Catch rate is not always what it appears to be
One of the major position battles in camp this year will be at WR. Terry McLaurin and Jahan Dotson don’t appear to have any challengers for the two starting WR positions. After those two, the depth chart heading into camp is likely to contain more questions than answers to a new coaching staff making a fresh start.
The casual fan is probably in the same boat, trying to make sense of seven returning players, one draft pick and four free agent additions. One thing that people will look at to try to get a handle on the camp hopefuls’ chances of earning a roster spot is their production in previous seasons.
A lot of attention gets paid to total production stats and athletic testing metrics when evaluating WRs. A skill that I feel tends to be underrated in WRs is the ability to catch the football. A receiver can have sub 4.3 speed and be great at separating from defenders, but none of that matters if he can’t be relied on to catch the ball when he’s targeted.
Catching ability is usually measured by Catch Rate, which is simply the percentage of targets that a WR converts to receptions. In recent seasons, I have developed opinions about some of the Commanders’ WRs’ ball skills, based on raw Catch Rates. However, it was recently brought to my attention how strongly dependent catch rate is on the depth of the route tree that a receiver is assigned to run. To check whether my perceptions were accurate, I set about to correct Catch Rates for Average Depth of Target (ADOT).
The results revealed that my perceptions of a few of the returning vets had been inaccurate. I decided to extend the analysis to all of the WRs in Washington’s camp. The ADOT-corrected catch rate data might help to explain what attracted the Commanders to a few of the new additions who don’t immediately stand out on paper.
Photo by Ric Tapia/Getty Images
2024 Commanders’ WR Raw Catch Rates
As a starting point, the following table ranks the Commanders’ current WRs by raw Catch Rate. To make the data as representative as possible, I attempted to use stats from each receiver’s most recent peak season.
Since there seems to be a bit of a cloud over the Commanders’ 2023 offense, I used 2022 data where possible for returning players. For NFL vets who haven’t seen a lot of action lately, I used the most recent season with 20 or more targets. In Dax Milne’s case, I had to resort to using 2021 data, when he had the most targets. When players had 0 or minimal receiving targets in the NFL, I used data from their most recent season in another league.
NFL Veterans
The raw Catch Rate (Rec%) data separate Washington’s veteran WRs into three tiers. Jamison Crowder is the runaway leader, with a Catch Rate of 80%. The next tier comprises Terry McLaurin, Dax Milne and 2024 FA addition Olamide Zaccheaus, clustered together with catch rates in the mid-60% range. They are trailed by Jahan Dotson and Dyami Brown, with catch rates in the 50’s.
These data seem to suggest that Dotson and Brown have sub-standard ball skills. However, you may notice that these two also have the longest Average Depth of Target (ADOT) on the team. Furthermore, the Catch Rate leader, Jamison Crowder, has an ADOT nearly 1/3 that of Dyami Brown. That raises the question of how much influence ADOT had on the WRs’ reception rates.
NFL Novices
FA acquisition Davion Davis has been in the NFL since 2019, but only has 3 targets and 1 NFL reception to his name. The bulk of his playing experience at the pro level was in the 2023 USFL season. None of the other WRs on the Commanders’ roster has an NFL receiving target. That obviously includes rookie draft pick Luke McCaffrey and UDFA Marcus Rosemy-Jacksaint. These players’ Catch Rate data were taken from their most recent seasons in another league.
Among the WRs with little to no NFL experience, Kaz Allen stands out with the highest catch rate. Like Crowder, though, he has the lowest ADOT among the comparable players. At UCLA, he was primarily used on short dump offs, and excelled at gaining yards after the catch. The next highest catch rate belongs to Davion Davis, who appears to have been used on short and intermediate routes by the USFL Birmingham Stallions. The other WRs have roughly similar ADOTs but widely varying catch rates.
Like the NFL vets, there are hints that the Catch Rates might be influenced by ADOT, but it is unclear by how much.
Effect of ADOT on Catch Rate
In order to separate a receiver’s catching ability from the type of routes they are asked to run, we need to correct Catch Rates for differences in ADOT. To do that, I used a technique called linear regression, which may be familiar to readers from their school days. The good news is, that you don’t need to know the mechanics, and there won’t be a quiz. All you need to know is that linear regression allows us to plot the central tendency (a.k.a local average) of a data set that varies as a function of another variable. This is commonly referred to as plotting the trend line.
In this case, we are plotting the trend line of Catch Rate as a function of ADOT. Doing so allowed me to calculate the Catch Rate Over Expectation (CROE) for each of the Commander’s WRs. The CROE quantifies how much better or worse the WR’s catch rate is than would be expected by their ADOT. It provides an indication of how they performed relative to the average NFL receiver running a similar selection of routes.
The following graph of NFL receiving data from 2022 illustrates the basic approach.
The graph plots Catch Rate vs ADOT for every NFL receiver with 20 or more targets. There is a clear trend for greater ADOT to be associated with lower catch rate. While a catch rate of 50% would be terrible for a WR with an ADOT around 5 yards, it wouldn’t be bad for a WR with an ADOT of 18 yds.
