Due to the variance in wide receiver touchdowns year-to-year and other stats only weakly predicting next-season receiving TDs, regression can be used to identify wide receivers whose 2017 receiving touchdowns didn’t align with their expected touchdown total.