Did Big Media just have its Dewey Won Moment?

Big Media's Dewey Wins Moment

I stayed up late last night watching the Democratic debate on CNN. Full transparency; I’m a big fan of Bernie Sanders. But very early in the campaign process I wanted to look at the objective data to see if social sentiment and other social indicators could predict the next president. There are several sites that simply aggregate polls to make prediction. These polls such as the Suffolk and Quinneppiac poll are typically culled from a bank of 500-1500 phone calls. The data I looked at comes, using a variety of online listening tools and APIs, from Twitter, public Facebook posts, blogs, Tumblr, new website comments. The tools show a variety of data points:
Twitter follower acquisition – how many followers they have or gain.
Twitter engagement – how relevant Twitter posts are to followers, determined by retweeting.
Potential Human Reachâ„¢ – the total number of people candidates’ Twitter account can reach.
Mention volume – how often people are mentioning the candidate.
Passion Index – how passionate the comments about the candidate are. Do they just not like them or do they hate them.
Social Sentiment – how a candidate is viewed; positively or negatively.

I’ve been keeping track of this information on a Tumblr blog called The Social Election. I’ve been posting numbers, charts, graphs, occasional news items related specifically to data such as the determination that millions of Clinton and Trump Twitter followers are robot accounts. The data has shown some very interesting information. But something else jumped out at me last night.

I’ve known that Big Media, the CNN’s and Washington Posts of the world, prefer newsmakers to news for awhile. After all, the bigger the story, true or not, the more they can sell advertising for. What I didn’t realize was how far away from the facts they’d stray to push their own agenda. Perhaps they still don’t understand the power of the Internet, perhaps they live in a bubble, perhaps they don’t realize how much data we have access too.

Last night’s Democratic debate was sponsored and hosted by CNN. Here is their current headline, and their take on who won the debate:
CNN Hillary is WInner

What makes this interesting is how out of touch with the opinions of voters this is. Here, from CNN’s own focus group:
CNN focus group says Bernie Sanders won the debate

Of course those crazy kids on Facebook love Bernie too:

And here are screenshots from public polls that CNN ran during the debate. On the economy, foreign policy and the overall debate, viewers chose Bernie Sanders as the winner by a HUGE margin. I took these screenshots with my phone lest they disappear. The overall debate winning poll is still live here.

CNN Poll: Sanders winsCNN Poll: Sanders wins
Bernie Sanders wins CNN debate

In fact, CNN wasn’t the only news organization with a similar opinion that conflicted with voters:

Slate gave the debate win to Clinton…

Slate says Clinton wins debate

Slate’s readers had a different opinion:
Slate readers says Sander won debate

Overseas, The Telegraph’s 7 political pundits handed the debate win to Hillary Clinton. Their readers also had a difference of opinion:

Screen Shot 2015-10-14 at 12.39.21 PM

Time readers also gave the win to Sanders…

Time viewers say Bernie won debate

While Time itself gave the win, at least on the economy, to Clinton.

FOX News, that bastion of fact-checking, gave Clinton the win…
Screen Shot 2015-10-14 at 12.59.38 PM

…even though their poll, of course, gave the win to Sanders.


Even local news channels polls showed Sanders won:


So, despite overwhelming poll numbers, social sentiment that looks like the chart below, how can Big Media get it so wrong? Are they living in a bubble? Is it driven by their corporate agenda? Are they judging by some standard unknown to voters? Are they merely reporting what their advertisers want them to? Whatever your party affiliation and no matter which candidate you back, you should be disgusted to know that the “journalists” you turn to for informed opinions and facts are completely out of touch with the American people.

Brand Passion Index


on “Did Big Media just have its Dewey Won Moment?
15 Comments on “Did Big Media just have its Dewey Won Moment?
  1. Pingback: Democratic Debate 2015 - Page 2

  2. While supporters yearn for traditional mainstream coverage (and recognition of Bernie’s true favorability status) it appears that social media will need to become our true media in this election if we are to truthfully convey his message and its resonance with the public. It would be worth considering how to circumvent the major media in reporting these results to the non-social media followers, to better convey what is really happening with our traditional media. I had originally believed that there were a few holdouts who we could trust (individual reporters on mainstream stations) but it appears that the corporate masters have pulled the plug on most of them (e.g., Ed Schultz, etc.) I’d be interested in learning how to better inform the average non-political voter about what is really happening and why not to trust their polls anymore. Your passion intensity metric and people opinions vs. media punditry opinion comparisons also offers damning testimony to this phenomenon.

  3. I have been pulling (and Elizabeth Warren, but she said no, at least for this cycle) for Senator Sanders since before our grassroots calling was heard of by most people. The one word I use to describe him is Integrity. I don’t believe any of the other candidates on either side can claim that descriptive.
    I am so sickened by the manipulations imposed by corporate America, including our mainstream news outlets. He is my guy!

  4. Saw all these same numbers for Paul last election. Didn’t matter. The establishment will get their candidates. But you get to pick whether they wear a elephant pin or a donkey pin! Democracy!

    • Hi Hamilton,
      You may be right. That’s exactly why I started gathering data on social and digital conversations surrounding all of the candidates; to see how closely public sentiment matches what corporate media is reporting and, in the end, if public sentiment matches what happens at the polls in November. Initially my curiosity focused on the predictive capabilities of social listening software. Interesting what it has uncovered. Keep your eye on thesocialelection.tumblr.com.

