Bruin Sports Analytics
The Sustainability of Lamar Jackson's Production
By: Vishal Sundaram
When tasked with winning football games, every head coach in the NFL talks about establishing an identity, as all the best teams have one. The 49ers are winning off of a multi-dimensional running game and a ferocious defense that gets to the opposing quarterback. The Saints are winning off a dink-and-dunk quick-hitter offense. The Chiefs are winning with speed on the outside and a lethal deep-ball passer. The Patriots have been "doing their job" for the last 2 decades. But out of all the teams in football, there's one team's identity which stands out as the most established, physical, and against-the-grain in all of football: the 14-2 Baltimore Ravens, the best team in the league.
It all starts with their stifling defense which, once again, is a contender for the league's best. While they have a good defensive line, where they really shine is at defensive-back, where they have the league's deepest and most dangerous group. Headlined by the elite ball-hawking Earl Thomas and suffocating Marlon Humphry, the Ravens also have three other feisty cornerbacks in Brandon Carr, Jimmy Smith and Marcus Peters. On the other side of the ball, the Ravens run the rock up the opposition's throat, with a rock-solid offensive-line anchored by Ronnie Stanley and the ageless Marshal Yanda, premier blocking tight-ends in Mark Andrews and Nick Boyle, and one of the league's angriest runners in Mark Ingram. But these assets aren't what makes the Ravens special; they've been this type of team for a long time. The addition of one piece has made their offense go from byzantine to innovative, from stagnant to electric. This one piece is arguably the best mobile quarterback in NFL history: Lamar Jackson.
Back in 2013, the last time the Ravens won the Super Bowl, the NFL was a very different place. For starters, the team they faced in the Super Bowl, the 49ers, had Colin Kaepernick installed as their starting quarterback following an electric run to the big game. Kaepernick was one of the hottest young quarterbacks in the game because of his legs, having gashed the Packers in that very same playoff run for 181 yards and 2 touchdowns on an 11.3 yards per carry clip. This performance set the record for most single-game rushing yards by a quarterback not only in the playoffs, but in any game in NFL history, a record that still stands today. On top of that, the Rookie of the Year that season was none other than current Ravens back-up, Robert Griffin III (RG3), who capped off a historic rookie season with 815 yards on the ground, a number that ranks 7th all time for quarterbacks. Another rookie that season was Russell Wilson, a player whose best rushing season ranks one rung above RG3's, at 849 yards. All of this is to say that the last time the Ravens were dominant, rushing quarterbacks were a trend. Since then, mobile quarterbacks like RG3, Kaepernick, Marcus Mariota, Tim Tebow and Johnny Manziel have flamed out, while remaining players with this skillset like Russell Wilson and the younger Deshaun Watson rely more heavily on their passing ability than their legs.
Because of this trend slowly dying out, Lamar Jackson was a polarizing prospect when he finally hit the NFL draft in 2018. Michael Vick, the consensus greatest mobile quarterback of all-time before Lamar Jackson, endorsed him as a better runner than himself, but due to the decline of this skill's usage, many teams didn't even view Jackson as a quarterback prospect. Hall of Fame General Manager Bill Polian referred to Jackson as too slight for the position, not enough of a thrower, and stated that he should switch to a position such as wide-receiver, which is more reliant on speed. Jackson ended up being selected 32nd overall in the NFL draft, and is currently revitalizing speed for his position. His 1206 rushing yards through 15 games this season is an NFL record, shattering Michael Vick's previous high of 1039, and making him a surefire bet to win the MVP. After seeing this historical production, I wanted to measure Jackson's improvement since his rookie season, see how he stacks up to past quarterbacks in this department, and decide how sustainable his current level of production is.
Last year, through 7 starts in his rookie season, Lamar Jackson had 556 yards on the ground, an average of 79.4 yards and 17 attempts per game. Despite the excellent production, 17 rushing attempts is an unsustainable number for a quarterback. To make this number tangible, look no further than this year's leading rusher Derrick Henry. Henry is built like a truck and led the league in rushing this season. Despite all this, Henry averaged a hair under 19 carries a game (18.93), less than 2 carries per game more than Jackson in his 2018 season. The Ravens knew this had to change too, and hired Kaepernick's former offensive coordinator, Greg Roman, to run the offense. This season, Jackson averaged an even better 80.4 yards per game over a 15 game sample, but averaged under 12 attempts (11.7) per game on the ground, a much more sustainable number. To put Jackson's production in context, Ezekiel Elliott, who had the 4th most rushing yards in the league this season, rushed for 1357 yards in 16 games, which averages out to 84.8 yards per game. He did this in 125 more attempts than Jackson. Obviously, Jackson has the element of surprise on his side, and defenses crowd the box on Zeke, but still, a quarterback rushing for close to the same yardage per game as the highest paid running-back in the league is just absurd.
