Should the AI build more workers?

[JamesNinelives]

Something that always confuses me is why the AI sometimes leaves a bunch of tiles unimproved when the cost/benefit of improving them is usually worthwhile. I'm not sure how the worker logic is written but I figure if they build more workers then they would probably get around to improving their land sooner rather than later. Is this a good idea/would it help the AI compete with humans?

 

[azum4roll]

They seem to not prioritize workers after some are kidnapped by barbarians. Otherwise (if they can keep their workers safe) they improve tiles fine.

 

[DaniSciB]

Generally speaking, how many workers do the AI build per city? Do they "replace" or try to rescue a worker when a Barbarian kidnap them?

If they make more workers they might end doing 2 things:

1) Lose more workers to barbarians lol.
2) Improve their infrastructure faster and earlier than humans, this would lead to higher difficulty due to better yields. Which is good, since is why we want this.

 

[looorg]

3) If they build a lot of works and they are not captured by barbs or other players they'll eventually end up with an absolute crapton of workers that eventually won't have anything to do. Or does the AI know how to, and when to, disband units?

I guess it will also act then as an added incentive for the humans to capture workers instead of building them.

 

[JamesNinelives]

they'll eventually end up with an absolute crapton of workers

If we take things to extremes, sure. I think it's possible to nudge things without going that far though.

 

[looorg]

Some slight nudging might be worth a shot. As long as it just doesn't end up as pile on more and eventually it will be better. It's sort of like the AI isn't better at war just cause it has more units. All it appears to have done is to make the AI more aggressive, not better at actually doing war. Unless better here is some kind of substitute for just sending wave after wave after wave. But that isn't being good at war, that is just attrition.

The AI still do improve the important tiles, even tho it sometimes not the proper improvement on each tile and it does seem to take it quite a while to fix tiles that have been "improved" but with the wrong improvement due to later revelations of resources etc. So perhaps a few more workers might be in order, just as long as it doesn't end up in the extreme, or that is to say it knows when to eventually stop or do other things and not get stuck just pushing out workers.

 

[Bryan317]

Does difficulty have anything to do with this? I don't seem to see this issue on Deity as the AI always seems to be more improved on tiles than me. However, this may be my particular playstyle because workers are actually late priority for me. Now, late game is a separate issue because the human is just flat out better.

 

[looorg]

Does difficulty have anything to do with this? I don't seem to see this issue on Deity as the AI always seems to be more improved on tiles than me. However, this may be my particular playstyle because workers are actually late priority for me. Now, late game is a separate issue because the human is just flat out better.

Not sure. I don't really think I noticed it either that they wouldnt improve tiles at all. They seem to improve the important tiles for each city and then eventually they start to do the others as needed. They are not just using or doing the normal player shenanigans. Also they are sometimes building the non ideal improvements on certain tiles, or well not the once I would build but they are mostly following what they consider the standard, or recommended, improvement for that specific tile. During long wars (which is no apparently the standard) they do try to send out workers from the cities to repair the things I pillage etc. So it's not that they are just sitting around twiddling their thumbs. It might look odd that they are running back and forth a lot depending on how close to them my units are, but eventually they'll manage the repairs etc.

I guess if you see a city, lets say size 10 you would expect to see at least say 7ish improved tiles at least and that the others could/would be worked by specialists or sometimes it's ok to actually work a tile that is not improved for a bit but you might not want to do it long term. They don't seem to have an issue with not improving sea tiles with workboats either, they always seem to have a high priority from my observations.

 

[tommytoxen]

I wonder if it depends on the civ & the location of that civ's cities? Or even difficulty?

Sometimes I've seen civs with an overabundance of workers, all merrily running around building improvements and roads. I wish there was a "hey why's your worker building a village where a farm should go?" question in diplomacy. Real headscratcher sometimes

Then other times I've seen a civ with a city that hasn't seen a worker since its inception and many hundreds of turns (marathon) can pass without any improvements at all. Whether it's because they lack the workers or if they do have workers but are all prioritising cities further into their empire I don't know or maybe they just don't see that city worth improving I don't know. But I have seen cities left to rot without any workers for very long time.

I do think civs should prioritise workers if they have lots of unimproved tiles, especially if they have luxuries/resources that need improving. Because obviously leaving them unimproved is hurting that AI. But it needs to be done in a way that won't have unintended consequences.

Maybe that civ isn't building a worker yet because he's busy building a world wonder, and his other city is busy building a lighthouse, and his other is building a spearman. I think artificially pushing civs to build more workers might mean they build a worker instead of that spearman, harming his defence. Or instead of that world wonder, or builds worker before the lighthouse leaving that city unconnected for longer etc. I'd hate for us to solve one problem by creating another.

I don't know if it can be done only if certain conditions are met.
If civ !war && city has zero improvements && city not require essential building[/!underdeveloped] then build worker.
That way they'll only build a worker if they're not a war, if the city has no improvements (to avoid building too many workers) and if the city wouldn't be better off finishing that granary first before the worker.

 

[tommytoxen]

Not sure. I don't really think I noticed it either that they wouldnt improve tiles at all. They seem to improve the important tiles for each city and then eventually they start to do the others as needed. They are not just using or doing the normal player shenanigans. Also they are sometimes building the non ideal improvements on certain tiles, or well not the once I would build but they are mostly following what they consider the standard, or recommended, improvement for that specific tile. During long wars (which is no apparently the standard) they do try to send out workers from the cities to repair the things I pillage etc. So it's not that they are just sitting around twiddling their thumbs. It might look odd that they are running back and forth a lot depending on how close to them my units are, but eventually they'll manage the repairs etc.

I guess if you see a city, lets say size 10 you would expect to see at least say 7ish improved tiles at least and that the others could/would be worked by specialists or sometimes it's ok to actually work a tile that is not improved for a bit but you might not want to do it long term. They don't seem to have an issue with not improving sea tiles with workboats either, they always seem to have a high priority from my observations.

Yep, they love sea improvements I've actually been meaning to bring this up, in times of war I'll be pillaging sea improvements, and the civ will immediately build work boats to replace the tiles. Even though my ships are still there, so I'll kill that unit and then some turns later another work boat will pop up. I can only assume that city keeps churning out work boats because of the pillaged sea tile when it would be better off building something to support its war..
I think they love sea improvements a little too much at a time that doesn't make sense for them to. If possible would like to see them take the war and presence of enemy ships in vicinity into consideration before producing/sending out work boats.

 

[Stalker0]

3) If they build a lot of works and they are not captured by barbs or other players they'll eventually end up with an absolute crapton of workers that eventually won't have anything to do.

Considering the production benefits AIs get, there probably is no such thing as "too many workers" for a good portion of the game. However, of course the AI would need to understand to disband down to 3-4 workers once no more improvements are available.

 

[tommytoxen]

Considering the production benefits AIs get, there probably is no such thing as "too many workers" for a good portion of the game. However, of course the AI would need to understand to disband down to 3-4 workers once no more improvements are available.

do AI already do this? I've seen civs with tons of workers before, only later for them to only have a few.

Always wondered if civs were smart enough to disband workers that aren't needed, if only to save gpt if nothing else

 

[JamesNinelives]

Considering the production benefits AIs get, there probably is no such thing as "too many workers" for a good portion of the game. However, of course the AI would need to understand to disband down to 3-4 workers once no more improvements are available.

This is pretty much how I feel. I don't meant that the AI aren't improving tiles at all - they are. But I've never seen an AI with no more (useful) tile improvements left to make. So I figure having a few more workers would probably benefit them over the long turn.