The Daily Update at the Com

[Servius]

The initial paragraphs sounded good for CAI and DAI improvements, but the bulletted list didn't. I'm assuming that perhaps I just didn't understand their lingo. In case it's not clear to CA yet, at least as I see it, the Black Knight issue is the most pressing. The core features of the Black Knight DAI that need to change are:

- AI will declare war on you without the means to gain anything from it. Sometimes they can't even reach you. For example, the emergant nation of Hungary declares war on Great Britain, even though Hungary is landlocked, surrounded by nations it's at war with, and has no port or a navy. GB has no continental regions. In this situation, it is physically impossible for Hungary to do anything to GB. Thus, this DOW is pointless. It is a problem of ability. In situations like this, the AI should not have declared war in the first place. In addition, if war is already being waged, and the AI gets put into this position, it should seek peace.

- AI will declare war on you and then not do anything. It has armies, and they could march on your regions, but they don't. The most often experienced example I get like this as GB is the Huron. When the game starts, they attack Rupert's Land. Once you take Huron Territory, they still have that other region way to the northwest. They build up armies there, and could march them down to Rupert's Land, but they don't. This is not a problem of ability, it's a problem of willingness. Again, if the CAI is not willing to attack nations it's at war with, the AI should not declare war in the first place, and if war is already joined and the AI hits this position, it should seek peace.

In short, the DAI should have to pass the following tests before it declares war, and should seek peace whenever it can't meet these tests. In them, "I" is the AI talking to itself.

Test 1) The other nation has something I want.
Test 2) I've tried to obtain the thing I want through diplomacy (trading techs, money, etc.) at least once in the last X turns, but my offer was declined.
Test 3) I have the military means to take what I want, including sufficient, excess (as in, not needed for defense) land forces to take and defend the region I want and sufficient, available naval forces capable of delivering my land forces to the target (if necessary).
Test 4) My CAI will actually use the aforemention land and naval units to execute the forceful capture of what I want.
Test 5) I am currently Hostile towards the other nation.

NOTE TO CA: I don't actually know if there are two different AIs (one for diplomacy and one for campaign unit movement). This assumption is based on situations where a nation will declare war on me, and seemily have the military power and physical access to attack me, but won't. That, to me, seems to be a disconnect between a decision-making AI (diplomatic) and an executive AI (campaign).

Associated DAI issues include:
- Faction disposition towards you provides no reliable guidance about the liklihood they will declare war on you. For example, nations that are Very Friendly towards you, will DOW you. If the dispositions don't reflect the liklihood of DOWs then they're useless. This is addressed in Test 5 above.

- Rejecting AI proposals (like, to trade Region A for techs B and C or something) should lead to worsening dispositions between those two nations.

- The disposition penalites applied to expansion should be somewhat balanced by disposition awards under 'The Enemy of My Enemy is My Friend'. For example, GB and France are at war with Spain. If GB takes Grenada, France won't like GB as much because it fears GB's territorial expansion. That's understandable. But at the same time, France should be happy with GB for weakening Spain. Right now I feel the penality for expansion is far greater than the boost you get through 'The Enemy of My Enemy...'. I think those two should be closer in value.