Efficient use of workers

[SlipperyJim]

As you mentioned at the end you lose worker turns. With the mine example its 1 turn to move onto hill and 12 to build = 13 turns. If you stack your workers you get 2 mines in 14 turns compared to 13 but you get an extra 6 turns of production from first mine.

Good point. I had forgotten about the movement turns. Thanks for correcting me! That said, I still think it could be worthwhile to lose a couple of worker-turns in order to get six extra turns of mine productivity.

When I mentioned stacking workers in my last comment, I meant actually grouping your workers together and assigning orders to the whole stack. The problem with this comes from the remainder that is left when you divide up the work.

For example, if it takes twelve worker-turns to mine a hill, then a stack of two workers can do it in six turns. No lost worker-turns, no muss, no fuss. (Except for the lost movement turn that I forgot earlier....)

However, if it takes fifteen worker-turns to farm a grassland, then a stack of two workers will finish the farm in eight turns. When the farm is finished, one worker will still have a turn to do something. However, because he is stacked with the other worker, he won't do anything unless you manually break the stack and give him an order. So you wasted that worker-turn. Larger stacks of workers will lose more worker-turns.

...

All of that information is great in theory, and it governs how I use workers in the early and mid-game. However, by late in the game, I often have a heap of workers with little to do. At that point, I'll automate them all. However, I should hasten to add that I always play with the "Workers Leave Old Improvements" and "Workers Leave Forests" options checked. That way, my automated workers won't pave over a Town for a Farm or something equally stupid.

 

[SlipperyJim]

This could be a fallacy, not that I dont do it myself. You wouldn't send a single swordsman to attack a city, why not group multiple workrs together? Getting that first tile improved in a new city in 1 turn is much better than waiting 5 turns while a single worker does it.

Your comparison isn't really valid. One swordsman will often fail to capture an enemy city. On the other hand, one worker can complete any improvement you want, but it will take longer to do it.

That said, I agree with your point. Many times, I'd rather have the new improvement online as quickly as possible, and I don't mind wasting a worker-turn or two in order to get it.

 

[UncleJJ]

This could be a fallacy, not that I dont do it myself. You wouldn't send a single swordsman to attack a city, why not group multiple workrs together? Getting that first tile improved in a new city in 1 turn is much better than waiting 5 turns while a single worker does it.

I am not sure what you think could be a fallacy. Perhaps it is something to do with how you edited my quote ... something I don't appreciate if you're going to make accusations.

I was only trying to indicate that I use my workers differently at marathon speed due to the lower proportion of time workers spend on movement costs. Obviously if I need a special tile improved urgently then I am much less concerned with the efficiency of worker turns. I do stack workers on normal, just a lot less often, which was my point.

What does a swordsmen have to do with anything I said? that's a complete red herring.

 

[budweiser]

I am not sure what you think could be a fallacy. Perhaps it is something to do with how you edited my quote ... something I don't appreciate if you're going to make accusations.

I was only trying to indicate that I use my workers differently at marathon speed due to the lower proportion of time workers spend on movement costs. Obviously if I need a special tile improved urgently then I am much less concerned with the efficiency of worker turns. I do stack workers on normal, just a lot less often, which was my point.

What does a swordsmen have to do with anything I said? that's a complete red herring.

My apologies. I meant no offense. I was only trying to point out that multiple workers may be better than singles. I dont like to play the slower speeds, but I think they way they work, they bring out core truths that are applicable at any speed.

 

[ParadigmShifter]

I play normal speed most of the time.

If I'm going to move 2 tiles to end my turn on a tile that needs improving, I'll only move 1 and put 1 turn into a road on the tile I move through.

I normally have a worker roading forests and hills on his own ahead of the rest so my workers can move to and improve these tiles later without losing movement.

 

[UncleJJ]

My apologies. I meant no offense. I was only trying to point out that multiple workers may be better than singles. I dont like to play the slower speeds, but I think they way they work, they bring out core truths that are applicable at any speed.

Accepted, sorry for being grumpy today .

The way workers are used on marathon is one of the biggest differences you'll notice about the speed change (although there are many others ). They're cheaper to build (2/3 effective cost) and significantly more efficient (10% to 20%, I'd estimate), which means you'll need less of them.

 

[Learningciv]

Saved worker turns only pay off in the long run. In the short term, it's better to pre-chop forests (if your workers haven't anything better to do) than to build roads everywhere in order to "save worker turns".

I think you misunderstood what I meant by saving worker turns. I didn't mention anything about roads..... I said saving worker turns (by smart micromanagement) has value.

If you can build improvements + prechop forests with 8 workers why build 9 if all you are going to do is waste time with them?

