In part 3 of this series, I’m switching gears and looking at positive regression – or players that unexpectedly underperformed last year when compared to their career output. If you want to read the players I think will negatively regress, or if you want to know how the definition of regression and how to spot it, click the links below to jump to the previous parts of the series:
Marcus Mariota, QB, Tennessee Titans
2017 Stats: 15 games, 281/453, 3232 yards, 13 TDs, 15 INTs, 60 carries, 312 yards, 5 TDs
2017 Finish: 210 points, 14 PPG, QB18
Marcus Mariota and the Titans offense were one of the worst passing offenses in the recent NFL history. They were inefficient and ineffective. But how did the Titans become so bad (they threw for 20 less yards per game and 1 touchdown less per game in 2017 compared to 2016), despite only trying to add to their decent passing offense in the draft via first round pick WR Corey Davis? Well, the answer lies in their quarterback, Marcus Mariota.
A lot of the issues that we saw with the offense last season were primarily due to Mariota’s injuries. While I normally don’t like to attribute poor play solely to injuries, it was clear that the hamstring and ankle injuries he played through last year were affecting his play as a mobile quarterback. He took a huge step back in efficiency to a level we had not seen from Mariota ever before – in college or the pros. As long as Mariota stays healthy, we should be able to expect him to bounce back to the numbers he was putting up before. So in what ways was Mariota so inefficient last year? Well, in about nearly every category.
Half The Usual
One of the biggest differences, both in fantasy points and in deviation from the mean, was Mariota’s inability to throw touchdowns and his ability to throw interceptions last year. In college, Mariota threw touchdowns on 9% of his passes. In his first two years in the NFL, Mariota threw touchdowns on 5.1% and 5.8% respectively. Then last year, Mariota threw for a touchdown rate of only 2.9% – half the rate as the previous year and a completely unprecedented number for him. This led him to throwing only 13 touchdowns all of last year and was a contributing factor to Tennessee being the only team in the NFL last year to throw more interceptions than touchdowns. We would have expected him to throw nearly twice the amount of touchdowns that he did, based on his career stats and ten years of NFL statistics, for a total of 23 touchdowns (compared to the 13 he did throw). This would have resulted in a 40 point swing, good enough for a 2.25 points per game difference. Now add in the fact that until last year Mariota had never thrown more than ten interceptions in one year since high school, and we can expect some positive regression for Mariota on this end as well. Based on attempts and his career interception rate in the NFL, we would have expected him to throw around 10 interceptions last year, but he instead threw one and a half times the expected total when he threw a total of 15 interceptions last season.
Mariota took a step back in his yards per attempt as well – likely due to his inability to scramble with injuries, he relied on making shorter throws in order to avoid getting sacked. This resulted in 7.1 yards per attempt, down from the 7.6 career average he had before last year. And while 0.5 yards per attempt doesn’t seem like all that much, when there are 453 attempts, that adds up quickly to nearly 250 passing yards over the course of the season, or a slightly bigger difference than Kirk Cousins’ passing numbers a year ago to Jared Goff’s.
With these regressions, we would have expected Mariota to finish as the QB11 last season instead of QB18 where he ended up. This would have been in line with his 2016 season when he was the QB12 before missing the last game of the season and looked ready to make the next step. While Mariota is never going to be an elite quarterback in fantasy football, he also provides exceptional value (currently being picked as the QB17) as a late quarterback pick that we can expect to step back into the efficiency that he had shown prior to last year and prove he can again be a low end QB1 for the course of the season like over the last few years.
Lamar Miller, RB, Houston Texans
2017 Stats: 16 games, 238 carries, 888 yards, 3 TDs, 36 receptions, 327 yards, 3 TDs
2017 Finish: 158 points, 9.8 PPG, RB14
Disappointment or Misinterpreted?
