War With Zululand
The AI won't generally attack armies. I did find out the hard way that they'll attack if an army is weak enough that they see a possibility of killing it with one attacker. But they won't throw a stack in to kill a strong (healed) army.
So I sent two Knight armies into Zulu territory on special missions. One was to hang around the unconnected horses and ensure they stayed unconnected. The other was to pillage a couple of luxuries.
I also sent a group of four Cavalry armies heading for the Zulu capital Zimbabwe. The Zulu had built the Great Wall, Sun Tzu's, and Leonardo's there. And there were three sources of wine beside it. If I could destroy Zimbabwe it would be a serious setback for the Zulu.
The rest of my forces moved to the ex Arabia/Zulu border, founded two towns, and dug in to fight the Zulu there.
From 910AD to 1040AD the situation in the warzone stayed pretty much the same. Each turn large stacks of Zulu would press forward and I'd destroy as many as I could, trying to at least kill all those who could reach one of my cities or pillage one of my transport routes next turn. My armies would attack the strongest units in a Zulu stack, then my Cavalry and then Knights would pick off the weaker and weakened units.
A picture at end of 930AD (yes, I'm about to lose some workers, I couldn't destroy the invaders near one group):
The eastmost Zulu Rifleman visible above is a stack consisting of 33 Longbowmen, 7 Medieval Infantry, 10 Musketmen, and 8 Riflemen. Two of the other visible stacks are about the same size. The stacks closest to my units are much smaller, have been whittled down by my forces.
Another picture, at the end of 990AD:
Over time the brunt of the attack had moved to the western part of the warfront. The Rifleman southwest of center above is a stack with 1 Crusader, 1 Warrior, 11 Longbowmen, 8 Medieval Infantry, and 13 Riflemen. It started out as a much larger stack but this is what I've chopped it down to as it moved south.
At this date I had 43 Knights (32 of them in 8 armies, the rest elites), 118 Cavalry (72 of them in 18 armies) and 7 Cannons (I'd just started serious production of these.) I'd killed 425 Zulu units so far and my military was still rated as weak compared with them.
I started to worry that the Zulu might be producing units as quickly as I was killing them. It seemed impossible but... Hmmm, a Zulu city with a factory, if they were mobilized, might produce 25 to 40 shields/turn. With their Sid level bonus they need just 32 shields for a Rifleman. So, many of their cities might produce a Rifleman per turn? A stalemate would be bad news for me - the Zulu were fast researchers who'd end up with better units than I could handle. I was gaining a bit in strength and was producing new armies. But the average quality of the incoming units had already improved - Medieval Infantry and Riflemen were arriving instead of Longbowmen and Musketmen. These were the kind of paranoid thoughts running through my mind as the bloodshed at the warfront continued...
In the meantime, my four Cavalry army group had penetrated to the Zulu capital. In 940AD they were in striking distance but instead of attacking (with just 1 strike each) they used their free (no movement cost) pillage ability to cut the capital off from any returning reinforcement attempt:
On the next turn my armies attacked, killed 7 defending Riflemen, and razed the Zulu capital. I don't recall ever razing Sun Tzu's and Leonardo's before. It has always been my goal to capture them. But my culture was so low there was no hope of holding them in this situation.
I got a few cannons from the razing of Zimbabwe and realized I should be collecting them. A Knight army from the warfront began the journey to the Zulu heartland. For the moment a Cavalry army guarded the Cannons - they transferred to the Knight army later on and it then followed the other armies around collecting cannons and slaves.
The ability of armies to heal in enemy territory is a great bonus. After resting a while my Cavalry armies started destroying additional targets in the Zulu core. In 1020 I sent a second group of four armies to do the same. By 1100AD my invaders had razed the Zulu cities housing The Sistine Chapel, Copernicus', and Newton's, and a couple of other core Zulu cities.
In 1050AD I finally felt reasonably safe at the warfront. I was destroying all arriving Zulu units before they entered my zone, and I was able to start pushing forward into their territory. My paranoia had been unfounded
On the bad side of things, I encountered a Zulu Infantry in 1030AD. It seemed that her research had hardly slowed at all, she was at Replaceable Parts already.
Next came a long phase where the fighting remained as intense, but I was razing Zulu cities at the border and pushing forward a bit. I found that a four Cavalry army could attack a city defended by a fortified veteran Infantry and expect to win with a loss of about 1/3 of its hitpoints. (Sometimes I'd lose almost no HP, sometimes go red, and very occasionally I did lose an army in such an attack.) So I could send 4 or 5 healed armies 3 tiles deep in enemy lands, each could still attack once in the same turn, and they'd take out an enemy city. (Typical defense was 4 Infantry after the Zulu got all units upgraded/replaced with their new technology.)
In 1140AD I captured the Zulu city which had JS Bach's. This was the first city I intended to keep. If it flipped I'd retake it. I moved all my remaining warriors (happiness MPs) to this city to quell resistors and started starving it down. It flipped to the Zulu at size six and I retook it. This wonder is one I will not raze!
In 1160AD my military, for the first time, was considered roughly equal to that of the Zulu. I finally felt in solid control. I'd taken a good part of the Zulu territory, taken her wines, taken JS Bach's, and put a huge dent in her production and research capabilities:
At this point I had 32 Knights (in 8 armies), 134 Cavalry (108 of them in 27 armies), 78 of my own cannon, and at a guess 40 captured cannon. I was paying just a bit for unit support - was allowed 428 units and had 441 including settlers and workers.
