The early game build out.

I've got a question for you experts. After reading a post by Sullla in another thread, I started wondering whether it's better to build a granary first, or a settler first. I'm sure the answer depends on the strength of the tiles in your first city, but I have a difficult time evaluating that because I don't really understand the impact of the granary, shields, etc.

What I'd like to know is how you folks determine which is the better move, and also, how many food does it take to grow from pop 1-2, 2-3, 3-4, 4-5, 5-6, with and without a granary. Ideally, of course, food and shields would balance and come out producing enough pop and shields to produce a new settler every 5 turns or so.

The 2nd question I'd have is how large is the best size to grow your first city to to start popping settlers? 3, 4, 5, 6?

I think Cracker needs to do an article on this. How to evaluate the food/shield balance and maximize settler production.

I dunno. 8P

I've posted a few times over the last year (I'm only a casual player, spending most of my time on Ultima Online). When I mention that I produce a settler at size 5, everybody jumps in and says I'm waiting too long and I should produce one at size 3. When I say I'm producing a settler at size 3, I'm told that's not efficient and should wait until I'm size 5.

That's my bad ripcord. I was the one saying that most people don't grow to size 5 before popping settlers. Then again, at least I didn't claim to be an expert. It was your thread (and one other thread) that really started me thinking about this. I am sure there is a fairly easy way to determine whether it's better to build granaries, early settlers, etc, just based on the shield/food balance in your first city. That's why I started this thread.

An interesting thing about civ3 is that it takes the same number of excess food for a town (e.g. size 1-6) to grow. So 1-2-3-4-5-6 all cost the same amount of food (20 I think), the amount is halved for a town with a granary (10). This greatly influences the "best" size for your city. Each citizen needs two food, many spaces produce only two food (then you get two for your city so you have two extra for growth). Thus, if you have any bonus food spots you should grow to work them, that will increase your extra food per turn and since the amount needed to grow didn't change you will now grow faster. More population should also mean more shields. Look at it this way: with no bonus food and no granary it will take 10 turns to grow and so 20 turns min between settlers; with a granary 5 turns to grow and 10 turns min between settlers. Even if it takes you 20 turns to build your granary you will catch up in total settler production on just your second settler.

I almost always build a granary first. The only reason I wouldn't is if there is a nearby spot that has alot of bonus food (this includes flood plains) and my starting spot doesn't. Then I would build a settler ASAP (e.g. hopefully just as I reach pop 3), send him over to the food and build my first granary there. I will even whip my granary (I almost never use the whip) to get it out faster and just push luxuries for a while.

The tiny map will modify this a little, if you are only going to get 4 cities anyway then it may not be worth building a granary.

I usually build a setter so that it completes when I have a population of 3. If I wait longer then I have to worry about happiness.
I suggest if you see an area nearby outside your city radius with wheat or cows create a settler quickly. This new city will often have a quicker growth rate than your capital
Usually there are only a couple of very productive tiles in a city radius and in the very early I think it is better to have several small cities milking the really productive tiles than one big one. Several small cities also allow you to get those resources outside the capital city.

My take: if barbs are expected to be active build one settler before the granary. Have the second city produce military units for defense. If barb activity is low (Monarch difficulty or below or sedentary setting) a player can go ahead and make a granary before starting on settlers.

Again, try to time the completion of the granary exactly one turn before the population increases. On Emperor and above, use the luxury slider to keep the citizens happy. Better to go out exploring than saving one gold per turn with the garrison. Maybe have the early warriors circle back and garrison after exploring the perimeter out to ten or fifteen tiles.

Some players like to build an early Temple to boost culture and allow one more happy citizen. This is more useful for peaceful players.
- Bill

I play on Diety and set barbs to Raging. Only once or twice in many games have I had a problem due to building a granary first. I don't think I have ever lost my initial city to barbs (maybe I did and am using selective amnisia here). I do build one military unit first, although I use him to explore I try to keep him close to home. Then I try to entice the barbs to attack him in a forest or on a hill or mountain to upgrade him. Once I have a granary my settler comes pretty quick and I do try to use my second city as a military factory (a couple more warriors and then a barracks).

Also I don't see how building a settler first would help this, they don't do well against barbarians

It all depend on the situation.

The best you can hope for is a situation where you can produce 5 extra food per turn and have a granary. Assuming that there is enough shield production you can pop a settler every 4 turns. You need either a haystack on a flood plain or two bonus squares (haystack, cattle, game, floodplain) to have a shot at this.

The next best situation is producing 4 extra food with granary then you can only produce a settler every 6 turns.

The next best situation is producing 5 extra food then you can only produce a settler every 8 turns.
The next best situation is producing 3 extra food with granary then you can only produce a settler every 8 turns.

The next best situation is producing 4 extra food then you can only produce a settler every 10 turns.
The next best situation is producing 2 extra food with granary then you can only produce a settler every 10 turns.

The next best situation is producing 3 extra food then you can only produce a settler every 14 turns.

