Granary or GP first in capitol?

[Elowan]

I can build GP from the get-go in my capitol city and I have one other city built (popped a settler from an adjacent GH).

Here's my question:

Should I push out: warrior,warrior,settler (repeat x 1) in the capitol first and then should I build a granary next(knowing that I'll sell it after GP) followed by the GP?

I guess the essence of my question is this: will the time taken up by building the granary be regained in the subsequent building of the GP (larger population)?

I know that some would reccomend building a granary before producing settlers.

Looking for advice from the scarred and aged Wise Elders of Strategy.

[DaveMcW]

For those who didn't figure it out, GP=Pyramids.

If my wonder city is on a river I build a granary first. It is usually size 9 by the time the wonder is finished, which definitely makes up for time spent building a granary, and it is in great shape to build a second wonder.

If no river, the city can only reach size 6 and a granary is not that useful. If it doesn't have bonus food I help its growth with a worker or two.

[ducki]

I generally don't even think about either Granaries or Pyramids until I've got 6 or 8 cities built.

I seem to have the opposite problem from everyone else in the world that plays Civ3 - my towns grow too fast, so I don't want/need granaries.

No, I don't restart more than a couple of times to get away from the middle of the jungle or tundra, I don't use any mods other than Snoopy's Terrain graphics. Maybe I just don't grow as fast as other people, but I'm not experienced enough to deal with it, who knows?

Anyway, my capital is generally warrior-warrior-settler-temple-warrior-settler or something similar, then I move on to either a wonder prebuild or start beefing up military and worker ranks.

If I end up with a city that can crank out the pyramids, I'll build it, but for the culture, not the granaries. I'd rather have the library and lighthouse(or gardens).

[nbarclay]

Even with non-expansionist civs, I'll often do a pre-build, research pottery, and switch to a granary. That's especially true if the capital has really good food production, and even more especially true if the capital has exceptionally good food production and no nearby sites would. My absolute favorite is using either three food bonus tiles or two with one on irrigated grassland to build a settler pump that cranks out settlers in four turns or workers in two. (That requires some use of the luxury slider to combat unhappiness, and often some micromanagement to keep the governor from undermining growth by working forests, but it's fantastic for cranking out a lot of cities in a hurry.)

In the AU 201 game, since it was a large map, I built two warriors before starting my granary but kept them in the city to combat unhappiness until I finished my research. That got the timing right on my research; had I started a prebuild for my granary first, I would have risked wasting shields if an expansionist civ didn't come along to sell/trade it to me. (Bigger maps mean slower research in the early one-city days.)

Nathan

[DaveMcW]

I always build a settler or two before starting a granary, if I even build one. For the cost of a granary you could have 2 settlers = 2 new cities. 3 cities with no granary makes settlers faster than 1 city with a granary.

If you have an bunch of bonus shields I can see how a granary would let you exploit them to their full potential, since your pop growth can't keep up with settler production. But in "normal" starts an early granary keeps you at 1 city for too long.

[nbarclay]

Originally posted by DaveMcW
I always build a settler or two before starting a granary, if I even build one. For the cost of a granary you could have 2 settlers = 2 new cities. 3 cities with no granary makes settlers faster than 1 city with a granary.

If you have an bunch of bonus shields I can see how a granary would let you exploit them to their full potential, since your pop growth can't keep up with settler production. But in "normal" starts an early granary keeps you at 1 city for too long.

The driving cost factor in building settlers is population points, not shields. Suppose you have an industrious civ with two grasslands with shields readily reachable.

Turn 0: Found city, move worker. (Hey, how do you expect a programmer to count?)
Turns 1-3: Work on warrior, build mine. 2x3=6 shields.
Turns 4-5: Work on warrior, build road. 3x2=6 shields.
We have our first warrior, now build a second. From here on in, we'll have as many mines as we need since our industrious worker can build faster than our city can grow.
Turns 6-9: Work on warrior, 3x4=12 shields.
We have our second warrior; now start a granary.
Turn 10: First 3 shields on granary.
Turns 11-20: 5x10=50 shields on granary.
This is the earliest it would be theoretically possible to build a settler, since until now, we were stuck at size 2.
Turns 21-22: Finish the granary at 6x2=12 shields.
Turns 23-28: Build a settler, 6x5=30 shields.

So we've only lost eight turns compared with the theoretical fastest we could have sent out our first settler, but from now on, our capital can turn out settlers twice as quickly. (The loss would only be seven turns if we had a third bonus grassland.) That's not a bad tradeoff if there isn't better terrain nearby (and if there is, we could have switched to a settler any time up to turn 16 without loss). Of course it wouldn't work quite as well for a non-industrious civ or if we didn't have the grassland+shield tiles handy, but I strongly favor industrious civs, and it's a good trick to have in your toolkit for when the terrain makes it practical. By the way, on standard or smaller maps, building fewer warriors can speed up the first settler at the expense of not knowing as much about where the best place to send him would be.

When there are multiple food bonus tiles around the capital, a different dynamic comes into play. A granary in effect doubles the bonus in addition to doubling the city's base growth, so two cattle are like four or three are like six. Unless a new city has similarly high food bonuses, extra population can be worth a lot more staying in the capital than they ever would if sent out to found cities of their own.

Nathan

[ducki]

Wow, what a great analysis, Nathan!
Thanks for taking the time to do that.

I almost never play anything but industrious, too, so I'll have to flag this thread and give that a shot.

[vmxa1]

Ducki, just cut and paste this type of thing into notepad.

[Theseus]

Good analysis Nathan.

I have a more simplistic, sorta rule-of-thumb approach... there are two situations in which I'll likely build a granary:

1) Settler Pump. That's a no brainer.

2) Bigger towns / cities. Given good terrain, I'll try for fewer but bigger towns rather than REX. This requires bonus food tiles and rivers, as well as 2-3 early accessible luxuries (being religious doesn;t hurt). 5-8 pop towns and cities in the ancient era can be powerhouses.

Otherwise, I've got too much to do / build.

[badman]

I really like the Pyramids (they are my favourite ancient wonder together with GL, the only thing I don't like about them, is the fact that they trigger the GA for the Egypts), but they are not essential for my strategy.

If I can't get the GL or I feel strong enough to do all the research by myself, I'll build them.

But I've never tried to build an early granary, I think I'm gonna build one in my next game. How many % is luxury rate, when you have a city that produces settlers every four turns? (emperor)

[ArmaGeddin]

I never bother with either. My cities always grow fast enough. I head straight for the Great Library and I don't try for any other GW until I have it or missed it.

[Random Passerby]

I rarely bother with granaries for some time after the start. Most people seem to favor having a couple cities act as settler/worker pumps and having every other city home-grow its own defenders; I do just the opposite and have a couple "spearman pumps" sending out defenders to cover the settlers that spread from my other cities.

[zorbop]

hmmm. my build order normally goes something like this.

warrior-warrior-settler
warrior-warrior-settler
spearman-settler
temple-settler
now mostly repeated spearman-settler a few times

i preffer to claim as much land as possible instead of trying for GP. normally the only early wonder i get is gardens, and maby great library

of coarse, if my cap city has a cow on grassland.... everything comes much faster

[vmxa1]

nbarclay settled the issue.

[badman]

Yes, but don't forget that during those 8 turns you lost because of building a granary, your second city can start building warriors and settlers.
If I build a granary (since I've read this thread I often do so ) I'll always build it _after_ the first settler. This way, I lose some of the benefits of the granary, but I still have a very early second city.