They said the Democrats were a solid bet to hang on to the Senate. They said many of the races were tight as a tick. They said these were primarily local contests in which a president couldn't have much impact.
So what do journalists have to say for themselves now? Were they, once again, slavishly wedded to a conventional wisdom that turned out to be wrong? How did so many of them miss Tuesday's Republican tide?
"It caught everybody by surprise," says Ron Brownstein, political correspondent for the Los Angeles Times. "Why limit it to the press? If you talked to Republican professionals on Thursday and Friday, they were not expecting a two-seat Senate gain."
There were exceptions, of course, but such conservatives as Weekly Standard Editor Bill Kristol, Wall Street Journal Editorial Page Editor Paul Gigot and columnist Peggy Noonan predicted a Democratic Senate, as did many liberals, including New Republic Editor Peter Beinart, Washington Monthly Editor Paul Glastris, Fox's Juan Williams and CNN's Al Hunt, Margaret Carlson and Mark Shields. NBC correspondent Campbell Brown also gave the Democrats the nod.
"If you're a conservative and give an optimistic scenario, you get such a drubbing," says Jonah Goldberg, editor of National Review Online. "Everyone says you're just in the tank. There's incredible social pressure to hedge your bets, to stay within the 40-yard lines in what you predict." Goldberg says other panelists gave him "grief" after he predicted on CNN that Norm Coleman would beat Walter Mondale in the Minnesota Senate race.
Kristol, who believes there was "a late break" toward the GOP, credits the president: "We missed the fact that it was a special year. No one seriously thinks this could have happened if you hadn't had 9/11 and Bush's reaction to 9/11. I underestimated how much Bush as commander-in-chief would be worth."
Josh Marshall, author of the liberal Web site Talking Points, says pundits should have placed more emphasis on a trend in some final polls. "It was the conventional wisdom momentarily blinding people to the data that was there in the last few days," he says. "I certainly didn't call this one. The night before, I really started wondering. But I definitely didn't think it would go this badly for the Democrats."
The polls, however, were all over the map. A Minneapolis Star Tribune survey in the final days had Mondale leading Coleman, 46 to 41 percent; a Zogby International survey gave Mondale a 47 to 44 lead; the St. Paul Pioneer Press had Coleman ahead, 47 to 41.
Many of the weekend papers said that while the Republicans seemed assured of keeping control of the House, Tom Daschle was more likely to hold the Senate reins than Trent Lott.
"Democrats seem positioned to maintain -- or enlarge -- their one-seat Senate majority," Brownstein wrote in the Los Angeles Times, while adding that a GOP takeover was not out of reach. "Republicans face stiff odds in their bid to reclaim a Senate majority," The Washington Post said. "Voters are as likely to elect a divided government in this era of anxiety as they did two years ago," the Philadelphia Inquirer said.
Journalists may not have had exit polls on election night, thanks to a computerized fiasco at Voter News Service, but they leaned heavily on a blizzard of conflicting preelection polls showing tight races from coast to coast. In the end, though, many Republicans won easily, including Florida Gov. Jeb Bush (by 13 percentage points), Texas Senate candidate John Cornyn (by 12), North Carolina Senate candidate Elizabeth Dole (by 9), Georgia Senate candidate Saxby Chambliss (by 7) and Colorado Sen. Wayne Allard (by 6).
The media don't have a great history of calling midterm elections. In 1998, when Bill Clinton was facing impeachment, the pundit consensus was that the GOP would gain five to 15 House seats; instead, the Democrats picked up five. In 1994, few predicted a Republican landslide that captured control of both houses of Congress.
This time around, some prognosticators -- including Michael Barone of U.S. News & World Report, the Weekly Standard's David Brooks and CNN's Tucker Carlson and Robert Novak -- correctly forecast that the Republicans would recapture the Senate. They get to enjoy bragging rights for at least a week.
But in an environment as tight as the 2002 campaign, says Brownstein, "races are very tough to call, and it's almost irresponsible to call them. There's so much polling at the end of a midterm election that it becomes more of a distraction than an illumination."
Dick Polman, political correspondent for the Inquirer, says reporters got caught up in "echo chamber wisdom" -- that, for example, Paul Wellstone's death would energize Minnesota's Democrats and that war-hero Sen. Max Cleland couldn't lose in Georgia. "Nobody really picked up a stiff breeze that could blow these Republicans into office," he says.
Newsweek's Howard Fineman, who declared on "Hardball" Monday night that the GOP had "an outside shot" at seizing the Senate, says journalists missed a rather large elephant in the room.
"We always overthink it," he says. "We always focus on the trees instead of the forest. The forest here was George Bush's approval rating, anywhere from 60 to 70 percent. We've all had a history of underestimating Bush at every turn."
Not that Fineman's crystal ball was perfect: "Based on covering Elizabeth Dole's 2000 presidential campaign, I confidently predicted on the 'Today' show that she was like a tire with a slow leak that would go flat on Election Day. Boy, was I wrong."
But is the press pack now overinterpreting Tuesday's results? Who really knows whether the president's frenetic campaigning made the difference? Besides, a switch of roughly 29,000 votes in Minnesota, 11,500 in Missouri and 9,500 in New Hampshire would have produced a Democratic Senate and gobs of stories about how the White House blew it.
"The cycle for the next few days," says Polman, "will be to beat up on ourselves for having said for days and days that this is a 50-50 nation: Boy, aren't we foolish! But the underlying fundamentals haven't changed all that much."