Protest vote did not decide the outcome

But out of those people most went to the Greens, not the Liberals. And of those who changed their vote from Labor last election to Greens this election, all preferenced Labor. The only seat to change hands on this dynamic was Melbourne, but it went to the Greens, who will support Labor. It changes the balance of power on the Left, but not the government.

There is more truth in the Rudd claim, but not because Kevin was well-liked. This is a proposition that can be tested in Queensland.

The median swing against the government in Queensland is 5.16 percentage points. The swing in Griffith, Rudd's seat, was 4.07 percentage points. So perhaps the sympathy vote for Rudd in his own seat was 1.09 per cent.

Or it could reflect the fact that the Liberals ran virtually no campaign (still waiting for my first Liberal letterbox drop) and that Rudd ran a positive local campaign divorced from federal issues.

This could be contrasted with that other discarded leader, Malcolm Turnbull, who in Wentworth received a swing to him of 11.53 per cent.

There was a Queensland effect, but it was small. Only 2 per cent of our sample rated it as the most important issue, although 50 per cent of those were from Queensland. They were also mostly Liberal voters in 2007.

Across the country, one reason for disliking the way Rudd was dispatched was that some voters wanted to do it themselves, meaning it was not a vote changer for them, unless it was back to Labor.

Others mention Rudd as a reason for voting against Labor because of the failings of the "Rudd-Gillard government". Then there were the complaints Gillard was disloyal, which fed the perception that Labor was manipulative. This was counterbalanced by those who thought a good prime minister needed to be ruthless.

Some approved of the achievements of the Rudd government, and said Labor wouldn't be there without him. So, a Rudd factor, but barely perceptible and moving votes in both directions.

The last claim is that it was a vote against Labor state governments. There has to be some truth in this because the two biggest swings were in NSW and Queensland, the states with the least popular Labor governments, but I'm not sure that it is sufficient.

Looking at Queensland, where the largest swing occurred, another explanation could be the ETS or the mining tax. This is possible, as generally speaking the largest swings were in rural areas, where opposition would be expected to be higher, although not in Herbert or Flynn, the two seats most directly affected by both.

All of which leaves specific legislation and poor perceptions of the Labor brand, along with the work of individual candidates, the most likely explanations for the result. Unfortunately for history, these are untidy factors and they don't fit the opportunistic narrative of public relations.