Polls Predictably Wrong... Again


So what was it that made it so hard to predict the recent US election that caught the data scientists; the media; the pollsters; the pundits and the bettors by surprise much like Brexit did in the UK a few short months ago?

Could it be as New York Times (who got it dreadfully wrong from the get-go) op-ed columnist David Leonhardt said:

There are a few likely suspects: Trump supporters who didn't admit it to pollsters; a late break of undecided voters moving his way; lower-than-expected turnout among Democrats, both white and non-white; and higher turnout among conservative working-class whites.

All of which still begs the question­-then what were the pollsters and pundits doing?

Geoff Garin, a veteran Democratic pollster who worked for the pro-Clinton super PAC Priorities USA, said many surveys had under-sampled non-college-educated whites, a group that Trump appealed to. He also argued there had been an over-emphasis on the belief that the country's rising demographic diversity would put Clinton over the top.

"There was too great a belief that demographics are destiny, and that demographics would lead to a certain outcome," he said. "The reality turned out to be much different that."

"The very premise of polling is based on the idea that voters will be completely honest with total strangers," said veteran GOP operative Ned Ryun. - Politico

This makes me wonder if non-college-educated whites are now part of the rising demographic diversity of the United States and whether polls take into account the notion that, as Otto von Bismarck famously said, "people never lie so much as before an election, during a war, or after a hunt."

Or perhaps it's about the Seduction of Data because of its glut and our blind belief that Big Data can solve anything...or even more...make me do anything...or could it be that, as one red-faced pollster was quoted as saying, "data is dead"?

The Lesson of Trump's victory is not that data is dead. The lesson is that data is flawed. It has always been flawed--and always will be...But this wasn't so much a failure of the data as it was a failure of the people using the data. It's a failure of the willingness to believe too blindly in data, not to see it for how flawed it really is...

"We sometimes fool ourselves into thinking we have a lot of data," says Dan Zigmond, who helps oversee data science at Facebook and previously handled data science for YouTube and Google Maps. "But the truth is that there's just not a lot to build on. There are very small sample sizes, and in some ways, each of these elections is unique."

"As a data scientist, I always think more data is better. But we really don't know how to interpret this data," Zigmond says. "It's hard to figure out how all these variables are related." -- WIRED

I'd posit, as I have been writing for a while, that the inability to actually listen to people; to derive real insight; to understand motivation; to feel the fears that keep people up at night is what made the polls and the pollsters useless.

"Supporters of Brexit and Trump were continually maligned by the dominant media narrative (validly or otherwise) as primitive, stupid, racist, xenophobic, and irrational. In each case, journalists who spend all day chatting with one another on Twitter and congregating in exclusive social circles in national capitals--constantly re-affirming their own wisdom in an endless feedback loop--were certain of victory. Afterward, the elites whose entitlement to prevail was crushed devoted their energies to blaming everyone they could find except for themselves, while doubling down on their unbridled contempt for those who defied them, steadfastly refusing to examine what drove their insubordination." ­­- Glenn Greenwald, The Intercept

As one pollster summed it up:

'This is worse than "Dewey Defeats Truman,"' said Sabato. 'There were only a few polls back then. At least 90 percent of hundreds of surveys were wrong. Those of us who model from polls use their data; garbage in, garbage out.' - Daily Mail

And there you have it...GIGO...garbage in, garbage out.

The lesson for marketers and the rest of us who rely on such information is simple. PEOPLE FIRST...it's not the numbers...it's what the numbers mean...it's about PEOPLE.

As I wrote a number of years ago, micro-trends have been irrelevant since the digital channels took off. Where once it meant something to be a soccer mom, today, that same soccer mom could be a gourmet chef; rock diva; deep water diver and blogger all in the space of a few clicks or swipes. The characterization of demographic constituencies in this campaign was a throwback to the days of old and proved its irrelevancy in the results.

"American voters are not poor or black or female or college educated. In reality, people fit into multiple different groups at once. Two facts are simultaneously possible-that the poorest voters chose Clinton and that the poorest white voters chose Trump." -The Guardian

KNEE JERK ALERT-There are many reason Trump won and Clinton lost-I will save some of my views for my next post-my only point here is to comment on the failure of the powerful digital technology to come even close...listen:

People are wonderfully predictable...you can create mathematical models that describe everything that is likely to happen to the finest detail, and then people go and do something you didn't expect." -- Martin Lee, cybercrime manager, Alert Logic.

Frankly, I bless human serendipity, am grateful for it and hope that you and I will always be able to surprise-the alternative is too scary to contemplate.

And no one will ever make me buy those purple whale pants because they have "my data"...

What do you think?

Read more at The Weekly Ramble

Follow David Sable on Twitter: www.twitter.com/DavidSable

Follow David Sable on Twitter: www.twitter.com/DavidSable