The LAT/USC poll has been a consistent outlier this election season. The same poll seemed to work well in 2012, so I always wondered if it was capturing something (enthusiasm maybe) that other polls were not.
Nate Cohn does some analysis of this poll
tinyurl.com/…
Basically, methodological problems are distorting this poll. I’m not a great fan of ‘unskewing’ polls, but in this case, I think some analysis of the issues are justified. Patrick Murray of Monmouth polling said on twitter after hearing of one of their weighting problems “Amateur’. I’m guessing that the folks running this are not really pollsters, more like researchers and aren’t aware of some of the subtle problems that can crop up in polling.
And the end result is that HRC should be leading with ‘normal’ weights.