Why granaries don't work

[Ribannah]

It is now 750 BC. We have 10 cities, and are #1 in population and literacy.

[Arrian]

What civ are you, Ribannah? I assume you are industrious, since you were able to start on the Pyramids so early.

Good work. The "big granary" is powerful. It's also a much bigger investment, and requires a tad longer to pay off. Same basic cost-benifit issues as a granary in the capitol, just on a grander scale.

-Arrian

[Ribannah]

I'm not supposed to give any spoilers until the submission deadline.
You can almost always start on the Pyramids early since you can pick Masonry as the first tech to research. You can prebuild with a Granary, Barracks, Temple or sometimes the Colossus.

[nbarclay]

The problem with waiting for the Pyramids is that you're way behind on REXing until they're finished, especially if you don't build any granaries in the interim. Edit: Of course if you're by yourself on a 'pelago map, that isn't nearly as big a deal as if you have to compete with other civs in a land grab.

And here's a strange thought: with a capital on a river, building a granary in the capital before starting on the Pyramids there could actually speed up the Pyramids considerably. You lose 60 shields up front, but the extra production from extra laborers would more than make up for that over time.

Nathan

[Arrian]

Nathan,

Do you really think the granary's benifit would recoup the 60 shields fast enough to catch the same city that just went straight for the Pyramids?

-Arrian

[Panag]

Originally posted by Ribannah
I had a similar starting position in my present Epic game at Realms Beyond (emperor level), except that it is an archipellago map. I built one Settler and then started my capital on that big granary known as The Pyramids.
Meanwhile I founded my second city on a food-rich spot and cranked out 3 Workers (5 turns each) before starting on more Settlers.
So I'm taking this strategy to the extreme.
I finished the Pyramids in 1150 BC. In 1000 BC, I have a total of 5 cities that are all growing fast. Thanks to my early Workers I am already first in production output.

hi ,

and that is what most people do on high levels , in several PTW , PBEM games it looks that some people dont grow a lot in the first couple of turns , but then they start to make a come-back that goes further then the level of the others , .....

it really pays to have more citizens , its good for income , production, science , etc , ....

its the key to victory in many games , bigger cities produce faster units and produce other buildings faster

so in the end it really works to build those graneries

have a nice granery day

[Ribannah]

Originally posted by nbarclay
And here's a strange thought: with a capital on a river, building a granary in the capital before starting on the Pyramids there could actually speed up the Pyramids considerably. You lose 60 shields up front, but the extra production from extra laborers would more than make up for that over time.

Yes, sometimes I do that, too. It depends. In this particular game, the faster growth from the Granary would have me work unimproved tiles and adjust the luxury slider, so the advantage would be limited. This while I would have had to invest many turns to discover Pottery, instead of only 4 after my 40-turn Writing research.

As it went, I finished the Pyramids right before growing to size 7, the optimal timing.

[Dominae]

Arrian, the situation you describe in your MP games sounds like a one in which to build an early Granary. But:

Originally posted by Arrian
I never used to like granaries. But I think I may be sold on them.

My point is that I'm convinced there are times when "Granary first" is not the best option. Consider that the Americans can scout you out and Archer-rush you before you're even completed your first Settler. Thus I'm warning against people "being sold" on Granaries, since relying on this heuristic is not always optimal. Like in AU203 where early Temples were not a wise decision (a blasphemous thought...pardon the pun), so too I think Granaries are not always best (especially in MP).


Dominae

[ducki]

Just a point of interest on early Temples in AU203 - I did build some early temples - not everywhere, but places where I definitely wanted the border bump and I did fairly well in REX - not great, just well.

I'm with Arrian and nbarclay in that I really do like the early granary...

...but I'm also with you Dominae in that it's situational, just like nearly everything else in Civ3.

Sometimes it's the right call, sometimes it's a good call, sometimes it could result in early retirement.

It all depends, IMHO.

[Dominae]

ducki, that's what I've been trying to say all along! I too like early Granaries; sometimes (most times?) they're just awesome. But less-experienced players who read these forums will get the take-home message that "Granary first" is one of the golden rules of civ. This would be at the very least misleading.


Dominae

[Theseus]

On that note, should this thread be in the "Must Read" category?

[nbarclay]

Originally posted by Arrian
Nathan,

Do you really think the granary's benifit would recoup the 60 shields fast enough to catch the same city that just went straight for the Pyramids?

I ran some numbers on different scenarios, and the idea didn't work as well as I'd thought it would. If a city has no food bonuses and starts at size 1 with essentially no food saved up (e.g. right after building a settler), a granary pays off. But if food bonuses, three grasslands with shields, and forests are available, the city reaches size 7 without a granary soon enough after the granary would be finished that the granary ends up costing more time than it saves. The real killer for granaries is the high food cost growing from size 7 to size 8, so even with a granary, strictly in terms of building one wonder as quickly as possible, it pays to switch laborers to forests instead of keeping growing much if any past size 7.

Of course if you want to build the Great Library right after the Pyramids (and don't mind cranking up the luxury slider), the payoff for the granary would be considerable. Ditto if you want to use the city as a mega production center for something else and are willing to pay the cost. The Pyramids would be slowed down just a little, but the city would be bigger and able to grow faster for whatever it builds next.

Edit: Note also that a capital on a river can get much better use out of the luxury slider than most cities. So slider settings that help the capital grow near size 12 might be just about right for having cities with just roads for gold and a little corruption reach size 6. The slider setting wouldn't necessarily waste huge amounts of gold everywhere else just to help a single city be big.

Nathan

[nbarclay]

Originally posted by Dominae

Like in AU203 where early Temples were not a wise decision (a blasphemous thought...pardon the pun), so too I think Granaries are not always best (especially in MP).

Straying off topic a bit, I rarely regard temples as a high priority in the early game unless I'm playing a religious civ. There's simply too much else you can do with 60 shields - like build five or six warriors for use as MPs.

[nbarclay]

Originally posted by Dominae
ducki, that's what I've been trying to say all along! I too like early Granaries; sometimes (most times?) they're just awesome. But less-experienced players who read these forums will get the take-home message that "Granary first" is one of the golden rules of civ. This would be at the very least misleading.

Every Civ 3 player who wants to be truly good needs to learn two skills: (1) tailoring strategies to particular maps, and (2) tailoring strategies to particular opponents. I fully agree that there is no magic formula that's best in every situation.

Consider the ramifications of building a granary right off the bat if you're stuck on an island with room for maybe eight or nine cities total at most (and then only if they're spaced tightly) and you share that island with Germany. That happened in (I think it was) one of the early Apolyton tournament games, and a more, shall we say, violent opening seemed prudent.

But in most situations, especially in SP, it's hard to go too far wrong with at least one or two early granaries. That's the real message: at least one or two early granaries are probably a good idea, but check out the map you're playing on to make sure!

Nathan

[nbarclay]

Originally posted by Theseus
On that note, should this thread be in the "Must Read" category?

That, or perhaps better, a "best of" thread could be started with a name less likely to mislead people. Suppose you're a newbie and you see a thread "Why granaries don't work" in the "must read" list but don't actually bother to read it. What impression would you come away with?

Nathan