Commentary Magazine


Topic: Nate Silver

Are the Polls Biased? Democrats Hope Not

The latest crop of opinion polls has generally brought good news for Democrats. The Real Clear Politics average of all the national polls has given President Obama a slight lead, after this poll of polls had shown him trailing since Mitt Romney’s post-Denver debate comeback changed the race. Even more important, polls of likely voters in the battleground states have given the president leads in most of them. This caused New York Times blogger Nate Silver to double down on his forecast predicting an Obama win. According to Silver, Obama now has an 83.7 percent likelihood of prevailing on Tuesday.

We’ll leave aside the arguments about Silver’s odds-making, which depicts what even most liberals concede is an extremely close election as a near certain Obama win. Suffice it to say, as I wrote on Thursday, Silver’s belief that Obama had a field goal lead with 3 minutes left in the game (which he may now think is more like a 4-point lead with 2 minutes left) is based on a belief that the polls he trusts are accurate. On Saturday, however, he returned to the question that has to be haunting his readers: what if these polls aren’t accurate? While he admits the possibility, he thinks it unlikely that so many surveys could be in error. That seems logical, even persuasive. But the problem with that assumption is the same as it has been for the past month. Most of the polls showing Obama ahead either nationally or in some states reflect a common bias: their sample reflects a picture of the electorate that resembles the 2008 Democratic advantage. But this year we expect the gap in party identification to be smaller. In short, unless the Democrats match or exceed the massive “hope and change” surge of four years ago, then what Silver and the Democrats who look to his column for encouragement fear will be true: all the pro-Obama state polls are going to turn out to be quite wrong.

Read More

Which Team Has a Field Goal Lead?

There’s been probably too much attention paid to New York Times blogger Nate Silver over the last few weeks. Some of the criticism he has received (including some from this page differing with his conclusions if not necessarily always with his methodology) has been justified. But in fairness to Silver, he appears to be sticking to his guns about the accuracy of his forecast that continues to show President Obama as a heavy favorite to win re-election. While not intending to belabor the issue of his accuracy more than necessary, I think it’s worth returning to the subject one more time both in order to clarify my differences with his approach.

Silver explained his forecast again this morning as he surveyed the latest round of polls on the presidential race:

Mr. Obama is not a sure thing, by any means. It is a close race. His chances of holding onto his Electoral College lead and converting it into another term are equivalent to the chances of an N.F.L. team winning when it leads by a field goal with three minutes left to play in the fourth quarter. There are plenty of things that could go wrong, and sometimes they will.

But it turns out that an N.F.L. team that leads by a field goal with three minutes left to go winds up winning the game 79 percent of the time. Those were Mr. Obama’s chances in the FiveThirtyEight forecast as of Wednesday: 79 percent.

Not coincidentally, these are also about Mr. Obama’s chances of winning Ohio, according to the forecast.

That is a reasonable sounding point of view, especially when it is coupled with Silver’s disclaimers about the possibility that his forecast could be wrong and noting that a lot of tossup states that he believes Obama will win are still closely contested. But the problem here is that despite Silver’s confidence that what we are looking at is a three-point lead for the president, it may be nothing of the kind, either in Ohio or in the country as a whole. The probabilities he alludes to in sports–such as those that can give us precise statistical odds about what happens when an NFL team has a field goal lead with three minutes to play or a Major League baseball team has a two-run lead in the ninth inning–are entirely accurate and reliable because there’s no doubt in a game as to what the score is. In politics there is no such certainty, rendering Silver’s rational Sabrmetric approach to political polling mere guesswork.

Read More

Latest Defense of Nate Silver: Even When He’s Wrong, He’s Right

At this point in the election, both the national tracking polls and, generally speaking, the battleground state polls show an extremely close race. This has inspired much criticism of stat-man Nate Silver. Liberals, as Jonathan wrote recently, may be looking for a scapegoat if Barack Obama loses, and will wonder why Silver insisted that statistically Obama was an overwhelming favorite to win re-election right up to the end. Conservatives say that statistics don’t factor in momentum, that any predictive model that ignores trends should be taken with a grain of salt, and that current trends don’t back up Silver’s predictions of Obama as the heavy favorite.

