[EM] A "Runoff" type method for resolution of cycles in condorcet

[Ken Taylor]

The other day I thought of an interesting way to complete condorcet if
there's no CW. But it seems simple enough that I figure it must have been
proposed before. So I'm wondering if this method has indeed been proposed
before, and so what it's generally called. In either case, I'm wondering if
there's any glaring problems with the method that any of you experts on this
list see right off :)

So it works like this, if there's no condorcet winner, then
1) determine the Schwartz set. From this point on, only deal with the
candidates inside the schwartz set -- those outside have no more effect on
the election.
2) For the candidates within the schwartz set, drop the candidate with the
least number of first-choice votes. This step is why i call it a "runoff"
type method.
3) Dropping that candidate drops all defeats that candidate may have over
the other candidates that are left in the schwartz set.
4) Calculate the new schwartz set. If there's a winner, stop. If not, repeat
until there is a winner.

Ken