Poll: Name the winner in the great Georgia U.S. Senate runoff

[Sam Spade]

With early voting over, here are the numbers:

Topline Numbers
491,540 voted - 388,222 voted in person; 103,318 voted by mail (presumably more mail-in ballots will arrive before or on election day)

Demographics
74.37% White
22.59% Black
3.04% Other

Top 5 Counties in Turnout
Cobb: 37,499
Dekalb: 32,891
Fulton: 31,194
Gwinnett: 23,638
Forsyth: 16,704

http://sos.georgia.gov/elections/earlyvotingstats08_runoff.htm

[Sam Spade]

Dekalb was +59% for Obama
Fulton was +35% for Obama
Gwinnett was +10% for McCain
Cobb was +9% for McCain
Forsyth was +58% for McCain

Oh, I know how these places vote alright. Just wondering about the number of early voters.

To actually answer your question, instead of ADD stuff...

http://sos.georgia.gov/elections/election_results/2008_1104/Earlyvotingstats08.htm

Fulton: 179,764
Dekalb: 173,069
Cobb: 149,869
Gwinnett: 117,366
Henry: 70,204

[Sam Spade]

To actually answer your question, instead of ADD stuff...

http://sos.georgia.gov/elections/election_results/2008_1104/Earlyvotingstats08.htm

Fulton: 179,764
Dekalb: 173,069
Cobb: 149,869
Gwinnett: 117,366
Henry: 70,204

Beyond the obvious collapse in the overall figures, I note the lack of Forsyth on that list, and it's presence on the runoff one... and Fulton being only third and Cobb being first...

Meaning?  Well, I know what it means to me... 

[Sam Spade]

I'm gonna predict Martin because no one else is, and if he wins, everyone will think I'm some sort of genius and if he loses I can just never post in this thread again.

Same here.

When one genius comes up with the same long-shot result, does this increase the geniusness aspect of it? 

I'm going to play a little on the edges here and say Chambliss by 11.

[Sam Spade]

Polls close when?

If it's Georgia, 7:00 PM EST

[Sam Spade]

PPP estimates that today's Election Day voters are more friendly to Martin than the Early Voters.
Chambliss leads 58-41 among the 35% who said they have already voted. But Chambliss leads "only" 51-48 among Dec. 2 voters.

That's a reversal to what we have seen on or before November 4. Early voters used to favor Martin, but Nov. 4 voters broke heavily for Chambliss.

I still think Chambliss will win though.

This is not big news.  In other words, things are returning to normal.  Of course - MOE of those things...

[Sam Spade]

I see that the AA percentage ended up being 21.80% in early voting.  My 11% margin (given that blacks should show up in higher percentages today, but not *that* much) may be an underestimation.

[Sam Spade]

Are we going to have an "official" results thread or just use this one?

I don't see any reason to have a special thread.

http://hosted.ap.org/dynamic/files/elections/2008/by_state/GA_Page_1202.html?SITE=AP

[Sam Spade]

Polls have closed.  No clue when we'll start getting results.

[Sam Spade]

One precinct from a Metro Atlanta exurban county is in (Henry):
Chambliss 5,861     
Martin 2,473 

Might be early voting numbers.

[Sam Spade]

Those Bacon results have got to be wrong. Turnout did not actually fall by 99.9%. 4000 votes to 18?

Probably misentry.  We'll see.

[Sam Spade]

Georgia is not a place you can predict off of two precincts here and there.

[Sam Spade]

Folks, you have to understand that some of this weird movement occurs as black/white precincts come in, bumping the numbers around.

[Sam Spade]

Echoing Sam: People stop acting as if the counties are uniformly distributed please

It's the very worst way to analyze this.  Wait for the counties to be completely reported!

That being said, as I stated before, I wouldn't be surprised with a Chambliss 10+ victory.

[Sam Spade]

I want to make sure there's no unknown incredible turnout in metro Atlanta that I missed, b/c otherwise this one's quite over and it will not be close.  Northern Georgia is a real massacre - you'd think Martin's name starts with an O.

[Sam Spade]

BACON (100%)
Chambliss 1,402 80% (+10%)
Martin 358 20% (-8%)
Margin: Chambliss +18%

[Sam Spade]

Unfortunately, got to go.

But this one's over with barring something strange from DeKalb/Fulton.

[Sam Spade]

You know, I was half-tempted to make the 15-point Chambliss call this afternoon because it felt like that type of race.  Decided to stay conservative with the calls - such is life.

[Sam Spade]

You know, I was half-tempted to make the 15-point Chambliss call this afternoon because it felt like that type of race.  Decided to stay conservative with the calls - such is life.

Have you ever not made a prediction error because of being too conservative in your call?

Oh, dear God, no.  I've made of errors with *out there* predictions.  In fact, I'm made plenty of errors period, regardless of what CfromNJ might say.

[Sam Spade]

Did anyone actually believe Martin would win after failing to do so on election day besides Lief?

I thought Obama's previously amazing turnout machine had an inside gut draw (1-8 is what I said, but I should have given it worse odds) at being transferable to a non-Obama candidate. 

The odds were more accurately about 1 in 1 million.

I want to see what the turnout in blacks was, but I can't see the Obama turnout effect among these folks being more than 5-7 points max (probably less than 5 points actually considering all the undervotes).  As for youths, maybe a couple of points?  Others?  The Republican/Democrat difference among these folks is usually not great.

The rest of the movement has to do with strong Republican turnout in the GA suburbs in my book.  And also that Martin looks like O in the white areas, which may just be turnout, but could also be something more problematic.

[Sam Spade]

There's really only one surprising feature to last night's result; Martin's all-too-obvious collapse with rural white voters.

No kidding.  I thought his name started with an O looking at the results there.  Something to watch for this upcoming Saturday in LA-04 in my book.

[Sam Spade]

Is there anyway to tell whether the shift in rural areas was due to voters shifting from Martin to Chambliss or rather Martin's voters simply not caring enough to show up?

You'd have to go precinct by precinct, and even that wouldn't be 100%.  Still, if there are Martin voters in these rural areas (and I'm thinking white people - look up north) who didn't show up for the runoff, how can we be so sure that they're going to show up to any other races in the next few years?  After all, this runoff's turnout was quite similar to mid-term turnout in 2006.

It's kind of like six of one and half-dozen of the other - and another reason why I'm interested in LA-04.