Using Workers Efficiently

[Syiss_]

Like many here, I micromanage my workers relentlessly during the early game, then gradually move them off to "automate trade network" during mid game, and then in the late game when it doesnt really matter much anymore I will set them to full automation so they go improve some of the less useful tiles that I didn't bother with earlier but the cities are finally growing in to. I have it set so they dont override existing improvements or forests as well.

 

[TheMeInTeam]

On to its fifth page of discussion, when the answer is so apparent.

Automated workers do not understand city specialisation. Automated workers do not know your game plan. Automated workers do not prepare themselves for new technologies. Automated workers replace useful terrain improvements for inappropriate terrain improvements.

I am alarmed that some of the more experienced members of this community even half-entertain the suggestion that workers be automated.

... and welcome to CivFanatics eagle-bear!

This game was played under the setting "workers start automated". 100% random fractal map (even my leader). I never took them over .

It's a very late finish time, but arguably the larger factor there was spawning with only genghis khan as a neighbor...although note that until 8 turns ago, he was friendly which is why I'm not a smoldering pile of rubble for doing that right next to him .

He was gifted a city very early in the game. A city that was razz'd by a barb archer the following turn . The diplo from it lasted however!

The other reason this game went so slow is that I didn't get a 2nd religion until AFTER I'd had the culture slider going 20-30 turns. Despite spawning next to GK and only having one religion until after 1500 AD, we still have a win on immortal with full auto workers from turn 0 X_X.

View attachment TMIT AD-1914.CivBeyondSwordSave

 

[Cam_H]

Yeah, but kemo sabe, you're a really good player who can overcome the handicap.

(I kinda don't want to bump this thread for fear of turning it into some sort of absurd debate. I really think that impressionable players at the low to mid levels should be encouraged at every opportunity to manually control their Workers if the rest of us actually want to help them improve. I am fully aware of the automated Worker options btw).

 

[TheMeInTeam]

Yeah, I do agree that it's best to learn how to use them. Hell, you won't even know what to tell the governor to do w/o some knowledge of manual control. Manual control is a good thing to practice and master...but that game should also point out the fact that strategic decisions, diplo, and tech choice are the most important factors to improvement until one is at very high levels of play. Although I'd generally not advocate auto workers from turn 1 I would recommend prioritizing other aspects of gameplay first generally (of course one can always work on both or more...).

 

[vicawoo]

People who can't deal with 10 control groups of workers are weenies. If you ever played starcraft, you have to cycle through 7 control groups of buildings in less than a second. Even if you can only manage one click every one second, that's 20 seconds, less than half a minute total, to deal with 10 control groups (assuming your computer can keep up with you).

You should learn to make the decision what to improve automatically. You should be able to plan longer term while you're going through the motions.

 

[vicawoo]

And as for individual worker micro, you want an odd number of movement turns to flatland improvements, an even number to unroaded forests and hills. When you're off modulo, that's a free turn of roading.

You don't need 1.5 or some arbitrary number of workers per city. You need enough workers to improve as fast as you grow. If you grow every 5 turns and are building cottages, one worker is enough. If you have a food/flood plains city that grows every two turns, you might need more than 2 workers. If you have jungles and grow every two turns and want to cottage, you might need 4 workers. If you don't want to calculate and are growing too fast and are working an unimproved forest, go whip a worker.

 

[Gliese 581]

Dirk: I mean preparing a switch between cottages/workshops for example or farms/cottages. When I run out of improvements to make and/or I'm anticipating a key civic like state property or emancipation, that I plan use, I use my workers to prebuild improvements to within one turn of completion. The whole empire can then switch focus very quickly.

 

[TheMeInTeam]

People who can't deal with 10 control groups of workers are weenies. If you ever played starcraft, you have to cycle through 7 control groups of buildings in less than a second. Even if you can only manage one click every one second, that's 20 seconds, less than half a minute total, to deal with 10 control groups (assuming your computer can keep up with you).

