From Nate Silver, reviewing today’s polls:
While this is not a terrible set of numbers for Clinton — and probably not the “game changer” that the Comey news was billed as on Friday — I’ve also seen analyses that go too far in the other direction and conclude that the news hasn’t really had any impact. Clinton’s popular vote lead is down to 4.7 percentage points in our forecast, as compared with 5.7 percentage points on Friday and 7.1 percentage points two weeks ago. And Trump’s chances of winning are 24 percent in the polls-only model, up from 19 percent on Friday and 12 percent two weeks ago. Trump’s chances are 26 percent in the polls-plus model, which is converging with polls-only.It’s not easy to tell how much of that shift reflects a reaction to Comey, as compared with a race that had been tightening already. And it remains the case that the margin is closing because Trump is gaining ground from undecided voters and third-party candidates, rather than Clinton losing support. The fact is, though, that the data we’ve gotten during the past few days is consistent with a reasonably competitive race — although one in which Clinton has the advantage — especially given the significant disagreement in the polls and the relatively high uncertainty surrounding the polling this year.But we also haven’t seen many recent high-quality state polls from states such as Pennsylvania, Michigan, and Wisconsin, which are a key part of Clinton’s electoral firewall. The next set of results from those states will go a long way toward determining just how nervous Democrats wind up being. But at this point, the election is a long way from being in the bag for Clinton. Both FiveThirtyEight’s models and betting markets give Trump about a 1 in 4 chance of winning the election, about the same chance as the Chicago Cubs have of winning the World Series.