CR pairwise

[Roy]

DEMOREP1 wrote:
>The highest median will suffice as a tiebreaker (as has been
>mentioned many moons ago on this list).

Then the strategic case (weak information notwithstanding) is almost 
the same as CR. Instead of voting full-approval and full-disapproval, 
you vote approved candidates at 100, 99, 98, etc. and disapproved at 
0, 1, 2, etc. The ranking is considered first, without regard for the 
numbers, so 100 vs. 99 is the same as 100 vs. 70 or 100 vs 2. The 
specific numbers only come into play to break a tie.

It's a rather clever way of transforming CR into the
approval-plus-ranking system that you prefer, but it has no appeal
for CR advocates.