[EM] French Presidential Second Ballot.

[Richard Lung]

The French Second Ballot illustrates what's wrong with Condorcet Pairing, in general.
As Simon Laplace first appreciated, you cannot add first and second preferences, because they are of unequal importance. Less than half the votes, in the first ballot, reveal a much greater level of support than do the extra votes gathered in the Second Ballot, exclusive to merely two candidates.
The Gregory method remedies this, with later no harm, unlike Borda count..
Gregory can be used for single vacancies, provided the form of keep values are used, and more extensively than in Meek method, where they were introduced.
This is my innovation of Binomial STV, with election keep values and exclusion keep values combined to give all the candidates average keep values.
The innovation of a rational exclusion count does away the need to exclude candidates, in a second round.

The French Second Ballot  has a good feature, in the recording of blank papers, as abstentions, or a plague on both your houses. 
BSTV has this feature, also. In fact every blank preference is counted towards a quota for an unfilled seat. 
This blank preference counting feature should reduce the relative importance of exclusions, compared to elections, abstentions occuring more towards the exclusion end of the BSTV Ballot paper.
Only actual elections should tell whether to apply statistical tests of a candidates proximity to the quota.

Regards,
Richard Lung.