In my recent article for The Fifth Down, I detailed why I thought Jason Garrett was wrong to run the ball three straight times late in the fourth quarter against the Patriots. Even with 3:36 to play and a three-point lead, I thought Garrett could benefit from being more aggressive in that situation.
Some of you have chimed in with your thoughts on the situation, two of which I will address here. The first is that we (I) vilified Garrett when throwing the ball in the third and fourth quarter against the Lions, yet I am also criticizing him here for not airing it out. Can I have it both ways?
The other idea is that, had the Cowboys been more aggressive and Romo thrown an interception, the damage to his psyche and the team’s season would be far worse than the current situation. With a relatively soft remaining schedule, the team still has every opportunity to get back into the playoff hunt, and the confidence of the quarterback is vital to that process.
Both are solid thoughts and have some merit, but I do not wholeheartedly agree with either. A 24-point lead is far different from a three-point lead (even in the fourth quarter versus with 3:36 to play), and the numbers show that. While the Cowboys’ win percentage peaked at 85 percent against New England, it was as high as 97 percent against Detroit, including 95 percent just before the start of the fourth quarter.
Courtesy of AdvancedNFLStats.com
You could make a solid argument that, if you do nothing but run and secure the football from the third quarter on against the Lions, you cannot lose the game. At worst, your chances of winning would be no less than that 97 percent mark at which they peaked.
How many times out of 100 can the Lions score 24 points in 25 minutes without the aid of takeaways? One or two? Whatever it is, it is certainly low enough that Dallas should not have been throwing the ball on first down in the second half.
Even with under four minutes to play, a three-point lead and possession of the football, the Cowboys should have attacked on Sunday. As I detailed in the New York Times article, the fact that the Cowboys failed in their aggressive attack against Detroit really has no bearing—or shouldn’t have—on their aggressiveness versus the Pats.
Even if the situations were the same, a sample size of one game is hardly enough to affect future decisions. You don’t stop going for it on 4th-and-1 because you failed once, or twice, or any number of times, because those numbers will level out in the long run. Similarly, you don’t scrap your offensive approach because it hasn’t worked of late.
Of course, statistics cannot account for everything, and one of those things (as it currently stands, anyway) is the impact a decision might have on the team’s future, particularly in a psychological way. If the psyche of the team, namely Tony Romo, would have broken following a potential interception, could it be possible that running the ball, although perhaps not statistically superior to passing in the narrow sense, was the correct decision when we take the bigger picture into account?
While I see why one might think this, I don’t think it should be a factor in a team’s decision-making. Would Romo’s confidence plummet following another late-game interception? Perhaps. Would it affect the team’s ability to win down the road? Maybe.
But he is a professional football player, and if his psyche is so fragile that he cannot block out past interceptions and respond well from adversity, the team probably wouldn’t be able to win with him anyway. Further, a player who could potentially lose confidence so easily surely wouldn’t feel good about his offensive coordinator taking the ball out of his hands late in a three-point game. If Garrett doesn’t believe in his quarterback, who does?
Luckily, I don’t think Romo’s psyche is truly that fragile. He’s a confident player, and he possesses every requisite trait to lead this team to the playoffs. Perhaps unfortunately for Dallas, he sometimes isn’t put in situations which are optimal for him to display those traits...and win football games.