College Route Running Versatility and the 2024 NFL Draft Class

by Quinn MacLean|May 29, 2024

Share

The 2024 NFL Draft Class was viewed as one of the most complete wide receiver classes in recent memory with a record tying seven receivers taken in the first round. In total we saw 34 receivers selected and a persistent public scouting note to describe what separated each WR was their ability to have a complete “route tree” (Matt Harmon at Yahoo Sports does a great job of charting WR success at each route group). For visualization purposes, the route tree (provided by Pro Football Focus) can be categorized as the following:

A WR with experience running many different routes at a high level gives defenses issues in coverage and can provide signal on the level of trust the offensive coordinators have on their route running ability, even though some college offenses end up siloing certain players to certain roles. One way to determine route running versatility is to utilize Shannon Entropy, which aims to provide a singular value on whether we can predict the receiver’s next route. In short, the higher the Shannon Entropy the more likely the next route ran by the player will be more unpredictable given a wider inferred distribution. Shannon Entropy has been used in a football context before, one such example is understanding snaps at different positions to measure player versatility.

This analysis can also be used in scouting receivers. To calculate our “Route Running Entropy” score, we used NCAA PFF Data from solely FBS players from the 2018 to 2023 season on every route ran. For the purposes of our analysis, we are solely focusing on WRs (outside and slot receivers).

As mentioned prior, those that have the trust of collegiate offensive coordinators in their route running ability may be used in various different ways to get the player targets in the passing game. As we can see below, the unpredictable nature of route running versatility leads to a higher target rate at the collegiate level (perhaps not surprisingly). Those with a higher target rate here are also more likely to be drafted (as shown in the gold shaded bubbles below).

We can also use this metric to understand more about route running ability and predictability at the college level as a way to scout and understand a receiver’s development. One way to observe this is to see what routes players who have lower versatility are asked to run and which routes you can see players running if they have higher versatility (as defined by our metric).

To start, we controlled for those WRs with at least 200 routes in a career. We can see players with a high frequency of “Go Routes” (Includes, Sluggo, Up, Seam, etc.) or “Hitch Routes”, have a trend to be more concentrated in those types of routes and thus have less perceived versatility. Given the design of “Go” and “Hitch” Routes, these can emphasize players with pure linear speed. Players with a higher frequency of “Crossing”, “Corner”, or “Out” routes are asked to run more combinations of routes. These types of routes may require more technical footwork (the ability to transition out of breaks).

The “sweet spot” for those indicative of being an NFL receiver seem to be those players with 20-30% of Go or Hitch routes and a mix of other routes. One counter argument is that the personnel a player is used in could dictate the types of routes they are used in. However, we found minimal correlation between career 11 (i.e., 3 WR sets) personnel usage rate and route entropy, which shows that coordinators will still try to get a player involved as much as they can if they have exemplified good route running ability.

The flexibility of the entropy calculation can be used to understand development over time and whether or not players are asked to take on a broader route tree as they gain more experience. We can see this represented in the development curve below. It is important to note that the level of uncertainty widens as a player surpasses their 36th game played (12 games, three years, assuming they play their first three years) as this is likely a key decision point for a collegiate WR to determine if they should declare for the NFL Draft or stay another year. This presents some survivorship bias as we do not observe if the players who have declared for the draft would have gotten better at the collegiate stage or not. We could potentially assume that if they declared for the next level that they would have either improved or stayed relatively consistent.

One way we see a change in players’ route running versatility (and potentially their development) is if they decide to change teams (i.e., through the transfer portal). In recent years, we have seen a number of notable WRs transfer to new programs and see success in improving their route running ability, most notably Puka Nacua (Washington to BYU) and Jameson Williams (Ohio State to Alabama). WRs can change receiving rooms for a variety of reasons: scheme fit, avoid an overcrowded room of talented receivers, coaching changes, or wanting to compete against a different level of talent to name a few.

