Welcome to the 39th installment of the “The Wrong Read.” This article series started as one that reflected on recent podcast episodes and extended the ideas discussed there to logical conclusions with broader applications. Since then it’s become a space for me to write about whatever I want, with irregular references to various podcast episodes. Nevertheless, I’ll link to the episode that started my train of thought if applicable.
Last week I started exploring average draft position (ADP). My suggestion in that piece was that ADP does not appear to be consistently beatable, if beating ADP means compiling more accurate rankings year after year. However, I also suggested there may be another way to beat ADP consistently that doesn’t require us to become expert rankers. It may be possible to beat ADP — that is, to win at fantasy football — by exploiting persistent inefficiencies within ADP. In other words, even though we may not be able to consistently be more right than ADP, we can take advantage of ways in which ADP is consistently wrong. The article you’re currently reading attempts to do just that by looking for ways ADP routinely gets rookie running backs wrong.
Finding Inefficiencies in Rookie RB ADPs
In a previous piece I explored the idea that rookie RBs might be the key to winning in best ball with a Zero-RB approach.1 Even rookie RBs who did not end the season with a high point total still did enough to end up in the top five at the position at least one week — and many did it multiple weeks. So rookie RBs were definitely valuable assets for best ball players last year. But one key question remains: Even though rookie RBs were valuable, does that also mean they were values? In other words, did they outperform expectations? And if so, do rookie RBs always outperform expectations?
That last question is actually the most interesting one. Last season, on the whole, rookie RBs did outperform ADP-based expectations by about 26 PPR points on average. But 2017 may have been an outlier. Do rookie RBs outperform ADP-based expectations every year? Turns out, they do not. Since 2000, rookie RBs have on average scored about 1.9 points below expectations. In other words, it’s not as simple as just drafting rookie RBs and then winning your league.2 Rookie RBs, as a group, tend to underperform expectations.
Does It Matter What You Pay?
But maybe the cost to acquire those rookie RBs makes a difference here. That is to say, maybe rookie RBs tend to underperform because they tend to be overdrafted? If so, then we might expect to see a difference between early-round and late-round rookie RBs.
Dividing a 20-round draft into five equal portions lets us distinguish between early, middle, and late parts of the draft. So how do rookie RBs perform based on where, roughly, they are drafted? ADP actually makes a big difference, as the following graph indicates:
Rookie RBs drafted in the middle rounds — Rounds 9-12, in this analysis — are the only group to significantly outperform their ADP-based expectations, while rookie RBs drafted in the early rounds vastly underperform, especially in Rounds 1-4.
What About Recent Seasons?
Of course, Y2K was a long time ago. Things have changed a lot since then. Are we sure we can really draw strong conclusions from seasons that took place almost 20 years ago? Do these trends hold if we limit our data only to more recent seasons? The following graph shows the same data, since 2010.
The shape of the data is identical, but the degree to which middle-round rookie RBs outperform ADP is much larger. Since 2000, rookie RBs drafted in Rounds 9-12 score on average about 9.3 points more than their ADP-based expectations. Since 2010, that number is up to over 25 points of outperformance. Indeed, it may be largely recent results that are skewing the results since 2000. That is to say, if anything, rookie RBs in the middle rounds have become more undervalued in recent years, not less.
Was 2017 an Outlier?
The idea that a portion of a sample could cause significant skew brings up another concern, however. We already saw that 2017 was a great year for rookie RBs, and the fact that Kamara was drafted in precisely this range raises the question about whether his performance in 2017 was so good that it accounts for most of the outperformance we’re seeing in the above graphs. However, middle-round rookie RB outperformance is actually a consistent trend.
There were no rookies drafted in Rounds 9-12 in either 2010 or 2013, meaning 2012 was the only season in the last eight years in which middle-round rookie RBs failed to outperform expectations. In all other seasons in which rookie RBs were drafted in this range, middle-round rookie RBs as a group beat their ADPs. Indeed, 2017 is an extreme case,3 but simply in terms of whether middle-round rookie RBs outperform expectations or not, it fits within a near-decade-long trend of outperformance.
What’s especially interesting is that these findings don’t exactly hold across the RB landscape. Although all RBs see their performance start to beat ADP in the middle rounds, the degree to which rookie RBs outperform in this range is drastic.
While all early-round RBs vastly underperform ADP,4 veteran RBs do not outperform expectations in the middle rounds the way rookie RBs do. The increase in PPR points over ADP-based expectations for veteran RBs is roughly linear, as we would expect, but this is not the case with rookie RBs. In other words, rookie RBs drafted in the middle rounds are consistently undervalued by ADP.
What Does This Mean for 2018?
Rounds 9-12 correspond to Picks 97-144 in 12-team leagues. If we grant some margin of error and expand the range one round in either direction, that means we should be looking for rookie RBs with ADPs between picks 85 and 156. Using RotoViz’s new MFL10 ADP App, we see there are currently two rookie RBs with ADPs in that range:
Kerryon Johnson’s ADP has been steadily rising, and it may soon rise out of the range we’re looking at. Over the past month his ADP is around Pick 91, though looking at only the most recent week shows an ADP around Pick 88. The time to draft Johnson is now, before he gets too expensive. If Johnson can run away with lead back duties in Detroit, you’ll probably be happy even if you draft him much earlier. But consider the difference between drafting Hunt in the third-round last season and drafting him in the ninth. Early Hunt drafters were at a distinct advantage.
Like Johnson, Nyheim Hines’ ADP has also been steadily rising, but it has actually risen into the range we’re looking at.5 Best ball players have started to notice that Hines was drafted into a clear role with immediate value, analogous to Kamara in 2017 (though unfortunately on a far weaker offense). If Andrew Luck can indeed begin throwing soon and make his way back for at least part of the 2018 season, Hines could provide extraordinary value at his current ADP.
- Zero-RB drafters had a better 2017 than most would assume, especially those who drafted Alvin Kamara or Kareem Hunt. (back)
- Well, in one way it might be: veteran RBs underperform ADP by about 8.1 PPR points. Yes, that means that, as a group, all RBs tend to underperform ADP. Even more reason to utilize a Zero-RB strategy? Not quite: it turns out that all positions exhibit about the same underperformance. The best explanation I can currently come up with for what’s going on here is that players who are not being drafted at all in MFL leagues (or in less than five percent of them — the cutoff for inclusion in this dataset) are having an outsized impact, pushing all drafted players’ production lower than expected. (back)
- In fact it’s the single best season for middle-round rookie RBs since 2000, with an average outperformance of more than 91 PPR points. The next best season was 2008, at nearly 85 PPR points. (back)
- which makes sense, given the high expectations for early-round RBs (back)
- Indeed, Johnson and Hines are both two of the biggest recent risers at the RB position. (back)