Let's Call It 279-259 and 51-48 Romney

First the caveats.

The averages of most national polls (from both Pollster and Real Clear Politics) show President Barack Obama ever so slightly ahead in the popular vote and with enough electoral votes to retain the Presidency tomorrow.

Nate Silver at fivethirtyeight.com gives Mitt Romney only a 13.7 percent chance of winning; he has it 307-231 in electoral votes and 50.6-48.5 in popular votes.

The RCP average has Obama up 48.5-48.1 in the popular vote and with a 201-191 lead in electoral votes (with 11 states deemed toss-ups).  Pollster (the Huffington Post site) has it even better for Obama, up 48.0-46.8 in popular vote and 277-191 on the electoral side with four toss-ups (not enough for Romney even if he won all four).

However, I'm going against the grain for two very good reasons.  The pollsters I most admire, Gallup and Rasmussen, each have Romney up by a point, 49-48.  Gallup had suspended polling for most of last week in the aftermath of Sandy, but is out with its latest numbers today, and Romney clings to a lead.  Other polls actually have a tie in the popular vote (CNN at 49, Monmouth at 48, and Politico at 48).  How then, you might ask, is Obama ahead in averages?

It's simple.  One or two polling firms--those which I have targeted in the past as not reliable--have the President ahead by more than one.  Pew, for example, has Obama up 50-47.  I believe both Pew and Marist and Survey USA are oversampling Democrats by such a large margin that they simply cannot be trusted, neither for national or state polls.

Fox example, while Rasmussen has a 49-49 tie in Ohio and Monmouth has Obama up only a point, Survey USA has him up five points (49-44) there.

There's also word that CNN is oversampling Democrats by 11 points (it was seven points four years ago) to get to a 49-49 tie today, and that Romney leads by 20 or more with Independents.  Anyone who thinks Democrats will turn out with four percent more than in 2008 simply cannot be trusted.

However, I'm not going with Republican hacks like Dick Morris who come up with 325 electoral votes for Romney or even the highly respected Michael Barone who puts Romney at 315.

I'm more in the Karl Rove camp.

Here's how I get to 279-259 for Romney.

Nearly everyone agrees that 39 states and the District of Columbia are not in play (Minnesota is the only question mark there and then only with the Morris types).

In that group, Obama leads 201-191.

6 of  the states remaining are:

North Carolina 15--R

Florida 29--R

Virginia 13--R

Michigan 16--OB

Pennsylvania 20--OB

Wisconsin 10--OB

Let's give Romney the first three (the only argument might be Virginia where the RCP average has Obama up 0.3 points, but Rasmussen has Romney up two 50-48) and Obama the next three (Republicans think they have a shot in Wisconsin and Pennsylvania where Bill Clinton is making four stops for Obama today).

That gives Romney a lead of 248-247 with only five states in play.

Nevada 6--OB

Iowa 6--OB

Ohio 18--R

New Hampshire 6--R

Colorado 9--R

While Iowa could be close (ARG has Romney up 49-48), it seems reasonable to expect Obama to win there and in Nevada to get his number up to 259.

If we give Romney Ohio, he would be up to 266, still three shy of a tie and four shy of a win.  Thus, even with Ohio, Romney would either need to pick up Colorado or New Hampshire.

I have him picking up both.  Although RCP has Obama up 1.5 points in Colorado, it's largely because unreliable Survey USA has him up five and PPP up six; I'll go with Rasmussen again--he has Romney up three in Colorado (hey, aren't there a lot of Mormons in that state?).

New Hamsphire could be more problematic.  Most polls show Romney behind by two which is in fact the average.  Even Rasmussen has Obama up 50-48.  ARG has it 49-49; the WMUR Poll (which has been all over the map) has Obama up 51-48 (unless it's changed again in the last minute or two); NEC has Obama up 50-46; and PPP, a Democratic pollster, has it only a two point lead 50-48. 

The reason I've stopped trusting UNH numbers (WMUR) is that the same poll which shows Obama up by only three has Maggie Hassan, unlike all other polls, up by 11.

Let's put it this way.  Even if Obama wins New Hampshire, Romney still wins nationwide 275-263 if he holds on to Ohio and Colorado.

That's my story and I'm sticking with it.

On the generic front, it's nearly dead even.  Rasmussen has it 46-46, but the way things break, that should be enough for Republicans to hold onto most of their advantage in the U.S. House (I still have Democrats picking up only six seats).

My U.S. Senate predictions from a year ago are--shall we say--inoperative however.  I had Republicans getting to 54 seats with a gain of seven, but now--thanks to unforced errors in Indiana and Missouri and Olympia Snowe's decision not to run in Maine--Democrats seem likely to hold control 51-49.  Kaine has pulled ahead of Allen in Virgina; Nelson leads Mack in Florida; only one poll (UMass) has Scott Brown ahead in Massachusetts; and even with a Josh Mandel upset in Ohio and wins in North Dakota, Nebraska, Montana (no sure thing) and Wisconsin (no sure thing), Republicans, thanks to the right wing fanaticism in primaries again, will come up short.

Where's Vince McMahon to rig things when Linda needs him?

Comments

[Joshua]

The 12 for '12 took a bit of a beating today. Well, at least 2 lessons can be gleaned: don't put too much stock in Rasmussen and don't dismiss PPP.

[Rep Steve Vaillancourt]

Hey, I was 24 for 24. I called all 24 State Senate races correctly, and my NH House predictions were among the most accurate in the state. The 12 for 12 had already taken a hit when Gatsas failed to run. I did very well in US House races, very badly in Senate races. Keep in mind,t he 12 for 12 were made last December and updated throughout the year.

[Joshua]

Steve, there was such a deluge of information last night that I neglected to look at your race until today. As far as I could determine from online sources, you are narrowly ahead right now. Is it settled or will it go to recount? I really hope that this works out and that you continue in the House for another 2 years. Although it appears that the Dems will be running the chamber, that can't be any worse than working under O'Brien.

[Rep Steve Vaillancourt]

Yes; there will be a recount, but with machine counts (such as in Manchester and most places in the state), usually only one or two votes change. I'm up by 15; if you are anyone would like to see how it works and help out as an observer, email me at steve.vaillancourt@leg.state.nh.us. I'm allowed to have people watch the recount. It's not terribly exciting, like watching grass grow, but you get to see how cast ballots look.