Endgame battles

[Anskiy]

So I think the most common complaint I've heard about CoM is that almost invariably, the end game ends up being many battles of attrition, thanks to the AI's tendency to spam crappy halberdier/uncommon summon stacks. Basically the player is forced to conduct many meaningless battles even though they have assured their victory a long time ago. What is the best fix to this problem, in your opinion?
I know zitro advocates for being able to turn on strategic combat at will, and that would be a partial fix to the problem.

[Seravy]

Strong enough garrisons that the AI can't attack your cities with junk units so you can ignore them all should be possible already in theory.
Does the AI still attack strong garrisons with weak stacks that have no hope of ever winning? That would indicate there is a flaw in the strategic combat calculations.

[Anskiy]

This isn't just about getting attacked in cities. It's about field battles and taking cities. The AI will often engage your armies on the field, or force an engagement by occupying space even if they have no chance of winning. And really, I think they should stop summoning uncommons or building halberdiers past a point of time, as they become irrelevant as anything but filler past a point.

[Suppanut]

I don't completely agree with Anskiy. AI mostly recruit halberdiers due to they are generally their best normal unit available as they lack of late game option for normal units as AI go for number. But I do see AI number is weakness to exploit for some strategy as mass number of weak stack make ai vulnerable to strategy to bleed AI's mana (especially as sorcery player) or to chaos global curses (especially Doomsday or other chaos curse stacking up)

[zitro1987]

The main problem is those frequent low-stakes field battles where your weak or small stacks (often moving to rally points) get attacked by mediocre AI stacks. AI has nothing to gain defeating a wolf rider unit on the field when you can produce 10 per turn. For game enjoyment, this could ve a mini sacrifice the ai could take

Other than the choice to turn strategic combat at will, the more important solution is to discourage these situations a bit

Both AI and player will have hundreds of units no matter what - that probably will never change

So then it is about encouraging AI to hold off on these inconsequential battles a bit and field mini combat decisions could be adjusted according to game year when things get excessive in unit production. Of course if a stack of 6 halberdiers encounters a lone hero or something of interest AI should attack But why should the stack of 6 halberdiers attack a lone swordsman? whats the point? These inconsequential attacks are just padding game length.

Long story short, AI to attack bigger threats it can win against

[Suppanut]

I think ai should prioritizing threat of each stack. It should still attack only stack it could win on auto-resolve but it should priority greatest threat around them that it could handle first and would take weak one only if no higher priority target for them.

[Seravy]

But why should the stack of 6 halberdiers attack a lone swordsman? whats the point?

The point is, it makes the tile free. The AI doesn't know which tiles it needs to reach a critical target and which they don't.
If I made the AI skip these battles, players could surround their cities or whole empire with 1 swordsmen on each tile and never get attacked.

It should still attack only stack it could win on auto-resolve but it should priority greatest threat around them that it could handle first and would take weak one only if no higher priority target for them.

They already do exactly that.

[Suppanut]

Actually, block your city with single cheap units is great strategy that you could do against AI or even player, as it is the way to trade your units for time you need for reinforcement to save the city.

[Anskiy]

But why should the stack of 6 halberdiers attack a lone swordsman? whats the point?

The point is, it makes the tile free. The AI doesn't know which tiles it needs to reach a critical target and which they don't.
If I made the AI skip these battles, players could surround their cities or whole empire with 1 swordsmen on each tile and never get attacked.

That seems.. odd. Can you not make the AI force its way through such obvious blockades? I know for example that rampaging monsters will charge into anything in their path when deciding to attack a city, surely such a state could be implemented for enemy wizard forces too.

[Seravy]

Why would they be obvious?

Finding paths is not part of AI code, it's a gameplay/system feature.
The AI only knows "A way exists to the target" or "A way doesn't exists to the target", the same way when the human player clicks on the tile and their units either move there or not.

Yes, the AI can walk though enemy armies to reach a destination but that is NOT using AI code. It simply goes into the battle blind without any decision like rampaging monsters. It's only activated when the units in question belong to a player currently casting the Spell of Mastery.

A complex system that weights the importance of freeing the tile, the AI's own risk taken based on enemy combat spells, the potential MP wasted on each side and who that benefits more based on income and available mana crystals, and an analysis of expected human player MP spending behaviour for that battle would be necessary to be able to tell apart battles the AI can safely skip from those that they should be doing. This goes way beyond what's reasonably possible to do.