How to keep up with diety's settler expansion

the articles on this site speak of making granaries to get settler factories. this approach doesn't seem to work very well at all especially with the patch and costlier pottery... either you must be an expansionist civ for pottery or you get pottery late and are even more behind. either way, you need to have a great starting location and even if you do, by the time you start making settlers you have already lost the race.

what i find works much better is just making settlers asap from each city. by the time you get your granary all set up you can make a settler every 4 turns if you have a lucky spot. this is not advantageous since in the same time period you could have set up at least 5 or more cities. 1 settler coming out every 14-20 turns from just 5 cities is better than one settler factory.

if you wait until 2500 bc or later to make settlers the diety usually has taken lots of your good spots anyways.

the fastest way to keep up pace (or close to it against the luckier enemies) is to have your starting worker join your city and make a settler right away. your scouting is slower (no warriors initially) but this strategy works best with compact city placement since you only need a minimum of 2 grassland (1 bonus) to spam settlers (warrior, settler OR warrior, warrior, settler = settler in 20). the strategy works fine with spacier city placement but you might have to guess a bit with your settlers.

if you have a starting high in food it's usually best to keep the worker since you can irrigate and stuff, but if you just have grassland or something, irrigation does nothing so the worker is useless since we focus on cities, not production.

this strategy is just for the most recent patched civ III without conquests or play the world or anything. also, if you are looking for some magical early military rush this strategy isn't best since you're focusing on cities, not shields. in the long run, more cities = more shields though. also, this is for sedentary barbarians, possibly roaming. not higher since you have nothing defending cities early on.

Welcome to CFC!

I haven't played vanilla civ in a long time, only C3Conquests, but I think these comments still apply...

One thing, it's virtually impossible to keep up with the early Deity AI expansion regardless of start position or strategy. Deity AI gets extra settlers, units, and huge bonuses for production and pop growth. They WILL out-expand you, so it's just a matter of making the best of it.

Settler factories are crucial for almost any style of play. In some situations, such as an extremely poor food map or a tiny map with nearby enemies, early barracks and units can be the better way to expand. But those times are much more rare.

Usually, you will have a fair bit of land to fill in--provided you are doing a tight city placement spread--even against Deity AI. In those cases, it MIGHT be true that you will get more early cities by not building a granary first, but just building settlers.

However, there's a catch. While you might get a couple of extra early settlers by not building a granary, you will GREATLY lose out on the overall growth curve in the long run. Your granary cities will be able to replace pop much quicker than other cities, meaning they will have higher production (more pop working more often) and they are way more efficient in pumping settlers. Having non-granary cities, especially relatively food-poor ones, pump out settlers early on means those cities stay small for a long time and lose a huge amount of production. Not good for your empire, I'm afraid.

Trying to build settlers first will quickly be outpaced by building the granaries first, and then building settlers. You start slower, but turn on the gas pretty soon and blow by. It's generally not even a contest unless you have some HUGE natural food bonuses such as a clump of flood plain wheat tiles.

Generally, most civs get pottery very early and you can trade for it, provided you have scouts out getting contacts . Otherwise, research it. Also, speed along the building of your granaries by chopping forests in the near vicinity. Another rule of thumb is to usually only build 1-2 granaries unless you have an amazing amount of food bonuses. These 1-2 settler (and worker) factories will be sufficient to efficiently fill in most starting areas unless the map is large to huge size.

Lastly, your early worker is extremely valuable. You almost always will benefit more by using it to improve the land rather than joining it. True, irrigating grassland doesn't work in despotism, but mining the grass does, and helps to build the granaries and settlers to come.

One thing, it's virtually impossible to keep up with the early Deity AI expansion regardless of start position or strategy. Deity AI gets extra settlers, units, and huge bonuses for production and pop growth. They WILL out-expand you, so it's just a matter of making the best of it.