You should learn to make the decision what to improve automatically. You should be able to plan longer term while you're going through the motions.

I was actually ladder ranked reasonably well in starcraft.

But the way unit cycling jerks one around and the incessant need for worker orders for over an hour wears on me. Games of starcraft wouldn't approach an hour unless they were tightly contested garbage 8 man FFA games or something (camping in BGH was ridiculous, much preferred fewer resources). Even wc3, overpowered though its passive base defenses are, takes a small fraction of a typical standard size civ map, even if I completely automate workers. There are hard bottlenecks in load times, and advisable slowdowns like thinking about one's general strategy (admittedly I spend a lot more time on that then how to move individual units).

 

[vicawoo]

Counterarguments:
TvTs.
Unless you're playing for in a tournament, people will play several games in a row, so you can easily plays hours.
There are breaks in between games, but you can take breaks in civ.
Just the sheer scale. If you have even near half the Actions Per Minute of a korean progamer, that's still near 200 APM, and you need 1/10 of that to perform all your worker actions in a minute.
If you're talking custom games, 150 APM for an hour long DOTA game is pretty standard, if not click spam heavy.

Don't your hands get bored?

 

[TheMeInTeam]

Even with perfect in last hitting, you don't need anywhere near 150 APM in dota . Well, excepting certain heroes.

I never got much over 80-90 APM, which is why the top players (in ladder, pro presence on the ladder is spotty) would still beat me. Usually...

I don't know...the worker actions just don't seem to carry the same intensity/meaning of moving marines around or getting that potion off in time. Or "Oh #$%# that's the sand king's ultimate! Stun him!".

I do see what you mean though. Worker micro variance isn't really any more complicated than unit micro in RTS once you know what to do in each situation. Sadly, I still don't, and that freezes me from time to time (which you've seen in live stream).

 

[Edzako]

I refuse to believe that you can be in reasonably high place in ladder (I assume ICCuP) with 80-90 APM

 

[TheMeInTeam]

It's just basic ratings, I wasn't in pro leagues or anything (league/pro players do play in ladder, but not all of them and not with the consistency that would get someone ranked in many cases).

And it's not that hard. APM drops a hell of a lot if you don't uselessly spam waypoints and crap all early game and constantly re-enforce identical orders. The only time you really need even > 60 APM is when you're fighting, where you need a LOT more than that sometimes. Granted, I would cap a little over 100, and that would limit me yes, but any strategic advantage is enough to cover that deficiency easily against most players.

 

[vicawoo]

But the way unit cycling jerks one around and the incessant need for worker orders for over an hour wears on me.

That cultural victory on emperor must have been very wearing. It does get difficult to split your concentration with worker micro when you have to deal with pressing shift-enter every turn.

 

[TheMeInTeam]

Come on now, by the time you're pressing shift-enter, the land is almost entirely (or often you can just drop the almost) improved.

And it's not like you're picking up techs that would prompt you to re-optimize improvements.

 

[NonPrayinMantis]

Once things are moving and I already have my critical improvements, I'll start grouping them. Three workers grouped on epic (2 on normal) will road most tiles in 1 turn. You will waste worker turns this way (e.g. something that would take 4 turns with one worker is going to use up 6 worker turns with a group of 3), but it's far more convenient.

I never do the 1.5 worker thing. I usually play epic, and I'll have multiples of three....9, 12 or 15 workers depending on empire size. Sometimes I even just disband a worker that I get in wartime, because he wont be "complete" unless I make 2 more, and I'm busy building an army. If I find I'm short on workers, I'll queue one up in three different cities, for another group.

Late game, when I do a lot of farm/workshop conversions, they're in groups of 6. I split them back in 3's for railroads, but if I'm not roading it's 6.

QFT

I usually play on marathon and I still use groups of three workers (although I used to use 6 per group). If I capture extra workers, I will use a group of two, or two groups of two, until I build/capture another one to complete the group of three.

NPM