NFL 2016 Week 3 Projections Roundup: Pittsburgh Steelers at Philadelphia Eagles

Sometimes, the stars align.

Sometimes, lightning is caught in a bottle.

Sometimes, despite all logic and contrary to every popular opinion, serendipity controls the universe.

This Philadelphia Eagles season has that type of “feel.” After wins against two teams that wouldn’t elicit fear against Peewee League opponents, the Eagles find themselves with a 2-0 record and look to make a statement against a more worthy adversary. And as serendipity would have it, they get to do it at home. The Pittsburgh Steelers roll into Lincoln Financial Field with a 2-0 record as well, with future Hall of Famer Ben Roethlisberger and All-Universe wide receiver Antonio Brown. What do statistical models make of it all?

First, here are roundups from prior weeks:

Week 1

Week 2

On to this week:

VEGAS (Steelers -3.5, 60.7% win probability)

Last week: Eagles +3, 40.2% win probability

Vegas lines for the Eagles have been volatile from week-to-week, and even within week. After opening as 5.5 point underdogs, the line is now a more Eagles-friendly 3.5 points. In other words, Vegas thinks the Eagles’ probability of winning has increased from 29.9% to 39.3%. Pittsburgh is rightfully favored, and home field advantage isn’t much of a factor here (just 1.7 points). Instead, Vegas is getting swayed by Carson Wentz and the Eagles’ defense. Throw in a 4:25 PM start time for an in-state rival and emotions in Philly should be running high.

538 (Steelers, 56% win probability)

Last week: Eagles, 51% win probability

The Eagles continue to creep up the Elo rankings. Currently they are thirteenth (1531 points), while the Steelers are ranked third (1635). The last two opponents the Eagles have faced are currently ranked 29^th and 30^th, respectively (or not), so this is just more evidence of the fact that the Eagles haven’t encountered a respectable opponent as of yet. And the Steelers are indeed that, if not considerably more. 538 continues to be a bit bullish on the Eagles, who beat the Steelers 44,000 times in 100,000 simulations.

ESPN FPI (Steelers, 59% win probability)

Last week: Eagles, 41% win probability

In the last two weeks, the relative “Matchup Quality” of Eagles games has been understandably low. This week, the game is the fourth best on the docket (score of 72 out of 100). The top three are Cardinals at Bills (I know, but read why), Broncos at Bengals, and Vikings at Panthers. ESPN’s FPI gives the Eagles only a 41% chance of winning, but you know what, they did last week as well. And this week is at home.

Football Outsiders DVOA

What to make of this, ugh. The Eagles are currently the top ranked team in the league in terms of DVOA. This means a lot. Since 2003, every team ranked number one in DVOA after the second week of the season made the playoffs with at least a 10-6 record, and has averaged 12.5 (!) wins. Before you get too excited here, none of those top-ranked teams defeated opponents as lowly as the Bears and Browns in consecutive weeks this early in the season. In addition, there’s not enough information for FO to make adjustments for quality of opponent. So take this with a grain of salt. The Eagles defense is ranked second in the league and face a Steelers offense that is ranked seventh. The Eagles offense is ranked ninth; the Steelers defense is ranked fourteenth. One could argue that FO’s numbers forecast a probable Eagles’ win, but not by much.

Here are some other notable win probabilities:

Dr. K NFL Forecasts – Steelers, 63%, 28-23 (Last week: Eagles, 56.6%, 26-24)

Dratings.com – Steelers, 65.3%, win by 5 (Last week: Bears, 51.4%, win by 0.5)

Conclusion

Unlike last week, this week’s statistical forecasts consistently favor the Steelers, whose win probabilities range anywhere from 56% to 65%. According to the Trend Finder below, teams historically attack the Steelers defense more with the pass (63.3% of plays), yielding more yardage outside and more touchdowns in the middle of the field. If that trend holds, expect a bigger role for Carson Wentz, especially if the run game isn’t effective.