Checking for conflicts

[Mike Ositoff]

As I understand it, the way Tideman's suggested procedure checked
for conflicts was by considering whether various pair-orderings
can be found in the same ranking of all the candidates.

So, to find out if a certain defeat conflicts with any sets
of larger defeats, find out if there's a possible ranking of
the candidates that agrees with that defeat's pair ordering,
and with that of every stronger defeat.

That might be easier than tracing every possible beatpath,
but when there are 20 candidates, there are something like
4E18 possible rankings of the candidates, so I don't know if
that procedure would be feasible with very large numbers of
candidates--or if any SD procedure would.

Norms procedure would, and so I hope that it does produce
the results of SD.

Mike