Questioning Market Share: The Hunter-Patterson Comparison
Let me start by saying that I see the great value of using “market share” as a tool for analyzing WRs. (Eventually I will get to the comparison of Justin Hunter and Cordarrelle Patterson. I promise.)At RotoViz, when we look at a college WR and his professional prospects we primarily consider not his raw stats but the percentage he accrued of his team’s receiving yards and touchdowns in his final year; that is, we examine the extent to which he dominated his team’s total receiving production. Unequivocally I believe this method of analyzing WRs deserves a central position in the total evaluation process.
Having said that, I want to note some drawbacks of looking only at market share—which RotoViz’s Jon Moore calls Productivity Grade and Shawn Siegele has termed Dominator Rating—when we consider collegiate statistics. First of all, one could question what market share even means. What is the significance of a player’s domination within his collegiate offense? After all, as fantasy players, we don’t care about what percentage a WR accrues of his NFL team’s receiving yards and touchdowns—we care only that he accrues lots of them. In fact, we don’t even care if a WR gets all of his yards and touchdowns through the air. If he manages to rush for a couple hundred yards and a couple more TDs, that’s cool too. And in some leagues even punt and kick return yards and/or TDs count. In the end, raw statistics are all we care about for fantasy.
As a result, one would be quite justified questioning the use of Dominator Rating—at least until one actually saw the evidence showing the correlation between collegiate Dominator Rating and NFL success. Then again, one could ask this question: If we replaced collegiate Dominator Rating with collegiate raw stats, would a similar correlation to NFL success still exist, and, if so, would it be weaker or stronger? [I anticipate that someone at RotoViz will answer this question, and the answer will likely be, “Yes, weaker, so that’s why we look at Dominator Rating and not raw stats.”] And one could even ask this question: Do market share and raw statistics, used together, provide a correlation stronger than that of either one used separately? [Again, I think someone at RotoViz will provide an answer, which will probably be “Yes.” But I could be wrong.]
In the most straightforward manner possible, the following logic may apply: If we want to know whether a WR can manage 1000 scrimmage yards and 10 total TDs in an NFL season, shouldn’t we just see if he managed to accomplish this feat as an undergrad?—regardless of whether he captured 20% or 60% of his college team’s receiving production? If two college WRs existed identical in age, height, weight, build, Speed Score, Agility Score, vertical and broad jump, collegiate statistics, collegiate conference, and route-running capability, I would prefer the guy with the higher Dominator Rating. (Then again, wouldn’t a low Dominator Rating mean that the WR, even with big stats, has room to grow within an offense?) But if two fairly similar WRs were both selected in the first round, and one had 1200 receiving yards and 12 TDs in his final year, with a Dominator Rating of 0.30, and the other had 900 receiving yards and 9 TDs, with a DR of 0.4—which would you prefer? Perhaps the guy who proved in college (regardless of offensive system) that he was capable of putting up a 1000-10 season.
To quote the logic used in this RotoViz article, “If he can do it, he can do it.” By using market share in place of raw statistics—or, more precisely, by not using raw statistics alongside market share—we risk potentially muddying the waters of a process that is already muddy enough. I’m not saying that we should abolish the use of Dominator Rating. I think, though, that we should look at raw stats too.
The overriding benefit of using market share in place of raw statistics is that it adjusts for factors such as quarterback play and college scheme, since (we can all agree) that not all receiving yards are created equal. But, by that same measure, not all Dominator Ratings are the same. For instance, a high Dominator Rating might not mean that a WR is good but that he was the best of limited options on his college team. And what is the worth of a high Dominator Rating that occurs in a subpar passing game?—it could mean that the QB or scheme sucked and that the WR overcame these difficulties, but it could also mean that the WR sucked (albeit less than the WRs who surrounded him) and that he proved unable to rise above the limitations surrounding him.
