AI behavior : religions, research, alignment

[MagisterCultuum]

Religion should be important, but alignment should be more important in deciding what techs to choose; especially since this would direct civs towards founding religions appropriate for their alignment, and limit the spread of religions in civs with different alignments. With large alignment weights, it may be possible for both the Python code forcing religious choices and the inability to trade techs for religions that would change alignments to be removed without any ill effects. Also, remember that some leaders won't ever adopt a religion (Cassiel, any new agnostic trait ones you may mod in, and Auric Ulvin even though he doesn't currently have the trait), and so would never gain the modifier based on religion (Cassiel really shouldn't be rushing for Malevolent Designs so much).

 

[Lord Yanaek]

And Auric shouldn't found religions as i often saw him do (sometimes he founds two religions on a total of 3 possible as runes always go to the dwarves and leaves to the elves in a single game)

 

[MagisterCultuum]

Yeah, he should probably just be given the agnostic trait again. The inability to discover religious techs or build their disciples was changed in .25 to depend of the trait, not the civ (changed to allow for more option with unrestricted leaders), so there is nothing stopping his from wasting his research on founding religions (of course, a human playing as Auric could do the same thing and it wouldn't be wasteful, since a human could convert but Auric's AI still refuses)

 

[Nikis-Knight]

This (changing the value of a technology depending on if it can still found a religion) is done (since Vanilla Civ), save in the couple of cases where the FfH II python code presently forces a religion choice. Whether or not the change is large enough is another issue. There are, in my opinion, 2 major problems in tech choices in FfH II:
1) The base weights for technologies are sub-optimal.
2) The situational modifiers for technologies are bad, and in many cases, missing.

From what I am seeing the ai is not bad at either building advanced units or building lots of units. So I think better directing their research is probably the ai tinkering that will get results. (protecting upgraded units such as heroes and choosing promotions important tactical considerations)