city growth for your little size 12 city. To see how big the city can grow, let's assume everything that can be irrigated is...

Think of each square supporting food/2 size units.
Flood plains (2.5)
Grassland/Jungles/ForestGrass (2.0)
Plains (1.5),
Desert/Sea/Coast/Lake (1.0)
Tundra/ForestTundra/Hills,Ocean (0.5)
Mountain (0)
Add another city size for wheat, cattle, game and fish.

Go through the tiles in order of high food to low, starting with "1" size for the center square. Round down at the end if needed. Stop if workers needed is less than food. Example: Wheat Flood, Wheat Grass, four grass (one cattle), one plains, 8 hills, rest mtns What is the max size it *could* grow to? 1+2.5(+1)+2(+1)+4*2(+1)+1.5+8*0.5 = 22
This city can reach size 22 if all improvements to food squares are irrigation. You can trade off one size for 2 shields by mining instead of irrigating a grass or plains square (under railroads)

Ex 2: Coastal city-one fish, one whale, 1 ocean, 1 grass, 2 plains, 13 sea/coast, 3 desert
Max size = 1(center)+1(fish)+1(whale)+2(grass)+2*1.5(plains)+13(sea)+3(desert)= 24
The city can reach size 24, or can swap 3 population for 6 shields if it wants instead.

V. Wonders, Wonders (Prebuilding, and predicting your opponents)

At medium difficulties, especially Monarch and Emperor, the timing of Wonder building is difficult. Sometimes you risk building something only to have it snatched out from under you with 1 turn to go. Is there any way to gauge who will win the race? The most direct method is, with an embassy established, investigate a city -- this is a peaceful act which just requires gold. To do this all the time is prohibitive, and sometimes you just need a rough estimate. When the message pops up that the French are building the Great Library, hit F7 and see specifically where they are building it -- Paris or Bonneville du Povertie? Take a peak at that city if it's visible on the world map. Use the methods of calculating their shield production in the last section.

Quick and dirty method: (In the modern era, if you actually live to see it, the human player often dominates and can have whichever wonders he chooses) Let's specifically stick to ancient and medieval wonders. Count the forests and hills of foes building it. Add the number of mountains offset by a high food square like a flood plain, cattle or wheat. Stop if you hit their city size. Make the same count for your city. The higher one will likely win the race if they start at the same time. Ah but when does that happen? Ok, in despotism multiply this number by 2, or by 3 for another government. This is the estimate of the shield production rate. The number of shields needed for the wonder divided by this estimate rate will give the number of turns to complete. Let's say Paris started the Great Library 6 turns before you, do you have a chance? Don't guess, count! If they're size 6 with a cattle, two hills, a forest and decent food, their rate in despotism is about 2*(1+2+1) = 8 shields per turn. If you have a wheat flood plain, a hill, a forest and a mountain, size 5, your rate is about 2*(1+1+1) = 6. You don't have a chance. Add a river to get you up in size and two more hills, and your rate would be around 10. The GL takes 400 shields, so it will take them around 50 turns, and take you 40. Enough to make it with a 6 turn deficit in the second scenario.

Think ahead. Notice those messages when AI's start something! Even if it's not what you're interested in. The 'cascade' of wonders as one gets made and everyone switches is a huge factor in the ancient and early medieval age. When they start one, the one you're interested in or ANY one in a large game with many people building, look at the tech tree (F6) and ask how long it will take to even get the tech to start this wonder. That should be added on to how long it will take you to build the wonder unless you prebuild.