In fact, a high Dominator Rating in a subpar passing offense could even mean that the WR himself was a limiting factor for the offense—since, if he were better, his high market share would translate into better raw stats and a larger general “market” for his team. From this perspective, wouldn’t the worst kind of prospect be a guy who dominated his offense and still couldn’t produce desirable raw stats? I think so, and as a result we need to consider raw stats alongside market share in order to weed out such a WR—because we have no guarantee that NFL GMs won’t select this kind of player in the first round. And, what’s more, if a high Dominator Rating that fails to accrue significant statistics doesn’t intrigue me—if I’m always going to look back to the raw stats to place a WR’s Dominator Rating in context—then maybe I should just acknowledge the privileged position of raw statistics in the first place, because we really can’t afford to examine market share without them.
Further, I am unsure if we should even want to adjust for quarterback play and college scheme. One of the purposes of market share is to make outsized receiving production seem the same as average production—but it’s not. Even if a WR’s production was partially the result of playing in a spread offense with a “QB who made the WR,” I like having that information, which is implied in the raw stats but not market share. I like knowing that a guy performed well with a good QB in a system designed to get the ball to WRs, because that is very well what he may be expected to do in the NFL. Yes, perhaps he benefitted from his school’s system, but he also contributed to that system’s accomplishments. Why would I want to deemphasize a WR’s success when comparing him to other WRs?
People who prefer market share to raw stats will sometimes say this about a WR: “He had inflated stats because he played with a good QB,” or “He had inflated stats because he played in a spread system.” Very possible, but at least now we know he can play effectively with a good QB and in a particular system—and I don’t see why I should hold that against him or seek to diminish the extent to which he dominated WRs on other teams, even if he didn’t drastically overshadow those on his own team—and that information is contained in the raw stats. Furthermore, if a guy has college stats that look inflated, maybe all that means is that he had an awesome year or career and is a strong prospect. Just because raw stats are outsized does not mean they are inflated, and we don’t need to look at market share in order to deflate them. Rather, we should look at a player’s strength of schedule, Speed Score, Agility Score, efficiency in turning catches into yards and TDs, ability to run the route tree in his workouts, etc. If a big-stat WR underperforms in these areas, then his stats are inflated. If not, then his statistical domination is legit, regardless of his Dominator Rating, and that comparative domination is made most evident via raw stats, not market share.
Hunter v. Patterson
I feel as if I could go on and on, but I’ll transition to the ultimate point—and here is where I move to Hunter and Patterson. Market share ignores all non-receiving yards and TDs, and I think the accumulation of those statistics is important, not only because it occurred within the context of competitive games but also because it suggests particular skills. Perhaps a coaching staff noticed that a player excelled once the ball was in his hands and so decided to get him touches in the most straightforward way possible: handoffs and kick and punt returns. Further, perhaps the guy’s non-receiving touches precluded the coaching staff from getting the ball to him more through the air, since he was already touching the ball in other facets of the game. All of this information is useful and is accessible through raw stats but not market share.
And, more to the point, rushing WRs seem to do well in the NFL. Right now, I say this anecdotally, but I intend to examine thoroughly at a later point the NFL careers of WRs who routinely run the ball in college versus those who don’t. As it is, I can say with confidence that, since the 2000 college season, if a WR drafted in Rounds 1-4 had more than 75 rushing yards in a collegiate season, then he had a strong (and I expect above-average) chance of achieving at least one top-30 season in the NFL. WRs to hit the collegiate 75-yard seasonal mark include Koren Robinson, Roy Williams, Santana Moss, Bernard Berrian, Anquan Boldin, James Jones, Mario Manningham, Jeremy Maclin, Darrius Heyward-Bey, DeSean Jackson, Eddie Royal, Austin Collie, Kenny Britt, Eric Decker, Mike Wallace, Ryan Broyles, Mohamed Sanu, Randall Cobb, T.Y. Hilton, LaVon Brazill, Justin Blackmon, Julio Jones, Denarius Moore, Rod Streater, Tandon Doss, Marvin Jones, Jeremy Kerley, Chris Givens, Steve Breaston, and Antonio Brown. (And, yes, dammit, Ted Ginn Jr.)