 

[Seanner]

Because in the time it took you to figure out exactly how to perfectly use 8 workers instead of 9 for the next 37 turns, you probably could've realized you didn't want more hammers and should've been building farms anyway--some larger part of the strategy was probably miscalculated during all of this (unless the type of player that plays 1 turn per day). Mental efficiency of the player is more important than efficiency of a single unit.

There is a thread by Gliese I believe where he posts a diety game in progress on an icey start. He goes in quite some length over the first few turns about exactly where to send what and what to research and why and so forth and saving a turn here and there by optimizing hammer vs food production at exactly the right moments and how this perfectly obtains a resource with the wheel simultaneously. Then someone said, but you already have the resource because of the coastal trade. "Ahh." Similar things like this happened shortly thereafter, and in fact happens in everyones games all the time. There is no offense meant by any of this, it's just that it seems odd to go into excrutiating detail on exactly how to implement a strategy when the strategy itself was faulty due to a lack of understanding of simple rules or mere oversight because of that very focus on microing.

People will lose far more likely because they attacked with too few units or at the wrong time or the wrong player or just shouldn't have attacked or specialized incorrectly or didn't forward build enough to block off or teched stupidly or whatever...not because they were one hammer short 14 turns later because the worker went half a square in the wrong direction.

Just figure out what you should probably do, figure out the approximately best overall way of doing it, and direct your own efforts as efficienty as possible towards this goal. Instead of saving one hammer on a 900 hammer build of a wonder, spend more time determing whether you need that wonder. Otherwise you failed to save 900 other hammers....

Edit: Don't spend time figuring out whether you need the wonder either, because you won't know with the billion other things you could've done in Civ4. But after you win or lose, you can look back and figure out how the wonder may have helped or hurt in those circumstances and adjust for your future games. A game should take no more than a few hours..

 

[PieceOfMind]

I haven't read all the replies but micromanagement of worker turns is very simple:

If your city is working unimproved tiles or is about to priority should be on getting the tiles improved ASAP and one by one. If the city is working improved tiles and is not about to work an umimproved tile then that is the time to be saving worker turns here and there through micromanagement.

When building long roads saving worker turns can mean the road gets built several turns earlier so it is worth doing IMO.

Example: you want to build a road on two forests, with two workers.

Case 1: with micromanagement.
Move worker onto each forest (1 turn). Build road (3 turns).
Total 4 turns.

Case 2: no micromanagement.
Move workers onto first forest (1 turn). Build road (2 turns). During second turn one worker moves to next forest. Worker begins road on second forest and second worker moves onto tile (1 turn). Finish road (1 turn).
Total 5 turns.

 

[cabert]

I'm very good at micromanaging my worker turns, but that more than anything else means I don't build enough workers!

If you have the need to travel your worker to build for another city, it clearly means you should have 1 more worker for that other city !

I micromanage to the point of prefering roading into forests to allow the forest to grow into adjacent tiles, and I almsot (not even on marathon) leave an improvement on a hill without a road.
I forced myself to build a worker each time I saw a tile needing improvement out of range of my current worker force, and that was a great improvement.

conclusion : build more workers

 

[PieceOfMind]

Remember, if you're working unimproved tiles and/or you don't have enough workers, you can always use the whip.

Personally I do not agree with the general wisdom on these forums that always "build more workers" is the best solution, at least not to the extent it's usually advocated. I don't think I've ever had 2 workers per city and I would rarely have 1.5 workers per city, except near the very start of the game when there are three (or fewer) cities.

 

[EmperorFool]

I actually do like optimizing my workers, being a builder at heart. I suspect I am at the far end of the micromanagement spectrum. My main efficiency wins are probably gained simply by planning my worker actions in such a way as to minimize travel time. As I get into the later ages, I will drop this in favor of spending time on war planning because I have plenty of workers.

These are some good tips. Here's one I haven't seen precisely, though the logic was hinted at in discussing pairing up workers to create two mines.

When you start seriously chopping down all that jungle, send single workers to road the tiles before you send in grouped workers to chop/improve them. You a) get a road, b) get shorter chop/improve times, and c) lose only 1 worker turn for moving onto the jungle.

 

[EmperorFool]

Example: you want to build a road on two forests, with two workers.

Case 1: with micromanagement.
Move worker onto each forest (1 turn). Build road (3 turns).
Total 4 turns.

Case 2: no micromanagement.
Move workers onto first forest (1 turn). Build road (2 turns). During second turn one worker moves to next forest. Worker begins road on second forest and second worker moves onto tile (1 turn). Finish road (1 turn).
Total 5 turns.

