Using the time history chart from 538.
On 23rd August, Obama lead by 54.6 Electoral College Votes (ECV) in 538's prediction. Since then there have been 4 inflection points, when the trend changed. I have tried to guess (at a high level) what caused these inflections. Given the nature of polls and Nate's methods, it takes a few days for an event to show up in the 538 predictions.
Date |
Obama lead |
Average Daily change |
Triggering Event?
|
8/23 |
54.6 |
--- |
Baseline
|
9/8 |
95.8 |
+2.6 |
Conventions
|
9/19 |
72.8 |
-2.1 |
Convention bounce fades + Benghazi?
|
10/5 |
97.4 |
+1.5 |
Trend resumes?
|
10/13 |
32.8 |
-8.1 |
Debacle in Denver
|
11/1 |
58.8 |
+1.9 |
2nd debate
|
The long term trend has been for the President to increase his predicted margin of victory by about 2 ECV per day. Benghazi and the first debate have interrupted this trend (especially the latter).
The best guess would be that the trend until Tuesday will revert to the long term mean. Absent some inflection-causing event, the President ought to win by about 59 + 5*2 = 69 ECV. Since that's not possible (margin of victory has to be an even number), call it 70. That would make it Obama 305, Romney 235.
The wild card is whether or not Sandy was in fact another inflection-causing event, but in the President's favor. Florida is very close, and people in Florida have some empathy regarding hurricanes. And Romney's epic floundering and flailing regarding FEMA can't have helped.
The conclusion is that what Team Obama (which probably includes YOU) has been doing is effective. Keep on keeping on, and a conclusive victory will be achieved.