Checking in.
I just finished updating my election forecast models with today's poll releases.
President: 200 Electoral College votes, 14% probability of winning.
Senate: 49 seats, 29% probability of majority.
I know the polls are crap, but I am interest in the weekly changes. Since Saturday:
President:
- Nationally, President Trump's probability of winning increased from 4.6% to 6.2%.
- Arizona: Increased from 25.4 to 31.1.
- Florida: Decreased from 36.5 to 35.2.
- Georgia: Decreased from 65.9 to 61.5.
- Michigan: Decreased from 18.5 to 14.9.
- Minnesota: Increased from 0.3 to 1.0.
- North Carolina: Increased from 41.4 to 44.4.
- Pennsylvania: Decreased from 14.5 to 8.9.
- South Carolina: Increased from 98.4 to 99.9.
- Texas: Increased from 82.3 to 83.3.
- Wisconsin: Increased from 6.9 to 9.9.
In the Senate:
- Arizona 3 (McSally): Increased from 7.0 to 12.4.
- Georgia 2 (Perdue): Increased from 81.2 to 88.7.
- Michigan 2 (James): Increased from 13.1 to 20.8.
- Minnesota 2 (Lewis): decreased from 1.9 to 0.6.
- North Carolina 2 (Tillis): Decreased from 13.8 to 10.7.
- South Carolina 2 (Graham): Increased from 52.7 to 56.9.
- Texas 2 (Cornyn): Increased from 94.7 to 96.2.
I've decided that Quinnipiac polls are outliers. The Perdue poll in Georgia is my clue. I'm excluding them when they are clearly opposite everyone else.
-PJ