Settler factories are crucial for almost any style of play. In some situations, such as an extremely poor food map or a tiny map with nearby enemies, early barracks and units can be the better way to expand. But those times are much more rare.

the only reason you never catch up to diety's expansion is because you insist on settler factories. you are correct in stating that with that approach it is impossible to catch up. i've tried it and have never met that goal with it. without that approach you have a good chance at keeping pace with the number of AI cities. i've tried it without the granary and usually always keep pace with the AI.

if you really need early military, then it works to keep the worker and mine. otherwise shields are useless for your goal of maximizing cities. if you have nothing irrigatable, then worker joining city is best for this purpose.

fpt = food per turn
wg = with granary, growth needed for a settler (2) per turn
wng = with no granary, growth needed for a settler per turn

fpt | wg | wng
5 | .25 | .13
4 | .17 | .1
3 | .13 | .07
2 | .1 | .05

5fpt capital with granary = .25 settlers per turn
5fpt capital + 3*(2fpt city) = .28 settlers per turn
5fpt capital + 4*(2fpt city) = .31 settlers per turn

2fpt capital with granary = .1 settlers per turn
2fpt capital + 2*(2fpt city) = .15 settlers per turn
2fpt capital + 3*(2fpt city) = .2 settlers per turn

it seems almost always better to make settlers instead of granaries. chances are you'll make more cities than the examples listed here by the time you get a granary giving you even more settlers per turn. it's also likely that your non-capital cities can get more fpt than your capital.

things that don't go into calculations:
with settlers coming out of non-capital cities, they reach their destination faster.
non-settler production from all your cities.

i may be new here but from playing the game and doing the math it seems evident that the granary approach doesn't work.

You may want to look at this

UnnamedPlayer said:

the only reason you never catch up to diety's expansion is because you insist on settler factories. you are correct in stating that with that approach it is impossible to catch up. i've tried it and have never met that goal with it. without that approach you have a good chance at keeping pace with the number of AI cities. i've tried it without the granary and usually always keep pace with the AI.

Click to expand...

What do you consider keeping pace? The Deity AI starts with extra settlers, so you start out behind regardless.

if you really need early military, then it works to keep the worker and mine. otherwise shields are useless for your goal of maximizing cities. if you have nothing irrigatable, then worker joining city is best for this purpose.

Click to expand...

Settlers take shields, too. I've had many cases where my settler output was not limited by how fast my cities grew, but rather by their lack of shield production.

fpt = food per turn
wg = with granary, growth needed for a settler (2) per turn
wng = with no granary, growth needed for a settler per turn

fpt | wg | wng
5 | .25 | .13
4 | .17 | .1
3 | .13 | .07
2 | .1 | .05

5fpt capital with granary = .25 settlers per turn
5fpt capital + 3*(2fpt city) = .28 settlers per turn
5fpt capital + 4*(2fpt city) = .31 settlers per turn

2fpt capital with granary = .1 settlers per turn
2fpt capital + 2*(2fpt city) = .15 settlers per turn
2fpt capital + 3*(2fpt city) = .2 settlers per turn

Click to expand...

I'm not totally sure where you're getting these numbers. I understand that a 5fpt city with granary will get 25% per turn of the 2 pop needed for a settler, but that again totally ignores the shields needed.

it seems almost always better to make settlers instead of granaries. chances are you'll make more cities than the examples listed here by the time you get a granary giving you even more settlers per turn. it's also likely that your non-capital cities can get more fpt than your capital.

Click to expand...

Often, the capital is on some of the best terrain and even if it's not, it doesn't suffer corruption, meaning it can get the 30 shields needed for the settler faster.

things that don't go into calculations:
with settlers coming out of non-capital cities, they reach their destination faster.
non-settler production from all your cities.

Click to expand...

True on #1, reaching their destination faster. One drawback to the settler factory. #2 is one of the big reasons why a factory works. Granary cities replace pop faster, so they keep more citizens working = more production. Non-granary cities can't replace pop as fast, leaving them poorer on production as well.

i may be new here but from playing the game and doing the math it seems evident that the granary approach doesn't work.

Click to expand...

I can't say I'm convinced, sorry. I'd be very interested in seeing a save game or some screenshots where you put your technique into practice.

