Managing uneven town growth

[Delphi456]

Assumptions: Republic, Emp/DG difficulty, no lux's, all land settled, adequate number of workers, first two towns have graineries (former settler and worker pumps), not strong enough to goto war, can't trade for lux's.

This is a scenerio I face nearly every game. The first 2 towns grow at such a fast pace that the lux slider is wasting gold trying to keep those two towns happy. Here are the options I see: (1) turn some pop to specialists until other towns catch up in pop. (2) slow research to store up gold for cash rushing graineries in other towns, (3) build workers to add to other towns. I wonder how other players solve this problem.

 

[Daeron]

It's a bit difficult to offer advice without a specific save as example. Generally, it depends on what you're trying to build. If the extra pop doesn't improve the turns to build whatever you're building and forces the lux slider up, you're better of putting them on science.

To be honest, I can't recall many games where I didn't get at least one luxury. Maybe the solution to your problem is in settling strategy and connecting your trade network to your neighbours?

 

[DWetzel]

Assumptions: Republic, Emp/DG difficulty, no lux's, all land settled, adequate number of workers, first two towns have graineries (former settler and worker pumps), not strong enough to goto war, can't trade for lux's.

This is a scenerio I face nearly every game. The first 2 towns grow at such a fast pace that the lux slider is wasting gold trying to keep those two towns happy. Here are the options I see: (1) turn some pop to specialists until other towns catch up in pop. (2) slow research to store up gold for cash rushing graineries in other towns, (3) build workers to add to other towns. I wonder how other players solve this problem.

Either 1 or 3 depending on the exact circumstances. Option 4, depending on how tight your city spacing is, is to abandon them and give the tiles to adjoining cities. I wouldn't do this all that often, but when it works, it works.

 

[Spoonwood]

What sort of VC do you have in mind? Do you need to research? I know that during the expansion phase of some games, I've had luxuries as high 60%. If playing a conquest or domination game, I wouldn't worry about taking the luxury slider as high as the sky, since you can use pointy stick research, or import a luxury from the AI for gpt, and then disconnect the trade route to your capital for a tech. Or perhaps steal them.

In a space, 20k, or diplomatic game where you need to do research, you can consider RoPs and roading up the AI's territory if they don't have extra luxuries for you.

A save would help us to help you.

 

[Delphi456]

This question is strictly theoretical. It's about the economic trade-offs of the game.
Pretend lux's don't exist! (for Spoonwood, assume VCs = Con, Dom and SS) Yes, I need to research.

Assume there are only 4 towns: A, B, C, D.

A: pop 5, grainery, 0% corruption
B: pop 5, grainery, 0% corruption
C: pop 2, no grainery, >0% corruption
D: pop 2, no grainery, >0% corruption

In order for A and B to be content, the lux slider must be set to x%. However, this causes C and D to waste gold on overhappiness. Since A and B will grow twice as fast as C and D, this will cause the lux slider to go up each pop growth for A and B(JUST ASSUME THIS IS TRUE!!), causing C and D to waste even more gold (even when corruption is taken into account).

Whether you realize it or not, this situation happens TO EVERYONE, EVERY GAME, to a greater/lesser degree. Either you sacrafice food, gold, or shields because of this uneven growth pattern, unless you have developed a formula for perfectly harmonious growth.

@Daeron: seems like you base your decision soley upon whether it affects number of turns to complete a build.

@DWetzel: Hmm, I have no idea how Option 4 has any bearing upon my question.

@Spoonwood: Sounds like you set the lux slider to the highest common demoninator, thus promoting maximum pop growth. Correct? Even if (in your scenerio) 60% causes all but a few cities to waste gold on overhappiness. Correct?

 

[Spoonwood]

You don't necessarily waste gold, because there exist other things at play in terms of what gets you gold. We also preferably will consider the time period in which gold "gets wasted" and whether that outweighs having more gold at a later date. I haven't done exact or inexact calculations in every game, but my general sense comes as that you really don't lose all that much gold by raising the luxury slider in the long term... especially when you want to do research. Why?

