The superstar effect on perceived performance in professional football: An online experiment

We conduct a novel experiment to investigate whether football superstars consistently receive more favorable evaluations than non-superstars. Engaging 500 participants from Prolific, we randomly assign them to evaluate the same football videos with either visible or obscured players. In the control group, where players are visible, superstars receive lower performance ratings than non-superstars, challenging common perceptions. This trend is more intensified in the treatment group, where obscured identities result in even lower ratings for superstars, relative to non-superstars, suggesting a diminished superstar premium. These findings provide causal experimental evidence contributing to the literature on evaluation bias and the superstar effect.