Deep dive into the Giants’ rushing attack

It doesn’t take a rocket scientist to conclude that the 2019 New York Giants’ offense had an ineffective rushing attack. Due to a strong finish, where Saquon Barkley rushed for 112, 189, and 92 yards in the final three games, Big Blue ranked 19th in rushing yards per game. However, much of the 2019 rushing attack was difficult to witness, and Warren Sharp’s 2020 Football Preview substantiated what we all knew as Giants’ fans.

The preview is extensive and has a ton of valuable information. The Giants’ inability to establish the run was evident on first downs where young signal-caller Daniel Jones was frequently forced into third and long due to Pat Shurmur’s predictable inside zone heavy scheme.

Many would surmise that Barkley was repeatedly facing 8+ men in the box on first downs in the first three-quarters of the game, but that wasn’t the case. League average for 8+ men in the box, on first down in the first three quarters, happened only 25 percent of the time and Barkley faced that same numbered front only 26% of the time - obviously not a substantial difference.

Before we go over the end result of the league average and Barkley’s rushing attempts into different box types, I want to explain EPA. EPA is Expected Points Added. It’s a football statistic that measures the value of an individual play based on down, distance, field position, situation, and the end result of the play. Expected points are calculated through the context of the play and then the beginning of the play and end result are contrasted next to each other. EPA can and will dip into the negatives but the higher the points the better the play.

With that said, the NFL average for first down runs into a loaded box had an EPA of - 0.08 per attempt with a 41 percent success rate of the run, and the running backs averaged 3.9 yards per carry (YPC). This isn’t a surprise; it’s difficult to run into loaded boxes when defenses expect the run, but Barkley’s fell far short of the league average.

Barkley averaged 2.7 YPC with a 27 percent success rate, and a -0.11 EPA per attempt. The Giants are considerably worse, but it doesn’t get much better when the box lightens up to seven men; league average is 4.7 YPC with a 47 percent success rate and an EPA of -0.02 per attempt. Barkley was at 3.3 YPC with a 44 percent success rate and an EPA of -0.15 per attempt. Against six-man boxes is a different story though.

league average against six-man boxes was 4.8 YPC, 47 percent success rate, and an EPA of 0.02 per attempt. However, Barkley was at 5.9 YPC, 50 percent success rate, and an EPA of 0.13 per attempt. New York was able to string much more successful plays against lighter boxes relative to the rest of the league in similar situations. So, how would the Giants go about getting lighter fronts on first down?

It’s a true rarity to have a running back as skilled as Barkley face light boxes in running situations. Why would the defense do that? For starters, the Giants had to utilize and run out of 11 personnel (3 wide receivers, 1 RB, 1 TE). In situations where the Giants weren’t in 11 personnel, on first down in the first three quarters, they faced 7+ men in the box 98 percent of the time, mostly in 12, 21, and 22 personnel. In those situations, Barkley averaged 2.5 YPC with a 35 percent success rate, and a -0.22 EPA per attempt.

There were 41 rushers with at least 25 attempts in the same situation out of non-11 personnel packages; Barkley had 43 attempts in these situations throughout the 2019 season and he ranks as such relative to his peers:

• 41st (out of 41) in YPC: 2.5
• 39th in success rate: 35 percent
• 37th in EPA/att: -0.22

The Giants couldn’t do much in terms of running the football on first down if they weren’t in 11 personnel, and that’s with one of the most dynamic running backs in the league. This excerpt from Warren Sharp should bring back unfortunate memories for Giants fans about the 2019 offensive strategy:

“On first down in the first three quarters, when defenses flooded coverage and left just six or fewer men in the box, the Giants played right into it, passing the ball 68% of the time. These passes averaged 7.5 YPA, but recorded -0.09 EPA/att. This can be compared to Barkley’s runs against light boxes which we just discussed, where he averaged 0.13EPA/att. That’s a 0.22 EPA/play swing.

“While the Giants didn’t make defenses pay for playing coverage, they didn’t make defenses pay for loading the box, either. When defenses had 8+ defenders in the box, the Giants ran the ball on 61% of these first downs, playing right into the defense’s hands. As mentioned, Barkley was terrible running on first down against loaded boxes, much worse than the NFL average (2.7 YPC, 27 percent success, and -0.11 EPA/att) while passes gained 8.4 YPA with 52 percent success and 0.0 EPA/att.”