  5. This is savvy and important stuff. Media has been used far too long, to set stages that need to be dismantled, never mind shored up.
    Are not these media venues breaking the law, culpable for telling lies if what really happened is being manipulated?
    Thank YOU for such close scrutiny and fact finding. Have you asked NPR, perhaps Diane Rehm, to let you share your findings with the thinking public?

    • Hi Laura,
      Thanks for the support. You should start keeping an eye on my blog dedicated to raw voter data culled from millions of social and digital conversations: thesocialelection.tumblr.com.
      As for NPR, I don’t have any contacts there, but to be honest, they’re not much better than CNN. The point of this blog post was not to support Bernie Sanders, though I do. The point was to call out chillingly obvious favoritism in the glaring light of contrary voter opinion. I get that the press is entitled to their opinion, even when, especially when, it goes against public opinion. But this is a different situation. Not only did CNN allow a Clinto page takeover of their homepage the day after the first debate, they want out of the way to dismiss their own focus group, discredit their Facebook poll as “just kids” (despite the average age of a Facebook user being 40) and deleted comments on their story pages calling them out on it.
      NPR isn’t that different; dismissing and ignoring Sanders despite ever increasing event numbers not only in battleground states like New Hampshire and Iowa but in deep red states like Louisiana, South Carolina and Texas of all places. I’d be shocked if NPR realized that the story isn’t Bernie Sanders, but how Big Media is attempting to anoint the challengers on both sides. If you know someone at NPR who doesn’t have a pre-scripted agenda and is t afraid to ruffle feathers among their colleagues, let me know!

  6. It almost seems as if traditional media is trying to self destruct. Perhaps everybody is just following the Fox model and saying what they think listeners want to hear rather than speaking the truth. But no matter what the rational behind their behavior, there can be little doubt that at this rate, it wont be long before nobody trusts any of the major networks for anything, other than the weather report.

  7. There’s problem with using social media polls line this, one that sociologists and medical researchers have known about since before there was social media. It’s called self-selection bias, a.d you don’t seem to mention it anywhere.

    Basically, it means that instead of getting a proper, well-distributed demographics, the respondents are instead those who are invested strongly enough in the topic to be motivated to take the poll. While in this case that includes the people who really like Bernie, or who really like Hilary, it also includes the people who really hate Bernie, and the people who really hate Hillary. And that second set is by no means identical to the first set. This skews the result from what it would be if you knew who all was watching and you could create a properly representative demographic subsets and then ask them their opinions.

    This is is on top of the already biased subset of population that are social media users. It’s a group which tends towards educated, well-off with a good job to be able to afford smart phones, computers and Internet service.

    And because of sexism and racism, poll respondents tend to be white and male, because those people are used to having their opinions heard when they want to make them hurt and nobody prevent them from doing so, where is women and minorities are often prevented from being able to give their opinions or people simply ignore them when they try to do so.

    Because of all this, what you’re doing here is simply compounding the problem by using it by a subset of a biased subset of a biased subset. That’s not to say that Big Media didn’t also have their biases in how they reported who won, but it does mean at your results are not necessarily more accurate than theirs just because you used a bunch of online polls.

    • Hi Cory,
      You bring up an interesting point. The thing is that these online polls only reinforce what I’ve been seeing in some pretty advanced social listening tools.

      So here are some questions for you:
      If the online polls that these news organizations used are unreliable and biased, why did they run them?
      Why have they used them in the past to back up their analysis?
      Explain the fact that CNN’s Facebook poll showed a different winner than they claimed.
      How is it that CNN’s own focus group came up with the same response as their Facebook poll, their online poll, everyone else’s online poll, and the social listening tools I’m using?
      You suggest that online polls are skewed because of racism and sexism (Does the poll have a sign that says “whites only”? I didn’t see it), wouldn’t phone polls be skewed as well?

      Yes, people are self-selecting their choices in online polls. Isn’t that interesting? When people who take the time to vote make their voice heard, you get a different picture than when a polling company calls a few hundred people during dinner. The folks that polling companies call are not necessarily going to vote; they are likely voters. The folks that go out of their way to vote in an online poll, with no urging from a pollster are definitely going to vote. As for your assessment that online poll respondents are likely to be white and male; the social listening tools I’m using show the breakdown of those commenting online is 60/40 male female. In terms of racial breakdown there is no data that I’ve seen that suggests that racial representation of those that respond to online polls inaccurately portrays the racial makeup of voters.

      BTW, here is a great article on why phone polling simply doesn’t work because of the small samples size and skewed demographics: http://www.nytimes.com/2015/06/21/opinion/sunday/whats-the-matter-with-polling.html?_r=0
      At the same time, every news organization that declared Clinton a winner, has been basing their reporting over the last few months on this same ill conceived polling apparatus. The entire reason I launched TheSocialElection.Tumblr.com is to see if it had more predictive capability than those used by Big Media. I launched the site several months ago and began looking primarily at conversation volume, passion and sentiment of each of the candidates using identical terms. With Republicans the magic number was volume; if a candidate was generating a great deal of online conversations, they were covered by the news media, in other words: Donald Trump. On the Democratic side the exact opposite has happened; the candidate with the most conversation volume, highest passion ranking, and most positive sentiment is Bernie Sanders, yet Clinton is getting all of the press. The fact that online polls show Sanders in the lead is the first indicator I’ve seen that my tools are accurate.

      • They also pulled a dirty trick afterwards. Since they didn’t get the results they wanted with the online polls, they did another poll afterwards, asking who won the debate in order to declare Hillary the winner of at least one poll. The problem with that poll was that if you looked at how it was run, most of the people who were polled either watched just part of the debate or they watched none of it at all. The reason thought that Hillary won was because they were influenced by mainstream media stories.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>