I agree with Civverguy that you should check out Bamspeedy's article.

BTW, why does it take you so long to build a granary in the capital? Using forest chops if available, it's generally quite fast, unless you joined the worker to the city rather than mine some tiles...

I usually only build a couple of early granaries at most. More than that IS a waste.

Lastly, you are free to play any way you feel. Please don't take my reply as an attempt to force you otherwise. I'm just interested in figuring out which approach to city growth is more efficient...

EDIT: One other thing: Again, I haven't played vanilla civ in a LONG while, so if my comments don't apply to vanilla because of game mechanic changes, my sincere apologies.

by keeping pace i mean catching up in # of cities by a few thousand years into the game. the AI will get 5 or 10 cities up faster than you do, but you can beat some civs to 15 and 20.

shields are only a concern in cities that grow very fast (flood plains and granaries being the only real problems). almost all your cities will be grassland or plains. there are very few instances where a settler will not be made by the time you grow 2. the few cases are usually unmined wheat grassland (in which case you still get the settler out in less than 20 turns) or very corrupt cities (which will be a problem no matter which start you choose).

all you need is one grassland and one bonus grassland in the city's radius out of 8 workable tiles. the probability of finding these locations is very very high. 20 turns of 2 shields/turn is more than enough for a settler.

in farther away cities you might need 2 bonus grassland out of 8 tiles. this is also very common and very easy to find. 10 turns of 1 spt and 10 turns of 2spt is enough for a settler. (1 spt goes to corruption)

this happens without working the tiles at all. if you keep a worker you can get more shields and have enough for making things other than settlers.

most of my cities end up having 2 bonus grassland tiles in the radius. i make 2 warriors and a settler in 20 turns. with a corrupt city, it becomes 1 warrior and a settler in 20 turns. with more corruption, i make mines (by this distance, i've usually run out of space and can't make more cities anyways).

i've read bamspeedy's article. his start was rare since he had a great capital and little grassland anywhere else. also with the patch there is no way to get granary that fast.

also, the starts for islands or continents is probably different. with pangaea you will almost always find enough grassland or plains and get adequate cities.

i'll take some screens comparing the 3 starts (granary, no granary, no granary worker join).

the numbers in the "table" i made are just how much of a double food box the city fills each turn. (40 without granary, 20 with).

there are articles like the babylon diety one and ones like this, but nothing about the fastest way to expand with a normal starting location.

Key thing is you need to balance the sheilds and the food. If you can build as fast as you grow with a granery then you should build one (8+spt with 5fpt, or 5+spt with 4 fpt). The sheilds you spend on the granery will usually be while you are letting the city grow anyway. This can be further improved with a forrest chop or two (this might not be worth it on vanilla as the chops take 10 turns rather than the 4 turns on C3C).

As for stats - I can do better than that

http://forums.civfanatics.com/showthread.php?t=185250

gives some spreadsheets for the beginning of a team game, Paper Beetle can analyse all sorts of starts, This one was unusual in that the cow couldn't be watered, so food was at a premium.

UnnamedPlayer isn't telling us anything new, and what he is saying isn't incorrect. In fact, it is true for all difficulty levels, not just deity.

I've advised someone to not build granaries in this thread here:
http://forums.civfanatics.com/showthread.php?t=226435
Post#10, and I explained it more in post#30

But I've learned it myself from forum-member WackenOpenAir, here is an example where he explains it:
http://forums.civfanatics.com/showthread.php?t=167249&highlight=granary+5fpt
He explains it in Post#8 in that thread.
And this thread:
http://forums.civfanatics.com/showthread.php?t=212031&highlight=granary+food+surplus
Post#6

The idea is that you should always seek to grow your civ-wide food surplus. Building a settler first compared to building a granary first depends on what the food surplus your future city will be compared to the food value of the granary, which is equal to the food surplus of the current city.


As you can see, assuming the future city site has a +2 food surplus, a granary is only a better choice if the current city has a +5 food surplus. And is almost equal when the current city has a +4 food surplus.
Then there are special situations, where the AI starts very close to you, and you'll be landlocked before you can build a granary...

