> C(A)>E(B)=H(B)>F(C)>A(E)|>D?G>I(F)>B(H) > We would just need to explain to voters what an > absence of grade means...
However, if there is only one non-voted candidate then a ballot using '?' will not identify erm. ------ Matt, yes there will be no ? identifying the ungraded candidate, but we do not care because we explained previously to voters if we were to place him just after the cut-off... Thus voting C(A)>E(B)=H(B)>F(C)>A(E)|D(?)>I(G)>B(H) or C(A)>E(B)=H(B)>F(C)>A(E)|D(F)>I(G)>B(H) is equivalent to me. If you really want to register how it was marked you can use the first notation that explicitly puts D(?) like we could write ... | D(?) = G(?) >... I can leave with placing (?) after the cut-off, but I find placing them at the end more natural like the usual truncation means with a preferential ballot. Finally why use a grade ballot to rank n candidates? Because it allows doing the job with n answers instead of n*(n-1)/2... Steph. ---- For more information about this list (subscribe, unsubscribe, FAQ, etc), please see http://www.eskimo.com/~robla/em