Vb. Prebuilding

Ah, prebuilding. This is what gives the player a crucial advantage. The AI gets rather large bonuses on higher difficulty settings. *BUT* (and this is a really big but) they never 'prebuild'. Prebuilding is putting a city on palace, forbidden palace, or something they have no intention of ever completing, to gain a head start in the race. Even on deity, a player can have a very solid chance of getting 'a' wonder by doing this. In the early game the capitol might be your only 'solid' production city, and there is no 'big' placeholder available. But using a aqueduct or library or whatever takes the longest as a miniplaceholder can help if you're just a few turns short of getting the tech and they announce that Perseopolis is building your favorite wonder. The real power comes in non-capitol cities. Think longer term. Sure 6 shields might not be much and won't get a 'currently fought' wonder, but it will add up and be an auto-win of one of the 'next tier' of wonders. For example, in the middle-end of the ancient era, you can put your best non-capitol production city on palace, and your next best on forbidden palace, AFTER the initial land grab when you have enough cities that the people want one, and when you can spare putting one or two cities "offline" of current needs. Building a temple there first is virtually a MUST. Whip it if needed, the memory will be erased in the long wait for the wonder. These two cities have outstanding shots at getting Leo's and Sistine's (or Sun Tzu's). The other nice thing about this is that it firmly breaks the cascade at start-of-medieval. You don't want to see the case where six nations fighting for Leo just about finish it then research the higher techs. Boom, you get Leo but you lose Smiths, Copernicus, Bach's, perhaps even Newton's. Instead, if Leo were "prebuilt", baam, when Leo is done six nations lose some major shields as they have nothing to switch to. Often having a coastal city prebuilding is important for this reason, even if its production is poor. (Again, start early) Otherwise... they switch off to Lighthouse and Magellan, go reeeeally slow with their weak-production coastal city, and the cascade goes on. By the time that wonder gets made by a slow city everyone has researched Astronomy and the cascade is fresh and on again. On higher difficulties the number of wonders you get drops, but you can go after your choice of ONE (or maybe two) with every expectation to get it, with proper prebuilds.

Vc. Delaying the finish.

Would you ever DELAY the finish of a wonder? Rarely, but sometimes, yes. Let's say you're working on two wonders and want to get both. Say the one with a larger number of shields required is due to finish a few turns before the other one. (This may happen when the 'smaller' one is in a lower production coastal city, like Colossus or Magellan). If the AI has been trying to build both, there is a problem. When you finish the bigger wonder, the AI will cascade and may finish the smaller one the very next turn. This is the time to actually 'delay' the bigger wonder by moving some workers off shield squares and onto food squares, or entertainers. Time it so they both *finish on the same turn*. Then all squares in BOTH opponent cities are toast :P Only do this if you're sure that delaying the larger wonder won't actually make you lose it! (Example: You're building Bach (600) and have about a 10 turn lead over Persia. You're also in a race with them for Copernicus (400) which is close. Bach is due in 2 and Copernicus in 8. Slow down Bach so it completes in 8 also.)

VI. Economy: Science and Luxury Rates

For quite a while I thought science and luxury was straightforward. The game shows the total gold income for the nation, and you get to move sliders for science and luxury. 1000 income should give spending levels of 100, 200, 300, etc, right? Well... they don't. In fact, at lower income rates, you might see (by looking at the 'science' or 'entertainment' breakdown of the expenses list, that the rate goes -4, -8, -11, -16, -32, -34, -38. Huh?? 40% is 30 turns and 50% is 15 turns and a huge deficit!! The questions are: i) why is that? ii) what can I do -- I want a rate with zero net deficit and around 22 turns.

VIb. How it works

The F1 screen shows total numbers, BUT zooming into each city shows how it really works, city-by-city! Take the commerce of the city, and subtract the corruption (lost gold). That's your 'net income' for the city. The sliders are applied to this per-city-net-income, highest percent first. Net income times the percent is then rounded *OFF* (not up, not down) and applied as science -> beaker and luxury -> happy face. The remainder is the net 'cash' from that city. Ex 1: 8 commerce, 2 'red' corruption gold, for 6 net. Science of 40% gives 2.4 beakers, which is rounded off to 2. Luxury of 20% gives 1.2, rounded off to 1. That leaves 6-2-1 or 3 gold going to surplus cash.
Ex 2: 12 commerce, 8 corruption for 4 net. 30% science gives 1.2 (1) beaker. But 10% luxury is 0.4 happy faces, rounded off to ZERO. Were you hoping for one happy face at this luxury rate? Don't put a worker on entertainer duty, see if there is a square in the city radius that has one more gold icon than one you're currently working. Does that give 5 net, or did it go to corruption? If the latter, shift another worker until you get 5 net gold income. Now 10% of 5 is 0.5 rounded to 1 happy face. So everyone can be happy just shifting a worker from a mined hill to a forest with a road. Ex 3: You have a ton of corrupt cities in your nation. The net income is tiny, just one or two gold. Setting the science to one rate gives 0.4 beakers, rounding down to zero, in not just one but many cities. The F1 slider says "36 turns". Moving it up a notch makes them all 0.6 beakers, and now it's "18 turns". That's why!

