The larger point is that Attrition doesn't work more realistically on the SIEGED unit. Paradox, for far too long, has this inane system in place that slow-walks attrition for the force INSIDE the walls of the castle. Those are the people who suffered, who resorted to cannibalism, who died of thirst (can only live 4 days on average without water). So, yeah - attrition isn't working both ways. We're used to this status quo system because it's roughly the same since EU3 and we are all a bit bear-in-a-cage about it, when playing an EU series game, but this attrition issue is more severe for the lack of losses to the castle, than to the sieging army.
The other issue is that we don't represent logistics "tail" following the armies, or being pushed to arrive after the siege starts. Look at what Xerxes did to siege - thousands of cargo "wagons" that went with his armies. The actual percentage attrition is NOT fixed in the real world, and shouldn't be factored from "where" you are sieging, but instead factored based on whether or not you spent the money/used your power to ensure the logistics tail was good enough to keep your attrition down. It's more dependent on the Attacker than the Defender (and defender's location) in the real world. See - Mariupol, Ukraine, right now, for a real world example of a siege in the modern day (hate to table that example, but - the Attackers are well supplied from Black Sea logistics support, so it's the receiving end in Mariupol who suffers attrition).
This is just another mismatch with the real world that Paradox Game Designers and Dev's need to improve upon for an EU5 build.
