Better AI nuke target selection - should prioritize strongest enemy over puppets and weak countries

[Celestial_Khagan]

I'm currently in a game that is essentially WW3, Allies vs Soviets, and the US keeps nuking Poland instead of the Soviet Union. Which is not helpful or smart, since Poland is just a weak puppet, whereas the Soviet Union is the main threat.

Furthermore, the AI should prioritize national provinces of the strongest enemy over others. Non-national provinces are targets for liberation, and don't contribute to the ICs of the enemy country much anyway. This is both better game-play strategy-wise, as well as makes more sense realistically.

If there aren't good targets, it would have been smarter to save the nukes for later than to use it as well.

 

[MartinBG]

The AI already prioritizes national provinces with VPs or high production, but there are other factors like targeting provinces with high presence of enemy units. Valuable SOV provinces could be out of range for USA bombers or the path to there could be too dangerous, so the AI picks an easy target instead.

 

[Celestial_Khagan]

The AI already prioritizes national provinces with VPs or high production, but there are other factors like targeting provinces with high presence of enemy units. Valuable SOV provinces could be out of range for USA bombers or the path to there could be too dangerous, so the AI picks an easy target instead.

In my games, the US nuked Budapest Hungary, and Breslau and Warsaw Poland, whom belong to low IC puppets of the Soviet Union. They aren't at the front and don't have troops in them, I know why they nuke those, it's because the provinces individually have relatively high ICs. The problem is these are poor strategic decisions. Hungary and Poland only had about 10 divisions and have less than 100 ICs, while the Soviets have 300+ divisions and 400+ ICs. Nuking their weak puppets does nothing to weaken the Soviets and help the war situation for the AI.

Nuking provinces with large number of enemy units would have been a much better decision if they did that, but they didn't. The front was near the France / Germany border, there were provinces that had 50+ Soviet divisions in them that were a major threat, but the US AI didn't nuke those.

The other problem is that every strategic bomber has escort fighters attached to them, which halved their range and greatly reduced the possible targets the AI has.

If I was the AI, I would prioritize either high value targets such as Moscow, or provinces with large number of enemy divisions first. Puppets and low IC countries should be deprioritized. If the AI can't find a good target, it's better to save the nukes for a better opportunity than to waste it on targets that doesn't help them. As of right now, all I saw was the AI wasting nuke after nuke.