The number of workers needed per city.

[Tesuji]

In many posts on this forum I've seen people express the rule of thumb that one needs (on average) a worker and a half per city. One worker when the civilization has the industrious trait. Of course this is just a rule of thumb, and as with many rules and strategies in this game, it depends heavily on the situation at hand.

Although 1.5 workers per city sounds like good advice, it also feels like medicine with a bad taste. After all, building workers goes at the cost of population. And we all know that population is power. And it also goes at the cost of unit support. The inevitable result in just about all my games is I rarely have that many workers. Still, I often find that my workers are idle considerable parts of the game. Actually, the only time when I seem to have too few workers is with the advent or rail-roads. But often just before that I send my workers to my capitol and have them wait for Steam Power. Then they're busy again for a few dozen turns, after which I hardly need them again.

So this made me think: if I don't build as many workers as is generally advised, how come they're idle this much? So decided to give this a bit more thought to see if I could get some rationale behind this rule of thumb orcome up with a better one.

Let's start with the premises, what are workers needed for:
- Improving tiles worked by citizens.
- Connecting cities.
- Connecting resources and luxuries.
- Clearing forest, jungle and marsh.
- Making fortifications: so rare I'm going to ignore this.

The first item is the most basic function. Ideally, every citizen only works an improved tile. How do we achieve this, with how many workers? I find it a rare occasion that I need to connect a city or a resource through a tile that doesn't get worked on eventually. Since there are so many variables in this game, I'm going to start with the most basic scenario. This can then be used as a basis to make adjustments according to your current situation.

During despotism it doesn't pay to work hills or mountains, so in this stage of the game you almost exclusively work grasslands, plains and flood-plains. Improving grassland takes 10 worker turns. One turn to move, six to build a mine and three to build a road. Plains and flood-plains take 8 turns to irrigate and road. Your average town will have two food surplus. Unless you build granaries everywhere, which is not advisable, it takes 10 turns for a town to grow in population. So one worker should be able to keep up with a towns growth. Taking into consideration that it takes time (and food) to build a settler + worker and water tiles don't get improved, one worker will generally even have some spare time to connect a resource or city. Or clear some forest.

By the time you get out of despotism it starts to make sense to work hills. Move, mine and road takes 19 worker turns on a hill. Most likely however, by this time the cities that are the most likely candidates to work hills will have already grown past 6 and will now take 20 turns to grow. You will also have towns that still don't have an aquaduct, harbor or both and therefore grow more slowly or not at all.

So my conclusion is: the basic number of workers each city needs is 1.

This is not all however, I've ignored two important issues. One is food surplus and the other is clearing jungle. Let's start with food surplus. Having three food surplus instead of the usual two makes a town grow in 7 turns. A surplus of four makes it grow in 5 turns. You'll see the pattern now: it's not how many workers per city, it's how many workers per food surplus. And the answer to that is 0.5. So a standard town or city with the two food from the city-tile needs one worker plus 0.5 for every other food surplus it can produce.

Clearing jungle (or marsh to a lesser extent) is a more complicated issue. Jungle takes 24 turns to cut. So a town that is all jungle would need two extra workers to keep up, a city would need one extra worker. If I average this out it gets to 3 workers per dozen jungle tiles. The numbers are half that much for marshes.

So my new rule of thumb for the number of workers needed is: (ta-ta-ta-taaa!)

0.5*food-surplus + 0.25*jungle-tiles + 0.12*marsh-tiles.​

So as you see, no number of cities in this equation. But I think this is a much more accurate rule of thumb. Since the industrious trait makes workers work faster (but they move at the same speed) I would make this 0.4*food-surplus + 0.2*jungle-tiles + 0.1*marsh-tiles for industrious civs.

 

[vmxa]

Well there are two ways of creating workers that are not very painful (slaves not counted). One is to dedicate a town to the task. This means we do not really worry about loss of pop as it is set up to be static and not get above town size.

The other (well these are my ways anyway, not all possibilities) is to have a newly founded town crank one out first thing, if it can be down in 10 turns. Again this will not do much damage to pop as it will be paid back by increased food and shields.

I would submit that if you have idle workers as much as you say, you need to get more land. I am acquiring land through the game and seldom have idle workers, no matter how many I have.

1 or 1.5 workers, does not matter to me. What matters is do you have citizens working unimproved tiles? If so you do not have enough workers or are not managing them well.