For more on how science works, see regoarrax's thread on Apolyton:
http://apolyton.net/forums/showthread.php?postid=600368#post600368

There he points out that...
- Base beakers is fMapSize*TechCost in editor, fMapSize=16 tiny, 20 small, 24 std
- Multiply by #Civs that don't have tech divided by #Civs
- He also gives a table with 'tech cost', e.g. BronzeWorking 3, Alphabet 5, Literature 10, Monarchy 24, Feudalism 26, Chemistry 52, Steam 96, Radio 140, Robotics 200.
- If the cost changes in the middle due to more civs learning it, the cost might shoot right down, even to 'next turn'.
- The 'going rate to 'buy' a tech seems to be 50% of the beaker cost to research it

VIc. What to do about/with this info?

"Oscillate." Switch your science slider to a very high rate for several turns then slide it low for an equal number of terms. When you have a situation where there is a large 'gap between two percentages, it's likely due to heavy corruption or a large number of small cities. Over the course of researching a tech, you will get a discovery rate between those two values. If you occasionally stay one extra turn at high burn rate it will be faster than 'average', and if you save a little extra cash one turn, it will be a little longer. So instead of just 36 or 18 turns by sticking at one rate, you'll find it discovered at 21, 24, or 27 turns, however many you like (and can afford).

In really tight situations, you can discover a tech a turn early where it doesn't seem possible! (Well, isn't that why you're reading a micromanagement guide??!) You have just 10 gold in the bank and really need something in 3 turns or your world will collapse. Three turns at 4 gold per turn won't work. Pump UP the science to a higher rate that says 2 turns at 9 gold per turn, followed by a low rate at plus two gold and 9 turns, and finally, 1 turn for minus three gold. Phew! Got it. That being said, I have to wonder if this is best. You could go at the rate that said 3 turns at -4 gold/turn for two turns, and then see on the last turn you can drop it down to see 1 turn at -2 gold. If the alternate oscilating strategy worked, chances are good it will work here too. If it didn't, it won't here. Why? Science is tracked by beakers. If you need 80 beakers and the '3 turn' rate is cranking out 30 beakers, you will 'waste 10' which do NOT carryover if you keep the slider constant. But two turns of 30 leave you needing only 20. So on that last turn you can crank the slider down a notch, less gold, and still put out 20 beakers. If you get down to "one turn" and you see, drat, two gold overbudget. Go to your cities and find some square NOT generating commerce that could be, and switch. You might lose 3 shields spread over 2 cities, but you gain that crucial research turn.

VId. Do I need to research at all??

This is a new topic added after the 1.17 patch. The answer in a number of situations, surprisingly, is NO. There are three reasons:
1) The cost of researching or buying tech goes down as other civs learn it
2) The cost to BUY the tech instead of research it is a factor of TWO in actual gold
3) A new-to-1.17 "Screw you, it's the world against the human" factor.

Factors one and two mean that being on the trailing end of the tech race, getting techs last, can give you *over 90% discount* in gold. Really! For example if you buy a tech in a game with 10 civs and only you and Rome don't have it, the cost is discounted to 2/10 or 20% of its original price. Cut the price in half again, buying instead of researching, and it costs 1/10 as much as being the first civ to research the tech. This deep discount is called "tech devaluation" and some would say, it's evil. That's what allows the AI that you've crushed throughout the game to "catch up" and not fall further and further behind in the modern age.

Factor three is awful. I hate it. In version 1.16, when you first met a civ, you could make a "fair trade" of your tech for their equal-value tech and both sides were happy. The AI is acting reasonbly. Now, however, it seems to devalue YOUR tech by a factor of THREE compared to when it sells a tech. Now it won't even accept two for one deals, which is insane. It WILL give fair deals however to the other AI. I agree with Sirian and others that this hurts the role-playing aspect of the game (or 'immersion') quite a bit, and is like Civ2 "me against the world". Two thumbs down for this change :(

Factor three in combo with one and two means it makes even LESS sense to be the first to research a tech because the AI will undervalue it by a factor of three. The Great Library is the ancient era answer to this problem, and Wall Street in the middle age.