Additionally, as Josh Jordan has pointed out at National Review Online, there are problems with Silver’s statistical model, which gives more weight to pro-Obama polls—even older polls—than to those showing Mitt Romney’s recent gains. Silver’s defenders respond that Silver makes no guarantees, and that giving Obama a high chance of winning doesn’t preclude the opposite outcome. This defense—essentially that no outcome can disprove Silver’s model—treads a bit too close for comfort to Chuck Klosterman’s classic rant about statistics:

Read More

Infallible Election Prognosticators Tend to Have Brief Careers

Back in May 2011, the leading liberal poll analyst of this election cycle returned to his roots in an op-ed published in the New York Times. Nate Silver, who had parlayed a brilliant record as an independent numbers cruncher in the 2008 presidential election into a gig as the paper’s political blogger in the age of Obama, first made his name as a writer as a baseball guy and one of the leading exponents of new and advanced ways of looking at baseball statistics. On May 9, 2011, Silver penned a piece for the Times explaining why New York Yankees captain Derek Jeter was finished as a baseball star. Given that that the Yankees shortstop had an uncharacteristically mediocre 2010 season and was off to a slow start in 2011, it was hard to argue with Silver’s conclusion.

Except the very same day that Silver was planting Jeter’s tombstone in the Times, the future Hall-of-Famer got four hits, including two home runs in a game. I noted this embarrassing development in a blog post here titled, “The Perils of Punditry: That’s Why They Play the Games.” For my pains, I was subjected to a chorus of abuse via e-mail and Twitter from Silver’s fans, most of whom knew nothing about Sabermetrics. Indeed, another Times blogger noted my criticism (which was laced with respect for Silver’s work on both baseball and politics) and ironically noted, “the jury was out” on whether the results of “one game” could disprove the great Nate.

The jury was out in May, but within a few months, Silver’s fans would be dropping that prediction of his down the proverbial memory hole as Jeter put together a stellar second half of 2011 and followed it up with a brilliant 2012 in which he led the Major Leagues in base hits. That didn’t mean Silver didn’t know what he was talking about, but it was proof that a proper understanding of what has already happened didn’t necessarily give even the smartest of researchers the ability to predict the future. Fast forward to the last days of the 2012 presidential election campaign, and it looks like that day in May wasn’t the only time Silver’s crystal ball has clouded up.

Read More

The New National Pastime: Poll Analysis

The 2012 election is the first presidential contest in the age of Twitter. It’s also the one that may be remembered as the one in which analysis of poll data became the subject of mass discussion as opposed to the domain of a few political science and statistical freaks. The New York Times’s Nate Silver is as responsible for this as anyone, but the trend is fed by the proliferation of national polls whose results are as varied as their methodologies. Silver has become something of a lightening rod in this election as his forecast which, has continually favored President Obama’s prospects, is now coming in for almost as much scrutiny as the policies of the man he’d like to see re-elected. As someone who has occasionally criticized Silver’s conclusions, I think the focus on him is unfortunate. Silver is a brilliant stat man who whose work attempts to bring the unsparing realism and devotion to accuracy and understanding that is the hallmark of sabermetrics — the study of baseball statistics that derives from the acronym for the Society of American Baseball Research — to political writing. That, like some baseball writers, he cannot always rise above his prejudices, is unfortunate but does not mean his work isn’t worthwhile. Silver is always a good read and even if he seems to have an agenda, I always learn from his posts.

Nevertheless, given the importance that Democrats are placing on his “Five Thirty Eight Forecast,” it was only a matter of time before Silver was given a thorough takedown and Josh Jordan of National Review has done it in a must read analysis. In “Nate Silver’s Flawed Model,” Jordan details how Silver’s partisan leanings have influenced his judgment about how much weight to give to various polls. As Jordan points out, Silver tends to assess the reliability of certain polls based on his feelings about whether they are right, which is to say sufficiently pro-Obama. While I don’t think Silver’s purpose is deception, his bias has created a model that seems designed to produce one result even if it contradicts what many see as a pro-Romney trend. As such, he’s become the geekiest yet perhaps also the most important cheerleader in the country these days as liberals look to his blog for comfort in trying times. But Silver isn’t the only one making mistakes out there.

