Firstly, to aelf: you're dead wrong. I am not somebody who spams military units and ignores the diplomatic aspect of the game. In fact, it's the exact opposite. I find the military aspect of the game to be both the most tedious, and the most skewed in the player's favor. Thus I generally avoid it as much as I can. So your assumptions about my personality based on our disagreements are ill-founded.
combat odds are within the rules of the game. Circumventing the diplomatic rules ... while it does exist in a very limited form in the game ... would just be undermining an aspect of the game.
The best way I could explain my argument, is to have you imagine that diplomatic "odds" were displayed in the game. Imagine that when you mouse over another civ's name, a percent chance to declare war was displayed. For a friendly civ, the "Chance to declare war per turn" might be 1%. This is no different than the combat odds.
Even if unit X versus unit Y displayed a 99% chance for victory, given enough battles unit X would eventually lose. Similarly, even though you are
almost guaranteed to never have a war with a friendly civ, there is always a small chance that war will be declared.
Further, even if you took these percentages, on their own, to be less than 100%, that still doesn't include contingencies that could occur on each turn - e.g. a rival bribes that civ to attack, that civ wants a resource that you have, that civ feels boxed in and wants to expand, et cetera.
And we can assume that that percentage is always less than 100%, otherwise diplomacy would be the be-all-end-all of the game. All you would have to do is keep a positive Diplo modifier and you would never have to worry about an invasion. Does that seem balanced or challenging? It doesn't to me.
So I don't see a big difference between this and combat odds. In both cases, you can do a great deal to "stack the deck" in your favor,
but there are no guarantees. And, again, things are even less predictable in the case of diplomacy, because there are many more random elements than in the case of combat odds, where you know all the variables in advance.
This is mostly conjecture on my part, because I'm not a programmer and I don't know how the AI was programmed. But clearly a friendly disposition is not a guarantee of peace, period. With that bit of information, I'm going out on a limb and suggesting that diplomacy and declarations of war are subject to the same variance as every other game mechanic. The complainers seem to think diplomacy should be
unique in being a guarantee against aggression from select civs, but there is no reason to have that expectation.