Hi -- I'm back. Thank you again, Yastreblyansky and Crank, for some great work while I was away.

A lot happened while I was gone and I'm still trying to sort it all out. For now I'll just address the results of the West Virginia GOP Senate primary. Don Blankenship, the racist felon coal baron, finished third, 15 points behind the winner, Patrick Morrisey. That shouldn't be surprising because the last two public polls, from GOPAC and Fox News, showed Blankenship in third, 8 and 9 points behind the leader, respectively. It was surprising because, shortly before the election, Politico and the Weekly Standard reported on internal polls from rival campaigns claiming that Blankenship was in the lead.

Nate Silver expressed skepticism about the polls in real time.

There probably is an agenda. Presumably the Morrisey campaign leaked the poll/"poll" that had it Blankenship 28/Morrisey 27/Jenkins 14, hoping to scare people into thinking the Morrisey train is the only one leaving the establishment station and otherwise Blankenship will win… https://t.co/LLej4Gfspa

— Nate Silver (@NateSilver538) May 7, 2018

Then the Jenkins campaign caught wind of it and released their own poll/"poll" to counteract the narrative, showing it Blankenship 31/Jenkins 28/Morrisey 27. Note that they put themselves exactly 1 point ahead of Morrissey, which is sometimes a "tell" that their position is weak.

— Nate Silver (@NateSilver538) May 7, 2018

Last night, after the polls closed and it was clear that Blankenship was losing badly, Silver wrote,

I’m not just being a smartass when I put “polls” in scare quotes; there are so many ways to manipulate polling results (e.g. fudge around with your turnout model) so as to produce a desired narrative that the results may as well be fake. And every now and then, internal “polls” probably are just made up on the spot....

In this case, it seems pretty likely that the internal “polls” showing Blankenship surging represented gamesmanship on behalf of the other two campaigns ... as much as they did anything real.

But why would the campaigns release polls showing that they're trailing when there was another narrative showing each one in the lead? (The GOPAC poll had Morrisey up by 4, while the Fox poll had Evan Jenkins up by the same amount.) If those campaigns wanted to release fudged polls, why didn't each campaign release a poll naming its candidate as the front-runner?

Who really benefited from the thwarted-Blankenship-surge narrative? Two people: Mitch McConnell and Donald Trump.

McConnell worked hard to defeat Blankenship, while Trump weighed in with a late tweet. Blankenship now believes that the tweet cost him ten or more points in the race.

I think McConnell knew that Blankenship was toast but wanted to convey the impression that crossing him (and Trump) was an extremely dangerous thing to do. The victory of Roy Moore in the Alabama Senate primary last year cast serious doubt on McConnell and Trump's ability to get an electable candidate through a contentious intraparty contest. Now -- even though Blankenship probably lost all on his own -- they can say they have their mojo back.

So I think this Blankenship-surge narrative came from McConnell, and maybe Trump's people, not Morrisey or Jenkins.