The graph displayed below shows Lamar Jackson's per-game stats as a rookie compared to other notable running-backs' in 2018. This dot-plot includes every single 2018 pro-bowler at the position, as well as a few other notable players, such as Mark Ingram, Leonard Fournette and Jordan Howard. Despite the smaller sample size, the fact that he ranked 3rd in carries and 4th in yardage per game from the quarterback position over his last 7 games is eye opening.
Jackson's development as a passer has been the key difference that has allowed him to drastically improve his production on a per-carry basis, as it doesn't allow defenses to blitz and key off on his running ability. At the end of last season, many were calling Lamar Jackson more of a running back than a quarterback. NFL.com's QB Index wrote that while Jackson arguably added more value as a runner in half a season than any quarterback in NFL history, his errant throws and league-leading 12 fumbles (in 7 games) kept his value in check, ranking Lamar23rd in the leagueout of 32 starting quarterbacks. At the end of this season, he was theconsensus number 1and an MVP shoo-in. What parts of his passing game improved? According to Pro Football Reference (PFR), his ANY/A (Adjusted Net Yards per pass Attempt) has gone up from 5.99 to 8.19, his completion percentage has gone up from a terrible 58.2% to a much better 66.1%, and his sack percentage has gone down from 8.6% to 5.4%. To rationalize this data, when analyzing PFR's passing index, where a score of 100 is league average and anything higher is better, Jackson has gone from below average values hovering in the 90s range to above average values in the early 100s in every notable passing stat, among them ANY/A, Touchdown Percentage (TD%), Interception Percentage (INT%) and Completion Percentage (CMP%). His touchdown percentage, in particular, was tops in the league, as he had a league-leading 36 passing touchdowns on the season.
These stats show his increased mastery of the passing game, which increases the opportunity to make misdirection plays and establish a dominant running-game for the Ravens. The opposing defense must respect the pass and can no longer crowd up the line-of-scrimmage to stuff the run. The ANY/A in particular is telling, as this statistic has a strong correlation with offensive production. It is essentially a formula which takes passing yards, adds the number of touchdowns times a multiplier, subtracts the yardage lost due to sacks and the number of interceptions times a multiplier, then divides the total by the sum of the passing attempts and number of sacks taken. While this sounds like complicated jargon, it's essentially just calculating the average yards per play with touchdowns and interceptions thrown factored in. Thus, the ANY/A is an effective way to see how well a quarterback is able to lead his offense to points. This and the rest of Jackson's passing statistics help explain why his yards-per-attempt while rushing the football was substantially higher this season, sitting at 6.9 yards-per-attempt this year compared to last season's 4.7 average. It also helps explain how the Ravens as a team set the NFL's all time record for rushing yards in a season, with their 3296 yards on the ground breaking a mark that stood for 41 years prior.
While Lamar Jackson's improvement as a passer has fueled his and his team's improved statistical performance, I wanted to see just how sustainable this improvement truly is. Visually, Jackson's uniqueness is obvious. Lamar is just built different. He has unreal speed, better shiftiness than most running-backs, and touch on his passes. But, we've seen players like Kaepernick and RG3 take the league by storm in the past, just to regress to league average or worse in their next season. In order to gain statistical insight on what level of regression to expect, I recorded the statistics of a group of mobile quarterbacks in their first full season as a starter, and compared this to their statistics in their next full season accrued as a starter, attempting to see whether defenses figured them out. Players like Lamar Jackson and Colin Kaepernick had seasons where they started 7 games before their first full season, so I set the threshold of a full season as a starter to be a Week 1 starter who played at least 14 games. In order for me to consider the quarterback mobile, I wanted them to average at least 3 carries per game in their first season.
