well, yes and no. We don't want to have 5 deity AIs jump on you because that is clearly an unwinnable position. Especially since AI has tremendous advantage in both production and research. In fact it that was the case, the earliest contact with AI would be deadly for human as AIs could just rush you with earliest units. Thats why you have things like diplomacy so you assume your friend who is pleased with you wont attack you. If every AI would ignore diplomacy like humans do (when I say ignoring I mean deciding to attack AI you have friendly status with because you know you can kill it), then every level where AI has advantage in terms of production and research would be unbeatable basically. Granted in this game we were surrounded with AIs that played exactly like humans would.
I appreciate this response, and the underlying reasoning. I've never even loaded a game at this level, let alone won one, and I've never used an AI better than that of the last official patch. So all of my questions are driven by pure curiosity, and an attempt to learn how the total game can work.
And I've learned a lot from this thread; in particular, it would seem that Blake has explained some aspects of the diplomacy system that most of us never knew (thanks, Blake!). The most pernicious of these are related to the (clunky) vassal system, but some of them have to do with something very close to AI "strategic thinking" -- perhaps better called "deeper tactics" -- in which major decisions are not determined exclusively by the current diplomatic relations. I know some people don't like this, and would rather be able to predict with complete accuracy what any AI's range of options is at any given moment. (E.g., if I get Isabella to Friendly, she'll never attack.) Others may just want more transparency in the diplomatic modifiers as a guide to relations, without necessarily being able to know, absolutely, how this will determine AI actions. When an AI has eyes on your land, and is mobilizing, should you be able to prevent an attack by ticking it to Friendly? For all leaders? I'm on the fence about this, but I'll agree that one of the great improvements in CIV4 over CIV3 is the much greater transparency of game mechanics, especially the all-important diplomacy numbers.
If nothing else, this SG has revealed some of the grey areas in the diplomatic system, and I hope that it suggests some ways to implement changes that make the AI more competitive at the highest levels without massive upgrade advantages, and preserving some distinct leader personalities. I think that the fact that none of the AI executed a massive early rush (as a human would) is a good sign; you guys came really close to pulling this off, which is another good sign! I hope that you take another kick at the can, so that it becomes clearer which game elements most need work.
Sorry if this is OT, but given the intentions of the thread, I hope it isn't.