≡ Menu

2016 Game Scripts in Review: Pass Identity Ratings

Did you know: the Patriots and 49ers both threw on 54% of their plays last season. Both teams ranked in the bottom five last year in pass ratio, i.e., their percentage of plays that were either pass attempts or sacks. But the teams both passed infrequently for different reasons: New England didn’t want to pass much because they were often playing with the lead and were milking the clock; San Francisco didn’t want to pass much because their passing game was not very good and because both Colin Kaepernick and Blaine Gabbert were two of the most run-heavy quarterbacks in football.

San Francisco was undoubtedly a run-heavy team last year, but the Patriots?  Of course not.  No team with Tom Brady is run-happy, but the game scripts incentivized New England to be run-happy.  Regular readers know about Game Scripts, which is simply the average points margin over every second of every game.  New England had a Game Script of +7.7 last year, the highest margin in the NFL.  This means if you were to write down the amount by which New England was leading for every second of every game last year, and calculated the average, you would get a 7.7 point lead.  The graph below shows the Game Script for all 32 teams last year on the Y-Axis, along with their Pass Ratio (pass attempts plus sacks divided by total plays) on the X-Axis.

As you can see, the Patriots were a big outlier. New England passed on 54.4% of plays, which ranked 28th in the NFL, but the team’s +7.7 Game Script was the best in the league. The solution is to normalize both sets of ratings and then compare the stats. Here’s how. New England was 1.30 standard deviations below average in terms of pass/run ratio, known as the team’s Z-Score. But the Patriots were 2.62 standard deviations above average in game script. If you add those two numbers together, you get +1.32, which is New England’s pass identity. That was the fourth strongest in the league last year.

RkTeamPass RatioPR RkGame scriptGS RkZ-Score PRZ-Score GSPass Identity

The Ravens were the most pass-happy team in the NFL and also has a pretty strong Game Script; therefore, Baltimore stands out as the most pass-happy team in the 2016 season. The Ravens ranked 1st in pass attempts and 30th in rushing attempts, which is remarkable for a team that outscored its opponents by over a point per game. But the Ravens couldn’t find a consistent healthy force at running back, and were content to have Joe Flacco drop back play after play and throw short passes. Flacco, who earned his reputation as a big armed downfield passer, averaged a pathetic 9.9 yards per completion last year. With Steve Smith (11.4 yards per completion last year), Dennis Pitta (8.5), and Kyle Juszczyk (7.2) now gone — and they were three of the Ravens four leaders in completions in 2016 — it will be interesting to see how the offense operates this year.

Speaking of Juszczyk, he went to San Francisco in the offseason, joining the most run-happy team in the NFL. But the 49ers also changed head coaches, with Kyle Shanahan — last seen as the OC in Atlanta — now in charge. And the Falcons had the second strongest pass identity last year. And with both Kaepernick and Gabbert gone, replaced by Brian Hoyer and rookie C.J. Beathard. The 49ers also added Pierre Garcon and Marquise Goodwin, so the San Francisco offense — while not necessarily any good — should still be a lot different this year.

  • I’m actually looking forward to San Francisco this year. Brian Hoyer was on his way to a solid season last year before Clay Matthews broke his forearm. When he and Kyle Shanahan were together in Cleveland in 2014, they were a fun team to watch and Hoyer led the league in yards per completion. I don’t necessarily think SF will be better than a 5-11 team, but they should produce some exciting action.

    • Richie

      Hoyer always seems to mix in a little bit of promise, then gets hurt.

      Except that playoff game for Houston. No promise there.

  • evo34

    Of all the things you do on this site, this is my favorite. Any chance of posting a YTD table for 2017 that is updated weekly?

    • It’s possible! Anything is always possible. 🙂

  • Pingback: Weekly Sports Analytics News Roundup - September 12th, 2017 - StatSheetStuffer()

  • Chris Sweeney

    Great article, trying to find some success in DFS (and season long) and really want to figure out strategy and how/why games play out. Does anyone know where to find a one-stop place to go where they discuss what kind of offenses are used in the NFL? I’m finally understanding the game a little more, and, for example how do you know if a QB favors a slot receiver or a tight end, if that makes sense. Again, really nice read here, thank you, Chase.