This time around we look at a model I’ve put together for attempting to predict successful fantasy NFL running backs based on their college resumes. It takes into account multiple different variables, applies a weighted multiplier to each variable then adds them together to give the final score. To more easily compare across draft classes I’ve also calculated the percentile rank of each player’s score compared to the whole database.
The model currently sits with an R correlation coefficient of 0.6 for Seasons 1-3 fantasy points per game. In saying that, for RBs draft capital is the single most important variable, as it has an R of 0.57. Similar to WRs, using thresholds has been more helpful in applying this model practically.
Since 2011-2018 (removed 2019 as players were only rookies), a threshold score of >16.9 brings us 28 players. Of those:
>1 RB1 season = 54%
>1 RB2 season = 71%
>1 RB3 season = 79%
⁃ Draft capital ie the single most important variable
⁃ Dominator rating (% of team’s total rushing yards and rushing TDs + receiving yards and receiving TDs)
⁃ Weight adjusted speed score
⁃ Final season market share of receptions
⁃ Yards per team pass attempt
⁃ A 3 year average of their total yards
⁃ Power 5 conference
Running back is a bit different to WR in the fact that there’s generally only one RB on the field at a time, so once they’re on the field they’re not competing with anyone else for running back touches, whereas WRs are constantly competing with other pass catchers for targets. For this reason RBs are more likely to be one off week winners if they get an opportunity ie a handcuff.Embed from Getty Images
On to the 2020 class. We have 7 that end up >16.90, which is a huge number considering in all other years there has only been an average of 3.5/year. Unsurprisingly, considering draft capital is so important we have the big 5, plus two extra faces: AJ Dillon, and Lynn Bowden (who I’ll specifically touch on later because he’s a bit of an outlier in the model).
Jonathan Taylor deservedly sits clear at the top and does his best attempt to break the model, only having two players since 2010 better his post draft score – the dynasty RB #1 and #2 – Christian McCaffrey and Saquon Barkley. Not sure if you’ve heard of any of these players? Pretty exceptional company! And don’t hate on Gio, his first 3 seasons he finished as RB16, RB18, RB21.
One of Jonathan Taylor’s knocks was his lack of pass catching, however amongst RBs drafted in the first 3 rounds, his final year market share of his team’s receptions was fairly middle of the pack, and interestingly was more than Swift and Dobbins.
No need to break down all of the rest of the big 5 (Dobbins BAL, Akers LAR, Edwards-Helaire KC, and Swift DET) as they’ve generally been the consensus top 5 RBs in the class, and they all exceed our threshold, so happy for situation to be the tie breaker between them. After Damien Williams’ decision to opt out of the 2020 season it’s pretty clear Clyde Edwards-Helaire’s situation is the best, and he deserves very strong consideration to be a first round pick in redraft leagues. In dynasty leagues I still think there’s a discussion to be had between Jonathan Taylor and Clyde Edwards-Helaire at the rookie 1.01 spot, with the edge being given to CEH as given his clear path to opportunity in the 2020 season he should see a full workload, and therefore a likely value increase, earlier than Jonathan Taylor. Additionally, a lot of CEH’s appeal is his pass catching ability, and I’m not sure if you’ve met Patrick Mahomes, but he’s pretty good at throwing passes.Embed from Getty Images
The Polar opposites of AJ Dillon and Lynn Bowden
On to the more interesting players who clear our threshold of 16.9. Let’s dive deep into AJ Dillon and Lynn Bowden.
AJ Dillon was drafted in the second round which is significant capital for a team to put into a running back. Aaron Jones has been historically efficient every season he’s played, in yards per touch his first two seasons, and in TD efficiency this past season. He is however in his contract year, and was “only” a 5th round draft pick.
AJ Dillon is a size speed specimen in the mould of Derrick Henry.
