Some questions about the AI...

[LintMan]

I have a scattering of questions about the AI and (some tangents) that I'm curious about and maybe you guys have some insight on...

1) Does the AI "cheat" at all? (ie: omniscient view, bonus income/units, etc) What effects do the various difficulty settings have on the AI? Is there any way to tell what level the assorted AI's are set to in a game? (Aside - what difficulty level are the AI's set to in the Walkthru game?)

2) What makes AI decide to attack or hold off attacking you? What are its criteria for doing so? In my current game, I've shared a long, peaceful border with Pangaea for many turns. Then, it attacked me unprovoked and started a war. I later reverted to an earlier saved game that happened to be from just a turn or two before that attack, and despite my actions largely being the same as before, many few turns later, we are still at peace. Along the same lines, what does the AI consider "provocation" for an all-out attack? (ie: is putting temples/preachers along the border offensive? What about peaceful global effect spells (ie: Gift of Health)? How about finding a scout of yours in their territory? What if you take one/a few of their undefended border provinces?

3) In some experimentation, it seemed like several of the AI's were actually frequently entering its Pretenders into the Deathmatch Arena contest! Is this common? Does the AI frequently enter its pretenders into the deathmatch event? That seems potentially a very rash move that risks afflictions/death and at best saddles your pretender with a (albeit powerful) cursed item. (As an aside - what triggers the deathmatch events - is it purely random, or is it X number of years?)

4) Does the AI handle/adjust/adapt to game settings like high/low magic site percentage, strong/weak independent forces, map size vs number of nations, etc, or do these throw off its game plan? I guess my biggest two questions here are:
a) Could adjusting the independent force strength to be high/low have the effect of crippling the AI game (if it can't properly adjust to these conditions)?
b) Can the AI remain competitive in a game where there is a fairly large period of "peaceful" expansion and research before nations come into direct conflict? i.e: a large map, with not a large number of pretenders. I ask this not because I'm some hotshot awesome player (definitely not - I'm still learning), but because this is my favorite mode of game, but it is a common weakness in the strategy games I've played, and usually results in disappointment. I'm thinking specifically of my game of Space Empires IV, where I played an epic game in a huge galaxy with only a relatively few races. In the late game, I came to discover that *all* those other races had simply failed to expand at all, and were hopelessly behind in economy, territory, and research by the time I encountered them.

5) How likely is the AI to try to dispel a non-offensive global enchantment like Gift of Health, Mother Oak, or Forge of the Ancients? How many extra gems are needed? If I can only manange the spell's basic gem cost, is it even worth casting these type spells, or will they go down quickly?

6) Does the AI forge magic items and equip them intelligently on its commanders? Does it target its research to its strengths (ie: avoiding spending research on spell levels it has no use for (if it doesn't have the proper mage types)). Will it empower and equip mages to raise their magic paths to have access to more powerful spells?