In order to conduct this experiment, I gathered the statistics in major passing and rushing metrics for 10 mobile quarterbacks in the 2000s. This group consisted of Colin Kaepernick, Robert Griffin III, Marcus Mariota, Mitchell Trubisky, Russell Wilson, Cam Newton, Michael Vick, Daunte Culpepper, Dak Prescott and Deshaun Watson. While many of these players obviously play differently from Lamar, I believe most of these players are more than capable of running the football. Mitchell Trubisky and Marcus Mariota have shown commendable speed rushing the football, while Vick, Culpepper, Griffin and Kaepernick have had huge games on the ground. Newton, Wilson, Watson and Prescott have all had spurts of MVP play and have shown the ability to run and pass effectively throughout their careers. When collecting data, I once again started off by focusing on the Pro-Football-Reference metrics for ANY/A, Completion Percentage, Touchdown Percentage and Interception Percentage, four of the most important statistics for a quarterback. I found the Pro-Football-Reference metric valuable because it measured a quarterback's statistics in these categories relative to their peers in that given year, providing a standardized metric that adjusts for the changes in the passing game from Vick and Culpepper's eras to the present. Regardless, I found that all the players in the sample, on average, had above average statistics in their first full season as a starter. In their next full season, after an entire offseason of defenses keying in on their tendencies, this population of quarterbacks, on average, was slightly below league-average. Considering that the list contains a handful of MVP level talents, the observation that these quarterbacks went from close to the top 10 players at their positions to barely average is shocking. Obviously, there were some exceptions, but almost all the quarterbacks showed a slight regression.
Next, wanting to also look at raw statistical change, I compared the raw values of each of these quarterback's rushing attempts, yards, ANY/A, TD% and INT% to see the noticeable per-player changes in each statistic. To do this, I found each of these quarterback's statistics in their first full year, and either found a proportional change (ANY/A), percentage change (TD%, INT%), or the raw change (rushing yards and attempts per game) in their statistics the next-season. This thus calculates the average individual performance decrease, regardless of a player's talent, as the player is now only being compared with himself.
The findings proved to be insightful. Firstly, most quarterbacks kept rushing the football the same amount, with around the same number of production. Exceptions included Trubisky and RG3, who both saw their number of rushing attempts decrease by over 1 attempt per game, as well as Culpepper, who saw an uptick in usage and production. The rushing attempts and yards per game of the average player remained almost the same, decreasing by a miniscule 0.16 attempts and 0.95 yards per game.
Moving on to the Adjusted Net Yards per Attempt (ANY/A), the ANY/A value in the following year was 0.84 times the value in the first year. Thus, the regression was 0.16 multiplied by the ANY/A in the first year, with the proportional decrease displayed. Essentially, all that means is that these players were 16% less efficient in using their arm to gain yards and lead their offense.
Finally, the touchdown percentage dipped by 1.1%, meaning 1.1% less of these quarterbacks' throws resulted in touchdowns. Culpepper had the most jarring drop in touchdown percentage, with 3.2% less of his passes ending in paydirt. Only two players had their touchdown percentages go up: Deshaun Watson and Michael Vick. On the flipside, the interception percentage went up, with these quarterbacks throwing picks, on average, on 0.58% more of their throws. Watson and Dak Prescott had the most sizeable increase in picks thrown, with 1.8% more of their passes winding up in the opposition's hands. That adds up to almost an extra 1 out of 50 passes ending up in a turnover. The average quarterback throws over 30 passes per game, so this number is significant over an entire season.
So what does all this mean in-terms of Lamar? If we go by the raw average changes calculated, Lamar's ANY/A will go down to 6.88, his league-leading touchdown rate will go down to 7.9%, and his interception rate will go up to close to 2.1%. The data also indicates that his rushing attempts and yardage per game would remain somewhat similar. But, as we all know, all players are different, and even with these statistics factoring in an expected regression, Lamar Jackson would still be a player capable of wrecking games.
So Lamar Jackson is still going to be a game-changer, and the Raven's defense is still going to be suffocating. Regardless, I would expect defenses to catch on to Jackson to a certain extent, and I expect his statistics next season to hover closer to the numbers listed above than to those he recorded in his 2019 MVP campaign.
When looking for a team to compare this Ravens group to, it's hard not to go back to the last time the Ravens hoisted the Lombardi Trophy and see parallels to the team they beat that fateful day. Both the 2019 Ravens and the 2013 49ers had a dynamic quarterback who took the league by storm, a stifling defense stacked with difference-makers, and most importantly, a Harbaugh at head coach. Regardless of the similarities between Lamar's and Kaepernick's supporting casts, one of these players is the presumptive league MVP, and one of them has never made a Pro-Bowl. Despite this season's disappointing end, I expect Lamar Jackson and the Ravens to keep on flying.