I just find this city specialisation hard to achieve.
You can't build the 'multiplier' Ancient-Age improvements -- Libs, Markets, Courthouses, Harbours -- until you've researched/ exchanged a lot of optional and/or late-AA techs anyway, so until then you can't really be specialising your cities beyond 'Settlers/Workers vs. Military,' either. Also, in the Ancient Age,
all mil-units cost less than Settlers to produce, in terms of shields+food. You'll need mil-units anyway to keep order in your cities, and intimidate/ attack/ defend yourself from the AICivs once you meet them, so without a 'good' bonus-start/ pump-site as your capital, mil-units (even if only Warriors) should almost always be your first build(s). You would then use your exploring reg-Warriors to look around for the local bonus city-sites while your capital grows a bit, and then build 1-2 Settlers to occupy those sites (see below). And if you're unlucky enough to get no visible food-bonuses in '1st-ring' positions at all, then that will tell you that your expansion will probably have to be military-driven, so all your cities can then be turned over to making mil-units (with the fastest grower perhaps also building the occasional Settler), and then sending those mil-stacks out to make the AI miserable.
In a lot of examples posted on these boards I find that most starts are with some food bonus or multiple bonus. I feel that food bonuses are incredibly powerful and change the game early on way too dramatically basically deciding it then and there
Well, yes. Having bonuses, especially food-bonuses, will clearly make any start more powerful. But I don't think they're absolutely essential for a win at Emp/DG, and possibly not even at Deity (AI gets a 60% bonus and lots of starting units), unless your criteria for a 'win'
also includes getting a HoF-worthy score (yes, Lanzelot, I'm looking at you...).
For example, even one 4 turn settler factory is sufficient to out expand Emperor level advantages.
In a normal game (i.e. not a 1CC!) you will almost certainly need to build at least the first Settler out of your Capital (because you will have no other place to do it), but there's no rule that says that your Capital
must become your pump -- or indeed that the first pump you build will be your only pump throughout the entire game. Doing so would almost always be counterproductive: while your first pump will probably be one of your 1st-ring towns, once you start building your 2nd-ring, you will certainly want to get your capital and most/all of your high-commerce/ high-shield 1st-ring cities to Pop12 ASAP (and build the necessary infrastructure), to be able to use them as e.g. Lib-towns, Barracks-towns, or your 20K Culture/ Wonder-city. The only circumstances where a 1st-ring town might then stay as a Settler-pump a little bit longer would be if it had no freshwater -- and you hadn't got Construction yet, so couldn't build a 'Duct.
I'm not sure how I can see that working out. Don't the outer cities take longer to develop, produce less shields and be less effective in building units?
Yes, the further away from your capital a town is, the more corruption/waste there will be, ramping up sharply once you're over your OCN. But I'm not using 'inner' and 'outer' as synoyms for 'core' and 'farms' here, I'm using them as comparative terms. In the very early game (first 50-70T), only your capital will be 'inner' and your 1st-ring cities will be 'outer' -- but so long as you've built all those at around distance 3-5 from your Capital, connected them, and improved all the tiles they're working, they should still be reasonably productive, even after the corruption/waste is accounted for.
Once you've started building/capturing your 2nd ring, your 1st-ring cities become 'inner', and by that time, they should also be reasonably well-grown (Pop5-7), and you should have switched govs, and therefore have the lowered waste/corruption and higher shield-output to build the more expensive 'specialist' multiplier-improvements (Libs, Markets, 'Ducts) or mil-units (Swords/ Maces/ LBMs, Pikes/ Muskets, Horses/ Knights, Trebs/ Cannon) as needed. This is also when you'll need to start (thinking about) building your FP somewhere to maximise your OCN, especially when this is reduced by built-in constraints (e.g. high difficulty-levels, Small/Tiny maps).
Your 2nd-ringers are now your 'outer' towns, and with their lower/ more significantly corrupted SPTs, these should be building only the cheaper improvements or mil-units (e.g. Barracks, Walls, Warriors, Chariots, Archers, Spears, Cats) -- until you start acquiring your 3rd ring...
...that's an effective doubling of your total production, income and growth. Then you build another city and that's increasing production, income and growth by one third. Then you build another city and that increases production income and growth by one fourth. However as you increase in cities and corruption starts to take hold, you get diminshing returns in gold...
