Anecdotal evidence is worthless.
Pretending that polls stay static from the Spring to the Fall in most election cycles.
You might want to do a study of that some day.
Here's some really good anecdotal advice:
If Ted Cruz had to handle one tenth of the negative advertising he would get in the general election (if he actually won the nomination) that Trump has taken already, he would buckle and be at 5% in the polls.
He will NOT be able to handle Hillary, or whoever they throw in there, maybe not even Bernie.
Anecdotal evidence is worthless.
True, and so are many polls, especially the ones that 1) are comprised of soft questions about feelings, 2) are referring to an event seven months away, 3) are filled with hypotheticals. These factors, combine to make these favorability polls meaningless. Polls that ask “who are you voting for tomorrow” are far more reliable, in part because the pollster has a vested interest in getting it right.
Carson said something interesting when Megan Kelly asked him whether Trump could learn something from Carson’s vastly better likability ratings.
He said [paraphrasing] “True, a lot more people liked me, but they didn’t vote for me. Nobody liked Trump, but they all voted for him (laughing)!