The next best situation is producing 2 extra food then you can only produce a settler every 20 turns.


Of course you can add workers to cities to increase the rate of growth of a city, having a city producing workers every 10 turns can turn help a city with only 2 extra food produce a settler every 10 turns. This method makes up for poor starting positions where none of the better solutions are possible. Faster settler production can really change your growth patterns. If you do not have bonus food production, cities helping cities with workers is more productive than building a granary cause you have increased commercial productivity.

In short, I build granaries for settler factories that have 4 extra food available or more. If the city has only three extra food you may be better using a feeder city to speed it's growth rather than invest in a granary.

@Cartouche Bee
In the best case examples you posted, in order for a town to get a settler in 4 turns, it would need to also produce 7 shields per turn for 2 turns, 8 shields per turn for 2 turns, unless you pop rush. That indicates to me that you really need to have a city going from pop 4 to pop 6 (ancient world, no aqueducts or river, of course). To find a city that can produce 7 shields and 5 extra food from only 4 tiles would be a miracle.

In any case, it's really starting to sound like there are very few cases where it isn't better to build the granary first, since food is normally the limiting factor in popping settlers.

@billchin
Why do you need to time the granary to finish 1 turn before growth?

Originally posted by billindenver
@Cartouche Bee
In the best case examples you posted, in order for a town to get a settler in 4 turns, it would need to also produce 7 shields per turn for 2 turns, 8 shields per turn for 2 turns, unless you pop rush. That indicates to me that you really need to have a city going from pop 4 to pop 6 (ancient world, no aqueducts or river, of course). To find a city that can produce 7 shields and 5 extra food from only 4 tiles would be a miracle.

In any case, it's really starting to sound like there are very few cases where it isn't better to build the granary first, since food is normally the limiting factor in popping settlers.

@billchin
Why do you need to time the granary to finish 1 turn before growth?

Click to expand...

It's self balancing, and yes normally the city will fluctuate between 4 and 6, but I have had a city or two that was so well endowned that it fluctuated between 3 and 5. Pop rushing might be good in some cases but I rarely ever pop rush, over time a city that has little corruption will produce more if you do not pop rush. In many situations adding workers to cities is more productive than granaries but I'll let you make your own decision on that.

Personally, I feel settlers need to be produced as fast as possible. Other ai civs even non-expansionist ones expand rapidly by building only settlers and warriors it seems. You need to maximize your territory in the beggining to have a good chance at winning. when you have about 3-4 cities then you build a granary in the capital. If you think about efficiency then once your capital becomes too far from the frontier usually about 8 cities you begin work on other things, ie workers and other improvements. Keep in mind that as you expand you need to build defense.

It does seem that AI's don't take full advantage of granaries (or adding workers in some cases as Cartouche Bee astutely points out). That is why it is possible to keep up in expansion even with the Diety bonus working for them.

Originally posted by Cartouche Bee


It's self balancing, and yes normally the city will fluctuate between 4 and 6, but I have had a city or two that was so well endowned that it fluctuated between 3 and 5. Pop rushing might be good in some cases but I rarely ever pop rush, over time a city that has little corruption will produce more if you do not pop rush. In many situations adding workers to cities is more productive than granaries but I'll let you make your own decision on that.

Click to expand...

I'd like to see that city, especially on my next random map.

If I had to guess, I would say in most of my games, my four most productive squares would produce on average 3 extra food, and 5 or 6 shields. My usual build-out has been warrior(explore) warrior(explore) settler, warrior (police), settler or temple. I guess I should try out a warrior(explore), granary approach. I suppose I could even emphasize shield production early on as well as forestry to get the granary out in a reasonable time.

BTW, I play Monarch level, with sedentary or raging barbarians, and they have almost never attacked me for 50 turns or more, unless they pop from a goody hut right next to my capital.

Oh yeah, I wanted to point out that pop rushing the granary works out fine for me (as I say I almost never pop rush), I only do one pop and so its only 20 turns. If I have one luxury, no problem at all, even if I don't its only a matter of MMing between 10 and 30% luxury as your city grows. Not so bad considering that you get more cities faster. Thus quickly erasing any deficit.

In the early game I never have an entertainer.

Originally posted by Gothmog
Oh yeah, I wanted to point out that pop rushing the granary works out fine for me (as I say I almost never pop rush), I only do one pop and so its only 20 turns. If I have one luxury, no problem at all, even if I don't its only a matter of MMing between 10 and 30% luxury as your city grows. Not so bad considering that you get more cities faster. Thus quickly erasing any deficit.

In the early game I never have an entertainer.

Click to expand...

What are the limits of pop-rushing? I.e., how many people convert to how many shields? Pop-rushing a granary makes more sense to me than most other pop-rushes, except for the occasional defender.

Someone did a thread on this long ago, comparing straight settler production, to rushed settlers, to rushed granary. Thought it was interesting.

http://forums.civfanatics.com/showthread.php?s=&threadid=7764

He got the fastest growth through 2000BC by rushing his granary then building settlers.

