Obama vs. Romney: Your guide to poll-watching

Between now and the presidential election in November, there’ll be hundreds of polls weighing Mitt Romney vs. Barack Obama. Here’s some advice on how to sort through the numbers.

By Brad Knickerbocker, Staff writer / April 22, 2012

Polls show a very close race developing between Republican Mitt Romney and incumbent Barack Obama.

(L.- r.) Jae C. Hong/AP, Carolyn Kaster/AP

Between now and the 2012 presidential election – just 198 days as this is written, folks! – you will see enough polls to make your abacus freeze up.

Related stories

At this point, it all seems a blur, this Monitor “DC Decoder” headline being the most brutally honest: “Mitt Romney vs. Barack Obama: a squeaker or a landslide?”

So what’s a political junkie to do as the daily numbers keep pouring forth? What should we be paying attention to, which ones rejecting?

On his FiveThirtyEight blog at The New York Times, poll watcher Nate Silver lays out his twelve-step approach to poll analysis. (Just coincidence that 12 is the same number as many addiction recovery programs? I don’t think so.)

Topping the list: “Be patient.”

“Many of the poll-watching habits you learned for the primaries you will need to unlearn for the general election,” he writes.

In the general election, “there are fewer swing voters, the candidates are better known, and voter preferences are more rigid,” which means that “polls have a much stronger tendency to revert to the mean, and what is perceived to be ‘momentum’ is often just statistical noise.”

Other bits of poll-watching advice from Mr. Silver: Take the poll average (Real Clear Politics is very helpful), pay attention to likely voters versus registered voters, keep paying attention to President Obama’s approval ratings, look at a robust array of economic indicators but be careful with economic forecasts, read the polls in the context of the news, and don’t over-learn the lessons of history.

“There have been only 16 presidential elections since World War II,” he writes. “That simply isn’t a lot of data, and overly specific conclusions from them, like ‘no recent president has been reelected with an unemployment rate over 8.0 percent’ or ‘no recent incumbent has lost when he did not face a primary challenge,’ are often not very meaningful in practice and will generally not carry much predictive weight.”

There are other things to keep in mind too – like how election winners actually are chosen and the demographic weights and balances at play.