I'm freakin' sure that most mainstream polls have deliberately inflated the Witch's margins, their goals being to control the process and discourage the opposition.
But by how much did they overstate the Witch's support and understate Trump's? For starters, most of them probably oversampled 'Rats and undersampled Pubbies. They are basing their models on turnout from 2012, but this year's voters will be proportionally more GOP than in 2012 and proportionally less 'Rat. That's because the voters turning out will contain smaller percentages millenials and smaller percentages of blacks (without Obama on the ballot). And in both of those groups, among those that will be voting, polls show Trump doing better than Romney did.
Now here's another conundrum. Most of these polls show Johnson and Stein combined getting close to 10%. There is no discussion of the fact that third-party candidates usually fizzle out as Election Day approaches. Johnson is doing several times better than any Libertarian candidate in history right now. I can't for the life of me figure out why, since he gets next to zero media coverage and, in my book, he's far from a true libertarian. Which raises the issue: where are more of Johnson's votes more likely to go to: Trump or the Witch? My guess is Trump.
Another conundrum: Most of these polls add up not to 100%, but more like 90-92%. That means 8-10% can't tell a pollster who they're going to vote for. The general rule is that most undecideds break for the challenger/outsider. That could be Trump's "trump card." (pardon the pun).
I mostly agree, but I am unwilling to quantify any of those effects. I just hope it’s enough to overcome this year’s epic levels of fraud.