 

[onomastikon]

I like your analysis very much, and I find it useful.
"However".... I think your main gripe with the "old" rule of thumb comes from (perhaps mistaken) conclusions drawn from (correct) premises, namely that population is power and workers lower population. The mistaken conclusion that having (x number) of workers lowers your power by lower the population arises, I think, from the fact that there will be many situations which cap population, namely sanitation. Before discovering Sanitation, you might as well let your big towns produce some population-killers anyhow, and before discovering Construction, those towns who need an Aqueduct (or those towns who cannot afford one at that time).
Once I get my RRs done, I tend to keep my workers around for cleaning up pollution later. But if you have way too much, you can always disband them or join them to cities?

 

[Soter]

Hi
I'm learning how to use workers more efficiently and now build about 1 for every city I found. In the past I would build far fewer, maybe 4 for 10-12 cities, then go to war and capture foreign workers.
In the case of newly founded cities, a worker is often the best build option. If you have to build them, and you're not on a tiny landmass, I find there's always work for the them.
Road-building for one, is a very important aspect of the game. You fill your coffers with gold and prepare for a Republic government.
Connecting your cities is of prime importance, as well. I find that it is a turning point, from a scattered people you become a nation, sharing resources and luxuries while reducing corruption. It becomes easier to control your citizens, thus work them to their full potential.
But what I find the most rewarding, is that I can defend my nation with a small but faster and better trained military. I'm not into amassing huge numbers of military units. I prefer a compact but versatile military force that is faster and better trained than his enemy.
All in all, I guess it depends largely on the game parameters. Like most things in Civ.

 

[vmxa]

The concept of capturing workers to get things done is going to cause delays on levels and maps where that is even an option.

Often they will not have workers enough to get or towns to raze for them and worse yet for those who play above demi is you will not be in a position to take things from them right away.

What one tends to see in most games posted for advice is that they did not do that well in getting tiles improved. Mostly this is a function of a lack of workers.

 

[Own]

I say 2 per city, 1.5 if industrious. Once you're done with all the worker jobs (which should be very fast) most of the workers can join cities. Not all, as you might need to change a mine to an irrigation, and vice versa. This way hardly any pop is wasted.

I usually in every city I build I have that city build 2 workers before anything else. If it would be done before growth, then warrior, worker, warrior, worker.

 

[VladTepes]

Don't forget about pollution cleanup. Before you have mass transits and recycling centers, workers are invaluable. I have over 130 as industrious so I can conquer and pollution anywhere in one turn.

A less painful way to create workers is to do it right after you finish aqueducts and hospitals. The city should have stored up a full amount of food, so it's ready to grow immediately. A size seven city with enough shields will produce the worker and not lose a pop (well, it stays the same anyways, at least for the near future.)

 

[SJ Frank]

A few comments:

1) If you think you're going to have too many workers, then don't wait for your cities to grow those last few populations, merge your excess workers into those cities for instant-population.

2) Building workers doesn't kill population, it just temporarily stores the population in the form of a worker, which you can recover later. In fact, building workers can be the fastest way to grow population, if you build them in a city that takes 2 turns per worker, and merge workers into cities that take 20 turns to grow.

3) Building workers doesn't kill unit support. Yes, it does mean higher unit support in the short term, but because workers allow you to build those things that increase unit support (and generally improving the economy) faster, having more workers is better for unit support in the long term.

For example, in almost every game, I enter Republic with unit support of well over 50gpt, with the bulk of it being taken up by workers. During the first 10 turns or so, I have hardly any excess gold even with science turned down to 0%. But, because I have so many workers, each city is able to receive the kind of irrigation / mining / forest-chopping that it needs. By the end of the first 10 turns of Republic, I can usually grow my economy by about 50gpt. That kind of rapid recovery from the "Republic Recession" is not possible without a lot of worker support.

4) Sometimes, that "food-surplus" term in the final equation can be huge. A very good "builder's" strategy that I use, is to build a granary in each core city after growth to size 7+, and irrigate enough tiles to give the city 5+ food per turn. This way, each city can grow a population every 4 turns, and reaching size 12 in no time. With this strategy, each core city needs about 5-6 workers per city, both to improve enough tiles for the city to use while it's growing, and to prepare the mountain and hill tiles for the city to use once it reaches size 12.

5) It's not just how many workers you have, it's also how early you have them. In your equation, it is assumed that the worker is born at the same time as the city. It is a very important assumption, and it points to a very important strategy. If every city builds a worker as the first thing that it does, then 1 worker per city is all they need. That's what I've been doing lately, and in the last few games that I played, I fully improved my core city before I switched into Republic, making a lot of what I wrote in points 1-4 moot.

