Anyway, thank you for the constructive criticism. If you have any concrete suggestions for improving the AI, I'd like to hear about those too.
I think one of the major problems now (which according to Mr. Z. is going to be addressed anyway) are the goal files.
I assume that the Provence +3 glory for Austria and the Tyrol +2 for France is only to keep them fighting each other (at least Provence, but even Tyrol was only given to Bavaria in 1805 but not annexed by France). However, I (playing 1792) usually end up with AI-France getting Tyrol and AI-Austria taking Provence in the next peace. Same with Britain and their goal-provinces in northern France: they usually get them pretty early. I think the option of keeping countries independent from enemies should be used much more often as a goal (that's what RL British did; furthermore if this is the goal than a country annexed by the enemy should be liberated in the peace agreement and not annexed by the victorious nation - which has keeping them independent as goal). May be Sea-Provinces could be given some attrition level ar something else, which can only be mastered reasonable with certain naval skills (like a total amount of docks or so, giving the British a huge advantage). The AI should maybe guard supply lines (depots) better. Maybe it's also possible to introduce feudal level as a goal: one goal for all nations should be giving France a higher feudal level modelling the return of the Bourbons. On the other hand revolutionary France should want to give other nations a lower feudal evel.
Such goals should rank higher on the agenda than getting provinces additional to your national core provinces (Prussia occupies Artois, although they have really no interests there).
@malcolm_mccallum
Concerning Eu and Victoria (which are my favourite games): They have their events to keep them somehow on track. And although I liked this idea in the beginning (EU2 was the first strategy game I really played, most others I couldn't even stand the demo), you realize the serious shortcomings after some time (insufficient triggers, advance knowledge). So from this point of view CoG is a completely different game.
AI not caring about their supply lines: Thats also very common in Victoria. Seems to be difficult to teach such things to AI.
So in total I really like CoG and I think those problems can be addressed. Especially games like this can show a lot of different weird outcomings and it might be difficult to find all those during Beta-testing. That's why there is usually a lot of feedback from players and lots of patches after the publishing of such games.
(Besides, what I liked better about Victoria were the more accurate province borders. It was possible to give my country a historical national shape. Bit i could be much worse I have seen screenshots from Cossacks2 and Imperial Glory campaign maps [:@])