In 2018, the two most-highly drafted RBs of the most recent drafts put up very different rushing yards per attempt. Saquon Barkley ran for 5.0 yards per attempt. Leonard Fournette only managed 3.3 YPA. Does either number tell us anything about their likely fantasy production in 2019?
There is a lot of research to show that YPA is not a good statistic for evaluating RB ability. For example, the excellent analyst Chase Stuart has written extensively on this topic. That has focused on real-life NFL performance; I wanted to examine the question from a fantasy perspective.
So, to answer my question, I looked at all RB seasons with a minimum of 100 rushing attempts since 1988. I calculated the YPA for all those backs in each year as well as their fantasy points per game (FP/G) in that season and the next, using point-per-reception (PPR) scoring (10 yards = 1 FP, TDs = 6 FP, reception = 1 FP).
Then I summed everything up in this table:
The descriptive categories are based on the ranges of YPAs in the second column. Those ranges were picked to make the top and bottom groups include about 10% of the total carries of all the RBs in the sample. Then I tried to make the middle five categories all about the same share of carries. The “Average” group came out a little bigger, so these seven categories form a flattened-bell curve[1]:
The “FP/G” column is the overall fantasy points per game by each group of running backs in the year when they posted the YPAs shown. The “FP/G Next Year” are the numbers by those same backs in the following season. The final column, “FP/G % Change” is the next year column divided by the baseline year FP/G. So RBs like Fournette who were “Bad,” with YPAs of 3.4 or less, averaged just 9.4 FP/G in that season, only 8.0 FP/G the next, and that next year’s FP/G was just 85% of the previous year’s FP/G.
Meanwhile, Barkley is in the “Excellent” category, and those backs scored more FP/G in both the current year (15.7 FP/G) and more the next season (14.3) than their Bad counterparts. They also retained a greater percentage (91%) of their previous FP/G.[2] Here’s a graph of those percentages:
The black trend line indicates that in general, RBs who had better YPAs retained more of their previous year’s FP/G. But it’s also worth noting that EVERY category scored fewer FP/G in the following year – about 90% on average. So, BOTH Barkley and Fournette look poised to decline in 2019; it’s just that Fournette is likely (based on this over-simplified analysis) to lose an even bigger share of his FP/G.
In fairness to Fournette, we know he was injured in 2018 and only played in 8 games. And his 15.1 FP/G last year was still RB12. In terms of the graph above, as a rookie his YPA was Below Average (3.9) while he managed 17.7 FP/G. His 2018 FP/G was 85% of his rookie performance: close to what the graph would predict even though his per game fantasy numbers were much better than the typical “Below Average” YPA back.
Of course, while on average RBs in every YPA category decline, some do get better:
About a third of all these RBs improve their FP/G in the following season. Most of the YPA categories are close to this average. Only the “Below Average” RBs seem to do much worse than that average and the “Good” RBs out-perform the norm. But those are probably not statistically significant differences. The trend is slightly up across all the categories:
The R-squared of that linear regression trend line is just 0.10, meaning that YPA category only explains about 10% of the improvement or decline in RB FP/G.
As a counter to the Fournette 2017-2018 example, look at his fellow highly-drafted RB from 2017: Christian McCaffrey. McCaffrey only managed 3.7 YPA as a rookie and 14.4 FP/G. If we looked just at his “Below Average” YPA, we would have stayed away from him in 2018 and missed out on his 24.3 FP/G (and 5.0 YPA).
The point of all this is that YPA is only one of many factors that go into predicting future fantasy performance and probably far from the most significant one. It does play a minor role: just not as much as some would argue.
[1] The actual average of all RBs in this sample is 4.15 YPA; the “Average” category used here is skewed a little below this. Also, the average YPA all RBs who had less than 100 carries in this period is 3.9, indicating that overall, coaches gave their best backs more carries than their lesser players.
[2] Here’s how good Barkley’s rookie year was: only 37 times since 1988 have just 28 different RBs averaged 5.0+ YPA on 250+ carries. A warning: only 5 backs did it more than once.
