Now that’s a fast response in the Meta-Analysis.
On August 18th-20th, the electoral vote estimator appeared to be at a new plateau of Obama 289-291 EV, McCain 247-249 EV. Then on Thursday the 21st, John McCain had a very bad day. The next day the estimator started creeping up, presumably because the last day of multi-day polling caught the beginning of the fallout. Now we’re at Obama 300 EV, McCain 238 EV, with a Popular Meta-Margin of 2.36% and rising.
Seeing this effect so quickly and clearly comes from having very high precision.
“Seeing this effect so quickly and clearly comes from having very high precision.”
I must say you’re very brave/confident. I haven’t yet spotted this trend being confirmed at other sites.
I was moved to comment when I saw the change continue in the same direction for several days. Other sites are less precise and therefore would have trouble picking up the change.
Caveat: there was at most one day of sampling after the houses event. It is also possible that some other event has driven the change, though I am not sure what that would be.
In any event, this uptick is about to be overwhelmed by the next two weeks’ events – whatever their net effect is.