Pollsters Clueless To the End

For much of the last year, I have written about the polling regarding job approval for President Trump and also for the generic vote for Congress.  There seemed to be two groups of pollsters:  those who showed major Democrat advantages and those who showed thing rather closer.  There was no way both groups could be right.  I expected that as we got close to the election, though, the results would converge as the pollsters who cooked the books to get their desired result had to move to more realistic polling to preserve their reputations for the future.

Putting that thesis to the test, I looked at the generic ballot polling that was released today.  There were four polls.  One, by CNN, gave the Democrats a lead of 13% which would be the harbinger of a landslide.  A second poll, by Rasmussen, had the Democrats ahead by 1%, and that would mean a Republican control of Congress of about the current size.  The other two showed a lead for the Democrats of about 6%.  That would indicate that the Democrats would pick up seats, but also that they would likely not gain control of the House.

So how can this be?  There's a simple answer.  At least some of the pollsters haven't got a clue how to deal with the data they obtain.  They're busy massaging the numbers to show a result that they desire.

Simply put, the pollsters are mostly clueless even the day before the election.