I'm looking to integrate higher strategic functionality into XConq, a hex-based strategy engine. It doesn't have a political system, yet, but we've got a decent roadmap in place for that. What I'm interested in is the simulation of something that I haven't seen in classic 4X games: national disintegration. So this is the manner I'm looking to deal with it (I've never worked on AI-related things, so I'm sure it's pretty simple): (As a note, this is originally posted on xconq.org's forums, and so refers to a few things that are specific to the code, my apologies for lack of clarity) Step 1 - Determine Variance XConq would poll a side's units based on a new unit property: political-value. As a designer, I would assign political-value to advanced units, with higher political-value scores given to larger or more important cities. The variance between the highest political-value and the lowest political-value is the varience. This is checked against a new global property: minimum-varience-for-instability. The idea here is to simulate the effect of a nation growing into two seperate nations. While the initial check would only be for political-value, later on the system could be amended to check for some kind of cultural dissimilarity. My point in creating this kind of system is to create a proper simulator of civilizations, wherein the nations change and adapt and split up rather than simply grow and grow and grow (Unless eaten by another nation). Step 2 - Bloc Segregation If the variance is high enough, then begin grouping units together based on similar variance and proximity. These blocs would then check against another new global property - minimum-strength-to-rebel. A new uu-table property: rebel-recursively, would allow this system to determine what units would rebel along with the political-value units being considered. This table would be percentile-based, so that there would be some riskiness in the calculation and therefore some variety in the rebellions. Somehow, the minimum-strength-to-rebel would then serve as a base-chance of rebellion, with any additional strength increasing that chance. Step 3 - Initialize New Side Ultimately, this rebellion would result in spawning a new side. That's another feature for another time. Initially, it will only render the units independent. A possibility for the rebellion to spread should be factored in. Also, once a political system is in place, the determination of new political stances should come into play. As an interesting feature, I think if this happens to a human player, that player should be given the option to pick which side they want to play, so that they can be the rebels or the patriots. For specific side AIs, the original AI would be more likely to be in charge of the larger nation. I think this might end up being a rather elegant way to simulate national disintegration. As it's described here, the main cause of rebellion would be for one geographic area of the nation to significantly lag behind or leap ahead of another (in economic/political power terms). While this isn't exactly the case as in, say, the American Civil War, it does in a way resemble the splintering of the Roman Empire as it focused more and more eastward. Thoughts?