Aggressiveness Index 2019

John Harbaugh set an all-time record for fourth-down aggressiveness in 2019. If you've paid any attention to the NFL this year, this fact will not surprise you.
Football Outsiders introduced the concept of Aggressiveness Index way back in Pro Football Prospectus 2006. The goal was to find a way to rank coaches based on their tendencies on fourth downs in a manner that was easy to understand but accounted for the different rates at which the average coach will choose to "go for it" in different situations. Although no NFL coach is as aggressive as the data suggests he should be, we discovered there is quite a wide range of fourth-down tendencies among coaches.
Aggressiveness Index numbers center around 1.0 and generally describe how much more (or less) likely each coach is to go for it on fourth down compared to his peers; for example, a coach with 1.20 AI is roughly 20 percent more likely to go for it than an average coach in equivalent situations. That's based on historical averages, of course; the numbers on aggressiveness have been much higher over the last two years. Take a look at the leaguewide numbers for the last six seasons. This year, the overall attempts on fourth down were almost 50% higher than expectation based on historical norms.
NFL Aggressiveness Index, 2014-2019 | |||||
Year | AI | Overall Att. |
Expected Att. |
Overall Opps. |
Rate |
2014 | 0.98 | 197 | 201 | 3217 | 6.1% |
2015 | 1.00 | 211 | 212 | 3265 | 6.5% |
2016 | 1.06 | 233 | 220 | 3193 | 7.3% |
2017 | 1.03 | 215 | 209 | 3259 | 6.6% |
2018 | 1.37 | 272 | 199 | 3049 | 8.9% |
2019 | 1.46 | 305 | 209 | 3044 | 10.0% |
The Aggressiveness Index excludes obvious catch-up situations: third quarter, trailing by 15 or more points; fourth quarter, trailing by nine or more points; and in the last five minutes of the game, trailing by any amount. It also excludes the last 10 seconds of the first half, and it adjusts for when a play doesn't actually record as fourth-and-short because of one of those bogus delay of game penalties that moves the punter back five yards. Only regular season is included.
From 1989 through 2018, only 10 coaches had a season where they went for it on fourth down more than twice as often as the expectation: in other words, an AI above 2.0. (Sean Payton did it twice.) No head coach had ever had a season with an AI above 3.0.
John Harbaugh's AI in 2019 was 3.95.
Harbaugh also set a record by going for it on 23% of qualifying fourth downs. AI is different than purely measuring go-for-it rate because coaches around the league are more likely to go for it in certain situations than in others. But Harbaugh set a record in both categories. He was one of four head coaches to put up AI above 2.0 in 2019.
Here's the full list of all the seasons with AI above 2.0 (minimum 60 opportunities). Yes, there are some surprises on here. Each coaches is listed with the number of times the coach went for it in qualifying situations, the number of qualifying opportunities, and the number of times an average coach would be expected to go for it based on historical norms given the down, distance, score, and location on the field.
Top AI Seasons, 1989-2019 | ||||||||
---|---|---|---|---|---|---|---|---|
Rank | Coach | Team | Year | AI | Go | Opp | Rate | Exp |
1 | John Harbaugh | BAL | 2019 | 3.95 | 20 | 87 | 23.0% | 5.07 |
2 | McCarthy | GB | 2018 | 2.63 | 7 | 77 | 9.1% | 2.67 |
3 | Coslet | CIN | 1999 | 2.49 | 10 | 98 | 10.2% | 4.02 |
4 | Seifert | SF | 1994 | 2.34 | 10 | 60 | 16.7% | 4.28 |
5 | Tice | MIN | 2003 | 2.31 | 13 | 95 | 13.7% | 5.63 |
6 | Nagy | CHI | 2019 | 2.26 | 11 | 112 | 9.8% | 4.88 |
7 | John Harbaugh | BAL | 2018 | 2.25 | 16 | 107 | 15.0% | 7.11 |
8 | Schottenheimer | KC | 1991 | 2.22 | 12 | 87 | 13.8% | 5.40 |
9 | Payton | NO | 2007 | 2.22 | 16 | 92 | 17.4% | 7.20 |
10 | Shurmur | NYG | 2019 | 2.21 | 9 | 84 | 10.7% | 4.06 |
11 | Parcells | NE | 1994 | 2.12 | 16 | 80 | 20.0% | 7.53 |
12 | Pederson | PHI | 2019 | 2.06 | 14 | 103 | 13.6% | 6.79 |
13 | M. Smith | ATL | 2009 | 2.02 | 16 | 101 | 15.8% | 7.93 |
14 | Payton | NO | 2014 | 2.01 | 7 | 81 | 8.6% | 3.49 |
15 | Fisher | TEN | 2005 | 2.00 | 17 | 107 | 15.9% | 8.49 |
Harbaugh has always been one of the more aggressive head coaches in the league on fourth downs, but nothing like what we've seen the past two seasons and especially in 2019. He had a career AI of 1.04 through 2017.
