A week ago I said Clinton would win by 10+. There has been some controversy in the past week over Obama's "bitter" remarks. I doubt those really have much of any effect. But I don't think that truly matters in the end. In order to make this a truly close primary, Obama needs not only to convince voters that Clinton might not be so grand, but also that he is a prime alternative. Even the "close" polls, Obama's polling only in the mid 40's (or even low 40's) with many undecideds. I'm not one to believe that there are a ton of undecideds out there. Many in the mainstream media including First Read from MSNBC refuse to take polls from automated voice polls (polls that don't use a live interviewer), but these polls tend to have lower undecideds. Those polls that use live interviewers tend to be those that have larger undecideds (whether it be a Bradley or reverse Bradley (the fear of a interviewee to give a response about a candidate due to the candidates race) or simply don't push the interviewee for an answer). In fact, the top three most accurate pollsters according to pollingreport.com are ALL IVR. The top pollster by a long mile is Surveyusa, which by the way is the poll most favorable to Hillary Clinton. Some may attribute Surveyusa's dominance over its ability to poll late in the contest (because interviews by live interviewers usually have to take place over a number of days). Well that wasn't the case in Ohio where Surveyusa was the most accurate pollster (and is the state most comparable to Pennsylvania in terms of its demographics).
Public Policy Polling is out tomorrow (they say a surprise is coming....)... they are the second most accurate pollster... they were also the second most accurate in Ohio.
They have been more favorable to Obama in PA than Surveyusa, so we'll see if there is any sort of a trend.
The point I'm trying to make is that these so call "undecideds" are going to come in and vote Clinton for the most part. They are going to propel her to a 10+ victory.