With the final Newspoll and Essential Report for the year coming in (the results of which can be seen over at The Polls page) and with the December Nielsen poll going AWOL and not coming out at all (shame on you Fairfax), we ought to update our Pollytrack numbers. The last poll released was the Essential Report of December 15th, but our Pollytrack numbers still shifted last week as the one off Galaxy poll taken back in mid November fell out of our 5 week rolling Pollytrack window.
The Pollytrack charts for the primary and two party preferred estimates come in like this (where the little error bars are the theoretical minimum size of the maximum margin of error of our polling super sample)
The vote estimates are really getting away from Turnbull at the moment, taking the ALP and the Coalition back to the same situation they were experiencing in May. Alexander Downer can at least lay claim to having something over Truffles as leader – his polling honeymoon was longer.
If we look at the way these vote estimates play out in terms of swing and the number of seats such a uniform swing would theoretically deliver according to the pendulum we get:
A big Merry Christmas goes out to Chris Pearce, the Member for Aston as he’d be the 27th seat to fall leaving Joe Hockey in North Sydney as the most marginal Coalition seat holder were this repeated at an election with a uniform swing.
Next up, the raw data including the weights used for Pollytrack (weighted by sample size) as well as the other measures we follow over the The Polls page (our Weighted Pollytrack where the phone polls have their weights pre-set to 0.25 a pop, the phone poll and the all poll average)
The polling year ends on a note of polling consistency, with all our metrics being nice and agreeing with each other.
Finally, our locally weighted polynomial regression lines of fit we run through our polling data:
Our high sensitivity regressions have gone a bit ape over the last couple of weeks as the ALP polling numbers surged a few points, suggesting that the calmer low sensitivity lines – the more realistic lines – should lift over the next month if the current polling values for Labor hold up.
UPDATE:
Oops, I forgot the sample size and MoE chart.
2 Comments
Would it be possible to note important events upon Pollytrack (i.e. when Turnbull took office etc)?
It would be, but it could turn into a rather large chart to fit it all in (depending on how many events we add). The Turnbull leadership change happened the week before where the ALP primary dropped below its 2007 election result on the very first chart.