≡ Menu

So far this season, one thing is obvious: the NFL scheduled Thursday night games every week for the express purpose of screwing with people who do data analysis. Even though week three has started, I’m not ready to close the door on week two, in which nearly every game was competitive into the fourth quarter. But that doesn’t mean the game script for every game was close.

Winner LoserBoxscorePFPAMarginGame Script
Green Bay PackersWashington RedskinsBoxscore38201817.9
Atlanta FalconsSt. Louis RamsBoxscore3124713.4
Oakland RaidersJacksonville JaguarsBoxscore199108.2
Seattle SeahawksSan Francisco 49ersBoxscore293267.7
New England PatriotsNew York JetsBoxscore131036.4
Denver Broncos@New York GiantsBoxscore4123185.5
Cincinnati BengalsPittsburgh SteelersBoxscore2010103.9
Miami Dolphins@Indianapolis ColtsBoxscore242043.3
San Diego Chargers@Philadelphia EaglesBoxscore333033.2
New Orleans Saints@Tampa Bay BuccaneersBoxscore161422.2
Houston TexansTennessee TitansBoxscore302461.3
Chicago BearsMinnesota VikingsBoxscore313011
Kansas City ChiefsDallas CowboysBoxscore17161-0.1
Baltimore RavensCleveland BrownsBoxscore1468-0.8
Buffalo BillsCarolina PanthersBoxscore24231-1
Arizona CardinalsDetroit LionsBoxscore25214-1.3

Steven Jackson was injured early in his revenge game against the Rams (and is expected to miss two-to-four weeks), but consider: Atlanta had 45 pass attempts against just 16 running plays in a game in which their average lead was 13.4 points. And that was with a gimpy Roddy White! Last year, I noted that the Falcons were the most pass-happy team in the NFL after adjusting for game script, and it appears that the model hasn’t changed in 2013.

There weren’t any huge comebacks this week, a byproduct of all the competitive games. The Cardinals scored nine points in the fourth quarter to beat the Lions, in a game where Detroit’s offense was shut out in the second half. Matt Stafford and company gained just 90 yards and four first downs on 24 second half plays, enabling the Cardinals to steal a win. Half of the team’s six second half drives were three and outs, one was a fumble on the second play, and the final drive was five plays and ended on downs. The only successful drive of the half was a 51-yard march that put the Lions at the Cardinals 27, but David Akers’ field goal attempt was blocked.

But while the offense had an off day, there’s a hidden factor that explains why Detroit didn’t score more than 14 offensive points (DeAndre Levy intercepted a Carson Palmer pass for 66 yard touchdown, accounting for the other seven points).

Let’s take a look at each team’s average yardline on each snap1; remember, having the ball on your opponent’s one-yard line gets recorded as a “99″ while having the ball at your own one is simply a “1.”

Team# playsAvg Yardline

Half of Detroit’s twelve possessions started at their own twenty, which is bad enough, but another five began inside the Lions own ten yard line! The 12th drive was an end-of-half kneel down that began at the Lions 21.

Chicks dig kickers

Chicks dig kickers.

How did that happen? Jay Feely produced a touchback on six of his seven kickoffs; on the seventh, a holding penalty negated a 36-yard Micheal Spurlock return, pinning Detroit at its own eight. Detroit’s other five drives began after a Dave Zastudil punt, as Detroit’s only forced turnover was returned for a touchdown. How did Zastudil fare?

  • His first punt came at the Arizona 47 and landed at the Lions 11; a holding penalty put Detroit back on their own six.
  • Zastudil’s second punt was on 4th-and-1 from the Detroit 43. In a scoreless game, you want to go for it in that situation, but Zastudil did his job: a fair catch pinned Detroit at their own 7.
  • The third punt preceded the end-of-half kneeldown drive. Zastudil punted 48 yards from his own 39, and Spurlock was only able to return it 8 yards. That net of 40 yards was enough — after a Reggie Bush run for no gain — to convince the Lions to pack it in for the half (despite having three timeouts and 45 seconds).
  • In the second half, Zastudil was even better. On his own 36, he boomed a 52-yard punt which, thanks to a great play by the coverage unit, was returned by Spurlock for -6 yards.
  • The final punt of the game came with the Cardinals down by two with 5:42 left in the game. Facing 4th and 7 at the Lions 44, Zastudil pinned Detroit at their own eight yard line, which set up a three-and-out.

All told, that’s six of seven kickoffs for touchbacks and four of five punts that pinned the Lions inside their own 11-yard line. That’s how Detroit’s average drive started (excluding the end-of-half kneel) at just their own 14-yard line. And a result, the average offensive play for the Lions over the course of the game took place at their 32-yard line. According to Pro-Football-Reference’s Expected Points analysis, Feely, Zastudil, and the Cardinals’ coverage units added 7.02 points on just those twelve plays. Feely was also perfect on four field goals, including a 47-yarder, and one of the plays of the day was Justin Bethel’s blocked field goal. All told, this was an outstanding day for the Cardinals special teams, and a key reason why they won the game despite throwing a pick six.

  1. Subject to the same caveats described in this post. []
  • ch

    Something of a stretch to credit Arizona’s special teams for two Detroit penalties.

    That expected points analysis doesn’t seem right. It’s way overvaluing touchbacks on kickoffs. Graham Gano is getting credit for creating a touchdown’s worth of value in week 2, in a League where 62% of kickoffs result in touchbacks. Better kickoffs end up inside the 20. All of Gano’s kickoffs have gone for touchbacks, but he’s ranked just 14th currently in net kickoff average.

    Blair Walsh is getting almost 8 points of value in week 2 despite allowing returns of 76 and 80 yards.

  • These are absolutely invaluable. Amazing how looking at the score tells you story Y but looking at the gamescript will show you story X. I hope we are fortunate enough to receive these each week.

  • If you run an internet business, you may not be paying customers yet but they marketing database model have
    shown interest in learning more about your offer straight to the point.
    In your message, and tweak it for various audiences.