Giants vs. Cowboys, Week 12: Stats and analytics from the Giants’ 28-20 loss

What can the data and analytics tell us about the Giants’ loss to the Cowboys?

The New York Giants slipped to 7-4 on the season Thursday after falling to the Dallas Cowboys 28-20 in a tough Thanksgiving road game.

This was a game the Giants were always expected to lose, and were 10-point underdogs by the time the game kicked off. They were ultimately able to cover the spread and make it a one-score loss, but it was still managed to be a disappointing loss.

The Giants were seemingly poised to stun Dallas when they went into halftime with a 13-7 lead. But then the flow of the game flipped and Dallas used a more measured offense to limit mistakes, pulling away over the second half.

While the Week 11 loss to the Detroit Lions was a bigger let down (considering the Giants were favored to win and were savaged by injuries), losing a divisional game just hurts more.

What can the stats and advance analytics tell us about the game?

Plays of the game

Lately I’ve been making a habit of looking at the biggest plays of the game from the perspective of Expected Points Added (or lost), as well as Winning Percentage added (or lost). Often we can get caught up in the emotion of a game and allow that to color our perceptions.

For instance, the Giants’ failed attempt to convert a fourth-and-1 with a pass to Saquon Barkley has been described by fans as the play of the game. And while it was undoubtedly an important play that could have reverberations off the field, it was only the fifth-most important play in terms of winning percentage and the sixth-most important play from an EPA standpoint.

The play that had the biggest impact on EPA was actually Rodarius Williams’ first quarter interception of Dak Prescott.

Greg Olsen posited that Prescott thought Kayvon Thibodeaux crossed into the neutral zone, giving Dallas a free play. But whatever the reason, Prescott threw a lazy pass and Michael Gallup failed to work back to the ball. That gave Williams the time and opportunity to jump in front for the pick. The Giants were unable to convert the interceptions into points, going three-and-out. But denying Dallas the opportunity for even a field goal was important at that stage of the game.

The biggest play of the game from a Winning Percentage perspective was Dallas’ third-quarter touchdown to Dalton Schultz. Dallas was trailing New York 13-7 at that point, and the game was basically a toss-up, with Dallas having a 54 percent chance of winning. The touchdown gave Dallas a 14-13 lead and the game snowballed for the Giants. That play was also the second-most Expected Points added of the game.

You can easily see the impact of those plays on the Winning Percentage chart.

Spinning their wheels on offense

The Giants managed to make this a one-score game by the end, but they spent much of the game spinning their wheels on offense. Frustrations are starting to bubble up with regards to Mike Kafka and the offensive scheme in general.

For the most part, the Giants’ offense was similar to what we’ve seen throughout the year. As we’ve seen earlier in the year, the Giants opened the game with a pass-heavy script, testing the Dallas defense and trying to create opportunities to run later in the game. Unfortunately, the Giants’ running game has proven ineffective over the last two weeks.

Saquon Barkley ran a bit better in this game than against the Detroit Lions, but he still struggled to find room.

Unlike last week, Dallas devoted significant resources to stopping Barkley. He faced a stacked box on four of his 11 carries (36.36 percent), which is up from his season average of 22.77 percent of runs.

That did create opportunities for the Giants to throw the pass and the team did a good job of scheming separation for their depleted receiving corps.

Per NFL NextGenStats, Daniel Jones only threw into coverage on two of his 35 pass attempts (5.7 percent).

I had hoped that last week’s game against the Detroit Lions was the beginning of a turning point in the Giants’ passing attack. They were forced to look much further down the field than they had in previous games, which typically makes for a much more efficient offense. The early 44-yard pass to Darius Slayton reinforced that hope.

Unfortunately, much of the Giants’ passing game after that play was in line with what we’ve seen from them this year.

Daniel Jones ultimately threw as many passes behind the line of scrimmage (seven) as he did beyond 10 yards down the field. Even with the 44-yard pass, Jones’ average of 5.5 converted air yards and 6.7 intended air yards was roughly in-line with his season average of 5.1 and 6.5 air yards (respectively).

The short passes lead to a poor -0.1 EPA per play and a net -4 EPA for Jones, which rank in the 28th and 27th percentile among QBs since 2010, respectively. Those numbers were driven down considerably by the Giants’ terrible late-down passing. The Giants averaged -1.05 EPA per play on their 11 third and fourth down passes, with a success rate of just 18 percent.

And despite the short passes to players who are schemed open, Jones held the ball quite a bit. His 2.98 seconds to throw the ball was the longest of Thanksgiving, and only just quicker than his 3.04 season average (3rd longest in the NFL).

Despite the injuries the Giants did a pretty good job of keeping the Dallas pass rushers away from Jones.

Credit to the Giants’ defense

The Giants’ defense was always going to struggle this game. Dallas has an excellent offense and the Giants are dealing with an injury-riddled secondary. Wink Martindale attempted to compensate by ratcheting his aggressiveness up even higher than normal.

While that put a lot of pressure on defensive backs who should probably be special teamers, the Giants also forced Prescott to make a lot of tight window throws. Despite the talent discrepancy, the Giants allowed very little separation for the Cowboys’ top receivers.

Prescott threw into coverage on 30 percent of his passes, which was by far the highest rate on Thanksgiving. Those tight window throws lead to both of the Giants’ interceptions and were a big reason why they went into halftime with the lead.

The Giants also did a good job of pressuring Prescott. While they didn’t have a sack, the Giants’ pass rushers were consistently in the Dallas backfield. In particular, Kayvon Thibodeaux had a very active game — even if it didn’t show up in the box score.

But, credit where credit is due, Prescott played well in the second half.

Dallas was able to adapt to the Giants’ pressure packages and key in on one-on-one matchups in the secondary. They were able to isolate Darnay Holmes on CeeDee Lamb with predictable results, while Prescott made some impressive tight-window throws to Michael Gallup. Prescott completed 70 percent of his passes on the day, which was an incredible 17.7 points higher than his 52.3 expected completion rate based on player tracking data.

The Giants did what they could with a (severely) depleted roster, but it wasn’t enough.