There is also a fair amount of variation in catch rate among WRs at each ADOT, resulting in a broad scatter of points around the trend line.
The values of ADOT ranged from -0.7 yds, for Carolina’s Laviska Shenault Jr., to 25.8 yds, for Washington’s Dyami Brown. The average ADOT across all WRs was 10.5 yds and the median value was 10.6 yds.
The trend line, also known as the regression line, indicates the central tendency of the distribution of Catch Rates as a function of ADOT. The regression is calculated by partitioning the variance in the dependent variable (Catch Rate) to ensure that it is evenly distributed on either side of the trend line. It shows the predicted value of Catch Rate at any ADOT value, based on the statistical properties of the sampled distribution. In the grand scheme of things, ADOT is only moderately predictive of Catch Rate. In the 2022 NFL sample, ADOT explains 37.1% of the variance in Catch Rate (R2 = 0.371). That might not seem like a strong relationship to a physical scientist, but in a complex system like NFL receiving, where outcomes are influenced by a multitude of variables, it is pretty compelling.
For our purposes, we can think of the regression line as illustrating the expected Catch Rate at each WR’s ADOT. Therefore, the vertical distance between each WR’s data point and the regression line provides an indication of how much their Catch Rate differs from expectation.
To illustrate how ADOT can help to explain differences in Catch Rate between WRs, let’s have a look at two of Washington’s WRs from the past three years: Curtis Samuel and Dyami Brown. Over the past two seasons, Samuel led the Commanders in Catch Rate, while Brown has had the lowest Catch Rates on the team.
I had thought that Brown’s 35.7% Catch Rate in 2022 was particularly abysmal. However, his 25.8 yd ADOT was the highest in the NFL, and over twice the league average. His Catch Rate was only slightly below the regression line, indicating that it was pretty close to expectation for a WR with his ADOT.
Samuel’s Catch Rate/ADOT stats from 2023 were nearly identical to those shown above. In both seasons after recovering from injury, his Catch Rate was right on expectations for a WR with an ADOT in the range from 6.6 to 6.7 yds. The difference in Catch Rates between the two receivers can be attributed nearly entirely to the routes they were asked to run, rather than differences in ball skills.
Adjusting Catch Rates for ADOT
Catch Rate Over Expectation
To quantify how good Washington’s receivers were at catching the football, I calculated Catch Rate Over Expectation (CROE). This is simply the vertical distance between each WR’s data point and the regression line of Catch Rate on ADOT in the chosen season, in units of Catch Rate (reception percentage). When the WR’s data point sits above the regression line, CROE is positive, indicating that their Catch Rate is better than expected for their ADOT. When the data point sits below the regression line, CROE is negative, indicating that their Catch Rate is worse than expected.
The CROE for each WR was calculated from the Catch Rate vs ADOT data in their recent peak season as listed in the first table.
Catch Rate Grades
To quantify how well or poorly the Commanders’ WRs stack up against others with similar ADOTs, I also used the Catch Rate data to grade them on a bell curve.
Each WR’s Catch Rate was converted to a z-score, relative to Catch Rates of WRs with similar ADOTs. The z-score indicates how much higher or lower the Commander WR’s Catch Rate is than the local average in units of standard deviation. For example, a z-score of 1 indicates that a WR’s catch rate was 1 standard deviation greater than that of the average WR with a similar ADOT. The local average Catch Rate was computed from WRs with ADOTs ranging from +/- 0.5 to 2 yards, using the smallest window possible to get a sample of at least 12 WRs. Damiere Byrd and Kaz Allen could not be graded, because there were not enough WRs with ADOTs within +/- 2 yds.
Z-scores were converted to Grades according to the following scale:
A z >1.5 exceptional
B z = 0.5 to 1.5 exceeds expectation
C z = -0.5 to 0.5 meets expectation
D z = -1.5 to -0.5 below expectation
E z < -1.5 well below expectation
NFL Veterans
The CROE data and Grades separate the veteran WRs on the Commanders’ roster separate into three tiers.
The A tier comprises Jamison Crowder and Damiere Byrd, who had catch rates well above expectation for their ADOT in their recent peak seasons. However, their data are based on small target samples and should be taken with a grain of salt. Crowder had a high Catch Rate in 2023, even for such a low ADOT. That represented a massive swing from 2022, when he had a Catch Rate of 46.2% (ADOT 9.3 yds) and one of the lowest CROEs in the NFL.
Even with the small sample caveat, Damiere Byrd’s Catch Rate seems impressive for a WR with an ADOT over 20 yds. His production stat line for his last healthy season in 2022 does not really stand out (13 rec/124 tgts, 268 yds, 2 TD). But perhaps his ball skills made him appeal to the Commanders as a potential deep threat to stretch the field in Kliff Kingsbury’s spread offense. He could not be graded, because there were not enough WRs with ADOTs within 2 yds of his.
The B tier comprises Terry McLaurin and Olamide Zaccheaus, each of whom had above average CROEs and above average ADOTs.
