While it’s more nuanced and less exact, it’s still their
The former’s social. The latter’s statistical. They aren’t linked.
As for keeping the ‘hard to read’ player alive, I quote;
1 Like
other way around
if YBW is maf RC, Zone is VT
if Zone is not a VT, YBW is not maf RC (but can still be town or maf goon)
Alright I’m gonna rephrase the scenario and see if it changes your answer if that’s alright
Rephrased scenario
Player A has a 20% chance of being mafia, You can correctly read them 25% of the time.
Player B has a 20% chance of being mafia, You can correctly read them 80% of the time
Who do you execute
Those aren’t exactly statistics even with those numbers, but Baye’s Theorem. It’s similar, so I’m going to assume that’s your point.
Even so, it doesn’t change anything in my eyes. A here is ‘hard to read’, and B is ‘easy to read’.
You’re 55% more likely to read Player B correctly, why dosent that change anything
Because that’s still social, not statistical. Bayes’ Theorem is a form of ‘conditional probability’. There’s no proper way of getting exact numbers, like multiplying thousands of numbers together in your head. You can’t do it, really nobody can do it, so they come up with a simplified version. As well…
They don’t, but they still need to be accurate. How do you quantify how often you can deduct a person’s role in a social deduction?