Trump really did a number on him. They've gone from exceedingly confident to exceedingly cowardly.
This
Nate Silver arguments often follow this sort of logic:
Some evidence might suggest there could be movement in a particular direction. Now, not all evidence is suggesting this. For example, Evidence A suggests this phenomenon isn't occurring at all. However, Evidence B and C do suggest such a trend, at least to an extent, so it's worth mentioning this pattern as a distinct possibility.
With that said, we shouldn't be too quick to conclude that this movement is actually occurring, as Evidence B and C both fall into Category Q. Category Q has sometimes been unreliable or shown different numbers from other categories. It's important to remember to consider all evidence and categories when analyzing the state of this race, and be aware that the numbers can change. And when the numbers do change, it could signify a larger movement, or it could merely be statistical noise.
Yep. That what probability theory is. There are indication that the race have tightened a bit, but not enough. And all coming from "similar" sources.
Good polls
JUL. 29-31 CNN, (A-, RV) C+8
JUL. 29-31 CBS News/New York Times, (A-, RV) C+5
JUL. 31-AUG. 2 Fox News, (A, RV) C+9
JUL. 31-AUG. 3 NBC News/Wall Street Journal (A-, RV) C+9
JUL. 29-AUG. 4 IBD/TIPP (A-, RV) C+4
AUG. 1-3 Marist College (A, RV) C+14
AUG. 1-4 ABC News/Washington Post (A+, LV) C+8
AUG. 4-7 Monmouth University (A+, LV) C+13
AUG. 5-8 Selzer & Company (A+, LV) C+4
AUG. 9-16 Pew Research Center (B+, RV) C+4
AUG. 18-24 Quinnipiac University (A-, LV) C+7
