Winston Churchill, winner of the 1951 general election despite coming second
One of the arguments raised by opponents of the Alternative Vote is that it allows the candidate with the second or even the third largest number of first preference votes to win the seat. This is a clinching argument against AV and in favour of First Past The Post only if you think that second and subsequent preferences ought not to be counted and that only first preferences really matter, i.e. you are making a basic assumption that lies at the foundation of the case for FPTP in the first place. People who think that second and subsequent preferences ought to count should be relaxed about the possibility that first preference votes alone might not be decisive.
One can go further, in fact, and look at the maths. Fullfact.org reports on research in Australia, where AV is used, which found that in 3,354 constituency contests since 1949, only seven, or 0.02 per cent, saw the third placed candidate on the first round finally win.
Compare this with the experience of FPTP in the UK. The 1951 general election saw Clement Attlee’s Labour party win 295 seats on the strength of 13,948,385 votes and the Conservatives, led by Winston Churchill, get 321 seats (and a working majority in the House of Commons) with only 13,717,850 votes, that is to say 230,535 votes fewer.
In the context of 45 general elections since the Great Reform Act of 1832, this means that the second-placed party, rather than the first-placed party, has formed the government on 2.2 per cent of occasions. So much for the claim that FPTP means that the first-placed winner always wins.
PDF 
Loading ...
And there’s more. American presidential elections are conducted using FPTP in individual states to allocate the electoral college votes of each state. The winner is the candidate with the most electoral college votes, not the most actual votes. In 4 out of 56 elections (7 per cent), the winner in the electoral college has not been the candidate who got the most votes but the candidate who came second.
In the UK general election of February 1974, the Conservatives won 297 seats with 11,872,180 votes, whereas Labour won more seats (301) with fewer votes (11,645,616). In 1929, the same happened: 8,252,527 Conservative votes returned 260 MPs while 8,048,968 votes elected 287 Labour MPs.
In neither of these two elections did Labour win enough seats to form a government, but that still makes three elections out of 45, or 7 per cent, in which the most votes did not lead to the most seats.
A central claim of the No campaign is that AV is not fair because the second or third placed on first preferences might actually win. The same is true of the electoral system they are defending.