Political Junkie's Guide To The Elections - Week Ending November 3, 2012 (Key Swings Missing!)

Rasmussen Reports Pollig Data ^ | November 4, 2012 | Political Junkie Too

[Political Junkie Too]

Previous Editions:

October 27, 2012
October 20, 2012
October 13, 2012
August 11, 2012
August 4, 2012
July 21, 2012
Premier - July 4, 2012

Here is the latest update of my Rasmussen state-by-state presidential model.

The model begins with the final results of the 2008 election, and updates those results with current Rasmussen state polls as they are published.

Final 2008 Electoral College results were Obama 365, McCain 173. My model begins with Obama 364, McCain 174, due to not handling the few split-win states.

Furthermore, 2010 apportionment favored Republicans, leaving us with a starting point of Obama 358, Romney 180.

This Edition's Updates:

The Race for the White House

In this final report, the anticipated Romney surge has not happened, at least according to Rasmussen's polls. Obama's Electoral College count remains at 243 Electoral Votes, and Romney's remains at 261, with 34 EV up for grabs. Romney's lead is soft, however, because he has to defend 46 Leaning Electoral Votes to Obama's 6. Obama has a stronger base of 237 Electoral Votes to Romney's 215.

Probabilistically, Romney has 266.9 Electoral Votes (P90 is 292), and a 45.4% chance of winning. If I add a Democrat poll bias adjustment to the model, Romney has 294.9 EV (and a P90 of 317) with an 86.7% chance of winning.

This week, Rasmussen polled Colorado, Indiana, Iowa, Massachusetts, Michigan, Montana, Ohio, and Wisconsin. Absent swing-states in the final week are Florida (R +2%), Nevada (R -2%), New Hampshire (R +2%), and Virginia (R +2%). Perhaps current polls in this state would show Romney pulling away, and would pull the model over the line for Romney? I think so, since this is what the bias adjustment essentially does.

In Colorado, Romney is holding his lead on Obama. Romney remained steady at 50%. Obama gained +1% to 47%. Colorado would move from Strong to lean for Romney, but the MOE moved from 4.5% to 4.0%, keeping Colorado Strong for Romney.

In Indiana, the race tightened in the closing week, but not enough for Obama. Romney lost -2% to 52%, and Obama gained +2% to 43%. Indiana is Safe for Romney.

In Iowa, Romney gained another +1% and Obama held his position, breaking the tie. Iowa is now 49%-48% for Romney, and remains a Toss-up.

In Massachusetts, a re-poll from two weeks shows Obama growing his 57% to 59%, while Romney shrinks his 42% to 40%. Massachusetts is Safe for Obama.

In Michigan, a re-poll from three weeks ago shows Romney gaining +2% to 47%. Obama stays at 52%. Michigan remains Strong for Obama.

In Montana, a re-poll from two weeks ago shows Romney keeping his 53% and Obama losing -2% to 43%.. Montana remains Safe for Romney.

In Ohio, a re-poll from last week shows each candidate gaining +1%, keeping the race tied at 49%. Ohio remains a Toss-up.

In Wisconsin, the fourth week of polls has the race unchanged from last week's 49% tie. Wisconsin remains a Toss-Up.

Below is the latest round of state poll movements. The first number is the change in GOP result from the last poll (or last election if no prior poll), and the second number is the separation from the Democrat candidate's result.

Watch List:

• Arizona: No Poll (R +8%)
• Colorado: -1% (R +3%)
• Connecticut: No Poll (R -6%)
• Florida: No Poll (R +2%)
• Georgia: Never Polled (R +5.2%) - 2008 result
• Iowa: +1% (R +1%)
• Michigan: +2% (R -5%)
• Montana: +2% (R +10%) - drops out of the margin of error
• Nevada: No Poll (R -2%)
• New Hampshire: No Poll (R +2%)
• North Carolina: No Poll (R +6%)
• Ohio: +0% (R +0%)
• Pennsylvania: No Poll (R -5%)
• Virginia: No Poll (R +2%)
• Wisconsin: +0% (R +0%)

