AI workers too few in numbers?

[shyuhe]

I just played through a monarch game on bts and I noticed that the AI seems to be VERY slow to develop its land. Perhaps this is the reason why higher level AIs are weaker than they were in warlords? It was not uncommon to come across undeveloped tiles near large cities in the 1400's in my game, which seems very odd.

 

[Psyringe]

Judging from the dozens of workers I tend to capture from destroying other civilizations, I don't think that the AI builds too few of them. Your screenshot does look odd though. Perhaps the area around lagos has been pillaged before?

 

[jimbob27]

Some of it comes from the fact the AI often use SE's. You don't need that many tiles improved for a SE, because a fair amount of the population is taken up in specialists. As long as theres enough food improvment, you don't need to improve every tile.

In fact, it might be AI actually playing more sensibly than the human...... a lot of the time I improve all the tiles around all my cities, years and years before they ever get used. I expect a lot of other human players are the same.

 

[shyuhe]

Some of it comes from the fact the AI often use SE's. You don't need that many tiles improved for a SE, because a fair amount of the population is taken up in specialists. As long as theres enough food improvment, you don't need to improve every tile.

In fact, it might be AI actually playing more sensibly than the human...... a lot of the time I improve all the tiles around all my cities, years and years before they ever get used. I expect a lot of other human players are the same.

look at the size of the cities and the number of food sources in the vicinity...

None of the tiles have been pillaged in this vicinity as I was the only one to war with Portugal and I didn't pillage a single tile. I'm more worried that the AI hasn't cleared the jungle by 1400 AD. The negative health impact should be hurting the AI somewhat.

 

[uncarved block]

I'd have to agree, shyuhe, with this caveat-- the bigger problem is that the AI can't balance expansion with improvement. If the AI gets boxed in early, there will be more Workers, and plenty of developed land. Give them land to spread out to- say, a Big & Small map- and you'll see plenty of cities working undeveloped tiles, especially if there's overlap between two BFCs. (Check Sisiutil's ALC 17 for good examples. IIRC, at one point Mao had a Worker fortified on Flood Plain, even though there were about five or six undeveloped tiles being worked by the first two cities.)

If the golden rule is "one Worker per city", the BtS AI seems to prefer two or even three cities for each Worker. Throw in Espionage and Spies, and it's no wonder to me some AIs lag so badly in techs.

 

[jimbob27]

Well, judging by the amount of grassland and the resources....... I expect that whole area was solid jungle at the beginning. Looking at the shot a bit more..... I'm sure the main failing, is the AI's inability to deal with large areas of jungle.

I wouldn't say the AI underproduces workers in general, but when they start near a large area of jungle.... they don't really know how to deal with it.

Pretty much alway the AI that starts near the jungle, ends up being near the bottom of the pile. They just don't realise that they need to pay special attention to workers in that situation, and get all the jungle cut as quick as possible.

 

[Minor Annoyance]

What does SE mean?

 

[Psyringe]

What does SE mean?

Specialist Economy - a strategy that focuses on specialists and great people to run your empire, as opposed to CE (cottage economy, where you try to maximize your commerce).