FOX: Crunching numbers to gain an edge

[bubba9497]

Crunching numbers to gain an edge

Dan Pompei /

http://www.foxsports.com/content/view?contentId=2487784

The NFL is a long way from undergoing a statistical revolution similar to what baseball is experiencing, but several coaches, in the quest of gaining an edge, are taking numbers interpretation to a new level.

Statistics are being used more than ever by men such as Chiefs coach Dick Vermeil, Titans defensive coordinator Jim Schwartz, Saints coach Jim Haslett and Patriots research director Ernie Adams. They rely on numbers to determine what their teams are doing right and wrong, to assess opponents' strengths and weaknesses and to make critical decisions.

The statistics they study aren't the typical league leaders you see printed in the newspaper. Schwartz, in fact, crusades against the widely accepted practice of ranking offenses and defenses by yards, much as Oakland Athletics general manager Billy Beane scoffs at judging baseball players by batting averages. "The way the league judges offense and defense is the most ridiculous thing I've ever heard," Schwartz says. "Yards allowed or gained has nothing to do with winning. It's about points allowed or scored. You see so many people, coordinators who have been in the league for a long time, who ... try to find ways to steal yards." If coordinators can manipulate statistics to make their units appear more effective than they are, they can justify shortcomings.

Most teams generate their own numbers to some degree. Some rely on specially produced packages from STATS Inc., Elias Sports Bureau and stats expert Bud Goode, whom Vermeil has been working with since 1969, when they were introduced by George Allen. Vermeil's reliance on statistics has increased nearly every year since meeting Goode.

Each week during the season, Vermeil spends about four hours with a 500-page book he receives from STATS Inc. He spends another two hours studying pregame and postgame reports from Goode that cover everything from predicting the scores of every NFL game to comparing what a given team's offense is producing with what its defense is giving up. Vermeil pays particular attention to Goode's Achilles' heel section, which points out the statistical weaknesses of an opponent.

During one afternoon in his wood-paneled office overlooking the playing field at Arrowhead Stadium, Vermeil keeps digging out red three-ring binders, charts and books full of statistics. By my unofficial count, he brings out 22 books of statistics, some of them as thick as a triple-layer cake. He talks about a 20-year profile of Super Bowl teams, comparing averages from every NFL victory with every NFL loss, a quarterback study that goes back to 1984, the point value of fumbles (minus-1.66) versus interceptions (minus-6.35), and how a strength-of-schedule study points out which games are most important to win.

There are so many numbers in football that the greatest challenge may be identifying the meaningful ones.

Haslett focuses on red zone efficiency. Schwartz emphasizes third-down efficiency — and it helped the Titans' defense rank first in the NFL in that category last season. Vermeil forever is emphasizing turnover differential to his players. Last year, as he points out in one of his statistical compilations, even teams with losing records won 74 percent of the time when they were plus-1 in turnover differential. Playoff teams won 93 percent of the time when they had more takeaways than giveaways.

Late in Kansas City's 38-5 victory over the Bills last season, Chiefs defensive end Eric Hicks tapped Vermeil on the shoulder and said, "Hey, coach, we're plus-5 (in turnover differential). What's the winning percentage on that?" Vermeil loved it. "You tell your team, 'We have to play smart today,' " he says. "Our guys know what smart is. We led the NFL (last season) with plus-19."

Vermeil shares his weekly turnover research with Cowboys coach Bill Parcells. In turn, Parcells shares his weekly statistical profile of NFL officials with Vermeil.

In the offseason, Vermeil spends hours poring over statistics. He earmarks some of them for speeches to be used in training camp and beyond. If the Chiefs lose a key starter to injury, for instance, he'll be prepared to present statistical evidence to his team that demonstrates it can overcome the loss.

By early May, Vermeil already had charted every practice session of training camp based on statistics from last year. For instance, Vermeil looked at where his offense ranked in third-and-1 situations, third-and-2 to 3, third-and-4 to 5, third-and-6 to 9 and third-and-10 or longer. He noticed the Chiefs needed most improvement in third-and-6 or longer, so he is devoting more practice time to those situations.

Schwartz, who received distinguished economics graduate honors at Georgetown, where he played linebacker for four years, has spent some of his free days this offseason doing regression analysis studies on when to accept or decline penalties, when to take safeties and the significance of sacks. He has networked with Dodgers general manager Paul DePodesta, a cum laude graduate of Harvard who is a leader in baseball's statistical movement, and Internet stats geeks. His references include a probability study by a Rutgers statistician and a dynamic program analysis by an economics professor from Cal, as well as Moneyball, the book about Beane's unique reliance on statistics.

"I've spoken with (NFL) general managers about Moneyball and some of them have read it, but they don't know what to do with it," Schwartz says. "Football people don't have the mathematical side. A lot of it runs counter to traditional NFL thinking."

Football, it seems, could benefit from a little more math.

[SkinsFTW]

Good read but it seems that with all of that knowledge Vermeil would have chosen to do something about that awful defense that KC has.

[jimster]

I can understand how some stats can be useful in practice, but a lot of it is useless.

Sportscenter did a piece on a mathmatician who said teams should go for it on 4th down more often due to the law of averages or something.

[China]

Originally posted by jimster

Sportscenter did a piece on a mathmatician who said teams should go for it on 4th down more often due to the law of averages or something.

Ala Steve Spurrier going for it on 4th deep in our own territory? Despite the success on that play, I don't think we'll see a trend toward that anytime soon.