Another very interesting piece of information from Sharp’s in-depth analysis is how ineffective the left side of the Giants’ offensive line was in terms of running the football. Here’s the YPC per running direction for the 2019 Giants.

• 2.8 YPC – Left Edge
• 4.3 YPC –LT
• 3.6 YPC – LG
• 3.6 YPC – C
• 5.2 YPC – RG
• 5.0 YPC – RT
• 5.7 YPC – Right Edge

2018’s is even more pronounced with the left side doing poorly and the right side doing much better. Obviously this was with Pat Shurmur and Hal Hunter (offensive line coach); now the Giants will have Jason Garrett and Marc Colombo who can attempt to stabilize the line and establish a viable rushing attack on early downs.

There’s a lot of speculation and prognostication about Garrett’s effect on the offense and what he may do with the 2020 New York Giants. It’s expected that Garrett may employ a more run-based offense; through the first half of 2019, Shurmur threw the ball 55 percent of the time on early downs, which was above league average. Garrett has consistently been more run-heavy over the years and that’s the expectation going into 2020.

A vital way to protect a young quarterback is to provide him with a stable offensive line and a rushing attack. The Giants spent the second overall pick in 2018 on a talented back and they just invested a lot into their offensive line, but the center position is still a concern. Once you get an effective rushing attack, then the Giants can develop a stronger play-action game and use it to keep defenses honest.

With a solid play-action game, on early downs in 2019, the Giants were significantly better. With play-action working well on early downs, the Giants had a 59 percent success rate with 9.0 YPA and an EPA of 0.20 per attempt. Without play action, while passing obviously, the Giants had a success rate of only 43% for an average of 6.0 YPA and an EPA of -0.14.

It’s not overly surprising that play-action passes work better than normal passes, although the Giants didn’t use play-action as much as they probably should, but the best situation for the Giants was play-action shot-gun passes, something that Jason Garrett has done with Dak Prescott. The YPA, on all downs, for play-action under center was 10.9 which was 4 yards better than play action from under center (on all downs). The EPA/att was 0.48 (on 60 attempts) with a 67 percent success rate. The EPA on play-action from under center was -0.11 on 55 attempts with a success rate of 50 percent.

Football is a very complementary game and the Giants were able to have success on play-action shot-gun passes despite having a middling rushing attack. Barkley is also a fantastic running back out of the shot-gun and he did it almost exclusively in college at Penn State.

Garrett should know the importance of play-action from 2019 experience. When Dallas used play-action more than the league average last season, they went 8-2; when they failed to do so, they went 0-6. Game script has to do with play-action utilization to an extent but those splits are still fascinating. The Cowboys had so much success early in the season when Dak Prescott’s name was being floated as a potential MVP, and there’s a reason for that success.

The play-action rate for Dallas and the aDot (average depth of target) on play-action passes was 11.5 YPA with the league average being 8.2. With that rate, Dallas was using play-action passes at a clip of 40 percent with a league average of 23 percent. For whatever reason, from weeks 4-6, Dallas reverted to their 2018 play-action usage, which was just under league average with a lesser aDot. Dallas ended up losing all three of those games. Was the reason they lost because of their avoidance of the play-action passing game and the longer depth of target? - No, but it certainly didn’t help contribute to their wins.

I’m excited that Garrett is able to solely focus on the offensive side of the football and doesn’t have to deal with everything that a head coach must focus on. He can help this rushing attack improve while employing more play-action passes and hopefully utilizing Saqyon Barkley more in the passing game, something he didn’t exactly do well with Ezekiel Elliot (the Cowboys had a below-average EPA and success rate when doing so).

Garrett will have a more difficult time establishing a rushing attack without the likes of Zack Martin and the rest of the powerful Dallas line, but that doesn’t mean Colombo can’t groom this young Giants’ line and hopefully help build the unit exceed their 2019 form. It’s a young team, a young offense, and a new start for the Giants. Now it’s time to see if they can fix a rushing attack that hasn’t been great for far too long.