Read More

Romney’s Lead: More Than Just One Poll

On Tuesday, New York Times blogger Nate Silver attempted to make sense of the latest round of polls that had been released on Monday. Silver, an astute political statistical analyst, took note of the post-debate trend that has tilted the presidential contest in favor of Mitt Romney, but argued that the average of the various polls that had altered his daily forecast of the outcome had been skewed by one poll. That poll from Pew Research showed Romney ahead of President Obama by four percentage points, a result that seemed out of line with other surveys.

But the problem with dismissing the Pew Research Poll is that as more data is coming in from other sources, it isn’t possible to pretend that what has happened in the last week is the product of one poll. With the latest Gallup Tracking poll and an Investors Business Daily/TIPP Tracking poll both showing Romney ahead by two points, as well as other polls showing Romney gaining ground in swing states, there is a clear trend that is showing up across the board in a wide range of surveys. Romney has spent most of the year trailing the president and looked to be in big trouble in September as his deficit grew. But the first debate was clearly a turning point in the race, and though Silver has tried to argue that the post-Denver bounce has already started to recede, there is now a wide body of evidence illustrating that Obama is losing ground and, at best, is locked in a dead heat with his Republican challenger. The fact that the Real Clear Politics average of major polls is showing Romney with an aggregate lead today for the first time all year must send chills down the spines of the Obama campaign.

Read More

CNN Moves Wisconsin to “Toss Up”

Following several polls that show the race tightening in Wisconsin, CNN has moved the state from “lean Obama” to “toss up” on its electoral map:

CNN Thursday turned the important battleground state of Wisconsin from “lean Obama” to true “toss up” on its electoral map, in the wake of Mitt Romney’s naming of House Budget Chairman Paul Ryan, a seven term congressman from the Badger state, as his running mate. One contributing factor behind CNN’s move was a new poll that matched two others from last week that indicate that the presidential contest in Wisconsin is close. …

With Wisconsin’s move to true “toss up,” the CNN Electoral Map now suggests Obama leading in states with a combined 237 electoral votes, Romney ahead in states with a combined 206 electoral votes, and states with 95 electoral votes up for grabs. 270 electoral votes are needed with win the White House.

Read More

Defining a “Very Slight Favorite”

Today, on his New York Times blog, Nate Silver published the first of what will be a regular series of updates on the forecast for the November election. Silver makes no secret of his liberal leanings, but his statistical work (forged in his beginnings as an outstanding baseball analyst) is straightforward and generally reliable. In his first 2012 presidential forecast, he establishes President Obama as a “very slight favorite” as of the moment. That means he gives the president roughly a 60 percent chance to win re-election but, as he notes, his current estimate of 290 electoral votes for Obama is flexible with “outcomes ranging everywhere from about 160 to 390 electoral votes are plausible, given the long lead time until the election and the amount of news that could occur between now and then.” That sounds about right as does the criteria by which he seeks to judge trends via an economic index that will decline in importance, as we get closer to election day and an average of polls taken.

The problem, as Silver notes, is that national polls have trended toward Mitt Romney while state polls tend to favor the president, which has given the incumbent something of an intrinsic edge in Electoral College projections. That edge may be offset if current negative economic trends continue and Obama’s numbers decline everywhere. But it also points out that the wisest course for both candidates, but especially Romney, is to concentrate his resources in the key swing states of Virginia, Ohio and Colorado. Those three are the states where the probability of an Obama victory is the most narrow and may therefore decide the outcome. But while Silver’s detailed breakdown of the numbers is well worth readers’ efforts, the problem with this and any forecast is that trends have a logic all their own. If the public perceives that the economy is failing, then the president’s slight edge can be thrown out the window.

Read More

Flotsam and Jetsam

Robert Gibbs thinks the administration made the right call Mirandizing the Christmas Day bomber. Dennis Blair said no one really thought it through. One of them is off the reservation. Unfortunately, I think in this case it’s Blair. The Obami never make errors, don’t you know?

Not even on health care. Gibbs also says that the Massachusetts election doesn’t prove nuthin’ about nuthin’. (Democrats have to be praying that this is an act and that the White House doesn’t truly believe this.)

