poll frequency means that the polling average trendline is likely to be extremely accurate.
the interesting thing about the trendlines is their consistency. both candidates operate within a 5% range. ranges that don't intersect. ranges that are rut-like in their longevity.
for instance, it is noticeable that trump is topping out for the third time in a row at 42% or so. his new campaign team has done a good job of maximising his share. but his message looks like it hits a ceiling at 42% (or 44% if you look back a year). shifting that ceiling is going to be difficult. he runs into the demographic issues he has been using to deliver the 42% less-educated, white, male voter he has. women, african-americans and non-english speaking latinos don't like him at all. more educated white people aren't with him either. he needs these people to have collective amnesia in order to break through to higher figures.
clinton's team needs to get her numbers back near the top of her range. that is, around 48 or 49%. in which case, it isn't even close. to reach those numbers she has to deliver a positive message and to avoid falling into any new integrity traps. she needs a theme that resonates with people. unlike trump, she also has lots of persuasive surrogates. they will keep reminding clinton's core constituencies of who trump is. immigration. birtherism. racism. bullying etc. about time to release them. their aim is to harden his ceiling around 42%.
maybe the debates will change the ruts that they are in. otherwise, the sum of the polls trendline numbers are not only accurate but also highly predictable. what's also predictable is the panic and exuberance of the response as the candidate trendlines move up or down within their rut.
http://elections.huffingtonpost.com/pollster/2016-general-election-trump-vs-clinton
| 
