Bob Barr should have learned that perhaps he can forget about running for elective office.

Although Barr kept reminding voters in the 11th Congressional District that he was one of those who impeached Bill Clinton in 1998, they didn’t seem to care anymore. They picked a younger, fresher face in Barry Loudermilk and gave him a two-to-one victory margin.

Karen Handel, likewise, should have learned that it’s time for her to think about a career path that doesn’t involve running for public office. Republican voters rejected her in statewide elections in both 2010 and 2014. They also turned down the candidate she endorsed in this year’s Senate GOP runoff.  

If you are a sitting member of the U.S. House of Representatives, you should have learned the lesson that it’s not a good idea to try to step up and run for higher office—not at a time when public approval of Congress is at historically low levels. House members Jack Kingston, Phil Gingrey and Paul Broun all thought they had a clear path to winning the Senate seat being vacated by Saxby Chambliss. All three of them were rejected by the voters.

Journalists and pundits should have learned that endorsements don’t really carry too much weight and don’t guarantee a transfer of support to the candidate who gets endorsed. If endorsements mattered, Kingston would have clobbered David Perdue in the Senate Republican runoff by at least 10 percentage points. Nearly every GOP official of any stature in Georgia endorsed Kingston, as did two of the candidates who finished behind him in the primary (Handel and Gingrey). Even some tea party types got behind Kingston in his runoff campaign, but all of these high-powered endorsements amounted to nothing. Perdue beat Kingston anyway for the Republican nomination.

Everyone involved in politics—voters, candidates, consultants and journalists—should have learned this valuable lesson from the runoff: you can’t always put a lot of faith in what the pollsters are telling you. During the week prior to the July 22 runoff, three major independent polls were released that showed Kingston leading Perdue by an average of 6 percentage points. Matt Towery of Insider Advantage released this statement after his firm conducted a poll that showed Kingston leading by 5 points:

“This survey indicates that Kingston has managed to retake momentum in the race with under a week to go… Perdue’s campaign and his supporters have emphasized their candidate’s ‘political outsider’ position as a candidate running against the ‘insider’ Kingston. While that theme has worked well with older voters, it has lost its magic, having run long enough to pull in all of the votes it can.”

Actually, Perdue hadn’t lost any magic at all.  As the events of July 22 subsequently proved, his strategy of running as an “outsider” was precisely the path to victory that his campaign needed.  

Inaccurate polling is not something isolated to Georgia. In Mississippi’s Senate race, runoff election polls had tea partier Chris McDaniel running several points ahead of longtime incumbent Sen. Thad Cochran. Cochran successfully persuaded voters from predominantly Democratic precincts to cross over and vote for him in the GOP runoff, giving him a surprise win over McDaniel.

In Virginia, U.S. House Majority Leader Eric Cantor’s internal pollster showed him to be 34 points ahead of tea partier David Brat heading into the June 10 Republican primary. In a very low-turnout race, however, Brat upset Can