Of course, after the initial growth spurt is over, and you start with specializing your cities (some for units, some for workers, some for settlers) then you build a granary in your settler pump, even if its food surplus is less than +5fpt.

The idea is that you should always seek to grow your civ-wide food surplus. Building a settler first compared to building a granary first depends on what the food surplus your future city will be compared to the food value of the granary, which is equal to the food surplus of the current city.

i don't understand your chart very well. i think you assume you have to make a choice between a settler and a granary. in reality, i find that the choice is different. future city highlighted above should be future cities.

the more food your capital has, the more advantageous a granary is. that means the choice is probably between 3-4 settlers and a granary in a city with +4/5 fpt.

with +3 fpt, the choice is between usually 2-3 settlers and a granary.

with +2 fpt, the choice is between 1-2 settlers and a granary.

all this depends on when you get pottery and whether you have pre-built something.

you said a granary is better if your capital can get +5 fpt and the 2nd city is only +2fpt, or about even if the capital gets +4fpt. in these two situations, i think you need to take into account at least a 2nd, 3rd, and 4th city vs. the capital with a granary.

the analysis in one of your links of empire wide fpt is good, but changes when granaries come into play. it's better to see things in empire wide growth per turn.

fpt = food in growth box per turn
gpt = growth per turn without granary
ggpt = growth per turn with granary

fpt | gpt | ggpt
2 | .1 | .2
3 | .14 | .25
4 | .2 | .33
5 | .25 | .5

i guess it depends on the situation. the best case scenario for a granary is having a 5 fpt city and 2 fpt in everything surrounding and being america (industrious + free pottery = fastest granary) with forests to clear.

however, according to the math, i think it's better to make settlers instead of granaries probably more than 90% of the time if we only count making cities as the goal.

in the best scenario in favor a granary, the choice is probably between:

+5fpt capital with granary = .5 gpt
+5pt capital no granary + 2fpt city + 2fpt city = .45 gpt

This is a very interesting thread.

@ MAS: I didn't want to start a new thread, but I am wondering how these concepts you, Wacken and UP speak of affect the slinger. I hit the Republic slingshot almost every time on Monarch level, yet I've not been able to hit it at all on Emperor level. I believe the lux slider is killing me with only one content face. Under your rules it seems that I'd be jacking up the slider even further since I would not have the MP support and I would not be using specialists, since food surplus is more important.

Care to offer up your take?

By constantly building settlers and workers you cities remain small, so there should be few unhappy people per city.

Its just like building a city in SimCity4, first build outward, then upward!

In my personal experience, I find it more difficult to do the republic slingshot on huge maps vs standard maps, than monarch vs emperor!

I don't even try to keep up with the Deity AI settling. I figure that once we both get cities that are more than the OCN, all they are doing is creating future cities for me.

I do create granaries, both to get settlers until I run out of space and then to create workers.... lots and lots of workers. I usually only create them in a couple of cities, though. the 60 shield investment is a lot, so if you can't get enough shields to build settlers/workers in half the time, why cut the growth time in 1/2?

On deity, though, you need a couple of barracks cities, too.

this strategy isn't for quick warmongering. for that you need to keep your worker and focus on production. however, it is good for longer-term war after you're done expanding because you'll end up with more cities. it also allows for victory without going to war.

i don't exactly know how it works, but i think making more cities than the OCN is still advantageous

All of a sudden I'm in the mood to play SC4.

I don't quite understand this. Shouldn't we simply compare how long it would take to make up the loss in food and shields (maybe commerce, though this is extremely difficult to calculate)?

Let's see: What should we assume?
No food bonuses? And 2 mined BG? And a size 2.5 start. And decent improvements so that all future terrain used gives 2 fpt and 1 spt.
Build order= Granary, Settler, Archer, Archer, Settler, Archer, Settler, Archer, Settler
OR
Build order = Settler, Archer, Archer, Settler, Archer, Archer, Settler (Basic difference is no granary 1 less settler and 1 more archer. Earlier settling may help too.)

