Has it really been six weeks since we last polled Virginia?
Research 2000 for Daily Kos. 9/14-16. Likely voters. MoE 4% (8/3-5 results)
Deeds (D) 43 (43)
McDonnell (R) 50 (51)
Creigh Deeds (D)
Favorable 47 (46)
Unfavorable 42 (40)
No Opinion 11 (14)
Bob McDonnell (R)
Favorable 56 (57)
Unfavorable 40 (38)
No Opinion 4 (5)
This is pretty much unchanged, from the toplines (just float within the MoE), to the favorability numbers.
These findings are quite different from Rasmussen's latest Virginia poll, which found McDonnell with a narrow 48-46 lead, and seeming evidence that McDonnell's college thesis had hurt his numbers:
The survey two weeks ago was conducted shortly after The Washington Post ran news stories about a thesis paper written by McDonnell in 1989. The thesis reflected very conservative views on the role of women in society and other topics. Fifty-two percent (52%) of voters now say the writings are at least somewhat important in terms of how they will vote. That’s up from 36% in the previous survey. The number of unaffiliated voters who consider the writings important is up to 47%.
Ironic that Rasmussen, which has worked hard to become the conservatives' favorite pollster this year, shows a tight race, and ours, which is as non-partisan as a pollster can be (no matter how much bias our critics imagine), shows no gains for Deeds.