WRs who experience the biggest change in their route running versatility are those who move from Power 5 programs (P5) to a Group of 5 (G5) program and vice versa. We see a flat change in route running versatility among those who stay within their same level. Although we see a drastic increase in route running versatility in the move from a P5 program to a G5 program, we do see a wider distribution of outcomes in that shift. WRs that move from a G5 to a P5 program see minimal change in their route running versatility with a narrower distribution (more certainty). This could be less about a player’s ability and more about what the WR moving from a G5 to P5 program is asked to do.

For example, if a player ran a variety of routes at the G5 level but has shown pure linear speed, they may be asked to fit more into a WR-Z spot (think Brandin Cooks) to provide a change of pace within a room of quality receivers. Some recent players that increased their route running versatility by moving from G5 to P5 were Jacob Cowing (UTEP to Arizona) and Jamari Thrash (Georgia State to Louisville) who were both drafted in the 2024 NFL Draft.

Another application of this metric is to observe a receiver’s NFL readiness based on their route versatility. Below, we have listed the top receivers in route versatility who were drafted since 2020 (lagged a few years to ensure we had an appropriate amount of college years). For these receivers listed below here all but one (Freddie Swain) had “Go” routes listed as one of their top route frequencies in their collegiate career and all had either “Out” or “Crossing” routes in their top 3 route frequencies. This further illustrates the point that those who had speed and route running ability were further thought to be unpredictable in future routes ran. The top college players may run more Go/Hitch routes to take advantage of their speed differential, but relying on solely that trait can cause issues at the NFL level.

Our main assumption of this entire analysis is that uncertainty about what routes a player will run can be translated to an interpretation about a receiver’s versatility. We observed what the players were asked to do, which in part can be dictated by a player’s scheme (as seen by the number of Alabama receivers with higher versatility). We also are missing context around a player’s capabilities, which we infer based on the routes they are asked to run. We believe that given minimal correlation in offensive personnel usage to route running entropy that the scheme likely has a small part to play in the unpredictability for all receivers outside of their ability.

Another area that goes into route running ability that isn’t discussed here is a receiver’s average separation on routes ran. We make the assumption that given a relationship of targets per route ran to route entropy that those who separated were like to be targeted. One way to extend this analysis is looking at route entropy to average separation by WRs.

However, there is still some insight in how frequency in usage among certain route concepts leads to more unpredictable nature, which we believe can be used to scout a receiver’s route running ability. Among the WRs drafted in the first round this past draft, here’s how the 1st rounders stacked up in route running versatility (listed top 3 route frequency in parentheses):

  • Marvin Harrison Jr.: 3.05 (Go: 28%, Hitch: 20%, In: 9%)
  • Malik Nabers: 2.94 (Go: 29%, Hitch: 22%, Out: 11%)
  • Rome Odunze: 3.03 (Go: 28%, Hitch: 18%, Crossing: 12%)
  • Brian Thomas: 2.89 (Hitch: 27%, Go: 23%, In: 14%)
  • Xavier Worthy: 3.16 (Go: 24%, Hitch: 15%, Crossing: 14%)
  • Ricky Pearsall: 3.20 (Go: 19%, Crossing: 18%, Hitch: 16%)
  • Xavier Legette: 3.15 (Go: 23%, Hitch: 20%, Crossing: 12%)

As mentioned previously, those typically of NFL caliber will be in the 20-30% of Go or Hitch % as part of their route tree, which we see consistently above with the exception of Ricky Pearsall (who transferred from Arizona State to Florida). The high unpredictability of the route running ability of all these players should translate at the next level.

Share

Related Posts

Justin Jefferson’s Massive Contract: What It Means for the Dallas Cowboys & CeeDee Lamb

June 13, 2024

Cutting Through Noise to Increase CPOE Stability

June 4, 2024

The Class Play – Answering Key Questions from Texas CFB Message Boards

June 1, 2024

College Route Running Versatility and the 2024 NFL Draft Class

May 29, 2024