Summary of Electoral College breakdown

Obama - 243								Romney - 261
Safe		Strong		Leaning		Toss-Up		Leaning		Strong		Safe
EV	State	EV	State	EV	State	EV	State	EV	State	EV	State	EV	State
55	California	7	Connecticut	6	Nevada	6	Iowa	29	Florida	9	Colorado	9	Alabama
3	District of Columbia	16	Michigan			18	Ohio	4	New Hampshire	16	Georgia	3	Alaska
3	Delaware	20	Pennsylvania			10	Wisconsin	13	Virginia	15	North Carolina	11	Arizona
4	Hawaii											6	Arkansas
20	Illinois											4	Idaho
4	Maine											11	Indiana
10	Maryland											6	Kansas
11	Massachusetts											8	Kentucky
10	Minnesota											8	Louisiana
14	New Jersey											6	Mississippi
5	New Mexico											3	Montana
29	New York											5	Nebraska
7	Oregon											3	North Dakota
4	Rhode Island											7	Oklahoma
3	Vermont											10	Missouri
12	Washington											9	South Carolina
												3	South Dakota
												11	Tennessee
												38	Texas
												6	Utah
												5	West Virginia
												3	Wyoming
194		43		6		34		46		40		175

2008 Final Results

Link to 2008 Final Election Map

Current State Leanings

Link to Current Electoral College Map

Using the most recent Rasmussen polls, the results of 32,000 simulated elections are listed in the table below. The definitions of the columns are:

• "GOP Electoral Votes P10” – the value where the probability of the actual number being lower is 10%.
• ”GOP Electoral Vote EV” – the expected value (mean) or probability-weighted average of the simulation.
• ”GOP Electoral Vote P90” – the value where the probability of the actual number being lower is 90%. Or, it is the value where the probability of the actual number being higher is 10%.
• ”Probability of 270” – The probability associated with value 270 (the Electoral College majority goal).

Week	GOP Electoral College P10	GOP Electoral College EV	GOP Electoral College P90	Probability of 270
2008 Election	180	180	180	0.00%
12-May-12	193	220.73	248	0.16%
19-May-12	197	225.09	248	0.23%
26-May-12	206	232.72	256	1.21%
02-Jun-12	216	242.46	266	4.43%
09-Jun-12	221	247.73	275	14.02%
16-Jun-12	230	257.37	282	28.95%
23-Jun-12	231	257.92	282	30.10%
30-Jun-12	225	254.58	282	24.61%
07-Jul-12	225	254.58	282	24.61%
14-Jul-12	225	254.58	282	24.61%
21-Jul-12	219	248.33	276	16.46%
28-Jul-12	215	244.1	272	11.78%
04-Aug-12	216	245.64	273	12.75%
11-Aug-12	215	245.24	273	12.05%
18-Aug-12	225	255.61	282	25.11%
25-Aug-12	223	252.39	281	21.52%
01-Sep-12	223	252.39	281	22.08%
08-Sep-12	224	252.45	281	21.66%
15-Sep-12	221	249.88	278	17.54%
22-Sep-12	225	252.98	278	20.34%
29-Sep-12	225	252.04	277	19.04%
06-Oct-12	225	253.31	279	21.00%
13-Oct-12	233	258.74	283	27.69%
20-Oct-12	242	264.04	285	35.65%
27-Oct-12	238	265.95	291	44.18%
03-Nov-12	238	266.85	292	45.37%

Link to Probability Chart

And over in the Senate...

The Republican GOP Senate campaign gave up the slight gains from last week, giving back the one seat they picked up. Republicans were looking at a near certainty to take control of the Senate before the conventions.

This is the latest run of my Rasmussen state-by-state Senate model.

The model begins with the final results of the last Class 1 election, and I will update those results with current Rasmussen state polls as they are published.

Final 2010 Senate results were Republicans taking 47 seats, and Democrats caucusing 53 seats. My model begins with here.

As of now, the Senate looks to be a 52-48 Democrat hold, with a probability of 7.4% for Republicans taking over. If I add the same Democrat bias correction as above, the Seneate becomes a 52-48 take-over for Republicans, with an 75% chance of occuring.

