Higher than all the other remaining options, after the algorithm started eliminating people, sure. But that's an internal detail of the voting system – if my favourite candidate was eliminated early, I won't consider that I got a candidate I preferred to all the alternatives!
If 1/3 of voters think A > B > C, another 1/3 think B > C > A, and the remaining 1/3 think C > A > B, then, for each of the three candidates, 2/3 of voters would prefer someone else. There's no candidate that 50% or more of voters prefer to the alternatives.
After one candidate is eliminated (say C, because those thirds weren't quite exact), the remaining vote between A and B goes heavily in favour of A, because all the people who had C as first choice preferred A to B. Fine, we have a result, and it's true that 2/3 of people preferred A to B in the notional second-round vote. But that was notional!
The only way you can say that more than half the voters preferred A to the alternatives is if you pretend C wasn't one of the alternatives, because of their elimination in the first round. And since the rounds are instant and happen entirely in the imagination of the algorithm, that hardly seems like a particularly interesting property, especially because you can't use it to compare with other preference-voting systems (like Condorcet-type things) that don't even have internal rounds.
If the winning candidate A pursues a policy that B was dead against, claiming their 2/3 of the vote as a huge mandate of popular support, then maybe they have a point. But if they pursue a policy that C was dead against, on the same grounds, it would clearly be untrue.
no subject
If 1/3 of voters think A > B > C, another 1/3 think B > C > A, and the remaining 1/3 think C > A > B, then, for each of the three candidates, 2/3 of voters would prefer someone else. There's no candidate that 50% or more of voters prefer to the alternatives.
After one candidate is eliminated (say C, because those thirds weren't quite exact), the remaining vote between A and B goes heavily in favour of A, because all the people who had C as first choice preferred A to B. Fine, we have a result, and it's true that 2/3 of people preferred A to B in the notional second-round vote. But that was notional!
The only way you can say that more than half the voters preferred A to the alternatives is if you pretend C wasn't one of the alternatives, because of their elimination in the first round. And since the rounds are instant and happen entirely in the imagination of the algorithm, that hardly seems like a particularly interesting property, especially because you can't use it to compare with other preference-voting systems (like Condorcet-type things) that don't even have internal rounds.
If the winning candidate A pursues a policy that B was dead against, claiming their 2/3 of the vote as a huge mandate of popular support, then maybe they have a point. But if they pursue a policy that C was dead against, on the same grounds, it would clearly be untrue.