Brotherly Wars and AI

[Repo Man]

The links and downloads are really helping understand the AI politics and anger. If all goes well, and time permitting, I’ll be submitting an article on how the AI politics works, (or at least how I understand them to work). However, based on one day of reading and some preliminary analysis, I think I may have found why brotherly races still fight.

To begin, I have made one assumption; that the anger scale (brotherly to murderous) is the anger scale referred to in the AI texts. If not, then someone please ‘splain.

The most common quirk and complaint on the politics is why brotherly races never make peace. For now, I’m limiting my thoughts to why they never accept peace .. offering will come at a later date.

The accepting treaties algorithms is found in the politics.txt file, and runs from the 3rd line “Highest Allowed Treaty” to the 16th line, “Accept treaty score percent … Protectorate”

The line which determines if an AI accepts a treaty is “Accept Treaty Minimum Anger Chance.” From what I have read, this is the percentage of accepting a treaty if the anger level is low enough to accept that particular treaty. From what I can see, this is a paltry 5 percent. So in theory, it would take 20 turns before a peace treaty was accepted. Perhaps raising this number would remove the quirk of brotherly races staying at war. It would no doubt open up other problems, but one step at a time.

This theory does not explain why AI accept other treaties so readily, since AI accepts trade treaties far more than one out of 20 times. I think that may be found at the end of the AI politics, under the Accept Trade lines. These seem to indicate an automatic acceptance of trade if the trade values fall within certain percentages.

Any insight would be appreciated.