We should also assume 20 turns until a new settler is spawned from all settled towns

Turn 0 Granary started
Size 2.5, 5 spt, and 0 shields
Turn 5
Size 3.0, 6 spt, and 26 shields
Turn 10
Size 3.5, 6 spt, and 56 shields
Turn 11 Granary Completes (upkeep starts from now). Settler started
Size 3.6, 6 spt, and 0 shields
Turn 15
Size 4.0, 7 spt, and 25 shields
Turn 16 Settler Completes. Archer started
Size 2.6, 5 spt, and 0 shields
Turn 20 Archer Completes. Archer started
Size 3.5, 6 spt, and 0 shields
Turn 24 Archer Completes. Settler started (actually an spt wasting strategy)
Size 3.9, 6 spt, and 0 shields
Turn 25
Size 4.5, 7 spt, and 7 shields
Turn 29 Settler Completes. Starts Archer
Size 2.9, 5 spt, and 0 shields
Turn 33 Archer Completes. Start Settler
Size 3.8, 6 spt, and 0 shields
Turn 35
Size 4.5, 7 spt, and 13 shields
Turn 38 Settler Completes. Start Archer
Size 2.8, 5 spt, and 0 shields
Turn 40
Size 3.5, 6 spt, and 11 shields
Turn 42 Archer Completes. Start Settler
Size 3.7, 6 spt, and 0 shields
Turn 45
Size 4.5, 7 spt, and 19 shields
Turn 47 Settler Completes
Size 2.7, 5 spt, and 0 shields

OR

Turn 0 Settler started
Size 2.5, 5 spt, and 0 shields
Turn 5
Size 3.0, 6 spt, and 26 shields
Turn 6 Settler Completes. Start Archer
Size 1.1, 3 spt, and 0 shields
Turn 10
Size 1.5, 3 spt, and 12 shields
Turn 15
Size 2.0, 5 spt, and 29 shields
Turn 16 Archer Completes. Start Archer
Size 2.1, 5 spt, and 0 shields
Turn 20 Archer Completes. Start Settler
Size 2.5, 5 spt, and 0 shields
(as you can see, this is a 20 turn settler and 2 archers cycle)
Turn 40 2 more archers and 1 more settler have been completed. Start Settler
Size 2.5, 5 spt, and 0 shields
Turn 46 Settler Completes.
Size 1.1

So in approximately the same time, you can get:
1 granary
4 settlers
3 archers
Average size 3.
OR
3 settlers
4 archers
Average size 2.

*There are better strategies for doing the no granary version, which makes more archers and maintains a higher on average citizen count.

Advantages of granary
In the long run, more production
Allows for extra settler building
Overall more commerce
Flexible; can make workers or extra units as needed very quickly.
Good for large maps and pangaea
Benefits from reduced corruption

Advantages of no granary
No maintenance fees
Allows for extra archer building
Earlier settlement and claim to other territory
Does not require (much or any) MP and no lux tax
Good for islands or small maps
Power advantage=less demands

Conclusion: Choose the granary if you're not planning to go to war any time soon and the nearby territory looks dull. Don't build a granary if you're on a high level (less free content faces) or see good territory nearby that an AI might nab.

Of course, that was the boring, grasslands start. Plains would pretty much be the same.

Settler Built (+granary)
Turn 16
Turn 29
Turn 36 (bonus settler)
Turn 38
Turn 47
Turn 49 (bonus settler)
Settler Built (-granary)
Turn 6
Turn 26
Turn 26 (bonus settler)
Turn 46
Turn 46 (bonus settler)
Turn 46 (bonus settler)
Turn 46 (bonus settler)

Don't feel bad for the lack of granary empire. The 10 extra turns from the 1st settler might be the tipping point since that's 10 turns in advance forever for all future towns it makes too. Wow, in this case, the granary won't be helping for a while, huh?

Well, I haven't finished. Posting for now.

