Which is the TRUE Winner then?

A common argument posed by those in favour of First Past the Post over a preferential voting systems is that First Past the Post gives you 'true" winners who've got all their support out of people making only one choice. They also say that First Past the Post is fairest as the one with "the most votes" wins. 

First of all watch out for the phrase "the most votes". It might look similar to "most of the votes" but it means something very different. Having more votes than any other individual does not mean a majority. A majority is defined as more than everyone else put together, i.e. more than half. Merely having the largest minority does not mean you have the most support. It does not mean that on the whole you are in the best position to represent everyone. To form a single party government in the House of Commons, a party needs a majority of seats, more than half. It SHOULD be the same to win a seat, a candidate SHOULD need a majority of votes to win, but under First Past the Post, they don't. How is that fair?

So, putting aside this wordplay slight of hand, let's have a look at a slightly more developed version of the pro-FPTP argument. First Past the Post supporters might say "Ah, but under AV the 50% comes from a mixture of preferences, so it's not a true majority." My reply to that is that First Past the Post votes are a mixture of preferences as well. You only get to make one choice under FPTP so you often have to make a decision about whether to give it to your 1st preference, or to give it to someone else for tactical reasons.

The difference is that under FPTP you don't know where the votes come in the voters' preferences. And FPTP X vote could be anything from 1st right down to 2nd last - whichever is required to keep out the least liked candidate. What AV does is make it clear what the preferences are. So while a majority in AV can come from a mixture of 1st, 2nd and 3rd preference votes, a FPTP victory often comes from a MINORITY that is ALSO a mixture of various preference votes, and may of them probably lower down the scale than they would be under AV.

To any FPTP supporter who says "only 1st preferences should count", I say, you have to know what the 1st preferences are to make them count. You can't know that without inviting voters to include as many preferences as they like. If you restrict voters to a maximum number of choices, especially only one choice then you'll just never know.

What it really comes down to, is that there are situations where a candidate would win under AV when he/she got fewer 1st preference votes than one of the others, but got a majority over all once the preferences of voters for elminated candidates are considered. FPTP-backers think this is unfair. They think this for two reasons, both of which I intend to show are incorrect:

Incorrect reason number 1) They assume that all the 1st preference votes are equivalent to the X votes in FPTP, and therefore whoever got the most 1st preference votes in AV would have won under FPTP.

This is wrong for the reasons I have already explained. FPTP votes aren't all first preferences. They are distorted by tactical voting and in fact are made up of a whole range of preferences, from people who really like the candidates, and from people who see candidates as the only realistic alternative to the ones they really don't like. Under FPTP many people vote for their 2nd-least preferred candidate as they see it as the only way to keep out their most disliked. FPTP votes are NOT all 1st preference votes, so you can't look at the 1st preferences in an AV vote and say that that's how the votes would have gone under FPTP.

Incorrect reason number 2) They believe that because the majority that leads to a victory in AV is made up of some voters 2nd and 3rd preferences (possibly even lower down in some case) it invalidates the victory.

Putting aside the fact that FPTP also includes a mix of preferences, I'm going to show mathematically how a winner under AV would always beat the FPTP winner (unless they're the same person as is very often the case) in a head-to-head.

We're describing a situation where a candidate a) wins under FPTP b) gets the most 1st preferences under AV and c) doesn't win under AV.

Under FPTP one of the candidate wins with the largest minority. Let's say it's Bob with a number of votes X. It's a minority, so X < 50%

Under AV the maximum number of 1st preferences Bob could get is X. He could get fewer (if, say, some of his FPTP votes were tactical and not real 1st preferences) but he wouldn't get more as there'd be no reason to vote tactically against a likely winner under FPTP.

So Bob has the most 1st preferences X, and he has <50% therefore so must everyone else. 

So no-one has a majority of 1st preferences and the process of elimination and redistribution of votes begins. The end result is one of the other candidates, let's say it's Anne, wins, getting a majority of votes after the other candidates have been eliminated. 

Anne's total is Y. As Bob comes second, he isn't eliminated at any stage so his total is either X or >X. But because he comes second he must have fewer votes than Y. Anne on the other hand has Y votes, >50% > X.

So Anne would have beaten Bob in a head-to-head. With just the two of them standing and the voters choosing one or the other, more would have chosen Anne and Anne would have won fair and square in a FPTP sense. And with only two candidates FPTP has meaning as there's a real Post: the 50% mark.

At this point it is clear that more voters preferred Anne to Bob than Bob to Anne. Some put Anne 1st, some put Anne 2nd, or 3rd, or even 4th. But wherever they put Anne, they put her higher than they put Bob. If a voter put Anne 3rd and his/her vote went to Anne, then Bob would have been placed lower than 3rd by that voter.

However little of an endorsement for Anne it may appear to be, it's LESS of an endorsement for Bob. If the voter really didn't care whether it should be Anne or Bob, he or she would have given neither of them any rankings at all. By including Anne and Bob, the voter indicated a preference of one over the other, even though they also declared that they really prefer someone else to either Anne or Bob.

Since Anne would win the straight head-to-head with Bob, Anne is the more democratic choice and the more valid choice. The fact that Bob would have won under FPTP, and that Bob gets more 1st preference votes is dependent entirely on the number of other candidates. Bob only APPEARS to be popular, because of the X votes in FPTP and the 1st preferences in AV being split across a number of candidates. But once preferences are considered with the least popular candidates eliminated one at a time, it boils down to a straight head-to-head between the two strongest, Bob and Anne, which Anne wins.

AV accounts of the number of other candidates. It removes the vote-splitting issue and makes sure that the majority view is considered. Because Anne would win the head-to-head between Anne and Bob, Anne is the true winner, and so it is AV, that produces the true winners, not FPTP.