Date of Completion


Degree Type

Honors Thesis - Campus Access


Marketing (MRKT)

First Advisor

Andrew Healy, Ph.D.


Firms need to predict what level of performance they will receive from employees, which they do through salary offers. Classical theory suggests that firms will effectively use all information provided to them in order to make the most informed decision possible. However, past research has showed that firms often commit memory-based errors, and only recall recent, salient information. This paper examines the effects of luck on employees' performances, using Major League Baseball as a laboratory. After accounting for luck in performance, this paper then revisits the question of whether teams correctly utilize past performance data. When accounting for luck, teams do a much better job of utilizing all relevant and available information in making salary offers. This paper also attempts to account for the previously mentioned memory-based errors, mainly by analyzing extra effort exerted right before new contract negotiations. While there is a lot of variation in the extra effort of players, league-wide there does not appear to be any phenomenon that would justify memory-based errors.