Automated workers still don't work.

What I take issue with are consumers who place absolutely no value on 3rd party extensions to products that are best improved by free 3rd party extensions vs the official, paid extensions, who then turn around and rave how much "better" the official stuff is that does things worse than the free things released earlier by 3rd parties. These people get many names but the most positive one is "Loyal Brand Evangelical," coined by Apple. They are guaranteed repeat power users. If a company (such as Electronic "Arts" Games) pushes out loads and loads of games and DLC of decreasing quality and increasing price, the last demographic to stop buying are the Brand Loyalists.
It's a good thing I never did that then, especially since as I'm a modder of Beyond Earth. Y'know, a creator of third-party extensions.

Not surprised, but kinda saddened, that the thread took a personal turn. That seems to be the fate of anyone who points out flaws in criticism of the product. Despite the fact that criticising flawed arguments isn't the same as defending the content in the game.
 
An improvement queue could be useful. Imagine it to be a bit like the building queue. Each worker could have such a list and work through it.

Another variant of this idea is to have the improvement queue in the city screen. I'd just like to click the tiles and be able to choose what improvement should be built on them in the future. Then just build the workers and let them swarm out and do their thing with my personally adjusted priorioty. This would make managing large colonies so much easier.

Giving workers orders just when they become ready is a tedious thing to do. It mostly goes in the way when I want to do other things - like invasions f.e. But no, I have to press "." and "." again and whoo a military unit - "." again and another worker order, while the screen's jumping around the map from left to right like a kid with ADHD.
 
you can just delete/sleep workers whenever they want orders; still better than automating them.

I think keeping your workers actually working is more important than gold or placement, but the whole point is that at the late stages of the game it really doesn't matter, automation just takes one click.
 
An improvement queue could be useful. Imagine it to be a bit like the building queue. Each worker could have such a list and work through it.

Another variant of this idea is to have the improvement queue in the city screen. I'd just like to click the tiles and be able to choose what improvement should be built on them in the future. Then just build the workers and let them swarm out and do their thing with my personally adjusted priorioty. This would make managing large colonies so much easier.

Giving workers orders just when they become ready is a tedious thing to do. It mostly goes in the way when I want to do other things - like invasions f.e. But no, I have to press "." and "." again and whoo a military unit - "." again and another worker order, while the screen's jumping around the map from left to right like a kid with ADHD.

The queue shouldn't belong to A worker, but your entire worker population. Whenever a worker finshes, it finds the nearest thing on the queue (resources get say 7 tiles priority)
 
The queue shouldn't belong to A worker, but your entire worker population. Whenever a worker finshes, it finds the nearest thing on the queue (resources get say 7 tiles priority)

I don't think this would work as well as a per worker queue and would likely be more difficult to program as the workers would be under some AI control. This could work for making automated workers better maybe, but I'm certain it still wouldn't be as good as fully player controlled.

An individual queue per worker still leaves the workers entirely under player control, just with less management time. Instead of giving a worker a command every 6 turns, it could be every 18 turns; and more so with advanced improvements. The route-to mode is a great example already in game of this mechanism, where a worker is given a command to make several improvements and multiple workers can even work together doing it. A global priority queue, I would likely not trust that anymore than current automated workers.
 
I like the idea of giving automated workers a priority (improve resources first, max radius of movement, food focus, energy focus, etc) when automating but I feel that the previously mentioned check marks in the settings menu are essential features. Automated workers would then behave more rationally without changing their AI.

Those 4 check marks combined with advanced directions when automating would then allow players to move workers into position, automate them with priorities, expand valid options with check marks in settings as needed, then leave them be until they run out of valid actions. That is the definition of "Set it and forget it!"
 
Automated workers have regressed since civ iv, and they were not good then. They should build improvements based on city focus, link cities with shortest distance roads, and improve special tiles by default. If no city focus is selected they should build based on yield and available buildings.

They can't be as good as player micro, but they are not in a viable state. Why even include features so awful that using them is a trap outright?
 
They're only building farms? Well, sounds like they work perfectly fine - because that's what "everyone else" also does. Manually tell them to build a "Road to" when you see fit and there you go, worker AI that does your job at almost 30% efficiency.

And

if you want to use a more specialized strategy. How shall the AI know if you want Domes or not?

I've seen quite a diverse number of Improvements in AI-Terrain though, so there's probably some trick to getting the workers to build them. Did you try using City Focus Settings?

I'm almost 75% sure that will work. I noticed in Civ 5 that automated workers spam Trading Posts around Puppets (who are defaulted to Gold focus).
 
I'm almost 75% sure that will work. I noticed in Civ 5 that automated workers spam Trading Posts around Puppets (who are defaulted to Gold focus).
I wasn't sure how it works in Beyond Earth because I never automated workers in Beyond Earth. However, when I had wonders to build in Civ5, I set city focus to production. When a lot of things (mostly war, sometimes exploration) were going on and gold was no big issue, I would automate workers to a certain extent (usually route between cities, sometimes auto-improve). I rarely cared about puppet cities (revolts or whatnot leading to nothing done for a few turns) so nearby workers were usually automated. Wonder cities tended to get mines and lumberyards, puppets usually got trading posts, hungry (stagnant, starving, slow-growth, or food focus) cities tended to get farms and that's all fine with me. Beyond Earth gave us more options, new features, and a spiffy new Science-Fiction setting (which I love) to the Civ5 engine. I suspect that very little changed "under the hood" because most everything is merely a variation upon Civ5 mechanics, with poor Quality Assurance.
 
