Where should the Football Team look to fill out its wide-out line up?
I have been hitting the quarterback theme fairly hard this draft season, and I decided it was time to spell it for a few days, so it’s fresh for the final push. While QB is the team’s biggest near-term need in most analysts’ and fans’ estimations, the team also has a fairly thin receiving corps, with no established starting quality X receiver to line up opposite of Terry McLaurin. When Rivera does eventually find his QB of the future, if he’s not already on the roster, he will need a full lineup of receivers to throw to if he hopes to keep defenses honest.
In an article published a few weeks ago, I had a look at how the order of selection of QBs in the draft aligned with player outcomes. In this article I will take a similar approach to wide receivers which, to my casual eye seems to be one of the hardest positions to get right in the draft. It has certainly eluded the Washington Football Team, who have only managed to draft three starting quality WRs in the last 40 years: Art Monk, Jamison Crowder, Terry McLaurin. Well, four if you count Gary Clark’s selection in the 1984 supplemental draft for USFL players.
While I am taking a break in this article from one of my main preoccupations this offseason, I will continue to hammer away at another one. That is the concept, which crops up in every draft discussion thread, that it is essential to pick one of the top few draft prospects at each position to have a reasonable chance of landing an impact player. As that train of thought often goes, the top-rated prospects at each position, or if you like the players projected to go in the top 10 picks, are the elite playmakers who will have the real impact at the pro level. Whereas players selected in the third or fourth round are generally solid contributors or backups.
The outcomes of past drafts demonstrate that that line of thought is based on a misconception about what draft evaluation actually is. As I might have mentioned before, projecting college prospects to the NFL level is not an exact science, and the margins of error are enormous. What the painstaking efforts of NFL scouting departments give their teams is not really a rating of each prospect’s NFL career potential, which can be used to accurately rank players in precise order of career potential. What scouts’ ratings actually give their teams would be better viewed as a set of probabilities for each player.
Any careful examination of past draft outcomes will reveal that the probability of even the most highly-rated prospects being total busts is non-zero. Conversely, WFT fans just need to look to the last draft, when the unheralded Kamren Curl, picked in the seventh round, ended up taking the job of one of the teams most highly-paid star players when given the opportunity after Landon Collins was injured.
That is not to say that teams’ draft ratings are not predictive. Overall, players selected earlier in the draft do better than players selected in the later rounds. What seems to get overlooked is how large the margins of error are around player ratings, at least to the extent that they are reflected in teams’ draft selections. That error margin is why we regularly see players selected in later rounds outplay more highly rated prospects selected near the top of the draft.
As I am about to show, no other position in the NFL draft illustrates these points better than wide receiver.
Methodology and Stats
Question: how often is the first WR picked the best in his draft class?
To answer this, and several related questions, I ranked all the wide receivers selected in each of the last 10 draft classes by total career performance. There is no generally accepted career performance statistic for WRs. Rather, there is a bewildering array of statistics I could have used ranging from single stats, like receiving yards and touchdowns, to advanced analytic metrics like Adjusted Catch Yards. To keep things simple, I used Pro Football Reference’s CarAV.
CarAV is an attempt to place a single number on a player’s career value, equalized across positions. It is the sum of the player’s individual season AV values, weighted to emphasize peak seasons. AV is pretty complicated to explain. Interested readers should refer to this page.
CarAV has some limitations when comparing players at different positions and when comparing players from different draft years who have been in the league for less than 5 years. But it performs much better when ranking players at the same position from the same draft year. To verify that it was providing sensible results, I compared rankings for a few WR draft classes generated with CarAV to those generated with standard WR stats career receiving yards and receiving TDs. CarAV correlated perfectly with receiving yards and close enough with receiving TDs to give me confidence in its rankings.
CarAV was used to rank the WRs in each draft class from highest value, indicating the best receiver in class, to lowest, indicating the worst. I will refer to that rank order as Class Rank throughout the article. The other statistics used in this article are Draft Round and Overall Pick Number (or just Pick Number), both self-explanatory, and Draft Order. Draft Order is the order of WRs picked in a particular draft class. For example, Mohamed Sanu has a Draft Order of 12, because he was the 12th WR picked in the 2012 draft class.
For each draft class, I will list the top five WR, ranked by CarAV, followed by any of the first five WRs selected who did not make the top five, in order of their CarAV rankings.
WR Rankings by Draft Class, 2011 to 2020
We will now have a look at the WR rankings for the last 10 draft classes. Please bear in mind that, because CarAV takes several years to build up, it is not useful for comparing receivers in different draft classes who have been in the league less than about five or six years. It is only used here to rank receivers within the same draft class.
Also, it is not very precise. Players within a few points of one another are essentially tied. But a player with a CarAV of 65 is probably better than one with a CarAV of 55 in the same draft class. I have treated it as being precise to the single-digit level for purposes of making these rankings. Just take that with a grain of salt when looking at players within 2 or 3 points of one another.
Finally, the WR rankings are more certain the further back in time we go, because of the build-up effect I mentioned above, and because it often takes receivers a few years to develop and come up to the speed of the pro game. Because of that, I would expect some movement in the rankings of at least the last three draft classes in the coming years.
With those caveats and disclaimers in mind, here are the WR rankings from the last 10 draft classes:
2011 Draft Class
2012 Draft Class
2013 Draft Class
2014 Draft Class
2015 Draft Class
2016 Draft Class
2017 Draft Class
2018 Draft Class
2019 Draft Class
2020 Draft Class
Now, to answer the question, how often is the first receiver off the board the best in his draft class? Once in the last 10 drafts – D.J. Moore, picked 24th overall in 2018. In fact, the first receiver selected has only turned out to be one of the top 3 receivers in his class four times in the last 10 drafts, while ending up outside the top five in five of the last ten drafts.
