Associated Press
In this Oct. 3, 2012 file photo, Republican presidential candidate former Massachusetts Gov. Mitt Romney and President Barack Obama talk after the first presidential debate at the University of Denver in Denver. The razor-thin race for the White House has overshadowed the fight for control of Congress.

With the Election just a day away, predictions about what will happen have gotten specific. In the New York Times, the leading paper cheering for Barack Obama, Nate Silver says Obama will win with 50.3 percent of the popular vote and 303 electoral votes. In the Wall Street Journal, the leading paper cheering for Mitt Romney, Karl Rove says Romney will win with 51 percent of the popular vote and 289 electoral votes. On Intrade, the website where one can back his or her opinion with money, Obama is favored two to one. Among seasoned Washington observers, who claim to be able to discern trends others might miss, the majority choice is Romney. With such firm differences made public, a lot of reputations have been put at stake.

The differences in these opposing views spring from different approaches that now apply to polling methodology. Once the process was very straightforward. Computers accessed phone books to draw up lists of appropriate names to be called, phone calls were made — one in three of which was answered — interviews were conducted, the answers were fed into a software program and a result came out that matched other poll's results, "within the margin of error."

Things are different this year. It now takes 10 calls, not three, to get an answer. Does that affect the accuracy of the result? A growing percentage of voters no longer have landline phones. What's the best way to adjust the sample to accommodate that fact? Many assume that the electorate in 2012 will be the same as it was in 2008. Isn't it possible that 2008's higher turnout in certain demographic groups was an abberation? Arguments about these and other questions relating to the validity of various polling methods have been as heated as arguments about the validity of the statements made by the two campaigns.

I have never run a national campaign. However, I have run enough state campaigns to learn to pay attention to what consultants call "anecdotal evidence." In 1962, when my father was seeking his third term and every poll showed him losing, I checked in with O.N. Malmquist, the political editor of the Salt Lake Tribune who oversaw that paper's polling to get his view. He used his own family to explain what he thought was happening, polls notwithstanding.

"My daughter is a Democrat who spends her evenings reading books or listening to music. My son-in-law is a Republican who spends his evenings knocking on doors. The Republicans are going to win. Intensity level." We proved him right by re-electing Dad and taking both House seats back from the Democrats.

Intensity level favors Romney. Like Malmquist, I cite my children as examples. I have two daughters who live in California who report that, four years ago, their neighborhoods were filled with Obama signs; now, none. Another daughter lives in Williamsburg, Va., where her husband teaches at William and Mary; same story. Four years ago, the campus was alive with Obama fervor, but now, all is quiet. My own experience at the George Washington University, where I have been a guest lecturer, is similar. Four years ago there was high Obama enthusiasm, but none now.

Intensity level cannot be quantified. One expert has said that trying to do it through polling is "witchcraft." Nonetheless, it is a real factor in elections. It helped Obama win in 2008, and it will be important in 2012. Thus, for tomorrow, the best prediction is the one sports commentators often use when covering a close game: "The winning team will be the one who wants it more."

In compliance with Deseret News policy, comments will not be posted on political stories and editorials from now until the polls close Tuesday, Nov. 6.

Robert Bennett, former U.S. Senator from Utah, is a part-time teacher, researcher and lecturer at the University of Utah's Hinckley Institute of Politics.