Reuters/Ipsos Natl: Trump +2

[SherlockHound]

Tied 38-38 with RVs. Also, interesting that even when Trump is leading he still can't get over 40% of the vote.

More people need to be talking about this. I'm really interested to see how the 20% of undecided voters will break. I assume at least 5% of them are Bernie supporters who have yet to come home.

What were the percentages of undecided voters before the DNC convention in 2008? I doubt it was anywhere near this large of a number.

[SherlockHound]

In 2012, after the RNC, Romney peaked at a tie with Obama on the RCP average after being down by 3% 2 weeks before the RNC ended .  Romney never exceeded Obama on RCP average by more than 1% the entire 2012 election.  Trump already exceeded that by being up 1.1 as of today after being down by 4.7 2 weeks before the end of the RNC and it could go up more if and when Rasmussen comes in with their poll.  It is clear that Trump's bounce is greater than Romney and all things equal Trump will most likely outperform Romney even if he ends up losing.

The way this election has been so far, I could see Trump winning a squeaker or losing in a landslide 359-179. Trump, just like the polls, has been so unpredictable that any kind of traditional methodology we use to predict elections is questionable.

[SherlockHound]

The difference between 2012 and 2016 is that the people decided on Romney or Obama because they liked the candidates and their political stances, while this election is all about chosing the lesser of two evils.

Reminds me of 2004, Bush v Kerry.