In terms of proportionality, you're right. But you could also look at at it this way: doubling your number of cities doubles your output of, well, everything! Here's an example, using a non-Agri (no city food bonus), non-Exp (no Scout), non-Ind (normal Worker-speed) Civ, starting on a 'minimally good' Grass/Plains site, with no food-bonuses, but still useable (i.e. +2FPT, 2SPT at Pop1), and building only Settlers, Warriors and Workers:
Turns 1-10:
- Capital founded, builds Warrior1+2 (5T+5T)
- Worker1 improves 'best' local tile (e.g. roads+mines (B)Grass or roads+irrigates Plains) (8-10T)
- Warrior1 begins exploring around Capital (e.g. 3W, 2N, i.e. expanding spiral)
- Capital --> Pop2, Cultural border expands
Turns 11-20:
- Capital (minimum +2FPT, 3SPT) begins Settler1 (10T)
- Worker1 roads+mines/irrigates 'second-best' tile near Capital (8-10T)
- Warrior1 continues exploration (e.g. 1N, 6E, 3S)
- Around 1/2 of the 1st-ring area will be explored by Turn 20
- Warrior2 keeps order in Capital (assuming Emp+, no Lux)
- Capital --> Pop3, builds Settler1 --> Pop1
Turns 21-25:
- Capital (+2FPT, 2SPT) builds Warrior3 (5T)
- Settler1 moves to 'best visible' 1st-ring site (distance 3-4), founds City2 (4-5T)
- In this example, City2 will therefore be in the northern (explored) half of the 1st ring
- Warrior1 continues exploring (3S, 2W)
- Warrior2 begins exploring, covering still-fogged ground (3S, 2W)
- Entire 1st-ring area now visible, so you should now have a pretty good idea where you want to place the rest of your 1st-ring cities -- and in which order -- to maximise return on Settler-investment ASAP
- Worker1 roads towards City2 (max. 12-16T)
Turns 26-30:
- Capital begins Worker2 (5T)
- City2 builds Warrior4 (5T)
- Warrior1 and/or Warrior2 can now
- Continue their outward spiral(s)
- Start moving directly away from Capital and City2 to contact your neighbouring AICivs (if they haven't sent anything your way yet)
- Fortify at your next planned city-site(s) -- with the next-built Warriors from the Capital and/or City2 heading northwards/ eastwards rather than accompanying a Settler
- Warrior3 fortifies in Capital
- Capital --> Pop2, builds Worker2 --> Pop1
- City2 --> Pop2
Turns 30-40:
- Capital (+2FPT, 2SPT) builds Warrior5+6 (5+5T) or Chariot/ Archer/ Spear1 (10T) or Barracks (10T if Mil, 10+7T if not)
- City2 (+2FPT, 3SPT) begins Settler2 (10T)
- Worker1 begins improving City2's best tile (8-10T)
- Worker2 begins improving Capital's 3rd-best tile (8-10T)
- [Warriors1+2 continue exploring, hopefully contacting AICivs]
- Warrior4 fortifies in City2
- Capital --> Pop2
- City2 --> Pop3, Settler2 built -->Pop1
Turns 40-50:
- Capital (+2FPT, 3SPT) begins Settler3 (10T)
- City2 (+2FPT, 2SPT) builds Warrior7+8 (5+5T) or Chariot/ Spear/ Archer2 (10T) or Barracks (10-17T)
- Settler2+Warrior4 move to best available 1st-ring site, founds City3 (4-5T)
- City3 begins Warrior9 (5T) or Barracks (10-17T)
- Capital --> Pop3, Settler3 built --> Pop1
- City2 --> Pop2
Turns 50-55:
- Settler3+Warrior7 move to next best available 1st-ring site, founds City4 (4-5T)
So by this point in the game,
at worst, with no bonuses or Luxes anywhere in your 1st ring, and no chopping/ rushing/ whipping anywhere, you could still have at least 4 cities (including your capital), and have done quite a lot of exploration (assuming no Barbarians!). You might also have contacted at least your nearest neighbours, hopefully been able to exchange a few techs, and (if you haven't already picked a specific Civ/ Map-settings to channel a specific VC) have a pretty good idea of what VC you want to aim for.
Using the same 'Pop2 city (+2FPT, 3SPT) makes a Settler, Pop1 city (+2FPT, 2SPT) makes Warriors/Workers' scheme outlined above, those 4 cities can in the next 20T make another 4 Settlers, and more than enough Warriors/Workers to guard/improve the resulting cities. Therefore by Turns 75-80
at the latest you could have your first ring finished, with ~8 cities (including your capital), and possibly some city-improvements already built (e.g. Walls/ Barracks/ Granaries/ Temples, depending on your city-sites, and your Civ-Traits/ starting techs).
Notice that in the above -- very conservative -- example, the second city was founded after 25 turns, the 3rd about 20 turns later (Turn 45), and the 4th city 10T after that (Turn 55), with the next 4 cities predicted within the next 25 turns -- now that's what
I call exponential!
And obviously, if your Civ
does have any of the Trait-advantages that I deliberately excluded, and/or you
do have any food-bonuses or Luxes in/near your 1st ring, and/or you
do start forest-chopping/ whipping etc., you should be able to complete a full 1st-ring in (much) less time than this.
Military on the other hand is slightly different right? If you have one lone horseman it could probably only be used for defence. However if you increase the numbers, one, two, three, four etc as the numbers go up, the increase in military value is at a higher rate compared to city growth. For example with an army of 10 horsemen you could probably take 2-3 cities (or even more depending upon circumstances).
Yes, so long as you have Horses and HBR...
But that example just serves to illustrate my point about more cities = more pop = more output = more options = more everything. Military is just one aspect of this: the more cities you have (up to the OCN, anyway), the faster you can churn out those mil-units for conquest (if that's the way you've decided to play), and the more cities you can capture/PT-extort from the AI (which will itself raise your free unit-limit). With only a few cities, how long will it take you to get the tech and build those 10 Horses? And how many mil-units/cities will the AI have been able to build/ plant in the meantime?
Something else to bear in mind: at every stage in the game, the most modern mil-units always cost you (a lot) less to build/maintain than the latest city-improvements. And at higher difficulty-levels, city-improvements cost the AIs (a lot) less to build than they'll cost you. So it's almost always cheaper to build a Barracks (or 3) and a stack of veteran mil-units, and then go and capture AI cities which already have infrastructure built, than to build it yourself.