Automated workers shouldn't even be a feature. It's never going to work right and it's way too difficult to get it to work correctly. A better worker interface would be a MUCH smarter improvement and something that ideally should've been done a long time ago.

The ability to just drop improvements on the map with a priority system, leash workers to certain areas, and draw roads/magrails wherever you need them would all be much easier to do on their end then magically making super automation code, and much much better for players dealing with late game large empires.

Hell the ability to just draw roads/select tiles for upgrade would save me SO much time its not even funny. I'll build a road in a strategic spot, but route to decides the shortest path is from a road that's somewhere else so i've got to do it manually, and half the time the only thing I want to automate is when i'm upgrading all roads to mag rails, and I wish like hell there was a button to ONLY do that.
 
My ideal worker system would be a mix of AI and player instructions. Each worker has a "base" city just like air units, and workers will prioritize working within the city radius of their base city. Then player instructions are given in City View, not given to individual workers.

Players designate what improvement goes where in City View, and the AI governor then gives orders to workers according to those instructions, managing the order by prioritizing resources and improvements that conform to the current city focus.

Then there should be an infrastructure view that is a lot like the orbital view, which let the player designate road/magrail routes. When a worker is set to "infrastructure mode", the worker will prioritize building that road/magrail system.

To me, worker automation should not make gameplay decisions for the player. It should be there to make gameplay efficient.

This kind of system also means we can have a very elaborate worker units view, that calculates the completion rates of each city's designated improvements and how many workers there are. We can know exactly the percentage of infrastructure is complete and how many workers are working on them.

Last night I tried worker automation and they all rushed to the north-western corner of my territory... :(
 
My ideal worker system would be a mix of AI and player instructions. Each worker has a "base" city just like air units, and workers will prioritize working within the city radius of their base city. Then player instructions are given in City View, not given to individual workers.

To me, worker automation should not make gameplay decisions for the player. It should be there to make gameplay efficient.

Last night I tried worker automation and they all rushed to the north-western corner of my territory... :(
Limiting what the existing AI can do by manually giving Worker Automation a radius & focus (instead of using the Civ5 one of inheriting the focus from the city) and restricting what it can't do by giving more checkmarks in the settings (besides no terrascapes) should be doable with a mod (I think) because it uses existing AI features without adding new number crunching. More number crunching = longer AI turns

The solution I advocate has a shorter, shallower learning curve. It gives almost the same control to the player as your idea. If it impacts AI turn times at all, it would shorten them.

Now, if only a modder would be so kind as to implement it. :) :religion: :thumbsup: :please:
 
Automated workers shouldn't even be a feature. It's never going to work right and it's way too difficult to get it to work correctly. A better worker interface would be a MUCH smarter improvement and something that ideally should've been done a long time ago.

The ability to just drop improvements on the map with a priority system, leash workers to certain areas, and draw roads/magrails wherever you need them would all be much easier to do on their end then magically making super automation code, and much much better for players dealing with late game large empires.

Hell the ability to just draw roads/select tiles for upgrade would save me SO much time its not even funny. I'll build a road in a strategic spot, but route to decides the shortest path is from a road that's somewhere else so i've got to do it manually, and half the time the only thing I want to automate is when i'm upgrading all roads to mag rails, and I wish like hell there was a button to ONLY do that.
The ai needs workers to play...
 
I've been wondering. Why not implement a feature that allows players to "mark" what types of improvement to be built on a tile? A worker nearby would then go there and start working. I believe it helps on city planning and provide another solution to "worker automation" that sometimes doesn't work.
 
The ai needs workers to play...

I don't think worker automation should include strategic decisions. At the sponsor level, the AI should decide on victory conditions, what is needed to achieve that and what is available, then designate the roles each city should fill; at the city level, the AI should decide how to best customize each city to fill that designated role, which would necessarily include what improvement goes where; at the worker level, the worker at most should be able to decide on the optimal route for completing the designated improvements and infrastructure.

I'm of the opinion that AI play shouldn't be tied to worker automation logic in the first place.
 
I'm of the opinion that AI play shouldn't be tied to worker automation logic in the first place.

That seems...strange. If the AI is using the development team's best implementation of worker improvement logic (given city focus settings), why wouldn't you use the same code for automated player workers?

It was one of my big pet peeves in Civ IV that automated player workers would suicide by moving + improving adjacent to enemy territory, while AI workers didn't do that. I know it isn't like that in practice from the design perspective, but it gives the feeling of a deliberately-created noob trap mechanism that doubles as an inconvenience. It's shoddy, and there's no reason new iterations of civ titles should be repeating shoddy implementations.
 
Top Bottom