If you want to do research, you want to become a Republic ASAP. In a Republic pre-size 6 cities have 1 unit support. The general advice on workers comes as that want at least 1.5 workers per town as industrious, and 2 workers per turn as non-industrious. With no military in a Republic whatsoever that ALREADY puts you over the unit support limit. So, if 1. you have no luxuries, and 2. you don't raise the luxury slider to at least get your core towns to size 7, then you still end up losing gpt in unit support. If you raise the luxury slider, you may lose money immediately by raising it, but you'll make at least some of it back (if not much or all of it) in less money going towards unit support. Also, with bigger towns you have more shields for say 1. producing a market which gets you cash directly, or 2. indirectly getting you cash via a library/university earlier which can help with selling techs for gpt earlier. You also have more cash via more squares you have citizens working.

So, even though I haven't done exact or inexact calculations, I simply don't see how you lose all that much cash over a significant period of time by raising the luxury slider under your assumptions Delphi456. And it at least seems as clear to me as the noonday sun that the most economical way to play the game when you want to do research comes as to raise the luxury slider as much as needed until your core towns hit size 7.

Thanks for spelling out your assumptions so neatly!

 

[Bucephalus]

@DWetzel: Hmm, I have no idea how Option 4 has any bearing upon my question.

What I suspect DWetzel means is that you could mm your cities so that the food tiles which are causing unwanted growth in one city are used by a second city which is lagging behind in growth; the effect would be that a) your fast growing city would remain stagnant (thus requiring neither specialists or raising of the lux slider) and, b) the other city would catch up in population more quickly.

 

[DWetzel]

What I suspect DWetzel means is that you could mm your cities so that the food tiles which are causing unwanted growth in one city are used by a second city which is lagging behind in growth; the effect would be that a) your fast growing city would remain stagnant (thus requiring neither specialists or raising of the lux slider) and, b) the other city would catch up in population more quickly.

Kinda sorta basically this-- though if you could do that easily, you'd have already done this.

If I have five densely-packed cities, extreme example:

A,B,C,D: +1 fpt, lots of city improvements, good production, with a mountain left to work
E: +9 fpt (four mined cows, irrigated wheat) and nothing else (settler pump)

it may be worthwhile to completely abandon E, give a cow each to the other four cities, which are now four +2 fpt, very productive cities that will be able to work a mountain, and give benefits to the commerce of those tiles via markets/libs/etc. This nets those cities about 5 spt each, which could be a Very Good Thing.

And yes, it'd be probably better even here to just rush a bunch of settlers (assuming you have somewhere to put them!!!) from E, and let it just have wheat + scientists. There may be rare occasions where this is suboptimal, mostly involving corruption and getting productive cities to round numbers. I'm probably overstating the case by even mentioning it though.

 

[rysingsun]

In your example of city E, I would ask why you did not settle other cities close enough to share in the food tiles. Your example demonstrates the wastefulness of not sharing. Food sharing is particularly necessary with floodplains, since there is practically no benefit to wasting, say, five floodplains all on one city that has hardly any shields to actually produce any settlers.

Another way to deal with a mismatch in food production is to be slack on producing workers from non-settler pumps. You gave the example where you already have enough workers, so continuing to produce them from the food king causes worker-overpopulation. But, getting to your first example, where you only had two food-cities, you could have planned it so that outliers produced fewer workers and you played catch-up later with the settler pumps.

 

[Delphi456]

Spoonwood pretty much spelled out what I had feared; there really is no fine-tuned method for combating this problem. I was hoping someone out there would have an elegant solution, as routinized as the settler factory. My new stategy involves building a grainery in every city (that's not highly corrupted), but only having one city building a grainery at a time.

 

[Spoonwood]

Spoonwood pretty much spelled out what I had feared; there really is no fine-tuned method for combating this problem.

I think I also tried to say that it might not really work out as a "problem". How much gold do you really lose that way? Have you done any calculations and compared it to not running the luxury slider?

