[Baron Munchausen]

The problem that everyone is dancing aorund here is the simplistic AI. It's got a single linear scale of how much it 'likes' or 'dislikes' you and some randomizing factors that make it jump the tracks occasionally. Not at all realistic.

At the very least we need an AI with two 'scales' one of "like/dislike" and another of "fear/non-fear". So, if you are really powerful you might be offered a treaty by the AI simply because they are afraid of you. This sort of treaty is likely to go 'poof' if your own military strength suffers a setback, of course. But if you can see the two Ratings (they ought to both be available just as the single 'attitude' is now) you can guess why you have been offered the treaty and plan accordingly.

Ideally, of course, we ought to have some sort of 'cultural ties' that also grow stronger with time and make good alliances (not 'fear' alliances) difficult to just throw away. Even when war breaks out between two empires with a long history of good relations the AI ought to be smart enough to try to keep it a 'limited' war for just one particular objective, not turn it into an instant campaign of genocide.

Of course, there also ought to be completely different types of treaties with widely varying terms. The rigid linear scale of treaties is also annoying.

[ November 01, 2003, 19:25: Message edited by: Baron Munchausen ]