Sid games C3C: how to improve our skills

[ErikM]

In my game I noticed that AI will basically never attack a "green" army in the field. They will attack a city with an army in it. It must be an AI programming thingy. No doubt each unit computes some sort of cost/benefit ratio and chooses an action that maximizes payoff. So when an army is guarding a city, high likelihood of losing is probably offset by small propability of winning a city, and they attack. But in the field, an attacking unit will die almost surely for (from AI perspective) basically nothing. and so they look for a way around it.

In my game on the conquest thread, that was a cornerstone of my military strategy. First, it allows to stay in war almost indefinitely while never leaving republic (since AI never attacks your units). Secondly, you do all the attacking, so there are much fewer nasty surprises. Third, with some luck landscape-wise, you can channel their forces towards a killing zone to reduce their numbers before making your move.

I did the same thing nbarclay suggested to prepare for an invasion. The only difference is that I played continents and did some fighting against ironless AIs on my continent, so I had a couple of armies which were loaded with only two units to make them fit into caravels. Finished assembling them after landing.

[nbarclay]

[Edit: this was aimed at Aeson, but there was some crossposting in between.]

Sounds like you're doing well. What year did your invasion force land?

I've been trading the Zulus horses for a long time, but they don't have Chivalry yet. If they get it in the ten turns before our current trade expires (which is also when my troops are scheduled to land if I can get them assembled in time to launch my invasion then), I might have to face a few knights, but they won't have time to build or upgrade a huge number. As for horsemen, I'm not too worried about them; a Swiss Merc has better odds against a horseman without walls than against a MedInf with.

Yes, my Zulus have Gunpowder. I'm tempted to try to go for their Saltpeter city first thing, and if I had the forward bases you do, that would be my strategy. But without forward bases, my Swiss Mercs would have to survive for a fair while in enemy territory before I could deliver MedInfs to support them, or else I'd have to cut back dramatically on my number of defenders. Either one of those options could get ugly. But if I build a city of my own on the Zulu continent, I'm hoping I can draw in most of the Zulus' offensive forces to wear them down and set up to hit their Saltpeter city once reinforcements arrive.

I've started using my gold surplus to pursue Military Tradition, but I don't know whether I'll get it in time to do much good in the Zulu war or not. I'll have to change from Republic to Monarchy, and as much as I hate the timing, I'll probably need to do that about as soon as my GA ends (in between now and the time I attack) so I can benefit from some luxury deals that won't be renewed. In Monarchy, with fewer luxuries, and with my GA over, research will slow down a lot.

Are you sure armies are so vital to an effective offensive? I'm thinking that even against musketmen, a MedInf stack with about 50 trebuchets in support and plenty of Swiss Mercs to defend it should be able to do some pretty nasty things. And once the trebuchets can be upgraded to cannons (and the Zulus can no longer build additional musketmen to replace losses), such tactics would be even more effective.

Nathan

[nbarclay]

Vmxa1, just to be clear, it's the initial attacks with whatever reserves the Zulus have built up at the time I invade that really concern me. I'll have enough of an advantage in concentration of force to deal with units that trickle in later, even in a fast Sid-level trickle. But if I get pushed back into the sea before my reinforcements arrive, I'm in trouble - especially if I lose a lot of trebuchets in the process.

Nathan

[vmxa1]

Ok sorry I misunderstood. I guess if you use the 8 times cities plus 24 you have a base. I can only guess after that, but the more time that has gone by the larger the over the limit they will be.
Add at least 150 to the base.

The big issue is how many can be sent and in what time frame?
Figure 3 for defense in each city off the base should be close to what they have to send. If no horse types, you will probably see 20-25% coming. It depends on so many things, the size of the island and terrain, is it all roaded and so on.

Aeson had over 200 intially manage to reach him (first two turns).

[Aeson]

The Zulu have 19 cities, and I'm guessing they all are going to be producing a unit every turn, or close to it. They got Monotheism recently, so some may build Cathedrals, but that will be a few turns at most. I would be very suprised if they haven't built every possible improvement other than that in every city by now. So even after beating down their reserves, there will be a constant flow of units streaming towards any attacking stack.

Eric's point about the AI not attacking healthy Armies in the field is the important thing. To go on the offensive you have to leave the city walls and hills behind, and losses would start to mount quickly. With Armies you can make it to each city without being attacked at all, and by leaving a good defense in your cities you still get to fight all your battles at best odds.

Given that I get a Leader on defense pretty much every turn, and can likely average around 1 every 3 turns on offense at this point (just don't have many Elite MI yet) it's just a matter of time till I can get the Armies necessary to go on the offensive. That also gives me time to get a good number of Trebs together and I could have Metallurgy in 10-15 turns if I choose to. With Cannons (or even just a bunch of Trebs as the Zulu don't have Muskets) things will definitely go smoother. I have every 15 shield city (GA just ended) currently building Trebs, so I should have a lot by the time my situation is ready to switch to offensive.