Since the last report, Rasmussen polled Connecticut, Florida, Indiana, Montana, Ohio, and Wisconsin Senate races. Again, the absence of swing-state Senate polling may be keeping the Republican count down.

In Connecticut, Republican McMahon lost -2% and Democrat Murphy gained +3%, making the race now a 51%-45% lead for Murphy. Connecticut is a Hold for Democrats.

In Florida, Republican Mack gained +3% from two weeks ago, while Democrat Nelson gained +1%. Florida is a 49%-46% Hold for Democrats.

In Indiana, Republican Mourdock gave up 47%-42% lead over Democrat Donnelly, now trailing by 45%-42%. Indiana is now a Pick-Up for Democrats.

In Montana, Republican Rehberg kept his 48% from last week, but Democrat Tester gained + 1% to break the tie, making Montana a 49%-48% Hold for Democrats.

In Ohio, a fourth week of polling has the Democrat Brown holding onto his 48%, while Republican Mandel gained +4%, making the race a 48% tie.

In Wisconsin, Republican Thompson held onto his 48%, and Democrat Baldwin gained back +2%. The race is now 48% tie.

Below is the latest round of state polls. The first number is the change in GOP result from the last poll (or seat election if no prior poll), and the second number is the separation from the Democrat candidate's result.

Watch List:

• Arizona 1: No Poll (R +6%)
• Connecticut 2: -5% (R -6%)
• Florida 2: +2% (R -3%)
• Indiana 2: -8% (R -3%)
• Massachusetts 1: No Poll (R -2%)
• Missouri 2: No Poll (R -8%)
• Montana 2: -1% (R -1%)
• Nevada 1: No Poll (R +5%)
• North Dakota 1: No Poll (R +5%)
• Ohio 1: +4% (R +0%)
• Rhode Island 2: Never Polled (R -7%)
• Tennessee 2: Never Polled (R +2.7%)
• Virginia 2: No Poll (R -1%)
• Wisconsin 2: -2% (R +0%)

Using the most recent Rasmussen polls, the results of 32,000 simulated elections are listed in the table below. The definitions of the columns are:

• "GOP Senate Seats P10” – the value where the probability of the actual number being lower is 10%.
• ”GOP Senate Seats EV” – the expected value (mean) or probability-weighted average of the simulation.
• ”GOP Senate Seats P90” – the value where the probability of the actual number being lower is 90%. Or, it is the value where the probability of the actual number being higher is 10%.
• ”Probability of 51” – The probability associated with value 51 (the Senate majority goal).
• ”GOP Gain” – The difference between the GOP EV and the current number of seats (47), rounded down.

Week	GOP Senate Seats P10	GOP Senate Seats EV	GOP Senate Seats P90	Probability of 51	GOP Gain
2010 Election	47	47	47	0.00%	0
12-May-12	50	51.24	52	77.01%	4
19-May-12	50	51.24	52	77.01%	4
26-May-12	49	50.75	52	58.53%	3
02-Jun-12	49	50.63	52	54.34%	3
09-Jun-12	49	50.37	52	44.50%	3
16-Jun-12	49	50.37	52	44.50%	3
23-Jun-12	49	50.05	52	34.52%	3
30-Jun-12	49	50.05	52	34.52%	3
07-Jul-12	49	50.05	52	34.52%	3
14-Jul-12	51	52.04	54	91.10%	5
21-Jul-12	51	52.19	54	92.42%	5
28-Jul-12	51	52.17	54	92.18%	5
04-Aug-12	50	51.33	53	75.39%	4
11-Aug-12	50	51.43	53	78.01%	4
18-Aug-12	50	51.76	53	84.33%	4
25-Aug-12	51	52.59	54	95.05%	5
01-Sep-12	50	51.64	53	82.53%	4
08-Sep-12	50	51.64	53	82.46%	4
15-Sep-12	50	51.19	53	73.42%	4
22-Sep-12	48	49.83	51	30.40%	2
29-Sep-12	47	48.79	50	9.26%	1
06-Oct-12	47	48.49	50	4.63%	1
13-Oct-12	47	48.74	50	8.42%	1
20-Oct-12	47	48.62	50	5.51%	1
27-Oct-12	48	49.37	51	19.18%	2
3-Nov-12

Link to Senate Probability Chart

[Political Junkie Too]

In this final week, I really hope that the polls are way off from what will happen on the ground. That said, I wonder what will happen to Scott Rasmussen's reputation as the most consistent pollster if he's this far off from reality.

