Several exercises in Chapter 7 showed that attendance at American League baseball games increased with the number of runs scored. But fans may respond more to winning teams than to high-scoring games. Here is a regression of average attendance on both the average runs per game and the number of games won during the 2016 season:

a) The regression using only Runs as a predictor has a slope coefficient of 5180.47. Can you account for the difference in coefficients when Won is included in the model? Here is a plot of the residuals:

b) The point plotted in red is the Cleveland Indians, who won the AL pennant that year. What does the residual say about attendance at Indian games?

Response variable is: Total Attendance cases selected according to American League R-squared = 24.0% s= 642476 with 15 3 12 degrees of freedom Variable Intercept Won Runs Coefficient -1272076 19711.9 2825.73