AI should exclude irrelevantly weak (subject) empires when evaluating Federations

[Alblaka]

When evaluating whom to ally up with, the AI tends to generate binary YES/NO choices for each empire in a Federation, and will refuse anything beyond NAP/FederationAssociation with a Federation that contains even a single NO choice (aka, any one empire this empire does not agree with).
This suggestion aims at improving the AI by specifically excluding Subject empires (aka, members of the Federation who, maybe unwillingly, were added to the Federation by virtue of being subject to a Federation member) from that consideration. Maybe restricted to only excluding subject empires with less than 20% of their overlords Diplo Weight.

Example case:

I'm the tech-giant Machine Empire up North, having more-or-less-recently elevated my alliance with the Technocrats to the west to Federation status. Another potential applicant is the depicted Materials empire, whom we both get along with and share rivalries with.

Buuuuut, I also (by event) got a one-planet protectorate up in the northern edge of space, and that is the sole reason the depicted empire will not join the Tech Federation (disregard the distance penalty, a year or two of envoy + a bribe will fix that). Ethics match up, political landscape matches up, joining the Federation would give a massive tech benefit and the strongest non-FE empire as ally... but nope, the AI says no because someone somewhere there exists an irrelevant empire that they will never have any meaningful interaction with.

Doesn't seem logical, please fix AI.

 

[C4st1gator]

I also find it strange, that a single unfitting empire would break a federation apart. Federations are more similar to IRL confederations or military alliances, such as NATO and these often have members, who don't necessarily like one another, but still work together. Especially in a large federation there's bound to be members with minor conflicts of interest, but that doesn't mean the whole thing falls apart at the drop of a hat. It would be better, if the algorithm determining federation membership didn't have a bunch of K.O. criteria, but would weigh the criteria against each other.

For instance: Let's assume a federation has formed to respond to a hostile power. An unlucky spawn allowed a Fanatic Purifier to take a quarter of the galaxy relatively unopposed. We're most likely talking about a human player Fanatic Purifier, who has a stable empire, that isn't running a deficit and will not collaps on its own.
The AI should not get the suicidal thought of leaving the federation, because it should be far more concerned about the purging empire, which has against all odds become a galaxy level threat.

Granted, at that point the galactic community should probably declare that player a Crisis whether or not the Fanatic Purifier empire has the corresponding ascension perk, since it would be a secondary concern to the Galactic Community at large whether the galaxy dies or whether it gets purified.

 

[GOLANX]

It should be possible to use the subjects policy to remove your protectorate from your federation, you could also choose to integrate them or release them should solve that problem too.