Daily Poll Watch: Right-wingers attack Nate Silver simply because they don’t like his numbers
Earlier today, I responded thusly to a reader who had implied that Rasmussen polls were more accurate in the 2008 presidential election than the respective projections by analysts Nate Silver and Sam Wang:
The Rasmussen record [of accuracy] in 2008 relates only to the national popular vote. But when it came to state races, where the electoral votes were, nobody was more accurate than Nate Silver and Sam Wang.
Silver correctly predicted the presidential winner in 49 of the 50 states. The only state he missed was Indiana, which went for Obama by 1%. He also correctly predicted the winner of all 35 Senate races that year.
Sam Wang’s final prediction in the presidential race of 2008 was off by only one electoral vote out of 538.
As for Rasmussen’s vaunted accuracy just before the presidential election of 2008, most other polls were also pretty much in the ballpark in predicting an Obama victory by a fairly comfortable margin in the popular vote. But Rasmussen wasn’t very close in projecting the margin in the Electoral College.
On the day before the 2008 election, Rasmussen reported this:
“Rasmussen Reports Electoral College projections now show Obama leading 260-160. When ‘leaners’ are included, Obama leads 313-160.”
The actual final tally showed this: Obama, 365 electoral votes; McCain, 173 electoral votes.
Nobody was closer than Nate Silver and Sam Wang in their final Electoral College projections.
This reader’s bad-mouthing of Silver and Wang is in keeping with a spate of such stuff from professional conservative pundits in recent days.
Fortunately, Wang, who has his own disagreements with Silver, has come to his fellow analyst’s defense in THIS PIECE, wherein he refers to number-crunchers like Nate and himself as “nerds”
Paul Krugman is calling out National Review Online for their attempted takedown of Nate Silver for biased methods and somehow cooking the books. Krugman writes:
This is, of course, reminiscent of the attack on the Bureau of Labor Statistics — not to mention the attacks on climate science and much more. On the right, apparently, there is no such thing as an objective calculation. Everything must have a political motive.
Now more commentators on the right, including Jay Cost (The Weekly Standard) and Jennifer Rubin (Washington Post), are getting in on the act…
A popular approach to undermining technical knowledge is to throw mud, assert expertise, make picky points, and sow doubts among the less savvy…
The human mind has a large capacity for finding reasons to reject a piece of disagreeable evidence. I’ve written about this in the context of how people form false beliefs in politics (“Your Brain Lies To You,” NYT, June 27, 2008). Polling internals lend themselves very well to such “motivated reasoning.” It is always possible to find something not to like in a poll. This is why I discourage all of you from chewing over single polls.
I believe Silver doesn’t extract all the information and tends to add unnecessary factors, which leads to blurry probabilities and poor time resolution. However, his intuitions about the data are excellent and he is very concerned with getting things right. For purposes of popular consumption, he is a fine and honest nerd.
Elspeth Reeve of The Atlantic also has a COLUMN defending Silver from right-wing attacks:
Perhaps the most telling critique of Silver’s model comes from the people most deeply invested in it being wrong. Romney aides “laugh and roll their eyes when reporters tease them with mentions of the model,” Buzz Feed’s McKay Coppins reports. One adviser, though, offers an analysis more closely tied to real data, saying, in Coppins’ paraphrase, “FiveThirtyEight could well give them a better chance of victory as the swing state polls tighten in the final days of the race.” In other words, if the state polls change, so will Silver’s model. Which is pretty much what Silver himself would say.
Meanwhile, Silver’s latest projections give President Obama a 74.6 percent chance of winning re-election and a likely electoral vote total of 296, which is 26 more than needed.
Mind you, Silver would be the first to tell you that these numbers can, and likely will, change one way or another between now and election day. He’ll adjust his numbers according to changes in state polls.
Wang’s analysis is even more favorable for Obama. He says the president’s chances of winning now range between 91 percent and 97 percent. And his Electoral College Projection, at this point, gives Obama 305 votes.
Here’s an example of a right-wing pundit deliberately misreporting the results of a survey that was conducted by two pollsters, one of them a Republican and the other a Democrat:
Politico.com had THIS STORY this morning:
With eight days to go until the election, President Barack Obama has recaptured a narrow national lead over Mitt Romney, riding increased support from women and an edge in early voting.
A new POLITICO/George Washington University Battleground Tracking Poll of 1,000 likely voters — taken from last Monday through Thursday — shows Obama ahead of Romney by 1 percentage point, 49 percent to 48 percent. That represents a 3-point swing in Obama’s direction from a week ago but reflects a race that remains statistically tied.
Ah, but Fred Barnes of The Weekly Standard didn’t like it that the Battleground poll showed a bit of a gain for Obama. So HERE‘s what he reported:
The bipartisan Battleground Poll, in its “vote election model,” is projecting that Mitt Romney will defeat President Obama 52 percent to 47 percent. The poll also found that Romney has an even greater advantage among middle class voters, 52 percent to 45 percent.
While Obama can close the gap with a strong voter turnout effort, “reports from the field would indicate that not to be the case, and Mitt Romney may well be heading to a decisive victory,” says pollster Ed Goeas.
Well, somebody apparently called Barnes out on his misreporting of the Battleground poll, so he later added this update to his piece:
Politico reports on the poll’s top-line: Obama 49, Romney 48 percent. But it is the noteworthy last paragraph from Goeas’s memo that we’re referring to: “In sum, this data indicates this election remains very close on the surface, but the political environment and the composition of the likely electorate favor Governor Romney. These factors come into play with our “vote election model” – which takes into account variables like vote intensity, voters who say they are definite in their vote, and demographics like age and education. In that snapshot of today’s vote model, Mitt Romney leads Barack Obama by five-points – 52% to 47%. While that gap can certainly be closed by the ground game of the Democrats, reports from the field would indicate that not to be the case, and Mitt Romney may well be heading to a decisive victory.”
What nonsense! Nowhere in his story does Barnes say that Goeas is the Republican member of the Battleground polling team. And he offers no slants at all from the team’s Democratic pollster, Celinda Lake.
Barnes didn’t like the topline numbers in the Battleground poll, so he ignored them and went for a biased angle from Goeas. And he also represented the numbers he liked as the product of a “bipartisan” poll, as if the Democratic member of the team agreed with the biased numbers.