Conclusion: I like your conclusion

I think 1 worker per city is good enough, if you bulid them early. But I would also say that it doesn't hurt to have more than 1 worker per city, especially if the worker comes cheap (via worker factory).

 

[Kuhal]

For the first 2500 years, and non industrious civs I may have 2 per city for my core cities (usually capital plus the 6 nearest). I like to have one on mining/irrigation while another is on local/inter-city roading.

for the next 6-10 cities and up till 1AD, all but my fontier towns are roaded and have enough improved tiles to prosper. Jungle clearing is saved for later if at all possible but if spare workers are around then they will start early.

I am biased towards building cities far apart so that they overlap minimally while not wasting any land tiles. I like having 6 or more 30+ size cities in the late game. No particular reason for this and it's hardly efficient but it's a bias and there's no rational reason for it

 

[Desertsnow]

I tend to play the opposite way, building fewer workers in the early game, when I tend to concentrate on settlers for expansion, and then train a bevy of workers in a fell swoop, usually at the height of the REX. I'm biased toward wider spacing and bigger cities myself, though not necessarily as huge as size 30+.

 

[Brain]

I'd like to mention another thing. When my workers are sitting idle I like to sign ROPs and send them to road/railroad my neighbors' land. This way the AI has more gold to trade with me.

 

[Khan_Asparuh]

I don't know if every player has my style, but for me the workers have another very important function too. They can transfer population. So when I see a 2 turn worker factory (ideally with no waste at all, 5fpt and 5spt), I cant resist. Just count. Those workers can improve tiles and then join them. So a metropolis grows (later) at a cost of 10 food. This means 50 food cheaper, you see
Sometimes, especially in republic, I keep some settler factories going for ages. A bit more expensive on the shields side but requires less support. The same thing, the same purpose, joining. And it's always usefull to have some settlers standing by.
And I always try to snatch a worker out of my maxed cities when they can't grow anymore.
So I find myself with several hundred workers on a standard map, this is awesome after sanitation. When steam comes, my lands are RRed in no time and then I can start joining. Actually that's why I usualy prefer Iroquois to Maya when I want to play just for fun. Industrious is nice at some moments, commercial is awesome the entire game.
Have I mentionned how much I hate ICS?

 

[Tone]

1 or 1.5 workers, does not matter to me. What matters is do you have citizens working unimproved tiles? If so you do not have enough workers or are not managing them well.

Spot on! I tend to follow Qitai's advice on this issue: if a city is working an unimproved tile, its next build should be a worker.

1-1.5 per city is a rough estimate and as has been stated elsewhere, it depends what is going on in the game. I think Moonsinger said that as a rough guide, you need around 2/city in the MA (presumably to correspond with the high population growth of a more advanced government) and 3/city in the early IA (for rails). Many of these can then be assimulated back into the cities with hospitals in place.

I like the opening post though. Some sound analysis, Tesuji and this should convince some doubters that population+workers=power. However I do enough mm already so I think I'll just stick to the rough guide for now!

 

[eldar]

Regarding jungle/swamp cities:
I keep building workers from these cities, who I just stack up on the first jungle tile, and keep chopping away. I'll usually stop at 6 (24-turn clearance dropped to 4 turns; a little less painful!). This has the nice side-benefit of giving me more workers for other tasks later on.

 

[Tesuji]

I must say I also follow the custom to have each new town first produce a worker. It means I have to spend little thinking in what a new town should do first, provided of course the growth is on time. But it's obvious that a town with a granary needs half the food to produce the same worker so that a worker-factory should be preferable. But somehow, whenever I see a good candidate for a worker-factory I can't resist the temptation of turning it into a second settler-factory. At least as long as the REX phase seems to require it. But after the REX phase, I'll already have enough workers anyway...

I agree that the proper measuring stick is whether you often have citizens working unimproved tiles. But I find that by following my own rule this rarely ever happens. If it happens it's usually only because somehow my worker-population has become concentrated to one side of my empire and need a few turns to migrate to the unimproved land. But usually this can be overcome with proper planning and micro-managing.

As to producing workers and join them elsewhere: it's not free of course, you waste 10 shields. But here too applies the logic that population generated by a town with a granary is cheaper than population generated by a town without granary. And the difference is even bigger for cities (pop >6) without granary. Since a granary costs 60 shields, it's easy to see that the 10 shield cost of a worker is easily recovered if you can do away with building the granary.

But population-management was not the subject of my post.

 

[Abdomination]

When a city reaches a max size, say 6 or 12 and you dont have the neccessary improvement, build workers out of that city. Then, when you produce an aqua or hospital, join workers back into the city for some instant growth