On the Tuesday two weeks before Election Day, Jason Miller, the senior communications adviser to Republican presidential nominee Donald Trump, had a poll from Rasmussen Reports highlighted on his Twitter account. Miller tweeted it Monday and then pinned it (in Twitter parlance), ensuring that any visitor to his Twitter page would see the poll results first. "Rasmussen: 43-41 TRUMP," it said.

But the preview for the link said something different: "Hillary Clinton has slipped one point ahead." That's because the traditionally Republican-friendly Rasmussen's daily tracking poll shifted between Monday and Tuesday, and Miller's tweet was suddenly out of date.

Rasmussen has been one of three polls that the Trump team has embraced of late. There's also the L.A. Times-USC poll, which has skewed toward Trump from the outset because of an unorthodox methodology, and the new tracking poll from Investor's Business Daily, which has similarly been Trump-friendly. The IBD-TIPP poll, as it is known, is often referred to as the "most accurate poll from 2012," presumably because it had the lowest error in FiveThirtyEight's election rankings.

AD
AD

Over the past week or two, those three polls became a centerpiece of arguments used by the Trump campaign for its underestimated strength. Team Trump regularly uses Twitter as a sort of running press-release-slash-bulletin-board system, and you can see how key staffers plugged numbers that they thought cast Trump in a positive light, including Miller's highlighting that result from Rasmussen.

All of the bright-red boxes on the chart below are mentions of the three pro-Trump polls. The light red boxes are mentions of other polls, usually state polls.

Miller tweeted about the three Trump-friendly polls regularly. Both the candidate and his social media director, Dan Scavino, have been more likely to tweet about crowd size, Trump's new preferred metric for gauging his success, given that there have been fewer and fewer polls showing him with a lead. (They both also tweeted the results of unscientific online surveys.)

AD
AD

Kellyanne Conway, the pollster-turned-campaign-manager for Trump, tweeted more about other polls, mostly state ones. Why? Perhaps because there's a risk in isolating a few polls as being the ones to watch.

For example, Trump made a reference to the L.A. Times/IBD/Rasmussen polls on Twitter on Friday.

Watch what happened.

All three of those polls for the first time show Clinton with a lead. Awkward.

Other Trump allies ended up in a weird spot, too. Matt Drudge ran a headline Monday morning proclaiming a shocking poll result showing the race tied — contrary to the prevailing narrative. Left unmentioned was that the tie was in the IBD-TIPP tracking poll, and followed a day in which Trump led by two points.

This demonstrates the benefit in looking at polling averages instead of cherry-picking individual polls. The RealClearPolitics average, for example, shows Clinton with a healthy lead, albeit one that's narrowing slightly. Instead of getting to talk about the polls narrowing, Trump and his allies isolated three helpful surveys from an ocean that showed Clinton with a lead only to watch her take the lead in those, too. If the IBD-TIPP poll was the most accurate in 2012, as Trump mega-surrogate Boris Epshteyn insisted on pointing out on Twitter, what does it mean now that Trump is losing?

AD
AD

Again, one assumes that this is why Conway — well-versed in polling to the extent that she reportedly got annoyed when a staffer tweeted online survey results from her account — focuses more on state polls than on these outlier national ones. Talking about a new poll showing a close race in Pennsylvania or a lead in Ohio is a safer bet than picking out one national poll that has consistently deviated from the overall trend. (Of course part of her job is to serve as Trump's we-can-win hype man, so even she declared Sunday that the IBD poll "should get more mention and traction.")

There's a sweet spot in politics that every campaign likes to suggest it's in: just behind. When you're just behind, you can exhort your followers how urgently they need to get to the polls and you can demand of your donors that a few extra bucks will make the difference. Campaigns would obviously much rather be way ahead than just behind, of course, which is where Clinton is now. But no one wants to be way behind. It's much harder to convince people that you trail slightly when you're actually way behind than it is to convince them that it's closer than it looks when you're way ahead. The Trump campaign's scattershot arguments for why it's doing better than it appears are a function of things looking as bleak as they do. These three polls helped make the case that the mainstream media was getting something wrong. That argument just got weaker.

Ever the communications professional, at some point Tuesday morning Jason Miller unpinned that tweet that showed Trump trailing in Rasmussen's new poll. He hasn't tweeted about any polls yet Tuesday.

AD
AD