Week 2 fantasy projections

Tony Romo has a difficult matchup against a good Seattle Seahawks defense this week. AP Photo/Bill Kostroun

There were several surprising performances in Week 1 by offensive players such as Kevin Ogletree and Stephen Hill, although you wouldn't have been surprised by the latter if you read our rookie projections in "Football Outsiders Almanac 2012." While those breakouts may get everyone's waiver-wire juices flowing, this is a defensive matchups piece, so we'll instead break down some surprising Week 1 performances by team defenses according to our play-by-play efficiency metric, defense-adjusted value over average (DVOA, explained here). Truth be told, we don't start adjusting for opponent until after Week 4, so the Week 1 rankings you see below are based on unadjusted value over average (VOA).

The purpose of this little exercise is to see if our preconceptions about the quality of NFL defenses in 2012 held in Week 1, and if not, whether things are likely to balance themselves out as the season progresses. Below is a table showing three different sets of rankings for each defense: its DVOA ranking at the end of last season, its projected DVOA ranking for this season and its actual VOA ranking based on Week 1.

So, just as an example, we see that Football Outsiders projected San Francisco's defense to fall from third to 11th this season, and it actually had the 18th-best VOA in Week 1. Of course, that was against the Green Bay Packers, so it's likely that the 49ers' ranking will improve as the season goes on. They're not playing Aaron Rodgers & Co. every week.

For other defenses such as the Pittsburgh Steelers and Kansas City Chiefs, league-low Week 1 performances can be explained away relatively easily. For the Steelers, they were without a couple of key defenders, and were facing Peyton Manning in Denver. Same goes for the Chiefs, who fielded a makeshift unit against a high-powered Atlanta offense.

A handful of defenses, though, played poorly -- sometimes much more poorly than we projected -- despite not having much justification for it. For instance, the Jacksonville Jaguars' defense has no business being as inefficient as it was against a Minnesota Vikings offense that finished 24th in DVOA last season, and had Adrian Peterson playing at less than 100 percent. Similarly, the New York Giants' defense, which we actually projected to improve this season, laid an egg against Dallas in Week 1. Forcing cornerback Michael Coe onto the field is a ready-made excuse, but many of the biggest passing plays for the Cowboys came against starters Corey Webster and Kenny Phillips.

On the flip side, several defenses played much better than expected, but their Week 1 performances can be explained away given the matchup. For instance, Philadelphia, Seattle, Houston and Arizona all excelled against what we knew to be below-average offenses heading into the season. Oakland, on the other hand, played a good offense that happened to be missing its star running back.

However, a few defenses played extremely well -- and above expectation -- but did so against teams with good offenses. This would suggest that the defensive improvements in Week 1 were likely legit. New England, for instance, had one of the worst defenses in the league last season, sometimes featuring wide receivers as defensive backs. Nevertheless, based on offseason changes and some statistical indicators, we projected them to improve by leaps and bounds. Their Week 1 opponent, Tennessee, while not an offensive juggernaut, did finish 15th in DVOA last season, and has enough talent to exploit a bad defense.

The other defenses that fit in this mold were Denver and Cleveland, although the latter unfortunately just lost Pro Bowl cornerback Joe Haden to a four-game suspension. The Broncos, however, had the third-best Week 1 VOA despite playing a Steelers offense that finished sixth in DVOA last year. True, their starting running back was out, but backup Jonathan Dwyer held his own, and no one would argue that Pittsburgh's passing game couldn't pick up the slack given its quarterback and pair of 1,000-yard receivers.

In sum, as we move forward until the opponent adjustments to DVOA kick in, it will be useful from a fantasy perspective to keep a watchful eye on the defenses that already seem much worse than expected (Jaguars, Giants) or much better than expected (Denver). If these trends continue over the next couple of weeks, you'll be able to play matchups with these defenses before most other people realize that their preseason conceptions were actually misconceptions.

Here are the best and worst fantasy matchups for Week 2:


Ryan Fitzpatrick (plus-3 points)