Poll Analysis: Final Predictions

Obama	Romney
98.9% probability of winning	1.1% probability of winning
Mean of 311 electoral votes	Mean of 227 electoral votes

Huh. Four years ago at this time, I was frantically entering the eleven new polls that had come out on election day. Today…not so much. We get only one rather inconsequential poll, although because it is Maine, and the Congressional districts are reported, we get three new polls for the price of one.

		start	end	sample		%	%	%
st	poll	date	date	size	MOE	O	R	diff
ME	Maine PRC	01-Nov	03-Nov	905	3.3	53.3	42.2	O+11.1
ME1	Maine PRC	01-Nov	03-Nov	469	—	56.7	39.0	O+17.7
ME2	Maine PRC	01-Nov	03-Nov	436	—	49.7	45.7	O+4.0

As a consequence, this analysis differs little from yesterday’s analysis.

After 100,000 simulated elections, Obama wins 98,946 times and Romney wins 1,054 times (including the 180 ties). Obama received (on average) 311 (+2) to Romney’s 227 (-2) electoral votes. In an election held now, Obama would have a 98.9% (+0.1%) probability of winning and Romney would have a 1.1% (-0.1%) probability of winning.

My prediction: Obama wins. It’s almost certain.

Here’s our look back over the race from a series of elections simulated every seven days using polls from 06 Nov 2011 to 06 Nov 2012, and including polls from the preceding seven days (FAQ).

The interesting thing is the very ragged, multimodal distribution of electoral votes seen in the graph below. The single most likely outcome in this race is an Obama victory with 303 electoral votes. There is a 9.2% probability of that happening.

Then it jumps to 332 electoral votes, with a 6% probability. And then to 318 electoral votes with a 3.8% probability. And so on.

The raggedness of the electoral votes distribution reflects that there are a non-trivial number of important states with large uncertainty. Florida is Romney’s but with only a 63% probability. Iowa is Obama’s but with only an 84% probability. North Carolina is in Romney’s column, but with a 74% probability, and Virginia goes to Obama, but with a 78% probability.

Ten most probable electoral vote outcomes for Obama (full distribution here):

• 303 electoral votes with a 9.19% probability
• 332 electoral votes with a 5.94% probability
• 318 electoral votes with a 3.80% probability
• 304 electoral votes with a 3.78% probability
• 290 electoral votes with a 3.49% probability
• 319 electoral votes with a 3.30% probability
• 297 electoral votes with a 2.97% probability
• 314 electoral votes with a 2.30% probability
• 312 electoral votes with a 2.23% probability
• 333 electoral votes with a 2.11% probability

After 100,000 simulations:

• Obama wins 98.9%, Romney wins 1.1%.
• Average (SE) EC votes for Obama: 310.7 (19.3)
• Average (SE) EC votes for Romney: 227.3 (19.3)
• Median (95% CI) EC votes for Obama: 308 (276, 347)
• Median (95% CI) EC votes for Romney: 230 (191, 262)

Each column of this table shows the electoral vote total aggregated by different criteria for the probability of winning a state (Safe=100%, Strong=90%+, Leans=60%+, Weak=50%+):

Threshold	Safe	+ Strong	+ Leans	+ Weak
Safe Obama	159
Strong Obama	124	283
Leans Obama	20	20	303
Weak Obama	0	0	0	303
Weak Romney	0	0	0	235
Leans Romney	46	46	235
Strong Romney	129	189
Safe Romney	60

This table summarizes results by state. Click on the poll count to see the individual polls included for the state.

	0	0	EC	#	Total	%	%	Obama	Romney
	8	4	Votes	polls	Votes	Obama	Romney	% wins	% wins
AL			9	1*	404	39.6	60.4	0.2	99.8
AK			3	0*				(0)	(100)
AZ			11	1	1069	46.5	53.5	5.1	94.9
AR			6	1*	571	34.9	65.1	0.0	100.0
CA			55	3*	2986	57.3	42.7	100.0	0.0
CO			9	5	4007	52.1	47.9	97.3	2.7
CT			7	1	1183	56.7	43.3	99.9	0.1
DE			3	0				(100)	(0)
DC			3	1*	1173	91.6	8.4	100.0	0.0
FL			29	8	7066	49.7	50.3	37.3	62.7
GA			16	1	1276	46.4	53.6	3.1	96.9
HI			4	1*	1157	64.2	35.8	100.0	0.0
ID			4	1*	563	30.0	70.0	0.0	100.0
IL			20	1	1174	58.2	41.8	100.0	0.0
IN			11	1	570	45.3	54.7	5.7	94.3
IA			6	7	5348	51.0	49.0	84.4	15.6
KS			6	2*	1143	39.4	60.6	0.0	100.0
KY			8	1*	557	42.4	57.6	0.6	99.4
LA			8	1*	2548	37.9	62.1	0.0	100.0
ME			2	3	3005	55.9	44.1	100.0	0.0
ME1			1	2	1250	59.8	40.2	100.0	0.0
ME2			1	3	1700	52.2	47.8	89.7	10.3
MD			10	2*	1538	61.6	38.4	100.0	0.0
MA			11	3	2556	58.2	41.8	100.0	0.0
MI			16	5	4920	51.7	48.3	95.8	4.2
MN			10	3	1953	54.4	45.6	99.8	0.2
MS			6	2*	1420	43.2	56.8	0.0	100.0
MO			10	3	2536	45.1	54.9	0.0	100.0
MT			3	2	1411	45.7	54.3	1.2	98.8
NE			2	1	1119	43.2	56.8	0.0	100.0
NE1			1	1*	389	45.5	54.5	11.1	88.9
NE2			1	1*	646	47.4	52.6	17.4	82.6
NE3			1	1*	284	35.9	64.1	0.0	100.0
NV			6	2	1298	52.5	47.5	90.6	9.4
NH			4	8	6212	51.2	48.8	91.9	8.1
NJ			14	5*	3773	55.6	44.4	100.0	0.0
NM			5	2*	1307	54.8	45.2	99.0	1.0
NY			29	3*	1770	63.8	36.2	100.0	0.0
NC			15	3	2709	49.2	50.8	26.3	73.7
ND			3	4*	2563	42.6	57.4	0.0	100.0
OH			18	9	8649	51.5	48.5	97.9	2.1
OK			7	1*	279	35.8	64.2	0.1	99.9
OR			7	1	903	53.0	47.0	90.6	9.4
PA			20	5	3892	51.6	48.4	91.9	8.1
RI			4	1*	523	62.1	37.9	100.0	0.0
SC			9	3*	4199	48.2	51.8	4.1	95.9
SD			3	1	595	43.7	56.3	1.4	98.6
TN			11	1*	566	36.6	63.4	0.0	100.0
TX			38	1*	752	41.5	58.5	0.1	99.9
UT			6	1	594	26.3	73.7	0.0	100.0
VT			3	1*	415	71.3	28.7	100.0	0.0
VA			13	6	5538	50.7	49.3	77.5	22.5
WA			12	1	923	53.5	46.5	93.5	6.5
WV			5	1*	361	42.1	57.9	1.7	98.3
WI			10	2	2409	52.5	47.5	95.7	4.3
WY			3	0				(0)	(100)

^* An older poll was used (i.e. no recent polls exist).

Details of the methods are given in the FAQ.

The most recent analysis in this match-up can be found from this page.