I can't get to anyone else safely for another 30-40 turns most likely. The Zulu lands won't be too productive, other than allowing me more city sites to raise unit support numbers. So no rush here. I want to farm as many armies as I can get, filling them up with 3 units, get to Galleons, and then try to beat down the Egyptians.

I landed in 640AD IIRC BTW.

[ErikM]

Originally posted by nbarclay
Sounds like you're doing well. What year did your invasion force land?

It's not this start. See Your Sid level experiences thread on the conquest forum for a game log. It is currently on hold due to lack of time.

Are you sure armies are so vital to an effective offensive? I'm thinking that even against musketmen, a MedInf stack with about 50 trebuchets in support and plenty of Swiss Mercs to defend it should be able to do some pretty nasty things. And once the trebuchets can be upgraded to cannons (and the Zulus can no longer build additional musketmen to replace losses), such tactics would be even more effective

Not so much that they are vital on the offensive but they are vital to block AI counterattacks.

[vmxa1]

I think that Aeson laid out the logic behind using armies. The extra moves can let you get in a city. I will sometimes send a few armies next to the city with no attacking that turn, especially if I get a terrain bonus. So I can attacking and leave one move left to get inside. Now I can hopefully send in backup using the roads.

Even superior units will take a pounding in the open as they cannot heal and the AI will send a lot of troops at them. The trickle in is more like a stream for awhile. The problem is that is will not be easy to replace large losses in a timely fashion, so I have to be more defensive.

I like to have a defensive army move with those stacks to absort most of the punishment, if they do attack it. Even a 4 man army of pikes can take down a bunch of troops as it will be able to fort up.

[Aeson]

I've seen about 400 Zulu units in the first 7 turns. My kill counts don't take into account the 60 or so Impi which retreated from battle, or the 'defensive' Impi that don't attack. The 7th turn counts were somewhere inbetween the 5th and 6th turn counts, and the Zulu still have roughly 100 more units in view.

Just a guess, but I think they had around 800 units when I landed. 5-10 in each city, and at least a couple hundred non-Impi up N that I wouldn't have seen yet. They were at war with Egypt, and that may have drawn quite a few units to the N, and the turn they signed peace with Egypt I signed Egypt up against them again too. The slow units from furthest N would be showing up for another few turns.

[nbarclay]

ErikM, I was actually responding to Aeson in that message, but you posted in between. Thanks for your input, though.

Aeson, I'm definitely thinking a Swiss Merc army or two to protect the trebuchet/MedInf stack would be useful. But you already had that as of your status report, which left me wondering why you were so eager for additional armies to use in the war with the Zulus. My inclination would be to build the armies but not populate most of them (other than defensive ones) until I got cavalry. Cavalry armies would be vastly more powerful than MedInf or knight armies for the assault on Egypt, and if we attack the Zulus with individual units rather than with armies, we get more chances to generate leaders offensively.

By the way, how does your culture compare with that of the Zulus at this point, and are you capturing or razing cities? I suspect I'll be mostly razing in the early stages; I'll probably have over a third of the Zulus' culture when I attack, but not by much.

Nathan

[Aeson]

My per turn culture is approaching that of the Zulu, but I will raze everything but the Wonder cities anyways. All those captured workers will make nice coastal blockades... and can be kept for rails later on without ruining my unit support costs.

The army thing is mainly "the more the merrier". I'd prefer to have most of the fighting be defensive on my part, and then have a giant nutcracker that takes no losses (or close to it) in pounding city after city into the dust. I like attacking Pikes in cities with Armies, even if I have them down to 1 hp. If you get hit by bombardment coming in, 3-4 vs 1 hp, with odds favoring the defender for each hp, isn't terribly efficient. 1 hp units in the open are good for the individual units, and there won't be any shortage of them. I'll probably hit the Army cap from defensive Leaders anyways. Then near the end of the war I can disband MI Armies as I get Leaders and save empty Armies for Cavalry.

You can also rush Armies from the Military Academy in C3C, so disbanding Armies to build new Armies will help keep the Armies you have as modern as possible.

[vmxa1]

That is about the size of it. You have so many battles that you can almost count on getting leaders for armies. The old ones can be disbanded to rush a struture or two if you get to the max for your city count.

Once you start using knights against those stragglers they get to popping leaders.

I had failed to mention that the AI often has large numbers of units that it will not attack with in the first phase and only used as defenders for its attackers. If you have two move units you can pick off a number of these after they get damaged.

