The numbers one, two and three are key figures in the presidential race as the last week of September begins.
One and two are the gaps between President Barack Obama and Republican challenger Mitt Romney in the daily Rasmussen and Gallup polls, respectively. This represents a very slight uptick for the president from late last week, when both polls showed the candidates tied. But even this gap is well within the margin of error.
Three is the lead in the latest Politico/George Washington University poll, a poll that has reasonable turnout assumptions, favoring the Democrats by three points.
As the Deseret News has noted previously, any poll is only as useful as the turnout assumptions that lie behind it. Most polls showing large gaps in the president's favor right now assume that Democrats will outnumber Republicans by significant margins, often at or in excess of 2008.
Some polls, like this National Journal poll that shows Obama leading by seven points, obscure this key variable by simply not releasing it at all. It is difficult to know what to make of a poll that doesn't reveal the partisan makeup of its sample.
The Politico/GWU poll shows some movement in Obama's direction, but still within the margin of error. In May, Romney led this poll by one point, and in August Obama led by one point.
The glass is half full or half empty, depending on who is seeing it.
Republican pollsters Ed Goeas and Brian Nienaber, writing in Politico, are pleased to see Romney "winning the middle class," and contend that the election remains razor close.
One measure of the poll's accuracy is that it closely matches the generic congressional ballot in Rasmussen's current findings. The generic ballot is the poll question that asks if they would vote for the Republican or the Democrat in their U.S. House race if the election were held today.
Rasmussen has the GOP leading by one point currently, while the Politico/GWU poll has the Democrats leading by one point. Discounting for statistical noise, the two are essentially tied, and identical, which suggests the sample is balanced.
Perhaps the most important marker in this poll is the Hispanic/Latino vote, which surprisingly has Romney trailing Obama only 58 to 40 percent.
This is at stark odds with other recent polls, including a Fox News Latino poll last week that showed the president leading Romney among Latinos 60 to 30 percent.
The Fox poll may be more accurate because of the small numbers in the Politico/GWU poll, which polled under 100 likely Latino voters, compared to over 800 for the Fox News poll.
John McCain tallied just 31 percent of the Latino vote. The Politico/GWU poll shows Romney with 55 percent of likely white voters, identical to McCain's 2008 results.
If turnout levels were the same as 2008, Romney would have to perform markedly better in either category to be competitive.
Turnout could be high but is unlikely to reach 2008 levels. Still, 80 percent called themselves "extremely" likely to vote, up from 73 percent in an August Politico/GWU poll.
“That’s normally what you get in the final days, usually the final weekend of the campaign,” GOP pollster Ed Goeas told Politico. “So we’re now looking at an extremely engaged electorate — whether you’re a Republican, Democrat or independent.”
It is turnout, ultimately, not the elusive undecided voter, that will determine the election, and the higher enthusiasm across the board in the Politico/GWU poll is not a good sign for Romney, who likely would benefit from a lower overall turnout.
Eric Schulzke writes on national politics for the Deseret News. He can be contacted at firstname.lastname@example.org.