Is A has a p chance of being malicious and B has a q chance of being malicious then the chance of them both being malicious is pq. pq <= p and p*q <= q.
I'm honestly not sure why its so hard for you to understand that TWO people being malicious at the same time is less likely than either being mailicous on their own.
The point is that trusting two is not better than trusting one when both of them have equal chance to be malicious.
> human society if built on the exact same vague fuzzy framework.
May be, but we can try to call a spade a spade and not pretend that something is more trustworthy than it is.