538 Update Thread

[Gass3268]

10/28/2012

11/6/2012 Projected Electoral Vote - Obama 296.6/Romney 241.4 (Obama +1.1)
11/6/2012 Projected Chance of Winning - Obama 73.6%/Romney 26.9% (Obama +1.0%)
11/6/2012 Projected Popular Vote - Obama 50.4%/Romney 48.7% (Obama +0.1%)

Likely Obama: Nevada (80%), Wisconsin (87%)
Lean Obama: Iowa (73%), New Hampshire (72%), Ohio (75%), Virginia (60%)
Tossups: Colorado (O-58%)
Lean Romney: Florida (62%)  
Likely Romney: North Carolina (82%)  

 % Chance of winning each state:

Projected vote share:

[Speed of Sound]

Hopefully this nice, slow movement towards Obama for the last week is a suggestion that the race is  slowly moving towards an 'indifferent' incumbent retention/"well, the challenger's losing" environment.

[HagridOfTheDeep]

Hopefully this nice, slow movement towards Obama for the last week is a suggestion that the race is  slowly moving towards an 'indifferent' incumbent retention/"well, the challenger's losing" environment.

So Obama voters don't feel like the need to come out and vote? Works for me.

[Lief 🗽]

I think, if anything, Obama's slight bump over the past week is a result of Democratic voters finally turning on.

[Gass3268]

10/29/2012

11/6/2012 Projected Electoral Vote - Obama 294.6/Romney 243.4 (Romney +2)
11/6/2012 Projected Chance of Winning - Obama 72.9%/Romney 27.1% (Romney +0.7%)
11/6/2012 Projected Popular Vote - Obama 50.4%/Romney 48.8% (Romney +0.1%)

Likely Obama: Nevada (80%), Wisconsin (87%)
Lean Obama: Iowa (71%), New Hampshire (70%), Ohio (73%), 
Tossups: Colorado (O-55%), Virginia (O-58%)
Lean Romney: Florida (65%)  
Likely Romney: North Carolina (81%)

Change: Virginia moved from Lean Obama to Toss Up  

 % Chance of winning each state:

Projected vote share:

[Gass3268]

Decent day for Romney, but there really wasn't a lot of polling released today. I would guess we'll see less tomorrow.

[Fmr President & Senator Polnut]

More than likely a reflecting of Romney being tied again on Gallup RV and the Ras OH poll

[LiberalJunkie]

A county map I made of Nate's prediction with county trends from 2008.

[Lief 🗽]

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

[Speed of Sound]

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

SUSA had Obama up 3 in OH and CBS/NYT had Obama+1 nationally. My guess is the former is the shifter.

[Lief 🗽]

Yeah, it's the SUSA poll apparently:

You must be logged in to read this quote.

Obama's overall winning percentage (77%) is basically identical to his Ohio win percentage (78%).

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

I imagine the Quinnipiac poll that's being leaked right now will also help in that regard.

[Antonio the Sixth]

Yeah, it's the SUSA poll apparently:

You must be logged in to read this quote.

Obama's overall winning percentage (77%) is basically identical to his Ohio win percentage (78%).

Even closer: actually, it's 77.4% and 77.6%.Obama's margin in Ohio is up at 2.4 from 2.1 yesterday, a pretty sharp improvement (the PPP poll only added 0.1 to it, for example). On the other hand, Obama's national margin is also up sharply, to 1.9 from 1.5 points. That explains why Romney's probability to lose the electoral college while winning the PV declined significantly while Obama's was slightly up.

[Yank2133]

Yeah, he may hit 80% tomorrow.

[Gass3268]

10/31/2012

11/6/2012 Projected Electoral Vote - Obama 299.0/Romney 243.4 (Obama +4.4)
11/6/2012 Projected Chance of Winning - Obama 72.9%/Romney 22.6% (Obama +4.5%)
11/6/2012 Projected Popular Vote - Obama 50.4%/Romney 48.5% (Obama Margin +0.3%)

Likely Obama: Nevada (83%), Ohio (78%), Wisconsin (88%)
Lean Obama: Colorado (61%), Iowa (74%), New Hampshire (75%), Virginia (58%)  
Tossups: Florida (R-59%)
Lean Romney:
Likely Romney: North Carolina (82%)

Change:

Colorado moved from Tossup to Lean Obama
Florida moved from Lean Romney to Tossup
Ohio moved from from Lean Obama to Likely Obama
Virginia moved from Tossup to Lean Obama

 % Chance of winning each state:

Projected vote share:

[Antonio the Sixth]

Yeah, he may hit 80% tomorrow.