VII. Culture flipping

This info is from Firaxis itself, on Apolyton forum:
http://apolyton.net/forums/showthread.php?postid=698322#post698322
But some folks haven't read that or aren't clear on the implications.

Factors affecting flipping...
- Base values are:
A) #foreignNationals in the city, counting resistors twice, and...
B) # of the 21 city radius squares that fall inside your cultural borders.

To this base, several factors are applied multiplicatively. In order of importance:
1) Ratio of distance to capitols of both parties
2) Ratio of total culture pts of both civs
3) Each city keeps a record of all culture generated by any civ who ever occupied it If the 'attacking' civ has a greater historical total, chance of flip doubles
4) City in civil disorder doubles the chance of it flipping
5) WLTKDay halves the chance
6) #landBasedCombatUnits are subtracted.

In summary,
((Citizens+Tiles) * CultureRatio * DistanceRatio * StatusFactor * HistoryFactor ) - Troops
... is the percent chance _per turn_ of a flip occurring.
What's NOT clear is if only laborers count as citizens. I've noted HUGE difference in chances to flip when I made them all entertainers.

Implications:
- Do NOT let them be in revolt. Starving the city down avoids this and decreases the number of foreign nationals at the same time.
- A city real close to their capitol is toast (or one of theirs near yours)
- Being weak or strong in culture as a nation can have a big role vs the opposite

This is the basis of 'culture' rushing and culture 'bombing'
- Rushing a temple or library can gain a decent number of squares within the 21 square radius to come inside your cultural borders. It also starts to build historical culture for 'your' civ
- Building a city two squares away from one of their upstart colonies will push back its borders and make less of their 21 in their cultural border.

Patch 1.17 change: now it seems that having "enough" of a garrison will definitely prevent flipping. With no equations known, nobody knows what "enough" is. My own guess... they moved the "minus garrison" term from outside the inner parenthesis to inside, so if you have garrison = foreign nationals + #squares in radius not in your border then the city will not flip. Don't hold me to that, but I've not yet seen a flip occur if I have garrison >= population + # squares not in control.

VIII. Forbidden Palace and Corruption

When can you / should you build the Forbidden Palace. It's an option after a certain number of cities which depends on the map size:

- 6 on tiny, 7 small, 8 std, 12 lrg, 16 huge are MIN #cities before you CAN build FP
- 12 on tiny, 14 small, 16 std, 24 lrg, 32 huge. ABOVE this size expansion cities and core cities will start to see corruption bite into their commerce and shield production. The effect is small for democracy and more for republic, then monarchy.
- Also, beyond a certain map-size-dependent distance from the palace or the forbidden
palace, the corruption can get so bad that even with a courthouse you get "total corruption". These "desolate" cities will only produce one shield and one coin no matter how big they grow. I'm currently doing some detailed tests on this, but for now let me say on a standard map, 12 squares away the corruption starts to become very noticeable, and ~24 squares away it's desolate.
- Below these limits at a medium distance away, corruption can be crippling
- So building 2 small clusters of booming metropolises is possible, or combined into one "dumbbell" shaped nation, but beyond that fringes city are hurting

Could it be actually COUNTER productive to put roads in corrupt cities? This posts suggests
http://forums.civfanatics.com/showthread.php?s=&threadid=10115

> Fringe cities can be built by converting food to shields. One population point is worth 40
> shields. Since forced labor creates unhappiness, military police is usually needed. Under
> communism, it is important to NOT improve the commerce of these cities (minimal roads/RRs)
> as the corruption will be shared with your main cities. Every road in a square worked by
> these cities will cost you 1gp per turn! Keep an eye on those automated workers.