-PJ

[Political Junkie Too]

Ping.

-PJ

[Political Junkie Too]

Sorry, the Senate history chart messed up the last line for the current week. Below is the corrected chart.

• "GOP Senate Seats P10” – the value where the probability of the actual number being lower is 10%.
• ”GOP Senate Seats EV” – the expected value (mean) or probability-weighted average of the simulation.
• ”GOP Senate Seats P90” – the value where the probability of the actual number being lower is 90%. Or, it is the value where the probability of the actual number being higher is 10%.
• ”Probability of 51” – The probability associated with value 51 (the Senate majority goal).
• ”GOP Gain” – The difference between the GOP EV and the current number of seats (47), rounded down.

Week	GOP Senate Seats P10	GOP Senate Seats EV	GOP Senate Seats P90	Probability of 51	GOP Gain
2010 Election	47	47	47	0.00%	0
12-May-12	50	51.24	52	77.01%	4
19-May-12	50	51.24	52	77.01%	4
26-May-12	49	50.75	52	58.53%	3
02-Jun-12	49	50.63	52	54.34%	3
09-Jun-12	49	50.37	52	44.50%	3
16-Jun-12	49	50.37	52	44.50%	3
23-Jun-12	49	50.05	52	34.52%	3
30-Jun-12	49	50.05	52	34.52%	3
07-Jul-12	49	50.05	52	34.52%	3
14-Jul-12	51	52.04	54	91.10%	5
21-Jul-12	51	52.19	54	92.42%	5
28-Jul-12	51	52.17	54	92.18%	5
04-Aug-12	50	51.33	53	75.39%	4
11-Aug-12	50	51.43	53	78.01%	4
18-Aug-12	50	51.76	53	84.33%	4
25-Aug-12	51	52.59	54	95.05%	5
01-Sep-12	50	51.64	53	82.53%	4
08-Sep-12	50	51.64	53	82.46%	4
15-Sep-12	50	51.19	53	73.42%	4
22-Sep-12	48	49.83	51	30.40%	2
29-Sep-12	47	48.79	50	9.26%	1
06-Oct-12	47	48.49	50	4.63%	1
13-Oct-12	47	48.74	50	8.42%	1
20-Oct-12	47	48.62	50	5.51%	1
03-Nov-12	47	48.49	50	7.35%	1

-PJ

[randita]

Below is a list of competitive or potentially competitive Senate races ranked roughly from the most to least likely GOP win. The rankings might change slightly as the election season progresses.

Note the differentiation between the seats that represent GOP Holds and those that represent opportunities for Gains. Democrats currently control 53 seats in the Senate (including two independents who caucus with the Democrats), Republicans 47. If Republicans hold all the seats they have currently, they will need to gain four more seats to have a majority.

Links to GOP primary winners' websites are included so you can learn more about the candidates and donate, if you feel so inclined.

If you're interested in all polling on Senate races, check out Real Clear Politics Senate Polls.

