It "feels like" at first touch Proportional approval. But here's the thing...
You use the term "pairwise matrix" with cells = 0 or 1 depending upon whether the alternative is approved or not. But the approval status is by voter (= by ballot), so are you suggesting a pairwise matrix for each ballot? That seems unnecessary. If you mean cell i,j is the count of ballots where i is ranked better than j, it is the same as various other methods as far as constructing the alternative pairwise-matrix is concerned. > -----Original Message----- > From: [EMAIL PROTECTED] > [mailto:[EMAIL PROTECTED] > ] On Behalf Of Russ Paielli > Sent: Saturday, March 26, 2005 2:49 PM > To: [EMAIL PROTECTED] > Subject: [EM] Pairwise Approval Voting > > Hi Folks, > > Let me define a variation of Approval Voting with the same voting > procedure as standard Approval Voting but a different tally > procedure. > I'll call it Pairwise Approval Voting (PAV). > > The tally procedure starts with the construction of a pairwise matrix > from the ordinal information in each vote. That is, if candidate i is > approved and candidate j is unapproved, then the (i,j) element gets a > one and the (j,i) element gets a zero. The "regular" approval > counts go > on the diagonal. The CW wins if one exists, otherwise the > least-approved > candidate is dropped until a CW is found. > > Has this or a similar method been proposed before? > > Do you consider it better than, worse than, or equivalent to, > standard > Approval Voting? Why? > > Russ > ---- > Election-methods mailing list - see http://electorama.com/em > for list info > ---- Election-methods mailing list - see http://electorama.com/em for list info