Craig Carey wrote: > Regarding the definition of "one vote per voter", it can have > 3 aspects: > (1) the vote (with weight 1) shan't have an effective weight > that is greater than 1, and > (2) the vote shan't have an effect consistent with it having > a negative weight [i.e. monotonic]. and > (3) subject to all else, a vote of weight 1 ought have an > effective weight that is near 1. > > The First Past the Post method passes (1) and (2) and it is not > so good on (3) if it has STV papers. The IRV method is easily > rejected for failing (1) and (2). Mike's Approval method novaes > in a dim dirty flash over rule (1). Great. - I'll normalise my "effective weights" such that an effective weight of one corresponds to the approval vote which approves half the candidates. That way, a vote of weight one will have an "effective weight" of between zero and one, and hence pass (1) and (2). What do I win?
