There are times when it's in town's best interest to choose to abandon the majority vote - even abandon voting according to their own reads - in the interest of increasing the chances of them getting scum lims through.
The clearest example of this to me is when the town is significantly behind. Let's say you're in a Mini Normal and you're town. There are 9 players left: 6 town, 3 scum. What actually needs to
happen
in order for scum to get eliminated? Assuming scum don't bus, 5 out of the 6 town have to agree on the same person being scum. That's... not very good! People underestimate how snowbally Mafia can be.But with creative thinking, we can work around those steep odds. If the town can settle on one strong consensus townread, that person can decide upon an elimination target, and everyone can agree
before the target is decided
that they will sheep the consensus townread, regardless of their own reads on the situation. All of a sudden, instead of 5 out of 6 town players having to unite against a wolf while avoiding getting sidetracked by scum, just one player needs to get a good read. The odds are much better for town than playing out that vote normally.That's just one example. There are similar situations when town would benefit significantly from changing the way it decides the elimination. It doesn't even need to be that formal - I think sheeping is incredibly underrated as a legitimate strategy in Mafia. Current site meta trends towards expressing more confidence in your own reads than you actually have and pushing your own unique perspective. But what if you aren't actually confident in your reads? What if your best bet of hitting scum is actually just by following that player who seems really confident that they've caught someone? I think that sort of situation comes up a lot more often than people like to acknowledge.
Basically, town should delegate decision-making much more often than they do. Majority vote is how the game is usually played but is not the most accurate method of hitting scum, especially when town is playing from behind. Players often struggle to put this kind of idea into practice because they overvalue their own individual reads; they would rather vote their own suspects in LyLo at any cost for example, even if their confidence level in those reads doesn't justify it. People need to learn to trust others' reads more often, sometimes even over their own reads. I think a shift in thinking around this topic could result in a general leveling-up of town play.