For context, I've posted three prior pieces on this distorted poll and the use/non-use of its results by RealClearPolitics.com (RCP), which you may want to read before reading this latest update. They are here, here, and here.
So, as noted in those pieces, RCP ignored the Reuters/Ipsos (RI) daily tracking poll, which has always systematically over-sampled Democrat voters, throughout the duration of the GOP primary process. Then, on June 4 of this year, RCP for some reason made the decision to begin including RI in its running average of recent polls, even though RI's results have been ridiculously outside of the overall distribution of results of all other polls. I first noticed this in late June, and inquired via email to RCP to try to understand why they would suddenly include a poll they had ignored for a full year, just as Donald Trump had essentially pulled even with Hillary Clinton in the average of all the other polls. To this day I have received no response.
Interestingly, though, shortly after I began blogging on this, the RI poll released a result on July 11 that was apparently too outrageous (Hillary +13 when the average of other polls was Hillary +2) for RCP to include. Not saying that my blogging necessarily had anything to do with it, but it is an interesting coincidence.
Even more interesting, RI just released its new update this morning that suddenly shows a far more reasonable Clinton lead of +6 over Trump. While still well outside of the average of other polls, where Clinton's lead has shrunk to between 2 and 3 points, it at least can be considered a fairly reasonable result, one that could be written off as just statistical noise.
As I write this on Wednesday morning, RCP has not yet included this latest RI result in its average. We'll just have to wait and whether or not they ultimately decide to do so.
No comments:
Post a Comment