 

10/28/12 Race for the U.S. Senate Ranked Roughly From Most to Least Likely GOP Win
State	Republican Candidate	Democrat/IND Candidate	Hold OR Gain
NE*	Debra Fischer	Bob Kerrey	GAIN
ND*	Rick Berg	Heidi Heitcamp	GAIN
WI*	Tommy Thompson	Tammy Baldwin	GAIN
NV	Dean Heller+	Shelley Berkley	HOLD
IN*	Richard Mourdock	Joe Donnelly	HOLD
AZ*	Jeff Flake	Richard Carmona	HOLD
MT	Denny Rehberg	Jon Tester	GAIN
MA	Scott Brown+	Elizabeth Warren	HOLD
VA*	George Allen	Tim Kaine	GAIN
OH	Josh Mandel	Sherrod Brown+	GAIN
FL	Connie Mack	Bill Nelson+	GAIN
CT*	Linda McMahon	Chris Murphy	GAIN
PA	Tom Smith	Bob Casey+	GAIN
MO	Todd Akin	Claire McCaskill+	GAIN
MI	Pete Hoekstra	Debbie Stabenow+	GAIN
NM*	Heather Wilson	Martin Heinrich	GAIN
HI*	Linda Lingle	Mazie Hirono	GAIN
ME*	Charles Summers	Angus King (Ind.)	HOLD
*Open Seat          +Incumbent

 

[djpg]

I have been watching these polls anxiously for months now and this is what I have learned: There will be an election Tuesday and either Obama or Romney will win it. Has anyone learned anything else?

What a waste of time.

[Political Junkie Too]

We'll see what we learn on Tuesday.

If all the swing states come in at 52-48 or 51-49, then we'll know that the polling was right.

If they come in at 55-45 or wider, then we'll know that the polling was meant to astroturf Obama's popularity.

-PJ

[newheart]

What a waste of time.

Are you kidding? Tell that to the political operatives on both sides who have to have some means of making decisions regarding the allocation of their resources in order to win. Trust me, at this level, a national campaign cannot simply act on its "gut." I don't disagree that polling is never an exact science but it is incredibly useful even though all poll results should be taken with a grain of salt.

PJ, thanks for all the hard work. It has been very insightful over the weeks. Now we just get to wait a couple of days to see how right or wrong the polling community was.

[verum ago]

In this final week, I really hope that the polls are way off from what will happen on the ground. That said, I wonder what will happen to Scott Rasmussen's reputation as the most consistent pollster if he's this far off from reality.

I'm pretty confident about the presidential election at this point, though not enough to stop me from being a train wreck all day Tuesday.

Your comment about Rasmussen aroused some worry in me about his reputation. He's really a darling to a lot of people on FR, hasn't he? But it went away with a little more thought. That's not an attitude we should have toward any pollster. If the election costs him his reputation, so be it, as if it happens then he'll deserve it just like the rest, won't he?

I'll take a dominant Romney win that hurts Rasmussen over a weak win (or loss) that keeps up Rasmussen's reputation.

• • •

Two more days!

[djpg]

It is fine for political professionals setting strategy day by day. They have to know what seems to be working and what is not. Of course they cannot work “ on their gut.” They have their own private polls for those purposes. But what good does it do us to watch polls jump around for weeks only to discover that they are inconclusive a few days before an election?

[Political Junkie Too]

Fool me once, shame on you. Fool me twice, shame on me.

Just like how the MSM knows that we're on to their schemes, we'll soon see if the pollsters need to learn the same lesson.

Are they reflecting sentiment or trying to create sentiment that isn't there?

-PJ

[ding_dong_daddy_from_dumas]

If I add a Democrat poll bias adjustment to the model, Romney has 294.9 EV (and a P90 of 317) with an 86.7% chance of winning.

I think you already know this, but that may be moving the "bias" too far in the other direction. Maybe this election will come out somewhere in between the 2008 and 2010 results. It's easy for conservatives to believe that O will not get anything beyond the 20% nutcase vote, but we have a lot of ignorant voters out there.

I am confident you know much more about the details than I do.

[Political Junkie Too]

I treat the bias as an uncertainty that varies during the simulation, rather than as a constant. It ranges from zero to 2.5%,with a likely value of 1%. I think this will generate an adjustment that is still lower than the amount that people say the polls are skewed by.

-PJ

[ding_dong_daddy_from_dumas]

I think this will generate an adjustment that is still lower than the amount that people say the polls are skewed by.

Sounds reasonable. This is a very strange election, and more surprises may be coming. A real challenge for pollsters.

"God created pollsters to make astrologers look good." - John Kasich