#TrumpWon is bogus: A reminder that online polls are mostly meaningless

A fact-check of the BS surveys suggesting Donald Trump scored a yuuuge victory over Hillary Clinton in the debate

By Nico Lang

Published September 27, 2016 10:59PM (EDT)

 (Reuters/Lucas Jackson)
(Reuters/Lucas Jackson)

Trump won.

That was the No. 1 trending topic on Twitter this morning following last night’s presidential debate, in which Republican Donald Trump and Democrat Hillary Clinton faced off for the first time at Hofstra University. The 90-minute discussion was the first of three contests between the CEO and former first lady, one in which the candidates sparred over national security, the North American Free Trade Agreement and Clinton’s use of a private email server during her tenure as secretary of state.

Online polls suggest that Trump claimed a decisive victory over his Democratic rival. Nearly every single survey shows that the billionaire businessman not only beat Clinton but also destroyed her. Trump overwhelmingly came out on top in polls conducted by a vast litany of outlets — including the Time, NBC, The Right Scoop, CNBC, Wired, Buffalo News, ABC, Las Vegas Sun, The Telegraph, Fortune, Roanoke Times and CBS New York. It’s shockingly difficult to find a single online survey in which Clinton, who went into the debate in a virtual tie with Trump, emerges the winner.

That may be surprising to anyone who was actually watching the play-by-play last night. Trump, clearly unprepared, gave a series of increasingly rambling responses when Clinton provoked him about his climate-change denialism, habit of stiffing contractors, and history of sexist attacks on women, including referring to the 1996 Miss Universe as “Miss Piggy.” “Donald, she has a name,” Clinton said, in a moment of defiant outrage. “Her name is Alicia Machado.”

Following the debate, political commentators struggled to make sense of his bizarre performance. Chuck Todd of NBC's “Meet the Press” called the debate, in which Trump interrupted Clinton 51 times, “a very surreal event.” Both Salon’s Heather Digby Patton and The Huffington Post’s Howard Fineman called it the worst showing in debate history — with Fineman remarking that Trump was “incoherent . . . reeling from blow after blow.” Even Fox News — whose former CEO, Roger Ailes, is serving as an unofficial adviser for the Trump campaign — thought Clinton won. Fox’s Douglas Schoen and William Whalen agreed that she was more presidential and simply better prepared.

“The debate, taken in its entirety, demonstrates how much preparation really matters,” Schoen wrote. Clinton “was ready for all of his quips with a litany of detail that may have bored the viewer at points, but showed why she is winning on qualifications, experience and temperament in every poll.”

The CNN/ORC poll appeared to confirm those assessments, handing Clinton a landslide victory. The esteemed survey found that 62 percent of likely voters who tuned in for Monday’s debate thought that the Democratic nominee boasted the stronger showing as opposed to the 27 percent who favored Trump. A separate focus group from CNN indicated an even larger margin of victory for Clinton. In the focus group comprised of voters in the crucial swing state of Florida, 18 members thought Clinton won the debate. Just two watchers believed that her Republican challenger performed better.

If the media reported that Clinton destroyed her opponent, while internet polls suggest Trump triumphed by equal margins, what gives?

The answer is right on the box. The Time poll warns users that the survey is strictly unscientific, relying on those who click the button rather than a representative cross sample of the American public, as done in Gallup and Pew surveys. As the website wrote, internet surveys “are not predictive of how the debate outcome will affect the election.” Instead, these attempts at reading the tea leaves “are a measure, however imprecise, of which candidates have the most energized online supporters, or most social media savvy fan base.”

In truth, online polls mean almost nothing. With the Time poll, there’s nothing stopping you from voting multiple times on different browsers. I was able to vote for Clinton three times — once using Google Chrome, another time on Safari, and yet a third time on Firefox. If I were truly passionate about my vote, I could also use my cellphone, iPad or Kindle to add another tally for behind former senator’s name. It was easy to do the exact same thing on the Drudge Report, Fox San Diego, Breitbart, The Hill, Shelby Star, and NewJersey.com surveys, all of which went handily for Mr. Trump.

To help solve the problem and make these surveys more accurate, pollsters should block users from the same IP address — the numerical code assigned to your computer by its network — from voting more than once. It’s a no brainer.

Even that, however, may prove a fruitless task. On Trump’s subreddit, Reddit users have been sharing links to online polls in a master thread, encouraging one another to stack the deck in the CEO’s favor. “Get voting,” one user wrote. “Let's correct our record.” Nearly every single one of the surveys mentioned above is cited, allowing Trump’s followers to effectively troll the polls. Unless one were to dismantle social media, there’s no way to ensure that these surveys — no matter how numerous they are — accurately reflect the feelings of the mass public.

Although you may take online polls with a grain of salt, Trump does not. The Republican nominee, has been trumpeting these faulty indexes on his Twitter account as a sign that, despite the poor reviews, he actually did win last night’s historically abysmal debate after all. It must be true. Twitter says so! At 11:47 a.m., Trump tweeted a screenshot of the Time, Fortune and CNBC surveys. “Such a great honor,” he wrote. “Final debate polls are in — and the movement wins!”

The only way that statement could be less accurate is if online pollsters were allowing dead people to vote, too.

Throughout the campaign, the businessman has used his trumped-up polls as an indication of his continued lead over his opponents. It has proven enormously effective at drumming up support for Trump’s candidacy, helping to legitimize him to the voting public. Although pundits suggested that effective attacks on the CEO helped Marco Rubio win the CNN-Telemundo debate on Feb. 25, the Florida senator never gained ground in the polls.

In Trump’s case, the appearance of success begets more success. Given that the race is getting down to the wire, allowing him to continue inflating his numbers could have a dangerous impact on the election.

If you need a final word on last night’s debates, check out Frank Luntz's focus group. Although Trump supporters have been using the discrepancy between the CNN and online surveys to cry media bias, Luntz is hardly a shill for the Democratic Party. The respected Republican pollster worked for former president Bill Clinton’s bitterest foes throughout the '90s, including Pat Buchanan and Newt Gingrich.

During last night’s debates, Luntz surveyed undecided voters in Pennsylvania, one of the most pivotal swing states in the 2016 election. By a margin of 16 to 6, the focus group went for Clinton. She won points on everything from her temperament to her handling of difficult questions — on topics including U.S. race relations and Trump’s continued interrogation of Obama’s birth records. Trump, however, scored poorly on his response about why he won’t release his tax returns and his comments on the “small loan” he received from his wealthy father, which actually amounted to $14 million.

Calling her victory “bigger . . . than almost any debate I've done in a long time,” Luntz concluded, “This is a good night for Hillary Clinton. It is not a good night for Donald Trump.”

That may be true, but on the internet, the truth is whatever you want it to be.

Nico Lang


Related Topics ------------------------------------------

2016 Presidential Debates Donald Trump Elections 2016 Hillary Clinton Online Polls