It needs to be possible to remove a player from any location, and thus it needs to be possible to force fundamental conflict.
So that would mean that the faction location and structure should be more dynamic. That would mean you need to put less emphasis on structure Q as it will be destroyed during clashes and make it easier to build them especially defensive structures.
War and siege should be a collection of informal, localized states, rather than formal and global. Formal states create distinct interfaces between states that are prone to manipulations, workarounds and exploits. The game devolves to a game about those states, rather than a game about siege warfare.
I do not think there is any form of formality within the conflicts that we have at the moment especially when the target of attacks are actually randomized.
That players actually make progress in a siege is not too important. What is important is the ongoing threat of being able to do so if not prevented or counteracted, and the ability to always/often act toward furthering this goal. As long as both sides have to/are able to continously take meaningful action on account of the siege, it does not matter whether either side actually breaks the other, but in fact a protracted siege is highly desirable. This in itself creates activity and a sense of action and purpose.
One thing I am slightly concerned with is that people that haven't subbed will be forced to use all of their available time to fend off a siege just to lose because the attacker that are subbed can essentially win the game of attrition, unless this is the intention.
Losing should ideally be fun. If the process of losing is drawn out over a longer period of time, it can potentially be fun.
(unless I misunderstood something) This is where I disagree with the most. This isn't the same as Dwarf Fortress, your adversary isn't some random factor that the computer come up with, nor some chain of event causing some drama within your creation; your adversaries are people. In some instances some of these conflicts are caused purely by vendetta and mostly of the time, the loser isn't really having fun when the winner is smearing it in their face for all its worth. The other issue with losing is that it sets you back tremendously while your adversaries advanced if they manage to seize your resource, while your character is now a shiny trout in a tank full of barracudas; this is the nature of a permadeath survival multiplayer game. This is regardless of the fact whether a siege is blitzkrieg or prolonged.
Sieges should be subject to a great deal of inertia.
Definitely, the degree of winner/loser is basically a boolean at the moment but getting that to work is pretty tricky unless one can somehow localize the damage that an attacker can do while the defender has to regain that back. This also should reduce (not eliminate) the viability of casual raids where you can simply faceroll over a few villages and get easy loot.
This however runs across into one problem. What if someone decides to make an onion layers of walls or the traditional "Sodom" plot style city layout? The attacker will end up attacking through layers of layers of brick walls and not actually get anything from it until they hit a good spot. It's digging mounds and mounds of dirt to try to find gold, except the dirt is shooting arrows at you and trying to stab you.
I want to bring one thing into this as we speak. Most structure in the game are pretty flimsy compare to the defensive ones. It does make sense that defensive structure should be tough to drive back invaders but when you compare the important ones to a gate or wall, Big Bad Wolf doesn't need to to huff and puff to blow it down, he can just sneeze.
