Over the past four years, Aaron Schatz and his staff at Football Outsiders have applied the same principles of critical thinking to professional football that Bill James used to revolutionize baseball analysis two decades ago. Pro Football Prospectus 2007, their third annual NFL guide, analyzes every team and is loaded with research on everything from player development to on-field strategy. Today's excerpt, the second in a three-part series, uses Football Outsiders' advanced stats to identify some surprise playoff contenders. Yesterday's piece explained why no NFL team will ever emulate the 2006 Indianapolis Colts. Tomorrow, Schatz explains why it's so difficult to analyze NFL games and how the league itself may be standing in the way of a deeper understanding of the pro game.
The NFL has more year-to-year change than America's other major sports leagues. Over the past decade, nearly 50 percent of the teams in the NFL playoffs didn't make the postseason the previous year. Every preseason, fans and football pundits struggle to predict which teams will surprise. At Football Outsiders, we identify teams on the rise with the help of our DVOA metric. Generally, our picks turn out pretty well. Last year, we missed on the Saints and Jets like everyone else, but we forecast a return to the playoffs for the Eagles and Chargers and were the only pundits to predict that the Oakland Raiders would become a top defense. And going back a bit, we won Salon's preseason predictions contest in both 2004 and 2005.
How does DVOA help us make predictions? DVOA, which stands for Defense-adjusted Value Over Average, is compiled by breaking down every play and comparing it with the league average based on situation and opponent. (You can learn more about the statistic here.)The most important aspect of DVOA is that, since it is based on play-by-play data, it can be broken down into smaller sets—plays that involve a certain player, down, or situation. Within those breakdowns, you'll find a number of trends that will help project the sleeper teams of 2007.
The Third Down Rebound
The 2004 San Diego Chargers were coming off a terrible 4-12 season, but our numbers told us they would win at least eight games in 2005 and have one of the best offenses in the league. Everybody thought there was an error in the equations … until the Chargers went 12-4 and finished third in the league in scoring. Why were we bullish on San Diego? The 2003 Chargers had a strong offense on first and second down but were terrible on third down. Poor performance on third down turns out to be a very important indicator that a team will improve. It's true whether we're dealing with offense or defense, and the reverse is true as well: A team that is weak on first and second down but excels on third down will tend to decline the following year.
Teams get fewer opportunities on third down than on either first or second, so third-down performance is more volatile. But success on third down is also a much bigger part of a team's overall success, since the result is usually either very good (four more downs) or very bad (losing the ball to the other team by punting). Over time, teams will tend to play as well in these situations as they do in other situations. If a team struggles horribly on third down, then, it's likely that its record the following season will improve.
The third-down-rebound theory helps to explain a lot of the surprise Super Bowl contenders of the past decade, including the 1999 Rams, the 2001 and 2003 Patriots, and the 2005 Seahawks. Who could be a big rebounder this year? Watch out for the Washington Redskins. Last year, Washington had the worst third-down pass defense of any team since we started tracking play-by-play data in 1996. As you might remember from yesterday's piece, the Redskins had more than their fair share of injuries last year. If Washington gets a little healthier this year, they should improve considerably on third down and make a strong push for the postseason.
Missing the Trees for the Forest
Last season, the average NFL team threw the ball more than 500 times, had 450 rush attempts, drew almost 100 penalty flags, attempted 30 field goals, and fielded more than 100 punts and kickoffs. That's around 2,200 unique events on a season, counting both offense and defense. To determine who makes the playoffs, we distill those 2,200 unique events down to 16 wins and losses. It won't surprise you to learn that a team's won-loss record does not always match up with its performance on the field, as seen over the course of thousands of plays.
Last season, the Jacksonville Jaguars came in sixth in our DVOA ratings but went just 8-8. The issue wasn't luck but consistency. The Jaguars went 5-3 against playoff teams, blowing out the Jets 41-0 and embarrassing the Colts 44-17 but lost twice to pitiful Houston. In Pro Football Prospectus 2007, we run a chart for each team that shows its week-to-week performance. Jacksonville's chart looks like a massive earthquake.
Going back to 1996, six teams have finished in the top 10 of the DVOA ratings without a winning record: the 1999 Raiders, 1999 Cowboys, 2002 Chiefs, 2004 Bengals, and last year's Jaguars and Giants. The 1999 Cowboys were an aging team on the way down and went just 5-11 the next season. The other three teams to accomplish the feat before last year all won their divisions the next year, with an average of 12 wins apiece. (For an explanation of why the Giants won't make a leap forward this year, click here.)
If you don't trust our DVOA ratings, just consider the Jags' point totals. In every sport, statisticians use a principle called the Pythagorean projection to estimate wins based on points scored and allowed. The Jaguars outscored their opponents 371-274 last season, which projects to 10.8 wins. That's a far cry from 8-8. From 1983 through 2005, 15 teams had a Pythagorean projection at least 2.5 wins better than the team's actual record. On average, these teams gained more than three wins the following year.
Jacksonville's road to the postseason got a little more complicated last weekend. After publicly backing quarterback Byron Leftwich all offseason, head coach Jack Del Rio suddenly cut Leftwich and handed the offense to backup David Garrard. Nevertheless, the quarterback switch doesn't change Jacksonville's good offensive line, great running tandem, and suffocating defense. If the Jags play just as well as last year, it will probably be good enough to put them in the playoffs. With just a little improvement, they'll challenge the Colts for the division title.
Process of Elimination
Sometimes the best way to find a surprise playoff contender is to look for the teams that are going to decline. Every team that falls out of the postseason leaves room for another team to join the party.
The Chicago Bears, for example, are very likely to decline in 2007, although it has nothing to do with the "Super Bowl Loser's Curse." The dominating defensive and special teams play that Chicago used to go 13-3 last year is simply not sustainable. (To understand why, check out the Bears chapter in Pro Football Prospectus 2007.) The Chicago offense is not good enough to drive the team to a 10-win season if the defense and special teams drop from brilliant to merely very good. That means another team has to pick up the slack. It's not likely to be Minnesota, with an inexperienced quarterback throwing to a cadre of below-average receivers, or Detroit, which has a high-flying passing game but no running attack and a very poor secondary.
The team that will benefit from Chicago's misfortune, then, is the Green Bay Packers. Brett Favre spent the entire offseason complaining that the front office wasn't doing enough to build the Packers into a contender. Perhaps he doesn't remember that Green Bay went 8-8 last year and was just a tiebreaker away from making the postseason. One or two more wins this season will almost guarantee them a playoff spot in the weak NFC.
The Packers have some problems on offense—a rookie running back, questionable receivers—but even in old age Favre is still an above-average quarterback. It's also worth noting that Green Bay broke in two rookie offensive guards in 2006; all five starters return this year, and as a rule, offensive lines gradually improve as players gain experience working together.
The heart of this year's Packers, however, will be the defense. The Pack has a number of young, improving players in its front seven—defensive ends Aaron Kampman and Cullen Jenkins, linebackers A.J. Hawk and Nick Barnett—backed by two very good, experienced cornerbacks. If Favre would stop bellyaching for a minute, he might notice that the Packers have built him a winning team. That winning team just happens to reside on the defensive side of the locker room.