From that point on my invasion picked up speed. I shifted my spending, using all available cash to rush settlers to claim the razed land (instead of using cash for armies as I had to this point.)
In 1220AD I was ready to consider another target. There was little left of the Zulu, just some bits of cleanup remained:
War With America
I'd given America peace in 1020AD in exchange for an alliance against Zulu - I didn't particularly care about the peace but I did want to keep America at war. The timing on this peace deal worked out very well. At the end of my 1220AD turn America tried to renegotiate peace. I refused and we were back at war.
America had Cavalry but did not have Infantry. And her forces weren't large - the wars with Arabia and then Zulu had probably cost her a lot. So this war was an easy one for my military. But time consuming. Each turn I'd regather all my units, then apply them to remaining Zulu holdings or to America, as seemed best. I used many of my workers at the warfront to railroad into newly razed territory, which often allowed me to bring some cannons to the front, weakening enemy units and providing more targets which single Cavalry units could dispatch.
It took just four turns to run through the center of America and to drive the Zulu entirely off the landmass. In 1255AD, which is my game date at the time of writing, the map looks like this:
Shifting Gears
In 1255AD, having taken all the Zulu territory which I could reach by land, I gave the Zulu peace for four island towns. Leaving her with just one city, her capital. (Note: Even at this point of near destruction, Zululand would not include any tech in the peace deal.)
I'd remobilized twice during the long fight, once when the Inca finished off Arabia in 1010AD, and once when I made peace with America.
This time, when making peace with Zulu, I decided not to remobilize. My existing military will easily take control of the remaining areas I can reach by land. It is high time to get back to work on aqueducts, marketplaces, courthouses, universities, etc! I've reviewed all cities and reassigned them to infrastructure and growth projects.
I've been mobilized for 39 turns. Have been at war with Zulu for 38 turns, and with America for the last part of that. During that time I have destroyed 1469 enemy units and lost 145 of mine. The lost units include 3 armies of four Knights, 2 armies of four Cavalry, and 120 additional units (mostly Cavalry, a few elite Knights, I didn't keep separate counts.)
I've taken over five cities (Mpondo with JS Bach's, plus four cities included in peace deal.) In all other cases I've resettled where I razed enemy cities.
My forces currently consist of:
7 armies of four Knights
27 armies of four Cavalry
63 additional Cavalry
79 of my own Cannons
about 60 captured Cannons
I also have:
132 Workers
13 Settlers (many new ones each turn)
hundreds (500 at a guess) of foreign Workers
I've had 30 Great Leaders so far. They've all been used to build armies, and I've built 9 additional armies with the Military Academy.
Future Plans
This post brings the story up to date.
My immediate plan is to finish taking over the American lands and to take the tiny area controlled by Mongols. At that point I'll own my entire landmass except for a few cities the Inca managed to establish.
I will then be pretty close to the domination limit. I'm currently at 57%. Nonetheless I'll want to destroy the Inca and Aztecs as soon as I can. They're both strong enough to win by culture or launch a spaceship if left unchecked. I expect I'll also wipe out Egypt and China just to get the best possible choice of milking lands. And to get more luxuries - there are just five in all the lands I currently control.
The Inca, and perhaps even the Aztecs, may be a challenge. My military advisor considers my forces (which seem huge to me) to be weak relative to both Inca and Aztecs. And along the way I've sacrificied something which was in my original thinking - I've created four unit armies. Initially I'd planned to limit them to three units so that they could travel on Galleons after completing their work on the home landmass. Hah! That was before I saw the Zulu strength. I used all the force I could muster against Zululand and I'm glad I did. But it leaves me with armies which require Transports to reach my remaining rivals.
So my plan is to begin research soon and work toward building Transports. If I can do that while my current armies are still useful, great. If not I'll have to make new plans to deal with Inca and Aztecs. Either way I'm not sure how tough they'll be - might be straightforward, might be a challenge.
While researching I'll build up infrastructure and score. I'm looking forward to having Replaceable Parts and being able to use Civil Engineers soon.
I may also declare war on Inca while researching, and drive them off my landmass. If I do this I'll also want to ally Aztecs against them, which might slow them both down a bit as they fight at sea and fend off small invasions from each other.
It looks like my 50 turn research of Electricity will pay off in getting one tech for free. My single devoted scientist expects to understand Electricity 7 turns from now. During that time I'll build some Libraries and Universities. After learning Electricity I'll go for Scientific Method and a prebuild of Theory of Evolution, unless a rival is already on that path. If I can be first to get that it will of course give me a nice research boost, as well as denying that boost to my rivals.
I'll post again when I've made some progress!
Although there are many ways to milk the cow, yours is the best I heard so far - most honorable one too.
I think I will follow your good example and stop robbing the AIs' treasuries from now on. I thought that if it was ok for Robinhood, it must be ok for me, but now I know that my victory so far on Sid was a hollow one beause you just show me a better way. I'm so looking forward to patch 1.21 so I can try out Sid again with your Great Library tactic. Thanks for sharing a very well thought and proven tactic for Sid!
Saving for later was pointless vs. ensuring that there would be a "later"! 
Though I started very early, the AI was already at war on a few fronts and probably ignored my weak military capability...