Let's try a flood plainsy start. I think these would give an advantage to those who don't want granaries.

City has as many flood plains as desired, and 2 plains, 2 hills, 2 deserts. At least 1 FP worked for the whole time. (FP irrigated, plains and deserts mined).

With granary: Granary->Settler->Settler->Settler...->Settler
Without granary: Settler->Settler->...->Settler

For about 45-50 turns like before.

Turn 0 Granary started
Size 2.5 4 fpt, 1 spt, 0 shields
Turn 3
Size 3.0 5 fpt, 1 spt, 5 shields
Turn 7
Size 4.0 2 fpt, 5 spt, 12 shields
Turn 16 (Delay growth by working a hill instead of FP)
Size 4.9 1 fpt, 6 spt, 57 shields
Turn 17 Granary Completes. Start Settler
Size 4.95 2 fpt, 5 spt, 0 shields
Turn 18
Size 5.5 3 fpt, 5 spt, 5 shields
Size 5.5 0 fpt, 7 spt, 5 shields
Turn 22 Settler completes. Settler started
Size 3.5 3 fpt, 3 spt, 0 shields.
Turn 24
Size 3.8 3 fpt, 3 spt, 6 shields
Size 3.8 4 fpt, 1 spt, 6 shields
Turn 25
Size 4.5 3 fpt, 3 spt, 9 shields
Size 4.5 2 fpt, 5 spt, 9 shields
Turn 30 Settler completes. Settler started
Size 3.5 3 fpt, 3 spt, 0 shields
(another loop. 8 turn settler cycle. How slow!)
Turn 38 Settler->Settler
Turn 46 Settler completes.

OR

Turn 0 Settler started
Size 2.5 4 fpt, 1 spt, 0 shields
Turn 3
Size 3.0 3 fpt, 3 spt, 5 shields
Turn 10
Size 4.0 2 fpt, 5 spt, 28 shields
Size 4.0 4 fpt, 3 spt, 28 shields
Turn 11 Settler->Settler
Size 2.2 2 fpt, 3 spt, 0 shields
Size 2.2 4 fpt, 1 spt, 0 shields
Turn 15
Size 3.0 3 fpt, 3 spt, 6 shields
Turn 22
Size 4.0 2 fpt, 5 spt, 29 shields (such a shame)
Size 4.0 4 fpt, 3 spt, 29 shields (well, it makes a cycle)
Turn 23 Settler completes
(12 turn settler factory)
Turn 35 Settler completes
Turn 47 Settler completes

You could build:
1 granary
4 settlers
Avg. Size 4

Or you could build
4 settlers
Avg. Size 3

Same advantages as before except for the military thing. There's no time for that when you lack shields.

Settler Built (+granary)
Turn 22
Turn 30
Turn 38
Turn 42 (bonus settler)
Turn 46
Settler Built (-granary)
Turn 11
Turn 23
Turn 31 (bonus settler)
Turn 35
Turn 43 (bonus settler)
Turn 47

More advantages for the no granary team. The later towns get an early bonus as well (number 2 settler has 7 turns. This early advantage may make up for the inability to build MP. Also, it takes longer to recover from the time it took to build the granary since this time, you get an extra settler (but lose out on the bonus settlers).

Actually, MAS, I think you may be getting a convert. I may once again play for the no granary team.

(Newbies=no granary. Players=granary. Advanced=no granary?! What an odd thing.)

The next case is for a grassy cattle start. Wheat should be very similar.

Basically for this one, it's an odd cycle.

Granary(11)->Settler(15)->Settler(20)->Settler(25)->Settler(30)->Settler(35)->Archer(39)->Settler(44)
in 44 turns
Avg. Size 4

OR

Settler(6)->Archer->Settler(17)->Archer->Settler(27)->Archer->Settler(37)->Archer->Settler(47)
in 47 turns
Avg. Size 2.5

Settler Built (+granary)
Turn 15
Turn 20
Turn 25
Turn 30
Turn 35
Turn 35 (bonus settler)
Turn 40 (bonus settler)
Turn 44
Turn 45 (bonus settler)
Settler Built (-granary)
Turn 6
Turn 17
Turn 26 (bonus settler)
Turn 27
Turn 37
Turn 37 (bonus settler)
Turn 46 (bonus settler)
Turn 46 (bonus settler)
Turn 47
Turn 47 (bonus settler)

Guess what? The no granary team is still doing better (expansion-wise and military-wise)! That's really a shame, considering how I'd have played it before.