[nbarclay]

In 650 AD, two landmark events occurred. First, the Dutch people emerged from anarchy into Monarchy. And second, Dutch forces landed in Zululand. The invasion force consisted of all twenty-six trebuchets currently available, the first seven knights upgraded from horsemen, five MedInfs, a settler to replace Swazi once it's razed or abandoned, an elite sword unit, and 32 Swiss Mercenaries - a total of 72 units. I decided to go after the Zulu saltpeter supply right away after all, since I had more galleys available for the operation than I'd originally been thinking in terms of. Even so, there is some risk that I might have to turn around and go home if I take too many losses the first turn. I checked into getting an alliance with Egypt, but it would have cost far more than I was willing to pay. I ended up able to trade one luxury for two to Egypt, so I ended up renewing that luxury deal after all. With the luxury slider at 20%, that put me at 11 turns to Metallurgy with a deficit of about 300 gpt.

Two Swiss Mercs were lost in the initial Zulu counterattack, several more injured, and about half a dozen promoted to elite (albeit all of them injured in the process). When it was my turn again, I started pounding the Zulu saltpeter city with trebuchets and got the three musketmen down to one hit point each, the pikeman to two, and the three or four impis and MedInf to three. There was also a longbowman defending the city. I lost one MedInf in the ensuing assault, but barely had enough offensive units to finish the job. A knight razed the city, and I then founded my own city of Saltpetersburgh on top of the saltpeter supply, moving the knight back along with the workers from the raze. I couldn't really have asked for the invasion to go more smoothly.

[nbarclay]

It took most of a century to beat off the Zulu reserves, but I finally did it. About three turns into the war, I enlisted Egypt as an ally hoping to keep the Zulus busy on their northern flank. Finally, on the one hundredth anniversery of the initial landing, Dutch forces celebrated by burning a second Zulu city to the ground.

I only got four leaders so far, three of them building armies and the fourth rushing a certain structure with five sides. Two of the armies are remaining unpopulated thus far; to build the Heroic Epic, I pinged a longbowman down to one hit point and attacked with a Swiss Merc army. I also just got Metallurgy, so I'll be upgrading my trebuchets to cannons before I continue my offensive.

[vmxa1]

What sort of numbers did you encounter?

[nbarclay]

When I thought I'd dealt with the initial counterattack, I was a bit too optimistic. I was able to raze a third Zulu city and replace it with one of my own, and then found a fourth city and a fifth pushing my borders forward. But then so many Impi started swarming, accompanying MedInfs and longbowmen and pillaging tiles, that my offensive forces and bombardment units were too busy dealing with intruders to push my borders forward.

That was complicated by another problem: the Zulu island cities controlled the sea lanes between our homelands. I'd forgotten that cannons require iron as well as saltpeter; it's never been an issue for me before that I recall, but it was this time. I was able to hook up an iron source on the Zulu continent briefly and upgrade a few of my trebuchets, but most of them were too busy. Then the Zulus cut the road, and later, my iron city of New Hague (one I'd built myself) flipped, costing me a 4x Swiss Merc army and a stand-alone elite Swiss Merc.

I could have lived with using trebuchets instead of cannons, but without a connection between the land mass with horses and the one with saltpeter, cavalry would be impossible. Since my forces on the Zulu homeland were busy, that meant my homeland had to build an entire second, separate invasion force to take the islands. Which, in turn, meant that my forces in Zululand would have to survive on their own without reinforcements for a while. With trebuchets and cannons to soften up the Zulu units, my knights and MedInfs could finish them off with minimal losses. But minimal losses are not the same as no losses, and over time, my units found themselves taking greater and greater chances attacking while injured rather than allow injured Zulu units to retreat and heal.

Finally, as the millennium drew to a close, I managed to concentrate enough healthy and almost-healthy offensive units to raze the Zulu city two tiles from where I'd earlier founded New Rotterdam, thereby taking control of most of the continent's wines. About the same time, my island invasion force started its work, capturing two cities and razing and replacing a third. At long last, thanks to the Great Lighthouse (since I still didn't have Astronomy), I was able to bring horses and saltpeter together to start upgrading knights to cavalry.

And that's where things stand as of 990 AD. I've started upgrading knights to cavalry, but with no more income than I have, upgrading the horsemen I've had blocking my homeland from invasion for centuries will take a lot longer. I finally got Education from the Great Library a few turns ago, but since I need gold for upgrades, I'm going to hope the Zulus research Astronomy for me and I can get it from them in peace negotiations.

Once my forces are consolidated, I'll have over ninety bombardment units in Zulu lands, enough to beat off Zulu counterattackers and push forward with offensive operations simultaneously. And I'll have over a dozen additional cavalry as well Maybe, at long last, I'll be able to make steady progress instead of just holding my own.

Egypt recently completed Newton's University but has not yet started on any industrial wonders, so I can hope there's still time to at least seriously prune them before they get infantry. Otherwise, I'm in trouble.

Nathan