To Sacrifice Granaries?

[Jon Shafer]

Under what situations do people see not building Granary(ies) a good idea?

[nbarclay]

I've been known to defer granaries on occasion when I had a nearby rival I wanted to take out early. The Gallic Glory AU game was a classic example. But it would take an extremely strange map for me never to build any granaries at all. Sooner or later, I'll almost certainly want to do some serious REXing, and I regard granaries as an integral part of that. (Note, though, that I don't play Deity.)

Nathan

[ZargonX]

As nbarclay says, I really only pass up on granaries if there is an immediate strategic reason for doing so, such as whupping up on an extremely close neighbor. Then, once I've cleared some extra land from my backyard, I can switch to a proper REX mode.

[Jaybe]

I will rarely BUILD granaries, maybe when food surplus shrinks to one in a town during the REX period. Perhaps I should change that?

[nbarclay]

Jaybe, you seem to view granaries as a way to help growth in cities with low growth rates. I view them as a way to help cities with normal or especially with high growth get even still higher growth to crank out settlers and workers quickly. Keep in mind that almost any core city at size five has the production to crank out workers in two turns and settlers in anywhere from four to six. The only thing stopping cities from having that kind of REXing power is the lack of enough food. Build granaries in high-growth core cities and you double (possibly minus some loss to rounding) the cities' growth rate, allowing them to build workers and settlers up to twice as quickly. The effect of that on REXing speed can be truly enormous. (Note that you'll have to use some combination of MPs, luxuries, and the luxury slider to avoid happiness problems.)

The trick is to build the right number of early granaries to fit the land you have available. If a granary won't be done in time to help you with your REXing, it's probably not worth building until you're ready to take the city past size 6 (when the food cost for growth is twice as high).

Nathan

[Aeson]

When playing CivIII 1.16f-1.17f

[nbarclay]

Originally posted by Aeson
When playing CivIII 1.16f-1.17f

Why?

[Mountain Sage]

Don't build granaries, build Pyramids instead

More seriously, granaries helps you with REXing, then you can use them for speeding up the size of cities to 6, 12 and 12+.
A bigger city gives you more shields (i.e. improvements and/or units) and gold (happiness, trading etc.).

[Arrian]

I have taken to building granaries more often (Nathan is actually the one who finally convinced me of their power), but I still will sometimes not do it in the early game.

Usually such a decision coincides with ultra-early warfare. Granary = 60 shields. Barracks + 2 archers = 60 shields (militaristic civ). Smacking the closest AI civ silly in 2500bc? Priceless.

-Arrian

[Dominae]

Somewhat counter-intuitively, if your capital produces a lot of Food, you're sometimes better off producing a quick Settler then focusing on the Granary next. This way you can have the second city take care of military units and such, while the capital focuses on just pure REX. To be really safe (especially in MP), you have to produce quite a few Warriors out of your capital city at the beginning of the game (unless you really want to live on the edge). But this puts back your Granary and therefore your first Settler even further. So sometimes I find it better to go Warrior-Warrior-Settler-Granary in a high-Food capital.

Expansionist civs are really advantaged with respect to Granaries (makes sense, no?). They can use a few early Scouts to determine if there are any better Granary-city sites around the capital. If so, the capital produces a Settler at the first opportunity, then more Scouts/Warriors. If there are no better sites, or if the capital is already a perfect site for a Granary, then the Granary may be started immediately in the capital. All the while the Scouts are contacting other civs and determining if there is going to be any threat while the Granary is being built, which is a time of weakness. Non-Expansionists do not have all these benefits, and most importantly have to decide to blindly research Pottery from turn 1, which may not always be the best option.


Dominae

[Aeson]

Because in those two versions you could expand faster by demanding cities from the AI. Certainly the ratio of player to AI expansion was much greater as the AI is building the cities for the player. The limiting factor was how fast you could get to a power advantage (mainly number of cities, in 1.16f also number of units, in 1.17f quality of military units). On anything other than Deity, that was the third city built. Even on Deity in 1.16f it could sometimes be that quick as well.

Building a granary cut out unit builds, and usually is a bit slower to the 3rd city (definitely slower in the case of a Settler from a hut). Once you got to the power advantage and started demanding cities, then Granaries are worthwhile.

[DaveMcW]

From chieftain to emperor, you can do a peaceful REX with as many granaries as you like, especially one in your capitol.

