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 5

Data only from Weeks 2-5. You can filter by team, position, or stat, and you can remove columns. Best viewed on desktop.

Week 6 Regression Candidates

• RB (Positive Regression) - Joe Mixon (-7.7 half PPR over expected), Kareem Hunt (-3.4), Najee Harris (-2.6).

• RB (Negative Regression) - Nick Chubb (+7.3), Austin Ekeler (+6.4), David Montgomery (+3.3).

• WR (Positive Regression) - Diontae Johnson (-4.9), Rondale Moore (-4.2), Allen Robinson (-4.0), Elijah Moore (-3.5), Ja'Marr Chase (-3.5), JuJu Smith-Schuster (-2.9), Zay Jones (-2.7), Adam Thielen (-2.6).

• WR (Negative Regression) - Dyami Brown (+6.6), Gabe Davis (+5.0), Tyreek Hill (+4.9), Jakobi Meyers (+4.7).

• TE (Positive Regression) - Zach Ertz (-4.9), Cameron Brate (-3.2), Tyler Higbee (-2.9), Evan Engram (-2.3).

• TE (Negative Regression) - Taysom Hill (+9.2).

RB Situational Usage

All per game numbers. "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.