Lamar Miller has never been a high-end running back, but he’s always been one that has been dependable week-to-week. In 2016, Lamar Miller scored less than 6 points in a standard league only twice while scoring over 11 points only four times throughout the season. The same held true in 2017 when he scored less than 5 points only twice and more than 11 points only three times. despite the lack of a high-ceiling, Miller has been able to use this consistency to finish as the RB14 in both 2016 and 2017. as a result of the high expectations placed on him coming into the 2016 season after a breakout year in Miami, Miller disappointed with his lack of long runs and high touchdown numbers. He became more of a floor play then the high ceiling pick that many thought he could be. His play over the last 2 years could be seen as disappointing due to misplaced and misguided expectations which has led Miller to become the RB22 in this year’s season draft.
Touchdown Rate Swap
Despite finishing as the RB 14 last year, Miller was rather inefficient compared to his statistics from the previous three years. We saw his touchdown rate go down as well as his yards per carry, and he has increased potential given the situation with the running backs and quarterback in Houston.
Last season, Miller scored only 3 touchdowns on the ground, his lowest since 2013. In 2014, 2015, and 2016, Miller scored 8, 8, and 5 touchdowns on the ground. his touchdown rate for those three seasons was 3.1%, while his touchdown rate from last year switched those numbers as he finished with a 1.3% rate on the ground. We would have expected him last year to score seven touchdowns in the course this season instead of the three that did score, which would have been more in line with the rest of his career.
Add in the fact that not only was Miller underperforming in scoring, but he also was underperforming in how many yards he was getting per carry. In the three years previous to last year, Miller was averaging 4.48 yards per carry, or a little bit above average. but last year, Miller had an outlier season when looking at his career when he averaged only 3.7 yards per carry. Considering he had never had a season below for in either college or the pros, it seems evident that Miller is due for a bounce back in yards per carry. And I firmly believe that the drop in efficiency last year is not due to age, as Miller is only 27, considered one of the last Peak years for a running back, and the normal RB drop in efficiency is at age 30. What was also missing from Miller’s season last year was long runs. Throughout his entire time in the NFL, Miller hadn’t had a season where he was the starter without a long run of at least 45 yards. Last year, his long was only 21 and it was his only carry of 20-plus yards. That resulted in 0.4% of his carries going for 20-plus, and 0% going for 40-plus. Compare that to his career as a starter, which is 3.2% of his carries going for 20-plus and 0.6% going for 40-plus, and you can see why his numbers were low last year.
Taking these stats into account, if Miller had merely performed at his career average, he would have finished not as the RB14, but as the RB9, ahead of other backs like Ezekiel Elliott, Jordan Howard, and Devonta Freeman.
But statistical regression is not the only thing playing into Miller’s potential as a low RB1 this upcoming season. Miller’s ADP is currently in the middle of the fifth round because of fear that D’onta Foreman will be taking some of the carries from Miller. But Foreman is recovering from an Achilles injury, one of the worst possible injuries for a running back to have. Time and time again we’ve seen running backs have their athleticism, speed, and agility sapped by an Achilles injury, and we have little reason to believe that Foreman is going to be substantially different. Foreman also may start the season on the PUP list, taking him out for at least the first six weeks of the season. This gives Miller very little competition in the backfield, and would allow him to receive good work on both the ground and in the air.
But also add to the fact that Watson will be returning after his injury last year, and Miller seems set up for success. When Watson was the starting quarterback for Houston last year, Miller averaged 88 yards from scrimmage per game and scored a touchdown for every 33 touches he received. But when Watson went out with an injury, Miller took a major step back in fantasy. Miller only averaged 66 yards from scrimmage, or 75% of what he was getting with Watson, and he only scored once every 62 touches on the ground and in the air – nearly half the rate as when he was with Watson.
With Watson, Miller was scoring once every 33 touches and averaging 88 yards from scrimmage per game. Without Watson, Miller was scoring once every 62 touches and averaging only 66 yards from scrimmage per game.
All of this, along with Miller’s proven history as a low-end RB1 or high-end RB2, leads me to believe that Miller will bounce back and significantly outperform his RB22 ADP. He’s the perfect type of back to pair with an early RB and WRs in the first few rounds.