How bad is the distance penalty? It seems that even small distance changes can play a big difference (ie a steep corruption-distance curve). Artax had a good post on the effects of corruption and how to plan cities accordingly.
http://forums.civfanatics.com/attachment.php?s=&postid=161641

He focused on a 'large' map, Monarchy govt, so scale accordingly for differents. He saw 10 squares away is where he starts to see corruption, up near 20%. By 12 it was up to 50%, at 13, 75%, and 14-15 or more, total corruption. Thus cities inside 10-12 range are treated fundamentally different from outside that range. The 'core' is then a box 21 by 21 (in terms of city centers) where diagonal and horiz distances each count 1, just like movement. Space them 4 squares apart (picture two up then two diagonal) for maximum coverage in minimum space. Done perfectly that will get you 32-37 cities per core, or up to 74 total, quite a lot. These can all expect to hit 20 pop if terrain is decent and 50-100 shields in peacetime. Where to put the FP? On a pangaea or large land mass, an inland site 24 squares from the palace is ideal, for the largest 'productive' and 'contingous' core nation. One core city founding, strongly prefer a spot 10 moves away over 12. Within this range the goal is dense packing.

On the other hand, OUTSIDE your ideal core radius of 10-12, everything changes. Production and commerce are total non-concerns, since you're stuck with one of each. Culture is important both for total nation culture and to expand boundaries. Why? You want to loosely pack, but have total coverage of cultural borders. It's all to deny AI settling space and to end up having control over resources. A rushed temple and library hit 100 culture in 25 turns. High food is good to get a better defensive bonus for city size, and to regrow quickly from rushing. Do NOT just go and build all city improvements in these cites. They can't even pay for their own improvements and will drain the core. Size 7 with aqueduct (or river) and granary make for a nice conscript or worker farm.

The non-core cities will even look different. Out there only irrigation is needed and a road connecting to the core, not a total network or mines. In expansion, you'll want to conquer anyone within your core radius asap. Beyond that radius, expansions are for the purposes of denial and for luxury/resource capture, definitely NOT for production or income.

VIIIb. What about Communism?

What about it? If you don't 'plan for it' but just drop into communism from democracy or republic thinking it's "the best" for extended war, you are in for a rude surprise. Not just a "hmmm, not quite as powerful as I thought" but a kick-you-in-the-teeth, I-can't-even-FATHOM-how-bad-this-reeks kind of surprise. I've been in games with large empires where we did this, and when we came to our senses and shifted to Monarchy, we went from running a DEFICIT with ZERO percent science to running a several *hundred* gold surplus, about 40% science, and a huge increase in production. Communism really is *that* bad. That's why you try to 'drag the AI down' into communism in the late industrial age. If they go there, they've lost, never to recover. Odd but true. This game where we went commie showed us why. Cash went into a deep hole, science went with it, and the production of our previously extremely productive core got slaughtered. If you're in a long war in a Republic or Democracy, here are some tips:
i) if you can ride it out and just need more time, boost luxury rate *way* up there
ii) if the war will take much longer than weariness will allow, go *Monarchy*
iii) if you're fighting two civs, make peace with one until you're done with the other
iv) don't make a large number of low-effectiveness attacks. Bombarding the landscape with catapults will cause a big increase in war weariness but will have NO real affect on the outcome.

There are advantages and factors that are beneficial in communism, but you really need to make fullest use of those benefits if that government is to be any use at all. Strict, strict avoidance of buildings that need maintenance in your captured little cities, use of the whip, heavy use of tax collectors, drafting, espionage, and careful thought about if you want a small or sprawling empire. I'll have a ton more to say about communism after my next Succession game, where it is going to get a major workout.

VIIIc. Courthouses and Police Stations.

What if you could build a factory in your best expansion cities in the ancient era? Would you build one? Hmmm, to expensive, not a good value? What if this 'factory' also increased revenue *100%*. Still not convinced? What if I throw in a "Solar Plant" for FREE?! Ready to buy? Wait!! There's more!! This magnficent factory- plant-economic center is yours for just... 80 shields!! Order TODAY and we'll throw in protection vs propaganda at no extra charge! :D

