Post-Patch (2.08) "Improved Spellcasting AI" discussion

[tinkthank]

I was the most excited about the "Improved Spellcasting AI" note on 2.08 and downloaded immediately.

I have only tested 2 nations with various scenarious since 2.08, and I have noticed following spellcasting improvements:

- AI mages cast Breath of Winter less often
- AI mage targeting slightly better
- AI mages cast defensive boosts (shields etc.) less often as battle continues

But I have not yet noticed much other AI improvements. Who has? Or: Can we get some details together on what has improved?

I have noticed

- Pythium Comm. Masters will *still* kill their slaves by casting spells such as "phantasmal warriors" when *independent* armies are *already routed*
- dual/triple/sextuple redundant buff castings (blessing, etc.)
- casting poison resistance / cold resistance when fighting no poison-inducing (cold inducing)troops or mages (indy archers and militias)
- casting totally useless spells (lone Theurg casting communion master on himself, them communion slave)
- prevalence for casting summoning spells no matter what the situation; some spells *never* get cast (such as orb lightning!)

Things I would like to see:
- increase value of casting summoning spells when enemy outnumbers you, otherwise not
- decrease value of casting high-fatigue spells on routing enemies
- decrease value of casting low-precision ranged and area-effect spells on routing enemies
- increase value of casting positive buffs during the beginning of battle
- decrease value of casting beneficial debuffs (like "sermon of courage") at the beginning of battle
- decrease the value exponentially of casting positive buffs when one is already cast
- decrease /increase value of specific "type only" defensive buffs depending on enemies (fire resistance vs. barkskin, etc.)

What do the others think?

[ February 17, 2004, 08:18: Message edited by: tinkthank ]