In 2018, the New England Patriots pulled an about-face on a long-standing team-building philosophy. With the selection of running back Sony Michel in the first round, New England made a hard pivot away from years of finding niche backs on the cheap to “do their job.”
2004 was the final season of peak Corey Dillon, at least by the measure of simple counting stats. Dillon finished that regular season with 1,635 rushing yards and 12 rushing touchdowns. The Patriots did not have another 1,000 yard back until 2010, when BenJarvus Green-Ellis barely eclipsed the arbitrary barrier, totaling 1,008 rushing yards.
In the years in between, the likes of Dillon, Green-Ellis, and Laurence Maroney would bare early-down responsibilities, while Bill Belichick would mix in Kevin Faulk and Danny Woodhead as receiving backs. That role-based approach to the RB position continued in the 2010’s, as Stevan Ridley, LeGarrette Blount, Dion Lewis, Mike Gillislee have been deployed as early-down backs (with an honorable mention to Jonas Gray), while Shane Vereen, James White, and Rex Burkhead have been featured in a receiving back capacity. (It should be noted that if any two backs from this list of names could be considered “all-around” in that they have been used as rushers and receivers, it would be Lewis and Burkhead).
Only twice since 2010 has New England featured a 1,000 yard back; Ridley in 2012 and Blount in 2016. It didn’t much matter, as the Patriots offense has ranked in the top 6 in offensive DVOA every year since 2010. The common thread among most of the names above is the low cost to acquire each of them. While New England did send a second-round pick to Cincinnati for Dillon and used a first-round pick on Maroney, the remainder of the running backs required a cheap asset to obtain their services. Between low priced free agency, mid-round picks, and late-round picks, the Patriots didn’t pay a premium for a running back since Maroney’s drafting in 2006.
Twelve years later, all that went out the window when New England chose Michel in the first round of the 2018 draft. Fresh off a Super Bowl loss in which the Patriots defense could not get a single stop to seal the victory, and a year where they ranked 31st in defensive DVOA, Belichick declined to select a player at any defensive position in order to take a RB.
His mysterious arrival onto the roster has proven to be the perfect foreshadowing for Michel’s career to date. His 2018 regular season was nothing special. His playoff run was outstanding. His 2019 regular has been abysmal. No one really understands the extent of his on-again, off again knee issue.
Michel on the field
Michel was limited to 13 games his rookie season but received 209 carries, 14th most in 2018. Michel compiled 931 rushing yards and six rushing touchdowns on those attempts. While ostensibly those numbers look good, they overstate how well the rookie runner fared.
Per FootballOutsiders, Michel ranked 24th in DYAR and 26th in DVOA among running backs in 2018. More specifically, his DYAR on the year was 24, meaning he produced 24 yards above what an average running back would have on his 209 carries, while his DVOA metric was a -2.7%, meaning he performed -2.7% worse than an average running back would have in the same set of circumstances. On the plus side, however, Michel had a 53% rushing success rate, the 11th best mark in the league in 2018.
Still pertaining to 2018, New England’s offensive line ranked 3rd in adjusted line yards. This would suggest that Michel got what was blocked for him in 2018 but nothing more, as well as evidence that his runs were successful in larger part due to the offensive line’s performance.
Michel then took the postseason by storm. In three games culminating in a Super Bowl victory, New England rode Michel. He touted the rock 71 times while racking up 336 yards and 6 touchdowns during the title run. The Patriots found it prudent to give their rookie RB an average of 23.67 carries per game while chasing a championship, a workload no running back in the NFL received in 2018, regardless of how small the sample size might be for an individual player.
Yet here we are, in 2019, and Michel’s production is as grim as can be through three games. So far this season, Michel ranks 34th and 33rd in DYAR and DVOA, respectively, among 39 qualifiers in each category. His success rate has dropped down to 23rd in the league.
In fairness to Michel, the Patriots’ offensive line is significantly worse than last year. In stark contrast from the 3rd best ranking in 2018, New England’s offensive line is 18th in adjusted line yards in 2019. Furthermore, Michel is being contacted earlier. In 2018 the Georgia product averaged 2.4 yards prior to contact, yet that number has dropped almost a full yard to 1.5 this season. This is also more evidence that his success rate is more indicative of the offensive line as opposed to the runner himself.
Still, Michel isn’t producing much on his own. On the year he has 1 missed tackle on 45 attempts. One. He is facing a stacked box (8+ defenders) on an insanely low 2.22% of his runs. Despite the light fronts he rushes against, he is sporting the second-worst efficiency in regards to how many yards he is traveling in order to gain a yard. The latter two numbers are both courtesy of the NFL’s Next Gen stats, which bring us to our final data point.
PFN has developed our own metric in which we attempt to quantify how much a player is personally responsible for his own output. Football is a team game played within a system. With the help of Next Gen stats, PFN can isolate how much an individual player contributed to his own stat line. The metric is called Offensive Share Metric (OSM). If pressed for time, keep this quote in mind during the remainder of this article; “OSM is formulated by looking at a combination of the NFL’s very own Next Gen stats and other algorithms that have been created. Doing this allows us to see how well a quarterback or receiver [or running back] did with what only he could control.”
With that said, Michel is currently sporting a 4.69 OSM grade, the 4th lowest mark in our database out of 47 running backs that have qualified for a single week of grading. The purpose of the metric is not for direct comparison of if one player is better than another, so despite not being scaled for volume, this remains a concerning data point for New England and the former first-round pick. What we’re saying is Michel is personally responsible for only 4.69% of his output.
Moving forward
It’s fair to defend 2019 Michel with “it’s only three games.” However, the postseason run was also “only three games.” The most simplistic of math would put the three outstanding games and three discouraging games as evening out to average, then pointing to how Michel was essentially an “average” running back in the 2018 regular season, per FootballOutsiders.
How the RB performs for the rest of the season or even the rest of his career is anyone’s guess. As of this writing, however, New England’s drafting of Michel is as bizarre as Michel’s on-field performance since.
Jonathan Rosenberg is a writer for PFN covering the AFC East. You can find him @frosted_takes on Twitter.