Poll position: Prognosticators punch it out over who'll prevail

By Mike Miliard
01:58 PM
Share

The electoral battle has been pitched this past month. Each side is fighting hard, confident in its rightness and rectitude. Each is convinced that truth and common sense are on their side. Each is sure that they will win this coming Tuesday.

You thought I was talking about Barack Obama and Mitt Romney?

No, I'm talking about the pollsters and pundits whose job it is to predict this race.

One one side, there are stat savants such as Nate Silver, who writes the New York Times' much-read FiveThirtyEight blog.

On the other, television talking heads such as MSNBC's Joe Scarborough, and political operatives like Karl Rove and Dick Morris.

Silver was originally a baseball statistician. He created the PECOTA (Player Empirical Comparison and Optimization Test Algorithm), which uses complex number crunching to try to predict ballplayer's future performance, and he managed the sabermetric stat bible, Baseball Prospectus, for six years.

Interestingly, Silver has recently been attracting a level of ridicule and dismissiveness from certain corners of the media, especially on the political right, that sounds very much like the heated divide in baseball – between advanced statistical research an old-school scouting and gut instinct – that was chronicled in the book and movie Moneyball.

At FiveThirtyEight, Silver runs daily simulated elections based on a specially-designed system that aggregates dozens and dozens of polls, weighing them depending on when they were conducted, their sample sizes and the pollster's accuracy rating.

Silver is an Obama supporter, but he insists his weighting methodologies are entirely nonpartisan. His model correctly predicted the outcome of the 2008 election in 49 of 50 states, and nailed all 35 Senate races that year.

So far this election season, despite a recent trend that has seen Romney running roughly neck and neck in national polls with Obama (and often a tick or two ahead of him), Silver's site has lately consistently shown that the president has a 70 to 80 percent chance of winning on November 6.

This is because Obama continues to maintain leads, even if sometimes slight, in the battleground states needed to win the Electoral College. As of Friday afternoon, the blog projects 303 electoral votes for Obama, versus 235 for Romney. He pegs Obama's chances of victory at roughly 81 percent.

(Since Romney's surge in the polls following Obama's zombie-like performance at the first presidential debate in Denver, many Democrats have hit the refresh button on FiveThirtyEight several times a day, clinging to his calculations like a calming cup of herbal tea – or a life raft in a scary sea.)

But Silver's complex and considered analysis of the race has attracted laughing derision from many observers – most of them, perhaps unsurprisingly, Romney supporters.

"Nate Silver says this is a 73.6 percent chance that the president's going to win. Nobody in that campaign thinks they have a 73.6 percent – they think they have a 50.1 percent chance of winning," said Scarborough on his show, Morning Joe, the other day. "Anybody that thinks that this race is anything but a tossup right now is such an ideologue [that] they should be kept away from typewriters, computers, laptops, and microphones for the next ten days, because they're jokes."

Some conservatives have been even more triumphalist. Not content to call the race the toss-up it very well may be, political operative Dick Morris crowed recently that it's going to be a "landslide."

Romney "will win this election by 5 to 10 points in the popular vote. And will carry more than 300 electoral votes," he said, describing a near-mirror image of Silver's model.

Others have resorted to weird, ad hominem attacks on Silver himself.

"Nate Silver is a man of very small stature, a thin and effeminate man with a soft-sounding voice," wrote conservative blogger Dean Chambers.

Silver has been somewhat nonplussed by the attacks.

"I think I get a lot of grief because I frustrate narratives that are told by pundits and journalists that don't have a lot of grounding in objective reality," Silver said on the Charlie Rose Show.

Indeed, The Atlantic raises a good point when it points out that most of Silver's critics are people "whose livelihoods depend on people caring about their subjective feelings about elections."

If some quibbling with Silver (and stat-crunchers like him) seem to be basing their projections on gut-level feelings based on imperfect polls, others have sought to deploy numbers of their own to illustrate their certainty of Romney's victory. Karl Rove penned an op-ed in the Wall Street Journal on Wednesday. He wrote:

It comes down to numbers. And in the final days of this presidential race, from polling data to early voting, they favor Mitt Romney.

He maintains a small but persistent polling edge. As of yesterday afternoon, there had been 31 national surveys in the previous seven days. Mr. Romney led in 19, President Obama in seven, and five were tied. Mr. Romney averaged 48.4 percent; Mr. Obama, 47.2 percent. The GOP challenger was at or above 50 percent in 10 polls, Mr. Obama in none.

The number that may matter the most is Mr. Obama's 47 percent share. As the incumbent, he's likely to find that number going into Election Day is a percentage point or so below what he gets.
 

But just as important as the numbers Rove cites are the numbers he ignores. Indeed, Talking Points Memo's Josh Marshall says the column is emblematic of what he calls "The Poll Truth Movement." Like the controversy over Obama's birth certificate, some on the right seem to hew to their own set of facts.

'What you hear nothing about in Rove’s argument is actual poll data. Actual state polls," writes Marshall. "There’s a heavy reliance on Gallup’s national numbers and various data points about early voting and crowd sizes and party ID. But actual polls? Not so much".

For Silver's part, annoyed by the ongoing poll wars, he recently offered Scarborough a wager: $2,000, payable to the Red Cross' Hurricane Sandy relief efforts. (Something for which his knuckles were rapped by the Times' ombudsman Margaret Sullivan.)

"He’s been on a rant, calling me an idiot and a partisan, so I’m asking him to put some integrity behind it," he told the Times. "I don’t stand to gain anything from it; it’s for charity."

Regardless of who ends up winning that wager, let's hope that next Wednesday morning shows a clear victor in this close and contentious presidential election.

After two exhaustingly long years of campaigning, the last thing this country needs is one of the nightmarish "Zombie Election" scenarios John Heilemann lays out in New York Magazine.

"Obama’s position has stabilized and he holds a small but significant advantage in terms of the electoral map – but his sub-50 percent support levels in all of the battleground states is a cause for real concern among Democrats," he writes. "All of which is to say ... that next Tuesday night is likely gonna be the emotional equivalent of riding the Cyclone at Coney Island: a nerve-­jangling, empty-out-the-liquor-­cabinet-and-stash-box sort of affair."

The only thing worse would be a 2000 recount redux, with months of rancorous partisan arguments that make these tussles over polling seem civil by comparison.