 

[Lanzelot]

You don't necessarily waste gold, because there exist other things at play in terms of what gets you gold.
...
So, even though I haven't done exact or inexact calculations, I simply don't see how you lose all that much cash over a significant period of time by raising the luxury slider under your assumptions Delphi456. And it at least seems as clear to me as the noonday sun that the most economical way to play the game when you want to do research comes as to raise the luxury slider as much as needed until your core towns hit size 7.

Spoonwood definitely hit the nail on the head with this answer. There may be exceptions, but I think as a general rule we can say that raising the luxury slider will always be the best solution to your problem. In general you will always gain more than you loose. Let's try a little hypothetical calculation on the example above (A & B size 5, 0% corruption and C & D size 2, a bit corruption):

• When raising the luxury slider by 10-20%, C and D can surely not "loose" more than 1-2 gpt each. Let's be pessimistic and assume we indeed loose a total of 4gpt. (Which are not even lost completely, because they raise your score, and that might be a factor, if you are playing for a high-score game! I assume the situation we discuss here is quite early, around 2000BC.)
• On the other hand, A and B each can now put one more citizen to work. Let's be pessimistic again and assume they can work only an "average" tile, e.g a roaded irrigated plain. So for the 4 gold that C & D "loose", we gain 4 food, 2 shields and 2 gold! If A or B still have an unworked bonus tile available, that would even be more. A shield is worth 4 gold, so we are even making profit here! And of course the extra food is invaluable! (As Spoonwood said: we want to reach size 7 asap, especially since A and B are uncorrupted!)
  In the early stage I am willing to go even 60-70% lux, if that allows me to use every food tile available, because it will mean more towns & workers in the near future. And more towns & workers mean "more everything" a little time later...

Lanzelot

 

[Lanzelot]

My new stategy involves building a grainery in every city (that's not highly corrupted), but only having one city building a grainery at a time.

I don't think that's necessary/advisable. It depends on VC of course, but in general I would say: build barracks and horsemen in the slow-growing towns.

 

[Delphi456]

I completely understand what you guys are saying, but I need a little more clarification on a couple points:

@Spoonwood: Since you only mentioned you would adjust the lux slider up until core towns reach 7, what do you do with the lux slider when cores grow >7? You were somewhat open-ended about this point.

@Lanzelot: Since you are in agreement with Spoonwood about letting towns grow even at cost of a few lost gpt, I don't see how you can be opposed to graineries, which would get all other towns to size 7 faster (so, less unit support, more gpt, more spt). And since I'm building them one town at a time, it doesn't preclude me from implementing your barrack/horseman idea. Can you elaborate on the downside of my strategy? [VC = Con, Dom or SS]

 

[Spoonwood]

@Spoonwood: Since you only mentioned you would adjust the lux slider up until core towns reach 7, what do you do with the lux slider when cores grow >7? You were somewhat open-ended about this point.

True, I didn't provide that detail. It would depend on victory condition. But, my general advice and rule of thumb comes as to not raise the luxury slider above 30% once towns hit size 7. At least if you want to do research. Going too high sacrifies current beakers for future ones. I rarely remember encountering a situation where I can't get luxuries by the time most core towns get to size 7. On an archipelago 80% map, but then I've probably get an upper level 20k game going. And in that sort of situation I WANT early research, so I definitely won't sacrifice current beakers for future ones. Maybe in a space or diplomatic game where I didn't get luxuries I might do that, if I didn't worry about getting an early tech in time.

Lanzelot has a very good point about shields converting to money and vice versa.

 

[Lanzelot]

@Lanzelot: Since you are in agreement with Spoonwood about letting towns grow even at cost of a few lost gpt, I don't see how you can be opposed to graineries, which would get all other towns to size 7 faster (so, less unit support, more gpt, more spt). And since I'm building them one town at a time, it doesn't preclude me from implementing your barrack/horseman idea. Can you elaborate on the downside of my strategy? [VC = Con, Dom or SS]

Well, when I gave that recommendation, I mainly had the early expansion phase in mind. Let's say we plan for Domination, Conquest or 100K, where fast (and early) expansion is important. In these cases two early horsemen make more of a difference than an early granary.

