Skip to main content

Running the Numbers: Reading Sunday's Win-Probability Graph


Despite an 8-8 record in 2012, the Cowboys held a lead just 22 percent of the time. That's really an astounding statistic and probably the biggest negative the team had heading into 2013. For at least one game, though, the Cowboys' luck shifted.

Looking at data on win probability, we can visualize the ebb and flow of the Cowboys' big opening night victory over the Giants. Using historic game data as a foundation, Advanced NFL Stats publishes visualizations that update in real time to display a team's probability of winning a game at any point. Down 10-3 and facing a third-and-10 at the opponent's 25-yard line with 3:20 to play in the second quarter? The win probability graphs can give you an indication of how likely you are to win, and an accurate one at that. I highly recommend monitoring them on game day.

The Cowboys' win probability graph from Sunday night's victory paints a picture we rarely saw last season, one with the 'Boys controlling the direction of the contest.


Due to a relatively fast pace from the Cowboys and a lack of rushing from both squads, there were an abundance of plays from scrimmage in this contest. And of those 133 plays, Dallas found themselves as an underdog on just five of them. That's a sharp contrast from a year ago.

You can see the Cowboys were the favorites to win from about two-thirds of the way through the first quarter until the final second of the game. At its worst point, Dallas still owned a 40 percent chance of winning.

I labeled three major plays that changed the course of the game – two that went in the Giants' favor and one for Dallas.

  • Play 1: Tony Romo Interception – 31 percent drop in win probability

Holding a 3-0 lead with a third-and-11 at the Giants' 14-yard line, the Cowboys owned a 71 percent chance of winning the game. That changed in a hurry when Terrance Williams ran the wrong route and Ryan Mundy picked off Romo, returning it to the Cowboys' 1-yard line.

We should really be talking more about DeMarco Murray's hustle on the play to stop the touchdown and the defense's ability to hold New York to a field goal. When the Giants tied the game at three, the Cowboys still had a 52 percent chance to win. Had they given up a touchdown, it would have been just 37 percent. It's those "tipping point" plays that really affect the outcome of games.

  • Play 2: Victor Cruz 70-Yard Touchdown – 15 percent drop in win probability

It looked like the Cowboys would head into halftime with a 10-point lead, but that changed when Dallas showed some confusion on defense and Cruz burned them for a quick 70-yard score. Because Dallas was already up 10 points, their odds of winning the game decreased by only 15 percentage points. I ran the numbers and had this game been tied, the 'Boys' probability of winning would have dropped by 20 percent.

  • Play 3: Brandon Carr Pick Six – 31 percent increase in win probability

There was no bigger play in this contest than Carr's interception return for a score. The play was mostly luck, but it's luck that Dallas is glad to see. After the score, the Giants had just a three percent chance of coming back for the win.

Beating the Giants' D in Week 12

These two teams meet again in Week 12. The Cowboys obviously can't count on again securing six takeaways, meaning they're going to lose if the offense performs as poorly as they did on Sunday night.

The lack of sharp jumps in win probability is really a reflection of how well the Giants played Dallas defensively. Sitting back in Cover 2 and Cover 2 Man-Under for much of the game, the Giants were content to allow Dallas to run the ball and throw it underneath, just as long as Dez Bryant was slowed down on the outside.

It worked. The fact that Dallas scored 36 points masked the fact that they played really poorly on offense. The running game was average at best and Romo totaled only 5.4 yards per attempt through the air. Last year – one that was considered a "down year" for Romo – he didn't have a single game with a total that low.

The Cowboys' biggest offensive play of the game, by far, was the 15-yard touchdown strike to Jason Witten. However, even that outcome increased Dallas' win probability by just seven percentage points. When the Cowboys travel to the Meadowlands in mid-November, the underneath passing game will be important, but they're absolutely going to need to find a way to get the ball downfield more effectively.

Once the bounce of the ball doesn't go in their favor, the dark side of randomness that always rears its head, will the Cowboys be able to stretch out a win?

This article has been reproduced in a new format and may be missing content or contain faulty links. Please use the Contact Us link in our site footer to report an issue.

Related Content