Thanks for the ping.
This is a poll of registered voters (over half of whom don’t vote in midterm elections, with young people, minorities, soft Democrats and independents being the likeliest not to vote) that oversamples Democrats and in which supporters of each of McConnell and Bevin have every incentive to claim that they won’t be voting for the other Republican in a matchup against Alison Lundergan Grimes (or Grimey, as she prefers to be called), given that supporters of each candidate want to project the notion that only their guy can defeat Grimey. If this was a poll of likely voters that matched the electorate that we’ll get this November, both McConnell and Bevin would be leading Grimey. Absent a big screw-up by the GOP nominee (which I humbly suggest is something that would be a hundred times likelier to occur with the untested, unknown Bevin as the nominee than with McConnell), Grimey can’t win: she’s a Democrat running for the U.S. Senate in KY in 2014, and so far has not proven to have very good campaigning skills.
McConnell would get 53%+ against Grimey; Bevin would get somewhere between 45% and 55%, depending on a large number of factors. Grimey certainly would have a much easier time “defining” Bevin in voters’ eyes than she would at trying to define McConnell.
I'm not so sure... The Rasmussen poll linked above (post 8) was Likely Voters. It showed McConnell and Grimes tied. Bevin was slightly ahead of Grimes.
But, I agree that the numbers are skewed now, for all the reasons you stated. I'm not worried about my state voting Republican in a State wide election... barring, as you say, an "Akin-like" self-inflicted wound. Which, yes... would be more likely from Bevin than from McConnell. But, we've got 3 months to go. Let's see how the campaigning goes.