At a time when NFL head coaches are increasingly willing to weigh the probabilities and be aggressive, the Cowboys’ Jason Garrett chose to play not to lose last night. It backfired spectacularly.
Faced with a fourth-and-1 from the Texans’ 42 with 5:40 to play in overtime, Garrett decided to punt. Chris Jones pinned Houston at its own 10-yard line, but Dallas never got the ball back and lost, 19-16. Garrett explained that his decision was rooted in establishing field position:
“Yeah, it was a long one [yard],” the Cowboys coach told reporters. “You know, we had a third-and-2 and we didn’t make much on it and we just felt like at that point in the game, the way our defense was playing, the idea was to pin them down there. Chris [Jones] did a great job with the punt. They got the ball on the 10-yard line and hopefully you make a stop and you win the game coming back the other way with a game-winning field goal.”
Hindsight is easy, and it’s true that the distance was a yard-plus, and that Dallas had attempted to run Ezekiel Elliott on the previous play, only to come up short. But the Cowboys seem to be built specifically for these types of scenarios: They drafted Elliott fourth overall in 2016, and even though he managed just 54 yards on 20 carries last night, the Cowboys are a league-best 94.7 percent (18 of 19) on fourth down and one yard or fewer since they picked Elliott. Even allowing for fourth and two or fewer yards, the Cowboys have converted 90 percent (18 of 20). The league average conversion rates in that time are 65.3 percent and 63.5 percent, respectively. Outside the red zone, the league average conversion rate on fourth-and-1 is 66.2 percent since 2016. And the Cowboys are an NFL-best 91.7 percent.
Just last week against the Lions, Garrett elected to go for it on a fourth-and-1, with Elliott running straight ahead and picking up two yards and a first down. The game circumstances were a bit different then: Dallas was leading by three in the third quarter and had the ball on Detroit’s 3-yard line. Garrett’s explanation at the time was interesting:
What about win probability? Inside the Star’s Kevin Brady determined that Dallas had a 52 percent chance of winning at the time of Garrett’s decision last night, and that the punt only dropped their odds to 50 percent—an insignificant difference. However, going for it and failing to convert would have dropped the Cowboys’ chances to 34 percent—a 16 percent drop. But going for it and picking up the first down with a two-yard gain would have boosted their chances of winning to 74 percent—a 24 percent spike. The Cowboys had more to gain than they stood to lose, by a difference of eight percentage points, or 50 percent.
Guess who else wasn’t happy about Garrett’s call?
Back in August, ESPN’s Seth Walder wrote a terrific piece about the NFL’s player-tracking data, which is being shared among all 32 teams for the first time this season. One of Walder’s conclusions was that a chasm is growing between NFL teams that apply quantitative analysis to their decision-making, and those that don’t. But there’s nothing new about coaches who have been getting smarter about going for it. Meanwhile, Jason Garrett is in his ninth season as Dallas’s head coach. He’s made the playoffs just twice.