In this paper, we use DEA to measure the NBA basketball teams’ efficiency in seasons 2006-2007,
2007-2008, 2008-2009 and 2009-2010. In this context, each team is a DMU; we select the payroll
and the average attendance to be the inputs while the wins and the average points per game to be
the outputs. First, in order to obtain benchmarks, we measure the DMUs efficiency through classic
DEA BCC model with an assurance region for each one of the four seasons individually and together.
When we consider the four seasons together, we may analyse whether the performance of each
team increases or decreases over time. Next, we evaluate the teams cross efficiency by DEA game
to consider that there is no cooperation among DMUs. This approach also improves the efficiencies
discrimination.