He may, but only if there are even better polls coming for Obama tomorrow. Notice that even the now-cast is still under 80%.

Still, very encouraging at this stage.

And Gass, PA and NM are over 90% actually.

[Gass3268]

Yeah, he may hit 80% tomorrow.

He may, but only if there are even better polls coming for Obama tomorrow. Notice that even the now-cast is still under 80%.

Still, very encouraging at this stage.

And Gass, PA and NM are over 90% actually.

NM was at 90%, but thanks for the heads up on PA

[○∙◄☻¥tπ[╪AV┼cVê└]

Yeah, he may hit 80% tomorrow.

He may, but only if there are even better polls coming for Obama tomorrow. Notice that even the now-cast is still under 80%.

Still, very encouraging at this stage.

And Gass, PA and NM are over 90% actually.

NM was at 90%, but thanks for the heads up on PA

The only states that aren't outside of the 95% confidence interval (less than 97.5% odds for both candidates) are NV, CO, IA, WI, OH, VA, NC, FL, PA, NH, plus ME-02 and NE-02.Those districts are probably just underpolled, so we really only need to talk about those 10 states.

Also, seeing as we are only a week from election day, the now cast and the election projection are pretty close.

[Adam Griffin]

Nate Silver has stated before that if any candidate's percentage to win in a particular state is below 60%, then the state would basically be classified as a tossup in his model. Comparatively, if a candidate has a 60% or more chance, then the state can be classified as 'leaning'. Today in the 538, we saw the following:

1) Colorado shifted from toss-up (~58%) to lean Obama (60.7%).

2) Virginia shifted from toss-up (~59%) to lean Obama (61.8%).

3) Florida shifted from lean Romney (63%) to toss-up (59.3%).

Another observation: Obama's chances in Nevada moved above 80% again today, putting it on par with Romney's chances of winning North Carolina.

[Antonio the Sixth]

Yeah, he may hit 80% tomorrow.

He may, but only if there are even better polls coming for Obama tomorrow. Notice that even the now-cast is still under 80%.

Still, very encouraging at this stage.

And Gass, PA and NM are over 90% actually.

NM was at 90%, but thanks for the heads up on PA

The only states that aren't outside of the 95% confidence interval (less than 97.5% odds for both candidates) are NV, CO, IA, WI, OH, VA, NC, FL, PA, NH, plus ME-02 and NE-02.Those districts are probably just underpolled, so we really only need to talk about those 10 states.

Also, seeing as we are only a week from election day, the now cast and the election projection are pretty close.

A 95% probability already means the State is within the 95% interval. If the model gives Romney a 5% chance to win, that means there is also 5% chances that Obama wins by more than what the margin of error indicates.

PA is at 95.4% so it's safe.

[Ty440]

Prediction Obama breaks 80% in the next update.

[Speed of Sound]

Didn't bump much actually, I'm guessing that FL polls dampened things, especially since today's polls were more confirmations of leads rather than growing ones (remember, unlike parts of this site, Silver has had Wisconsin over 85% for a while now. Polls there don't move the needle much). Obama is just shy of 300 EVs in the average now though, at a 1.9% PV lead, and 78% chance of victory.

[Lief 🗽]

Only a 1% winning chance bump. Lame.

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

Only a 1% winning chance bump. Lame.

It's because Silver's been bullish on Ohio and Wisconsin for Obama anyway. Intrade, on the other hand, has been showing a major surge.

[Speed of Sound]

Nate re-ran the numbers at 1:00 am, and now it's:

Obama-300.4 EVs, 79% chance to win, no change in PV (still 50.5-48.6%)

I believe he added the Marist polls. Gravis' 4 point Obama lead in VA isn't in this.