The New York Giants hoped to stop their recent 1-3 slide with a win over division rival Washington Commanders in Week 13.
As it turned out, the Giants’ didn’t exactly stop the skid, but they didn’t lose either. A long, hard-fought but sloppy game resulted in a frustrating tie. The momentum swung wildly throughout the game and it seemed as though both teams had chances to run away with the game. This was a very back-and-forth game between two very similar squads, but had a wholly unsatisfactory ending.
Can the advanced stats and analytics shed any light on why the game had the outcome it did?
Plays of the game
The winning percentage chart for this game was absolutely wild. Immediately after the game we said that both teams had a chance to take the game over and turn it into a blow-out victory. However, while both teams made big plays, they also made some back-breaking mistakes and the chart bears that out.
Just take in how both teams seemingly snatched defeat from the jaws of victory throughout the game.
Just looking at the chart, it shouldn’t surprise that the two biggest plays in terms of EPA took place at the beginning of each half.
The biggest play of the game was the fumble by Daniel Jones to start the game. That one play was worth -5.0 EPA for the Giants, as well as -10 points in winning percentage (from 44 percent to 34 percent). It set the stage for Washington’s early lead and made the game look as though it was quickly heading towards a bad loss for the Giants.
The second-biggest play of the game (from an EPA perspective) was the sack-fumble by Azeez Ojulari to start the third quarter. That play was worth 4.4 EPA for the Giants and +10 points in their winning percentage (from 44 percent to 54 percent).
However, that touchdown also marked the beginning of (about) 35 minutes of almost complete futility for both offenses.
And as you can tell from the winning percentage graph above, the sudden nature of overtime makes for massive swings in winning percentage. And as such, the list of plays with the biggest impact on winning percentage is dominated by overtime.
For instance, the biggest swing in winning percentage was Washington’s punt with 37 seconds left in overtime. That took the Giants’ winning percentage from 98 percent (on a failed attempt to convert a fourth down) all the way down to 19 percent.
The biggest non-overtime change in winning percentage was Taylor Heinicke’s late-game touchdown to tie the score at 20. That was worth a 14-point swing in winning percentage, dropping the Giants’ odds of winning from 75 percent down to 61 percent.
About the Giants’ offense
I wanted to write about the Giants’ offense after the game, but I realized it was going to take more time and space to properly digest what happened and why. That opens the door for narratives to run wild, but that’s just the nature of the beast.
Over the last 18 hours or so (and really going back to the loss to the Lions), I’ve seen a lot of frustration with Mike Kafka and Brian Daboll with regards to the Giants’ offense. Fans want to know why Kafka and Daboll have suddenly become more conservative and predictable in their offense. And the honest answer is ... Nothing’s really changed. The Giants’ offense is still structurally the same as when the team was winning.
But as B.J. Armstrong once said, winning covers up a multitude of sins.
The Giants coaches have had to go to some pretty great lengths to protect the offense as a whole. We understand that the Giants have a narrow path to victory and can’t afford to beat themselves or give their opponents any extra opportunities. We also understand that the Giants have limitations on their roster that need to be coached around. The Giants’ offense is the end result of coaching to the players’ strengths and weaknesses.
The state of the Giants’ receiving corps, their offensive line (and Washington’s pass rush), and Jones’ strengths as a passer lead to a very quick offense that didn’t look very far down the field.
Kafka primarily called quick-strike passing plays that didn’t involve prolonged progression reads. Jones had the third-fastest time to throw in the NFL on Sunday, averaging just 2.47 seconds. By comparison, his season average coming into this game was 3.01 seconds. They schemed a lot of short, defined reads with schemed separation for the Giants’ receivers.
The Giants receivers all averaged good separation from Washington’s coverage players, in large part due to how the Giants’ plays were drawn up.
In fact, Jones only threw four passes to players who were considered “covered” by NFL NextGenStats.
All told, the speed of the Giants’ passing attack and the play design made life easy for Jones and the receivers while minimizing the time the offensive line needed to block Washington’s formidable front.
Despite giving up four sacks, the Giants did a pretty good job of keeping Washington’s defenders away from Jones.
So, why are people frustrated?
The problem is that most of the Giants’ passing attack was very short ranged, and Jones averaged just 3.1 air yards on completions. That made for a great completion percentage, but those passes didn’t accomplish all that much. Jones averaged just 0.02 EPA per drop back and added just 0.8 expected points through the air all game.
Much of the Giants’ offense is predicated on the run, building the passing game off of that with play-action bootleg rollouts. But without much of a vertical attack to keep the Commanders’ defense honest, they were able to play downhill.
We could see the impact of that on the Giants’ running game, as Saquon Barkley struggled to find much running room anywhere along the line of scrimmage.
Barkley’s 3.5 yards per carry on 18 runs is roughly the same as his 3.55 yards per carry against Dallas and roughly in-line with the 3.0 yards per carry he’s averaged over the previous four games. Defenses have generally done a much better job of maintaining gap discipline against hand-offs, rallying to the ball to prevent Barkley from making a single tackler miss.
Unfortunately for the Giants, this is pretty much the style of offense their roster forces them to play.
The flip side of the offensive struggles from both teams is that this was primarily a defensive struggle. Both the Giants and Commanders field among the best defensive lines in the NFL. Both the Giants duo of Leonard Williams and Dexter Lawrence, and the Commanders’ Jonathan Allen and Da’Ron Payne combine power, explosiveness, and pass rush ability to be dangerous up the middle.
But where the Giants went to some extreme lengths to limit the risk to which they exposed Jones, the Commanders didn’t do much to protect Taylor Heinicke.
Heinicke had a longer time to throw at 2.68 seconds, and he looked further down the field on average.
Washington also suffered a pair of significant injuries along their offensive interior, finishing the game with their second string center and third string right guard.
The Giants also got some great play from Kayvon Thibodeaux and Azeez Ojulari, who were both consistent presences in the Commanders’ backfield. Between the longer-developing offense, and the play of the Giants’ young linemen, Heinicke was under some heavy pressure.
The Commanders’ run game did give the Giants’ trouble. Brian Robinson, in particular, was able to find running room despite how well the Giants’ front played.
There was significant concern before the game about the match-up between the Giants’ injury-depleted secondary and Washington’s receiving weapons. But even so, the Giants didn’t back down from their typical aggressive coverage schemes.
The Giants benefitted from an officiating crew that allowed very physical play downfield. The Giants’ coverage players were able to maintain contact much further down the field than typically allowed by NFL officials, to the point where it looked more like college coverage than NFL. And while there was still a mismatch between Washington’s receivers and the Giants’ corners, they still limited the separation Washington was able to generate.
(note: Curtis Samuel is clearly an error. NextGenStats didn’t record any receptions or carries for him)
The tight coverage helped keep the ball in Heinicke’s hand, giving the Giants’ pass rushers that much more time to get home. Kayvon Thibodeaux had a lightning-fast sack of 2.4 seconds when he was left unblocked at the end of the game, but the Giants’ also benefitted from coverage sacks as well.
Frustration with the Giants’ offense is understandable. But Wink Martindale’s defense played a big role in the Giants turning the game around early on and having opportunities to win throughout the second half.