The graph below shows the Game Script (X-Axis) and Pass Ratio (Y-Axis) for every game this season. For new readers, a team’s Game Script is simply its average margin of lead (or deficit) over every second of a game. Pass ratio is simply passing plays (pass attempts plus sacks) divided by all offensive plays (pass attempts + sacks + rushing attempts).

As you can see, there’s a clear relationship between the two variables: on average, the better the Game Script, the lower the Pass Ratio.

We can also create season ratings of Game Scripts and Pass Ratios for each team. Let’s use the Patriots and Eagles as examples.

New England has had an average Game Script across its 8 games of +13.1. This year, New England’s pass ratio in those 8 games is 58.2%. Philadelphia has had an average Game Script of -2.2, and a pass ratio of 55.0%. It might strike you as odd to see that New England has a higher pass ratio — i.e., it’s passed more frequently — than Philadelphia. It should! That’s because New England has the strongest passing identity in the NFL, while the Eagles have the strongest rushing identity in the NFL.

The Patriots have, by far, the best average Game Script this season; all else being equal, you might expect New England to therefore have the lowest pass ratio in the NFL. Instead, the team is barely below average, ranking 19th in percentage of passing plays. Philadelphia has the 25th-best Game Script this year, as the Eagles had a -4.4 Game Script against Atlanta, a -4.9 GS against Detroit, a -9.9 vs. Minnesota and a -14.8 against Dallas. And yet the Eagles have just the 25th-best highest passing ratio in the league! That’s very run-heavy, as noted yesterday.

The graph below shows the Game Script (X-Axis) and Pass Ratio (Y-Axis) of each offense this season. I have shaded in team colors the Patriots and Eagles data points: [continue reading…]

On Friday, I looked at each team’s offensive pass identity. Today, the reverse: a look at defensive pass identities.

The Jacksonville Jaguars were one of the best teams in the NFL last year. Jacksonville had the 3rd-best points differential in the NFL in 2017 after 1 quarter (+41), the 5th-best after 2 quarters (+86), the 4th-best through 3 quarters (+109), and tied for the 3rd-best points differential overall. Unsurprisingly, Jacksonville had the 4th best average Game Script last season, which means you should expect the Jaguars to be run-heavy and Jacksonville’s opponents to be pass-heavy.

On the offensive side, things held to form: Jacksonville rushed on 49.4% of plays, the highest ratio in the NFL last season. But on defense, that wasn’t the case: teams passed on only 56.4% of plays against the Jaguars last year! Consider that opponents passed on 65.4% and 62.3% of plays against the Eagles and Vikings, teams that finished 3rd and 5th in Game Script last season.

On the other side, the Tennessee Titans.  Last season, the Titans were an average team, finishing with a slightly negative Game Script. And yet teams passed on them like they were the Patriots! In fact, opponents passed against New England on 61.6% of plays and against the Titans on 61.7% of plays.  The Titans Game Script was 0.27 standard deviations below average, while the opponent pass ratio was 1.42 standard deviations above average. As a result, the Titans have a Defensive Pass Identity of +1.69, making them the defense teams were most likely to pass against. [continue reading…]

Yesterday, we looked at offensive Pass Identity grades. Today, we are going to use the same process to analyze the data for defenses.  Yesterday’s post is required reading to understand how Pass Identity grades are calculated, but here’s one update.  While we can use the same numbers for Game Script (including the 3.27 number for standard deviation and 0 for average), that’s not the case for defensive Pass Ratio. There, while the average is roughly the same at 58.29%, the standard deviation is much smaller at 2.84% (it was 4.66% for the offenses).

Let’s use the Lions as an example.  Detroit had an average Game Script of +0.4 last year, meaning the Lions were leading by, on average, 0.4 points during every second of every game.  That was 0.11 standard deviations above average. [continue reading…]

As we did last year, today I’m going to calculate the final 2014 Game Scripts and Pass Identity data.  Every week during the season, I write about the Game Scripts from the previous weekend. For new readers, the term Game Script is just shorthand for the average points differential for a team over every second of each game. You can check out the updated Game Scripts page, which shows the results of all 256 games from 2014, and you can read the history behind the metric here.

Let’s begin by looking at the 2014 Game Scripts numbers. The Packers held an average lead of 6.9 points during their regular season games, the highest average in all of football. Because Green Bay was so good, Aaron Rodgers and the Packers weren’t very pass-happy; in fact, the Packers ranked just 21st in pass attempts. That’s why Jordy Nelson and Randall Cobb, as good as their raw numbers were, look even better in some advanced metrics. In some ways, the Packers were the victims of their own success last year, as Green Bay was — by far — the best first half team in the NFL in 2014. That led to the high Game Script number, and a lot of casual dress second halves. [continue reading…]

I’ve published the Game Scripts data from every game this year at the 2014 Game Scripts page, available here. What would it look like if we plotted Game Script score (on the X-Axis) against Pass Ratio (on the Y-Axis) for every game this year? Something like this:

As you move from a more negative Game Script to a more positive one, the expected Pass Ratio decreases. But the relationship is not purely linear: in extreme cases, the Pass Ratios tend to move a bit more towards league average, and I think that trend is probably even stronger than it might appear on this graph. In any event, you can derive a best-fit polynomial equation from that data, which could give us an expected Pass Ratio.

