[Arryn]

Originally posted by Boron:
especially when being ermor / pan cw you don't care about the population anyway.

True. But playing Ermor is so much different than playing the other, living nations it's almost like you're playing a different game. Certainly different rules apply. I should have been more specific so that people won't nit-pick the basic argument, especially given how many people on this forum play as Ermor.

If you play Ermor, disregard the rest of this post.

so even if you don't conquer the province you do serious harm to your opponent.

As I said, I generally avoid attacking unless I have a fair degree of confidence I'll win. I'm not big on Pyrrhic victories, either. Besieging a castle that I don't expect to take, with the intent of "seriously harming" my enemy falls into the Category of "raiding", and I already said that I do tax at 200% when raiding. Generally, though, if I have enough force to crush the local defenders and reach the castle walls, then I have enough force to take the castle too, without a long siege. (I consider a long siege anything over 2-3 months.) It's not worth it to me to screw up a province's unrest for battles I expect to win, and provinces I want to keep.

but unless you are 100% sure you conquer the province it is always a good idea .
you have slightly reduced income / population later and to patrol or lower taxes for the following turns if you conquer the castle but that's not a bad deal for the nice extra money you get and the trouble your enemy gets through taxing .

The income is not "slightly reduced". That's bad assumption #1. 200% tax very quickly raises unrest, which very quickly lowers income. A more than 50% loss of income (after just 1 turn) is not a "slight" matter.

Population in Dom 2 dies much easier than it can be replaced. Bad assumption #2 is thinking that losing a little population can be ignored. It might not hurt much, but it's still an income loss. Losses add up. About the only time pop loss can be ignored (to a degree) is if you have growth scales, and most players don't take growth scales.

Patrolling ties up troops that can usually be used to better effect doing something else (like conquering other provinces). Having to waste time and resources (troops) to patrol away all that excess unrest is bad assumption #3. The long-term cost of not having those troops doing something else is almost always worse.

It takes longer to reduce unrest via lowering taxes than it took for the unrest to go up when you increased taxes. The lowered income for the next several turns is always more than the income you gained in the one turn in which you "got rich quick". It's sort of like betting against the house in a casino. You always lose in the long run. So assumption #4 is thinking you're costing your enemy more than you're costing yourself. Not true if you expect to be the victor in the battle.

Now if you have troops that have nothing better to do than patrol, AND gold isn't all that important to you (for some odd reason), then what I've said above is moot.