I would call case 2 "minor micromanagement" because it involves treating the workers separately. There is a 3rd case:

Case 3: Zero Micromanagement

Group the workers. Move them onto first forest (1 turn). Build road (2 turns). Move them onto 2nd forest (1 turn). Build road (2 turns).

Total: 6 turns​

I mostly do (1), occasionally do (2) if in semi-dangerous neutral territory with military support a few turns behind or I just didn't plan ahead. I will never do (3) because there is no payoff--it doesn't take any more time for my brain to do (1) or (2)--to offset the cost (2 turns without a road and 4 lost worker turns).

One thing that you need to consider is how efficient your mind is at micromanagement. I have some OCD traits that make my brain particularly good at micromanagement (and programming), so the cost is less for me than others. Others are very good at simple arithmetic, making their brains suited to quick combat or whipping calculations. It pays to know your strengths and weaknesses and play to them just as you play to the map.

 

[popejubal]

This could be a fallacy, not that I dont do it myself. You wouldn't send a single swordsman to attack a city, why not group multiple workrs together? Getting that first tile improved in a new city in 1 turn is much better than waiting 5 turns while a single worker does it.

It depends on the resource.

I'll be much more willing to gang up workers on Marathon because the relative inefficiency is much lower.

Throw 2 workers onto a hill to build a road on Normal and you've spent 2 worker turns building the road and 2 worker turns moving. Throw 2 workers onto a hill on Marathon and you spend 2 worker turns moving and 6 worker turns building. Obviously, if you throw 6 workers onto the hill, you lose out on that improved efficiency.

On Normal or even Epic, I'll often send 1 worker to build a road while the other workers keep busy on other projects. When the road is complete, I'll send the rest of the gang to speed construction. On Marathon, I don't want to spemd 6 turns waiting, so I'll either send 2-3 road workers or I'll skip the road entirely.

 

[JonathanStrange]

Efficient AI workers are a blessing; I like capturing cities that have had efficient workers improving my new tiles.

 

[StrategeryBush]

On normal speed I rarely have a worker leave a tile without building a road, particularly a 2MP tile. Frequently that road which nominally costs 3 worker turns actually only costs 2, as the road built allows the worker to reach and begin some other improvement on the same turn he leaves the hill/forest; further, going back to build a road on that hill/forest in the future actually costs 4 net worker turns, one to move to the hill/forest and three to build the road.

For the same reason I will not sent two workers to build a mine or chop a woods without a strong reason, it costs that extra turn where the worker has done nothing but pure movement. I do use multiple workers to build farms or cottages when the second can reach the tile and start work the same turn. Saving on worker movement is important on normal speed, much more so than on slower speeds. This is the reason for the higher relative value of the Indian Fast Worker on normal, rather than slower speeds.

 

[Woodreaux]

I generally play marathon but my points should be valid on other speeds. My early game (ancient & classical eras) worker management isn't especially systematic. I just build workers when I have the opportunity (in food heavy cities that at their :happy: cap) and steal them from rush/choke victims. I employ them individually usually one per tile in the rear echelon, 2-4 high stack escorted near or at the front or wilderness. When building roads through the wilderness, I keep worker stacks evenly divisible into the number of worker turns needed for a road. On marathon that's 2,3 or 6 (for grassland and plains). I do this so that the entire group of workers stays busy each turn all in the same tile, so that the escort need only protect one tile.

I use this same principle for most improvements, it cuts back on the need to mm. As the discovery of Calendar approaches, I build more workers (about double). The plantations take a bit more time. Their resources raise the population caps, so more BFC tiles need to be improved. During medieval era, I'll usually have workers group in clusters of 4-6. The idea being, I want them to finish their job all at the same time. When I need to do anything that involves entering woods or jungle, if I can plan ahead I send a single worker to build the road (so only 1 worker turn is lost). By concentrating my public works into a small number of stacks, I only have mouse over a few tiles each turn to see what is finished.

The most worker intensive part of my game is the discovering of Replaceable Parts. If my game is going well, I'll have a horde of workers grouped into 2 or 3 stacks of 8 or 16. I usually have roads through the forests before hand, so my workforce will create 1 or 2 new lumbermills each turn. With steam power, their efficiency increases, so I decrease the stack sizes. They usually finish lumbermilling the woods about the time railroads are ready. So I usually have most of my workers in stacks of 6.

So to sum it up, I group workers in stacks whose height divides evenly into the number of worker turns needed for what they'll be doing. I try to keep the stacks doing tasks require a multiple of worker turns a multiple of that stack's height. For odd tasks, I just check the stacks for a mixture of red and green dots.