Green Bay's Greg Jennings catches a pass in Sunday's game at Kansas City (photo from NFL.com)

Green Bay's Greg Jennings catches a pass in Sunday's game at Kansas City (photo from NFL.com)

11/05/2007

Is Football's Conventional Wisdom Unwise?

by Joseph Brownstein

Halfway through the 2007 NFL season, the Green Bay Packers have won all but one of their eight games. But many aren’t sold on the team.

Led by future Hall-of-Fame quarterback Brett Favre, Green Bay has the fifth best offense in the league, averaging 362.2 yards per game; but skeptics point to the fact that almost all of that yardage comes through the passing game. No one does a worse job of running the football: Thirteen individual players average more running yards per game than Green Bay’s 72.1.

“Every defense knows exactly what's coming. The Packers have the worst running game in football, easily the worst in Favre's 16 Green Bay seasons. All the defensive coordinators sitting in their batcaves Monday and Tuesday planning to face the Packers are saying: We don't have to load up against the run now, because we know Green Bay can't beat us unless Favre beats us,” wrote Sports Illustrated’s Peter King in an October column.

While many say a team like this is destined for collapse, is the conventional wisdom—that a team needs to balance running and passing—outdated? In recent years, several publications have arisen which try to look at how the conventional wisdom of sports holds up under the scrutiny of numbers—and often the numbers show the wisdom is wrong.

In a 2005 article for the Journal of Quantitative Analysis in Sports, which he founded, Dr. Benjamin Alamar looked at the problem of how often a team should run and how often it should pass.

Looking at average gains of passing and running plays, Alamar concludes that, in fact, teams should pass much more, as the reward is often greater. Teams gain more yardage, on average, when they pass.

That would not always have been true, he wrote, noting that years ago—when a lot of the conventional wisdom was formed—quarterbacks had a lower percentage of their passes caught by their wide receivers. Additionally, they had a higher percentage of passes intercepted by a defender.

But now, for most teams, he thinks passing more is a better strategy—especially since, as he documented, the element of surprise seems to have little effect on passing gains but does effect rushing gains—which would now benefit since they would happen less frequently.

“For most teams, the passing attack is significantly more efficient than the running attack,” said Alamar, although he noted that “some [teams] are so bad there should be some balance.”

Outside the Box

The problem of how teams run the ball was the problem that motivated Aaron Schatz to start his website, FootballOutsiders.com. By looking at game statistics, he showed that, contrary to what announcers would often say, teams were not winning because they could run, but running at the end of games when they were ahead, in order to run out the clock.

Schatz said that media analysis of football today suffers from the fact that announcers grew up watching football in the 1970s, and expect the game to be played the same way—much like people still listen to the same music they listened to in high school and college.

“It’s not the era of Franco Harris anymore,” said Schatz, referring to the 1970s Pittsburgh Hall of Fame running back.

On Football Outsiders, Schatz and other writers use numbers to determine how much players contribute to their team’s winning—numbers which are often different from the raw statistics certain players put up.

“I’m trying to change the way reporters talk about the NFL,” said Schatz.

Schatz also edits the annual Football Prospectus, which predicts team performance and looks at a variety of statistical problems each year.

Out of Sight, Out of Mind?

Part of the challenge in understanding statistical analysis of sports lies in getting people to understand numbers which don’t directly correlate to what they see.

 “Generally speaking, analysis of sports is done by people watching—determining by what they see,” said Alamar.

He explains that quantitative analysis looks at hundreds of instances in order to determine what is really happening—something that is difficult to pick up on from just watching.

Alamar, for his part, believes that coaches are reluctant to pick up new strategies because of media criticism that will follow. He points out that coaches who attempt to pick up a first down on fourth down rather than punt the ball away are often criticized by the media, despite the fact that the numbers show that it is a winning strategy.

“It’s so ingrained in our culture that you punt on fourth down, that you have to have a run-pass balance,” he said. “Unless you’re extremely successful in deviating from that, it causes you more trouble than playing what everybody else does.”

But while he believes in the value of numbers, Alamar doesn’t think they can replace scouting and other traditional means of making decisions by sports teams.

“What you can deal with is measuring how good different strategies are and how good a performance is on the field,” he said. “This kind of information is always a complement to traditional sources of information; it’s not a replacement for sources of information used now.”