I'm sorry you feel that way, Lohrenswald.
The set-up:
One bloodthirsty vig based on item
One one-shot town thief (Cuthillius). His steal, if used, was the first thing that happened. A failed theft revealed the thief to their target. This was a probably unnecessary (given the scum's weakness) hedge against a day one lynch of snerk, since the sword, while it nominally would go up for vote if its holder was lynched, would invariably be stolen in the night..
Two scum. They could either scan one person for the sword, or individually steal (their steal happened at end of night). A failed theft revealed the thief to their target. They also automatically won if they ever held the sword for two day phases (Didn't have to be consecutive). This also probably wasn't needed, since it was unlikely they would go two days and not be at the win threshold by numbers.
When I set out to design this game, I had a few things in mind:
My bete noire has always been designing a kill-less scum team that was suitably powered. I don't think they were, but they won, so I thinK I'm going to move the goalposts and declare victory.
In a Sidar game, I will always have a post-lynch participation feature, so that was never in doubt. I considered more complicated or direct methods, but decided talking was sufficient.
I wasn't going to have confirmed town participants, so no role-reveals on lynch followed automatically. I decided to keep role-reveals on NK, as a slight boon to the town and hint that it was plausible that a scum could be the target of the NK.
I thought about giving the scum more power, but on this site, no role-reveals tends to throw the town for a loop, so I called it good.
If I had it to do over, I might have let each scum scan separately, but overall, I'm actually pretty satisfied with the balance and the outcome.