Oh, I almost forgot—Wes Welker and Percy Harvin.
When emailing with other writers at RotoViz, I asked this question: “Why should Patterson be punished in a MS system just because not all of his yards are receiving yards?” Shawn Siegele sharply responded with his own rhetorical question: “Because unless whoever drafts Patterson intends to use a 6’2”, 216 pound guy almost exclusively behind the line of scrimmage like Harvin then his non-receiving totals don’t affect his projection as an NFL receiver?” I understand Shawn’s point, but the example of Wes Welker demonstrates the importance of considering a WR’s collegiate rushing contribution. Welker hardly ever rushes in the NFL or receives the ball behind the line of scrimmage, but with the Patriots he did enable the short passing game to function as an appendage of the running game, and his collegiate rushing prowess hinted at what he could do if given the ball quickly and in space. Rushing WRs do well in the NFL not because they run the ball, although more teams are increasingly looking for hybrid players who can be used in a variety of packages. Rather, rushing WRs do well because the skills needed to run the ball are transferrable to the NFL and can be exploited within the passing game. In general, these WRs are so dynamic that, however they get the ball, they do well once they get it.
I’m sure I’m not the only person to have noticed the relative NFL outperformance of rushing WRs. In this piece on T.Y. Hilton, I talk about the drafting proclivities of Bruce Arians, the “Wide Receiver Whisperer,” particularly his penchant for selecting effective short and slender WRs later in the draft. Particularly, I said this: “From 2007 to 2012, spanning his time as the Offensive Coordinator of the Steelers and then the Colts, Arians oversaw the drafting of the following players in Rounds 3-6: Mike Wallace (2009, 3rd round), Emmanual Sanders (2010, 3rd), Antonio Brown (2010, 6th), T.Y. Hilton (2012, 3rd), and LaVon Brazill (2012, 6th). For a guy who swings only at short and slender WRs after the top rounds, Arians hits a lot of homeruns.”
How does Arians hit these homeruns? For the most part, he swings only at the short and slender WRs who also have at least one collegiate season with 75 rushing yards. Wallace, Brown, Hilton, and Brazill all had at least one such season, and the only player who didn’t (Sanders) is perhaps the least exciting of the five. If rushing WRs are good enough for Arians, they should be good enough for us too, and when he selects Markus Wheaton in the 3rd round or Chad Bumphis (or, God forbid, Marquise Goodwin) in the 6th round of the 2013 Draft, we should pay attention, because, yes, all those guys produced at least one season of 75 rushing yards in college, as did Tavon Austin, Keenan Allen, Robert Woods, and, finally, Cordarrelle Patterson.
That last guy on the list has taken an awful lot of heat from the staff at RotoViz—and almost all of it is warranted. His list of comparable players is not ideal. He very well could be the next Darrius Heyward-Bey. Jon Moore quite ably makes the case against Patterson in two pieces, here and here. And Shawn Siegele even goes out of his way to get in on the action in a piece on not Patterson but Christine Michael! That’s some loud non-love.
While I think that Patterson is a problematic prospect, I want to suggest, via a comparison with his teammate Justin Hunter, that Patterson is not as horrible as he may seem. I acknowledge that his combined market share of receiving yards and TDs is horribly low, but so is Hunter’s, and he doesn’t receive nearly the vitriol that Patterson does even though he too has a chance of being drafted in the first round.
Here is a table comparing Patterson and Hunter’s 2012 receiving performances, highlighting how they both performed as 21-year-olds in their last FBS seasons.
If one looked only at this table, Hunter would appear to be the superior prospect. He certainly had the better receiving season.
But what if one looks at the exact same season using this table?
All of a sudden we see that Patterson, with two fewer touches, acquired three more scrimmage yards and—including his punt and kick returns—one more total touchdown. Particularly impressive is that Patterson outgained Hunter with a third of his touches coming in the form of carries, generally low-yardage touches. All of a sudden, Hunter does not seem much (if at all better) than Patterson, who—the raw stats indicate—was more dynamic of a performer.
