IMO the one off polls lag the tracking polls which have a better input into impetus. I have no doubt that as we get closer to election day the one off polls will 'herd merge' together as they wish to keep whatever shred of credibility they still possess.
****************************************************************
One week has passed since the media's "Shock Horror" foaming at the mouth frenzy with female television talking heads taking to their fainting couches and Donald Trump's obvious strong performance in the second debate.
The polling firms rushed out their post Trump Tape 'results' with grossly distorted poll components (Dem +16 in the most egregious example) and advised Trump was up to 11 points behind Clinton.
That is of course ridiculous and even as a 'snapshot in time' it is ridiculous.
The honest tracking polls, each one which has its own bias which is cancelled out in the aggregate, which is what counts, told a different story. Lets take a look;
On Friday, the day of the Trump Tapes release, Trump had an aggregate lead (the last column on the right) of 0.1 He had lead from 09/02 without interruption from between 0.1 to +4.0 from time to time. Towards 10/7 Trump had been declining slightly and the post tape release media frenzy accelerated that sharply.
As can be seen UPI/C which has a Clinton bias jumped from Trump being 0.9 behind to 6.4 behind in three days! This was mirrored to a lesser extent by PPD and hardly at all by the USC/LATimes poll, which has a Trump bias, which kept him ahead for three more days before, even there, he surrendered to being a tiny 0.4 points behind.
The most alarming poll for Trump supporters was Rasmussen which is now a daily tracker (USC is a 7 day and PPD a three day average which explains a lot) and is the most reactive to the media.
Trump went from -1 on 10/7 to -7 !! on 10/7 and then to +2 on 10/13 a 9 point turnaround in his favor.
This is why any one poll should be taken with the utmost caution and aggregation with direction indication is the best analysis we have.
Nate Silver follows "one off "polls which give Clinton a much larger lead than the current tracker aggregation of Clinton +0.7. As I have written either he is right or I am there is no middle ground the results being so far apart at this stage.
So there you have it, after all the sound and fury from the media over basically nothing, nothing has really changed. Trump led by 0.1 on 10/7 Clinton leads by 0.7 today and Trump could well be in the lead at the start of the next 7 days.
In the end, if Trump had performed disastrously in the second debate we might be looking at a different polling picture, but even then it may not have had a major effect long term.
It is the state polls that matter and a little more post craziness results are needed to get a handle on things as they stand with them.
No comments:
Post a Comment