Storm Raven said:
We aren't talking about discouraging adventurers. We are talking about discouraging murderous cutthroats. The PCs in question have engaged in murder of people under the authority of the government multiple times already. I think that just about any government is going to want to discourage people like that from hanging around. There are probably other adventurers out there who are willing to forego the pleasure of going on murder sprees, and those are going to be the ones desired and rewarded.
Heck, those are the adventurers who will probably be hunting down the murderous cutthroats.
As I said, it depends. If you have a large population of civic-minded adventurers, the value of murderous cutthroats who mostly kill demons and orcs drops dramatically. On the other hand, I see little reason to assume a sufficient population of civic adventurers. Most of all, I see no reason to posit that a particular scenario will necessarily generate adventurers to fill it; if the world has a large active adventuring population, then they should be constantly aware of each other and sometimes stepping on each others toes. It seems to me that if there is a group of adventurers who are liable to come after the group who killed the Arbiter-person in the OP, they should be well-known and previously integrated into the setting, with their actions clearly visible. To do otherwise smacks of "This scenario requires characters who will do what I want. If PCs don't, then NPCs will."
Again, I'm not questioning that governments would want to limit the actions of high-level murderous thugs. I just question their ability to do so in most D&D universes.
On the DM-enjoyment aspect: DMs are just as much entitled to fun as players. I don't buy the increased-workload-means-increased-fun-weight argument; if I didn't enjoy DMing and worldbuilding, I wouldn't do it. That being said, coming up with a simulationistic world and then finding out how to cleverly pull genre expectations out of neutral rules is great fun, to me.
My view is that, as in every cooperative endeavor, compromise is necessary. Players need to balance their choice of actions against the demonstrated in-world consequences, and DMs need to sometimes look for edge cases to explain why the expected, unfun result doesn't happen. The party is captured by a ruthless villain, say. The predicted, simulationistic, and unfun response is "You are slain before you awake. Game over." This is not an ideal outcome. So, you need to explain why this hasn't happened in the context of the world. No one but an idiot wouldn't kill the party outright? OK, that means that the villain necessarily is an idiot. He's been shown to not have been an idiot in the past? OK, then something changed between then and now. Perhaps he was being advised by one of the PCs diabolical mastermind enemies earlier on, has since been backstabbed and is now working against that enemy, has rejected much of said enemy's previous (good) advice as obviously designed to lead him into bad decisions, and along with it, has decided that rather than killing the PCs, he's going to keep them alive but imprisoned to thwart the enemy's plans.
Simulationism is a two-edged sword; you need to not only enforce the rules consistently, but populate your world cleverly. Elements in the game world do not have necessary existence, if an evil overlord or a system of laws exists, it's because you chose to put it into the world, presumably because you thought it would be interesting for the PCs to interact with. If the simulationist rules of the world indicate that the existence of demon lords will result in an unfun campaign setting, you don't introduce demon lords as written.
