Regression candidates, situational deep dives, and the occasional model backed by the same play-by-play and Next Gen Stats data behind the query tool.
Actually we were wrong because of the carries. Our Week 1 running back projections missed by 34 yards a back. Three quarters of that miss was workload we assumed would repeat and in fact did not. Six seasons say half of any week's carries are gone by the next one, glad we know now.
Every RB's Week 1 graded against the yards and touchdowns their carries were actually worth, and a Week 2 projection that treats one hot game the way six seasons of history say it should be treated.
A model built on six completed seasons, projecting every team's 2026 win total from its 2025 record and how its offense and defense actually played. What it says, and where it is weak.
Which RBs scored fewer rushing touchdowns in 2025 than their carry locations implied and are primed to bounce back in 2026. An expected-touchdown model on play-by-play, corroborated by Next Gen Stats.