For my latest ratings chart, click here.

U.S Senate
Republicans +6, Independents +1
GOP 51, Democrats 47, Independents 2

While I have no ratings changes this week, an explanation is in order in two races in which my current ratings run contrary to the public polling averages.

In Colorado, I continue to stick with Democratic incumbent Mark Udall, despite the fact that Republican challenger Cory Gardner has led every public poll since Oct. 1st. However, this is a close call and I am monitoring this race every day.

At the moment, Real Clear Politics indicates that Gardner’s average polling lead is 3.8%. This is a crucial number. In at least the last two election cycles, the final RCP average has shown Democrats in key statewide races—Sen. Michael Bennet in 2010, and President Barack Obama in 2012—outperforming those final averages by exactly 3.9% each. Additionally, recent reports indicate Colorado’s turnout might be unusually high for a midterm, and the state recently went to a system by which all voters may choose to vote by mail, which might aid turnout. Polling models tend to dismiss “unlikely voters,” but the ability to vote by mail has the potential to change the equation for individuals who are more willing to fill out a ballot at home rather than trekking to the polls. If so, we won’t learn that until Election Day; at least some of the polls won’t pick it up, and that could affect the averages. But the bottom line is that higher turnout almost always benefits Democrats, whose voters tend to be less reliable, so if high turnout does materialize in Colorado, it is almost certainly going to be to Udall’s benefit.

My final call in this race next week will come down to whether Gardner’s average polling lead is at least four points. If Gardner is leading by four points or more on Nov. 1st, I will change the rating. Otherwise, I am going to rely on past experience and stick with Udall.

In Georgia, I continue to rate the Senate race as “Leans Republican” despite the fact that Democrat Michelle Nunn has surged into a small lead over Republican David Perdue. Because of Georgia’s peculiar runoff rule—initially instituted by conservative segregationists decades ago to help prevent liberals or political outsiders from winning elections—a candidate must secure at least 50 percent of the votes on Election Day, or the top two vote-getters advance to a January runoff. At this point, I still do not see Nunn getting to that magic number of 50 percent, and even if she places first on Election Day, the dynamics of a runoff—with much lower turnout to be expected, and the fact that cash-rich conservative groups will spend the ensuing two months blasting her repeatedly over the airwaves—lead me to believe that she ultimately will not win the seat. Nunn’s hopes of winning this seat probably rest on her getting to 50 percent by Nov. 4th.

U.S. House of Representatives
Republicans +5
GOP 239, Democrats 196

Today, I am changing the rating in Minnesota’s 8th District from “Leans Democratic” to “Leans Republican.” A recent SurveyUSA poll, released Oct. 16, shows Republican challenger Stewart Mills with an eight-point lead over incumbent Democrat Rick Nolan. Word has been filtering out from Washington, D.C., for months about Nolan’s weak fundraising performance, and this is by no means a safe Democratic district in which a Democratic incumbent can simply coast. My best guess at this point, based on the available data and anecdotal information, is that Mills ousts Nolan on Nov. 4th.

I am also making a rating change in Arkansas’ 2nd District, where it now appears that Democrat Pat Hays has the lead over Republican French Hill in an open-seat race.  Hays has gone from losing 44%-43% to Hill in a Hendrix College poll released July 31st to leading in a poll by the same pollster, 46%-42%, on Oct. 20th. This is the most Democratic district in Arkansas, and the frantic efforts to boost Democratic turnout on behalf of endangered Sen. Mark Pryor may end up benefiting Hays more than they benefit Pryor.

As these two ratings changes cancel each other out, I am making no change this week in my projection of a modest, five-seat gain for Republicans, but again, I caution that this is almost certainly a floor and not a ceiling for the GOP.

Additionally, I am moving the race in Nevada’s 4th District from “Likely Democratic” to “Leans Democratic” and placing it on the Watch List. I learned yesterday, in a Twitter exchange with Nevada political journalist Jon Ralston ( that early Republican voting in Nevada is so strong, GOP challenger Cresent Hardy may have the potential to upset incumbent Democratic Rep. Steven Horsford. I am also reading from numerous sources that outside Republican groups are flooding money into this district, hoping to push Hardy over the top. I have read enough of Mr. Ralston’s work in recent years to know that he has his finger on the pulse of Nevada politics and is, in fact, one of the best state-level political analysts in America, so I trust his assessment. I am going to keep a close watch on this race over the next week to see if a further change in its rating is in order.

Additionally, I am placing two races back on my list due to dramatic closes in recent polling, and adding another not previously listed.

In Arkansas’ 4th District, a Hendrix College poll released Oct. 21 showed Republican Bruce Westerman leading former FEMA director James Lee Witt, a Democrat, 44%-42%. This reflects a dramatic narrowing from a previous poll on July 31st showing Westerman ahead 48%-34%. The only other poll I can find for this district was a Wes Anderson poll on Aug. 21, showing Westerman ahead 47%-29%, but with a large number of undecideds (24%) and Westerman still under 50%. While Westerman is clearly still ahead, there has been enough of a narrowing in two polls by the same pollster that I can no longer consider it “Safe Republican,” and I am going to re-list this one as “Leans Republican.” It may well be that the ongoing efforts to boost Democratic turnout for Sen. Mark Pryor are helping Witt also.

In California’s 21st District, Democratic challenger Amanda Renteria has made a major turnaround in a little over a month. A SurveyUSA poll released Oct. 21 showed her trailing incumbent Republican David Valadao 47%-42%, as compared to a survey released Sept. 9 by the same pollster with Valadao ahead 56%-37%. I think it is clear that Valadao remains ahead, but I am putting this race back on my list as “Likely Republican.”

A surprisingly close race has developed in Massachusetts’ 9th District, in which two Emerson College polls in the last few weeks have produced differing results. The first, released Oct. 8th, showed Republican challenger John Chapman with a surprising 45%-40% lead over incumbent Democrat William Keating. The most recent survey, released Oct. 20th, showed Keating back in front, but only by three points, 47%-44%. I am adding this race to the list as “Leans Democratic” and placing it on the Watch List.

Finally, there are two southern California races I am watching very closely, and while I am not changing their ratings this week (both currently rated “Leans Democratic”), I am going to do a final, in-depth analysis on them over the coming week. CA-26, based mostly in Ventura County, features a tight race between incumbent Democrat Julia Brownley and Republican challenger Jeff Gorell, currently a state legislator. CA-52, based in the San Diego area, also has a tight race between incumbent Democrat Scott Peters and Republican challenger Carl DeMaio, most recently a San Diego city councilman. In both cases, the Republicans have made a point of demonstrating that they are social moderates, and both districts are close to an even GOP/Democratic split. Along with AZ-2 and IL-10, these are among the races in which I am having the biggest struggles with making a call. If Republican performance nationally is better than expected, these four will be among the first currently “Leans Democratic” seats to flip to the GOP.

Democrats +2, Independents +1
Republicans 26, Democrats 23, Independents 1

I am making one rating change this week. In Massachusetts, it is now clear that Democrat Martha Coakley, who infamously faded at the end of a special election for the U.S. Senate and lost to Republican Scott Brown, is in the process of another late collapse. Republican Charlie Baker has led in four of the last six polls and is now clearly ahead. Coakley may yet win, if Massachusetts’ Democrats turn out and save her in this heavily Democratic state, but at this point, it does not look promising.