Hi guys, this is my first time in here posting...
To note, my point is stricly on playing the AI, but may apply for multipleyer in some cases.
This strategy is interesting and worth looking at. When reading the art of war, it states that a siege can be harmfull for all parties involved, but if i remember correctly, it also states that a only way to attack walled cities is by a siege.
It also states "when doing battle, if victory takes long, it blunts the military and grinds down it's sharpness." And "If soldiers are long in the field, the state's resources are insufficient."
This is quite true for Civ, if doing battle for to long (like with a siege), the production in the cities becomes crippled and eventually the solders holding out will get killed one by one. As a conciquence, the player attemting the siege will get behind in development, because he is to busy producing military inits to replace the killed one's.
As an experiment, I tried this on prince level. I started building warriors and sent four of them to siege an Egiptian city, that went quite well. He never made any attemt to expand and only build military to attack my army. After this was set, I desited to turn to the next opponent, the germans. They had two cities, so it required some more units. This is where the plan got wrong.
The germans had stronger units than the egiptians. By this time I had the skirmisher (or whatever it's called), and desited that they would be sufficent for the job, since the would enjoy 50% - 75% defence bonus, depending on the terrain. I was deadly wrong, the plan failed badly.
At this point I was producing a unit every third turn or so and sending it to siege the germans, the germans answer was to produce axmen and chop of my men one at a time. At first I looked to have the upper hand, but my production was slowing down all the time because of unhappyness in the city, remember, I was only producing military units. Eventually the germans killed most of my men. The end of the battle was when he sent three strong units to take my city, then I was forced to settle for peace with a terrible score and way backward in research.
I tried another experiment with this, this time the AI cheated, they got support from three strong units that appeared out of nowhere just by my city and that was the end of it, I had no troops to defend.
In conclution. I would say that this method is ok for a fiew turns against one enemy. To pillage the enemy, but if it goes on for to long, it will harm city production and research just goes crumbling down. In one of thoes experiments a actually went bankrupt and the workers went on strike.
Please excuse my spelling, english is not my first language.
Siff
... I always have extra muscle hanging around 
