Bugfix : Siege progress from assaults was not scaled up due to the recent changes. Didn't save the txt before uploading..

It is now, though I also decided to nerf it to 75% of what it was, effectively speaking, because of the bigger Sane Warfare levy/retinue numbers, and the proportionally-not-that-much-bigger garrison sizes.

Siege Tweaks

- Siege progress and fort defense, alongside all their sources, were scaled up an order of magnitude (i.e forts have 1000 defence + 500 per level beyond the first, rather than 100/50, and base siege progress is +10, rather than +1).

- Non-cavalry Men-at-Arms now give a little bit of siege power - 0.2 per regiment for heavy infantry and pikemen, and 0.1 for archers and skirmishers, except some culturally strong variants - longbowmen, chu-ko-nu and abudrar, who also get 0.2. It doesn't show in the stat screen.

- Nerfed actual siege engines' siege power (relative to the new scaled up values).
Bombards by 33%, Trebuchets by 40%, Mangonels by 20% and Onagers by 16.6%. Do note that 1 bombard regiment is still worth 50 heavy infantry regiments siege power-wise, so its not like they're useless. Reasoning : Sieges were over too fast when actual artillery got involved. Like 3 months for max level forts fast.

- AI will devote 160% of the garrison to siege tasks, up from 150%, since fort levels are higher in the mod, and that causes more attrition, and AI is/was always bad at getting a siege to start again when its besiegers attrition to under the garrison size, causing AI realms to get stuck sieging objectives for months to years longer than it should take.

- "Nerfed" sappers strategy perk. It now only gives one-fifth as much siege progress per men-at-arms regiment. Reasoning : It used to give a lot, and retinues / mercenaries are a much more common sight in Sane Warfare. This led to lategame armies supported by mercs getting extra siege power equal to 40 bombard regiments (i.e 400 bombards) from men-at-arms, leading to sieges of level 40+ castles (the REALLY tough ones) being over in 2 months if the ruler just so happened to have it unlocked.

- Base siege phase length increased (20->30 days). Had too much random impact.

- Tweaked AI % desire of men-at-arms regiments as siege machine regiments, so that if an AI realm can afford full cap men-at-arms lategame, it finishes up with 9 full size 20 regiments, rather than 7 full size 20 regiments of men-at-arms and one size 20 + one size 6 regiment of bombards.

These changes help the AI siege, since it is currently unable to actually devote its siege regiments to... well, sieging. It literally tasks which soldiers siege at random, so the bombards often just sit there in the next barony while the lategame hillforts are getting sieged for five+ years while it eats attrition.

Economy update next

A bunch of number tweaks so AI builds more thematic retinues and is better at economy :

- Generic light infantry from 17/17 10/16 -> 15/15 damage/toughness and 15/15 screen/pursuit. They have pursuit and screen, which bowmen don't, and are quite a bit cheaper to maintain, both of which are advantages I neglected to take into much consideration.

- Pikemen (and their cultural variants) got 20 screen (was 0). Reasons obvious.

- Generic Light Cavalry lost 2 toughness, since it was still a bit too good with its 50 pursuit 30 screen (now 28/20/50/30, from 28/22/50/30).

- Slightly reduced AI weighting for the pursuit stat, which should result in it deciding to use retinues based around light horsemen/light infantry less (unless thematically appropriate).

- AI will now try to budget for upkeeping Men-at-Arms equal in their unraised maintenance to 40% of its income (120% if raised), rather than 50% of its income (150% if raised). It will also desire to keep a war chest of 3-years income, rather than 2. This should make it go bankrupt less (it did that frequently).

- Converting to local culture no longer costs any prestige, because in the (rather plausible) scenario that an AI character who is out of prestige is restructuring their retinue, they might take a very long time to be able to convert, and we can't have that.

The next patch or the one after will adress the economy. I'm still gathering data for now, and based on observer games, the AI empires do seem to (on average, not always) have a bit too much gold. Not by that much, but too much nonetheless. So I will see where that goes.