An exclamation point on two days’ worth of rumors about the RNC threatening to cut Trump off, with an “emergency meeting” happening today between Team Reince and Team Trump to help get his campaign back on track. I hope Trump fans appreciate Reince’s performance here. It’s aimed squarely at them, to reassure them that the RNC is fully behind its nominee. You guys will still vote Republican in 2018 if Trump loses by eight points and then spends two years insisting Priebus stabbed him in the back, won’t you?

Why Reince felt obliged to mention the polls as part of his shpiel, I don’t know. He’s duty bound to be enthusiastic as part of his pep talk, but he could have sidestepped the subject with the ol’ chestnut that the only poll that matters is on Election Day. As it is, he’s right that the LA Times poll shows a tight race, with Clinton up one point. What he doesn’t mention is that it’s the only national poll tracked by RCP that does.

There are a few other things he doesn’t mention. For starters, the LAT poll has always tilted unusually heavily towards Trump. In mid-July, at a moment when the RCP poll average had Clinton ahead by four points, the LAT had Trump up three. At one point after the Republican convention, Trump surged to a seven-point lead in the Times poll. He surged in other national polls too, but the biggest bounce he got anywhere else was a three-point lead. And of course his seven-point margin has now been completely erased by Clinton’s post-convention bounce. FiveThirtyEight sees a “house effect” in the LA Times poll worth fully five points to Trump, meaning that a one-point lead for Hillary there suggests an actual lead nationally of six. And of course, even a six-point national deficit doesn’t capture the deep trouble Trump is having in battleground states now. In Pennsylvania, he’s been down 10-11 points in four straight polls. The new survey of North Carolina has him trailing there by nine. A lone national poll that’s tight means next to nothing when the weight of battleground polls is heavily against him, and of course Reince knows it. But when you’re stuck introducing your nominee at a rally, even a thin silver lining will have to do.

One other note about the LA Times poll, since we’ll probably hear it cited a lot in the weeks to come as a hopeful outlier from the other national surveys. Its methodology is unusual in that, instead of polling a bunch of different Americans each time, it uses a fixed group of the same 3,000 people and re-polls a subset of them every day. The idea is to capture shifts in opinion among a set population to detect which way the general public might be shifting. It worked well in predicting the 2012 outcome, but the problem with using a fixed group is that if there’s a “lean” towards either candidate among that group, it’s going to show up every time you poll them. If instead you’re using a sample of different Americans each time you conduct a survey, then you might get a “lean” towards one candidate in one poll but it won’t necessarily show up in the next one you do. That might help explain why the LA Times data is durably pro-Trump. As to why there might be a lean, Nate Cohn notes that the LA Times weights its sample according to how people claim they voted in 2012, with a roughly accurate split of 27 percent for Obama and 25 percent for Romney. The flaw in doing things that way, though, is that some people in every poll reliably misstate whom they voted for in the last election. Inevitably some small but meaningful chunk insists that they backed the guy who won the election even though, in reality, they voted for the loser. When you account for that, Cohn estimates, a more accurate sample for the LA Times poll would be something like 33 percent who claim they voted for Obama versus 25 percent who claim they voted for Romney. The fact that the Times’s sample is far more evenly balanced suggests that it skews too heavily Republican. Which is exactly what the data showing a “tight race” supports.

Anyway. Here’s Reince being a good soldier. followed by recent video of Sharon Day suggesting that the election might be rigged. Sharon Day is co-chair of the RNC.

Y/day, RNC co-chair, Sharon Day, suggested Nov election could be rigged. This wasn't Trump. Weeds grow deep at RNC. pic.twitter.com/neiYPhNzep

— Old Rant Dump (@TrumpinTheShark) August 12, 2016