Analytics didn’t actually support Brian Daboll’s aggressive 2-point gamble in Giants’ win. Here's why

·7 min read

Brian Daboll had himself a debut on Sunday.

The new New York Giants head coach oversaw a rally from a 13-point halftime deficit to a 21-20 upset of the Tennessee Titans, punctuated by a go-ahead two-point conversion in the waning minutes in the game. In search of a stamp to mark the transition from the Joe Judge era, Daboll found an emphatic one.

But analytics writers immediately pushed back. Was it the optimal decision? It worked this time, but would his gamble pay off enough in the long run if made repeatedly? Did the upside of an aggressive Week 1 overshadow the downside of a potential failed conversion? Yahoo Sports reached out to analytics experts who provided insight on both fronts.

Here's how it happened

With 1:06 remaining in the fourth quarter, Giants quarterback Daniel Jones found tight end Chris Myarick in the end zone for a touchdown. The six-point score left Daboll with his fateful decision. A successful extra point would tie the game at 20-20. A successful two-point conversion would give the Giants a 21-20 lead. Failure on either effort would leave the Giants trailing 20-19 with a kickoff looming.

With the likely outcome of his first NFL game as a head coach in the balance, Daboll chose aggression. He sent his offense out for a two-point conversion. Jones took a shotgun snap, rolled to his right, then shoveled a pass underneath to running back Saquon Barkley, who barreled his way through Titans defenders into the end zone for a successful conversion and a 21-20 lead.

But that wasn't the ball game. Tennessee had a chance to respond, and it came close to converting. After a touchback on the ensuing kickoff, the Titans drove the ball 41 yards in 1:02 to leave kicker Randy Bullock with a 47-yard field goal attempt that would win the game if successful.

It was not. Bullock's attempt sailed wide, and the Giants celebrated a 1-0 start for the first time since 2016.

Analytics don't support Daboll's call

Seasoned bettors know that playing the result of any given decision is folly. So did the fact that it paid off mean that Daboll's gamble was the right decision?

On a strictly analytical front, it wasn't. Ian O'Connor of Football Outsiders told Yahoo Sports that attempting an extra point would have produced a 4.5% greater win probability than a two-point conversion attempt.

The Giants entered Sunday's game as 5.5-point underdogs. O'Connor told Yahoo Sports that if they had successfully converted an extra point, they would have been approximate 2-to-1 underdogs from that point forward considering that the Titans had a chance to respond in regulation and would have remained favorites in the event of overtime.

Meanwhile, Football Outsiders projects that a successful two-point conversion would give the Giants just a 55% probability to win the game. Again, the Titans had a chance to respond in regulation in either scenario. In the end, the Giants needed Bullock — a career 76% kicker from 40-49 yards — to miss a 47-yard kick. Football Outsiders also projected that a below-average Giants offense had a below-average chance of converting a two-point conversion.

The analytics don't support Brian Daboll's late-game gamble in Sunday's win over the Titans. But that doesn't mean that it was the wrong decision.
The analytics don't support Brian Daboll's late-game gamble in Sunday's win over the Titans. But that doesn't mean that it was the wrong decision. (Christopher Hanewinckel/Reuters)

At the same time, an extra point is likewise no guarantee. Just ask the Bengals. NFL extra-point success rates have ranged from 94.2% to 92.9% since the NFL moved them back 13 yards in 2015. Given that the Titans would have a chance to respond in regulation in either scenario, doesn't it make sense for the underdog Giants to go for two?

Again, no. As O'Connor explained, teams that are trailing play differently than teams that are tied in the late stages of a game.

"Teams behave differently when down by one vs. when tied," O'Connor told Yahoo Sports. "If they’re trailing, they have to score or they lose, so they take more risks and are more aggressive. However, when tied, they still have overtime to fall back on, so they don’t usually play as aggressively, especially with Tennessee being favored in overtime."

Could Daboll's call still be right call in spite of the analytics?

So the analytics are definitive. They don't support Daboll's decision. But are the hard numbers the end of the discussion? Certainly not. O'Connor acknowledges as much while making the analytics case.

The Football Outsiders equation considers a number of game-specific factors — most notably who's on the field, in this case Jones, Barkley, the Giants offensive line and the Titans defense. It doesn't, however, adjust for in-game performance.

In this instance, Barkley returned from three injury-marred seasons on Sunday to look like the difference-maker selected second overall in the 2018 draft. Barkley torched the Titans defense to the tune of 9.1 yards per carry, 194 yards from scrimmage and a touchdown in addition to the two-point conversion. There's something to be said for riding the hot hand, especially when that hand is as talented as Barkley's. The hot-hand concept isn't factored into the analytics equation.

Meanwhile, there's only one chance to make a first impression while initiating a culture change. And Daboll seized it.

"In this case, we know Brian Daboll is trying to build a winning culture, which almost certainly played into his decision to trust his guys and go for the win," O'Connor said. "We won’t fault him for this decision as much as we do some of the egregious 4th down calls we’ve seen in the past and will certainly come in the coming weeks."

Saquon Barkley doesn't care what the analytics say about the Giants' decision to go for two. (Wesley Hitt/Getty Images)
Saquon Barkley doesn't care what the analytics say about the Giants' decision to go for two. (Wesley Hitt/Getty Images)

Daboll's aggression paid off in more ways than the win column. On a Giants franchise reeling from the Judge debacle (not to mention Pat Shurmur, Ben McAdoo and ex-architect Dave Gettleman), Sunday's win meant more to a locker room in the wake of five straight losing seasons. They gushed about Daboll postgame, with Barkley delivering his rookie head coach the game ball.

"When we scored, I was on the field and I saw him put up the two sign, and we kind of made eye contact," Barkley told reporters of Daboll on Sunday. "He gave me that look. I knew what the play was going to be. He gave me that look, and I kind of looked back at him and said, 'F yeah."

"We called it up, and we were able to execute the play and get in. He's a man of his word. He told us he's gonna be aggressive. He told us he's gonna to lean on the players to make plays. In that situation, he did exactly that. When you have a coach like that, it's definitely going to make you go out there and fight for him and execute in those situations."

Game theory and analytics are and should be significant factors in modern coaching decisions. But as displayed on Sunday, they aren't the end of the discussion. Good luck making this particular analytics case to Barkley — or anyone else in the Giants locker room.