Morgan: 53.5-46.5

Out of bed now, so I’m a bit late with the news that Morgan has completed a hat-trick of polls that have produced their closest result for the year just in time for polling day. Conducted last night, it has Labor ahead 53.5-46.5, their primary vote at 43.5 per cent (the lowest since January) and the Coalition’s at 41.5 per cent (highest since last November). The odd pollster out, ACNielsen, conducted its poll much earlier in the week.

In administrative news, I just did cleared out a huge volume of comments from moderation, so apologies to those who were held up there. Most of you were first timers – you should be able to comment freely now. I’m off to perform my civic duty, and will have a think about how to handle the election night traffic on my return. I’m much preferring IRC to Meebo (see “Dress Rehearsal 3” post below) and beginning to think that alone should be enough to divert enough blog comment traffic to prevent my server from melting down.

Author: William Bowe

William Bowe is a Perth-based election analyst and occasional teacher of political science. His blog, The Poll Bludger, has existed in one form or another since 2004, and is one of the most heavily trafficked websites on Australian politics.

595 comments on “Morgan: 53.5-46.5”

Comments Page 6 of 12
1 … 5 6 7 … 12

  3. You’ve gotta love Howards bizarre tenacity “If people dont vote for me they risk not getting a tory govt” urrrrr yes and that would deter us why???

  6. yes Grover, I was shocked when I first heard her also. Is that the best they could do?? In a tight marginal?? The talent is really drying up.

  7. Responding to Amaranthus (text below mine)

    Do you know far and wide I’ve been looking for this? Thank you!!

    If you have several samples, you may as well (i) do whatever possible to make the estimates commensurate (ii) POOL them … and (iii) try to deal with anomalies if and where possible.

    The weighted mean is the result of pooling these samples, and is likely a better indicator, given we don’t know which is actually the best.

    The Galaxy is an anomaly, only if you take a whole lot of things as being ideal that never are inpractice. Anomalies and differences are far more likely the product of very specific methodlogical points and/or weaknesses in implementing random sampling, than any genuine differences in voter intentions.

    The standard deviation of the means (percentages) is about 1.4%. The Galaxy poll is within about 2 standard deviations of the mean of the means, without weighting because I’m considering between samples variance a bigger factor given the sample sizes.

    In short, if you’re realistic that there is more variation between polls than among, and admit you don’t know which is best, they’re not really that inconsistent and the best estimate available remains right around 54%.

    The nice Monte Carlo simulations based on the same mean gave about a 0.97 chance of a Labor win.

    *********************************
    Amaranthus Says:
    November 24th, 2007 at 2:29 pm
    As I said earlier, the polls are actually quite consistent with each other if you do your own preference weighting and take account of sample size correctly.

    If you allocate preferences as 75% green to ALP and 50% other. This gives the following table (with sample size as 2nd column):

    ALP n Poll
    53.75 2615 Newspoll
    55.63 2071 ACN(p)
    56 1421 ACN(o)
    52.13 1200 Galaxy
    54.25 1300 Morgan1
    53.13 2115 Morgan2

    The weighted mean is ALP 54.2% of TPP vote. That would yield around 88-92 seats.

    If you use the a 63% weighting for the Greens, it is ALP 53.2%. This yields 81-85 seats.

    *******************************

  8. I’m finding it very hard not to pour a drink before vote-counting begins. I’m forcing myself to drink cups of tea.

    Reading about Overington and Chijoff makes me want to open the champagne now. Best laughs of the campaign arrive shortly before the polling booths close.

  10. I think it’s especially ironic that Ovaries wrote a piece about there not being enough humour in the campaign yesterday. She has been absolutely hilarious.

  11. One correction. I meant if …. if you’re realistic that there is more variation between polls than within for these sample sizes; i.e. than if they were all done in exactly the same way by same people …

  16. Anyone watched the Sky News channel of Rudd? They have camera and mics on him when he goes to vote – and you can clearly hear his address. Not sure if they should have done that.

  18. Don’t be too hard on Chijoff. Remember, Jackie Kelly left it till pretty late to decide not to stand again. The Liberals would have struggled to get a new candidate, with the polls (and Kelly’s decision not to re-nominate) suggesting the seat was gone anyway.

    Chijoff, Newhouse and McKew are the three candidates from this election who won’t be easily forgotten. For different reasons, of course.

  19. In McEwen we had the mildly amusing situation of the Family First candidate manning the booth all by himself while our entire family were handing out Greens cards 😉

  21. Kina (241) I think it’s a very risky thing for Turnbull to do. If he is challenged and the result over turned Labor don’t have to select the same candidate in the re-run. They could come up with someone a lot more competent and anything could happen. If Labor wins and there are by-elections as a result of resigning Liberals I’m tipping the ALP will do very well in them.

  25. I don’t think ESJ and Glen are the same person. ESJ does fuddy duddy oldster brilliantly and Glen does young and naive the same. I have Glen’s wide variety of punctuation and mood for exactly the same reason I expect he does – late night drinking!

    All of them have a right to be heard and I’m eith the redoubtable William, this site is a Pseph site not a left wing love in site.

    General Wenck had better turn up with those divisions pronto!

  27. 14 minutes ago. Award-winning journalist Caroline Overington has hurled abuse at Wentworth ALP candidate George Newhouse before slapping him across the face at a polling station in eastern Sydney, according to witnesses.

    At first we thought who was this woman yelling at Newhouse, then she slapped him and we realised it was Caroline Overington.

  30. Appeal for sanity:

    Please, please don’t read too much – or even much at all – into the Sky News exit polling to be released any moment.

    This is simply a marketing gambit by Sky to get people tuned into their coverage early, as opposed to the various free-to-air options.

  32. William. Computerless tonight but I’ll be thinking all night of this fine forum and it’s excellent citizens. Thanks to you and all others for both the (at times) compelling insight and great entertainment. I’ve lost far too many hours here due to the exceptional quality of this blog. Have a great night.

  33. Anyone watched the Sky News channel of Rudd? They have camera and mics on him when he goes to vote – and you can clearly hear his address. Not sure if they should have done that.

    It’s OK, he’s changing it soon.

  35. Important to remember to make the distinction between meaningless online ‘exit’ polls, and phone polls it would seem, to avoid unnecessary palpitations at this stage I’d say

  37. At my booth the Lib HTV man was standing 50m away from the booth and on the edge of the footpath. While in line, I saw him almost run over twice by passing traffic looking for parking. I guess they mistook him for a car jacker…or mybe not.

  38. It otherwise would be a good tactic. For any candidate who has worked for govt you could advertise in the paper, just before an election, saying they are invalid. The defence [and no doubt Turnbull’s] is that if they hadn’t resigned properly they would be invalid.

    I guess you could say someone is invalid due to a possible criminal history – since they haven’t provided evidence of not having one.

    I suspect if Turnbull wins Newhouse will have grounds to challenge.

  47. They are dribbling out the exit polls results .. first they are giving most important issues to voters. 1. health and hospitals 2. the economy 3. IR/environment

  48. Did anyone else see Liberal HTV cards which had a Bold 9 Point Plan on them and the Labor HTV card I saw just had a pic of Kevin on the back….

    It was a bit concerning – a lot of people were reading through the Bold Plan as they were lining up….

    I hope Labor was a bit smarter about this in more marginal seats – I was in a fairly safe Labor seat.

  49. The problem with the weighted mean approach is that it assumes that the population proportions remain constant over time and that variation between the polls is just sampling error. This may be invalid if the underlying proportions are changing over time.

    still things looking somewhat good for the alp

    cheers

Comments are closed.

Comments Page 6 of 12
1 … 5 6 7 … 12