4.53s 40yard dash✅
117 weight adjusted speed score (97th percentile) ✅
41” Vert ✅
131” Broad Jump✅
A three year starter at Boston College he did nothing but run the ball and score TDs. His lack of receiving production, 0 receptions in his Freshman year despite 300 carries, is a concern as there’s not many 2 down workhorses left in the league (Henry, Howard are the main two). In saying this though, despite his measly box score receiving stats, his final season reception market share of 8.07% was only marginally below Dobbins share of 8.46% despite their actual receptions being 13 and 23 respectively.
In addition to having the second highest model score in the class (I’m still ignoring Bowden, you’ll see why soon), I believe there are three main variables working in Dillon’s favour:
1. Aaron Jones is out of contract after this season
2. He was drafted in the 2nd round
3. Green Bay drafted a blocking TE/FB(Josiah Deguara) in the mould of Kyle Juszczyk in the 3rd round
Alright, we can now talk about Mr 98th percentile Lynn Bowden. Bowden is an intriguing name and his journey at Kentucky was a mixed one. He started as a WR but actually ended up playing as a wildcat QB in his final season due to injuries to their QB room, and lead the SEC in rushing yards. So technically never played as a running back.
Given catching the ball is an important part of becoming a fantasy RB superstar, it’s an important part of my model, therefore it can be gamed by RBs who catch a lot of passes, or in this case, a player who literally played as a WR for two out of his three college seasons. So he’s kind of an RB, kind of a WR, more an offensive weapon, and really it’s a model for running backs as opposed to offensive weapons, so an impressive score but one we should take with a massive amount of salt (unless of course he develops into an absolute stud in which case I’ll claim all credit), or disregard his model score as to be honest, it’s probably not applicable.
The Raiders drafted Bowden as a RB/WR and have even referred to him as a “Joker” in that they envision him lining up all over the formation. Given Bowden’s WR background, if there is an injury to Josh Jacobs, Bowden could truly provide a 3 down skillset and rake in the receptions.Embed from Getty Images
Time for the players who missed the cutoff.
Moss, Benjamin and Vaughn make up the first tier of this group.
Moss provides some interesting competition for Singletary as Gore leaves 179 touches up for grabs. Murky backfields are often opportunities for value and murky is exactly what the Bills backfield is as Moss joins Singletary who was just drafted in the 3rd round in 2019.
Eno Benjamin is incredibly interesting to me. He graded very highly pre draft (97th percentile) but was hurt by the lack of draft capital (7th round). He also owns the highest final season reception market share (18.03%) out of the real (ie not Bowden) drafted RBs, over 4% more than the next highest who was Ke’Shawn Vaughn (13.86%). Eno is in a high scoring offense where the RB1 is a highly coveted fantasy position. Kenyan Drake is on the transition tag so is not signed past 2020. If Eno is able to beat out Chase Edmonds for backup duties then I think he has the skills as evidenced by his college production make a splash if Drake was to go down with injury, or go on the COVID reserve list which is a very real possibility.
Ke’Shawn Vaughn has been a huge riser being drafted by the Bucs and has good enough college production to back it up. His main knock being that he’s an older prospect, already being 23.
We then drop down to sub 60th percentile. I wouldn’t be counting on these players on as anything more than potential role players or bye week fill ins. Sustained, meaningful, season long production is unlikely.
I do however want to address Antonio Gibson, as he has garnered plenty of buzz. Similar to Lynn Bowden, Gibson was never really used as a pure running back. Across 4 years of college he played as a hybrid player, skewing towards being more of a receiver. In his senior season he was employed fairly evenly between a receiver and runner – 38 catches and 33 rushes. His score is buoyed mainly by his 3rd round draft capital, his reception share, and his elite speed score. In college he was very efficient (11 yards per carry in a limited 33 carries in his senior season) although he simply didn’t command a high workload, presumably because his coaches didn’t think he was worthy of it. I’m a Derrius Guice truther to the end so if you asked me I’d say Gibson had an elite running back he needs to overcome in order to carve out a significant workload in Washington, however if you ask a lot of others they may say the backfield is up for the taking.
Thanks for reading and I hope you got something out of this article. Moral of the story with rookie running backs generally is, fade running backs drafted from round 4 onwards, target anyone drafted higher than that.
Find my Rookie WR article here.