Going into the 2012 cycle, the Republicans had an excellent opportunity to gain in the Senate, with far more Democrats up than Republicans -- and several Democrats who probably won last time thanks mainly to the national tide for the Dems in 2006.
Since then, it's only getting better for Republicans, with the latest today's announcement by Jim Webb of Virginia that he won't seek a second term. Virginia now joins North Dakota, with Kent Conrad retiring, as prime GOP pickup opportunities; Republicans will also have a chance in Connecticut to win Joe Lieberman's seat, while the only GOP retirement to date, Kay Baily Hutchison, is from relatively safe (for Republicans) Texas. Of course, it's still early, and it's very possible that Tim Kaine (if he runs) could do better than Webb, but still, the early days of the cycle have been good for the GOP.
Larry Sabato tweets "If Obama wins VA again, a credible D will win Senate. If R POTUS nominee carries VA, G Allen will win (assuming he gets R nod, as is likely)." I suppose that's the way to bet, all else equal, but if he'd like to give me odds, I'd be happy to take the other side -- in other words, he gets Obama/Dem and GOP/Allen, while I get Obama/Allen and GOP/Dem. Or, to put it another way, I'd be willing to bet that there will be at least a five point difference between Obama's vote in Virginia and the vote for the Senate Democratic nominee. For what it's worth, Mark Warner ran a dozen points ahead of Barack Obama in 2008, although George Allen and George W. Bush had almost identical wins in 2000.
Hmmm...actually, I'd love to know the answer to this: what's the average difference between presidential vote and open seat Senate/governor vote, and how has it changed over the years? You would want to separate out the south, at least from 1948 through 1992 or so. I feel as if I should know this, but I'm blanking on it just now...anyone?