[Edratman]

This thread has some interesting depth and ideas in it. If (big word), it can be determined which pretenders are optimal for the AI, that has the potential to be an additional level of difficulty, ie. a best pretender level, and I see the opportunity for another super impossible level where the AI gets best pretender and two provinces with castles and maybe an extra 50 or so scale points.

I made pretenders for the AI, but they didn't perform as well as the randoms did. Of course, as I realize now, I was making pretenders that would jibe with my play style. This weekend I'm going to try immobile pretenders, good/great scales and throw the balance into rainbow magic.

Gandalf clued me into the low resource game several weeks ago and that is all I play now (resource 50%). It plays quite differently than the high resource games I used to play. The AI is certainly less self destructive because it is not fielding armies with several hundred chaff. I assume that the AI decision tree uses number of units as a key decision factor, so it seems to play a bit better. It definately builds more castles and better units, on the whole. On the other hand, it does seem to be less aggressive, probably because it doesn't have as many units. In my current game with half the AI at level 3, half at level 4 (difficult and mighty?) I am getting a good run for my money. (I am not a superstar player, just someone muddling through.)