Assignment:
Revenues, Payrolls and Winning
Problem Statement
One of the most important debates in pro sports concerns the relationship between payrolls, winning, and revenues. Since data exist on all of these factors, you can examine this relationship for yourself. The data you need are at the Sports Business Data Directory.
Procedure
A. Use all available NHL team payroll, income-expense, and total points data from the 2001-2002 season to the 2011-2012 season (You should have 10 years, the 04-05 season was cancelled). For each year where all data are available, produce a separate MSExcel spreadsheet for each year with the following columns: team name, total points, payroll, and revenue (this is overall revenue (Revenue $mil) from the income-expense file, not TV revenue). Convert payroll and revenue to $millions for charting.
B. Sort the data by payroll in ascending order. For each year, produce a separate scatter chart with payroll on the x-axis and total points on the y-axis. Calculate the simple correlation between total points and payroll for each year.
C. Using the same spreadsheets, sort the data by total points in ascending order. For each year, produce a separate scatter chart with total points on the x-axis and revenue on the y-axis. Calculate the simple correlation between total points and revenue for each year.
Hand In
Hard copy of all MSExcel files, tables, and charts created, plus a paragraph for each of the following questions:
1. Explain the chain of analysis from payroll to total points (in part B) and then on to total points and revenue (in part C). What other factors might affect these relationships that are outside of the data looked at in this project?
When looking at the data over the time period, you can note that there seems to be a stronger correlation of payroll on points than there is of total points on revenue. This can be partially explained by economist Berri who argues that teams reward points more than they do performance. Although there are instances such as in 2002-2003 and 2005-2006 seasons where there is a stronger correlation between total points scores and payroll, these two seasons are the seasons before and after the season that wasn't played because of bargaining. Other factors not looked at in this project that may also effect these relationships are the amount of star players on the team and the MRP of players, which could be hard to measure.
2. In part B, what do the graphs tell you about the relationship between payroll and total points? Do the simple correlation statistics add any more information about this relationship? Explain.
Upon reviewing the graphs, although some seasons have a stronger correlation than others, on average there appears to be a moderate to strong relationship between payroll and total points. Simple correlation statistics support this statement with correlation between payroll and total points landing more or less around a .5 correlation, which represents a moderately positive relationship. In fact, the average correlation among all seasons for payroll on total points in .478, indicating a moderately positive relationship.
3. In part C, what do the graphs tell you about the relationship between total points and revenue? Do the simple correlation statistics add any more information about this relationship? Explain.
When viewing the graphs for the relationship between total points and revenue, we note that there is still a positive relationship but it appears to be much weaker in most seasons than the relationship between payroll and total points. Simple correlation statistics support this claim with correlation fluctuating but remaining under .5 for the most part and even showing negative relationships in a season, indicating a weak positive relationship. Looking a bit deeper, we can see that the average correlation among all seasons for total points on revenue is actually .283 which indeed is a weak positive relationship overall. This again can be explained by economist Berri who claims that teams seem to focus more on players that score more points and pay them more money which doesn't always render higher revenues.