Such a ridiculous building doesn't exist you say? Think again! It's called a... Courhouse. Ok, not in every city, but in a large number of your cities, the courthouse (and police station) can work just as described -- let's look at an example. You're doing fairly well and have a small but growing empire. You now have a half-dozen or dozen cities which are 12-24 squares away from the capital. Let's look at one 16 squares away. You find about 66% corruption, losing 14 of 20 coins, and just over 50% waste, losing 8 of 14 possible shields. Add a courthouse and the loss becomes 7 of 20 coins and 5 of 14 shields. Your revenue increases from 6 to 13, over a 100% increase, and shield output from 6 to 9, a solid 50% increase. This is the equivalent of a marketplace, bank, and factory *combined*. And the shield increase will help you build each of those quicker! No solar plant? You're feeling cheated?? Go out to 18 or 19 squares away where you're losing 11 of 14 shields. The courthouse reduces loss to 7 of 14 shields, bringing output up from 3 to 7. That's huge! Roughly speaking, the courhouse seems to reduce corruption (revenue) by about 50%, and the waste (shields) by 33%. A police station on top of that cuts corruption by about 67% and waste by 50%.

So the question becomes not, "Should I build a Courthouse?" but rather "Should I ever NOT build it??!" That answer too is yes, sometimes:
i) If your commerce loss is 2 coins, building a courthouse pays for itself. So even with low corruption, it's not 'wrong' to build. But you may have higher priorities. This very low corruption will only occur within a distance of about 2 city radii.
ii) If the city is "desolate", building a courthouse or police will have no effect. In fact, it will COST you by adding to the maintenace cost with zero benefit. The problem is, you can't tell before you build it if the city is desolate or just 'corrupt' (able to be helped by a courthouse, vs hopeless) Again, I hope to have more numbers later, but for now suggest that you count the distance to your palace or forbiden city (whichever is closer). On a standard map it it's 20+ away, it would have to be a size 12 city and have a police station as well as courthouse to see any benefit whatsoever. Above 25 squares, forget it.

I would be remiss if I didn't point out that until playing succession games and seeing others build courthouses, I was not a big fan. Even then, after some gentle prodding by Sirian (Thanks!), it took a while for the 'lightbulb' to come on. When I did the math, that lightbulb became a shining beacon of radiant energy.

IX. "Opening Game" -- how to start a civilization

There many different ways to 'open' a game, and the strategies are very very difficulty dependent. On Emperor you'll reach unhappiness the very first time your cities grows (to size 2) unless your first unit built is a warrior. On the easiest setting, happiness is a total non-issue at the start, and sometimes not until the late medieval era. There is also the map size to factor in. On a huge map with less-than-maximum civilizations, you'll likely get a cluster of a half dozen cities or more before you even make contact. Military prowess is optional. On a tiny map with max opponents and parked next to the Zulus or Aztecs, if you ignore your military it's "game over". But I'm bothered by not knowing the answer to a simpler question: where getting wiped out early is not a big fear, what is the "optimal" way to start a civilization's group of "core" cities? The capital and a ring of six around it? First settler asap? Settler 'farm' or use several secondary cities? Granary first? Can you get the job done fast and still go for an early wonder? There's so many situations and varieties it's hard to answer, much yet cast as a mathematical model. I'm trying to build a simulator to answer these questions, but that's not an easy task either :P

Some principles, based on what's been presented so far in the guide...
1. In a city with tiles with greatly different food values, aim to pop out the settler as the city grows to the size hitting a 'dip' in the excess food supply. The food box doesn't change from size 1 to 6. So if you can grow one bigger and get an extra flood plain, you can put out settlers faster by not doing it right away at size 3. When you grow to a size where your "excess" is less than before, that's a great size to hit as a settler completes.

