538 Update Thread

[pbrower2a]

Obama had a huge bump today for some reason, up to a 77.4% chance to win. Maybe because Gallup wasn't in today?

Clock running out.

[Speed of Sound]

Another day, another bump for the President. A slew of polls (and another day off the calendar for the challenger) shifted the % chance of victory to 80.8-19.2, Romney's lowest numbers since the 6th of October. Furthermore, the PV projection is now 50.5-48.5, sending Obama for the first time in weeks back to a full 2 point lead there. VA and CO are at ~65%, IA at 79%, OH at 80%, NV at 87%, WI 91% (and PA 96%).

[Lief 🗽]

Not looking good for Mittbot.

[Niemeyerite]

Another day, another bump for the President. A slew of polls (and another day off the calendar for the challenger) shifted the % chance of victory to 80.8-19.2, Romney's lowest numbers since the 6th of October. Furthermore, the PV projection is now 50.5-48.5, sending Obama for the first time in weeks back to a full 2 point lead there. VA and CO are at ~65%, IA at 79%, OH at 80%, NV at 87%, WI 91% (and PA 96%).

What about NC? (I can't enter the site.)

[Lief 🗽]

Another day, another bump for the President. A slew of polls (and another day off the calendar for the challenger) shifted the % chance of victory to 80.8-19.2, Romney's lowest numbers since the 6th of October. Furthermore, the PV projection is now 50.5-48.5, sending Obama for the first time in weeks back to a full 2 point lead there. VA and CO are at ~65%, IA at 79%, OH at 80%, NV at 87%, WI 91% (and PA 96%).

What about NC? (I can't enter the site.)

80.0% Romney

[Antonio the Sixth]

Finally OH and WI break 80% and 90%, respectively.

[Yank2133]

Another day, another bump for the President. A slew of polls (and another day off the calendar for the challenger) shifted the % chance of victory to 80.8-19.2, Romney's lowest numbers since the 6th of October. Furthermore, the PV projection is now 50.5-48.5, sending Obama for the first time in weeks back to a full 2 point lead there. VA and CO are at ~65%, IA at 79%, OH at 80%, NV at 87%, WI 91% (and PA 96%).

What about NC? (I can't enter the site.)

80% for Romney.

[Antonio the Sixth]

I wonder if Obama's record odds (87.1%) will be beaten come Nov 6. The now-cast is already at 83%...

[Okay, maybe Mike Johnson is a competent parliamentarian.]

My suspicion is that it'll stay between, say, roughly 78 and 85 from here on out.

[Lief 🗽]

Does anyone remember what Obama's odds were the day of the election in 2008?

[sobo]

98.9%

http://web.archive.org/web/20081106113055/http://www.fivethirtyeight.com/

[Speed of Sound]

And on the 30th (closest it gave me to the 1st), it was 95.5-4.5. Understandably, we aren't seeing quite those numbers today.

[DemPGH]

O now at 80.8% national and 303 EVs and Nate may have darkened a couple of the blue ones although not sure. He has more shades than most Looks pretty good there.

[Ty440]

LOL at Nate telling Scarborough to put up  $2000 on the election, Scarborough backed down.

[Gass3268]

10/1/2012

11/6/2012 Projected Electoral Vote - Obama 300.4/Romney 243.4 (Obama +2.8 EC's)
11/6/2012 Projected Chance of Winning - Obama 79%/Romney 21% (Obama +1.8%)
11/6/2012 Projected Popular Vote - Obama 50.5%/Romney 48.2% (Obama Margin +0.2%)

Likely Obama: Iowa (79%), Nevada (87%), New Hampshire (78%) Ohio (80%),
Lean Obama: Colorado (64%), Virginia (66%)  
Tossups: Florida (R-55%)
Lean Romney:
Likely Romney: North Carolina (80%)

Change:

Iowa moved from Lean Obama to Likely Obama
New Hampshire moved from Lean Obama to Likely Obama
Wisconsin moved from Likely Obama to Safe Obama

 % Chance of winning each state:

Projected vote share:

[Starbucks Union Thug HokeyPuck]

Obama - 51%, 332
Romney - 47.5, 206

...and there you have it.  I'm not changing a thing. 

that's what I locked in as my prediction literally 2 mins after the 3rd debate ended.  538 is certainly starting to move towards this map.  Nate has Florida at 55% for Romney and the closest state.  Still think Obama pulls it off. 

[Lief 🗽]

Yup, that map is the thinking man's prediction at this point. The rakish daredevil's map is that + Obama winning North Carolina.

[Antonio the Sixth]

Yup, that map is the thinking man's prediction at this point. The rakish daredevil's map is that + Obama winning North Carolina.

Yeah. I'd say Obama's best-case scenario at this point is winning every State he's favored + FL and NC. Romney's is winning the States where he's favored + CO, VA, IA and NH.

OH and WI might not be completely lost, but they would require a systematic bias in State polls + massive GOP enthusiasm gap + late momentum for Romney. Not exactly the kind of combo you'd like to bet on.

[Ljube]

You must be logged in to read this quote.

Interesting to see that he is worried about his reputation. I don't know if his job is at stake too.

[You kip if you want to...]

Letting the haters get to him

[Filuwaúrdjan]

What's the problem here?

[Franzl]

You must be logged in to read this quote.

Interesting to see that he is worried about his reputation. I don't know if his job is at stake too.

Why should he be worried about his reputation? You understand his projections are entirely based on polling?

[Ljube]

You must be logged in to read this quote.

Interesting to see that he is worried about his reputation. I don't know if his job is at stake too.

Why should he be worried about his reputation? You understand his projections are entirely based on polling?

Yes, I do. And he won't lose reputation in my eyes. I check his blog daily.
But the majority of people don't understand. They will ask him why he was so off. And they will blame him, not the polls.

[Silent Hunter]

You must be logged in to read this quote.

Interesting to see that he is worried about his reputation. I don't know if his job is at stake too.

Why should he be worried about his reputation? You understand his projections are entirely based on polling?

They're also based on factors like approval and economic indicators, but those get a lower weighting as the election approaches.

[afleitch]

It's worth noting that Nate simply analyses the polls. If the polls are systematically wrong, no matter who the pollster is he can miss a race. However polls, on the whole, tend to be accurate. What complicates matters is not necessarily the polls, but what pollsters do with the data.

In 2010, Silver had Sharron Angle with an 83.4% chance to win in Nevada based on the polls that had her ahead. Were the polls wrong? Yes and No.  Reid was up 3 and up 4 on the last YouGov and CNN polls with RV's but the LV model had him down 3 and 4. Infact nearly all the RV data with pollsters had Reid ahead but the LV model had him down.

The other race he called wrong was Colorado. He had Bennet close the gap on his Oct 25 model but then had Buck break away. Again the same thing happened; Bennet was ahead or tied on the RV models with CNN, YouGov, Marist  but behind in the same polls under the LV screen.

So Silver now takes the differences between RV and LV into consideration. Gallup is a prime example of this. He uses Gallup but stuck with their RV figures rather than their LV screen which started to diverge from the mean very early on.