Average Depth of Target Carries Less Weight Than it Used to

Photo: Corey Perrine/Florida Times-Unio/USA TODAY NETWORK

Share

Justin Herbert’s 2022 season was full of highlights but still frustrated fans. His Average Depth of Target (ADoT) – how far he was targeting his receivers downfield – was looked at as something inconsistent with his extraordinary passing ability. Herbert had the 35th highest ADoT out of 36 qualifiers at 6.4 air yards per pass attempt. This was only ahead of Colts quarterback Matt Ryan.  

Often considered the best quarterback in the league, Patrick Mahomes was first in Expected Points Added (EPA) per pass attempt while having the 26th highest ADoT. Continuing the trend of success despite a lower ADoT, Kyle Shanahan’s offense was ranked 2nd in EPA per pass while Brock Purdy and Jimmy Garoppolo had the 22nd and 29th highest ADoT respectively.  

In the current NFL, the path to having an efficient passing attack does not need to stem from throwing deep shots excessively. 

The Relationship Between Average Depth of Target and Efficiency

There is, however, credence to the thought that a higher ADoT is connected to better passing efficiency. Since 2015, almost every season has seen a strong correlation between the average air yards of quarterbacks and what their EPA per pass attempt was on the season. While 2020 saw the strongest correlation, it was high from 2017-2021 as well. This occurred until the 2022 season where the correlation hit its lowest mark in just under a decade. We do also see that, as most things in football are, the correlation between ADoT and EPA is cyclical: It was not very correlated from 2011-2014, then was from 2015-2021, and now 2022 might usher in a new era where shorter passes are seen as efficient again. 

Looking at the quarterback level, we can see that every year since 2014 had a positive trend between a quarterback’s aggressiveness and their efficiency. 

The rise of two high safety looks took firmer shape in 2018 under Vic Fangio in Chicago but became more widespread after Brandon Staley had so much success in 2020 in Los Angeles. The goal of having two high safeties (instead of the traditional one high safety that became popular with the Seahawks’ Legion of Boom) is to limit explosive passing plays by putting a ceiling over the offense while forcing a quarterback to make consistently accurate short throws. The defense trusts their ability to capitalize on one mistake before the offense can march down the field on a ten-play drive. 

In 2020, when single-high was played more often, quarterbacks could throw it deep successfully as the defense was less designed to deal with multiple vertical threats. Now in 2022, offenses are finding other ways to be efficient passing the ball by throwing shorter and relying more on yards after the catch. 

2022 still showed a positive trend between ADoT and EPA per pass but not to the extent we have seen in the past.

The League’s Decrease in Average Depth of Target

As mentioned earlier, Patrick Mahomes led the league in passing efficiency while having a below league average ADoT (selection bias is also in play here, but more on that later). Some of the league’s top play callers in Kyle Shanahan, Ben Johnson, and Brian Daboll led effective passing attacks by keeping their quarterback’s pass attempts shorter. This is not to say there is a more preferred version between what Joe Burrow did and what Josh Allen did last year but that a low ADoT does not deserve the same negative association it once held 

With more snaps, quarterbacks and their play callers have adjusted the passing game to fit the times. The median (center of the box in the boxplot) ADoT has lowered from 8.8 in 2011 to 8.0 in 2018 to now 7.7 in 2022.

This seems to be mostly explained by the increase in frequency in passes behind the line of scrimmage (LOS) going from 19.7% in 2017 to 22.0% in 2022. This comes at the expense of 1-5 and 6-10 air yard passes continuing to dip. As former Atlanta Falcons Director of Football Data & Analytics and current TruMedia Director of Football Strategy John Taormina theorized, “Teams are designing high value EPA passes with negative air yards, and that’s often an inverse relationship. I truly think that’s the entire effect. 

Adjusting Average Depth of Target for Situation

This brings us to the next point that average depth of target is heavily influenced by selection bias. ADoT is often higher in situations like 3rd and long or when a team is trailing as opposed to 2nd and short when a team is leading. A quarterback who is putting their team in advantageous situations is likely going to have a lower ADoT 

Expected Points is a commonly used metric for evaluating team and player strength because it adjusts the number of yards gained on a play by the situation that the team is in (e.g., gaining 3 yards on 3rd and 2 is better than gaining 5 yards on 3rd and 10). This is not commonly done for ADoT. Since someone like Patrick Mahomes is often leading and ends up on 3rd down less often than the average quarterback, we would expect his ADoT to be low naturally. Using a simple linear regression model to account for down, distance, yard line, score differential, and time remaining we can get an expected average depth of target, which is a predicted ADoT if a league average quarterback were in this exact situation. 

What This Means Going Forward

Returning to the beginning, Justin Herbert did have a very low ADoT, but this was not a surprise based on the situations he was in. That same principle applied to Patrick Mahomes, Joe Burrow, Jalen Hurts, and Josh Allen. Those 4 quarterbacks led often and therefore were not expected to have a high ADoT.  

The point being made is two-fold: 

  1. While still positively correlated, the popular belief is that a higher ADoT leads to significantly more passing success. That correlation is still there to an extent but not as strong as it has been in the past couple of seasons. 
  2. Average Depth of Target is something useful (as are most data points) but somewhat flawed due to the influence situation has on it. A quarterback also can also contribute to the influence on their own ADoT because better quarterbacks should be leading more often. When discussed, using an adjusted version or keeping in mind what is expected of the passer is important. 

The overall premise of this article is not to say that teams should avoid throwing the ball down the field or that ADoT is completely useless. It is to say that we should critically examine what a lower ADoT might mean. There could have been many reasons for Herbert’s low ADoT last year, but the results should not cloud the process. As Mahomes and Allen’s vastly different styles showed, there are multiple ways to lead an efficient passing attack. For now, we have moved past the days of 2017-2021 where ADoT and passing effectiveness went hand-in-hand. As we know, however, the NFL is nothing if not cyclical.  

Share

Related Posts

Class Play – What To Look For at The 2024 Combine

February 29, 2024

Stats & Scheme – Combine Metric Predictiveness

February 27, 2024

SumerSports Show – Episode 145

February 27, 2024

NFL Combine: What Influences Production and Draft Position?

February 27, 2024