One of the topics of discussion during 2019 fantasy drafts had to do with Joe Mixon’s ADP. If you remember those early days of September just before the start of the season, Mixon’s ADP was 20.6 overall, or RB11. That wasn’t a great deviation from what Mixon had done in 2018 finishing the season as the RB10. The difference is that he finished there while playing only 14 games, so his rank would have been higher on a full 16-game season.
Some folks wondered why Mixon was getting off the board a little late compared to other players with less upside. One reason often given was Cincinnati’s struggling offensive line. In fact, the Bengals OL ranked 24th (2017) and 22nd (2018) in Adjusted Line Yards per Football Outsiders. They ranked 26th in 2019. That’s three years in a row in the bottom-third of the league. Meanwhile, Mixon has finished the past three seasons as RB34, RB10, and RB13. For someone playing behind such a bad OL, those finishes don’t look that bad.
Anyway, all of this Mixon discussion (and also his current ultra-expensive ADP of 8.2 and RB6 in early FFPC leagues) got me thinking about the true impact of offensive lines in fantasy football. Should fantasy owners take OLs into consideration when drafting players? Have OLs defined season-long results and final running back rankings over the years? Let’s explore!
How Do Offensive Lines Perform Year-to-Year?
The first thing I wanted to know was how variable offensive lines are on a year-to-year basis. Obviously, every positional unit changes from one season to the next, so some variance is to be expected. Players get traded, cut, signed, and retire. And even if the personnel remain the same from one year to the next, there will always be changes in performance level.
Using data from Football Outsiders I have plotted every team’s OL in Adjusted Line Yards (a weighted average of how many yards per carry a running back gains; the higher the number, the better the OL) in Year N (starting in 2000) and in Year N+1.
The R-Squared is just 0.11, which isn’t zero but isn’t significant enough to consider it a serious indicator of anything. Simply put, there are big swings in OL performance from one season to the next, let alone multiple seasons.
That is the first factor that makes me think relying on offensive line performance when drafting rushers might be overthinking things a little.
Is There Any Relation Between OL-Talent And RBs ADP?
The next step was to see if the rank of offensive lines in a year truly affects RB ADP. So I basically looked at OL ranks from Year N-1 and plotted them against RB ADPs from Year N (no players from 2000 are included as I don’t have 1999 OL data, but RBs from 2019 are included as we indeed know 2018 OL data).
I guess I have bad news for those reasoning Mixon’s ADP last season resulted from Cincinnati’s bad OL. Turns out the relation between offensive line strength in one year and RB ADP the next is basically zero — the actual R-Squared is a dismissable 0.007.
To What Extent Does OL-Talent Affect RB Scoring?
We know that offensive lines are volatile from one year to the next. We also have discovered that offensive line play doesn’t impact RB ADP. But what about in-season performance? This is what we really care about, after all. Do offensive lines impact fantasy scoring?
Those are all the RB1/RB2/RB3 (green/yellow/red) seasons since 2000, with each positioned in terms of the total PPR on the season and how each player’s offensive line ranked in Adj. Line Yds. There is, indeed, some relation between both variables. Sadly, the R-Squared sits at a paltry 0.08 making the relationship effectively nonexistent.
But the variables used might be skewing the data as PPR for running backs are also factoring in the passing game, and OL rankings only comprise 32 different spots that might not truly reflect the difference in actual Adj. Line Yds between times. Let’s refine the results.
Quite a mess, that is. The relationship hasn’t improved much. In fact, the R-Squared is now at 0.09 — still rather insignificant.
Even while I’m accounting for just the top-36 players in each season, you can see how there are a bunch of dots sitting on or below the zero-line for ruPPR. That’s because of the presence of some pure pass-catching RBs in the bunch, so let’s adjust the data again. Instead of total ruPPR, I calculated the ruPPR per carry for each of the players in the data set.
Finally, some pattern appeared! Adjusting the variables at work I think I’ve reached the strongest possible relationship between offensive line talent and running back fantasy production on the ground. Putting everybody on a per-carry basis, the R-Squared is now a much better 0.20 — not incredible, but high enough to give it at least a second thought when debating who to pick between two close players.
Historical Data and Early 2020 Outlook
With the takeaway that offensive line performance is not that impactful in how RBs perform in fantasy leagues, it makes sense to take a look at how things have gone during the past few years, splitting the data by season. I have plotted the last five years (from 2015 to 2019, both included) of data next, including the final RB-rank and OL Adj. Line Yds.
As you can see, the low relationship between OL-talent and final running back-rank shows. There are dots all across the board showcasing absolutely different profiles of RB-seasons:
- While both Todd Gurley and Saquon Barkley finished RB1 in back-to-back seasons, Gurley was “helped” by the third-best OL in the league while Barkley had to endure the fourth-worst without much impact in his final fantasy tally.
- From 2015 to 2019, the RB1 played under OLs ranked 15th, 7th, 3rd, 29th, and 27th.
- In the same span, the last of RB3 (top-36 running back) each season played under OLs ranked 30th, 6th, 15th, 12th, and 10th. Again, lines all across the board.
- The R-Squared values for each of the five seasons are, from 2015 to 2019: 0.0, 0.17, 0.19, 0.02, and 0.04.
As expected, looking at ruPPR per attempt improves the results a bit in terms of the relationship between both variables, but not to a great extent. The R-Squared values are, in order: 0.25, 0.19, 0.30, 0.21, and 0.29. OL performance explains 30% of the fantasy rushing results in the best of cases, which is not high enough to consider it critical in assessing any player’s potential upside.
Finally, for those of you getting into 2020 drafts already, this is how ADP in early FFPC drafts compares to 2019 OL performances:
That’s right: virtually no relation at all (0.002), and if there is one, it happens to come out as negative!
Joe Mixon has raised his ADP for the fourth consecutive season while playing behind an offensive line that gets worse by the day. Luckily for his owners, this shouldn’t affect his performance too much.