Econometrics quiz of the week
From the WSJ's ProfessorJournal.com
In this week's By the Numbers column, Allen St. John reports on the use of statistical analysis in NBA managerial decisions about player effectiveness. "The NBA season just tipped off, and finally one of the league's most intriguing stories doesn't involve Shaquille O'Neal or LeBron James. This year's guy to watch may be an MBA [Daryl Morey of the Houston
Rockets] from MIT's Sloan School of Management." Mr. Morey's approach includes
examining how an entire team's scoring rises or falls when a given player is added into the mix. Comparing the statistical analysis of player contributions in baseball and basketball: "Baseball and basketball lend themselves to statistical analysis in different ways. Mr. Morey says baseball's linear nature makes it easier to 'isolate a player's contribution.' But in basketball, there are a lot of what he terms 'interaction effects.'" These interaction effects are essential to the analysis of basketball player contribution. The column offers nice examples of these interaction effects and the importance of considering them when doing statistical analyses of player contribution.
1.) In econometrics, what are "interaction effects"?
2.) Interpret the statement: "Mr. Morey says baseball's linear nature makes it
easier to 'isolate a player's contribution.'" In baseball, does one player's on-base percentage or batting average depend on the batting of his teammates? In basketball, does the number of points that a team scores when a player is on the court depend on how the player interacts with his teammates?
3.) To consider the theory of these interaction effects, contrast the marginal productivity of labor in the Cobb-Douglas and perfect substitutes production functions. In these production functions, the inputs are the various players,
and the output is the number of points scored. Which production function corresponds to baseball offense? Which production function corresponds to
4.) Why are professional sports teams willing to use statistical analysis that is more sophisticated than points per game, for example, to determine player contributions?
Reviewed By: James Dearden, Lehigh University