That is pretty disappointing, but it might not be that relevant to the WFT, holding the 19th pick in the first round of the coming draft. On average, 1.7 WRs have come off the board before the 19th pick in the last 10 drafts. The highest number was three, which happened three times; while zero or one WRs were picked in the top 18 four times. Also each of the last 10 draft classes has produced at least three WRs who would represent significant upgrades to the WFT’s receiving corps, making the focus on just the best WR in class seem a bit too restrictive.
Therefore, to broaden our sample and make the analysis more relevant to Ron Rivera’s decision making at pick #19, let’s have a look at the first three WRs and the best three WRs picked in each draft class.
How have the first three WRs off the board fared?
The first figure shows how the first three WRs selected in each of the last 10 drafts have ended up ranked in their draft classes. The good news is that nine of the thirty WRs selected first through third in their draft classes have become one of the top three performing WRs in their class, for a hit rate of 30%. On the other hand, 14 out of 30 ended up ranked as the 6th ranked WR in their class or worse, for an over-draft rate of 47%.
That might not be such good news for the teams picking the first few WRs off the board, but if you think about it, it means that other teams must be finding bargains later in the draft.
Where were the best WRs taken within their draft classes?
The previous figure suggests that the first few WRs off the board might not have proven to be the surefire difference makers that the teams picking them thought they were getting. That raises the obvious question, where were the best receivers in their draft classes selected? The next figure shows the order within their draft classes at which the top three receivers in each of the last 10 draft classes were taken:
An amazing 17 out of 30 (57%) of WRs who ranked amongst the three best in their classes were selected after five or more receivers had been taken. To fully appreciate the significance of this finding, it is worthwhile to take a look at their names:
Randall Cobb (Draft Order: 7th), Alshon Jeffery (7th), Keenan Allen (8th), Davante Adams (9th), Tyler Lockett (9th), TY Hilton (13th), Marvin Jones (23rd), Stefon Diggs (19th), Michael Thomas (6th), Tyler Boyd (7th), Tyreek Hill (18th), Juju Smith-Schuster (6th), Cooper Kupp (7th), Chris Godwin (11th), Michael Gallup (9th), DK Metcalf (9th), Chase Claypool (11th).
That is an astonishing amount of talent left on the board after five players have been picked in a position group almost every draft. To put those numbers into perspective, they equate to around 1.7 players most of us would rate as very good to “certified difference makers” per draft.
What is the best draft round for picking wide receivers?
I mentioned above that no more than three WRs have come off the board before the 19th pick in the last 10 drafts. I also found that the majority of WRs ranked in the top three of their draft classes were selected after five or more receivers had come off the board. Putting those two facts together suggests to me that it might not be that critical to select wide receivers early in the draft to have a shot at landing a real difference maker.
To see if that’s correct, let’s have a look at the draft picks where the best three wide receivers were selected in the last 10 drafts:
It is true. While a respectable 11 of the 30 (37%) best three receivers in class in the last 10 drafts were first round picks, seven of those were selected in the second half of the first round. The real surprise here is that the majority (63%) of top-three-in-class WRs were selected in the second round or later. In fact, more of the best three WRs in their draft classes were selected in the second half of the second round or later (19) than in the first round and a half of the draft (11).
If there was a “sweet spot” for WRs, in terms of value per draft capital expended, it would seem to be the second half of the second round through the fourth round, where 13 of the 30 top three receivers in their classes were selected. That is great news for the WFT, looking to add another starting WR in this draft, because they hold four picks in this range.
How have WRs selected in the top 10 fared?
From the previous graph it appears that only four of the WRs selected in the top 10 overall picks have wound up being amongst the best three receivers in their draft classes. That doesn’t sit very well with the concept that the top three or four prosects in each draft class are where you find the difference makers. In total, eleven WRs have been drafted in the top 10 of the last 10 draft classes. What happened with the other seven?
The final figure shows the distribution of Class Ranks for WRs drafted in the top 10 picks of the last 10 drafts. Only four of the 11 ended up as one of the top three receivers in their draft classes (Julio Jones, A.J. Green, Mike Evans, Amari Cooper), which is surely the result that teams picking in the top 10 would have wanted to see. The remaining six (54%) were major disappointments, since they were not as good as five or more other wide receivers in their draft classes. The last figure is particularly alarming, when we consider that the highest number of WRs taken in the top 10 is three (2017), and that only happened once.
The results of this analysis reinforce the view that wide receiver is amongst the most difficult positions to project from college to the NFL, if not the most difficult. My last two articles for Hogs Haven (QB draft order, Sweet Spot for QBs) would have reminded readers about how difficult quarterback is to project. While the hit rate for drafting QBs is notoriously low, at least with quarterbacks, the best players tend to come off the board in the first round; and, as a general rule, quarterbacks taken earlier in the draft tend to be better than those taken later.
Wide receivers take the difficulty of draft projection to another level, with more of the top receivers in their draft classes coming off the board after the first five have been selected than before. That is so crazy that it suggests that NFL teams might actually be looking at the wrong things when they pick WRs toward the top of the draft. Someone who knows more about scouting players than me, which doesn’t narrow down the field that much, should really look into that.
To the extent that the first few WRs selected in each draft class are the most highly rated by NFL teams (we only really know how one team rated each of them), I don’t think the most highly-rated WR prospects are the ones you want to pick. I’d rather take the ones who are still on the board in the second round or later, after the first few have been taken, and save my earlier picks for positions that are easier to project… once I figure out which ones those are.
Thanks to James Dorsett for his usual thorough editorial assistance.