A few notes on some of the coaches on the table above with lower rates but high AI:
We never ran a specific Aggressiveness Index article on the website with the 2018 numbers, though they were featured in Football Outsiders Almanac 2019. Yes, Mike McCarthy led the league in AI in the 12 games he was the Green Bay head coach in 2018, followed by three of the same head coaches who are on the list for 2019: Harbaugh, Doug Pederson, and Pat Shurmur. McCarthy only went for it on 1 of 4 qualifying fourth-and-1s in 2018, but he was aggressive in situations where the historical baseline is really low including fourth-and-4 in the second quarter against both San Francisco in Week 6 and Arizona in Week 13. McCarthy is known for getting very conservative with a lead and for kicking short field goals on a couple of specific fourth-and-1s in the 2014 NFC Championship Game. But he has a career regular-season AI of 1.17, making him one of the more aggressive head coaches of the last few years on fourth downs.
Bruce Coslet in 1999? Over his career, Bruce Coslet had a higher AI (1.56) than any other head coach with at least three years of experience between 1986 and 2017. (If we include 2018 and 2019, he is passed by Doug Pederson.) Bruce Coslet simply did not want to kick long field goals in 1999. In two different first quarters, Coslet went for it from the opposing 34 on both a fourth-and-9 and a fourth-and-11 instead of bringing in either a kicker or a punter. He also had Jeff Blake attempt a pass on fourth-and-3 from the Pittsburgh 33 with a fourth-quarter lead. Blake was sacked but the Bengals held on to win, which brought their record to 2-10 that year.
How are Matt Nagy and Pat Shurmur above 2.0 for 2019 despite lower raw go-for-it rates? Nagy only went for it on 4 of 9 qualifying fourth-and-1s this year but late in the season he was going for it in some long-distance situations such as fourth-and-6 from the Detroit 32 in Week 13 and fourth-and-7 from the Green Bay 41 in Week 15. Both of those were in the second quarter. Shurmur only went for it on 3 of 7 qualifying fourth-and-1s but went for a couple of first-half fourth-and-2s and a fourth-and-4 against the Jets in Week 10 where Daniel Jones threw an 8-yard pass and Darius Slayton tacked on another 31 yards of YAC for a touchdown. (The play result doesn't matter for AI, I just figured it was notable.)
Here are the full Aggressiveness Index results for 2019. With the dramatic change in frequency of going for it on fourth down over the last couple years, 28 of 32 main head coaches come out with an AI above 1.0. That includes both Super Bowl coaches, but Andy Reid and Kyle Shanahan are still down near the bottom of the league in Aggressiveness Index. Reid has historically been one of the most conservative coaches in the league on fourth downs, although he's been more aggressive the last couple years including an AI of 1.55 in 2018. Shanahan is very conservative, and was 31st with an AI of 0.78 in 2018. He does have a high "go for it" rate, but he tends to go for it in situations where historically the average head coach has also gone for it. Kyle's conservative nature is a surprise because his father was a more aggressive coach, with a career AI of 1.10.
