A Look at Zero RB Results Through 3 Weeks
I am going to take a look at how running backs and wide receivers drafted inside the first five rounds of fantasy football drafts have performed so far. I will look at average performance, performance relative to ADP, positional variance, average games played, and proportion of players who have played in all games so far. I’ll look at these metrics for both PPR and standard scoring.
Performance to Date
Overall, we can see that WRs have had the points edge in PPR, while RBs in standard. But that doesn’t really tell the whole story. For example, what if the early round RBs failed and the RBs drafted in rounds 3-5 picked up the slack? To check this out, I plotted points vs. ADP for each position. First the plot for PPR scoring and PPR ADP.
We can see the situations with Adrian Peterson, Le’Veon Bell, and Jamaal Charles have negatively affected the RB point totals. Obviously this is skewed by the fact that I am taking a look at it in Week 3, and we knew Bell would be suspended for the first three weeks (however, that is reflected in his ADP). As the season progresses, what we want to see from Bell is whether he moves closer to the expected trendline for his ADP. The RBs and WRs so far have put up pretty comparable point totals from about Rounds 3-5. Only two WRs taken inside the first 30 picks have a point total below 28 points, those being the injured Keenan Allen and Sammy Watkins.
In standard scoring, using standard scoring ADP, the picture isn’t quite as clear.
It appears even in standard scoring early RBs are affected by the trio mentioned earlier along with Doug Martin, combined with a shift to earlier ADPs relative to PPR scoring. After ADP 30, it appears RBs are faring better than WRs as a whole. While that’s true, you can also see six red dots above 35 standard points and two red dots below 10 points. Only Jeremy Langford has an ADP after 30 and falls within that 10 to 35 point chasm.
We can measure this variance by fitting a curve to these graphs and then taking the root-mean-square-error and dividing by the average points by position (listed in the table at top). This is called the CV(RMSE). Doing so gives us the following:
The running backs have seen higher variance relative to ADP, even early in the season. That fits the narrative that early round RBs can either win you a league when they hit, or lose you a league when they don’t. The premise behind Zero RB is avoiding this early round variance at RB, and taking the more consistent WR position. In turn, you benefit from RB variance in the later rounds and your roster also becomes stronger as early round RBs succumb to performance or injury issues.
Speaking of injuries, we can also look at games played data to suss out whether RBs or WRs have missed more time and why. Just using average number of games played, we get the following:
On average, RBs have played fewer games. Removing Bell because his absence was not due to injury, the RB numbers change to 2.68 for PPR and 2.65 for standard scoring. In terms of percentage of players,3 21.1 percent of RBs have missed at least one game in PPR and 25 percent in standard, compared to 9.7 and 12.5 percent respectively for WRs. Here’s the full list of injuries to date and expected number of games missed:4
Putting that in terms of a proportion of current expected games missed (expected missed divided by the product of 16 and the number of players drafted in the first five rounds, by position) we get the following, first for PPR then for standard.
In PPR, we can already see that RBs are expected to miss more than 2x the fraction of total games than WRs. That falls in line with Josh Hermsmeyer’s findings on injuries for early round players. In standard, the gap is a bit closer, mostly because there are fewer WRs taken in the first five rounds.
Undoubtedly injuries will continue to pile up at both RB and WR. But we can expect from Josh’s findings, and what we’ve seen so far in 2016, that the injury rate for these high usage RBs will outpace the injury rate for WRs of similar ADP.
While it’s too early based off performance alone to conclude Zero RB is a home run, some signs are there that it’s shaping up to be one of, if not the optimal strategy for PPR in 2016. High variance at the RB position, consistency at the WR position, and injury rates are all a part of the reasons for using Zero RB, and all are present in this early look at the data. However, there have been several RB hits so far in PPR, such as David Johnson and Matt Forte. But we do see that every year. The real test will be what fraction of RBs drafted in the first five rounds are hits by the end of the year, relative to the fraction of WRs.
For standard scoring, the picture isn’t quite as clear. The injury gap is closer, thanks to fewer WRs being drafted early, and as a whole RBs have scored more points than WRs. It will be interesting to keep track of these results for both standard and PPR as the season progresses. I will give an update every third week, so look for the next one after week six.
Subscribe for a constant stream of league-beating articles available only with a Premium Pass.
- For PPR I used ADP data from the Best-Ball App, and setting the cutoff to be one week before MFL10s closed. For standard scoring I used this query. (back)
- I am ignoring the three TEs that fall in this range because they don’t produce a meaningful analysis. (back)
- So again, Bell is removed from this analysis. (back)
- Including games already missed (back)