[EM] Condorcet Criterion for plurality.

[Martin Harper]

David Catchpole wrote:

> Votes:
>
> 3 A>B>C
> 2 B>A>C
> 4 C>A>B
>
> Using plurality, C wins. If we assume that voters have rankings, _whether
> or not they can express them on their ballots_, then plurality fails a
> Condorcet criterion.

But, in the above case, the B>A>C voters 'should' vote strategically for A, and
this means that A will win, so the sincere Condorcet winner should win. I'm missing
something here, aren't I? ;-)