Ohio Polls: Looking At the Numbers, Here is How Obama Could Win This Swing State
We did the Virginia poll breakdown, now it's off to the Buckeye State. Remember, we are not paying attention to who leads and margin of error and instead we applying the polling data to possible “likely voter” turnout models in a new way to look at polling numbers, especially the internal numbers from polls. The following three latest Ohio polls, Survey USA, Gravis, and CBS/NY Times offer insight into voting characteristics and how it breaks down according to party affiliation. Ohio is a little more volatile than the Virginia polling data, but all the data is consistent across all three polls in terms of party affiliation relating to voting tendencies.
Ohio shows a greater amount of undecided voters which makes modeling even harder and there is another poll, the Ohio Poll, which actually gives President Obama a lead in the independent category and using that poll would undermine the consistency characteristics we are looking for so we left it out. Ohio shows some movement possibilities which are not evident in the Virginia polls. Those movement characteristics are likely to be positive to President Obama this close to the election with about 11% of independents undecided and if they stay undecided that pushes up Democratic advantage. Let's take a look at the percentage breakdowns.
Gravis
Romney
Obama
Rep
85
14
Dem
16
83
Ind
53
41
Survey USA
Romney
Obama
Rep
90
6
Dem
8
88
Ind
48
37
CBS/NY Times
Romney
Obama
Rep
91
7
Dem
6
92
Ind
49
43
Average
Romney
Obama
Rep
88.67
9.00
Dem
10.00
87.67
Ind
50.00
40.33
Now let's see how the weighted polls breakdown
Ohio Poll Weighted Breakdown
D+3
Weight
Romney
Obama
Rep
33
29.26
2.97
Dem
36
3.60
27.18
Ind
31
15.50
12.50
48.36
42.65
D+4
Weight
Romney
Obama
Rep
33
29.26
2.97
Dem
37
3.70
32.44
Ind
30
15.00
12.10
47.96
47.51
D+2
Weight
Romney
Obama
Rep
33
29.26
2.97
Dem
35
3.50
30.68
Ind
32
16.00
12.91
48.76
46.56
D+6
Weight
Romney
Obama
Rep
30
26.60
2.70
Dem
36
3.60
31.56
Ind
34
17.00
13.71
47.20
47.97
We can breakdown the weights anyway we want to, but the ideal scenario for President Obama is the D+4 or D+6 scenarios with the likelihood that undecideds remain undecided pushing up the weights for the Democrats and Republicans in the final tally. Let's take a look at what happens in such a scenario when undecided voters stay undecided to a certain degree.
Ohio Undecideds Stay Undecided
D+3
Weight
Romney
Obama
Rep
38
33.69
3.42
Dem
41
4.10
35.94
Ind
21
10.50
8.47
48.29
47.83
D+4
Weight
Romney
Obama
Rep
37
32.81
3.33
Dem
41
4.10
35.94
Ind
22
11.00
8.87
47.91
48.15
D+2
Weight
Romney
Obama
Rep
38
33.69
3.42
Dem
40
4.00
35.07
Ind
22
11.00
8.87
48.69
47.36
D+6
Weight
Romney
Obama
Rep
35
31.03
3.15
Dem
41
4.10
35.94
Ind
24
12.00
9.68
47.13
48.77
What looks to be a potentially opportunity for Romney in Ohio is really moot if he can't close the deal and favors President Obama in the end. It also looks to be the only way Obama receives more 47% of the vote. It's a possibility undecided voter votes for the incumbent or they do not vote at all rather than voting for the challenger at this point in the election cycle. Two of those 3 scenarios favor Obama. For a battleground state, D+4 is certainly a model one can use and play with the weights to see gains and losses, but as you approach 50 with the weights the greater the likelihood the candidate on the plus side will win assuming he has a vast super-majority support in his own party and Obama does.
While polling data is not infallible, if you start seeing consistency in the data across polls then averaging those internals according to party affiliation and applying a range of turnout models show what likely has to happen on election day. Ohio is interesting because it still has polls which show outlier data and being mindful of those outliers can show you a direction the race can head towards.
President Obama likely has to have 37% or better Democratic turnout to win Ohio, but Mitt Romney likely only needs to keep that gap within D+3 with Republican turnout to win Ohio as long as Democratic turnout is less than 43%.
