Todd Bowles stood in front of the Buccaneers defense Thursday morning and issued a warning:
Don’t get complacent.
He wasn’t referring to the team’s recent wins against Atlanta and Jacksonville. He was looking ahead to this Sunday’s game against Indianapolis and referring to the offense’s approach to fourth down, particularly its recognition that it doesn’t have to be just a kicking down.
Stop the Colts on third down? Great. Now get up and stop them again.
“Coach told us today, do not run off the field,” said Bucs linebacker Shaquil Barrett, who has recorded a team-high four sacks on third down. “They might go for it. The first meeting this morning, that’s what he said.”
Public perception is that a revolution has taken hold in the NFL, that teams aren’t as afraid to go for it on fourth down, that coaches are getting more aggressive.
Frank Reich, whom the Colts hired from the staff of the analytics-driven 2017 Super Bowl champion Eagles, has developed a reputation for being one of those coaches.
Indianapolis has attempted fourth-down conversions at a 19.8 percent clip this season, the seventh-highest rate. Once the Colts cross midfield, they’re even more likely to go for it. Their 32.5 percent attempt rate ranks fourth.
Even so, Reich’s aggressiveness is relative. NFL coaches are a conservative, risk averse lot. They tend to give excess weight to potential losses over potential gains. Reich’s not immune.
Revolution? This isn’t it.
Frank Frigo, co-founder of Louisville-based sports analytics and consulting firm EdjSports, says coaches have a long way to go. They continue to make decisions that hurt their teams’ chances of winning. Still, he sees the shift in the way analysts and fans talk about fourth downs.
“I will say that I’ve never seen the discussion at this level,” he said. “Just watching games, the way teams go for it now, there’s not a lot of head-scratching about it.”
Case in point: During last Sunday’s game between San Francisco and Baltimore, Ravens quarterback Lamar Jackson ran on third and 6, typically a passing situation. When he gained only 4 yards, Fox play-by-play announcer Kevin Burkhardt asked, “And now do the Ravens go for it?”
“Likely, yes,” broadcast partner Charles Davis said. “Because that’s what they do. Remember, their third-down run was to set up this fourth-down attempt. … No surprise at all, Baltimore’s going to go for it because that’s their MO.”
The Ravens did go for it, and they succeeded. They ultimately capped their drive with a go-ahead field goal.
“That is commentary you would have never heard before a couple of years ago,” Frigo said.
Reich is sometimes wrong
Heading into Week 5, the Colts were heavy underdogs against the unbeaten Chiefs. Kansas City’s explosive offense was averaging 34 points per game, and it was a foregone conclusion that it would hang at least another 30 against Indianapolis.
Reich steered the Colts to victory thanks in part to a pair of fourth-down decisions late in the game. In the fourth quarter, with a three-point lead, he went for it twice — on fourth and 1 from the Kansas City 39-yard line with 13:52 left and on fourth and 1 from the Kansas City 27 with 11:15 left. Both calls were correct, according to EdjSports’ game-winning chance model, which measures, in real time, how a coach influences his team’s win probability at the point of decision. In other words, the model evaluates the process, not the result. (The results, by the way, were good. Indianapolis converted both times.)
The advantage the model has over us humans? It isn’t swayed by emotion.
“It will indicate which choice, on average, will produce the most wins,” Frigo said. “What you find out that can be a little bit counterintuitive is that retaining possession even in situations where the regret is very high — in your own territory, on fourth and short — is often right by a lot.”
Reich’s calls against the Chiefs were relatively easy decisions. He made a much more courageous call in Week 2 against the Titans when, with a 19-17 lead and 2:24 left, he chose to go for it on fourth and 1 — from his own 35-yard line. Fail, and the opponent takes possession already in range for a game-winning field goal. Succeed, however, and you maintain possession and drain the clock. The Colts converted and did just that, but the decision alone increased their win probability by 27.8 percentage points, according to EdjSports.
Though Reich is seen as a devotee to fourth-down analytics, he is prone to making the same serious blunders that most coaches make. In Week 11 against Jacksonville, his fourth-down decisions cost Indianapolis 8 percentage points in win probability. In Week 12 against Houston, his decisions cost Indianapolis 11 percentage points. The Colts beat the Jaguars but lost narrowly to the Texans and come to Tampa fighting for their playoff lives.
As it turns out, Bucs coach Bruce Arians’ “no risk it, no biscuit” mantra isn’t rooted in reality. On fourth downs, he has been more bashful than bold, and it has cost his team dearly.
The bill, according to EdjSports: 102 percentage points of win probability, the third-highest total in the NFL and 21 more percentage points than Reich. Put another way, Arians’ fourth-down management alone has cost Tampa Bay an entire game’s worth of win probability, and then some. A caveat: Not all fourth-down decisions are the same, or as Frigo puts it, not every coach receives the same test. Some coaches face more high-leverage situations than others.
Still, Arians has made a series of clear and obvious mistakes — 26 by EdjSports’ count (Reich has made 18). Take, for instance, his decision to kick a field goal on fourth and 2 late in the Bucs’ Week 3 game against the Giants. The kick was good, but it hurt more than it helped. It extended Tampa Bay’s lead to only six points and essentially forced New York to play for the win when it otherwise might have played to tie and force overtime.
Less than three minutes later, the Bucs faced fourth and 1 from the Tampa Bay 31 — a nearly identical situation to the one the Colts faced against the Titans. Arians chose to punt, and the Giants scored a game-winning touchdown on their next possession.
And you thought Tampa Bay had issues at quarterback. And at running back. And at offensive tackle. And at defensive back.
The team’s in-game risk management and decision-making lay bare a more foundational flaw: its failure to seize the potential competitive advantage that data analysis provides. There’s a common refrain in analytics circles: We don’t hear from the Bucs.
On Sunday, when the Colts approach the line of scrimmage on fourth down, the Bucs will be ready to stop them. In the long term, however, here’s what they cannot stop: Math. Slowly but steadily, it’s taking over the game. Those who embrace it will win. Those who avoid it will fall further behind.