Then for a lake fish start.

A consistent 3 fpt situation. This also gets 2 mined BG.

Granary->Settler->Settler->Archer->Settler->Settler->Archer
Settler->Archer->Settler->Archer->Settler->Archer

Turn 0 Granary started
Size 2.5, 3 spt, 0 shields
Turn 4
Size 3.0, 5 spt, 14 shields
Turn 11
Size 4.0, 6 spt, 50 shields
Turn 13 Granary completes. Settler started
Size 4.3, 6 spt, 0 shields
Turn 18 Settler->Settler
Size 3.5, 5 spt, 0 shields
Turn 22
Size 4.5, 6 spt, 21 shields
Turn 24 Settler->Archer
Size 2.8, 3 spt, 0 shields
Turn 26
Size 3.5, 5 spt, 8 shields
Turn 29 Archer->Settler
Size 3.95, 5 spt, 0 shields
Turn 30
Size 4.5, 6 spt, 6 shields
Turn 34 Settler->Settler
Size 3.5, 5 spt, 0 shields
Turn 38
Size 4.5, 6 spt, 21 shields
Turn 40 Settler->Archer
Size 2.8, 3 spt, 0 shields
Turn 45 Archer Completes
Size 3.95, 5 spt, 0 shields

OR

Turn 0 Settler started
Size 2.5, 3 spt, 0 shields
Turn 4
Size 3.0, 5 spt, 14 shields
Turn 8 Settler->Archer
Size 1.6, 1 spt, 0 shields
Turn 11
Size 2.0, 3 spt, 5 shields
Turn 16 Archer->Settler
Size 2.75, 3 spt, 0 shields
Turn 18
Size 3.0, 5 spt, 8 shields
Turn 23 Settler->Archer
Size 1.75, 1 spt, 0 shields
Turn 25
Size 2.0, 3 spt, 4 shields
Turn 31 Archer->Settler
Size 2.9, 3 spt, 0 shields
Turn 32
Size 3.0, 5 spt, 5 shields
Turn 37 Settler->Archer
Size 1.75, 1 spt, 0 shields
Turn 45 Archer Completes
Size 2.9, 3 spt, 0 shields

You could build:
1 granary
4 settlers
2 archers
Avg. Size 3-4

Or you could build
3 settlers
3 archers
Avg. Size 2

Settler Built (+granary)
Turn 18
Turn 24
Turn 34
Turn 38 (bonus settler)
Turn 40
Turn 44 (bonus settler)
Settler Built (-granary)
Turn 8
Turn 23
Turn 28 (bonus settler)
Turn 37
Turn 43 (bonus settler)
Turn 48 (bonus settler)

The odd thing about this one is that it's almost equal even though it only has 3 fpt. It must be the fact that I kept the no granary team too small (not enough production).

Finally, the ultimate: the 5 fpt situation. 2 food bonuses (1 cow 1 wheat/game).

Cow is mined, wheat is irrigated. That way we know what's what for sure. More flexible in terms of shields vs food, anyway. 2 mined BG. The rest mined grassland.

Granary->Settler->Settler...

OR

Settler->Archer->Settler->Settler->Archer->Settler->Settler...

