Any square that produces trde for any reason produces one more if there is a road in that square.

DaveV, Wm, & MarkEverson: the added trade arrow from roads also means that it is advantageous to "forest" river squares and cut down non-river forests. That way you can get trade in every square.

I am new to civ(less than 3 months) and this thread is a great resource.

Keep up the good work.

Martial Law can be imposed under Anarchy, Despotism, Monarchy, and Communism. Each of the first three military units stationed in a city makes one unhappy citizen content, and this effect is doubled under Communism (so each unit makes two unhappy citizens content--very nice!).

Under Republic and Democracy, military units in the field that are not in a fortress within 3 spaces from a city you control also can cause unhappiness. For Republic, each roving unit beyond the first makes one citizen unhappy. For Democracy, every roving unit makes two citizens unhappy. This is why I build fortresses for my military before switching to Republic or Democracy.

Also, there are certain units that make people unhappy in their home city, regardless of where they are stationed. Under Republic, each Bomber, Stealth Bomber, Helicopter and Missle Units makes one citizen unhappy. This effect is doubled for Democracy.

One way to wage war under Republic and Democracy is to build Shakespeare's Theater. By reassigning your military units to Shakes before sending them in the field, you can ignore the unhappiness effects.

Now that I really understand the unhappiness factors, I can effectively prepare for government changes throughout the course of the game. I hope this answers your questions adequately!

Here is a repost of some comments on happiness made by Xin Yu that I believe are of great value. Thanks Xin!:

There are two levels of disorder control: global and local. Do the global first until most of your cities are OK, then deal with the few problematic ones.
First, choose the government. If most of your cities can grow using 'WLTKD' then obviously you should choose democracy or republic. The former is preferred since the courthouse gives one more happy citizen (crucial if you play deity and have a little bit more than a dozen of cities) and has no corruption. In other cases it's a trade-off between more trade vs. losing martial-law/free support and unhappiness due to troop out of city. Basically you add the increased trade for democracy/republic (adjusted for city improvements and trade route revenues), and compare with the total of lost martial-law luxuries (12 for commie, 6 for monarchy, 2xpopulation for fundy in deity), supporting shields (3 shields=6 taxs), and the out of city unhappiness (1 unhappy=2 luxuries). I would choose fundy or commie in the early ages but go to democracy when I can build superhighways. I may stay in commie/fundy if playing 2x production.

Second, decide the luxury rate. Since science rate can be set to zero (you use caravans to get sciences, hire one scientist), it's a split between tax and luxury.

Third, build improvements. For bigger cities, prefer marketplaces to temples, banks to cathedrals due to the following reason: a temple with mysticism keeps two citizens content, which is equivalent to 4 luxury. A marketplace increased both tax and luxury by 50%, so if you originally have 8 trade icons or more your total tax+luxury is better with a marketplace than with a temple. Same for banks vs. cathedrals. Of course wonders are different--they cost less and need no maintainence.

Fourth, use martial law.

Fifth, use Elvis if necessary.

Sixth, assign your troops properly in democracy/republic to reduce unhappyness. Two wonders can help: Shakespear and J.Bach. Actually the latter is better. If you have J. Bach, build a city on a mined hills and put a worker on another mined hills. Keep the city at size one, and with factories, hydro plant, and manufacturing plant it can support tons of units. It will not go riot due to JB's effect.

In addition to all of the standard bits of info regarding diplomats and spies, here are a few lesser known items.

1. Unless you have a good reason to build something other than a caravan, build a caravan.

2. Unless you have a good reason to attack something else, attack the capital with sufficient force to take it on the first try.

3. Unless you have a good reason to save your money, spend it about as fast as you get it.

Comments on these laws.

Early caravans are the fastest way to build early wonders like Copernicus and Michelangelo. By the time the early Rennaisance wonders have been built caravans are more profitable when used for trade. Also trade caravans delivered to a distant continent boost science far better than any library or university. Build other items if you need to stave off riots or need defenders. Diplomats and spies are good because they don't cost shields. If you have some money you can buy extra defenders.

If you can take the capital it will often put other cities into disorder or cause the AI to build another palace. But then it will be cheaper to bribe other cities. Sometimes a capital will be in an awkward location and you will have to take some preliminary cities.

Keep enough money so that it is expensive for AI civilizations to bribe your units or cites. Have enough money so that you can bribe attackers. Beyond that spend the money to build wonders or caravans.

Capturing a useful wonder in another city is a good reason to go after a city other than the capital. If the Great Wall or Pyramids is built in a city other than a capital the first two triemes together with three caravans and a diplomat that I build will sail to that civilization deliver the caravans and use the proceeds to bribe the city with the wonder.

If a capital is in a really awkward location, and this is common, you may have to capture nearby cities either with bribes or attack. AI cities frequently go into disorder. If you are not in a hurry and have sufficient money, station a diplomat in a boat just off the coast, wait for the city to go into disorder and then bribe it.

The AI civilization spends a lot of money when it moves the capital. That makes other cities easier to bribe. If the AI capital moves to an easily defended location, wait until you have superior attacking units and then go after the city. Veteran cavalry are acceptably effective against pikemen.

I'm probably reinventing the wheel here, but in the absence of takers, it's DIY (do it yourself) time. Here's the empirically obtained figures for *selected* terrains and terraformations showing the number of turns for engineer and settler, where x = undoable (us poor weak settlers )

grassl => hills 9 x
grassl => forest 4 9
forest => grassl 19 x
forest => plains 2 4 2 (Eng+Settler, both free to move next turn)
plains => forest 7 14 4 (Eng+Settler, both free to move next turn)
hills => plains 19 x
desert => plains 9 x 4 (2xEng, both free to move next turn)
tundra => desert 10 x
glacier => tundra 19 x
jungle => forest 7 14
jungle => grassl 7 14
jungle => plains 19 x
swamp => forest 7 14
swamp => plains 19 x

(Hope the simple tabling comes out OK - hard to judge from the limited input box)

Certainly selected for the OCCasion , the extra third colum shows some combined efforts - interesting how times are "halved" for odd numbers.

The experimental error appears to be +- 0.5. I'd be interested to get feedback to confirm or deny the occurence of spurious results. I'm sure others would also have encountered anomalous eg accelerated processes in normal play!

Turns taken for terrain improvements (roading, irrigation...) to come.

Sure, most experienced civers would have the figures in their head, but numbers are helpful particularly in the context of the challenge games for us novices.

Interesting your quote of 5 turns for the terraform, forest to plains (by a settler) compared to my empirical value of 4. This is an example of my observation about the error range for a given operation when the number of turns is greater than about 4 or so.