Super Bowl LII will forever be known for one play — the Philly Special. The Eagles faced a fourth-and-goal late in the second quarter against favored New England leading 15–12. Rather than take the sure three points from the one-yard line, Eagles coach Doug Pederson dialed up a trick play that Nick Foles caught for a touchdown, pushing the lead to 10.
Pederson’s aggressiveness was not limited to that play. He also converted a fourth-and-1 at his own 45 with 5:39 left trailing 33–32. That drive gave the Eagles the lead for good. Those two fourth-down decisions helped the Eagles to a long-awaited Super Bowl title.
For a long time, mathematical applications of going for it on fourth down were met with shrugs of indifference — at best — from coaches. Outright scorn was often the response.
But statistical analysis suggesting that head coaches don’t go for it often enough has been part of the conversation since at least the 2000s, if not way earlier. As football analysis has grown more sophisticated, this was a hanging curveball just waiting to be hit.
Football Outsiders has a statistic called the Aggressiveness Index that measures how much more (or less) likely a head coach is to go for it on fourth down compared to his peers. The decisions are compared against the historical average rather than the current season. The number centers around 1.0, so a coach who was at 0.8 would be 20 percent less likely to go for it, for example.
Super Bowl LII happened at the conclusion of the 2017 season. Since then, the league-wide Aggressiveness Index has exploded:
NFL Aggressiveness Index, 2014-19
Not only did Pederson end a long Super Bowl drought for the Eagles, but he also brought the power of aggressiveness to the attention of his peers. Coaches love to copy a winner.
John Harbaugh’s aggressiveness index in 2019 was 3.95. Matt Nagy, Pat Shurmur and Pederson all went for it twice as often as the historical league average. For the first time in modern NFL history, coaches started going for it on fourth-and-short on their own side of the field. Even coaches you maybe wouldn’t call analytical savants tried it, coaches like Houston’s Bill O’Brien and Minnesota’s Mike Zimmer. (Zimmer, of course, would also say that going for it against Washington at home was his dumbest decision as Vikings head coach. Minnesota won the game, 19–9.)
Various head coaches were criticized last year for not being aggressive, but perhaps none more so than Pete Carroll. Carroll and offensive coordinator Brian Schottenheimer put together an overtly 1980s game plan. They had more rushing attempts than any NFL team that trailed as often as it did, and they won 11 games mostly thanks to Russell Wilson’s four fourth-quarter comebacks. They won games in which they trailed 20–6 (Cleveland), 21–7 (Tampa), 20–14 (Rams), and 10–0 (49ers) by being forced to abandon their run-heavy game plan and having an ace quarterback run a hurry-up offense.
Carroll never learned his lesson, explaining loudly throughout the season about how much he preferred to be conservative. With the Seahawks down 21–3 to Green Bay in the Divisional Round, Wilson led another furious comeback to get the score within 28–23. But with 2:41 left in the game, on fourth-and-11, the Seahawks punted and never got the ball back.
“We were thinking about going for it in that sequence, but not at fourth-and-11,” Carroll said in his postgame presser. “We thought our odds were so low. We had all the clock, we had the time, we had all the opportunities to stop them to get the ball back.”
Carroll was one of only four full-time head coaches to have a below-average aggressiveness index last season, and the only one of the four to have a winning record. (The other three were Jon Gruden, Vic Fangio and Matt Patricia.)
We should go for it more — but when?
The NFL is increasingly becoming a league that relies not on time of possession but rather on total number of possessions — and what you do with those possessions.
The Chiefs-Ravens game in Week 3 last season was a case in point. Concerned about giving Patrick Mahomes any extra possessions to score, the Ravens aggressively went all-in on going for it on fourth down. The game featured one of the worst performances of Lamar Jackson’s season; he completed just 22-of-43 passes for no touchdowns and 267 yards. But the Ravens had a dominant run attack, converted three-of-four fourth-down attempts and came back to make a 17-point deficit a five-point deficit in a 33–28 loss. In the context of that quarterback performance, it was a wildly impressive outing for Baltimore.
It’s also worth exploring the downside of such a strategy. When the Ravens exited the playoffs after a 14–2 regular season, they failed on two key fourth-down conversions and four total fourth-down conversions in their 28–12 loss to the Titans. It is true that the Titans were playing well, and it is also true that the Titans had spent a lot of their offseason importing run-stuffing players after struggling at times in short-yardage situations.
But it could also be argued that the Ravens were a little too eager to start trying to convert those fourth downs. Facing a 7–0 deficit against the Titans at the start of the second quarter, the Ravens attempted to convert a fourth-and-1 from their own 45-yard line, which seems like something an underdog would do and not a heavy favorite playing at home. Their failure gave Tennessee a short field and led to a quick touchdown. But you can understand the decision — the Ravens had been eight-for-eight on fourth-and-1 during the regular season.
There’s a reason fourth-down conversions have tended to mainly be championed as an underdog tactic by analytics: Increasing variability can sometimes have dire consequences in a low-possession game.
Interestingly, despite his reputation, Bill Belichick was one of the more conservative coaches last year. He had more qualifying opportunities (124) to go for it on fourth down than any head coach, but he went for it only 13 times, putting him 17th in the NFL in fourth-down aggressiveness. He explained his philosophy to ESPN’s Greg Garber back in the early 2000s:
“The more third-and-1s you make, the more likely you are to go for it on fourth-and-1,” Belichick said then. “You also factor in the defensive side of the ball. How good of a short-yardage or goal-line team is your opponent? When you add those two up and you’re real good and you don’t think the other team is that proficient at it, that’s one thing. If it’s vice versa, then maybe that skews you.”
Belichick’s team last year struggled to convert third downs and lost fullback James Devlin to injury in mid-September, further complicating their short-yardage game plan. On third-and-4 or less, the Patriots converted only 43-of-83 attempts. For perspective, the Ravens converted 48-of-78. The Patriots converted only twice in eight fourth-down attempts of four yards or less.
In short, while it’s true that on a general level, teams should be going for it on fourth down more often than they have been — and that adjustment has begun — it’s also true that the particular situation may affect a coach’s thinking. Your quarterback may be nursing a bum knee and unable to execute your money play — how does your play-caller react? Is his second-favorite play call in this situation something that he has confidence in? If it’s not, will he be able to see past his gut feeling to look at the percentages?
If there was a poster child for fourth-down attempts last year, it was Harbaugh, who pushed that strategy to the limit and showed more success with it than any other coach; the Ravens converted 17-of-24 fourth-down attempts (70.8 percent) to lead the NFL. Jackson’s read-options keyed an attack that was a case study in NFL run game success, and when you combined that with a historic level of aggressiveness, it yielded an offense the likes of which we haven’t seen in the NFL in a long time. Unlike Pederson’s Eagles, though, the Ravens sputtered at the end of the season.
Their aggressive fourth-down strategy turned on them.
It took a Super Bowl win to get head coaches to be more aggressive in the first place. It would probably take a Ravens Super Bowl win to get fourth-down rates to approach the actual optimal numbers that statisticians would suggest.