And The Winner Is...

About a week ago, the Lord of Truth asked for some election prognostication from yours truly.  I think this is mostly so he can mock me for being wrong (let it be noted that I'm still catching flak for a missed question during a Greek Week quiz bowl event 15 years ago... appropriately so, I might add). 

Let's start with the caveat that other people do this for a living and do a better job of it.  I would start with Nate Silver over at the New York Times, not because I agree ideologically with Silver, but because his model was damn near perfect in 2008, he's not afraid to take counter-intuitive positions, and the analysis is fair even if it tilts to the left.  Plus, his site compiles a ton of good information that's helpful to analysis.  In particular, I recommend the Senate and Governor's race tables if you're into this stuff.

Silver also has a useful hour-by-hour guide to tracking the House races, which would be great for use in designing a drinking game.

A similarly detailed analysis (including race-by-race stuff for the House) from Patrick Ishmael is found here.  Obviously, he reaches different conclusions than Silver.  So does Jim Geraghty, whose prediction of a net gain of 70 seats for the GOP would go pretty far down the line to making me break out some celebratory Scotch tomorrow night.  Silver posits a pretty wide confidence interval on the House, and I don't blame him one bit. 

So what does this all mean?  It means that predictions are a game for suckers, and the only guys whom you should trust to make an accurate prediction would be the gys like Silver, Ishmael, and Geraghty, who devote real time and analysis to what they're doing.  And since they're still not sure, anyone else who can claim they have a good prediction for tomorrow is either lying or a fool.

On the other hand, one could just guess...

 

House:

Me trying to do a race-by-race analysis in the House would be pretty useless, like asking Andy Reid why he doesn't run the ball more.  Most of the forecasters seem to be settling around GOP +50-55 or something, but it keeps sliding up.  I'll go with GOP +59, but with a couple caveats:

1.  I pulled that number out of thin air.  Seriously.  I might revise the estimate tomorrow by having my three year old take a guess.

2.  My guess is that if the GOP outperforms to where its gains go above +59, it might be in the mid-70's somewhere before the Republican "wave" stops.  Likewise, if the GOP underperforms in the House and drops below +50, I think it's more likely to end up right around +40 than the mid-40's, which would lead to some serious conflict with a one-seat GOP House majority.  The Gallup generic ballot numbers are really important to House races, because House districts are gerrymandered to be safe for one party; once the other party gets a big lead in the generic ballot, that means the lead can counter the gerrymandering, particularly when one side is more enthused and if there are no races at the top of the ballot or would break the same way. 

3.  That number is a net number -- the GOP will lose about 4 seats (my guess is that they drop only two if the wave turns into the type of wave that almost killed Greg Brady during the surfing contest).

4.  I'm avoiding district-by district analysis, although I am joining all decent Americans in hoping Alan Grayson loses and loses big.  If the seat in my district flips from Dem (Connelly) to GOP (Fimian), then I may need to raise a few extra glasses in celebration, because that may be an indicator that the wave is going to be big.

Senate:

I'm going to focus only on races where the seats had any chance of flipping this cycle.  I'm following Silver's model and using percentages, but will go more in detail below:

LA: GOP Hold (Vitter wins); 100% likely

AZ: GOP Hold (McCain wins); 100% likely

OH: GOP Hold (Portman wins); 100% likely

ND: Switch from Dem to GOP (Hoevan wins); 100% likely

AK: Switch from Dem to GOP (Boozman wins); 100% likely

IN: Switch from Dem to GOP (Coats wins); 100% likely

DE: Dem Hold (Coons wins); 99% likely

MO: GOP Hold (Blunt wins); 99% likely

NH: GOP Hold (Ayotte wins); 95% likely

NC: GOP Hold (Burr wins); 95% likely

KY: GOP Hold (Paul wins); 95% likely

CT: Dem Hold (Blumenthal wins); 95% likely

WI: Switch from Dem to GOP (Johnson wins); 95% likely

FL: GOP Hold (Rubio wins); 95% likely

AK: GOP Hold (Miller/Murkowski wins); 90% likely

PA: Switch from Dem to GOP (Toomey wins); 90% likely

NV: Switch from Dem to GOP (Angle wins): 75% likely

CA: Dem hold (Boxer wins); 75% likely

IL: Switch from Dem to GOP (Kirk wins); 70% likely

CO: Switch from Dem to GOP (Buck wins): 60% likely

WV: Dem Hold (Manchin wins); 55% likely

WA: Dem seat: Toss-up

GOP +5 is pretty much guaranteed, and I feel pretty good about a +7, with +9 very much in play.  +10 might be asking too much (they would need everything to break as above, win Washington, and get a flip on either West Virginia or California), although that's what they need to swing control. 

Some personal thoughts on individual races:

Florida -- The Dems are so desperate to stop Marco Rubio that they've been trying to cut their own candidate off at the knees to get Dems to vote for Republican-turned-Independant Charlie Crist.  The reason?  Young Hispanic Republican conservative from a key state, with a great lifestory, spurred on by the Tea Party, and without the baggage of an Angle or O'Donnell or even Joe Miller in Alaska.  And they haven't figured out a way to stop him.  Joy.

Wisconsin: Pat Toomey was my favorite GOP candidate of this cycle because he's from PA, ran the Club for Growth, and had the balls to take on Specter back in 2004, but Ron Johnson is another Tea Party star.  Loved his ad where he noted that there were tons of lawyers in the Senate, but no manufacturers. 

Pennsylvania: This is one of the key early races -- if Toomey is comfortably ahead, then the wave in the House might be unstoppable.  If Sestak somehows beat him, the Dems might be able to stop the wave from hitting Nevada, Illinois, and Colorado.  But if Toomey wins, the GOP is pretty well set for at least a 7-seat pickup, and the Dem firewall is in West Virginia.

Nevada: Please, beat Harry Reid.  Illinois, Pennsylvania, and Nevada have me nervous because of the chance of voter fraud.  The dead are a big component of the electorate in each of those states.

Alaska: Please, beat Lisa Murkowski.  Write-in establishment candidates who got their job from their daddy and claim to be doing it "for the people" should be deported to Canada.

Illinois:  See my voter fraud point, and multiply because it's Obama's old seat and it's Illinois.  The dead not only vote there, they each vote multiple times. 

Colorado: If Ken Buck wins (and he should), the GOP probably gets to at least +8, and I'm expecting Washington's vote count to be so tight that we won't know the result Tuesday night.  Which means...

West Virginia:  If Manchin somehow loses, the GOP will win Colorado, Illinois, and Nevada, and will need only Washington or California for a majority. And even if they don't get either one... Joe Lieberman instantly becomes the most powerful man in the Senate (which would be split 50/50).

Delaware:  I think Christine O'Donnell has no shot.  It is a shame, because if this seat was in play, the GOP would have a really good shot at Senate control.  Plus, if she's really a witch, she could turn Democratic Senators into toads.

Connecticut: If this race is close or Blumenthal loses, the Dems are toast.  And we get to have WWE wrestlers probably working on McMahon's staff.  Win-win, America.

California:  I doubt Fiorina pulls this off, but she's still close enough that there's a real chance.  But California's so crazy that I'd discount the chances of an upset.

Governors

Pennsylvania will flip tp the GOP.  So will Michigan, Wisconsin, Iowa, Illinois, and Ohio.  Minnesota will buck the Big Ten trend by flipping the other way, and so will California and Connecticut.  Tennessee will flip to GOP, New Hampshire to Dems, New Mexico to the Republicans. 

And Florida will be tight, but will stay GOP after a recount.  Hey, we need at least one, so why not in Florida?