But on deity and multiplayer, sinking 60 of your first 100 shields into an improvement will leave you defenseless against the early rush. I don't buiild granaries at all if I start next to an aggression level 5 AI, and I build settler, granary for aggression level 4. In multiplayer I do the same, based on the estimated aggression of my opponents.

[trev]

Mostly I do not build early granaries, my preference is to build armies in between worker and settler builds, this is necessary to keep a reasonable parity of military power with nearby rivals and send armies to explore and defend settlers and new towns from barbarian attack

[Kirby]

Early granaries are too slow on deity. If you don't constantly pump military and settlers the other AI's will be all over you because you are too weak.

[Yahweh Sabaoth]

Would someone here be so kind as to go into greater detail as to what areas you would build a granary in, and which you would not?

For instance, take two games where you're playing as Russia. On one game you start out next to a lot of wheat on grassland. In the second game, you start out in the middle of a forest, with game.

I would build a granary in the second instance, and just another settler in the first (with a scout, or warrior, of course).

But I'm often wrong in these matters. What would you all do?

[Jon Shafer]

Ideally I want to get a +4 food bonus with a Granary for my Settler pump(s) with nearby cities to take up the extra food on the 3rd turn.

If that helps any.

[Cartouche Bee]

Pyramids aside (a wonder that I do shoot for), I currently only build granaries when I have 5 surplus food in a city. Otherwise the expense is too great, early in the game. Now, 60 shields produces 2 small cities, farming workers off those 2 cites (assuming 2 surplus food) triples the rate of growth in the capitol or any other city, if you want, instead of the doubling effect of a granary in the city where it was built.

[Jon Shafer]

You lose shields, but gain cities long-term. And if you look at things differently, you gain those shields back from the cities you've built (sooner), so you gain all the way around as long as you can get an efficient and well-defended REX going.

[Frank Johnson]

I tend to put cities in a pattern of spearman -> settler....which is usually good enough for the standard growth rate. Having a city with a grainery seems to throw that balance out of whack.....giving me population points which are unhappy and sitting on unproductive / unimproved terrain, which my worker force is unable to make into productive land while taking advantage of the new cities as well.

I'm unwilling to have unescourted settlers, and warriors just don't cut it, even against barbarians, so am I going at this the wrong way? Would I see a signifigant change in my growth rate as a civ if I had a city just for settlers, and others specializing in the spearmen? If so, I doubt it would be as radical as the benefit of changing my city spacing to 3 squares, but.....what kind of effect does it have?

[Dominae]

Originally posted by Frank Johnson
Having a city with a grainery seems to throw that balance out of whack.....giving me population points which are unhappy and sitting on unproductive / unimproved terrain, which my worker force is unable to make into productive land while taking advantage of the new cities as well.

You can use the Luxury slider to ensure your populace in your Granary city are Content. As for unimproved terrain, remember that your Granary city does double-duty as a Settler and Worker pump, so you should not be falling too far behind in tile improvements (quite the opposite). Getting more Workers faster also means that you can hook up Luxuries faster (in general), providing another way to deal with the unhappiness caused by "extra" citizens.

I'm unwilling to have unescourted settlers, and warriors just don't cut it, even against barbarians, so am I going at this the wrong way?

For me, Warriors always cut it against Barbs. Remember that if you're using close city-spacing, escort duties are short, and therefore you should have enough Warriors roaming around to explore and escort.

Would I see a signifigant change in my growth rate as a civ if I had a city just for settlers, and others specializing in the spearmen?

The idea is to supplement the standard REX strategy (finish a Settler in a city whenever it hits size 3) with a Granary/Worker pump strategy. You expand the fastest if you use both. If you feel your expansion is going well, your highest production cities can stop building periodic Settlers, and be allowed to grow in order to increase their production even more. Expansion and the early-game are not (at least in my mind) "all or nothing" propositions.


Dominae

[Cruddy]

Originally posted by Frank Johnson
I tend to put cities in a pattern of spearman -> settler....which is usually good enough for the standard growth rate. Having a city with a grainery seems to throw that balance out of whack.....giving me population points which are unhappy and sitting on unproductive / unimproved terrain, which my worker force is unable to make into productive land while taking advantage of the new cities as well.

I'm unwilling to have unescourted settlers, and warriors just don't cut it, even against barbarians, so am I going at this the wrong way? Would I see a signifigant change in my growth rate as a civ if I had a city just for settlers, and others specializing in the spearmen? If so, I doubt it would be as radical as the benefit of changing my city spacing to 3 squares, but.....what kind of effect does it have?

