In my opinion the land bridges are a requirement. Many of the modders who remove land bridges do so for the wrong reasons. To understand the problem with landbridge removal you first have to understand how the AI sees shipping and how invasions actually work.
A line of vessels in adjacent sea regions, linking one province to another allows the movement of army stacks between these provinces. If the destination is an enemy province, an invasion is possible if a blockage does not exist. The AI knows nothing of this. The AI only knows that movement/invasion is possible when the fleets are in the right places at the right time. So for example, if a route exists from Scotland to Palestine, the AI will see a neighbouring province and will assume that this province will always be there. This is why the AI makes risky invasions with it's king. When the route is cut off - which the AI faction can easily do by itself as it moves ships around - Palestine will be isolated (with the usual results). The same goes for Ireland and the Mediterranean isles (everyone has seen the byzzie emperor marooned on crete or rhodes at least once?).
So instead of cutting off landbridges, creating them is actually the better option. I usually link Ireland to Wales, Cyprus to Antioch and/or Lesser Armenia, Rhodes to Nicaea and Crete to Greece. If anything it gives more useful provinces. Landbridges also help AI crusades. It means that there are no regions which cannot be reached by land, but does not decrease the value of fleets in the slightest.
In the past we looked at different solutions to "silly invasions" problem, one of which was suggested by gollum, was to break the "ocean zone" into three larger sea zones, with the usual sea regions within these. The idea was to decrease mobility and reduce the silly invasions. As with any mods to the game, this had "side effects" and created more problems than it solved. I'm nut sure if gollum included it in his mod (I think not).
I wasn't suggesting my opinion was "the right one", just giving an opinion, which is all anything posted on a messageboard is. ;)
The border forts vs spies issue is another paradox. For years I used to remove border forts, but the AI still does not counterspy effectively and falls victim to spies and assassins far too easily. The problem with leaving border forts is that they act as the primary counterspy, meaning that your agents in the province do nothing and don't gain valour from catching rival agents. You can just not build border forts and let the AI continue building them - this is what I used to do and seems like the best compromise.
With them border forts in the game, rival assassins/spies under about valour 5 stand next to no chance of survival. Which means that you have to micromanage agents, call off their missions etc - that's no fun, in fact it's a pain in the arse, it's an even bigger pain when an agent you've spent time and effort training up gets caught by a border fort. It all boils down to tedious micromanagement, which the AI has no hope of being able to do.
As I have said in the previous post, the AI also sends his own agents against his own generals/agents. I have seen assassins targeting other assassins and spies, which is not a huge problem, but also their own factions generals - generals with good stats and V&Vs... inquisitors can behave similarly.
There are also agent functions which the AI cannot use at all or uses very seldom. AI spies in particular very seldom use any of their abilities and will often just sit in the province where they were trained. Unlike the overpowered shinobi in STW, they're also not very effective at instigating revolts as their effects do not stack. The AI is not clever enough to get 20 zero valour spies and one 4 valour, drop them in an already close to revolt rival factions province - without a border fort - and reap the rewards. I have only once seen an AI spy conducting a treason trial (before that I believed they didn't do it at all) on a very high loyalty general...
Bookmarks