Aggressiveness Index, 2019 | |||||||
---|---|---|---|---|---|---|---|
Rank | Coach | Team | AI | Go | Opp | Rate | Exp |
1 | John Harbaugh | BAL | 3.95 | 20 | 87 | 23.0% | 5.07 |
-- | Fewell | CAR-2 | 3.70 | 2 | 21 | 9.5% | 0.54 |
2 | Nagy | CHI | 2.26 | 11 | 112 | 9.8% | 4.88 |
3 | Shurmur | NYG | 2.21 | 9 | 84 | 10.7% | 4.06 |
4 | Pederson | PHI | 2.06 | 14 | 103 | 13.6% | 6.79 |
5 | Zimmer | MIN | 1.92 | 9 | 89 | 10.1% | 4.68 |
6 | Marrone | JAX | 1.84 | 7 | 104 | 6.7% | 3.81 |
7 | Kitchens | CLE | 1.84 | 11 | 93 | 11.8% | 5.99 |
8 | Kingsbury | ARI | 1.82 | 13 | 98 | 13.3% | 7.14 |
9 | B. Flores | MIA | 1.74 | 11 | 94 | 11.7% | 6.30 |
10 | Rivera | CAR-1 | 1.70 | 9 | 84 | 10.7% | 5.31 |
11 | Reich | IND | 1.65 | 15 | 94 | 16.0% | 9.09 |
12 | O'Brien | HOU | 1.62 | 13 | 91 | 14.3% | 8.04 |
13 | Taylor | CIN | 1.57 | 11 | 98 | 11.2% | 7.00 |
14 | Payton | NO | 1.56 | 10 | 96 | 10.4% | 6.42 |
15 | McDermott | BUF | 1.40 | 13 | 114 | 11.4% | 9.29 |
16 | Lynn | LAC | 1.40 | 6 | 76 | 7.9% | 4.30 |
17 | Belichick | NE | 1.39 | 13 | 124 | 10.5% | 9.36 |
18 | Callahan | WAS-2 | 1.37 | 3 | 68 | 4.4% | 2.19 |
19 | LaFleur | GB | 1.29 | 10 | 108 | 9.3% | 7.77 |
20 | Vrabel | TEN | 1.27 | 6 | 92 | 6.5% | 4.72 |
21 | Quinn | ATL | 1.25 | 8 | 86 | 9.3% | 6.38 |
22 | Arians | TB | 1.24 | 8 | 91 | 8.8% | 6.43 |
23 | Tomlin | PIT | 1.23 | 8 | 103 | 7.8% | 6.52 |
24 | Gase | NYJ | 1.17 | 10 | 101 | 9.9% | 8.58 |
25 | Garrett | DAL | 1.16 | 6 | 83 | 7.2% | 5.19 |
26 | McVay | LAR | 1.15 | 6 | 96 | 6.3% | 5.20 |
27 | Reid | KC | 1.09 | 8 | 90 | 8.9% | 7.37 |
28 | K. Shanahan | SF | 1.06 | 12 | 94 | 12.8% | 11.31 |
29 | Patricia | DET | 0.93 | 6 | 103 | 5.8% | 6.47 |
30 | Jon Gruden | OAK | 0.90 | 8 | 85 | 9.4% | 8.85 |
31 | Carroll | SEA | 0.86 | 6 | 98 | 6.1% | 6.94 |
32 | Fangio | DEN | 0.68 | 5 | 105 | 4.8% | 7.40 |
-- | Jay Gruden | WAS-1 | 0.00 | 0 | 21 | 0.0% | 0.22 |
Thanks to Jim Armstrong, who came up with the idea of Aggressiveness Index and computes it for every season.
Comments
13 comments, Last at 30 Jan 2020, 8:11am
#4 by meep 42@hotmail com // Jan 28, 2020 - 5:54pm
I wonder what the correlation between AI and team record is. It appears to me that teams that were out of it most of the year tend to have more aggressive coaches -- because it doesn't matter. Teams like KC and SF who were strong throughout the year simply have less incentive to be aggressive.
#5 by A-a-ron // Jan 28, 2020 - 7:53pm
If I'm understanding, fake punts and fake field goals are marked as "go". Fakes so often seem like a bad idea compared to leaving the offense on the field, e.g. the Texans' fake punt in the divisional round. It would be interesting to see the difference if fakes were counted "no go".
#8 by Laserblast // Jan 29, 2020 - 3:23am
What was Lynn's score before the last game of the season (so, excluding the last Chiefs game)?
He seemed to go several times to win against the Chiefs, even though the Chargers were already out of the playoffs. At the time, I felt it was only to project his reputation as a "play to win" coach.