It really depends on the starting terrain. The best settler producer cities have ample food AND the resources to build the settlers. If there isn't a suitable site, then a granary can tip the balance - if you have grassland but no cows, wheat or flood plains is the situation here.

However, the extra wait for the granary to be built can count heavily against you, as the opponents who have more favourable starting positions will out build you (they do anyway on Deity...).

Unesecorted settlers will USUALLY be OK. If you have scouts or warriors out exploring, they'll find the barbs and other Civs. Having barbs pillage an unprotected city isn't good - but you can bounce back from it (unlike Civ2). Remember, military units wandering about will prevent barb uprisings.

Putting you city spacing down again, depends on the terrain. Before you get luxuries hooked up with roads, your biggest city will be a 4 anyway - assuming they can get the food. If you have a lot of improvments that NEED doing (plains irrigated, hills mind rather than forest or worst case lots of mountains) then the extra workers more than make up for the inconvenience of abandoning it later.

[Yahweh Sabaoth]

And never forget that, despite what civ you are, there's a time to stop REXing and building workers and just conquer! You will get about 50% of the needed worker supply from those conquests, I find. Then again, I play regent, so maybe I'm totally off-base.

I would agree that it's best to get out and get the iron and horsies before the others do... if that means putting off the granary and accepting fewer, but more well-placed, cities at first, then so be it.

[gunkulator]

The problem I have with close-pack is that these first few cities constitute your empire's core. They'll have the longest oppurtunity to grow, the lowest corruption and the longest time to generate culture, gold and science. Large overlap hampers your future.

But then again, I'm not that great of a player so what do I know.

[Yahweh Sabaoth]

Originally posted by gunkulator
The problem I have with close-pack is that these first few cities constitute your empire's core. They'll have the longest oppurtunity to grow, the lowest corruption and the longest time to generate culture, gold and science. Large overlap hampers your future.

But then again, I'm not that great of a player so what do I know.

This was definitely an issue in Civ2, but depending on the number of other civs you're up against, I think it's worth sacrificing a little largesse for more early-on power.

But I'm not the greatest player either.

[Frank Johnson]

Their growth really isnt hampered untill the industrial age. Plus, if cities are "packed" together near your capital, all the squares that are in the "good" zone of low corruption can be worked in the anicent era instead of the industrial era.

The only problem is higher maintance costs involved in making lots of city improvements.....but....for taking over people early in the game and simply pumping units, its ideal.

[Jon Shafer]

Just design your city grid with the plans of disbanding your capital later on. That's what I tend to do now when I have the chance to build up more than a few cities. Later on when you reach size 12 or so with your core cities you can disband/move your capital and the spacing will be evened out.

[MPatton]

I have never built a granary. And definitely don't need em now cause most of my major cities are maxed. I also have a better way of increasing my cities.
raze large enemy city send all those workers to current cities and add to city
Also find two cities that will produce reaonably well and have them build cities. when they build a settler send two new conquerewd workers to add to city.

When I used WWI style it was about even now with WW2 style I am seeing more and more open land
Cause I build 3 settlers every 5 turns and raze 3 to 6 cities every turn
workers acquired from razing = half of current city
Once I gained 18 workers from one city. It sure beats the hell of squelching all them resisters
My current enemy has many large cities for me to raze. Smallest is probably about 22 and biggest is like 40
Figures it would be the chinese that I am attacking too.
I figure I razed a city the size of L.A.
I also have enough workers to work my terrain so any new ones get added to city

[Jon Shafer]

What about before there are cities the size of LA to steal Workers from?

Granaries are most important early on to help your REX along. How do you do that?

[Dauphin]

Granaries halve the amount of time you need to increase a population, but then settlers building new cities have a similar value of bonus.

A size 3 city with each tile working 2 food will have an excess of 2 food. With a granary growth is in 5 turns.
Two size 1 cities with each tile working 2 food will have an excess of 4 food. Neither have a granary but you grow two pop points in 10 turns.

My decision to build granaries revolves around which method produces the greatest benefit in food, commerce, production and strategic position.

[MPatton]

yeah but it also increases costs. You must pay 1 gold piece for each one. Which makes it harder to upgrade, forces you to have a higher tax rate so you research slower. As is without additional help I would break even at 30% and a 70% science for a tech every 5 turns. But with trading, a wonder and tax specialists I make 300+ gp a turn which I need a steady income to upgrade. I just spent 7000 gold upgrading infantry to MI
One option would be to build em early and then when it maxes out sell em. Once they max out you don't em anymore. Or capture or build pyramids. What is REX?