Underdog
Network

;

Fantasy Usage Model - Week 6 Aftermath

Oct 18th 2022

Hayden Winks

Posted every Tuesday, The Fantasy Usage Model combines my expected fantasy points model with his film notes to recap the previous week's game. An explanation of this is here, but it basically turns targets, carries, air yards, goal line touches, and other variables into expected fantasy points.

Fantasy Usage Model - Post Week 6

Data only from Weeks 3-6. You can filter by team, position, or stat, and you can remove columns. Best viewed on desktop.

Week 7 Regression Candidates

  • RB (Positive Regression) - Jonathan Taylor (-9.5 half PPR over expected), Joe Mixon (-5.9), Kareem Hunt (-3.5), Alvin Kamara (-3.3), A.J. Dillon (-2.9), Aaron Jones (-2.7).

  • RB (Negative Regression) - Austin Ekeler (+6.0), Jamaal Williams (+4.8), Derrick Henry (+4.5), Nick Chubb (+4.2), Josh Jacobs (+4.2), Dalvin Cook (+4.0).

  • WR (Positive Regression) - Diontae Johnson (-5.4), D.J. Moore (-4.9), Curtis Samuel (-4.6), Allen Robinson (-3.9), Rondale Moore (-3.8), Chris Godwin (-3.2).

  • WR (Negative Regression) - Jakobi Meyers (+7.2), Gabe Davis (+4.9), Chris Olave (+4.0).

  • TE (Positive Regression) - Darren Waller if healthy (-5.6), Zach Ertz (-3.1), Evan Engram (-2.4).

  • TE (Negative Regression) - Greg Dulcich (+5.8), T.J. Hockenson (+3.0).

RB Situational Usage

All per game numbers from Weeks 3-6. "Dream Opps" are inside the 5-yard line opportunities plus "passing game targets", which are targets in obvious passing downs like the 2-minute warning, while trailing big, or on 3rd-and-long.

WR & TE Situational Usage

4.0 targets per game and 2 games this season to qualify. "EPA" is an efficiency stat. "CPOE" is completion percentage over expected. "TDOE" is touchdowns over expected. "aDOT" is average depth of target. "Deep Perc" is percentage of targets traveling 15+ yards downfield. "Deep Catch Rate" is how often those deep targets are caught. "Sideline Perc" is percentage of targets near the sideline, where targets are less valuable.