A drawback of grouping workers?

[guermantes]

Hmmm, I got the tip that grouping my workers would make the worker management less all over the place and thus offer me a way of turning off automation without being overwhelmed with having to give worker orders every turn.

But I just want to check with you people to see if I have correctly identified an issue that I need to keep in mind if grouping workers.

Building a railroad on grass takes 3 worker turns. If 2 grouped workers are to build this, does this mean that 1 worker will be idling during the second turn as the first worker can finish the railroad by himself? Or will the 2nd worker start to build on the adjacent square during the second turn (given I have given them orders to build from A to B)?

Equally, having 3 grouped workers build 2 camps that takes 4 turns each to finish, will I lose 4 worker turns due to 'locked-in' workers?

[Jaybe]

With grouped workers and jobs ending with a worker still available, that worker GROUP comes up in the unit orders queue. You will have to separate the group to utilize the available worker because some of the group will be in red status.

If you are hurtin' for workers, then to not lose efficiency either:
add a worker when necessary so it comes out even, or
subtract workers so it comes out even (as in, just one worker).

[Blake]

Unfortunately you are correct that grouping workers is a Bad Idea. In limited cases you can group exactly the required number - railroading for example.

There are other cases of worker micro, like when doing an improvement two tiles away (flat land), you should stop halfway and pre-build something.

[Jaybe]

One must also remember that if you interrupt building a tile improvement, you can resume it at a later time with no loss. I have occasionally been surprised by sending a single worker to do something and find it taking only 1 turn.

[Blake]

I sometimes use that to defeat the barbarians.
Like it's marathon speed and I need a copper mine hooked up. The problem is if i build the mine then it'll get pillaged before the road gets finished, and vice-verca. So the solution is to build both until they're one turn from completion, then complete them both to give only a very narrow window of oppurtunity for the barbs to pillage.

[Urban Ranger]

Why don't you just stick an archer or two on the tile to defend it?

[Blake]

Because axemen rip up archers on flatland.
This would be raging barbarians plus a lapse of judgement which results in not founding on the copper .

[Urban Ranger]

On flatland? Oh. Unlucky.

[Blake]

No... unlucky is no copper at all .

[fed1943]

Forgive me, but I just cannot understand Blake here:

How do you get to begin build the mine and road? But not the last 2 turns? What is the diference (safety concern)?

May be better, if at all possible, one or two (possible sacrificial) archers and one or two meek warriors close waiting... (and some gold).

Best regards,

[Blake]

Unfortunately I think I deleted my screenshots of raging hordes of barbarian axemen crawling all over everything. Sacrificing units to them is remarkably ineffective.

What I refer to is building something, like a mine, or a cottage, until the worker shows 2 turns to go, then cancel the worker. At any time in the future that improvement can then be completed with a single worker turn. The partial progress cannot be pillaged so it's a way to have workers do work while the barbarians are being driven out.

Similer is pre-chopping, this is when you have a worker chop a forest until it shows 2 turns to go, then cancel the worker. the forest can then be chopped in a single turn. This is sometimes used to complete wonders right after getting the tech.

[Yosho]

Actually, I don't think grouping workers really costs you anything.

If a ralroad would take 1 worker 3 turns, then it takes 2 workers 2 turns, but on the second turn one of the workers finishes early and you can have him move and do something else. Doesn't matter if you use one worker or two, it still costs you the same number of worker-turns.

[Blake]

Yeah sure if you're willing to constantly regroup them. it can be faster than moving them one by one (using ctrl-click to select all still-active workers) but it's not as good as it could be.

[Generaldoktor]

I've frequently spoken in favor of micromanagement. I guess being immortal, I don't worry about wasting time.

Regrouping is the key, if your sanity can handle it. Click the worker stack again, sometimes it shows movement points left even if the top unit won't move. When you see the group displayed, click on one that shows green or yellow; then click on the "same type" group orders key down below; this reassembles the group without the guys that are totally expended. Move the new, smaller group to the next job; some of the "yellow" workers may run out of movement on the way, so repeat the regrouping until everybody expends all their movement/activity for the turn. You can finish two or sometimes even three projects this way out of a big stack, but the stack is, at least initially, easier to move and you have a better chance of having enough labor to finish everything, with an initially large stack, or at least, a lot.

Some of you probably know this (and just don't want to micromanage ; ) repeating this for the benefit of newbies who may not.

[Orky]

Eh, I play Marathon speed, and a road takes 6 turns, and most of the other improvents take multiples of 3 turns too, so I can group three or even two workers and have things done quickly with minimal micromanagement or loss of worker-turns.