Karl Rove + New York Times (Update 2)

On Oct. 2 and Oct. 7, I wrote about Karl Rove’s U.S. Senate predictions and then juxtaposed his with the NY Times/Nate Silver model.

On Oct. 2, Rove had the Dems losing 10 seats, the GOP gaining six, and four as toss-ups.

I examined the four toss-ups using the NYT model.  Again, at that time (Oct. 2), the model predicted that the Dems would retain the majority in Senate with 51 seats to the GOP’s 49.

On Oct. 2, West Virginia’s Joe Manchin (D) had a 63% chance of winning, according to the NYT.  But on Oct. 7, John Raese moved to a 64% probability of winning, meaning if everything else remained the same, the Senate would be split 50-50.

What does the NYT say about those four toss-ups on the eve of the election?

Washington:
Oct. 2
Murray (D) – 50%
Rossi (R) – 47%
Probability Murray wins – 74%

Nov. 1
Murray – 52%
Rossi – 48%
Probability Murray wins – 84.7%

MURRAY IMPROVES BY 11%

Nevada:
Oct. 2
Reid (D) – 48%
Angle (R) – 48%
Probability Angle wins – 55%

Nov. 1
Reid – 47%
Angle – 50%
Probability Angle wins – 82%

ANGLE IMPROVES BY 27%

Illinois:
Oct. 2
Giannoulias (D) – 48%
Kirk (R) – 47%
Probability Kirk wins – 54%

Nov. 1
Giannoulias – 48%
Kirk – 49%
Probability Kirk wins – 61%

KIRK IMPROVES BY 7%

West Virginia:
Oct. 7
Manchin (D) – 48%
Raese (R) – 50%
Probability Raese wins – 64%

Nov. 1
Manchin – 50%
Raese – 47%
Probability Raese wins – 23%

MANCHIN FLIPS PROBABILITY, IMPROVES PROBABILITY BY 71%

If this holds, the Senate will be 51 – 49, in favor of the Democrats.

Let’s see how the model holds up.