Granary when playing agricultural civs?

[NickyH]

I'd like to know what you think about granaries when you play agri civs.

I personally build them (almost) everywhere when I'm not agri. I like to get my towns to grow as fast as possible.

Now I'm playing an agri civ for the first time in a very long time, and it seems like the extra food is enough to make me prioritize units or other buildings. Once my cities are at size 8 or 9, and I still haven't built a granary, it feels like a waste to build it, since the city will grow to size 12 quite soon anyway. And after I've reached size 12 I'll still pay for the upkeep for a granary. (Now when I think about it, I should start to sell my granaries once my cities are at size 12. Or does anyone see a downside to that?)

In the REX phase my settler factories got granaries though. That seems to me like the only option for agricultural civs as well.

Edit: I made the poll public, so everyone will see how you voted. It doesn't strike me as a sensitive question with need to hide one's vote.

 

[Othniel]

I think about granaries the same way regardless if I'm AGRI or not.

I will generally have 1-3 granaries in high food location in the core. Depending on what I need, those granary cities will function as settler or worker factories. When building granaries, I'm looking for spots that can get to +5fpt. This is ideal because a city with +5fpt and a granary can grow every 2 turns with zero food waste. Of course, sometimes the core area is food poor and lesser spots have to become the factories.

Usually I don't see any need to have more granaries than this. One or two settler factories, even if they are slow ones such as 10-turners because the core area is food-poor, is often enough to fill out the core area and give me enough settlers to start filling in any captured lands. Once I've expanded enough to get some highly corrupt cities, those corrupt cities will all be on settler or worker duty themselves. This strategy gives me plenty of settlers.

The most efficient way to make cities grow is to let them self-grow from size 1-size 6. Then, to make them grow larger, the best thing is to join workers. With that in mind, I always am setting up worker factories and churning them out. Frequently in my games, if a core city makes for a good 2-turn worker factory, it will stay that way for most of the game, churning out workers and never getting above size 6.

The extra food for being AGRI or if I acquire the Pyramids, is certainly helpful. It can make setting up worker or settler factories more simple. But the critical thing is just getting factories set up in the first place.

 

[NickyH]

Thanks for an answer that gave me something to think about. Has anyone compared the efficiency of joining workers from a worker factory, against building granaries, as a strategy to grow? It certainly sounds like it could be a more efficient way, although it's likely that I'd personally find the extra micro management too boring in most of my games. Unless the advantage of joining workers from a factory is really big, of course…

One thing to consider is that a city that's good enough to become an efficient worker factory is probably not too bad in waste and corruption, and could quickly become one of the really productive cities. (You won't just spend population points only to feed your other cities, you will also spend shields and gold that's not very corrupted.) If you keep using one city as a worker factory for the purpose of joining workers, you'll abstain from a fairly productive city. And before rail roads, the upkeep for workers on the way to their destination would cost a bit. And once there, joining the worker is a waste of 10 shields built in the factory.

Edit: One thing that would probably be a big advantage for Othniel's proposed method is unit support in republic. Consider cities where booth an aqueduct and a granary would be needed for a fairly quick growth above size 6. If joining workers, only the aqueduct is needed to gain the extra troop support. That would probably pay for the extra workers heading for the city where they'll be joined before rail road, and render a decent gain in troop support on top of that.

 

[Spoonwood]

You can pay the upkeep for workers or pay the upkeep for more granaries. I'll pay for more workers, especially since I can develop territory (think chop jungle/marsh) faster. I myself didn't quite get building so few granaries for a while and didn't understand the mass add-in worker strategy. Chamnix helped me along with some nice analysis here http://forums.civfanatics.com/showthread.php?t=277648 I even used that idea in a 5CC Emperor game with Sumeria and it worked out quite well. You can find that in my "Z for Zulu" thread if you like. On wet, warm Huge maps I'd think the basic idea works really as second to none, as I played a Deity-Builder game (you can find it in a thread called "Deity Builder") with the Maya where I simply didn't have a GA until I built Hoover's Dam. I didn't need one as a I lead basically lead in tech from sometime in the high middle ages. I *do* usually end up building lots of granaries everywhere in the industrial age... but that's a bit different since I usually irrigate everything I can, granaries don't take as many turns to build, and I want specialists everywhere I can get them to speed up my science rate. If I warmongered during this period more, I'd probably re-think that. I haven't read the thread, but the "granary or settler" thread here http://forums.civfanatics.com/showthread.php?t=244860 might have some intersting information also.