Armies and alternative units

[Cassius Critzer]

I adjusted the Attacker class army some to hopefully encourage the AI to use them. They take slightly more shields for a significant boost, but perhaps in the AI code it's not sufficient to switch to them.

I tested Imperial class armies (which are regiments) versus Defender class armies (which are battalions) on rich farmland (which has an excellent defenses even without a fortress. That resulted in a win but with as little as 8 hp of damage (ie you crushed them) to as much as 96 (you won but it was a nail biter)! Whatever the unit value which is 12 HP is multiplied times ten ie a range from 0-120.

An Imperial regiment attacking a Defender class battalion within a fortress resulted in the reigiment sometimes able to kill two who were stacked in a fortress( which is excellent to see the AI doing) but also 50% of the time resulted in the regiment dying ie they used up every last bit to take them out.

Defensive battles with battles and regiments are not predetermined outcomes. They are delightfully risky with some spectacular outcomes.

Rich farmland is a misnomer. It's the pinacle of terrain modification that incorporates military engineering with agricultural upgrades. I need a better name. I had considered an intermediate so it takes a longer time to fully upgrade but some of Japan has swamps that were historically drained for habitable use. So I may still use that swamp terrain.

 

[Cassius Critzer]

November 1478

The AI has created more than a thousand Defender class battalions. It's using them to ruin infrastructure upgrades to slow down invasions, challenge enemy probes, and now defending from multiple invasions all over Japan.

67 Defender class have died. 67 x 650 equals 43,550 samurai deaths.

The Imagawa who got booted out of Kyushu early, then wisely spent their money, are in last place on tech progression, but built a massive Defender class force. They then swarmed an Oda port city and took it. That's the first totally upgraded city/citadel to fall since 1448. They utilized mixed naval and army use to whittle down the defenders. Meanwhile they simultaneously are succeeding in beseiging Takeda territory by occupying fortresses and ruining farming and trade income. It's great.

Were I not testing simultaneously as I have an alliance with the Takeda, they might kick Takeda Shingen's butt and he has twice the terrain...but foolishly pushed tech acquision and neglected his defenses. Now he's playing catch-up and the Hojo are invading, the Oda, Mori, and Uesugi have declared war. He's being squeezed.

It's so great!

 

[Cassius Critzer]

After breaking the siege and occupation of tiles by the Imagawa in one part of the borderlands of Takeda territory, they just shifted elsewhere to the southeast. The AI is smart enough not to keep risking its Defender class battalions as it lost several that I eliminated so repulsing the rest.

It was a very tough battle with severely hurt regiment class Imperial armies mostly surviving. In a real game under such an alliance, most human players would just pay lip service to the alliance and at most just use naval forces to harass the shared enemy.

But in a real game, you would absolutely have to take a bordertown as a staging area so you could heal these battalions and regiments because otherwise they are targets as they have been severely injured. Then cycle out some harassing company class units to attack the enemies weaker occupying company class units and counter-invade the shared enemy territory.

These Defender class battalions are able to attack as higher value attack and defense units but are designed for garrison forces. They can hold fast in a tight spot but ultimately will lose. Think of the scene in the battle between the Imperial forces and the standard samurai in The Last Samurai film. They might win some engagements when brilliantly employed using tactics, but cannot ultimately conquer or even adequately defend such as to be unassailable. Nobody wins in chess by constantly defending.

The AI's using an occupying strategy. There are up to 21 slots for a city "grid". This means if occupying and the city is not a port city, then agriculture is severely curtailed. And the AI chooses to not put every citizen to work but generally has more specialists than required.

Using the modified terrain values, and at maximum levels, the city can create a surplus of up to 120 per season, but the AI will often create surpluses of 50. This means with occupation of these tiles that they will go into a hunger situation and draw down on their storehouse. At maximum they can hold out for two years.

This seems historically correct as the samurai of this era had dried preserved harvest from the land and sea, which they rotated, and kept in storehouses. You can see this in Twilight Samurai as the protagonist's responsibility is there.

Now if the city is one of the 20 city/citadels that I created as the base for turn one, then in most cases these were either rich farmland or worked upon to transfofrm from grassland to rich farmland. This means that even if every single tile is occupied in a non-port city, then they still will have adequate amounts to draw from to support several Defender class battalions. They will after two years begin dying one citizen per season but may of these can be as high as 36 in inhabitants by 1479. This means you cannot wait them out but the beseiged are in a tight spot.

If I alter the food and shields generated by reducing this, then such occupations after two years put the city in disorder. That may happen...it may need to happen.

It's interesting to see the AI use this tactic with little cost to its invading forces. Battles may resolve by attrition and be very grindy.

You can adjust it so cities have inadequate resources in such situations as if the generated units are "not supportable" so they dissipate. The way it's set up now, a remarkable amount of units can be supported by an urban maxed out city/citadel.

 

[Cassius Critzer]

If you were to occupy a city region's "grid" of 20 tiles and the city was founded on a mountain tile, and some are, then zero food but only shields are generated. This means if defending such cities you cannot allow an occupation to occur as after two years, there will be no military units that can be 'supported". They will wither away.

With swarming happening of multiple tribes then human players of the Jidaigeki mod may face tough invasions that are hard to displace. This means adequate fortresses are built and garrisoned to prevent that from occurring. Army units change tactics.

As above, the Takeda were in a situation with inadequate garrisons and a genuine invasion by the Imagawa, a halfhearted less orchestrated one, possibly as a "feint" by the Hojo, and war declared by many. The AI are using alliances and secret alliances when vendetta are generated most likely due to the threat of battalion and regiment armies.