Bill back then a pop rush yeilded 40 shields and now only 20 shields, but rushing a granary does have a quicker up front settler production and is the way to go if you have 4 or more extra food in the city and you time it with city growth to preserve food at the time of the granary production.

You have to be careful with those old posts cause the rules changed alot between the initial release and now.

As usual, the maximum speed of expansion will be highly dependent upon the specific tiles in and around your first city. With that said though, I would venture to say that in the great majority of cases it would be beneficial to build a granary before any settlers and keep your capital city at as large as possible in producing settler. Cartouche Bee already outlined the ideal situation above, with settlers produced every 4 turns. That's kind of a rare situation, but I managed to run it in a private game of mine where I had 2 grassland wheats in range of the starting city (two turns at size 5 producing 7 shields/turn, then two turns at size 6 producing 8 shields/turn, grow to size 7 and go fall back to size 5). While that's a rare case, you can see from the excellent info Cartouche Bee posted above how much faster settlers can be produced with granaries as opposed to without them. It's like making an initial investment of 60 shields in order to get a higher return on every other settler produced for the remainder of the game from that city. Obviously it's not worth it on tiny maps or on an island where you will never have more than one or two cities, but in most cases it's well worth it.

The real reason why granaries are so effective is that they half the food necessary to grow (duh) and food, NOT shields, are the limiting factor in expansion in almost all cases. Unless you start surrounded by floodplains or something like that, you will almost certainly be able to produce settlers faster than your city can grow to a size large enough to pop them out. Without a granary it will take 20 turns to grow from size 1 to 3, assuming no bonus food tiles. Barring a really weird starting spot, you should be able to get 2 shields/turn at size 1 and 3 shields/turn at size 2. So it's clear that production outstrips food resources in a normal starting position. Once you build a granary, it only takes 10 turns to grow from size 1 to size 3, over which span you would produce approximately 25-30 shields. In short, a granary allows food to keep up with production in terms of settler production. Since settlers cost 30 shields each and a granary is only 60 shields, once you've built two settlers post-granary, you have essentially gotten your worth from it and anything else is a bonus. As for the issue of what size to run settler production (1/3, 2/4, 3/5, etc.), that is too highly dependent on the individual situation to make many comments. Generally speaking though, the more bonus food tiles you have, the higher the size you should run your settler production.

And for those of you who still don't believe in the value of the early granary... Well, when PTW comes out you can see if you grab more land than me or not.

Barring a really weird starting spot, you should be able to get 2 shields/turn at size 1 and 3 shields/turn at size 2

Click to expand...

And in some starts you can get 3 shields at size 1, and 5 at size 2. If you have game on forests, you'll get 2 food/2 shields. Then if you mine a bonus grassland square, you'll get that granary finished 4 turns before you hit size 3 and get a settler out only 2 turns later than if you hadn't built the granary. And after that first settler, the next settlers out of that city will come so much faster. Even if you don't have game on forests, if you mine 2 bonus grassland right away you'll get that granary built pretty fast (helps alot if you are industrious!) Sometimes I build 1 settler, then a granary, so that I have a chance to build some warriors for garrison/exploring, and have time for my worker to get the terrain improved before starting on that granary.

I guess I have neglected granaries for the most part, but that is because of my particular play style and map conditions. Because I need to pump out all those scouts (ALOT of scouts) to get techs, rather than build a granary. And I use ICS, so my settlers build a new city, either the turn they were created, or the very next turn, and these new cities are working on settlers, also. Because of this, I can't sit around waiting for my first city to get to size 5 or whatever, and I'm not going to build a granary in every city.

If you aren't going to do an ICS 'settler flood' a granary is a good investment in most cases.

As to why the granary one turn before pop increase? The granary takes effect after the next pop increase. Having several turns idle time burns gold for no benefit. A player often has a choice of a food tile or a forest tile to increase growth or shields. By timing it exactly at one turn the granary provides max benefit. Pop rushing is an option, especially if the capital is in a shield poor location such as flood plains.

As for building one settler before the granary, I suggest trying the AU 107 game, standard rules, Emperor difficulty.
http://apolyton.net/forums/showthread.php?s=&threadid=63643

With this set up, everytime I tried one warrior then the granary, barbs come calling before the granary is completed. If they manage to defeat my one warrior or he is out exploring the game is lost.

By building a settler first, I have more gold for research (no luxury slider needed). A settler is 30 shields, so I can build the second city and have that second city build two or three warriors in the same that the 60 shield granary is ready.

In my experience, Barbs are often a problem for me on Emperor or Deity, standard size maps, version 1.29. I don't know if it is average starting locations (no cows), bad luck, map size or what, but going for warrior, granary has often meant barbs (often several of them) approaching the capital to ruin the game before it even starts.

Granaries are also great on any river city, allowing them to get to pop 12 very quickly.
- Bill