I’ve already seen a multitude of opinions as to why the Cowboys lost a game they clearly should have won in Detroit yesterday. There are lots of secondary reasons as to why Dallas is now sitting at 4-4, but the fundamental underlying issue is something I’ve discussed in the past: the Cowboys are a results-oriented franchise.
By that, I mean they aren’t proactive in making changes, waiting until something goes wrong until they try to fix it. Coach Jason Garrett preaches the importance of “the process,” but he doesn’t practice what he preaches.
If you want to more thoroughly understand why the Cowboys lost on Sunday, let’s take a look back to the Cowboys’ win over the Eagles just eight days ago. Here’s a series of tweets I posted within about 40 minutes of each other:
“Cowboys will lose right around one expected point and four percent win probability by punting on fourth-and-one here.”
“Cowboys lost right around 0.8 expected points and another three percentage points of WP by punting on fourth-and-five at the opponent’s 37.”
“Another punt, another 0.9 expected points and three percentage points of WP lost. This is fun, isn’t it?”
“You can see how these decisions can quickly add up. We’re talking about 1/10th of a loss expected from just one type of decision in one quarter.”
These days, we have this newfangled invention called “math” that we can use to help us determine the best calls in certain situations. Using tools like the Fourth Down Calculator at Advanced NFL Stats, you too can play along. And although a nerdy guy like me put this stuff together, it’s based on real NFL game data—what has actually worked for teams in the past.
As has been the case since he was named head coach, Garrett either didn’t know the percentages or simply ignored them. But the Cowboys won the game, so all of that didn’t matter, right?
The decisions to punt in the first quarter against the Eagles were just as bad as some made by Garrett on Sunday, regardless of the outcome. The Cowboys need to identify and fix these weaknesses before they manifest themselves in the loss column.
The Cowboys’ handling of the end game is consistently among the worst of any team in the league, and we saw that again in Week 8. I’ll get to that in a minute, but they might not have even been in that situation had they more appropriately managed earlier choices.
The worst of the bunch was a second quarter field goal try on a fourth-and-two at the Lions’ 35-yard line. Using the Fourth Down Calculator, we can establish some baseline stats for the situation. Again, these are based on how offenses have performed in the same situation in the past.
In attempting the field goal, the Cowboys lost 0.98 expected points. Another way of thinking about that is if the Cowboys were to play out that situation 1,000 times, they would score right around a full point more, on average, by going for it over kicking a field goal. The ‘Boys lost a decent chance to score a touchdown on that drive instead of coming away with three points—points that ultimately decided the game.
You might argue that kicker Dan Bailey made the field goal, justifying Garrett’s decision to kick it. I have a feeling many people within the Cowboys’ organization would propose that rationale, but it’s just wrong. It’s that sort of “ex post facto” thinking that has resulted in mediocrity in Big D.
Further, the numbers might be even more in favor of going for it when we factor in the specifics for Dallas. Bailey is 9-for-14 in his career on 50-plus yard field goals. This one was from 53 yards out, and we wouldn’t expect Bailey’s expected conversion rate to be much higher than the 50 percent used in the calculator. But even if we bump Bailey’s expected conversion rate to, say, 70 percent, the Cowboys should still have gone for it.
That’s especially true when you consider that the Cowboys have an above-average offense. They might have been playing poorly at that time, but it’s hard to think their chances of converting a fourth-and-two were worse than that for the typical NFL offense.
The End Game
Let’s start with the series before the field goal that everyone is discussing: the one that started on the Dallas 19-yard line with 3:27 left in the game. Up by three points, the Cowboys had possession of the ball and Detroit had all three timeouts.
With the way the Lions were moving the ball, it was obvious the Cowboys weren’t in a position to do everything in their power to run clock. A first down was crucial, so Dallas really should have been running close to their normal offense.
Instead, they kept the ball on the ground on both first and second down, losing two yards in the process. The first down run was understandable in that situation, but why run a draw on second-and-13? Although the contest wasn’t completely about point-maximization at that point, Dallas really should have been treating the situation as if a three-and-out loses the game.
The Cowboys punted and the Lions went four-and-out. Again, that doesn’t justify the Cowboys’ earlier play-calling by any means; it was wrong to run on second down when they did it, and it was wrong after the Lions turned it back over to Dallas.
Sitting at the Lions’ 31-yard line with a first-and-10 and 1:24 on the clock, the Cowboys basically could have taken a knee three times and ensured giving back the ball to Detroit deep in their own territory with under 30 seconds to play and no timeouts.
I don’t agree with the assessment that kneeling was the right move—they should have run to try to get a first down, as they did—but when Dallas faced third-and-14 at the Lions’ 35-yard line with 1:14 to play, the coaches should have told the offense that there’s no way they can commit a penalty or fumble the ball.
Instead, Tyron Smith got flagged for holding. The Lions declined the penalty, but the clock still stopped at 1:07. At that point, the Cowboys faced a fourth-and-five at the Lions’ 26-yard line with a three point lead. Detroit had no timeouts. What’s the right move?
The Cowboys attempted a field goal, which was absolutely the worst choice they could have made. In kicking the field goal, the Cowboys’ win probability decreased from a possible 94 percent to 87 percent. That might seem small, but it’s a huge difference. If the Cowboys were in that situation 1,000 times, the Lions would win nearly twice as often with Dallas attempting a field goal over going for it, or even punting (yes, punting!).
The numbers appropriately capture why a field goal is the wrong choice, but there are two major factors that influence the stats being the way they are. First, the downside of missing a field goal is monumental. Kicking at the 26-yard line, a miss would set up the Lions at around the 34-yard line with over a minute to play, needing three points to tie.
Second, the value of a six-point lead in that situation isn’t that much greater than a three-point lead, and there are two reasons for that. First, most coaches in that situation mistakenly think “we can’t give up three points or we lose” (or in the case of Detroit “all we need is three points and we win”). But that’s not true. The Lions actually needed three points to have roughly a 50 percent chance to win the game in overtime. There’s a big difference.
Second, because of how coaches think about three-point leads, the Cowboys almost helped Detroit in kicking the field goal. In addition to getting the ball back, the Lions also knew they needed a touchdown. And that touchdown wasn’t to force overtime, but to win outright.
Had Dallas punted or even gone for it and failed, the Lions likely would have had a much different mindset. They probably would have been playing for a field goal instead of the win.
There’s some game theory involved with such a choice—an aspect of decision-making the Cowboys have messed up time and time again. Until the Cowboys truly embrace analytics—until they stop judging the value of their decisions based on the outcome—they’re going to be a talented team with a mediocre record.