The C tier of the veteran WRs comprised Jahan Dotson, Dyami Brown and Dax Milne, whose Catch Rates ranged from right around expectation (Dotson) to moderately below expectation (Brown, Milne) for their ADOTs.
None of Washington’s veteran WRs had Catch Rates well below expectation for their ADOTs.
NFL Novices
The WRs with little or no prior NFL experience also sorted themselves out into three tiers based on the CROE data from their most recent seasons in other leagues.
The A tier comprises Davion Davis and Marcus Rosemy-Jacksaint, who both excelled at catching the ball in 2023. Davion Davis had a breakout season with the Birmingham Stallions in the USFL, after having bounced around the NFL since being signed by the Vikings as a UDFA in 2019. In addition to standing out with an exceptional CROE, Davis was fifth in the USFL regular season in receiving yards (575). Among USFL WRs with a minimum of 30 targets, he was third in yards per route run (2.12) and fourth in yards after catch per reception (7.3 yds). I haven’t ranked him among the other NFL novice WRs, because he played in a pro league.
UDFA signee Marcus Rosemy-Jacksaint had the highest CROE of any of Washington’s WRs, while playing for the Georgia Bulldogs in 2023. His ball skills might be his saving grace because his draft profile and his athletic testing scores (RAS 1.68, 4.81 sec 40 time, 1.66 sec 10 yd split, 4.45 sec shuttle, 7.35 sec 3-cone) raise questions about his ability to gain separation against NFL defenders. Nevertheless, Rosemy-Jacksaint posted a respectable stat line of 29 receptions, 472 yds, and 4 TDs, with 2.45 yards per route run and 5.1 yards after the catch per reception against SEC competition. His best hope is that his superior ball skills and willingness as a blocker will be enough to overcome his athletic limitations and allow him to claw his way onto the bottom of the Commanders’ roster.
None of the neophyte WRs graded in the B range. The C tier comprises 2023 UDFA holdovers Kaz Allen and Mitchell Tinsley, with Catch Rates around 2% over expectation. Kaz Allen had the distinction of having the lowest ADOT among FBS WRs in 2022 (min. 50 targets), and there were not enough WRs within 2 yds ADOT to grade him against. But I think it’s fair to group him by proximity to Mitchell Tinsley, who is a solid C+.
The D+ tier included Luke McCaffrey and Brycen Tremayne, whose Catch Rates were more than a little below expectations for their ADOTs in their final college seasons. Before anyone panics about the Commanders’ third round draft pick, it is worth pointing out that 2023 was only Luke McCaffrey’s second season at WR, and his Catch Rate was down from 65.9% (ADOT 11.4 yds, 88 targets) in 2022. Hopefully he can rebound in 2024 with better coaching and QB play at the NFL level.
Photo by Bryan Lynn/Icon Sportswire via Getty Images
Conclusion
Catch Rate is fairly strongly influenced by the depth of routes that a WR is assigned to run. As a result, Catch Rate is not very meaningful on its own. In order to appreciate whether a WR’s Catch Rate is good or bad, it is necessary to adjust for ADOT.
Adjusting for ADOT reveals misconceptions that have arisen around two of Washington’s veteran WRs. Jahan Dotson and Dyami Brown have had Catch Rates below 60% throughout their early careers in Washington. While those numbers may appear low when viewed out of context, both WRs’ Catch Rates are right around league average for their ADOTs. Brown suffers, in particular, as a result of being assigned to run one of the deepest route trees in the league.
The CROE data also highlight three of Washington’s newcomers as being particularly good at catching the ball.
Damiere Byrd
5-9, 175 lbs, 4.27 sec 40, RAS 9.18
Byrd has bounced around the league since signing with the Panthers as a UDFA in 2016. In his most recent stint with the Falcons, he was used sparingly as a receiver, primarily on deep routes. He might provide competition in that role to Dyami Brown. Early in his career in Carolina, he also demonstrated explosiveness as a kick and punt returner, which could provide another avenue to earn a roster spot.
Davion Davis
5-11, 195 lbs, 4.64 sec 40, RAS 3.15
Davis has been in the NFL since signing as a UDFA with the Vikings in 2019. His career NFL stat line is 3 targets and 1 reception for 17 yards. He had a breakout season in 2023 with the Birmingham Stallions and was one of the top five receivers in the USFL. Davis is another player who could earn a roster spot on special teams, if he is unable to replicate his success as a WR in the NFL. At Sam Houston State, he was used as a punt kick returner. In 2017, he averaged 21.6 yds on 13 punt returns with 2 TDs, and 25.3 yds/return on kickoffs.
Marcus Rosemy-Jacksaint
6-1, 195 lbs, 4.81 sec 40, RAS 1.68
Rosemy-Jacksaint is not going to win with speed. But he was able to use his length and exceptional ball skills to carve out a role as the second leading WR for the Georgia Bulldogs in 2023. In fact, he tallied more receptions and receiving yards than Chargers’ second round pick Ladd McConkey. If his ball skills are enough to overcome his athletic limitations, he could eke out a roster spot as a possession receiver and short yardage specialist.