Jay Ajayi, RB, Philadelphia Eagles
2017 Stats: 14 games, 208 carries, 873 yards, 1 TD, 24 receptions, 158 yards, 1 TD
2017 Finish: 111 points, 7.9 PPG, RB33
A Unique Situation
Jay Ajayi had a unique situation last year – despite being the lead back in Miami, he was dealt after week 7 to the Philadelphia Eagles, who then proceeded to use the ‘16 Pro Bowler very sparingly until week 12 when he got 15 carries in the game. From week 8 to week 12, Ajayi never carried the ball more than 9 times. But from week 12 to the end of the playoffs, Ajayi had double digit carries in every game except the Super Bowl (where he finished with 9 carries).
Ajayi’s time in Miami throughout the 2017 season was less than pretty. He failed to score on the ground once throughout those first seven games (indicative of Miami’s poor offense in general, as they became the first team in five years to not score a rushing touchdown until week 8 or later). Ajayi came more into his own after he settled down in Philly and became a more important part of their offensive game plan, which leads me to believe that regression is due for Ajayi.
Ajayi struggled mightily in two main categories that are crucial for fantasy football – rushing touchdowns and yardage. As mentioned before, the fact that it took Ajayi until his first week in Philadelphia to score a rushing touchdown is not a good result for a RB that was being taken in the first round at tenth overall, according to FantasyFootballCalculator.com. But we should expect regression given Ajayi’s touchdown history in college and the NFL. In college, Ajayi scored on 13.5% of his carries, including two years of 18+ touchdowns in a season, whereas that number dropped to a more league-average 2.9% in his career after entering the NFL. But despite being just average at scoring touchdowns, Ajayi put up a measly 0.5% touchdown rate last season – or scoring just once every 200 carries. He was the only back last year to score only once and have at least 200 carries, and only four other running backs scored just once and had at least 100 carries (with three of those having 120 carries or less). Players with more rushing touchdowns than Ajayi include WR Cordarrelle Patterson, QB Drew Brees, and RB Peyton Barber. Part of this was due to the poor Miami offensive line that killed all of his output for the first half of the season, because we would have expected him to score roughly six touchdowns throughout the season, rather than the one that he did score.
Failure to Run
On top of not reaching the end zone often, Ajayi also struggled to move the ball well on the ground compared to his previous years on college and the NFL. At Boise State, Ajayi ran for over 5.6 yards per carry – an unattainable statistic for the NFL. But even during his time in Miami before last year, Ajayi was averaging a 4.72 yards per carry rate. Last year, Ajayi rushed for only 4.2 yards per carry – a much bigger drop in efficiency compared to his career. This resulted in an over one hundred yard difference. We saw some of the positive regression in Philadelphia for Ajayi as he had a 4.06+ yards per carry rate in seven of the ten games, and a 5+ yards per carry rate in five of the ten games. This seems to indicate much of his issue with his season long stats were a result of the poor offensive play in Miami and that a full season in Philadelphia should bode better than what we saw out of him just a year ago.
If we take that regression and apply it to a full 16 game season last year, Ajayi would have finished as the RB16 rather than the RB33 that we saw him finish at last year. If we take his last seven weeks in Philadelphia (when they started using him more and got him caught up with the playbook) along with his career touchdown rate and extrapolate that for 16 games, he would have finished with 220 carries for 970 yards and 6 touchdowns on the ground along with 30 catches for 362 yards and 1 touchdown in the air. Those numbers would have put him in the low end RB1 tier as the RB10 on the season – a far cry from the RB33 finish. This RB10 area is where we can expect Ajayi’s realistic ceiling to be, and his floor will be the RB20 range. Considering Ajayi is being taken as the RB19 in drafts, he’s being selected near his floor and presents a solid value with upside as your RB2 or RB3.
So with that, those are three players that I think we need to watch for positive regression this year in fantasy football. Make sure you watch for part four to find out four more players due for positive regression, coming later this week!
Think I’m wrong about a player, have other players you think are due for regression this year, or just want to ask a question about this upcoming fantasy football season? Leave a comment below or email me at email@example.com.