Who's Closer to the Truth? CNN or Nate Silver?



As election day ticks away, it's hard not to notice the drastically different forecasts from two mainstream news outlets: CNN and Nate Silver's FiveThirtyEight blog, a licensed feature of The New York Times online.

On the morning of the 2012 election, according to Silver's calculations, Obama has 90.9 percent chance of winning the election.

According to CNN, it's dead heat. Just yesterday, CNN posted this story. The beginning:

You want close? We've got close!

Seven of the eight national polls released since Sunday indicate the race for the White House is in a dead heat, like most have shown for weeks.

More importantly, it's a similar story in the key battleground states that will decide whether President Barack Obama or Republican nominee Mitt Romney will be victorious on Tuesday.

According to the final CNN/ORC International poll released before the election on Sunday night, 49% of likely voters questioned in the survey say they support the president, with an equal amount saying they back the former Massachusetts governor.

So, clearly, the two media outlets are painting vastly different pictures. But which one will be proven to be closer to the truth? And are their underlying motivations coloring their figures?

Nate Silver has become a whipping boy for right-wing pundits. But those pundits miss the point: Silver is staking his reputation as a reliable prognosticator on his numbers. My gut says his desire to be taken seriously and be respected far outweighs any perceived bias. Not to mention, someone hoping to re-elect Obama wouldn't be telling folks that it's very likely he'll win. That would only serve to make Obama voters feel that the need to vote is less urgent.

Meanwhile, CNN just seems to say "dead heat dead heat dead heat" over and over again. And I was particularly struck by this story posted Sunday night. The title: "CNN National Poll: Dead heat between Obama and Romney." Though it contains a link to CNN's electoral map, it never once mentions that the electoral college, not the popular vote, determines the winner. I mean, I knew that in third grade! Twenty paragraphs, and nothing about swing states, either. Just a long, supposedly informative story about how the race is, yes, a dead heat. Even though most polls, even theirs, suggest swing states favor the president. Assuming their coverage is misleading, which it seems to be, what could be their motivation? Covering their asses so they can't be proven wrong? Making it seem closer than it is so they get better ratings and paint a more dramatic narrative?

I guess time (and not very much of it) will tell who is right. I've spent approximately two minutes and 12 seconds developing this admittedly arbitrary criteria to determine who was closer to the truth.

Nate Silver's figures:
50.8 for Obama
48.3 for Romney

49 for Obama
49 for Romney

Silver also says Obama will have 313 electoral votes, Romney 225. That's an 88 vote difference.

If Romney wins, CNN is clearly more right.

If Obama wins the popular vote by 1 percent or less, and the electoral college victory is under, say, 40 votes, CNN is still closer to accurate.

But if Obama wins by, say 1.5 percent or more, and wins the electoral college by 50 or more, Silver is the clear winner.

Anything in the middle, we'll call a draw.

And here's another reason CNN is really starting to look like a bunch of ratings-hungry hacks to me — this recent story, which began like this:

When the race is done, the balloons have wilted, and the confetti has been swept up, Campaign 2012 may be marked more by its failures than its triumphs.

But here's the starkest failure in these final days before the vote: Neither candidate has made a convincing enough argument for his presidency to break free of the margin of error in the polls.

No matter who is elected, close to as many Americans will have voted against him as for him.

Well, that's pretty much been true for five of the past 13 elections (38 percent of them) that had margins of victory of less than 2.5 percent. How is a close race a stark failure? Because the public is fairly evenly divided, the candidates have failed? Perhaps the author, Tom Foreman, is new to American politics? Or is he correct that nearly 40 percent of all presidential elections are stark failures?

Of course, as a co-worker said, "CNN jumped the shark a long time ago." So why should I be surprised?

Comments (16)

Showing 1-16 of 16

Add a comment

Add a comment