10 steps for a successful builder strategy

[alexman]

Yes, it's a fact. I don't know how much of an effect it has, but keeping workers from an AI civ without joining them to a city makes that civ mad. Soren mentioned this in the Feb 15 chat here on Apolyton.

[Arrian]

Hmmm *ponders*

Well, normally if I have workers from another civ, they were taken in war, so it's no big deal if it makes them mad - they don't have long to live. But if buying a worker and using it hurts diplomatic relations, that's unfortunate, and worth considering. I always thought that buying a worker early on was a good idea, because you get a boost to your road/mine/irrigation building speed, and the AI you buy from loses a worker, for a pittance of gold. I don't have time or pop points to build workers usually until I've built 4 cities or so.

-Arrian

[Aeson]

It does make the AI think a bit less of you to have "slaves" of their nationality. I would doubt that how you aquired the worker makes any difference. It doesn't hurt that much though, it makes the AI think you are stronger (more units) and they are weaker (less units). Any change in how the AI deals with you because of reputation is offset by the change in power balance. I buy every AI worker I see throughout the game.

I once gave a small AI civ 50 of its workers back on the same turn. I didn't want to add them to my population because my culture was still relatively low. The AI became polite (from cautious), but the change in unit count led them to declare war the next turn. When I reloaded and didn't "free" their workers, they remained cautious and didn't declare war.

[Arrian]

Aeson,

That's really wierd, because I specifically recall Soren posting that workers didn't count toward the AI's calculation of "power." He said that the military advisor you see is totally different (and more simplistic) than the AI equation. Apparently, the AI does take into account the combined attack/defense values for units, plus numbers. I'll see if I can find the post. Your example may prove there is something going on that Soren hasn't considered... which I think is important.

-Arrian

[Arrian]

Here's the link:

http://apolyton.net/forums/showthread.php?s=&threadid=44074

[Aeson]

I know I'm not the only one who has experienced this (1.16f). I recall at least a few posts by people who had the same thing happen in their games. I remember a thread where someone suggested this as a way to get the AI to declare war. There could be reasons other than counting workers towards unit counts though.

At the time the AI had 1 city, so I'm reasonably sure that they didn't add those workers to their city and think they were bigger than me. I had well over 100 cities and at least 200 military units. Maybe they were just holding off on conquering me so that I wouldn't torture their captives!

In any case, buying workers is still a good idea, especially early on. It slows down the AI's expansion and speeds yours up a bit. I usually check every AI in diplomacy anyways, and most of the time can get a worker for the price of a useless map and 10-20 gold. You can always add the workers to your cities if you're worried about the change in AI attitude. As long as you don't add too many of one nationality to a single city it shouldn't make much difference culturally.

[dac]

I wonder if the AI has to pay upkeep on the workers you gave back? I know you don't pay for captured workers, but you pay for those of your own nationality. If you gave him workers back of his own nationality, maybe you tanked the AIs economy and he had to go to war to welsh on some debt?

Or maybe there's that last little bug in the algorithm nobody has found yet, and workers, even though they aren't supposed to be counted, sometimes are. Wouldn't be the first time somebody fixed a bug by changing code in all 27 places it's used, only to find out that it's used in a 28th.

[Aeson]

There could very well be a "we have too many units, lets start a war to get them killed" switch in the AI.

[Hurricane]

I tried my strategy on Emperor, and, as expected, it was much harder. I finally managed to do it on a standard map, where much of the world was water. The problem with large maps is that the AI is much better than the human player to expand, and that luxuries and resources are much more scattered around. This means that you often end up with only 1 or perhaps 2 luxuries on a huge map, while you easily can get 3 on a small or standard map.

The ROP-selling strategy didn´t work, either. It seems the extra units the AI gets at the beginning makes them feel so strong that they won´t pay for ROPs. This makes it harder for you to keep up technologically. In the earlier games, I got the tech lead during the medieval ages, but in this game only during the early industrious ages.

The failed ROP-selling instead meant that I had no problem declaring war on the Romans (5 cities), which my Mounted Warriors had no problem taking out. I didn´t get any leaders, though. This meant I got the FP rather late in the game.

I am now 4 techs ahead of the other civs (researching combustion), who have been fighting each other for some time. I have a steady income of about +0 ... +100 when at 100% science. I´ll probably take out the Americans when I get tanks and then head for a space-race victory.

[alexman]

Originally posted by Hurricane
The AI sells workers for 25-30 gold (lump sum). This is cheap considering you save 1 pop point, get a maintenance-free worker and that the AI used up one pop point for a measly amount of gold (which you of course quickly get back from him by trading).

From the AI perspective, Why would he ever trade you one of his workers for 20-30 gold in the first place, especially at the beginning of the game? I think I understand how the game calculates the worker's trade cost, but I also think it's flawed.

A worker costs 10 shields and one population point. A city of size less than 6 with a plus two food surplus takes 10 turns to regenerate that population point. if we say that during those ten turns an average tile with two food, one shield, and one commerce would have been worked, that's another 10 shields and 10 commerce lost.

So the AI trades you 10 commerce and 20 shields for about 25 gold. That would have been fair enough, except it doesn't take into account the fact that after the 10 turns that it takes him to replace the population, that's still one population point less than it would have otherwise been (thus losing more production and commerce until the city can no longer grow), not to mention the fact that in 10 turns he could have made several tile improvements that would have further benefited his economy.

Considering all the ridiculous deals the AI gives you for tech and resource trading, the value it places on workers seems out of character!

[Ribannah]

Reasons why the price might be fairer than it looks:

(a) The worker may have been an idle entertainer, or the city was at maximum size
(b) The AI has trouble paying for its upkeep

Since slave workers are slow, I prefer adding them to my capital. Early on that can be quite a boost!