And here is my view of SV (even though it is highly unlikely that anyone is going to take the slightest notice)!

It really is time we recognised the nonsense that is the Supplementary Vote (SV). What we are talking about here is not first and second "preferences", as is generally being alleged, but a first and second VOTE. Not only that, the only second votes that count are those for the first two in the first, FPTP, vote (and even worse of course is the fact that the voter has to GUESS who the first two are).

Moreover, the true position is that in this election neither of the first two candidates achieved a majority on any rational calculation. On the first VOTE Johnson achieved a mere 42% and Livingstone a mere 36%. On the second VOTE Johnson achieved a mere 10% of the vote while Livingstone achieved 12%. It is sleight of hand just to add these two votes together, and then pretend that the "winner" has majority support.

The candidate who achieved the highest second vote was a candidate shut out on the first vote, namely Paddick of the LibDems with 26% - more than twice what either Johnson or Livingstone achieved. Who can say that under the much more rational Alternative Vote (AV) with real first, second and later PREFERENCES, and the voter's consciousness that his SINGLE transferable vote might really count, the LibDem candidate might not have been the overall favourite, with more than the required 50% of the total vote?

The nearest comparable system is the French two round system. Here there are also two votes but on separate occasions. On the first occasion (if they vote at all) the French tend to scatter their votes - especially on the left - in a series of "protest" votes for splinter parties, hoping to come back to the "real" candidate in the second round. But, they could find that he/she is shut out. And so, in 2002, instead of being able to vote for Jospin who was, by a tiny margin, just behind Le Pen, they found themselves faced with a choice between a fascist and a right-winger even though Jospin was probably the overall favourite for President. But at least they could SEE what their choice was, which is more than can be said for SV. And of course there is none of the SV jiggery pokery of adding two VOTES together, since all the second vote is split between just two candidates so one of them is bound to have a majority

I wonder who devised SV and why there is not a clamour to have it replaced in future by AV?

AV is *not* the system to hope for - it has its own disastrous problems, in particular that you can only afford to vote for your "third party" candidate if you're confident they're not going to get too many votes. Any sane system will meet the "Condorcet critereon", which states that if there is a candidate who would win a pairwise contest with any other candidate, they should be elected the winner. It will also meet a bunch of other criteria, such as monotonicity (moving someone up your preferences cannot hurt them) that AV does not meet.

However, in this instance I don't think I can blame the voting system; everyone knew the runoff was going to be between Ken and Boris, and so I fear that those who did not express a preference between them did so on purpose.

I guess the strongest argument to the contrary you could give would be that if the voting system *forced* you to choose between Ken and Boris then maybe a lot of "as bad as each other" idiots would have grudgingly admitted to the polling form that no, one of them is worse.

Unfortunately the Condorcet system has it's own fatal flaws. The worst flaw is that a later preference can count against an earlier one. So thare would be good tactical reasons for Ken and Boris's supporters not to give a later preference at all or even to express a false later preference. This reason alone is enough to remove it from serious consideration as a voting system.