I will lol if D's win AL, but lose VA.
Elections for governor are just different. They can be uncorrelated with the national environment or the state's partisan lean.
VA-GOV 2013 went the complete opposite way of the 2014 midterms, same thing with VA-GOV 2001 and the 2002 midterms. And VA-GOV 2005 was a Dem winning by basically the exat same margin as 2001. That leaves only the McDonnell landslide in VA-GOV 2009 and the 2010 GOP wave midterm that remotely match. It's not well correlated.
There were a couple of situations that made 2001 and 2013 non-predictive of the next midterm.
In 2001, 9/11 basically froze the political environment to where it was prior to 9/11 with Warner leading. Rove was not yet able to nationalize the elections around terrorism.
In 2013, you had the government shutdown that briefly made the environment pro-Dem until the healthcare website issue turned it back to Republicans, which didn't happen quite early enough to swing the race to Cuccinelli.
Other than those years, the VA GOV race is a pretty good predictor of the lean of the next midterm. You could argue that Gilmore's win in 1997 should have led to 1998 being a good GOP year, but I would argue that Gilmore's win was more a function of the pro-incumbent environment (Gilmore basically ran as George Allen's clone) that continued through the 1998 election (the most pro-incumbent year in both Houses other than 1988).