Back on planet Earth, Sen. Evan Bayh “gets cold feet” about pushing unpopular health-care legislation through Congress using parliamentary tricks on a party-line vote. It’s not clear whether he’s an outlier or the beginning of a trend toward political sanity in his party.

In a similar vein, Allahpundit catches Chris Matthews being sane, arguing for “reality” and against reconciliation to pass health care. Well, he was going up against Alan Grayson.

Noemie Emery thinks there’s a split on the Left: “Those edging their way toward the lifeboats are those members of the House and Senate who sooner or later have to be in touch with the voters. Those who want the bill passed (i.e., pushed down the throats of the howling public) are White House officials and pundits, bloggers, academicians, talk show hosts, and others who don’t face reelection in this year or any, and will even find their business improving if the bill passes and all hell breaks loose. The pundits, who have no skin in this game since they will not get fired, have transferred their soaring contempt for the American people to their beleaguered House members. ‘Jump! Jump!’ they cry to the quivering congressfolk. No sacrifice is too great for others to make for their dreams.” Unfortunately for the Democrats, the White House so far is with the “Jump! Jump!” crowd, raising the question as to whether Obama really wants a second term or simply thinks he’s immune to the same forces that are knocking down fellow Democrats one by one.

If the elections were held today, Larry Sabato and Nate Silver think the Democratic majority would shrink to 52 seats in the Senate (h/t Michael Barone). But the elections aren’t being held today, and lots can change in 10 months.

It’s Republican confidence and the loss of all those seats that may spare the country any more noxious legislation. The Washington Post agrees: ”Obama’s biggest priorities — overhauling health care, expanding college aid, reducing climate change — are now in limbo, facing dim prospects as Republicans show little interest in cooperating, and Democrats brace for a 2010 midterm election year potentially as volatile as 1994, when the GOP captured the Senate and the House two years after Bill Clinton was elected president.” Probably didn’t help that, as Democrats now complain, Obama was “too hands-off, too absent.” Or that the country tuned him out.

Mickey Kaus points out that “comparative effectiveness” research is a crock. Obama, Kaus argues, either “has an average President’s shallow understanding of the subject,” is trying to make “bending the cost curve” look painless when it really involves making value judgments to deny care, or is practicing willful ignorance. Could be some combination of all three, of course.

In California, front-runner Meg Whitman is narrowing the gap with Jerry Brown in the gubernatorial race. Hey, if Massachusetts is in play, California is in play.

ObamaThink

It’s sometimes hard to figure out why the Democrats are so anxious to pass such awful health-care legislation — in fact indefinite, awful legislation — which is also so unpopular. Taking issue with liberal pollster Nate Silver, who thinks passing ObamaCare would be courageous, James Taranto concludes that the legislation has become highly unpopular because the public has figured out that it’s a “monstrosity.” He explains:

Whose job was it to make ObamaCare popular? The politicians who backed ObamaCare, of course. If 61% of Americans oppose the Senate bill, it is because the senators who support it have failed to make their case. It’s hard to see how someone who thinks they had a good case to make can excuse this failure, much less present it as an achievement of near-courage.

Moreover, Democrats are convinced, in a groupthink exercise outmatched only by the global-warming hysterics, that passing a hugely unpopular bill is the only chance to save themselves from a 2010 wipeout. There are several explanations for this delusion. First, they really don’t believe it’s politically wise but want universal, government-run health care so badly that they’re willing to take a dive politically. (This is the rationalization of the ideological true believer.) Second, they think the voters are dopey and don’t understand how wonderful life under the monstrous tax, mandate, and rationing scheme will be, only that, in the end, they’ll learn to love it. Third, Democrats are petrified of their liberal base and worry about primary challengers and donor rebellion. (As far as the 61 percent of Americans go, however, they seem willing to take their chances.) Of course, more than one explanation may be valid here.

We’ll see if the Democrats manage to cobble together a bill that can attract 60 senators. That seems to be the sole focus of their energies. Then they’ll have to worry about the tens of millions of enraged voters who will be very unhappy with them. Like Scarlett O’Hara — they’ll worry about that another day.