Just consider the following point: unless you are very lucky, many of your 1st and 2nd ring towns won't be on a river. So they can grow only to size 6. (Construction, and therefore aqueducts, is still a long time ahead.) Usually by the time the town finishes the granary, it has already grown to size 4. So the granary will help only with the last two growths. After that it will just eat 1gpt and doesn't do you any good, until you've also built the aqueduct.
So basically you invested 60 shields, just to save 20 food! Instead of requiring 50 turns to grow from 1 to 6, the town needed 40 turns. That's not enough to justify a 60 shield investment (and a 1gpt expense). The two early horsemen on the other hand will by now have gained you 2-3 extra towns! (At that early point, it's much easier to conquer land, because the AI towns are not yet well-defended.)

Usually for a military VC I have at most 2-3 core towns with a granary, which do all my settler/worker production. All the other towns don't get a granary and start with horsemen production right away. There may be exceptions to this rule: if I'm lucky and get another first rate city location (river, food bonus and shield bonus), then this city might get a granary too.

With a military VC, the game is usually over by the time the expansion phase ends... So no need for any more granaries. (Especially if one of my neighbors was kind enough to build the Pyramids for me... If you start with Masonry, you should always gift it around to your neighbors: they will spend hundreds of shields on the Pyramids then, instead of on spearmen. So their towns will be easier to take over, and will provide more value to you... ) For a non-military VC there's still time to start building granaries now (if you didn't get the Pyramids...). In my opinion, the granary is almost useless for the growth phase from 1-6 (as outlined above), but in the growth phase from 7-12 (after you built aqueducts) it becomes worth-while. The only exceptions are worker/settler factories or powerful city sites, that you want to build up early.

Another tip that might be helpful: in some situations it may not even be necessary to build a granary for the 7-12 growth phase: if the city doesn't have a food bonus, just two irrigated grasslands, it will need 50 turns from 7 to 12 without a granary and 25turns with a granary. Even 25 turns is a very long time! If you still have your worker/settler factories up and running, they can probably build 3 workers every two turns (assuming 3 pumps going at the rate of 1 worker every two turns). So if you join workers into a city, it can go from 7 to 12 in only 3-4 turns!! Just leave the pumps running for 20 more turns before shutting them down, and then all core cities should be at size 12, and you've saved half a dozen granaries!

Lanzelot

 

[Daeron]

Specializing a town as a workerpump is a lot more cost effective than building granaries everywhere. One 2-turn worker pump can double the growth rate of 5 other +2 food cities over 10 turns. Basically, it has the same effect as five granaries.

A city at floodplains might even be capable of a 1-turn worker pump, barring high corruption of course. Compared to that, building granaries is atrociously wasteful.

 

[Delphi456]

Lanzelot, what level do you play at where you can overrun the AI so quickly with just horsemen? I can't believe this would work at DG level. Also, since you said you can have a military VC by the end of the expansion phase, you must be playing pangea?

You did remind me that I can irrigate 2 grasslands instead of building a grainery. Can't believe I forgot that trick. Thanks.

I only build the grainery AFTER I build aquaducts (if aquaducts are required).

If you assume the building of a grainery is 60 'needlessly wasted' shields, I could also make the claim that building 6 workers in city A to join back into city B is also 60 lost shields (even though I join workers all the time, too).

 

[Bucephalus]

You did remind me that I can irrigate 2 grasslands instead of building a grainery. Can't believe I forgot that trick. Thanks.

This only true if you are no longer in Despotism.

 

[Lanzelot]

Lanzelot, what level do you play at where you can overrun the AI so quickly with just horsemen? I can't believe this would work at DG level.

Well, a pure horsemen Domination/Conquest will certainly work up to Monarch. It may still work on Emperor (see some of ignas' excellent games in GOTM/COTM a while back). Personally, on Emperor I start with horsemen, but then switch over to Knights towards the end. (For a recent example see my GOTM 99 submission from last month.)

On Demigod/Deity it still works ok for the first (sometimes the second) AI, but by the time you start on the second/third AI, they will have pikemen, so you need Knights, and sometimes (large maps or archipelago, where it takes a while to get to the last AIs) even Cavalry.

Lanzelot