While I was watching Jay Cutler make the Denver Bronco's management look brilliant (3 interceptions with 4:39 left in the first half: go, Jay, go!), the Chicago Bears scored their first points of the season with a safety. The NBC announcer noted that since Lovie Smith took over the coaching duties, the Bears have scored 21% of their points with special teams or defense. The league average is 9%. The announcer than said that this means the Bears score twice as many points off special teams/defense than the rest of the league. Wrong. It means as a percentage of their own scoring, they double the league average. If a team scored twice as many point per game than the lackluster Chicago offense, but had 11% of their points coming from st/d, they'd score more st/d points than the Bears. Given how many points the Patriots scored in their 16-win 2007 season, it's entirely possible that they did just that.

Of course, this doesn't matter much in covering a football game. But when so many people go to the polls, attend rallies, give donations and otherwise participate in democracy with similarly shoddy statistical skills, it creates problems. They base their opinions on either false understandings of statistics, or too readily buy into poor/manipulated statistics.

Often, people with poor statistical education will quote the old warhorse that statistics can say anything you want them to. That's only true if you don't know enough to see through bad statistics. A marginally savvy reader will spot obvious statistical manipulation, ask for clarification about methodology when results seem counter to expectation and otherwise have the capacity to see the difference between good statistics and crap ones. But so few Americans have a clue how to think about statistical information that it's easy to mislead them.

Boring as they may be, basic statistics should be taught in high school. Give future voters a chance at understanding the information they'll be dealing with throughout their adult lives.