Estimated Reading Time:4 minutes
Last year, I introduced a new framework for understanding wide receiver play, quarterback decision-making and offensive potential. This framework relies on an XGBOOST model and PFF’s impressive collection of route-level data. Using machine learning and this breadth of PFF data, we can create models with the goal of predicting where a target should go on a given play.
The resulting metrics, Share of Predicted Targets and Share of Predicted Air Yards, are both more stable than their “actual” counterparts.
Wild-Card Recap
- Christian Kirk: 9 targets, 28.13% target share
- Ladd McConkey: 4 targets, 13.33% target share
- Josh Whyle: 1 target. 2.7% target share
- Joshua Palmer: DNP
Out of the three players that were in the model and played in wild-card weekend, we were able to identify the WR3 fantasy finish in Christian Kirk. Kirk also led his team in target and air yards share while accumulating more than three times the receiving yards as the next-closest Texans receiving threat.
Potential Breakouts: Divisional Round
These are players who were open far more often than they were targeted in the wild-card round. In general, players who show up on this list see an uptick in targets per route run and target share relative to both themselves and all players with similar target shares.
EPA Capture Rate and Potential EPA Per Attempt
As we have previously discussed, using the predicted EPA model, we can derive two very powerful and meaningful metrics.
- Potential EPA Per Attempt looks at the sum of the maximum predicted EPA on every play divided by attempts. The goal is to proxy receiver ability and offensive playcalling.
- EPA Capture Rate explains the sum of a quarterback’s actual EPA they amass divided by their potential EPA. The goal is to understand how well a quarterback is operating their system and utilizing the talent at their disposal.
Of the remaining eight playoff teams, six are in the upper-right quadrant of capture rate and potential. Drake Maye has by far the best EPA capture rate, while Matthew Stafford has the highest potential EPA per attempt. This makes intuitive sense, as Maye is operating with incredible efficiency and, of the teams remaining, Stafford likely has the best combination of receiving threats and offensive system.
Josh Allen and Bo Nix show up as the only two divisional-round quarterbacks not in the top quartile for potential EPA per attempt. Per PFF Premium Stats, Both Allen and Nix tie for the ninth-lowest average depth of target for all quarterbacks with at least 100 dropbacks. While they are doing a good job capitalizing on the potential EPA of their respective systems, their potentials are being limited compared to the other remaining playoff teams.
Wild-Card Round “Coach, I Was Open” Review
WR Christian Watson, Green Bay Packers
Green Bay trailed 31-27 after a 25-point fourth-quarter rally, which set up a final drive with 1:36 remaining — down 4 and needing a touchdown.
About two seconds after the snap, Jayden Reed breaks open on a crossing route, while two Bears’ defenders flow toward Josh Jacobs and Josh Whyle. Reed’s target probability exceeds 20% with 0.62 predicted EPA. Jordan Love holds the ball briefly and begins to feel pressure from Montez Sweat. It is unclear why Love didn’t throw this pass, but after Tremaine Edmunds shifts toward covering the middle of the field, Reed is no longer open.
Love does not move off his initial spot; the pocket is relatively clean until release. As pressure arrives, Christian Watson gains clear separation on Jaylon Johnson, but Love checks down to Jacobs for three yards, costing 23 seconds (24% of the remaining clock). An incompletion likely leaves a better game state.
Watson earns a +1 play-level PFF grade and 1.81 predicted EPA, the highest predicted EPA this week for any non-targeted player. With an accurate throw to Watson, the Packers could have taken the lead.