The problem is that even when you get past the technical issues (margin of error, confident interval, good sampling, etc) and try to read through some of the general polling bias many polls use, we still have a lot of things working in our favor that polls can't show.
The pollsters aren't reflecting the massive GOTV effort we will put in place, the fact that undecideds break against the incumbent, and that overall our base is mobilized more than ever before (and that we actually really like Kerry and don't just hate Bush).
Mark Shields helps put things in perspective.
That is the advice of respected analyst and Democratic pollster Guy Molyneux, who argues that the point-spread between any incumbent president and his challenger is meaningless. What does matter in pre-election polls, he insists, is the percentage of the vote the incumbent president is receiving.
History shows that the percentage of the vote that an incumbent president gets in the major polls before Election Day is an accurate predictor of the percentage of the vote the incumbent will win on Election Day. Thus, in Molyneux's judgment, the "incumbent who fails to poll above 50 percent is in grave danger of losing his job."
In the four most recent elections where an incumbent president sought re-election -- Bill Clinton in 1996, George H.W. Bush in 1992, Ronald Reagan in 1984 and Jimmy Carter in 1980 -- in the three major polls of the broadcast networks and their newspaper partners, the incumbent won the same percentage of the actual vote (or less) as he had received in the polls in three elections. Only Ronald Reagan, who had polled 58 percent of the vote and then actually won 59 percent of the vote, exceeded the average of the three network surveys.
Here is the president's case: "Things have never been worse in the United States, and I'm the only guy who can get us out of the mess we're in."
But eventually, as Carter in 1980 and the first Bush in 1992 discovered to their regret, when an incumbent president seeks re-election, the voters make that election a referendum on the incumbent's record and first decide whether the incumbent deserves a second term.
History tells us that undecided voters break overwhelmingly on Election Day in favor of the presidential challenger -- unless the incumbent president's campaign has been able through negative attacks to "disqualify" the challenger.
One analyst has concluded that since 1976, 86 percent of undecided voters have voted for the challenger candidates.
Unless the polls begin to give the incumbent more than 50 percent of the vote, then the result on November 2 could well be the second one-term Bush presidency in a dozen years.
And, now the best part: Bush is making of dangerous habit of repeatedly polling under 50%.