2. On Monarchy and below, think food, food, food. Unless you're going for an early ancient war, build absolutely as little military as you can stomach. Rather than go for the typical two-defender-for-MP-duty approach that is balanced, consider the "Farmer's Gambit". Get one 'scout' warrior going (or use your scout) to survey the area, then focus on city building, and when you hit size 3, finish a settler. Unhappiness isn't a problem (in Monarchy) if you don't go above size 3, and less aggressive AI's or distant AI's don't bother you. By the time they do, the goal is to have two or three more cities than you would have without going food-crazy, and so have the "city base" to switch to a military game quickly if you need it. Check out the rbd2 succession game for a really prime example of this, which is both the best case and the worst case scenario rolled into one game!
http://forums.civfanatics.com/showthread.php?s=&threadid=13653

In this game we played Japan and had a very food rich starting position. We went food-high, irrigating, forsaking defense and grew-grew-grew. And it worked well, that was the "best case" part. "Worst case" is when a civ normally rather peaceful, decided even on Monarchy to come after us and our smoke-and mirrors defense in the southern city. Well, we had more cities than they, and they lacked a key resource, and so we dove right into full-press mode. Soon they were pressed back, and we kept at it, steamrolling them, stealing a wonder, and taking the whole continent. All we wanted to be was peaceful farmer! This so upset the people of our nation that we set our sights on conquest :P

3. On Emperor, the AI is not so kind, or so weak. Neither are the rules. Unhappiness will occur right away when you grow, at size 2. So the first unit is very likely a warrior, and it may have to stay on MP duty. Better may be to temporarily move the luxury slider up and send him scouting, then build a worker or a settler next. You can still do a "Farmer's Gambit" going for quick cities and low military, just not 'no military'. Also, when the AI finds you, it won't have any qualms about attacking if you're *really* weak, and with its unit and production advantage, it will win. If RBD2 were Emperor diff, Egypt would have won (or more likely, we would not have made the same choices for no defense)

4. In a high-food low shield start area, don't neglect the whip. Especially if you are religious (cheap temple) and have Ceremonial Burial, fight the problem of unhappiness early, whipping a temple. Yes, you'll make one unhappy citizen, but he'll get over it if you don't whip the city to death. Whipping a granary might also be the best thing you could do, losing one unit to be able to double your growth rate and get more settlers quicker.

5. Your first move with the starting worker should be to irrigate the most appropriate square (if there is one). Then build a road on it, then irrigate a second square (if flood plains around) or start building a road to your second city and irrigate a spot for him. Food = growth = power in the start.

6. Start on a river. Almost always a good idea. Extra commerce (in fact, DOUBLE the science rate when it's just you and your capital), not stuck at size 6, and future home of the Hoover Dam :p

Worker and settler factories
[1.16]
Take the last example and slips in a granary on this last city. If it were on a river it could grow naturally above size 6. You get an interesing situation here due to the change in the size of food boxes. At and below six food boxes are 20 squares, while they're 40 from size 7-12. Going from 7 down to 6 the granary means you'll have enough to be completely full on the next turn, ready to grow to 7. This can be a "worker factory". Move a worker from a food square to a shield square to get 10 shields per turn. This city will pump out a worker *every* turn! If only 5 shields, every two turns. (In both cases, one excess food needed) The city grows to seven and gets a half full food box, then it pops the worker and those 20 food boxes remain, which fills the size 6 box. This repeats next turn. What if you want settlers? IF you can manage 15 shields per turn, build on the worker idea. With granary in place, at size 7, build a worker. This drops it to six, immediately back at 7 next turn. Start settler and have the worker join the city. In two turns (30 shields at 15/turn) a settler will pop out. That's one settler per three turns (per four turns if your shield rate is only 10). (This idea crysallized by LaFayette in Aeson on the Apolyton strategy forum as far as I can tell. It was a common theme in Civ2 due to changing food box sizes.) Alas it's not perfect for higher difficulties. With a temple and two mil units first, your luxuries and entertainment would have to be +2 or +3.

Vers 1.17 - The worker factory has been 'fixed'. They've shut down on using the size 7 to 6 loophole, but haven't killed it completely. The situation described above will now end up cranking out a worker every OTHER turn instead of every turn, if you have 5 food excess. (The number of turns it takes to pop out another worker is now 10 divided by food excess, rounded up.) The nice thing is that it still takes no micromanagement, but will produce a steady flow of workers.

Acknowledgements- much of what I've learned and presented here has come from gaming with the RBD crew (Realms Beyond) So much thanks to them for a lot of fun, and learning. Special thanks to Sirian, who from the start has not been shy about calling me on the boneheaded moves I've made :D and teaching me tons about Civilization III.

Feedback is very welcome, I hope this is useful.
Thanks,
Charis