For example, with a Game Script of +3.0, teams should be expected to pass on 56.3% of all plays. But in the Eagles/49ers game, Philadelphia passed on 78.6% of all plays. At the time, I thought it was an oddly pass-happy performance, as it turns out, it was the most pass-heavy game of the year, with a pass rate 22.3% higher than expectation.

If we perform that calculation for every game this year, we can derive season grades. Let’s look at the Colts line in the table below. In 7 games this year, Indianapolis has an average Game Script of +7.9, which happens to be the highest in the NFL (the table is fully sortable). Based on how each game has unfolded, Indianapolis would be expected to pass on just 52.7% of all plays if it was an average team; however, the Colts have passed on 59.4% of all plays. That means the Colts have passed 6.7% above expectation, the second highest rate in the NFL this year. The table below lists that data for each team through week 7: [continue reading…]

Steve Buzzard has agreed to write another guest post for us. And I think it’s a very good one. Steve is a lifelong Colts fan and long time fantasy football aficionado. He spends most of his free time applying advanced statistical techniques to football to better understand the game he loves and improve his prediction models.

---

Last month, I wrote about how to project pass/run ratios using offensive Pass Identities and the point spread. However, this methodology only considers one side of the ball. Can we actually improve our projections model using both offensive and defensive Pass Identities? As it turns out the answer is yes.

First, I started off by creating defensive Pass Identities using the same methodology found here. The first thing I noticed was the standard deviation of pass ratios for defenses was only 3.0% compared to 5.1% for offenses. This led me to believe that offenses control how much passing goes on in a game more than defenses. I was glad to see this as it confirmed most of my previous research as well. Given this, it wasn’t appropriate to use a standard deviation of 3.0% for defenses in my projection while using a standard deviation of 5.1% for offenses. Instead, I used the combined standard deviation of all 64 offensive and defensive pass ratios, which turned out to be 4.17%. This doesn’t change the order of passer identities much but obviously does increase the deviation from the mean for the offensive side of the ball and decrease it for the defensive side. [Chase note: Determining the best way to handle the differing spreads between offensive and defensive pass ratios is a good off-season project; in the interest of time, I advised Steve to split the difference and move ahead with the analysis.]

Now that we have a Pass Identity grades for both sides of the ball, we can add a strength of schedule adjustment, too. To make the SOS adjustment, I simply took the average of the defensive Pass Identities played by each offensive unit and the average of the offensive Pass Identities played by each defensive unit. As expected the SOS adjustments had a much larger impact on the defensive Pass Identities than the offensive Pass Identities.
[continue reading…]

Every week this season, I’ve written about the Game Scripts from the previous weekend. For new readers, the term Game Script is just shorthand for the average points differential for a team over every second of each game. You can check out the updated Game Scripts page, which shows the results of all 256 games this year. Week 17 saw some big blowouts and some tight finishes: Peyton Manning, Andrew Luck, and Drew Brees all led their teams to convincing wins against overmatched opponents, while Green Bay and Philadelphia clinched playoff berths with close wins.

Week 17 was unremarkable from a Game Scripts perspective, although I’ll note that Denver’s win over Oakland produced a Game Script of 21.6, the fifth highest average margin of the year (and the best by the Broncos this year). On the comeback side, only three teams won with negative Game Scripts, and two of those wins (Green Bay, Carolina) were back-and-forth contests. That means we should all take a moment to reflect on the resolve and grit of the San Diego Chargers, who overcame an average deficit of 4.6 points (in regulation) to force overtime and eventually defeat the Chiefs B team.

The full Game Scripts data from week 17: [continue reading…]

I’ve been posting the Game Scripts numbers each week this season, and now have a full page dedicated to the results from every game at the top right of your screen. But the best use of Game Scripts is to adjust Pass ratios for teams to understand their true Passing Identity. Here’s how you do it.

1) Calculate how many standard deviations above/below average each team is in Game Scripts. The average Game Script, of course, is zero. The standard deviation through five weeks is 4.69, so the Broncos (8.43 Game Script) are 1.80 standard deviations above average in Game Script.

2) Calculate how many standard deviations from average each team is in Pass Ratio, defined as pass attempts (including sacks) divided by total plays. The average Pass Ratio through five weeks is 59.8%, while the standard deviation among the thirty-two teams is 6.7%. The Giants (excluding last night’s game) lead the league in Pass Ratio at 71.8%, which is 1.79 standard deviations above the league-average Pass Ratio.

3) Add how many standard deviations above/below average each team is in both Game Scripts and Pass Ratio. To convert these into an Index (and a more intuitive number for folks), multiply that result by 15 and add it to 100. So a team that has a Pass Identity that is 1 standard deviation above average will be at 115, while a team that is 1.6 standard deviations below average will be at 76.

Here are the results:
[continue reading…]