Turn 0 Granary started
Size 2.5, 3 spt, 0 shields
Turn 2
Size 3.0, 5 spt, 8 shields
Turn 6
Size 4.0, 7 spt, 30 shields
Turn 9 Switch wheat to mined grass.
Size 4.75, 7 spt, 52 shields
Size 4.75, (3 fpt), 8 spt, 52 shields
Turn 10 Granary->Settler
Size 4.9, 8 spt, 0 shields
Turn 11 Switch back to wheat
Size 5.5, 9 spt, 9 shields
Size 5.5, (5 fpt), 8 spt, 9 shields
Turn 13
Size 6.5, 9 spt, 26 shields
Turn 14 Settler->Settler
Size 4.75, 7 spt, 0 shields
Turn 15
Size 5.5, 8 spt, 8 shields
Turn 16
Size 5.75, 8 spt, 16 shields
Turn 17
Size 6.5, 9 spt, 25 shields
Turn 18 Settler->Settler
Size 4.75, 7 spt, 0 shields
(As you can see, this is the basic 4 turn settler factory. It probably would be best to reduce its size by making a few workers, but whatever).
Turn 22,26,30,34,38,42,46 All make settlers

OR

Turn 0 Settler started
Size 2.5, 3 spt, 0 shields
Turn 2
Size 3.0, 5 spt, 8 shields
Turn 6 Settler->Archer
Size 2.0, 3 spt, 0 shields
Turn 10
Size 3.0, 5 spt, 14 shields
Turn 12 Archer->Settler
Size 3.5, 5 spt, 0 shields
Turn 14
Size 4.0, 7 spt, 12 shields
Turn 17 Settler->Settler
Size 2.75, 3 spt, 0 shields
Turn 18
Size 3.0, 5 spt, 5 shields
Turn 22
Size 4.0, 7 spt, 27 shields
Turn 23 Settler->Archer
Size 2.25, 3 spt, 0 shields
Turn 26
Size 3.0, 5 spt, 11 shields
Turn 28 Archer->Settler
Size 3.5, 5 spt, 0 shields
(This is a 16 turn settler settler archer cycle)
Turn 33 Settler
Turn 39 Settler
Turn 44 Archer

You could build:
1 granary
9 settlers
Avg. Size 4-6

Or you could build
5 settlers
3 archers
Avg. Size 2-4

Settler Built (+granary)
Turn 14
Turn 18
Turn 22
Turn 26
Turn 30
Turn 34
Turn 34 (bonus settler)
Turn 38
Turn 38 (bonus settler)
Turn 42
Turn 42 (bonus settler)
Turn 46
Turn 46 (bonus settler)
Settler Built (-granary)
Turn 6
Turn 17
Turn 23
Turn 26 (bonus settler)
Turn 33
Turn 37 (bonus settler)
Turn 39
Turn 46 (bonus settler)

A granary clearly does much better when you have 5 fpt. Looks like my ways must change.

lots of work there.

the whole case for the no-granary start is that you make settlers and stuff from cities other than your capital. by just comparing 1 city, granary should win.

with no granary, each non-capital city does the same stuff as the city that made the settler, just 3-5 cycles behind. making a settler doubles growth per turn on average meaning it doubles the rate at which settlers come out.

there are too many variables to calculate shields and stuff. with workers joining to get a settler out faster, and making almost nothing but settlers + warriors (to keep order and fill in the gaps until settler can be made) this strategy isn't for a quick war. it's just to rapidly expand and get as many cities placed as possible before you run out of space. in the long run, this has to equal more shields and commerce and food.

the only valid comparison would be to calculate how going to war would increase productivity via captured cities or something, and you know we can't do that.

there is always the tradeoff between settlers and growth. of course if you grow to size 5 you can get out 5 archers and a settler in the same time it takes to get out 2 archers and a settler. there are infinitely many permutations and combinations of what can be done. the only thing i think you can measure is how much food you get each turn.

also shields are not an issue as long as the settler can come out in 20 turns. all that means is you need 1 shield other than the city square. this is very easy to accomplish for every city you'll think of settling. the only exception is flood plains and plains, in which case you need to have a worker.

Settler Built (+granary)
Turn 16
Turn 29
Turn 38
Turn 47
Settler Built (-granary)
Turn 6
Turn 26
Turn 46

in practice, there's definitely more than a 10 turn difference as to when the first settler comes out. you have to magically get pottery. you need to keep the worker to get such fast production also.