And what about this table?
|Name||Year||Age||Ht||Wt||40 Time||Drop Rate||Speed Score||Vert Jump||Broad Jump||FBS Years||Injury|
Hunter is taller and more explosive in his jumps, but he is also supermodel skinny, runs slower, drops many more passes, and has already torn an ACL. I grant that he was returning from a major injury in 2012, but he needed more time to do what Patterson did in only his first year—record 1000 scrimmage yards—and he never did what Patterson did in 2012: Lead the team in total (all-purpose) touchdowns.
And here’s the kicker—Hunter scored SEVEN of his nine TDs against schools from non-BCS conferences. Against SEC teams, Hunter underwhelmed at turning touches into touchdowns. And how did Patterson do in his 8 SEC games? Let’s see.
Despite recording the exact same number of scrimmage yards, Patterson drastically outscored Hunter. In fact, through a variety of touches, he managed to score 0.875 TDs per game against SEC opponents—and that’s not including the legitimate TDs he scored that were negated by penalties, which Evan Silva calls attention to in this piece. In fact, that mark of 0.875 TDs is the best for all SEC WRs except for the unheralded and RotoViz favorite Chad Bumphis, who scored 1 TD per game as his offense’s focal point. That Patterson, as his team’s second option—relegated to accepting Hunter’s leftovers and supplementing his production with rushes and kick and punt returns—produced almost as many TDs as Bumphis is incredible. And when you consider his size, speed, and his first-round (potentially top-ten) stauts . . . yeah, he may turn into DHB, but Patterson’s still worth a draft pick if he reaches a certain spot in rookie drafts. Hunter?—maybe (not).
Here is how Patterson compares to Santana Moss, Peter Warrick, and Ted Ginn Jr. in their last college seasons. All these players had at least one season of 75 rushing yards in college and were drafted in the top-half of the first round.
|Ted Ginn Jr.||2006||21||5-11||178||4.38||13||29.19%||62||798||9||11|
Patterson has the worst market share of the group, but his raw stats are comparable, as his athleticism. Many scouts think of Patterson as raw, but Evan Silva has noted his ability to run routes, and he has better routes and hands than Ginn ever did. Of the three comparable players, Moss and Warrick seem most like Patterson, and, with athletic ability between the two, Patterson could very well have a professional career situated along the continuum they creat. In twelve NFL seasons, Moss has produced five top-30 years, two of which have been top-10 finishes—and one of those was a top-3 affair. Warrick, meanwhile, produced in six NFL seasons only one top-30 finish, in his fourth year, although his rookie year did produce a positional ranking of 31. Those two fates are the most realistic ends of Patteron’s NFL spectrum—with the sad potential that, yes, he could also “develop” into Ginn.
Patterson is likely to be drafted in the top 20 picks, and (as my upcoming series on first-round WRs will reveal) 71.23% of the WRs taken that high from 1978 to 2007 produced at least one top-30 season. If Patterson is taken with a top-ten pick, his historical odds (based solely on draft position) improve to 78.38%. Patterson is not A.J. Green or Julio Jones, but even DHB has produced a top-30 year. If Patterson is selected in the top-half of the first round by a non-dysfunctional franchise, he may be very draftable in dynasty leagues as a low-risk pick. He may not provide many top-30 (or any top-10) seasons, but he’ll probably provide one to three in the next five years—and, if you find yourself in a rookie draft with no first-round-worthy RBs on the board, such as Knile Davis, and if Geno Smith and a couple of WRs have already been selected, then you could probably do a lot worse than taking Cordarelle Patterson with your (presumably low) first-round pick, because particular first-round WRs have some value, and I think Patterson may be one of those guys.
Justin Hunter? Even if he is a first-round pick, he’s probably not one of those guys.
What do those guys look like and what exactly is their value? I explore these and many more questions in the upcoming series, “Everything You Wanted to Know About Sexy First-Round WRs But Were Too Afraid To Ask” . . . or some other lame title.