But look, do yourself a favor: go back and read the coverage of ACNielsen’s polling from six years ago. You can find it here and here, and my commentary from the time is here.

It was bad for the government (a Coalition government, of course, at the time). Very bad – in fact, rather worse than this week’s (58-42 compared to 56-44). And the commentary certainly reflected that. But reading it again I can’t see any suggestion that the government’s fate was a foregone conclusion, or that it was dead in the water.

And indeed, although the polls got even worse for it, that government didn’t lose in a landslide. It was decisively beaten, but it still managed 47.3% of the vote.

Governments fall behind in the polls all the time; sometimes they recover, sometimes they don’t. For what it’s worth, I don’t think this one will, but the idea that it’s impossible is sheer nonsense.

The Howard government’s recovery in 2001 is probably sui generis, but Peter Brent yesterday graphed its performance against the Gillard government’s polls to date and the two look eerily similar.

There’s a real story in the opinion polls at the moment: the way that the trend towards Labor from about the last eight months has gone into reverse. (See Mark the Ballot for the history.) But the media don’t want to tell that story because it would mean (a) admitting that the pro-Labor trend existed in the first place, which would conflict with their narrative, and (b) admitting that what matters is the trend averaged across a number of polls, not relying on a single poll or a single pollster.

It’s less than a year since I quoted Brent’s wise words about opinion polling, but it’s worth doing it again:

If news items were given the emphasis they deserve, political polls would [be on] say, around page eight … But opinion polls cost a bomb to produce, so onto page one they must go. Then everyone must pretend that’s where they belong, adding several hundred words of interpretation — turning them over, looking for meaning, interpreting them as good or bad for someone or other, pretending you can identify why the numbers move over a fortnight.

Don’t get excited about a dramatic movement in any single opinion poll.

Each is just an imprecise dip in the ocean. Wait for another, and then another.

If you’re writing a front page story based on a single opinion poll, you’re doing it wrong. No matter how interesting it seems, no matter how much it agrees with your preconceived notions.

And one more thing: the preferred prime minister numbers are still meaningless, no matter how many of them you’ve got. They were meaningless when Julia Gillard was ahead, and they’re meaningless now that Tony Abbott is ahead. Just ignore them.