No.12309724 ViewReplyOriginalReport
This whole 50/50 thing seems very weird.
In a situation where one given persons vote is loosely correlated with any other given persons vote, can we estimate the chance of a 5.000.000 people vote between two options only differ by less than 3000?